Universitext

Series Editors
Sheldon Axler
San Francisco State University, USA

Carles Casacuberta
Universitat de Barcelona, Spain

Angus MacIntyre
Queen Mary, University of London, UK

Kenneth Ribet
University of California, Berkeley

Claude Sabbah
École polytechnique, CNRS, Université Paris-Saclay, Palaiseau

Endre Süli
University of Oxford, UK

Wojbor A. Woyczyński
Case Western Reserve University, USA

Universitext is a series of textbooks that presents material from a wide variety of

mathematical disciplines at master’s level and beyond. The books, often well
class-tested by their author, may have an informal, personal even experimental
approach to their subject matter. Some of the most successful and established
books in the series have evolved through several editions, always following the
evolution of teaching curricula, to very polished texts.
Thus as research topics trickle down into graduate-level teaching, first
textbooks written for new, cutting-edge courses may make their way into
Universitext .
More information about this series at http://www.springer.com/series/223
Mark J. D. Hamilton

Mathematical Gauge Theory

With Applications to the Standard Model of Particle
Physics
Mark J. D. Hamilton
Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

ISSN 0172-5939 e-ISSN 2191-6675

Universitext
ISBN 978-3-319-68438-3 e-ISBN 978-3-319-68439-0
https://doi.org/10.1007/978-3-319-68439-0

Library of Congress Control Number: 2017957556

Mathematics Subject Classification (2010): 55R10, 53C05, 22E70, 15A66,

53C27, 57S15, 22E60, 81T13, 81R40, 81V19, 81V05, 81V10, 81V15

© Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher,
whether the whole or part of the material is concerned, specifically the rights of
translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or
by similar or dissimilar methodology now known or hereafter developed.

The use of general descriptive names, registered names, trademarks, service

marks, etc. in this publication does not imply, even in the absence of a specific
statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.

The publisher, the authors and the editors are safe to assume that the advice and
information in this book are believed to be true and accurate at the date of
publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made. The publisher remains neutral with
regard to jurisdictional claims in published maps and institutional affiliations.
Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Dedicated to my father and my mother
Preface
With the discovery of a new particle, announced on 4 July 2012 at CERN, whose
properties are “consistent with the long-sought Higgs boson” [31], the final
elementary particle predicted by the classical Standard Model of particle physics
has been found. The aim of this book is to explain the mathematical background
as well as some of the details of the Standard Model. It is directed both at
students of mathematics, who are interested in applications of gauge theory in
physics, and at students of physics, who would like to understand more of the
mathematics behind the Standard Model.
The book is based on my lecture notes for graduate courses held at the
University of Stuttgart and the LMU Munich in Germany. A selection of the
material can be covered in one semester. Prerequisites are an introductory course
on manifolds and differential geometry as well as some basic knowledge of
special relativity, summarized in the appendix. The first six chapters of the book
treat the mathematical framework of gauge theories, in particular Lie groups, Lie
algebras, representations, group actions, fibre bundles, connections and
curvature, and spinors. The following three chapters discuss applications in
physics: the Lagrangians and interactions in the Standard Model, spontaneous
symmetry breaking, the Higgs mechanism of mass generation, and some more
advanced and modern topics like neutrino masses, CP violation and Grand
Unification.
The background in mathematics covered in the first six chapters of the book
is much more extensive than strictly needed to understand the Standard Model.
For example, the Standard Model is formulated on 4-dimensional Minkowski
spacetime, over which all fibre bundles can be trivialized and spinors have a
simple explicit description. However, this book is also intended as an
introduction to modern theoretical physics as a whole, and some of the topics
(for instance, on spinors or non-trivial fibre bundles) may be useful to students
who plan to study topics such as supersymmetry or superstring theory.
Depending on the time, the interests and the prior knowledge of the reader, he or
she can take a shortcut and immediately start at the chapters on connections,
spinors or Lagrangians, and then go back if more detailed mathematical
knowledge is required at some point.
Since we focus on the Standard Model, several topics related to gauge theory
and fibre bundles could not be covered, such as characteristic classes, holonomy
theory, index theorems, monopoles and instantons as well as applications of
gauge theory in pure mathematics, like Donaldson and Seiberg–Witten theory.
For those topics a number of textbooks exist, some of which can be found in the
bibliography.
An interesting and perhaps underappreciated fact is that a substantial number
of phenomena in particle physics can be understood by analysing representations
of Lie groups and by rewriting or rearranging Lagrangians. Examples of such
phenomena, which we are going to study, are:
symmetries of Lagrangians
interactions between fields corresponding to elementary particles (quarks,
leptons, gauge bosons, Higgs boson), determined by the Lagrangian
the Higgs mechanism of mass generation for gauge bosons as well as the
mass generation for fermions via Yukawa couplings
quark and neutrino mixing
neutrino masses and the seesaw mechanism
CP violation
Grand Unification
On the other hand, if precise predictions about scattering or decay of
particles should be made or if explicit formulas for quantum effects, such as
anomalies and running couplings, should be derived, then quantum field
calculations involving Green’s functions, perturbation theory and
renormalization are necessary. These calculations are beyond the scope of this
book, but a number of textbooks covering these topics can be found in the
physics literature.
The references I used during the preparation of the book are listed in each
chapter and may be useful to the reader for further studies (this is only a
selection of references that I came across over the past several years, sometimes
by chance, and there are many other valuable books and articles in this field).
It is not easy to make a recommendation on how to fit the chapters of the
book into a lecture course, because it depends on the prior knowledge of the
audience. A rough guideline could be as follows:
One-semester course: Often lecture courses on differentiable manifolds
contain sections on Lie groups, Lie algebras and group actions. If these
topics can be assumed as prior knowledge, then one could cover in gauge
theory the unstarred sections of Chaps. 4 to 7 and as much as possible of
Chap. 8 , perhaps going back to Chaps. 1 to 3 if specific results are needed.
Two-semester course: Depending on the prior knowledge of the audience,
one could cover in the first semester Chaps. 4 to 6 in more detail and in the
second semester Chaps. 7 to 9 . Or one could cover in the first semester
Chaps. 1 to 5 and in the second semester Chaps. 6 to 8 (and as much as
possible of Chap. 9 ).
Mark J. D. Hamilton
Munich, Germany
July 2017
Conventions
We collect some conventions that are used throughout the book.

General
Sections and subsections marked with a in front of the title contain
additional or advanced material and can be skipped on a first reading.
Occasionally these sections are used in later chapters.
A word in italics is sometimes used for emphasis, but more often to denote
terms that have not been defined so far in the text, like gauge boson , or to
denote standard terms, like skew field , whose definition can be found in
many textbooks. A word in boldface is usually used for definitions.
Diffeomorphisms of manifolds and isomorphisms of vector spaces, groups,
Lie groups, algebras and bundles are denoted by .
We often use the Einstein summation convention by summing over the
same indices in an expression, without writing the symbol ∑ (we also sum
over two lower or two upper indices).
If A is a set, then Id A : A → A denotes the identity map.

A disjoint union of sets is denoted by or .

The symbols Re and Im denote the real and imaginary part of a complex
number (and sometimes of a quaternion).

Linear Algebra
We denote by Mat( n × m , R ) the set of n × m -matrices with entries in a
ring R .
n × n -unit matrices
 are denoted by I n or I .
The conjugate of a complex number or quaternion q is denoted by and
occasionally by q . The conjugate of a complex or quaternionic matrix
A is defined by conjugating each entry: .

The transpose of a matrix A is denoted by A T . A matrix is called symmetric

if A T = A and skew-symmetric or antisymmetric if A T = − A . For a
complex or quaternionic matrix A we set . A complex or
quaternionic matrix A is called Hermitian if A † = A and skew-Hermitian if
A†=−A.
Note that the Dirac conjugate of a spinor ψ , defined in Definition 6.
7.15 , has a special meaning and is in general not equal to ψ † or ψ .
For a matrix A we denote by det( A ) the determinant and by tr( A ) the trace
.
In , where , we denote by e 1 , e 2 … , e n the standard basis
vectors with n entries

If f : V → W is a linear map between vector spaces, we sometimes write fv

instead of f ( v ) for a vector v V to reduce the number of brackets in
formulas.

Groups
The neutral element of a group G is usually denoted by e G (if it is not a
matrix group, where the neutral element is usually denoted by I ).
By a group without preceding words like “topological” or “Lie” we mean a
group in the algebraic sense, without the additional structure of a
topological space or smooth manifold.
 We usually write the group operation as ⋅ (multiplication). Occasionally, we
write the operation for abelian groups as + (addition). The neutral element
is then sometimes denoted by 1 or 0, respectively.

Manifolds
All manifolds in this book are smooth ( ), unless stated otherwise.
On a manifold M we denote by the set of smooth functions on M
with values in and by the set of smooth functions with values
in a vector space W .
The differential of a smooth map f : M → N between manifolds M and N at
a point p M is denoted by D p f or f (push forward) . If g : N → Q is
another smooth map, then the chain rule in p M is

or simply

If X is a vector field on a manifold M and p M a point, then we denote by

X p or X ( p ) the value of X in p .
By a curve γ through a point p in a manifold M tangent to a vector X Tp
M we mean a smooth curve γ : I → M , defined on an open interval I
around 0, such that γ (0) = p and .

Suppose f is a real or complex-valued function on a manifold M , p M a

point and X T p M a tangent vector. Then we denote the directional
derivative of f along X by ( L X f )( p ), df ( X ) or D p f ( X ). The same
notation is used if f takes values in a real or complex vector space.

Diagrams
Feynman diagrams can be read with time increasing from left to right. Arrows
on fermion lines indicate particle-flow and point in the direction of momentum-
flow for particles, but opposite to the direction of momentum-flow for
antiparticles [125, Sect. 9.2 ]. The Feynman diagrams in this book have been
prepared with feynMF/feynMP. Commutative diagrams have been prepared with
TikZ and function plots with MATLAB.
Acknowledgements
There are several people and institutions I would like to thank. First, I am
grateful to Dieter Kotschick and Uwe Semmelmann for their academic and
scientific support since my time as a student. I want to thank Tian-Jun Li for our
mathematical discussions and the invitation to conferences in Minneapolis, and
the Simons Center for Geometry and Physics for the invitation to a workshop in
Stony Brook. I would also like to take the opportunity to thank (belatedly) the
German Academic Scholarship Foundation (Studienstiftung) for their generous
financial support during my years of study.
I am grateful to the LMU Munich and the University of Stuttgart for the
opportunity to give lecture courses on mathematical gauge theory, which formed
the basis for this book. I want to thank the students who attended the lectures, in
particular, Ismail Achmed-Zade, Anthony Britto, Simon-Raphael Fischer, Simon
Hirscher, Martin Peev, Alexander Tabler, Danu Thung, Juraj Vrábel and David
Wierichs, as well as my course assistants Nicola Pia and Giovanni Placini for
reading the lecture notes and commenting on the manuscript. Furthermore, I
would like to thank Bobby Acharya for his excellent lectures on the Standard
Model and Robert Helling and Ronen Plesser for our interesting discussions
about physics.
Special thanks to Catriona Byrne, my first contact at Springer, to Rémi Lodh
for his excellent editorial support and suggestions while I was writing the
manuscript, to the anonymous referees, the editors and the copyeditor for a
number of comments and corrections, and to Anne-Kathrin Birchley-Brun for
assistance in the production and publication of the book.
Finally, I am grateful to John, Barbara and Patrick Hamilton, Gisela Saalfeld
and Inge Schmidbauer for their encouragement and support over the years, and
to Guoshu Wang for her friendship.
Contents
Part I Mathematical Foundations

1 Lie Groups and Lie Algebras: Basic Concepts

1.1 Topological Groups and Lie Groups

1.2 Linear Groups and Symmetry Groups of Vector Spaces

1.3 Homomorphisms of Lie Groups

1.4 Lie Algebras

1.5 From Lie Groups to Lie Algebras

1.6 From Lie Subalgebras to Lie Subgroups

1.7 The ExponentialMap

1.8 Cartan’s Theorem on Closed Subgroups

1.9 Exercises for Chap. 1

2 Lie Groups and Lie Algebras: Representations and Structure Theory

2.1 Representations

2.2 Invariant Metrics on Lie Groups

2.3 The Killing Form

2.4 Semisimple and Compact Lie Algebras

2.5 Ad-Invariant Scalar Products on Compact Lie Groups

2.6 Homotopy Groups of Lie Groups

2.7 Exercises for Chap. 2

3 GroupActions

3.1 Transformation Groups

3.2 Definition and First Properties of Group Actions

3.3 Examples of Group Actions

3.4 Fundamental Vector Fields

3.5 The Maurer–Cartan Form and the Differential of a Smooth Group

Action

3.6 Left or Right Actions?

3.7 Quotient Spaces

3.8 Homogeneous Spaces

3.9 Stiefel and Grassmann Manifolds

3.10 The Exceptional Lie Group G 2

3.11 Godement’s Theorem on the Manifold Structure of Quotient

Spaces

3.12 Exercises for Chap. 3

4 Fibre Bundles

4.1 General Fibre Bundles

4.2 Principal Fibre Bundles

4.3 Formal Bundle Atlases

4.4 Frame Bundles

4.5 Vector Bundles

 4.6 The Clutching Construction

4.7 Associated Vector Bundles

4.8 Exercises for Chap. 4

5 Connections and Curvature

5.1 Distributions and Connections

5.2 Connection 1-Forms

5.3 Gauge Transformations

5.4 Local Connection 1-Forms and Gauge Transformations

5.5 Curvature

5.6 Local Curvature 2-Forms

5.7 Generalized Electric and Magnetic Fields on Minkowski

Spacetime of Dimension 4

5.8 Parallel Transport

5.9 The Covariant Derivative on Associated Vector Bundles

5.10 Parallel Transport and Path-Ordered Exponentials

5.11 Holonomy andWilson Loops

5.12 The Exterior Covariant Derivative

5.13 Forms with Values in Ad( P )

5.14 A Second and Third Version of the Bianchi Identity

5.15 Exercises for Chap. 5

6 Spinors
 6.1 The Pseudo-Orthogonal Group O( s , t ) of Indefinite Scalar Products

6.2 Clifford Algebras

6.3 The Clifford Algebras for the Standard Symmetric Bilinear Forms

6.4 The Spinor Representation

6.5 The Spin Groups

6.6 Majorana Spinors

6.7 Spin Invariant Scalar Products

6.8 Explicit Formulas for Minkowski Spacetime of Dimension 4

6.9 Spin Structures and Spinor Bundles

6.10 The Spin Covariant Derivative

6.11 Twisted Spinor Bundles

6.12 Twisted Chiral Spinors

6.13 Exercises for Chap. 6

Part II The Standard Model of Elementary Particle Physics

7 The Classical Lagrangians of Gauge Theories

7.1 Restrictions on the Set of Lagrangians

7.2 The Hodge Star and the Codifferential

7.3 The Yang–Mills Lagrangian

7.4 Mathematical and Physical Conventions for Gauge Theories

7.5 The Klein–Gordon and Higgs Lagrangians

7.6 The Dirac Lagrangian

 7.7 Yukawa Couplings

7.8 Dirac and Majorana Mass Terms

7.9 Exercises for Chap. 7

8 The Higgs Mechanism and the Standard Model

8.1 The Higgs Field and Symmetry Breaking

8.2 Mass Generation for Gauge Bosons

8.3 Massive Gauge Bosons in the SU(2) × U(1)-Theory of the

Electroweak Interaction

8.4 The SU(3)-Theory of the Strong Interaction (QCD)

8.5 The Particle Content of the Standard Model

8.6 Interactions Between Fermions and Gauge Bosons

8.7 Interactions Between Higgs Bosons and Gauge Bosons

8.8 Mass Generation for Fermions in the Standard Model

8.9 The Complete Lagrangian of the Standard Model

8.10 Lepton and Baryon Numbers

8.11 Exercises for Chap. 8

9 Modern Developments and Topics Beyond the Standard Model

9.1 Flavour and Chiral Symmetry

9.2 Massive Neutrinos

9.3 C, P and CP Violation

9.4 Vacuum Polarization and Running Coupling Constants

 9.5 Grand Unified Theories

9.6 A Short Introduction to the Minimal Supersymmetric Standard

Model (MSSM)

9.7 Exercises for Chap. 9

Part III Appendix

A Background on Differentiable Manifolds

A.1 Manifolds

A.2 Tensors and Forms

B Background on Special Relativity and Quantum Field Theory

B.1 Basics of Special Relativity

B.2 A Short Introduction to Quantum Field Theory

References

Index
Part I
Mathematical Foundations
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_1
Chapter 1 Lie Groups and Lie Algebras:
Basic Concepts
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

Gauge theories are field theories of physics involving symmetry groups.

Symmetry groups are groups of transformations that act on something and leave
something (possibly something else) invariant. For example, symmetry groups
can act on geometric objects (by rotation, translation, etc.) and leave those
objects invariant. For the symmetries relevant in field theories, the groups act on
fields and leave the Lagrangian or the action (the spacetime integral over the
Lagrangian) invariant.
Concerning symmetry groups or groups in general we can make a basic
distinction, without being mathematically precise for the moment: groups can be
discrete or continuous. Both types of symmetry groups already occur for
elementary geometric objects. Equilateral triangles and squares, for example,
appear to have discrete symmetry groups, while other objects, such as the circle
S 1, the 2-sphere S 2 or the plane with a Euclidean metric, have continuous
symmetry groups. A deep and less obvious fact, that we want to understand over
the course of this book, is that similar symmetry groups play a prominent role in
the classical and quantum field theories describing nature.
From a mathematical point of view, continuous symmetry groups can be
conceptualised as Lie groups. By definition, Lie groups are groups in an
algebraic sense which are at the same time smooth manifolds, so that both
structures – algebraic and differentiable – are compatible. As a mathematical
object, Lie groups capture the idea of a continuous group that can be
parametrized locally by coordinates, so that the group operations (multiplication
and inversion) are smooth maps in those coordinates. Lie groups also cover the
case of discrete, i.e. 0-dimensional groups, consisting of a set of isolated points.
In theoretical physics, Lie groups like the Lorentz and Poincaré groups,
which are related to spacetime symmetries, and gauge groups, defining internal
symmetries, are important cornerstones. The currently accepted Standard Model
of elementary particles , for instance, is a gauge theory with Lie group

There are also Grand Unified Theories (GUTs) based on Lie groups like
SU(5). We shall see in later chapters that the specific kind of Lie group in a
gauge theory (its dimension, whether it is abelian or not, whether it is simple or
splits as a product of several factors, and so on) is reflected in interesting ways in
the physics. For example, in the case of the Standard Model, it turns out that:
The fact that there are 8 gluons , 3 weak gauge bosons and 1 photon is
related to the dimensions of the Lie groups SU(3) and SU(2) × U(1) (the
SU(5) Grand Unified Theory has 12 additional gauge bosons).
The fact that the strong , weak and electromagnetic interactions have
different strengths (coupling constants) is related to the product structure of
the gauge group SU(3) × SU(2) × U(1) (GUTs built on simple Lie groups
like SU(5) have only a single coupling constant).
The fact that gluons interact directly with each other while photons do not
is related to the fact that SU(3) is non-abelian while U(1) is abelian.
Our main mathematical tool to construct non-trivial Lie groups will be
Cartan’s Theorem, which shows that any subgroup (in the algebraic sense) of a
Lie group, which is a closed set in the topology, is already an embedded Lie
subgroup.
Besides Lie groups, Lie algebras play an important role in the theory of
symmetries. Lie algebras are vector spaces with a bilinear, antisymmetric
product, denoted by a bracket [⋅ , ⋅ ], satisfying the Jacobi identity. As an
algebraic object, Lie algebras can be defined independently of Lie groups, even
though Lie groups and Lie algebras are closely related: the tangent space to the
neutral element e G of a Lie group G has a canonical structure of a Lie
algebra. This means that Lie algebras are in some sense an infinitesimal,
algebraic description of Lie groups. Depending on the situation, it is often easier
to work with linear objects, such as Lie algebras, than with non-linear objects
like Lie groups. Lie algebras are also important in gauge theories: connections
on principal bundles, also known as gauge boson fields, are (locally) 1-forms on
spacetime with values in the Lie algebra of the gauge group.
In this chapter we define Lie groups and Lie algebras and describe the
relations between them. In the following chapter we will study some associated
concepts, like representations (which are used to define the actions of Lie groups
on fields) and invariant metrics (which are important in the construction of the
gauge invariant Yang–Mills Lagrangian). We will also briefly discuss the
structure of simple and semisimple Lie algebras.
Concerning symmetries, we will study in this chapter Lie groups as
symmetry groups of vector spaces and certain structures (scalar products and
volume forms) defined on vector spaces (in Chap. 3 on group actions we will
study Lie groups as symmetry groups of manifolds). Symmetry groups of vector
spaces are more generic than it may seem: it can be shown as a consequence of
the Peter–Weyl Theorem that any compact Lie group can be realized as a group
of rotations of some finite-dimensional Euclidean vector space (i.e. as an
embedded Lie subgroup of SO(m)).
We can only cover a selection of topics on Lie groups. The main references
for this and the following chapter are [24, 83, 142] and [153], where more
extensive discussions of Lie groups and Lie algebras can be found. Additional
references are [14, 34, 70, 77] and [129].

1.1 Topological Groups and Lie Groups

We begin with a first elementary mathematical concept that makes the idea of a
continuous group precise.

Definition 1.1.1 A topological group G is a group which is at the same time a

topological space so that the map

is continuous, where G × G has the canonical product topology determined by

the topology of G.

Remark 1.1.2 Here and in the following we shall mean by a group just a group
in the algebraic sense, without the additional structure of a topological space or
smooth manifold.

We usually set e for the neutral element in G. An equivalent description of

topological groups is the following.

Lemma 1.1.3 A group G is a topological group if and only if it is at the same

time a topological space so that both of the maps

called multiplication and inversion , are continuous.

Proof Suppose that multiplication and inversion are continuous maps. Then the
map

is continuous and hence also the composition of this map followed by

multiplication. This shows that

is continuous, hence the group G is a topological group.

Conversely, assume that G is a topological group. Then the map

is continuous and hence the map

is also continuous. This proves the claim. □

The concept of topological groups is a bit too general to be useful for our
purposes. In particular, general topological spaces can be very complicated and
do not have to be, for example, locally Euclidean, like topological manifolds. We
now turn to the definition of Lie groups, which is the type of continuous groups
we are most interested in.

Definition 1.1.4 A Lie group G is a group which is at the same time a

manifold so that the map

is smooth, where G × G has the canonical structure of a product manifold

 determined by the smooth structure of G.

Remark 1.1.5 Note that we only consider Lie groups of finite dimension.

Remark 1.1.6 Here and in the following we mean by a manifold a smooth

manifold, unless stated otherwise.

Of course, every Lie group is also a topological group. We could define Lie
groups equivalently as follows.

Lemma 1.1.7 A group G is a Lie group if and only if it is at the same time a
manifold so that both of the maps

are smooth.

Proof The proof is very similar to the proof of Lemma 1.1.3. □

A Lie group is thus a second countable, Hausdorff, topological group that can be
parametrized locally by finitely many coordinates in a smoothly compatible way,
so that multiplication and inversion depend smoothly on the coordinates.

Remark 1.1.8 (Redundancy in the Definition of Lie Groups) A curious fact

about Lie groups is that it suffices to check that multiplication is smooth,
because then inversion is automatically smooth (see Exercise 1.9.5).

Remark 1.1.9 (Hilbert’s Fifth Problem) It was shown by Gleason [65],

Montgomery and Zippin [97] in 1952 that a topological group, which is also a
topological manifold, has the structure of a Lie group. This is the solution to one
interpretation of Hilbert’s fifth problem (see [135] for more details). We will not
prove this theorem (the existence of a smooth structure), but we will show in
Corollary 1.8.17 that on a topological group, which is a topological manifold,
there is at most one smooth structure that turns it into a Lie group.

We will see in Sect. 1.5 that there is a deeper reason why we want symmetry
groups to be smooth manifolds: only in this situation can we canonically
 associate to a group a Lie algebra, which consists of certain left-invariant
smooth vector fields on the group. Vector fields are only defined on smooth
manifolds (they need tangent spaces and a tangent bundle to be defined). This
explains why we are particularly interested in groups having a smooth
structure.

We consider some simple examples of Lie groups (more examples will

follow later, in particular, in Sect. 1.2.2).

Example 1.1.10 Euclidean space with vector addition is an n-dimensional

Lie group, since addition

and inversion

for are linear and hence smooth. The Lie group is connected,
non-compact and abelian.

Remark 1.1.11 Euclidean spaces can also carry other (non-abelian) Lie
group structures besides the abelian structure coming from vector addition. For
example,

with matrix multiplication is an example of a (so-called nilpotent), non-abelian

Lie group structure on , also known as the Heisenberg group .

Example 1.1.12 Every countable group G with the manifold structure as a

discrete space, i.e. a countable union of isolated points, is a 0-dimensional Lie
group, because every map G × G → G is smooth (locally constant). A discrete
group is compact if and only if it is finite. In particular, the integers and the
finite cyclic groups for are discrete, abelian Lie groups. The Lie
group can be identified with the 0-sphere

with multiplication induced from .

Example 1.1.13 The circle

is a 1-dimensional Lie group with multiplication induced from : multiplication

on is quadratic and inversion on is a rational function in real and
imaginary parts, thus smooth. Both maps restrict to smooth maps on the
embedded submanifold . The Lie group S 1 is connected, compact and
abelian .

Example 1.1.14 The following set of matrices together with matrix

multiplication is a Lie group:

As a manifold . We have

and

showing that multiplication and inversion are smooth maps and that SO(2) is
closed under these operations. The Lie group SO(2) is one of the simplest
examples of a whole class of Lie groups, known as matrix or linear Lie groups
(the Heisenberg group is another example). The Lie group SO(2) is isomorphic
to S 1.

We want to generalize the examples of the Lie groups S 0 and S 1 and show that
the 3-sphere S 3 also has the structure of a Lie group. This is a good opportunity
to introduce (or recall) in a short detour the skew field of quaternions .

1.1.1 Normed Division Algebras and the Quaternions

The real and complex numbers are known from high school and the first
mathematics courses at university. There are, however, other types of “higher-
dimensional” numbers, which are less familiar, but still occur in mathematics
and physics. It is useful to consider the following, general, algebraic notions (a
nice reference is [8]):
Definition 1.1.15

1. A real algebra is a finite-dimensional real vector space A with a bilinear

map

and a unit element 1 A such that 1 ⋅ a = a = a ⋅ 1 for all a A. In

particular, the multiplication on A is distributive, but in general not
associative.

2. We call the algebra A normed if there is a norm | | ⋅ | | on the vector space A

such that | | ab | | = | | a | | ⋅ | | b | |.

3. We call the algebra A a division algebra if ab = 0 implies that either a = 0

or b = 0.

4. The algebra has multiplicative inverses if for any non-zero a A there is

an a −1 A such that aa −1 = a −1 a = 1.

It follows that every normed algebra is a division algebra and every associative
division algebra has multiplicative inverses (this is not true in general for non-
associative division algebras).
The following is a classical theorem due to Hurwitz.

Theorem 1.1.16 (Hurwitz’s Theorem on Normed Division Algebras) There

are only four normed real division algebras:

1. The real numbers of dimension 1.

2. The complex numbers of dimension 2.

3. The quaternions of dimension 4.

4. The octonions of dimension 8.

We want to describe the quaternions in this section and leave the (non-
associative) octonions to Exercise 3.12.15.
Recall that there is an algebra structure on the vector space , so that this
vector space becomes a field, called the complex numbers : the multiplication
is associative and commutative. Furthermore, every non-zero element of has a
multiplicative inverse.
The complex plane is spanned as a real vector space by the basis vectors 1
and i, with i 2 = −1. By distributivity this determines the product of any two
complex numbers. We define the conjugate of a complex number z = a + ib as
and the norm squared as . The multiplicative inverse
of a non-zero complex number is then

We also have

for all complex numbers , so that is a normed division algebra.

There is a similar construction of an algebra structure on the vector space ,
so that this vector space becomes a skew field , called the quaternions : the
multiplication is associative and every non-zero element of has a
multiplicative inverse. However, the multiplication is not commutative. As a real
vector space is spanned by the basis vectors 1, i, j and k. The product satisfies

Using associativity of multiplication this determines all possible products

among the basis elements i, j, k and thus by distributivity the product of any two
quaternions. We have

showing in particular that the product is not commutative. The products

among i, j and k can be memorized with the following diagram:

We define the real quaternions by

and the imaginary quaternions by

We also define the conjugate of a quaternion w = a + bi + cj + dk as

and the norm squared as

The multiplicative inverse of a non-zero quaternion is then

We also have

for all , so that is a normed division algebra.

Remark 1.1.17 One has to be careful with division of quaternions: the

expression for is not well-defined, even for w ≠ 0, since multiplication
is not commutative. One rather has to write zw −1 or w −1 z.

Multiplication of quaternions defines a group structure on the 3-sphere:

Example 1.1.18 The 3-sphere

of unit quaternions is a 3-dimensional embedded submanifold of and a

Lie group with multiplication induced from . As a Lie group S 3 is connected,
compact and non-abelian (it contains, in particular, the elements ). The
3-sphere and thus the quaternions have an interesting relation to the rotation
group SO(3) of , to be discussed in Example 1.3.8.

Remark 1.1.19 We shall see in Exercise 3.12.15 that there is an algebra

structure on the vector space , so that this vector space becomes a normed
division algebra, called the octonions . This multiplication induces a
multiplication on S 7. However, the multiplication does not define a group
structure on S 7, because it is not associative.
1.1.2 Quaternionic Matrices
Certain properties of matrices with quaternionic entries cannot be proved in
the same way as for matrices with real or complex entries, because the
quaternions are only a skew field. In particular, it is not immediately clear how
to make sense of the inverse and a determinant for quaternionic square matrices.
Even if this is possible, it is not clear that such a determinant would have the
nice properties we expect of it, like multiplicativity and the characterization of
invertible matrices as those of non-zero determinant. Since we are going to
consider groups of quaternionic matrices as examples of Lie groups, like the so-
called compact symplectic group Sp(n), we would like to fill in some of the
details in this section (we follow the exposition in [83, Sect. I.8] and [152]). Like
any section with a star this section can be skipped on a first reading.

Definition 1.1.20 We denote by the set of all m × n-matrices with

entries in .

The set is an abelian group with the standard addition of matrices.

We can also define right and left multiplication with elements :

and

This gives the structure of a right or left module over the

quaternions ; we call this a right or left quaternionic vector space (since is
not commutative, left and right multiplication differ). In particular, the spaces of
row and column vectors, denoted by , each have the structure of a right and
left quaternionic vector space.
We can define matrix multiplication

in the standard way, where we have to be careful to preserve the ordering in

the products of the entries. All of the constructions so far work in exactly the
same way for matrices over any ring.
 We now restrict to quaternionic square matrices . It is sometimes
useful to have a different description of such matrices. The following is easy to
see:

Lemma 1.1.21 Let be a quaternion. Then there exist unique complex

numbers such that

Let be a quaternionic square matrix. Then there exist unique

complex square matrices such that

Definition 1.1.22 For a matrix , with A 1, A 2

complex, we define the adjoint to be the complex square matrix

Example 1.1.23 For the special case of a quaternion with

, considered as a 1 × 1-matrix, we get

The quaternions are sometimes defined as the subset of

consisting of matrices of this form.

Remark 1.1.24 Another definition of the adjoint, used in [152], is to write A =

A 1 + A 2 j with A 1, A 2 complex and define

We continue to use our Definition 1.1.22, which seems to be more standard.

Using the adjoint we get the following identification of with a

subset of .

Proposition 1.1.25 (Quaternionic Matrices as a Subspace of Complex

Matrices) We define an element by
Then J −1 = −J and the image of the injective map

consists of the set

The proof is a simple calculation. The following proposition can also be verified
by a direct calculation, where for the second property we use that for a
complex matrix C:

Proposition 1.1.26 (Properties of the Adjoint) The adjoints of quaternionic

n × n-matrices A, B satisfy

Corollary 1.1.27 If quaternionic n × n-matrices A, B satisfy AB = I, then BA =

I.

Proof If AB = I n , then

Hence by a property of the inverse of complex matrices

and thus BA = I n . □

We can therefore define:

Definition 1.1.28 A matrix is called invertible if there exists a

matrix with AB = I = BA. The matrix B is called the inverse
of A.

Corollary 1.1.29 A matrix is invertible if and only if its

adjoint χ A is invertible.
Proof If A is invertible, then χ A is invertible by Proposition 1.1.26. Conversely,
assume that χ A is invertible. We can write χ A −1 in the form

for some complex n × n-matrices B 1, …, B 4. Then

Setting

we get

We conclude that A is invertible with inverse B. □

Definition 1.1.30 The determinant of a quaternionic matrix

is defined by

Remark 1.1.31 Note that the determinant is not defined for matrices over a
general non-commutative ring.

The determinant on has the following property.

Proposition 1.1.32 (Quaternionic Determinant Is Real and Non-negative)

The determinant of a matrix is real and non-negative, det(A) ≥
0.

Proof According to Proposition 1.1.25

for all . Since det(J) = 1 we get

hence .
The proof that det(A) is non-negative is more involved. Any quaternionic
matrix can be brought to the form

for some k ≤ n by applying elementary row and column operations. The

corresponding elementary quaternionic matrices E act by left or right
multiplication on A, thus the adjoint χ E acts by left or right multiplication on χ A
. The elementary matrix E for switching two rows (columns) or adding a row
(column) to another one are real, thus

and det(χ E ) = det(E)2 = 1. It remains to consider the elementary matrix E

that multiplies a row (column) by a non-zero quaternion a. Writing a = a 1 + ja 2
with a 1, a 2 complex, the elementary matrix E is of the form E = E 1 + jE 2 with

and

Interchanging twice a pair of rows and twice a pair of columns we can bring
χ E to the following form, without changing det(χ E ):

It follows that

This proves the claim, since χ A is the product of matrices of non-negative

determinant. □
Remark 1.1.33 A second, independent proof of the second assertion in
Proposition 1.1.32 uses that the subset of invertible matrices in
is connected, cf. Theorem 1.2.22. Therefore its image under the
determinant is contained in . The claim then follows, because is
dense in .

Example 1.1.34 According to Exercise 1.9.8 the determinant of a matrix

is equal to

The next corollary follows from what we have shown above.

Corollary 1.1.35 (Properties of the Quaternionic Determinant) The

determinant is a smooth map

where has the manifold structure of . For all matrices

the following identity holds:

A matrix is invertible if and only if det(A) ≠ 0.

The determinant of quaternionic matrices, defined above via the adjoint,

therefore has at least some of the expected properties.

Remark 1.1.36 Note that the determinant of quaternionic matrices does not
define an -multilinear, alternating map on (n factors).

1.1.3 Products and Lie Subgroups

There are certain well-known constructions that yield new groups (and
manifolds) from given groups (manifolds). Some of these constructions are
compatible for both groups and manifolds and can be employed to generate new
Lie groups from known ones. We discuss two important examples of such
constructions: products and subgroups.

Proposition 1.1.37 (Products of Lie Groups) Let G and H be Lie groups.

Then the product manifold G × H with the direct product structure as a group is
a Lie group, called the product Lie group .

Proof This follows, because the maps

and

are smooth. □

Example 1.1.38 The n-torus

is a compact, abelian Lie group. Similarly multiple (finite) products of copies of

S 3 and S 1 are compact Lie groups. A particularly interesting case is S 3 × S 3,
because it can be identified with the Lie group Spin(4), to be defined in Chap. 6
(see Example 6.5.17).

Definition 1.1.39 Let G be a Lie group.

1. An immersed Lie subgroup of G is the image of an injective immersion

ϕ: H → G from a Lie group H to G such that ϕ is a group homomorphism.

2. An embedded Lie subgroup of G is the image of an injective immersion

ϕ: H → G from a Lie group H to G such that ϕ is a group homomorphism
and a homeomorphism onto its image.
We call the map ϕ a Lie group immersion or Lie group embedding,
respectively. In both cases, the set ϕ(H) is endowed with the topology, manifold
structure and group structure such that ϕ: H → ϕ(H) is a diffeomorphism and a
group isomorphism. Then ϕ(H) is a Lie group itself. The difference between
embedded and immersed Lie subgroups ϕ(H) G is whether the topology on
ϕ(H) coincides with the subspace topology on ϕ(H) inherited from G or not. The
group structure on ϕ(H) is in both cases the subgroup structure inherited from G.

An embedded Lie subgroup can be described equivalently as an embedded

submanifold which is at the same time a subgroup. Most of the time we will
consider embedded Lie subgroups (immersed Lie subgroups appear naturally in
Sect. 1.6). Note the following:

Proposition 1.1.40 If ϕ: H → G is a Lie group immersion where H is compact,

then ϕ is a Lie group embedding.

Proof Since H is compact and G is Hausdorff, it follows that the injective

immersion ϕ: H → G is a closed map, hence a homeomorphism onto its image. □

Hence immersed Lie subgroups which are not embedded can only be non-
compact.

Example 1.1.41 Consider the Lie group G = S 1 and an element x = e 2πiα S 1

with . The number α is rational if and only if there exists an integer N such
that Nα is an integer. This happens if and only if x N = 1. Hence if α is rational,
then x generates an embedded Lie subgroup in S 1, isomorphic to the finite cyclic
group , where K is the smallest positive integer such that Kα is an integer. If α
is irrational, then x generates an immersed Lie subgroup in S 1, isomorphic to .

Example 1.1.42 Similarly, consider the Lie group G = T 2, which we think of as

being obtained by identifying opposite sides of a square [−1, 1] × [−1, 1]. It can
be shown that the straight lines on the square of rational slope through the
neutral element e = (0, 0) define embedded Lie subgroups, diffeomorphic to S 1,
while the straight lines of irrational slope through (0, 0) define immersed Lie
subgroups , diffeomorphic to .

Example 1.1.43 The sets of invertible elements, for ,

together with multiplication are Lie groups of dimension 1, 2, 4, respectively.
The spheres S 0, S 1, S 3 are embedded Lie subgroups of codimension 1.

The proof of the following (non-trivial) theorem is deferred to Sect. 1.8.

Theorem 1.1.44 (Cartan’s Theorem, Closed Subgroup Theorem) Let G

be a Lie group and suppose that H G is a subgroup in the algebraic sense.
Then H is an embedded Lie subgroup if and only if H is a closed set in the
topology of G.
By a closed subgroup of a Lie group G we always mean a subgroup which is a
closed set in the topology of G. The difficult part in the proof of this theorem is
to show that a closed subgroup of a Lie group is an embedded submanifold.
Cartan’s Theorem allows us to construct many new interesting Lie groups by
realizing them as closed subgroups of given Lie groups.

Remark 1.1.45 Groups can also be constructed by taking the quotient of a

group G by a normal subgroup H. However, for a Lie group G and a subgroup H,
the quotient space G⁄H will not be a smooth manifold in general (at least not
canonically). We will show in Sect. 3.8.3 that if H is a closed subgroup of a Lie
group G, then G⁄H is indeed a smooth manifold. In this case, if H is normal, then
G⁄H will again be a Lie group.

1.2 Linear Groups and Symmetry Groups of Vector

Spaces
The most famous class of examples of Lie groups are the general linear groups
over the real, complex and quaternionic numbers as well as the
following groups:

Definition 1.2.1 A closed subgroup of a general linear group is called a linear

group or matrix group.

By Cartan’s Theorem 1.1.44 linear groups are embedded Lie subgroups and, in
particular, Lie groups themselves. We are especially going to study the following
families of linear groups, which are called classical Lie groups:
the special linear groups in the real, complex and quaternionic case
the (special) orthogonal group in the real case
the (special) unitary group in the complex case
the compact symplectic group (also called the quaternionic unitary
group) in the quaternionic case
the real pseudo-orthogonal groups for indefinite scalar products, like the
Lorentz group.
The general linear groups are the (maximal) symmetry groups of vector
spaces. The families of linear groups above arise as automorphism groups of
certain structures on vector spaces. They can also be understood as isotropy
groups in certain representations of the general linear groups.
There are two classes of Lie groups we are interested in which are not (at
least a priori) linear:
the exceptional compact Lie groups G2, F4, E6, E7, E8 (we will discuss G2
in detail in Sect. 3.10)
the spin groups , which are certain double coverings of (pseudo-
)orthogonal groups.
All Lie groups that we will consider belong to one of these classes or are
products of such Lie groups. Most linear groups are non-abelian and certain
classes of linear groups – the (special) orthogonal, (special) unitary and
symplectic groups – are compact . Lie groups like the Lorentz group and its spin
group are not compact.
There are several reasons why linear groups are important, in particular with
regard to compact Lie groups:

1. First, it is possible to prove that any compact Lie group is isomorphic as a

Lie group to a linear group, see Theorem 1.2.7. In particular, the compact
exceptional Lie groups and the compact spin groups are isomorphic to linear
groups.

2. Secondly, there is a classification theorem which shows that (up to finite

coverings) any compact Lie group G is isomorphic to a product

of compact Lie groups, all of which belong to the classes mentioned above
(classical linear, spin and exceptional Lie groups); see Theorem 2.4.23 and
Theorem 2.4.29 for the classification of compact Lie groups.
Even though the classical linear groups look quite special, they are thus of
general significance, in particular for gauge theories with compact gauge groups.
Spin groups, such as the universal covering of the Lorentz group and its
higher-dimensional analogues, are also important in physics, because they are
involved in the mathematical description of fermions.
Finally, the exceptional Lie groups appear in several places in physics: E6,
for example, is the gauge group of certain Grand Unified Theories , E8 plays a
role in heterotic string theory and G2 is related to M-theory compactifications.

Remark 1.2.2 It is an interesting fact that there are non-compact Lie groups
which are not isomorphic to linear groups. One example is the universal
covering of the Lie group (see [70, Sect. 5.8] for a proof).

1.2.1 Isomorphism Groups of Vector Spaces

The simplest and fundamental linear groups, perhaps already known from a
course on linear algebra, are the general linear groups themselves. Let
.

Definition 1.2.3 For n ≥ 1 the general linear group is defined as the group of
linear isomorphisms of :

where denotes the ring of n × n-matrices with coefficients in .

Group multiplication in is matrix multiplication.

For we act with matrices on the left of the right vector space , so that
the maps are indeed -linear:

The following alternative description of the general linear group follows

immediately:

Proposition 1.2.4 The general linear group is given by

Clearly, for n = 1, we have

We first consider the real general linear group.

Proposition 1.2.5 (The Real General Linear Group Is a Lie Group)

is a non-compact n 2 -dimensional Lie group . It is not abelian for n ≥ 2
.

Proof It is clear by properties of the determinant that is a group in the

algebraic sense.
Note that by continuity of the determinant, is an open subset of
In particular, is a smooth manifold of dimension n 2. Multiplication
of two matrices A, B is quadratic in their coordinates, hence a smooth map.
According to Remark 1.1.8 this shows that is indeed a Lie group (we
can also see directly that inversion of a matrix is, by Cramer’s rule, a rational
map in the coordinates, hence smooth).
The manifold , as a subset of with the Euclidean norm, is not
bounded, because it contains the unbounded set of diagonal matrices of the form
rI n with (these elements actually define a subgroup of the general
linear group, isomorphic to ). By Heine–Borel, is not compact.
To show that is not abelian for n ≥ 2, note that contains the
subgroup H isomorphic to , consisting of matrices of the form

It therefore suffices to show that is not abelian: it is easy to find two

matrices in which do not commute. □

Using similar arguments it can be shown:

Proposition 1.2.6 (Complex and Quaternionic General Linear Groups Are

Lie Groups) Over the complex numbers and quaternions we have:

1. is a non-compact 2n 2 -dimensional Lie group . It is not abelian for

n≥2.

2. is a non-compact 4n 2 -dimensional Lie group . It is not abelian for

n≥1.

As an application of the Peter–Weyl Theorem, the following can be shown (for a

proof, which is beyond the scope of this book, see [24, 83, 129]):

Theorem 1.2.7 (Compact Lie Groups Are Linear) Let G be a compact Lie
group . Then there exists a smooth, injective group homomorphism ϕ of G into a
general linear group for some n.
According to Corollary 1.8.18 the map ϕ is a Lie group isomorphism onto a
linear group (see Sect. 1.3 for the formal definition of Lie group
homomorphisms and isomorphisms).
The Peter–Weyl Theorem shows that every compact Lie group can be
considered as a linear group. If we assume this result, we shall see later as a
consequence of Theorem 2.1.39 that a compact Lie group G can even be
embedded as a closed subgroup in a unitary group U(n) for some n and thus
in the rotation group SO(2n) by Exercise 1.9.10. As a consequence any
compact Lie group G can be literally thought of as a group whose elements
are rotations on some .

Remark 1.2.8 By comparison, recall that Cayley’s Theorem says that any finite
group of order n can be embedded as a subgroup of the symmetric group S n .

1.2.2 Automorphism Groups of Structures on Vector

Spaces
We want to consider specific classes of linear groups that arise as automorphism
groups of certain structures on vector spaces.

Definition 1.2.9 We define the following scalar products :

1. On the standard Euclidean scalar product

2. On the standard Hermitian scalar product

3. On the standard symplectic scalar product

 Here

are column vectors in , and , respectively.

Definition 1.2.10 Let . We define the standard volume form vol

on by

where (v 1, …, v n ) is the n × n-matrix with columns v 1, …, v n .

Remark 1.2.11 For this form is -multilinear and alternating, but not
for .

Definition 1.2.12 Let n ≥ 1.

1. The special linear groups for are defined as the automorphism

groups of the standard volume forms:

2. The orthogonal , unitary and (compact) symplectic (also called

quaternionic unitary) groups are defined as the automorphism groups of
the standard Euclidean, Hermitian and symplectic scalar products:

3. The special orthogonal and special unitary groups are defined as

These Lie groups are called classical groups .

Remark 1.2.13 There are additional classes of classical groups, called (non-
compact) symplectic groups, which are defined as automorphism groups of
skew-symmetric forms. We will not consider these Lie groups in the subsequent
discussions.

Remark 1.2.14 In the physics literature the Lie group Sp(n) is sometimes
denoted by USp(n) (and occasionally by Sp(2n) or USp(2n)). For example, [149]
uses the notation Sp(n), whereas [90] uses the notation Sp(2n) and [138] uses the
notation USp(2n). We continue to use the notation Sp(n).

It is often useful to have the following alternative description of these groups.

Proposition 1.2.15 (Matrix Description of Classical Groups) Let n ≥ 1.

1. The special linear groups are given by

2. The orthogonal, unitary and symplectic groups are given by

where A T denotes the transpose of A and .

3. The special orthogonal and special unitary groups are given by

Proof

1. For and column vectors the following identity

holds:

This follows because as matrices

 and the determinant is multiplicative: det(AB) = det(A)det(B). This implies
the formula for .

2. A matrix satisfies

if and only if

Choosing v = e i , w = e j this happens if and only if A T A = I and thus AA T =

I. The complex and quaternionic case follow similarly.

3. This is clear by the results above.

□

We did not define a quaternionic special unitary group, because it turns out that
there is no difference between such a group and the quaternionic unitary group.
This follows from the next proposition.

Proposition 1.2.16 Let A Sp(n). Then det(A) = 1.

Proof Let with A 1, A 2 complex. According to

Proposition 1.1.32 the determinant of A is real and non-negative. We have

and

Thus for A Sp(n)

Therefore det(A) = 1. □

We now want to prove the main result in this subsection.

Theorem 1.2.17 (Classical Groups Are Linear) The special linear,

(special) orthogonal, (special) unitary and symplectic groups are closed
subgroups of general linear groups, i.e. linear groups . They have the
following properties:
 1. The special linear groups have dimension

These Lie groups are not compact for n ≥ 2 .

2. The orthogonal, unitary and symplectic groups have dimension

These Lie groups are compact for all n ≥ 1 .

3. The special orthogonal and special unitary groups have dimension

These Lie groups are compact for all n ≥ 1 .

Proof
Closed subgroups: It is clear that all of these subsets are subgroups of
general linear groups. The maps

for , and

are continuous (polynomial in the coordinates), hence the preimages det

−1(1) are closed subsets. This shows that is a closed subgroup of
the general linear group . Similarly, the map

is continuous (quadratic in the coordinates), hence the preimage of I is a

closed subset. This shows that O(n) and the intersection
are closed subgroups of the general linear group
. Similarly for U(n), SU(n) and Sp(n).
Compactness: To show that is not compact for n ≥ 2 it suffices to
show that the subgroup is not compact. This follows by
considering the unbounded subset of matrices of the form

To show that O(n) and hence SO(n) are compact it suffices to show by
Heine–Borel that O(n) is a bounded subset of the Euclidean space

Let A O(n). For fixed i = 1, …, n we have

hence | A ij | ≤ 1 for all indices i, j. This implies the claim. Compactness of

U(n), SU(n) and Sp(n) follows similarly.
Dimensions of special linear groups: Finally we calculate the dimensions.
We claim that the smooth map

for has 1 as a regular value. This implies again that and

are smooth manifolds and also the formulas n 2 − 1 and 2n 2 − 2 for
the dimensions (the argument for is slightly different and is given
below).
To prove the claim, let for and write A as

where the v i are column vectors in . For fixed consider the

curve

in . Then C(0) = A and since the determinant is multilinear

thus

This shows that the differential of the determinant is surjective in every

and 1 is a regular value.
We now prove the formula for the dimension of . We want to
show that

has 1 as a regular value. This implies that is a smooth manifold

of dimension 4n 2 − 1. Let with A 1, A 2 complex,
written in terms of complex column vectors as

For consider the curve

in . Then D(0) = A and the adjoint of D(t) is

with determinant

It follows that

This shows that the differential of the determinant is surjective in every

and 1 is a regular value.
Dimensions of O(n), U(n) and Sp(n): To calculate the dimension of the
orthogonal group O(n) consider the map

where denotes the space of symmetric, real n × n-matrices. Then

O(n) = f −1(I). The differential of this map at a point A O(n) in the
direction is

Let and set

Then (D A f)(X) = B and thus I is a regular value of f. This shows that O(n)
is a smooth manifold of dimension
We can calculate the dimensions of U(n) and Sp(n) in a similar way,
utilizing for the map

where denotes the space of Hermitian n × n-matrices (the set of

all matrices B with B † = B). Again I is a regular value and thus U(n) and
Sp(n) are smooth manifolds of dimension

where k = 2, 4 for . This implies

and

Dimensions of SO(n) and SU(n): We claim that SO(n) is a submanifold of

codimension zero in O(n). The determinant on O(n) has values in { + 1,
−1},

This map obviously has 1 as a regular value.

Similarly, we claim that SU(n) is a submanifold of codimension one in
U(n). The determinant on U(n) has values in S 1,

We claim that this map has 1 as a regular value. Let A SU(n), written
in terms of complex column vectors as

For consider the curve

It is easy to check that C(t) is a curve in U(n) and

with
 This proves the claim.
□

Example 1.2.18 It follows directly from the matrix description that

and

We also saw that

and it is not difficult to check that

We will discuss less trivial identifications between Lie groups, like ,

after we have defined the notion of a Lie group isomorphism.

Remark 1.2.19 Some of the linear groups appear directly in gauge theories: For
instance, as mentioned before, the gauge group of the current Standard Model of
particle physics is the product

There are Grand Unified Theories based on Lie groups like SU(5) and SO(10)
(or rather its universal covering group Spin(10)).

Remark 1.2.20 (Classical Linear Groups as Isotropy Groups) Note that

and its subgroups act canonically on the (column) vector space by
matrix multiplication from the left. This is an example of a representation ,
called the fundamental representation. We will consider representations in
more detail in Sect. 2.1.
Representations are special classes of group actions , see Chap. 3. It is not
difficult to see that the linear groups can be realized as isotropy groups of certain
elements in suitable representation spaces of the general linear groups.
Representations and isotropy groups will also be used in Sect. 3.10 to define the
exceptional Lie group G2 as an embedded Lie subgroup of .

1.2.3 Connectivity Properties of Linear Groups

Proposition 1.2.21 (Connected Components of Lie Groups) Let G be a Lie
group. Then all connected components of G are diffeomorphic to the connected
component G e of the neutral element e G. In particular, all connected
components have the same dimension.

Proof Let g G and denote the connected component of G containing g by G

g . Consider the left translation, given by

restricted to the connected component G e containing e. The image of this

smooth map is connected and contains g, therefore the image is contained in G g
. By the same argument

has image contained in G e . It follows that

is a diffeomorphism and thus all connected components of G are diffeomorphic.

□

We want to understand how many connected components the classical Lie

groups have.

Theorem 1.2.22 (Connected Components of Classical Groups) Let n ≥ 1.

1. The Lie group has two connected components ,

determined by the sign of the determinant. The Lie groups and
are connected.

2. The special linear groups are connected for all .

3. The Lie group O(n) has two connected components O(n)± , determined by
the sign of the determinant. The Lie group SO(n) = O(n)+ is connected.

4. The Lie groups U(n), SU(n) and Sp(n) are connected.

There are direct proofs for these assertions that the reader can find in the
literature. We chose to give a different argument using homogeneous spaces in
Sect. 3.8.3, which is conceptually clearer and simpler. The proof with
homogeneous spaces utilizes certain actions of the classical groups on
and spheres S m . The assertions then follow by induction over n from the
corresponding (trivial) statements for the groups with n = 1.
The number of connected components of a Lie group G can be identified
with the number of elements of the homotopy group π 0(G). In Sect. 2.6 we will
discuss higher homotopy groups of Lie groups.

1.3 Homomorphisms of Lie Groups

Lie groups have two structures: the algebraic structure of a group and the smooth
structure of a manifold. A homomorphism between Lie groups should be
compatible with both structures.

Definition 1.3.1 Let G and H be Lie groups. A map ϕ: G → H which is smooth

and a group homomorphism, i.e.

is called a Lie group homomorphism . The map ϕ is called a Lie group

isomorphism if it is a diffeomorphism and a homomorphism (hence an
isomorphism) of groups. A Lie group isomorphism ϕ: G → G is called a Lie
group automorphism of G.

Remark 1.3.2 We will show in Theorem 1.8.14 as an application of Cartan’s

Theorem that continuous group homomorphisms ϕ: G → H between Lie groups
are automatically smooth, hence Lie group homomorphisms.

Remark 1.3.3 Occasionally we will call a Lie group homomorphism just a

homomorphism if the meaning is clear from the context.
We consider some examples of Lie group homomorphisms.

Example 1.3.4 Let G and H be Lie groups. The constant map

is always a Lie group homomorphism, called the trivial homomorphism .

Example 1.3.5 Consider with addition and the Lie group S 1. Then

is a surjective homomorphism of Lie groups. The kernel of this map is

Taking products, there is a similar surjective homomorphism .

Example 1.3.6 It is easy to check that the map

is an isomorphism of Lie groups.

Example 1.3.7 We have

We also have

It can be checked that the map

is an isomorphism of Lie groups. This follows from a direct calculation or from

Proposition 1.1.26.

Example 1.3.8 (Universal Covering of SO(3)) We define a Lie group

homomorphism
in the following way: Let be a unit quaternion, | | w | | = 1, and consider
the map

This is an -linear isomorphism of the 4-dimensional real vector space . Since

the map τ w is orthogonal with respect to the standard Euclidean scalar product
on . The map τ w clearly fixes

and therefore restricts to an orthogonal isomorphism

on the orthogonal complement

This shows that ϕ(w) O(3). Since the map ϕ: S 3 → O(3) is continuous, S 3 is
connected and ϕ(1) = I, it follows that ϕ has image in the connected component
SO(3) and hence defines a map

It can be checked that this map is a surjective homomorphism of Lie groups with
kernel { + 1, −1}, cf. Exercise 1.9.20. The homomorphism ϕ defines a connected
double covering of SO(3) by S 3 (this is the universal covering of SO(3), since S
3 is simply connected).

1.4 Lie Algebras

Lie algebras are of similar importance for symmetries and gauge theories as Lie
groups. We will begin with the general definition of Lie algebras and describe in
the next section their relation to Lie groups.

Definition 1.4.1 A vector space V together with a map

is called a Lie algebra if the following hold:

 1. [ ⋅ , ⋅ ] is bilinear.

2. [ ⋅ , ⋅ ] is antisymmetric:

3. [ ⋅ , ⋅ ] satisfies the Jacobi identity :

The map [ ⋅ , ⋅ ]: V × V → V is called the Lie bracket . We will only

consider Lie algebras defined on real or complex vector spaces. Unless stated
otherwise the vector spaces underlying Lie algebras are finite-dimensional.

We collect some examples to show that Lie algebras occur quite naturally (we
discuss many more examples in Sect. 1.5.5).

Example 1.4.2 (Abelian Lie Algebras) Every real or complex vector space
with the trivial Lie bracket [ ⋅ , ⋅ ] ≡ 0 is a Lie algebra. Such Lie algebras are
called abelian. Every 1-dimensional Lie algebra is abelian, because the Lie
bracket is antisymmetric.

Example 1.4.3 (Lie Algebra of Matrices) The vector space of

square matrices with is a real or complex Lie algebra with bracket
defined by the commutator of matrices A, B:

The only axiom that has to checked is the Jacobi identity. This example is very
important, because the Lie algebras of linear groups have the same Lie bracket,
cf. Corollary 1.5.26. It even follows from Ado’s Theorem 1.5.25 that any finite-
dimensional Lie algebra can be embedded into such a matrix Lie algebra.

Example 1.4.4 (Lie Algebra of Endomorphisms) In the same way the vector
space V = End(W) of endomorphisms (linear maps) on a real or complex vector
space W is a real or complex Lie algebra with Lie bracket defined by the
commutator of endomorphisms f, g:
Example 1.4.5 (Lie Algebra Defined by an Associative Algebra) Even more
generally, let A be any associative algebra with multiplication ⋅ . Then the
commutator

defines a Lie algebra structure on A.

Example 1.4.6 (Cross Product on ) The vector space is a Lie algebra

with the bracket given by the cross product:

Again, the only axiom that has to be checked is the Jacobi identity. We will
identify with a classical Lie algebra in Exercise 1.9.14.

Example 1.4.7 (Lie Algebra of Vector Fields on a Manifold) Let M be a

differentiable manifold and the real vector space of smooth vector fields on
M. It follows from Theorem A.1.45 that together with the commutator of
vector fields is a real Lie algebra, which is infinite-dimensional if the dimension
of M is at least one.

As in the case of Lie groups we have two constructions that yield new Lie
algebras from given ones.

Definition 1.4.8 Let (V, [ ⋅ , ⋅ ]) be a Lie algebra. A vector subspace W V is

called a Lie subalgebra if for all w, w ′ W the Lie bracket [w, w ′ ] is an
element of W.

Example 1.4.9 Every 1-dimensional vector subspace of a Lie algebra V is an

abelian subalgebra.

Example 1.4.10 From the geometric interpretation of the cross product it

follows that does not have 2-dimensional Lie subalgebras.

Example 1.4.11 (Intersection of Lie Subalgebras) If W 1, W 2 V are Lie

subalgebras, then the intersection W 1 ∩ W 2 is again a Lie subalgebra of V.

Definition 1.4.12 Let (V, [ ⋅ , ⋅ ] V ) and (W, [ ⋅ , ⋅ ] W ) be Lie algebras over the
same field. Then the direct sum Lie algebra is the vector space V W with the
Lie bracket

Remark 1.4.13 Note that if V, W are Lie subalgebras in a Lie algebra Q which
are complementary as vector spaces, so that Q = V W, it does not follow in
general that Q = V W as Lie algebras. For v, v ′ V, w, w ′ W we have

hence we need in addition

We finally want to define homomorphisms between Lie algebras.

Definition 1.4.14 Let (V, [ ⋅ , ⋅ ] V ) and (W, [ ⋅ , ⋅ ] W ) be Lie algebras. A linear

map ψ: V → W is called Lie algebra homomorphism if

A Lie algebra isomorphism is a bijective homomorphism. An automorphism

of a Lie algebra V is a Lie algebra isomorphism ψ: V → V.

Example 1.4.15 Let V and W be Lie algebras over . The constant map

is always a Lie algebra homomorphism, called the trivial homomorphism .

Example 1.4.16 The injection i: W ↪ V of a Lie subalgebra into a Lie algebra is

of course a Lie algebra homomorphism.

The following notion appears, in particular, in physics:

Definition 1.4.17 Let V be a Lie algebra over and T 1, …, T n a vector space

basis for V. Then we can write

where the coefficients are called structure constants for the given basis
{T a }.

Because of bilinearity the structure constants determine all commutators

between elements of V. The structure constants are antisymmetric in the first two
indices

and satisfy the Jacobi identity

(here we use the Einstein summation convention and sum over d). Conversely,
every set of n × n × n numbers satisfying these two conditions define a
Lie algebra structure on .

1.5 From Lie Groups to Lie Algebras

So far we have discussed Lie groups and Lie algebras as two independent
notions. We now want to turn to a well-known construction that yields for every
Lie group an associated Lie algebra, which can be thought of as an infinitesimal
or linear description of the Lie group.
Recall that for every smooth manifold M, the set of smooth vector fields
on M with the commutator forms a Lie algebra, which is infinite-
dimensional if dimM ≥ 1. We could associate to a Lie group G the Lie algebra
of all vector fields on G. However, as an infinite-dimensional Lie algebra
this is somewhat difficult to handle. It turns out that for a Lie group G there
exists a canonical finite-dimensional Lie subalgebra in which has the
same dimension as the Lie group G itself. This will be the Lie algebra associated
to the Lie group G.

1.5.1 Vector Fields Invariant Under Diffeomorphisms

We first consider a very general situation. Let M be a smooth manifold and Γ an
arbitrary set of diffeomorphisms from M to M.

Definition 1.5.1 We define the set of vector fields on M invariant under Γ by

We have:

Proposition 1.5.2 For every set Γ of diffeomorphisms of M, the set A Γ (M) is

a Lie subalgebra in the Lie algebra with the commutator.
Proof Suppose Γ = { ϕ} consists of a single diffeomorphism. If X, Y A {ϕ}
(M) and , then

and

according to Corollary A.1.51. Hence A {ϕ}(M) is a Lie subalgebra of . The

claim for a general set Γ of diffeomorphisms then follows from

1.5.2 Left-Invariant Vector Fields

We now consider the case of a Lie group G. There exist special diffeomorphisms
on G that are defined by group elements g G.

Definition 1.5.3 For g G we set:

These maps are called left translation , right translation and conjugation by g,
respectively.

Example 1.5.4 For with vector addition and we have

This explains the names left and right translation.

The following properties are easy to check:

Lemma 1.5.5 (Properties of Translations and Conjugation) For all g G
we have:

1. The inverses of left and right translations are given by

The inverse of conjugation is given by

In particular, L g , R g and c g are diffeomorphisms of G.

2. L g and R h commute for all g, h G.

3. .

4. The conjugations c g for g G are Lie group automorphisms of G, called

inner automorphisms .

Remark 1.5.6 Note that left translation L g and right translation R g are not Lie
group homomorphisms for g ≠ e, because

Example 1.5.7 If G is abelian, then L g = R g for all g G and c g = Id G for all

g G. Each of these two properties characterizes abelian Lie groups.

We now set:

Definition 1.5.8 A vector field on a Lie group G is called left-

invariant if L g X = X for all g G. In other words, the set of left-invariant
vector fields on G is A Γ (G), where Γ is the set of all left translations.

We get with Proposition 1.5.2:

Theorem 1.5.9 (The Lie Algebra of a Lie Group) The set of left-invariant
vector fields together with the commutator [⋅ , ⋅ ] of vector fields on the Lie
group G forms a Lie subalgebra
 in the Lie algebra of all vector fields on G. We call the Lie algebra of
(or associated to ) G.

Remark 1.5.10 We could also define the Lie algebra of a Lie group with right-
invariant vector fields. Using left-invariant vector fields is just the standard
convention.

Remark 1.5.11 We defined Lie algebras in general on vector spaces over

arbitrary fields. The Lie algebra of a Lie group, however, is always a real Lie
algebra.

As mentioned before, vector fields, their flows and the commutator are only
defined on smooth manifolds. This is the reason why only Lie groups have an
associated Lie algebra and not other types of groups.
We want to show that there is a vector space isomorphism between the Lie
algebra and the tangent space T e G.

Definition 1.5.12 Let G be a Lie group with neutral element e and associated
Lie algebra . We define the evaluation map

Lemma 1.5.13 The evaluation map is a vector space isomorphism.

Proof The evaluation map is clearly linear. To construct the inverse of a vector
x T e G under the map ev define a vector field X on G by

To show that X is smooth, consider the multiplication map

with differential

Then the following map is smooth

which is just the vector field X.
The vector field X is also left-invariant, because

for all g G and thus L g X = X. The map

is the inverse of ev. □

We can therefore think of the tangent space T e G of the Lie group G at the
neutral element e as having the structure of the Lie algebra . In particular, we
get:

Corollary 1.5.14 The Lie algebra of a Lie group is finite-dimensional with

dimension equal to the dimension of the Lie group. Furthermore, a left-invariant
(or right-invariant) vector field on a Lie group is completely determined by its
value at one point.

It is a non-trivial theorem that any abstract real Lie algebra can be realized by the
construction above (for a proof, see [77, 83]):

Theorem 1.5.15 (Lie’s Third Theorem) Every finite-dimensional real Lie

algebra is isomorphic to the Lie algebra of some connected Lie group.

Lie’s Third Theorem was proved in this form by Élie Cartan. Note that there
may be different, non-isomorphic Lie groups with isomorphic Lie algebras: a
trivial example is given by the Lie groups and (S 1, ⋅ ) whose Lie algebras
are one-dimensional and hence abelian. The orthogonal and spin groups provide
another example, to be discussed in Chap. 6.

1.5.3 Induced Homomorphisms

Just as we get for every Lie group an associated Lie algebra, we get for every
homomorphism between Lie groups a homomorphism between the associated
Lie algebras.

Definition 1.5.16 Let G, H be Lie groups and ϕ: G → H a homomorphism of

Lie groups. If X is a left-invariant vector field on G, we can uniquely define a
left-invariant vector field ϕ X on H by

This defines a map

called the differential or induced homomorphism of the homomorphism ϕ.

Remark 1.5.17 Here are two remarks concerning this definition:

1. Note that ϕ(e) = e for a homomorphism, so that is well-defined.

2. As the notation of the theorem indicates, the push-forward on vector fields is

defined in the case of Lie groups not only for diffeomorphisms, but also for
Lie group homomorphisms acting on left-invariant vector fields. This
definition is possible, because left-invariant vector fields on Lie groups are
determined by their value at one point.

Theorem 1.5.18 (The Differential Is a Lie Algebra Homomorphism) The

differential of a Lie group homomorphism ϕ: G → H is a
homomorphism of Lie algebras.

Proof We have to show that

By Proposition A.1.49 this will follow if we can show that ϕ X is ϕ-related to

X, i.e. that

We have

because ϕ X and X are left-invariant and ϕ is a homomorphism. This proves

the claim. □

Note that it is essential for this argument that the map ϕ is a Lie group
homomorphism.
Corollary 1.5.19 Let H G be an immersed or embedded Lie subgroup.
Then is a Lie subalgebra.

Proof The inclusion i: H ↪ G is a homomorphism of Lie groups and an

immersion. Thus the induced inclusion is an injective homomorphism
of Lie algebras. □

We can ask whether it is possible to reverse these relations:

If G is a Lie group with Lie algebra and a Lie subalgebra, does there
exist a Lie subgroup H in G whose Lie algebra is ?
If is a Lie algebra homomorphism between the Lie algebras of Lie
groups G and H, does there exist a Lie group homomorphism ψ: G → H
inducing ϕ on Lie algebras?
Both questions are related to the concept of integration from a linear object
on the level of Lie algebras to a non-linear object on the level of Lie groups. We
shall answer the first question in Sect. 1.6 and briefly comment here, without
proof, on the second question. The following theorem specifies a sufficient
condition for the existence of a Lie group homomorphism inducing a given Lie
algebra homomorphism (for a proof, see [77, 142]):

Theorem 1.5.20 (Integrability Theorem for Lie Algebra

Homomorphisms) Let G be a connected and simply connected Lie group, H a
Lie group and a Lie algebra homomorphism. Then there exists a unique
Lie group homomorphism ψ: G → H such that ψ = ϕ.

Example 1.5.21 Without the condition that G is simply connected this need not
hold: every Lie algebra homomorphism induces a unique Lie group
homomorphism . However, ϕ does not always induce a Lie group
homomorphism SO(2) → H (see the discussion after Corollary 2.1.13).
Similarly there are homomorphisms for n ≥ 3 (so-called spinor
representations, see Sect. 6.5.2) that do not integrate to homomorphisms SO(n)
→ H (it can be shown that SO(n) has fundamental group for n ≥ 3).

1.5.4 The Lie Algebra of the General Linear Groups

We have defined the Lie algebra associated to a Lie group, but so far we have
not seen any explicit examples of this construction. In this and the subsequent
subsection we want to study the Lie algebra associated to the linear groups,
i.e. closed subgroups of the general linear groups. We can understand the
structure of the corresponding Lie algebras by Corollary 1.5.19 once we have
understood the structure of the Lie algebra of the general linear groups.

Theorem 1.5.22 (Lie Algebra of General Linear Groups) The Lie algebra
of the general linear group is and the Lie bracket
on is given by the standard commutator of matrices:

An analogous result holds for the Lie algebra of for .

We want to prove this theorem. The Lie group is an open subset of

, therefore we can canonically identify the tangent space at the unit element I,

with the vector space

Lemma 1.5.23 If , then the associated left-invariant

vector field on G is given by

where ⋅ denotes matrix multiplication.

Proof To show this, let γ X be an arbitrary curve in G through e and tangent to

X. Then

The last equality sign in this calculation can be understood by considering each
entry of the time-dependent matrix A ⋅ γ X (t) separately. □

Lemma 1.5.24 Let be vector fields on an open subset U of a Euclidean

space and curves tangent to and at a point p U. Then

Proof Let e 1, …, e N be the standard basis of the Euclidean space and write

Then, because of [e k , e l ] = 0, we get

We can now prove Theorem 1.5.22.

Proof Since is an open subset of a Euclidean space, we can calculate

the commutator of the vector fields at the point I by

This proves the assertion. □

We would like to mention the following conceptually interesting theorem

concerning Lie algebras (for the proof in a special case, see Proposition 2.4.4;
the general proof can be found in [77, 83]).

Theorem 1.5.25 (Ado’s Theorem) Let be a finite-dimensional Lie algebra

over . Then there exists an injective Lie algebra homomorphism of
into for some n.

As a consequence of Ado’s Theorem every Lie algebra over is isomorphic

to a Lie subalgebra of for some n.
We will later show in Theorem 1.6.4 that for a given Lie group G, every Lie
subalgebra is the Lie algebra of a connected Lie subgroup H G.
Therefore Lie’s Third Theorem 1.5.15 follows from Ado’s Theorem 1.5.25 and
Theorem 1.6.4, applied to some general linear group .

1.5.5 The Lie Algebra of the Linear Groups

As a corollary to Theorem 1.5.22 and Corollary 1.5.19 we get:

Corollary 1.5.26 (Lie Algebra of Linear Groups) If the Lie algebra of an

embedded or immersed Lie subgroup of is identified in the canonical
way with a Lie subalgebra of , then the Lie bracket on the Lie
subalgebra is the standard commutator of matrices.

As simple as this corollary may seem, it is in fact very useful. In general it can
be quite difficult to calculate the commutator of two vector fields on a given
manifold. Corollary 1.5.26 shows that this is very easy for left-invariant vector
fields on Lie subgroups of general linear groups.

Theorem 1.5.27 (Lie Algebras of Classical Groups) We can identify the

Lie algebras of the classical groups with the following real Lie subalgebras of
the Lie algebra .

1. The Lie algebras of the special linear groups are:

2. The Lie algebras of the orthogonal , unitary and symplectic groups are:

3. The Lie algebras of the special orthogonal and special unitary groups
are:
Remark 1.5.28 We can check directly that these subsets of are real
vector subspaces and closed under the commutator; see Exercise 1.9.16.

We prove Theorem 1.5.27.

Proof

1. Let and suppose that . We shall show in Sect. 1.7

(without using the results here) that and that

for all . We get

hence

Since we already know the dimension of from Theorem 1.2.17, the

assertion follows by calculating the dimension of the subspace on the right
of this inclusion.
The claim for follows, because under the adjoint map χ, the
group gets identified according to Proposition 1.1.25 with the
submanifold

The tangent space to the neutral element I is contained in

which corresponds under χ to

since

by Proposition 1.1.26. The claim follows by a similar dimension

argument as before.

2. If A(t) is a curve in O(n) through I with , then A(t)A(t) T = I, hence

 i.e. M is skew-symmetric. The claim then follows by comparing the
dimensions of O(n) and the vector space of skew-symmetric matrices. The
cases of u(n) and sp(n) follow similarly.

3. The case of su(n) is clear by a similar dimension argument as before. The

case of follows, because if , then automatically tr(M) = 0.
□

Example 1.5.29 The Lie algebra has dimension 1 and is equal to ,

spanned by i.

Example 1.5.30 The Lie algebra has dimension 1 and consists of the
skew-symmetric 2 × 2-matrices. A basis is given by the rotation matrix

The Lie algebra is isomorphic to , because both are 1-dimensional and

abelian.

Example 1.5.31 The Lie algebra has dimension 3 and consists of skew-
symmetric 3 × 3-matrices. A basis is given by the rotation matrices

These matrices satisfy

where ε abc is totally antisymmetric in a, b, c with ε 123 = 1.

Example 1.5.32 The Lie algebra has dimension 3 and consists of the
skew-Hermitian 2 × 2-matrices of trace zero. We consider the Hermitian
Pauli matrices :
 Then a basis for is given by the matrices

The commutators of these matrices are

The map

is a Lie algebra isomorphism.

Example 1.5.33 The Lie algebra has dimension 8 and consists of the
skew-Hermitian 3 × 3-matrices of trace zero. We consider the Hermitian Gell-
Mann matrices:

Then a basis for is given by the matrices for a = 1, …, 8. The

matrices iλ a for a = 1, 2, 3 span a Lie subalgebra, isomorphic to .

Example 1.5.34 The Lie algebra has dimension 3 and is equal to ,

spanned by the imaginary quaternions i, j, k. If we set

then

The map
is a Lie algebra isomorphism.

Example 1.5.35 The Lie algebra is equal to .

Example 1.5.36 The Lie algebra has dimension 3 and consists of the
real 2 × 2-matrices of trace zero. A basis is given by the matrices

with commutators

Example 1.5.37 The Lie algebra has dimension 6 and consists of the
complex 2 × 2-matrices of trace zero. It is also a complex Lie algebra of complex
dimension 3. A complex basis is given by the same matrices H, X, Y as above for
. In analogy to the quantum angular momentum and quantum harmonic
oscillator, X is sometimes called the raising operator and Y the lowering
operator . According to Exercise 1.9.18, as a complex Lie algebra, is
isomorphic to the complex Lie algebra .
The Lie algebra plays a special role in physics, because as a real Lie
algebra it is isomorphic to the Lie algebra of the Lorentz group of 4-dimensional
spacetime (see Sect. 6.8.2).

Example 1.5.38 (The Heisenberg Lie Algebra) The Lie algebra of the
Heisenberg group Nil3 is

A basis is given by the matrices

satisfying

We see that z commutes with every element in the Lie algebra , i.e. z is a
central element. Furthermore,

so that is an example of a nilpotent Lie algebra.

1.6 From Lie Subalgebras to Lie Subgroups

Let G be a Lie group with Lie algebra . In this section we want to show that
there exists a 1-to-1 correspondence between Lie subalgebras of and connected
(immersed or embedded) Lie subgroups of G (we follow [142]). We need some
background on distributions and foliations that can be found in Sect. A.1.12.

Definition 1.6.1 Let be a Lie subalgebra. If we consider as the set of

left-invariant vector fields on G, then is a distribution on G, denoted by .
Equivalently, if we think of as the tangent space T e G and as a vector
subspace, then the distribution is defined by

Lemma 1.6.2 The distribution associated to a Lie subalgebra is

integrable .

Proof Let V 1, …, V d be left-invariant vector fields on G defined by a vector

space basis for . Then V i is a section of for all i = 1, …, d and since is a
subalgebra, the commutators [V k , V l ] are again sections of . If X and Y are
arbitrary sections of , then there exist functions f i , g i on G such that
We get

This is a section of . Thus the distribution is integrable. □

Definition 1.6.3 For a Lie subalgebra , let H denote the maximal

connected leaf of the foliation through the neutral element e G.

Theorem 1.6.4 (The Immersed Lie Subgroup Defined by a Lie

Subalgebra) The immersed submanifold H is the unique, connected,
immersed Lie subgroup of G with Lie algebra .

Remark 1.6.5 The subgroup H is sometimes called the integral subgroup

associated to the Lie algebra .

Proof We first show that H is a subgroup in the algebraic sense: Let g H.

Since is left-invariant, we have

hence is a connected leaf of , containing g −1 g = e. By maximality of H

we have

Hence if g, h H, then g −1 h H, showing that H is a subgroup.

We want to show that the group operations on H are smooth with respect to
the manifold structure. The map

is smooth and has image in H. Since H is the leaf of a foliation, it follows

from Theorem A.1.56 that

is smooth. We prove the remaining statement on the uniqueness of H below.

□
It remains to show that H is the unique connected immersed Lie subgroup with
Lie algebra . We need the following:

Proposition 1.6.6 (Connected Lie Groups Are Generated by Any Open

Neighbourhood of the Neutral Element) Let G be a connected Lie group
and U G an open neighbourhood of e. Then

where

Proof This follows from Exercise 1.9.4. □

Here is an immediate consequence.

Corollary 1.6.7 Let K and K ′ be Lie groups, where K ′ is connected. Suppose

that ϕ: K → K ′ is a Lie group homomorphism, so that ϕ(K) contains an open
neighbourhood of e K ′ . Then ϕ is surjective. In particular, if K K ′ is an
open subgroup, then K = K ′ .

We now prove the uniqueness part in Theorem 1.6.4.

Proof Let K be another connected, immersed Lie subgroup with Lie algebra .
Then K must also be a connected leaf of the foliation through e G, hence
by maximality of H we get K H. The differential of the inclusion i: K ↪ H is
an isomorphism at every point, hence K H is an open subgroup. The assertion
then follows from Corollary 1.6.7. □

Example 1.6.8 According to Example 1.1.42 the 1-dimensional Lie

subalgebras of the Lie algebra of the torus T 2 define embedded Lie subgroups if
they have rational slope in and immersed Lie subgroups if they have
irrational slope.

1.7 The Exponential Map

We saw above that the tangent space T e G of a Lie group G at the neutral
element e G has the structure of a Lie algebra . In this section we want to
study the famous exponential map from to G, which is defined using integral
curves of left-invariant vector fields (we follow [14, Sect. 1.2] for the
construction).

1.7.1 The Exponential Map for General Lie Groups

For the following statements some background on integral curves and flows of
vector fields can be found in Sect. A.1.9.

Theorem 1.7.1 (Integral Curves of Left-Invariant Vector Fields) Let G be

a Lie group and its Lie algebra. Let

denote the maximal integral curve of a left-invariant vector field through

the neutral element e G. Then the following holds:

1. ϕ X is defined on all of .

2. is a homomorphism of Lie groups, i.e.

3. ϕ sX (t) = ϕ X (st) for all .

Definition 1.7.2 The homomorphism is called the one-parameter

subgroup of the Lie group G determined by the left-invariant vector field X.

We prove Theorem 1.7.1 in a sequence of steps. Let

denote the maximal integral curve of the vector field X through e, satisfying

Lemma 1.7.3 For all s, t I with s + t I the following identity holds:

Proof Let g = ϕ X (s) G. Consider the smooth curves

and

It is easy to show that both η and are integral curves of X with .

Hence by the uniqueness of integral curves (which is a theorem about the
uniqueness of solutions to ordinary differential equations) we have

This implies the claim. □

Lemma 1.7.4 We have t max = ∞ and t min = −∞.

Proof Suppose t max < ∞ and set α = min{t max , | t min | } < ∞. Consider the
curve

It is easy to check that γ is an integral curve of X with γ(0) = e. However,

by construction, hence γ is an extension of ϕ X , contradicting the choice of t max .

This shows that t max = ∞ and similarly t min = −∞. □

Lemma 1.7.5 ϕ sX (t) = ϕ X (st) for all .

Proof Fix and consider the curve

It is easy to show that δ is an integral curve of the vector field sX with δ(0) = e.
Hence by uniqueness ϕ X (st) = ϕ sX (t). □
Definition 1.7.6 Let denote the integral curve through e G for an
element . Then we define the exponential map

See Fig. 1.1.

Fig. 1.1 Exponential map

Remark 1.7.7 The reason for the name exponential map will become apparent
in Sect. 1.7.3. Note that by definition the exponential map of the Lie group G
maps the Lie algebra to the connected component G e of the neutral element e.
Elements in other connected components can never be in the image of the
exponential map.

Example 1.7.8 The simplest example is the exponential map of the abelian Lie
group with vector addition. We can canonically identify the Lie algebra
with . Then the exponential map is the identity map, since the left-invariant
vector fields on G are the constant (parallel) vector fields. In this particular case,
the exponential map is therefore a diffeomorphism between the Lie algebra and
the Lie group.

Proposition 1.7.9 (Properties of the Exponential Map) The exponential map

has the following properties, explaining the name one-parameter subgroup:

1. exp(0) = e

2. exp((s + t)X) = exp(sX) ⋅ exp(tX)

3. exp(−X) = (expX)−1

for all and .

Proof This is an exercise. □

Remark 1.7.10 One-parameter subgroups are the immersed or embedded Lie

subgroups determined as in Theorem 1.6.4 by 1-dimensional (abelian)
subalgebras of .

Definition 1.7.11 Let X be a left-invariant vector field on a Lie group G. Then

we denote its flow through a point p G by ϕ t X ( p) or ϕ t ( p). It is
characterized by

The one-parameter subgroup ϕ X (t) determined by X is in this notation ϕ t (e).

Proposition 1.7.12 (Relation Between the Flow and the Exponential Map)
Let G be a Lie group and X a left-invariant vector field. Then its flow ϕ t ( p)
through a point p G is defined for all ,

and given by

Proof Define ϕ t ( p) for all by the right-hand side. It is clear that

Furthermore,

since X is left-invariant. This implies the claim by uniqueness of solutions of

ordinary differential equations. □
Remark 1.7.13 Integral curves of vector fields are defined for all times usually
only on compact manifolds. It is a special property of Lie groups that integral
curves of left-invariant vector fields are defined for all times, even on non-
compact Lie groups.

We want to prove a property of the exponential map that is sometimes useful in

applications.

Proposition 1.7.14 (Exponential Map Is a Local Diffeomorphism) Under

the canonical identifications

the differential of the exponential map

is the identity map. In particular, there exist open neighbourhoods V of 0 in

and U of e in G such that

is a diffeomorphism.

Proof Let and γ(t) = tX the associated curve in . Then and

Remark 1.7.15 In general, the exponential map is neither injective nor

surjective and hence not a global diffeomorphism between the Lie algebra and
the Lie group (of course, the exponential map can only be a diffeomorphism if
the Lie group itself is not compact and diffeomorphic to a vector space). See
Corollary 1.7.20 for a situation when the exponential map is surjective. It is
known that the exponential map is a diffeomorphism in the case of simply
connected nilpotent Lie groups (see [83] for a proof).

Recall from Theorem 1.5.18 that every homomorphism between Lie groups
induces a homomorphism between Lie algebras. The exponential map behaves
nicely with respect to these homomorphisms.

Theorem 1.7.16 (Induced Homomorphisms on Lie Algebras and the

Exponential Map) Let ψ: G → H be a homomorphism between Lie groups
and the induced homomorphism on Lie algebras. Then
i.e. the following diagram commutes:

Proof Consider the curve γ(t) = ψ(exptX) for . Then

and

Here we used that ψ is a homomorphism and X left-invariant. We conclude that γ

is the (unique) integral curve of the left-invariant vector field ψ X through e
H and therefore

Corollary 1.7.17 (Exponential Map for Embedded Lie Subgroups) Let G

be a Lie group and H G an embedded Lie subgroup with exponential maps

Then for the following identity holds

Proof This follows from Theorem 1.7.16 with the embedding i: H ↪ G. □

The following generalization of Theorem 1.7.1 to time-dependent vector fields is

sometimes useful in applications.

Theorem 1.7.18 (Integral Curves of Time-Dependent Vector Fields on Lie

Groups) Suppose that G is a Lie group and a smooth map. Let
X(t) denote the associated left- (or right-)invariant time-dependent vector field
on G. Then there exists a unique smooth integral curve g: [0, 1] → G such that

Proof We only indicate the idea of the proof in the case of a left-invariant
vector field X(t). Details can be found in [14] and are left as an exercise. Let Z be
the vector field on defined by

On the interval [0, δ], for δ > 0 small enough, there exist integral curves (g(t), t)
and (h(t), t + δ) of Z with g(0) = h(0) = e. Then

defines an extension of g(t) to an integral curve of X on [0, 2δ]. □

1.7.2 The Exponential Map of Tori

It is important to understand the exponential map of the torus T n , because
compact Lie groups always contain embedded Lie subgroups isomorphic to tori
(see Exercise 1.9.11 for explicit examples of tori contained in the classical
groups).

Proposition 1.7.19 (Exponential Map of Tori) The exponential map of every

torus T n is surjective.

Proof This follows, because according to Example 1.7.8 the exponential map
of is surjective. □

Every element of a compact connected Lie group G is contained in some

embedded torus subgroup (for a proof, see [24, Sect. IV.1]). Then
Corollary 1.7.17 implies:

Corollary 1.7.20 (Exponential Map of Compact Connected Lie Groups)

The exponential map of a compact connected Lie group G is surjective.

Remark 1.7.21 As Exercise 1.9.27 shows, this is not true in general for non-
compact connected Lie groups like . See Exercise 1.9.28 for a statement
which is true in the general case.

Example 1.7.22 Every one-parameter subgroup of SO(3) is given by the

subgroup of rotations around a common axis in (see Example 1.7.32).
Corollary 1.7.20 then implies that any rotation of can be obtained from the
identity by rotating around a fixed axis by a certain angle.

The following assertion was discussed in Example 1.1.42.

Proposition 1.7.23 (Embedded and Immersed One-Parameter Subgroups)

A torus of dimension at least two has both embedded and immersed one-
parameter subgroups.

Corollary 1.7.24 Every Lie group G that contains a torus of dimension at

least two has both embedded and immersed one-parameter subgroups.

Example 1.7.25 The Lie groups SO(4) and SU(3) contain embedded tori of
dimension two and thus immersed one-parameter subgroups . The Lie groups
SO(3) and SU(2), on the other hand, only contain embedded tori of dimension
one. It can be shown that every one-parameter subgroup of SO(3) and SU(2) is
closed, hence isomorphic to S 1 (cf. Exercise 1.9.25 for the case of SU(2)).
This is intuitively clear for SO(3), because the one-parameter subgroups are
rotations around a fixed axis and thus return to the identity after a rotation by 2π.
The result can also be interpreted for SO(4): we can define an embedded Lie
subgroup T 2 = SO(2) × SO(2) in SO(4) as those rotations which preserve a
splitting of into two orthogonal planes and only rotate each plane in
itself. If the velocities of the two rotations have an irrational ratio, then both
rotations never return at the same time to the identity. This corresponds to an
immersed one-parameter subgroup of SO(4).

1.7.3 The Matrix Exponential

In this section we want to determine the exponential map for the linear groups.
Let .

Definition 1.7.26 Let be a square matrix. Then we set

Lemma 1.7.27 (Convergence of Exponential Series) For any square matrix

 the series converges in each entry.

Proof Let | | ⋅ | | denote the Euclidean norm on . Define the operator norm of
a matrix by

Then | | ⋅ | | is indeed a norm on the vector space of square matrices and satisfies |
| MN | | ≤ | | M | | ⋅ | | N | |. Since the exponential series for real numbers
converges, it follows that the exponential series is Cauchy and hence
converges. □

Lemma 1.7.28 (Exponential of a Sum for Commuting Matrices) If

matrices commute, AB = BA, then

In particular, e −A is the inverse of e A , so that .

Proof This is Exercise 1.9.23. □

The following is immediate.

Theorem 1.7.29 For every the map

is smooth and satisfies

We get:

Corollary 1.7.30 (Exponential Map of Linear Group Is Matrix

Exponential) Let

Then

where exp on the left denotes the canonical exponential map from Lie algebra to
Lie group. The same formula holds for the exponential map of any linear group.

Proof For let denote the associated left-invariant vector

field on . According to Lemma 1.5.23, is given at a point
by

This shows that the map

from Theorem 1.7.29 is the integral curve of the vector field through I. The
first claim now follows by Definition 1.7.6 of the exponential map. The second
claim concerning linear groups follows by Corollary 1.7.17. □

Example 1.7.31 The simplest non-trivial case of this theorem is the exponential
map

Example 1.7.32 A slightly less trivial example is the matrix exponential of tr,
where and

is the generator of from Example 1.5.30. It is easy to see that

and thus

for all n ≥ 0. Hence

This is just the matrix for a rotation in by an angle t. Similarly the matrix
exponential of tr 3, with
one of the generators of from Example 1.5.31, is

which is the matrix for a rotation in around the z-axis. Rotations around other
axes in are given by one-parameter subgroups conjugate to the one defined
by r 3, showing that all one-parameter subgroups of SO(3) are closed.

The proof of the following well-known formula uses that the determinant is
multilinear in the columns of a matrix and thus only holds for real and complex
matrices.

Theorem 1.7.33 (Determinant of Matrix Exponential) Let

where or . Then

Proof We first calculate the differential D I det. Let X = (x 1, …, x n ) be an

arbitrary n × n-matrix with column vectors . We have by multilinearity
and antisymmetry of the determinant

Then
Consider the curve

Then

and for all

The unique solution of this differential equation for γ(t) is

This implies the assertion with t = 1. □

Example 1.7.34 Let

be a real or complex diagonal matrix. Then

and the equation

is trivially satisfied. The same argument works for upper triangular matrices.

Using Theorem 1.7.16 we can write the statement of Theorem 1.7.33 as follows:

Corollary 1.7.35 The determinant

 is a group homomorphism with differential given by the trace

Notice that the trace is indeed a Lie algebra homomorphism to the abelian Lie
algebra .

1.8 Cartan’s Theorem on Closed Subgroups

Our aim in this section is to prove Cartan’s Theorem 1.1.44 (we follow [14, 24]
and [142]). This theorem is important, because we used it, for example, to show
that closed subgroups of the general linear groups are embedded Lie subgroups.
We will also employ it later to show that isotropy groups of Lie group actions on
manifolds are embedded Lie subgroups. The proof of Cartan’s Theorem is one of
the more difficult proofs in this book and follows from a sequence of
propositions. One direction is quite easy.
Let G be a Lie group.

Definition 1.8.1 Let H G be a subset. A chart of G such that

for some k < n is called a submanifold chart or flattener for H around p.

Proposition 1.8.2 Let H G be an embedded Lie subgroup. Then H is a

closed subset in the topology of G.

Proof Suppose that H G is an embedded Lie subgroup. In a submanifold

chart U of G around e for the embedded submanifold H, the set H ∩ U is closed
in U.
Suppose y is a point in the closure and let x n be a sequence in H
converging to y. Then x n −1 y is a sequence in G converging to e. Hence for
sufficiently large index n, the element x = x n H satisfies x −1 y U and thus
y xU.
Since the group operations on G are continuous, the closure is a subgroup
of G. It follows that and . Thus y H and H is
closed. □

The converse statement is more difficult. Assume from now on that H G is a

subgroup in the algebraic sense, which is a closed set in the topology of G. To
show that H is a k-dimensional embedded Lie subgroup of the n-dimensional Lie
group G, we have to find around every point p H a chart of G
which is a submanifold chart for H around p. The following argument shows that
it suffices to find a submanifold chart for H around e.

Proposition 1.8.3 (Submanifold Charts) Let be a submanifold

chart for H around e H. Suppose p H. Then

is a submanifold chart for H around p. Here L p denotes left translation on G by

p.

Proof Note that

since H is a subgroup of G, hence

This implies

and

We want to find a submanifold chart for H around e. It turns out that we first
have to find a candidate for the Lie algebra of the subgroup H.

Proposition 1.8.4 (The Candidate for the Lie Algebra of H) Let H G be

a subgroup in the algebraic sense which is a closed set in the topology of G.
Then

is a vector subspace of .

To prove Proposition 1.8.4 note that if , then for all , since we

can choose . It remains to show that if , then . We have
to understand terms of the form

This is the purpose of the following proposition.

Proposition 1.8.5 (Special Case of the Baker–Campbell–Hausdorff

Formula) Let G be a Lie group with Lie algebra . Then for arbitrary
vectors we have

where ε > 0 is small enough and O(t 2) is some function of t such that O(t 2)⁄t 2
stays finite as t → 0.

Proof According to Proposition 1.7.14 the exponential map is a

diffeomorphism from an open neighbourhood V of 0 in onto an open
neighbourhood W of e in G. We can thus introduce so-called normal coordinates
on W: Choose a basis (v 1, …, v n ) for the vector space . Then there is a unique
chart (W, ϕ) of G around e with

such that

Let

Let μ denote group multiplication in G:

Utilizing the chart ϕ this induces a map

where is a small open neighbourhood of (0, 0). The map is defined by

We then have to show that

This follows from the Taylor formula for since

hence

and

hence

We will use this proposition in the following special form.

Corollary 1.8.6 (Lie Product Formula) For arbitrary vectors and all

Proof This follows from Proposition 1.8.5 with the general formula exp(Z) n =
exp(nZ) for any .□

We can now finish the proof of Proposition 1.8.4.

Proof If , then

since H is a subgroup of G. Corollary 1.8.6 together with the assumption that H

is a closed subset implies exp(t(X + Y )) H for all and thus .□

Let be the vector subspace of from Proposition 1.8.4 and an arbitrary

complementary vector subspace of , so that

We now start to construct the submanifold chart for H around e. We fix an

arbitrary norm | | ⋅ | | on the vector space .
Proposition 1.8.7 (Choice of the Open Subset ) There exists an open
neighbourhood of 0 in , so that

Proof Suppose that for every open neighbourhood of 0 in we have

Then there exists a non-zero sequence in converging to 0 so that exp(Z

n) H. Let K denote the set

and choose for every a positive integer such that c n Z n K. Since

K is a compact set, we can assume (after passing to a subsequence) that
converges to some Z K. Then

Since exp(Z n ) H for all we get with Lemma 1.8.8 below that .
However, and is complementary to , therefore Z = 0. This
contradicts that 1 ≤ | | Z | | ≤ 2. □

In the proof we used the following lemma.

Lemma 1.8.8 Let be a sequence of non-zero vectors in with exp(Z n )

H and Z n → 0 as n → ∞. Suppose that the limit

exists. Then

and thus .

Proof Let be fixed and define

We claim that

Note that
 This implies the claim, because Z n → 0. We get

However,

Since H is a closed subset it follows that exp(tW) H. □

The following map will be the (inverse of the) submanifold chart.

Lemma 1.8.9 (Definition of the Map F) The map

is smooth and its differential at 0 is the identity. Thus F is a local

diffeomorphism on a neighbourhood of 0.

Proof The differential of F maps

We will show that this map is the identity. For we have

A similar argument applies for .□

Proposition 1.8.10 (The Map F Defines a Submanifold Chart) Let

be the map from Lemma 1.8.9 . Then there exists a small neighbourhood V
of 0 in such that F(V ) = U is an open neighbourhood of e in G and

is a submanifold chart for H around e.

Proof According to Lemma 1.8.9 we can choose an open neighbourhood

of 0 in so that

is a diffeomorphism. According to Proposition 1.8.7 we can choose small

enough such that

Note that exp(Y ) H for all , hence

and

We conclude that

and thus
(1.1)
This implies the claim. □

With Proposition 1.8.3 this finishes the proof of Cartan’s Theorem. From the
proof we see:

Corollary 1.8.11 (The Lie Algebra of an Embedded Lie Subgroup) Let H

G be as in Cartan’s Theorem 1.1.44 . Then the Lie algebra of H is given by

Proof We denote the Lie algebra of H for the moment by L(H). It is clear that
. We know from Proposition 1.8.4 that is a vector subspace in and
from Eq. (1.1) that has the same dimension as H. This implies the claim. □

We collect some consequences of Cartan’s Theorem:

Theorem 1.8.12 (Kernel of Lie Group Homomorphism) Let ϕ: G → K be a

Lie group homomorphism. Then H = ker ϕ is an embedded Lie subgroup of G
with Lie algebra .

Proof It is clear that H is a subgroup of G and closed, because H = ϕ −1(e). By

Cartan’s Theorem 1.1.44, H is an embedded Lie subgroup of G.
 Let X be an element in the Lie algebra of H. Then exptX H for all ,
hence

This implies that

Conversely, let with ϕ X = 0. Then

This implies for all

Therefore, the curve ϕ(exptX) is constant and equal to ϕ(e) = e. This implies
exptX H for all and thus .□

Proposition 1.8.13 (Image of Compact Lie Group Under Homomorphism)

Let ϕ: G → H be a Lie group homomorphism. If G is compact, then the image of
ϕ is an embedded Lie subgroup of H.

Proof This is clear, because the image of ϕ is compact, hence closed. □

Theorem 1.8.14 (Continuous Group Homomorphisms Between Lie Groups

Are Smooth) Let ϕ: G → K be a continuous group homomorphism between
Lie groups. Then ϕ is smooth and thus a Lie group homomorphism.

We use the following lemma from topology, whose proof is left as an exercise.

Lemma 1.8.15 Let X, Y be topological spaces and f: X → Y a map. Define the

graph of f by

If Y is Hausdorff and f continuous, then Γ f is closed in the product topology of X

× Y.

We also need the following.

Lemma 1.8.16 Let ψ: K → G be a Lie group homomorphism, which is a

homeomorphism. Then ψ is a diffeomorphism, hence a Lie group isomorphism.

Proof It suffices to show that is injective, because K and G have the

same dimension. Suppose is the kernel of ψ and H the kernel of ψ.
According to Theorem 1.8.12 the subalgebra is the Lie algebra of H and thus
H ≠ {e}. This shows that ψ is not injective and hence not a homeomorphism. □

We now prove Theorem 1.8.14.

Proof The graph Γ ϕ G × H is a closed subgroup of the Lie group G × H,

thus an embedded Lie subgroup by Cartan’s Theorem 1.1.44. The projection
pr1: G × H → G restricts to a smooth homeomorphism

on the embedded submanifold Γ ϕ with continuous inverse

It follows by Lemma 1.8.16 that p is a diffeomorphism and thus

is a smooth map. □

Corollary 1.8.17 (Uniqueness of Smooth Lie Group Structure) Let G be a

topological manifold which is a topological group. Then there is at most one
smooth structure on G so that G is a Lie group.

Proof Suppose G ′ and G ″ are smooth Lie group structures on G. The identity
map

is a group isomorphism and a homeomorphism. By Theorem 1.8.14 this map is a

diffeomorphism. □

Corollary 1.8.18 (Embeddings of Compact Lie Groups) Let G, H be Lie

groups, G compact, and ϕ: G → H an injective Lie group homomorphism. Then
ϕ is a Lie group embedding, i.e. a Lie group isomorphism onto its image, an
embedded Lie subgroup of H.

Proof Since G is compact, the image of ϕ is compact, hence closed in H. This

shows that the image of ϕ is an embedded Lie subgroup by Cartan’s
Theorem 1.1.44. Moreover, ϕ: G → ϕ(G) is a closed map, hence a
homeomorphism. Lemma 1.8.16 implies that ϕ is an isomorphism onto its
image. □

1.9 Exercises for Chap. 1

1.9.1. Let G be a topological group and G e the connected component containing
the neutral element e. Prove that G e is a normal subgroup of G.
1.9.2 (From [ 135 ]). Let G be a topological group which is locally
Euclidean. Prove that G is Hausdorff.
1.9.3. Let G be a connected topological group and H G an open subgroup.
Prove that H = G.
1.9.4. Let G be a connected topological group and U G an open
neighbourhood of e. Prove that the set

where

contains an open subgroup of G. Deduce that W = G.

1.9.5 (From [ 129 ]). Let G be a group which is at the same time a manifold
so that the multiplication map

is smooth.

1. Show that the multiplication map μ is a submersion.

2. Prove that the map

is smooth and therefore G is a Lie group.

1.9.6. Prove Proposition 1.1.25 that realizes the space of
quaternionic matrices via the adjoint as a subspace of the space of
complex matrices.
1.9.7. Prove Proposition 1.1.26 on the properties of the adjoint for
quaternionic matrices.
1.9.8. Show that the determinant of a matrix with is
equal to

1.9.9. Show that every Lie group homomorphism between the Lie
groups (S 1, ⋅ ) and is the constant map to .
1.9.10.

1. Find an explicit Lie group embedding

2. Write as A = A 1 + iA 2 with A 1, A 2 real matrices and find

Lie group embeddings

and

3. Identify the image of Sp(n) in under the adjoint map χ and

find a Lie group embedding

1.9.11. Let T n denote the torus of dimension n.

1. Find Lie group embeddings

2. Find Lie group embeddings

 1.9.12. Find an explicit Lie group homomorphism

with discrete kernel.

1.9.13. Show that the Lie group homomorphisms from Example 1.3.7 and
Example 1.3.8 together give a Lie group homomorphism ψ: SU(2) → SO(3)
equal to

Deduce that

1.9.14. Recall that according to Example 1.4.6 the vector space is a Lie
algebra with bracket given by the cross product:

Find an explicit isomorphism of with the Lie algebra .

1.9.15. Prove that for n ≥ 1 the sphere S 2n does not admit the structure of a
Lie group.
1.9.16 (From [ 24 ]). Consider the Lie algebras of the classical linear
groups G from Theorem 1.5.27.

1. Show directly that the subsets of defined in Theorem 1.5.27 are

real vector subspaces and closed under the commutator of matrices.

2. Show also by a direct calculation that these subsets are closed under the
following map:

where is an element of the Lie algebra and g G an element of the

corresponding linear group (we will identify this map with the adjoint
representation in Sect. 2.1.5).
1.9.17.
1. Prove that there are, up to isomorphism, only two 2-dimensional real Lie
algebras.

2. Show that is not isomorphic to .

1.9.18. The Lie algebra is spanned as a real vector space by the
matrices τ 1, τ 2, τ 3 from Example 1.5.32. The Lie algebra is spanned as a
complex vector space by the matrices H, X, Y from Example 1.5.36.

1. Show that the matrices τ 1, τ 2, τ 3 are a complex basis for and express
this basis in terms of H, X, Y.

2. Show that as complex Lie algebras is isomorphic to , where

the Lie bracket on the right is the complex linear extension of the Lie
bracket of .
1.9.19. Consider the Lie group U(n) with Lie algebra .

1. Find an explicit Lie algebra isomorphism

2. Find an explicit group isomorphism

where is a normal subgroup.

1.9.20. Recall that

We identify quaternions , where , with the following

matrices:

Consider the following isomorphism of real vector spaces:

The Euclidean norm of an element is given by | | x | |2 = detX. Under
this identification we set:

1. Prove that this map is well-defined and yields a homomorphism

of Lie groups.

2. Show that ϕ is surjective and that its kernel consists of {I, −I}.
1.9.21 (From [ 98 ]). We identify the Lie group SU(2) and the quaternions
with subsets of the complex 2 × 2-matrices as in Exercise 1.9.20. Consider the
following isomorphism of real vector spaces:

The Euclidean norm of an element is given by | | x | |2 = detX.

Under this identification we set:

where ⋅ denotes matrix multiplication.

1. Prove that this map is well-defined and yields a homomorphism

of Lie groups.

2. Show that ψ is surjective with kernel {(I, I), (−I, −I)}.

1.9.22. Prove that every Lie group homomorphism ρ: S 1 → S 1 is of the form
for some .
1.9.23. Show that if matrices with commute,
AB = BA, then

1.9.24. Calculate exp(sτ a ) SU(2) for and the basis τ 1, τ 2, τ 3 of the

Lie algebra from Example 1.5.32.
1.9.25. Consider the Lie group SU(2) with Lie algebra .

1. Show that every element can be written as

with ,A SU(2) and

2. Prove that every one-parameter subgroup of SU(2) is closed, i.e. its image is
isomorphic to U(1).
1.9.26. Consider the Lie algebra from Example 1.5.33 with the basis v
1, …, v 8, where and λ a are the Gell-Mann matrices.

1. Prove that the following three sets of basis vectors

where α, β, γ, δ are certain real numbers, span Lie subalgebras of

isomorphic to . Determine α, β, γ, δ.

2. Prove that the one-parameter subgroups generated by each of the basis

vectors v 1, …, v 8 are closed.
1.9.27 (From [ 24 ]). Consider a matrix
1. Calculate e A and .

2. Prove that the exponential map is not surjective.

1.9.28.

1. Let G be a connected Lie group. Show that every group element g G is of

the form

for finitely many vectors X 1, …, X n in the Lie algebra of G.

2. Let ϕ: G → H be a Lie group homomorphism, where G is connected.

Suppose that the induced Lie algebra homomorphism is trivial.
Prove that ϕ is trivial.
1.9.29 (From [ 77 ]).

1. Calculate the k-th power of the nilpotent matrix

2. Calculate the k-th power of a Jordan block matrix λI n + N with .

3. Calculate e tA for a matrix A in Jordan normal form and .

1.9.30 (From [ 77 ]). Let .

1. Use Exercise 1.9.29 to show that the set

 is bounded in if and only if A is diagonalizable with purely
imaginary eigenvalues.

2. Show that e A = I if and only if A is diagonalizable with all eigenvalues

contained in .

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
 [MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
 [MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
 High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)
59.
Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
 (1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]
77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9, 83–92 (1983)
93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
 plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123.
 Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]
139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions. Chin.
Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_2
Chapter 2 Lie Groups and Lie Algebras:
Representations and Structure Theory
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

At least locally, fields in physics can be described by maps on spacetime with

values in vector spaces. Since symmetry groups in field theories act on fields, it
is important to understand (linear) actions of Lie groups and Lie algebras on
vector spaces, known as representations.
For example, we shall see that, in the Standard Model , three Dirac spinors
for each quark flavour are combined and form a vector in a representation space
of the gauge group SU(3) of quantum chromodynamics. Similarly, two left-
handed Weyl spinors, known as the left-handed electron and the left-handed
electron neutrino, are combined to form a vector in a representation space of
the gauge group SU(2) × U(1) of the electroweak interaction.
It turns out that every Lie group and Lie algebra has a special representation,
known as the adjoint representation. The adjoint representation can be used to
define the Killing form, a canonical symmetric bilinear form on every Lie
algebra. Both the adjoint representation and the Killing form are important tools
for the classification of Lie algebras. The adjoint representation is also important
in physics, because gauge bosons correspond to fields on spacetime that
transform under the adjoint representation of the gauge group.
The purpose of this chapter is to describe representations of Lie groups and
Lie algebras in general as well as the structure of semisimple and compact Lie
algebras. We also discuss special scalar products on Lie algebras which will be
used in Sect. 7.3.1 to construct Lagrangians for gauge boson fields. We only
cover the basics of the representation and structure theory of Lie groups and Lie
algebras. Much more details can be found in the references mentioned at the
beginning of Chap. 1, which are also the references for this chapter.

2.1 Representations
2.1.1 Basic Definitions
We begin with the basic concept of representations of Lie groups and Lie
algebras.

Definition 2.1.1 Let G be a Lie group and V a vector space over the real or
complex numbers. Then a representation of G on V is a Lie group
homomorphism

to the Lie group GL(V ) of linear isomorphisms of V. One sometimes writes

GL(V ) = Aut(V ), the Lie group of linear automorphisms of V. The Lie group
GL(V ) is by definition isomorphic to a general linear group of the form
, where and n is the dimension of V.
If the representation is clear from the context, we sometimes write

for g G, v V. A representation ρ of a Lie group G is called faithful if ρ

is injective.

For a Lie group representation ρ the identities

and

hold for all g, h G. Note that the definition of a representation ρ requires that
the map ρ is a homomorphism in the algebraic sense and differentiable (in fact,
by Theorem 1.8.14 it suffices to demand that the map ρ is continuous).

Example 2.1.2 By Theorem 1.2.7 any compact Lie group has a faithful
representation on some finite-dimensional, complex vector space.

Definition 2.1.3 Let ρ V , ρ W be representations of a Lie group G on vector

spaces V and W. Then a morphism of the representations is a G-equivariant
linear map f: V → W, so that
i.e.

Such a map f is also called an intertwining map . An isomorphism or

equivalence of representations is a G-equivariant isomorphism.

Definition 2.1.4 Let ρ: G → GL(V ) be a representation of a Lie group G.

Suppose that H G is an embedded Lie subgroup. Then the restriction

of the Lie group homomorphism ρ to H is a representation of H, called a

restricted representation .

We define representations of Lie algebras in a similar way.

Definition 2.1.5 Let be a (real or complex) Lie algebra and V a vector space
over the real or complex numbers. Then a representation of on V is a Lie
algebra homomorphism

to the linear endomorphisms of V (linear maps V → V ). If the representation is

clear from the context, we sometimes write

for . A representation ϕ of a Lie algebra is called faithful if ϕ is

injective.

For a Lie algebra representation the following identity holds:

Example 2.1.6 By Ado’s Theorem 1.5.25 any Lie algebra has a faithful
representation on some finite-dimensional vector space.

Remark 2.1.7 Note that if the Lie algebra is complex, then we require the
representation to be a complex linear map.

Definition 2.1.8 Let ϕ V , ϕ W be representations of a Lie algebra on vector

spaces V and W. Then a morphism of the representations is a -equivariant
linear map f: V → W, so that
i.e.

Such a map f is also called an intertwining map . An isomorphism or

equivalence of representations is a -equivariant isomorphism.

Definition 2.1.9 Let be a representation of a Lie algebra .

Suppose that is a Lie subalgebra. Then the restriction

of the Lie algebra homomorphism ϕ to is a representation of , called

restricted representation .

Remark 2.1.10 Unless stated otherwise we only consider representations of Lie

groups and Lie algebras on real and complex vector spaces and these vector
spaces are finite-dimensional.

Remark 2.1.11 Both types of homomorphisms are called representations,

because we represent elements in the Lie group or Lie algebra by linear maps on
a vector space, i.e. (after a choice of basis for the vector space) by matrices.

Representations of Lie groups and their associated Lie algebras are related:

Proposition 2.1.12 (Induced Representations) Let ρ: G → GL(V ) be a

representation of a Lie group G on a vector space V. Then the differential
is a representation of the Lie algebra .

Proof The proof follows from Theorem 1.5.18, because the differential of a Lie
group homomorphism is a Lie algebra homomorphism. □

With Theorem 1.7.16 we get the following commutative diagram:

Note that the exponential map on the right is just the standard exponential
map on endomorphisms (defined in the same way as for matrices, using
composition instead of matrix multiplication). We can thus write the
commutativity of the diagram as
This means: if we know how a Lie algebra element acts in a
representation on the vector space V, then we know how the group element expX
G acts on V.
Assuming Theorem 1.5.20 we get the following:

Corollary 2.1.13 (Integrability Theorem for Representations) Let G be a

connected and simply connected Lie group. Suppose is a
representation of the Lie algebra of G. Then there exists a unique representation
ρ: G → GL(V ) such that ρ = ϕ.

The discussion in Example 1.5.21 shows that this may not hold if G is not
simply connected. In particular, if

is a representation and X the generator of with exp(2πiX) = 1, then a

necessary condition that ϕ comes from a representation

is that

Example 2.1.14 For any constant there is a complex 1-dimensional

representation

We say that these representations have winding number k. In the Standard

Model these representations appear in connection with the weak hypercharge
gauge group U(1) Y .

Example 2.1.15 The Lie groups (and ) have canonical

representations on (and ) by matrix multiplication on column vectors from
the left. These representations induce representations for all linear groups, called
standard , defining or fundamental representations (by a fundamental
representation we will always mean the defining representation). There are
similar, induced representations of the corresponding Lie algebras.
Definition 2.1.16 A representation of a Lie group G (or Lie algebra ) on a
vector space V is called irreducible if there is no proper invariant subspace W
V, i.e. no vector subspace W, different from 0 or V, such that G ⋅ W W (or
). A representation is called reducible if it is not irreducible.

Example 2.1.17 The 0-dimensional and every 1-dimensional representation are

irreducible, because in these cases there are no proper vector subspaces at all.

Definition 2.1.18 A singlet representation is a representation of a Lie group

or Lie algebra on a 1-dimensional (real or complex) vector space. Similarly, a
doublet or triplet representation is a representation on a 2- or 3-dimensional
vector space. A representation of a Lie group or Lie algebra on an n-dimensional
vector space is sometimes denoted by n, in particular, if the dimension uniquely
determines the representation.

Example 2.1.19 (Trivial Representations) Let G be a Lie group and V a real or

complex vector space. Then

where every group element gets mapped to the identity, is a representation,

called a trivial representation . It is irreducible precisely if V is 1-dimensional.
Similarly, if is a Lie algebra, then

is a trivial representation. Again, it is irreducible precisely if V is 1-dimensional.

We will later study a class of Lie algebras where every representation is either
trivial or faithful, see Exercise 2.7.9.
It is a curious fact that the fundamental and trivial representations of
SU(3) and SU(2), together with the winding number representations of U(1)
in Example 2.1.14, suffice to describe all matter particles (and the Higgs field
) in the Standard Model ; see Sect. 8.5. The gauge bosons corresponding to
these gauge groups are described by the adjoint representation that we
discuss in Sect. 2.1.5.

Example 2.1.20 The fundamental representation of the Lie algebra is an

(irreducible) doublet representation on the vector space . Recall from
Example 1.5.32 that there exists an isomorphism . The fundamental
representation of thus also defines an (irreducible) triplet representation of
on (and ). It can be proved that has a unique (up to equivalence)
irreducible complex representation V n of dimension n + 1 for every natural
number n ≥ 0 (see, e.g. [24]).

Example 2.1.21 (The Heisenberg Lie Algebra and Quantum Mechanics)

Recall from Example 1.5.38 that the Heisenberg Lie algebra is a 3-
dimensional real Lie algebra spanned by vectors p, q, z with Lie brackets

Let be some real number. A central representation of is a

representation

on a complex vector space V such that z gets mapped to i ⋅ Id V . If we denote

the images of q and p in End(V ) by and , then

and the other two commutation relations are satisfied trivially (on the right-hand
side we do not write the identity map of V explicitly). This is the canonical
commutation relation of quantum mechanics.

2.1.2 Linear Algebra Constructions of Representations

There are several well-known constructions that yield new vector spaces from
given ones. If the given vector spaces carry a representation, then usually the
new vector spaces carry induced representations. We first recall the following
notion from complex linear algebra.

Definition 2.1.22 Let V be a complex vector space. Then we define the

complex conjugate vector space as follows:

1. As a set and abelian group .

2. Scalar multiplication is defined by

If f: V → V is a complex linear map, then the same map (on the set ) is
denoted by and is still complex linear. The identity map is
complex antilinear.

Definition 2.1.23 Let V and W be real or complex vector spaces with

representations

of a Lie group G. Then there exist the following representations of G, where g

G and v V, w W are arbitrary:

1. The direct sum representation ρ V W on V W, defined by

2. The tensor product representation ρ V W on V W, defined by

3. The dual representation on V , defined by

4. The exterior power representation on Λ k V, defined by

5. The homomorphism space representation ρ Hom(V, W) on Hom(V, W),

defined by
6. If V is a complex vector space, then the complex conjugate representation
on is defined by

Suppose in addition that

is a representation of a Lie group H. Then there exists the following

representation, where h H is arbitrary:

7. The (outer) tensor product representation ρ V τ W on V W of the Lie

group G × H, defined by

for g G, h H.

It is easy to check that each of these maps is indeed a representation.

Remark 2.1.24 The direct sum representation ρ V τ W on V W of the Lie

group G × H, defined by

is less important, because it is can be reduced to the representations ρ V and τ W ,

each tensored with the trivial 1-dimensional representation.

Remark 2.1.25 If V is a complex representation space for a Lie group, we then

get in total four complex representations which have the same dimension as V:
V, V , and .

The representations of the Lie group

that appear in the Standard Model of elementary particles are direct sums of
outer tensor product representations of the form

where U, V, W are certain representations of the factors SU(3), SU(2), U(1) of

G. See Sect. 8.5 for details.
Example 2.1.26 We describe these constructions using matrices. Consider the
column vector spaces , where . Representations ρ V and ρ
W of a Lie group G take values in the matrix Lie groups and .
We can then identify the canonical representations of G on the vector spaces

with the following representations:

1. V W can be identified with . For a column vector the

direct sum representation is given by

2. V can be identified with a row vector space that we here denote by .

For a row vector the dual representation is given by

3. Hom(V, W) can be identified with the vector space . For a

matrix the representation on the homomorphism space is
given by

4. Λ 2 V is the space of skew-symmetric, bilinear maps

and can be identified with , the space of skew-symmetric n × n-

matrices, by sending λ to the matrix A with coefficients A ij = λ(e i , e j ). The
representation on Λ 2 V is then given by

5. If , then as an abelian group and every complex scalar (and

hence every complex matrix) acts as the complex conjugate. For a column
vector the complex conjugate representation is given by
There are analogous constructions for representations of Lie algebras:

Definition 2.1.27 Let V and W be real or complex vector spaces with

representations

of a Lie algebra . Then there exist the following representations of , where

and v V, w W are arbitrary:

1. The direct sum representation ϕ V W on V W, defined by

2. The tensor product representation ϕ V W on V W, defined by

3. The dual representation on V , defined by

4. The exterior power representation on Λ k V, defined by

5. The homomorphism space representation ϕ Hom(V, W) on Hom(V, W),

defined by

6. If V is a complex vector space and a real Lie algebra, then the complex
conjugate representation on is defined by
 Suppose in addition that

is a representation of a Lie algebra . Then there exists the following

representation, where is arbitrary:

7. The (outer) tensor product representation ϕ V ψ W on V W of the Lie

algebra , defined by

for .

Remark 2.1.28 Perhaps the most interesting case in the proof that these maps
define representations is the dual representation for both Lie groups and Lie
algebras. To check that the formulas here define representations is the purpose of
Exercise 2.7.1.

Both constructions are related:

Proposition 2.1.29 Let G and H be Lie groups with Lie algebras and . Let
ρ be any of the representations of G on V W, V W, V , Λ k V, Hom(V, W)
or (or of G × H on V W) from Definition 2.1.23 . Then the induced
representation ρ of (or of ) is the corresponding one from Definition
2.1.27 .

Proof The proof follows by differentiating the representation of G (or of G ×

H). □

2.1.3 The Weyl Spinor Representations of

We discuss an extended example that is relevant for some theories in physics,
like the Standard Model or supersymmetry (see reference [146, Appendix A]).
Let . As we will discuss in Sect. 6.8.2 in more detail, the group
is the (orthochronous) Lorentz spin group , i.e. the universal covering of
the identity component of the Lorentz group of 4-dimensional spacetime.
 We denote by the fundamental -representation. Then we get
the following four complex doublet representations, where and ψ
ℂ 2:

1. The fundamental representation V:

2. The dual representation V :

3. The complex conjugate representation :

4. The dual of the complex conjugate representation :

Here we denote the elements of the vector spaces V , and for clarity
by ψ T , and .

Remark 2.1.30 In physics the components of the vectors in the spaces V, V ,

and are denoted by ψ α , ψ α , and . We could denote these
representations by 2, 2 , and .

Definition 2.1.31 In this situation the representation of on V is called

the left-handed Weyl spinor representation and the representation on is
called the right-handed Weyl spinor representation. Both representations are
also called chiral spinor representations.

We want to show that the remaining two representations are isomorphic to the
left- and right-handed Weyl spinor representations.

Definition 2.1.32 We define

Proposition 2.1.33 We have the following equivalent description of :

Proof The proof is an easy calculation; see Exercise 2.7.2. □

Proposition 2.1.34 The map

is an isomorphism of representations. Similarly the map

is an isomorphism of representations.

Proof We only have to show -equivariance of the maps. This follows

by applying Proposition 2.1.33:

and

See Sect. 6.8 and Lemma 8.5.5 for more details about these isomorphisms.

2.1.4 Orthogonal and Unitary Representations

It is often useful to consider representations compatible with a scalar product on
the vector space. Recall that a scalar product on a real vector space is called
Euclidean if it is bilinear , symmetric and positive definite . A scalar product on
a complex vector space is called Hermitian if it is sesquilinear (complex linear
in the second argument and complex antilinear in the first argument), conjugate
symmetric (exchanging the first and second argument changes the scalar product
by complex conjugation) and positive definite .

Definition 2.1.35 A representation ρ: G → GL(V ) of a Lie group G on a

Euclidean (or Hermitian) vector space (V, ⋅ , ⋅ ) is called orthogonal (or
unitary) if the scalar product is G-invariant, i.e.

for all g G, v, w V. Equivalently, the map ρ has image in the orthogonal

subgroup O(V ) (or the unitary subgroup U(V )) of the general linear group GL(V
), determined by the scalar product ⋅ , ⋅ .

In an orthogonal representation the group literally acts through rotations (and

possibly reflections) on a Euclidean vector space. There is a similar notion for
representations of Lie algebras.

Definition 2.1.36 A representation of a real Lie algebra on a

Euclidean (or Hermitian) vector space (V, ⋅ , ⋅ ) is called skew-symmetric (or
skew-Hermitian) if it satisfies

for all , v, w V. Equivalently, the map ϕ has image in the orthogonal Lie
subalgebra (or the unitary Lie subalgebra ) of the general linear algebra
, determined by the scalar product ⋅ , ⋅ .

We can similarly define invariance of a form on a vector space under

representations of a Lie group or Lie algebra in the case where the form is not
non-degenerate or not positive definite.
Invariant scalar products for Lie group and Lie algebra representations are
related:

Proposition 2.1.37 (Scalar Products and Induced Representations) Let

ρ: G → GL(V ) be a representation of a Lie group G and ⋅ , ⋅ a G-invariant
Euclidean (or Hermitian) scalar product on V, i.e. the representation ρ is
orthogonal (or unitary). Then the induced representation of the
Lie algebra is skew-symmetric (or skew-Hermitian).
Proof We have by Theorem 1.7.16

and hence by Corollary 1.7.30

Differentiating both sides by t in t = 0 and using the product rule we get:

This implies the claim. □

Let be a unitary representation of a real Lie algebra on a

complex vector space V. Then ϕ(X) is a skew-Hermitian endomorphism for all
, hence iϕ(X) is Hermitian. This implies that the endomorphism iϕ(X) can
be diagonalized with real eigenvalues (and ϕ(X) can be diagonalized with
imaginary eigenvalues).

Definition 2.1.38 The eigenvalues of − iϕ(X) are called charges of in the

unitary representation ϕ.

The minus sign in − iϕ(X) is convention: we can write ϕ(X) as iA X , where A X is

a Hermitian operator, and the charges are the eigenvalues of A X .
If is an abelian subalgebra, then the operators iϕ(X) for all
commute and can be diagonalized simultaneously. This idea is related to the
notion of weights of a representation and used extensively in the classification of
representations of Lie algebras and Lie groups (in a certain sense, that can be
made precise, irreducible representations are thus determined by their charges).

Existence of Invariant Scalar Products

It is an important fact that representations of compact Lie groups always admit
an invariant scalar product.

Theorem 2.1.39 (Existence of Invariant Scalar Products for Representations

of Compact Lie Groups) Let G be a compact Lie group and ρ: G → GL(V ) a
representation on a real (or complex) vector space. Then we can find a G-
invariant Euclidean (or Hermitian) scalar product on V, hence the given
representation ρ becomes orthogonal (or unitary ) for this scalar product.
The proof uses the existence of an integral over differential forms σ of top
degree n on oriented n-manifolds M:

If ϕ: M → N is an orientation preserving diffeomorphism between oriented n-

manifolds, then we have the transformation formula

We now prove Theorem 2.1.39.

Proof Suppose G has dimension n and let X 1, …, X n be a basis of T e G. We set

for the corresponding right-invariant vector fields on G, defined by

This basis has a dual basis of right-invariant 1-forms ω 1, …, ω n . Then the

wedge product

is a nowhere vanishing, right-invariant differential form on G of top degree. We

can assume that the orientation of G coincides with the orientation defined by σ,
so that

which is finite, because G is compact. Let ⋅ , ⋅ denote an arbitrary Euclidean

(or Hermitian) scalar product on V. We construct a new scalar product by
averaging this scalar product over the action of the group G:

where τ v, w is the smooth function

(here the representation ρ is implicit and we use that G is compact, so that this
integral is finite).
We claim that ⋅ , ⋅ is a G-invariant Euclidean (or Hermitian) scalar product
on V: It is clear that ⋅ , ⋅ is bilinear and symmetric (or sesquilinear and
conjugate symmetric in the complex case). For v ≠ 0 the function τ v, v is strictly
positive on G. As a consequence the integral is
with equality only if v = 0. Therefore ⋅ , ⋅ is a positive definite Euclidean
(or Hermitian) scalar product on G.
We finally show G-invariance of the new scalar product: Let g G be fixed.
Then

This follows from a short calculation:

where we used that ρ (which is implicit) is a representation. This implies

because σ is right-invariant. Since is an orientation preserving

diffeomorphism from G to G we get:

for all g G and v, w V. □

Decomposition of Representations
The existence of an invariant scalar product for every representation of a
compact Lie group has an important consequence.

Theorem 2.1.40 (Decomposition of Representations) Let ρ: G → GL(V )

be a representation of a Lie group G on a finite-dimensional real (or
complex) vector space V. Suppose that there exists a G-invariant Euclidean
(or Hermitian) scalar product on V (this is always the case, by Theorem
2.1.39 , if G is compact). Then V decomposes as a direct sum

of irreducible G-representations (V i , ρ i ).

Proof The proof follows, because if W V is a subspace with ρ(G)W W,

then the orthogonal complement W with respect to a G-invariant scalar
product also satisfies ρ(G)W W . We have
We can thus continue splitting V until we arrive at irreducible representations
(after finitely many steps, since V is finite-dimensional). □

Remark 2.1.41 One of the aims of representation theory for Lie groups G is to
understand irreducible representations and to decompose any given
representation (at least for compact G) into irreducible ones according to
Theorem 2.1.40.
For instance, for G = SU(2), we can consider the tensor product
representation V n V m , where V n , V m are the irreducible complex
representations of dimension n + 1 and m + 1 mentioned in Example 2.1.20. The
tensor product V n V m is reducible under SU(2) and its decomposition into
irreducible summands V k is determined by the Clebsch–Gordan formula. This
formula appears in quantum mechanics in the theory of the angular momentum
of composite systems.

Remark 2.1.42 One of the basic topics in Grand Unified Theories is to study
the restriction of representations of a compact Lie group G to embedded Lie
subgroups H G. If the representation ρ of G is irreducible, it may happen
that the representation ρ | H of H is reducible and decomposes as a direct sum.
The actual form of the decomposition of a representation ρ under restriction
to a subgroup H G is called the branching rule .
For instance, there exist certain 5- and 10-dimensional irreducible
representations of the Grand Unification group G = SU(5) that decompose
under restriction to the subgroup H = SU(3) × SU(2) × U(1) (more precisely,
to a certain quotient of this group; see Sect. 8.5.7) into the fermion
representations of the Standard Model. Details of this calculation can be
found in Sect. 9.5.4.

Remark 2.1.43 Suppose a Lie group G has a unitary representation on a

complex vector space V and e 1, …, e n is some orthonormal basis for V. If we
decompose V into invariant, irreducible subspaces according to Theorem 2.1.40,
then we can choose an associated orthonormal basis f 1, …, f n , adapted to the
decomposition of V (spanning the G-invariant subspaces) and related to the
original basis by a unitary matrix. In general, the basis { f i } will be different
from {e i }.
 In the Standard Model where G = SU(3) × SU(2) × U(1) this is related to the
concept of quark mixing . The complex vector space V of fermions , which
carries a representation of G, has dimension 45 (plus the same number of
corresponding antiparticles ) and is the direct sum of two G-invariant subspaces
(sectors): a lepton sector of dimension 9 (where we do not include the
hypothetical right-handed neutrinos) and a quark sector of dimension 36.
Counting in this way, the Standard Model thus contains at the most elementary
level 90 fermions (particles and antiparticles).
The quark sector has a natural basis of so-called mass eigenstates , given by
the quarks of six different flavours u, d, c, s, t, b, each one appearing in three
different colours and two chiralities (6 basis vectors for each flavour), yielding in
total 36 quarks. However, the basis given by these flavours does not define a
splitting into subspaces invariant under SU(2). The SU(2)-invariant subspaces
are spanned by a basis of so-called weak eigenstates that can be obtained from
the mass eigenstates by a certain unitary transformation. The matrix of this
unitary transformation is known as the Cabibbo–Kobayashi–Maskawa (CKM)
matrix , which has to be determined by experiments. The CKM matrix and quark
mixing will be explained in more detail in Sect. 8.8.2.

Unitary Representations of Non-Compact Lie Groups

It is an important fact that certain non-compact Lie groups do not admit non-
trivial finite-dimensional unitary representations according to the following
theorem (a proof can be found in [12, Chap. 8.1B]):

Theorem 2.1.44 A connected, simple, non-compact Lie group does not admit
finite-dimensional unitary complex representations except for the trivial
representation.

See Definition 2.4.27 for the notion of simple Lie groups. For example, the Lie
group is simple and non-compact, hence every non-trivial unitary
representation of G is infinite-dimensional. This has important consequences for
quantum field theory, see Sect. B.2.4. Of course, admits non-trivial
finite-dimensional non-unitary representations, like the fundamental
representation on .

2.1.5 The Adjoint Representation

We want to define a particularly important representation of a Lie group and its
Lie algebra. The vector space carrying the representation has the same
dimension as the Lie group or Lie algebra (we follow [142] in this subsection).
Recall that for an element g of a Lie group G we defined the inner
automorphism (conjugation)

The differential is an automorphism of the Lie algebra , in

particular a linear isomorphism.

Theorem 2.1.45 (Adjoint Representation of a Lie Group) The map

is a Lie group homomorphism, i.e. a representation of the Lie group G on the

vector space , called the adjoint representation or adjoint action of the Lie
group G . We sometimes write Ad G instead of Ad.

Proof Note that

Hence

This shows that Ad is a homomorphism in the algebraic sense. We have to show

that Ad is a smooth map. It suffices to show that for every the map

is smooth, because if we choose a basis for the vector space , it follows that Ad
is a smooth matrix representation. The map Ad(⋅ )v is equal to the composition
of smooth maps

given by

where we set

This implies the claim. □

The following identity (whose proof is left as an exercise) is sometimes useful.

Proposition 2.1.46 Let G be a Lie group with Lie algebra and ρ a

representation of G on a vector space V with induced representation ρ of .
Then

Example 2.1.47 The adjoint representation is very simple in the case of abelian
Lie groups G: if G is abelian, then c g = Id G for all g G and thus for
all g G, hence the adjoint representation is a trivial representation.

We consider a more general example: Let with be a

closed subgroup of a general linear group with Lie algebra . Fix Q G.

Proposition 2.1.48 (Adjoint Representation of Linear Groups) The adjoint

action

is given by

where ⋅ denotes matrix multiplication and we identify elements Q G and

with matrices in the canonical way.

Proof Define a curve γ(t) = e tX and take the derivative

In this situation, the Lie algebra on which the adjoint representation acts is
naturally a vector space of matrices.

Example 2.1.49 We consider the adjoint representation of the Lie group SU(3).
The Lie algebra consists of the skew-Hermitian, tracefree matrices. As a
real vector space, has dimension 8 and is spanned by iλ a , with a = 1, …, 8,
where λ a are the Gell-Mann matrices from Example 1.5.33. We can define an
explicit isomorphism
On such a matrix X the group element Q SU(3) acts as

Using the isomorphism we could write this as an explicit

representation on .

The following observation is sometimes useful.

Lemma 2.1.50 (Adjoint Representation of Direct Product) Let G = H × K

be a direct product of Lie groups. Then the adjoint representation of G on
is the direct sum of the adjoint representations of H on and K on :

Proof Let γ be a curve in H through e, tangent to . Then for (h, k) H×

K

Similarly for a vector in . □

Example 2.1.51 We consider the adjoint representation of the Standard Model

Lie group

We can write a group element Q H as a block matrix

with Q K K for K = SU(3), SU(2), U(1). We can similarly write the elements
of the Lie algebra of H as a block matrix: with the notation from Examples 1.5.
29, 1.5.32 and 1.5.33, the Lie algebra is spanned by iλ a , where λ a are the
Gell-Mann matrices , the Lie algebra is spanned by iσ a , where σ a are the
Pauli matrices , and the Lie algebra is spanned by i. We can then define an
isomorphism

According to Lemma 2.1.50 the adjoint action is given by multiplication of

block matrices:

The representation Ad H describes the representation of the gauge boson

fields in the Standard Model . The coefficients G a , W a and B (possibly with
a different normalization) are known as the gluon fields , weak gauge fields
and hypercharge gauge field ; see Sect. 8.5.5 for more details.

Like any other representation of a Lie group, the adjoint representation of G

induces a representation of the associated Lie algebra.

Theorem 2.1.52 (Adjoint Representation of a Lie Algebra) The map

given by

is a Lie algebra homomorphism, i.e. a representation of the Lie algebra on the

vector space , called the adjoint representation of the Lie algebra . We
sometimes write instead of ad. We have the following commutative diagram
according to Theorem 1.7.16 :

The map ad satisfies the formula

Proof We only have to prove the formula ad X Y = [X, Y ]. For left-invariant
vector fields X, Y on G, where X has flow ϕ t , we have according to the
commutative diagram

Here we used Proposition 1.7.12 and Theorem A.1.46. □

We can write the formula given by the commutative diagram as

A direct consequence of Example 2.1.47 is the following:

Corollary 2.1.53 If G is an abelian Lie group, then the adjoint representation

ad is trivial, hence the Lie algebra is abelian.

It can be shown that the converse also holds (for connected Lie groups), cf.
Exercise 2.7.7.

Remark 2.1.54 We can define for any Lie algebra , even if it does not belong
a priori to a Lie group, the map

by exactly the same formula

Then this map is a representation of (by the Jacobi identity), again called the
adjoint representation.

Remark 2.1.55 One should be careful not to confuse the fundamental and the
adjoint representation for a linear group. In general, the dimensions are already
different. For example, in the case of SU(n) the dimension of the fundamental
representation is n, while the adjoint representation has dimension n 2 − 1. For a
linear group the fundamental representation acts canonically on a vector space of
column vectors, while the adjoint representation acts on a vector space of
matrices.
Example 2.1.56 The homomorphism ϕ: S 3 → SO(3) from Example 1.3.8 is the
adjoint representation of S 3 = SU(2).

2.2 Invariant Metrics on Lie Groups

Since a Lie group G is a manifold, we can study metrics (Riemannian or pseudo-
Riemannian) on it. We are interested in particular in the following types of
metrics.

Definition 2.2.1 Let s be a metric on a Lie group G.

1. The metric s is called

left-invariant if L g s = s for all g G
right-invariant if R g s = s for all g G.
Equivalently, either all left translations or all right translations are
isometries.

2. The metric s is called bi-invariant if it is both left- and right-invariant.

It is clear that every metric induces a scalar product on . On the other

hand, given an arbitrary scalar product ⋅ , ⋅ on , it is easy to construct
a left-invariant metric on G by

a right-invariant metric on G by

for all g G and X, Y T g G.

However, in general we only get a bi-invariant metric in this way if G is
abelian (if G is not abelian, then L g ≠ R g for some g G). Bi-invariant metrics
have the following characterization:

Theorem 2.2.2 (Bi-Invariant Metrics and Ad-Invariance) Let s be a left-

invariant metric on a Lie group G. Then s is bi-invariant if and only if the scalar
product ⋅ , ⋅ on defined by the metric s is Ad-invariant , i.e.
for all g G and .

Proof Let X and Y be vectors in T p G. Then we can calculate:

and

where in both equations we used that s is left-invariant. This implies the claim,
because is an isomorphism of vector spaces. □

Theorem 2.2.3 (Ad-Invariant Scalar Products for Compact Lie Groups)

Let G be a compact Lie group . Then there exists a Euclidean (positive definite)
scalar product ⋅ , ⋅ on the Lie algebra which is Ad-invariant . The adjoint
representation is orthogonal with respect to this scalar product.

Proof This follows from Theorem 2.1.39, because Ad is a representation of the

compact Lie group G on the vector space . □

The existence of positive definite Ad-invariant scalar products on the Lie

algebra of compact Lie groups is very important in gauge theory, in particular,
for the construction of the gauge-invariant Yang–Mills Lagrangian; see Sect.
7.3.1. We will study such scalar products in more detail in Sect. 2.5 after we
have discussed the general structure of compact Lie groups. The fact that
these scalar products are positive definite is important from a
phenomenological point of view, because only then do the kinetic terms in the
Yang–Mills Lagrangian have the right sign (the gauge bosons have positive
kinetic energy [148]).

Here is a corollary to Theorem 2.2.2 and Theorem 2.2.3:

Corollary 2.2.4 Every compact Lie group admits a bi-invariant Riemannian

metric .

Remark 2.2.5 It can be shown that the geodesics of a bi-invariant metric on a

Lie group G through the neutral element e are of the form γ(t) = exp(tX), with
 . The notions of exponential map for geodesics and Lie groups thus
coincide for bi-invariant Riemannian metrics.

2.3 The Killing Form

We want to consider a special Ad-invariant inner product on every Lie algebra ,
which in general is neither non-degenerate nor positive or negative definite. This
is the celebrated Killing form.

Theorem 2.3.1 Let be a Lie algebra over . The Killing form on

is defined by

This is a -bilinear, symmetric form on .

Remark 2.3.2 Note that the Killing form for complex Lie algebras is also
symmetric and complex bilinear and not Hermitian.

Proof For we have

In particular, is indeed bilinear. To show that the Killing form is symmetric,

recall the definition of the trace tr( f) of a linear endomorphism f of a vector
space V: If v 1, …, v n is a basis of V and we define the representing matrix of f by

then

This number does not depend on the choice of basis for V: If ϕ: V → V is an

arbitrary isomorphism, then

We also have
for all endomorphisms f, g: V → V. This shows, in particular, that is
symmetric. □

Theorem 2.3.3 (Invariance of Killing Form Under Automorphisms) Let

be a Lie algebra automorphism of . Then the Killing form satisfies

If is the Lie algebra of a Lie group G, this holds in particular for the
automorphism σ = Ad g with g G arbitrary.

Proof Note that

Since σ is a Lie algebra automorphism we have

Thus

We get for the Killing form:

Corollary 2.3.4 The Killing form defines a bi-invariant symmetric form on

any Lie group G.

Remark 2.3.5 We will determine in Sect. 2.4 when the Killing form is non-
degenerate or definite (in the case of a real Lie algebra).

Proposition 2.3.6 (ad Is Skew-Symmetric with Respect to the Killing

Form) Let be a Lie algebra with Killing form . Then

Proof This follows from Theorem 2.3.3 and Proposition 2.1.37 if is the Lie
algebra of a Lie group G. In the general case we use the formula

which follows from the Jacobi identity. The definition of the Killing form
implies

because the trace is invariant under cyclic permutations. □

2.4 Semisimple and Compact Lie Algebras

In this section we discuss some results concerning the general structure of Lie
algebras and Lie groups (we follow [83] and [153]). There are two elements that
play a key role in the theory of Lie algebras:
The adjoint representation of the Lie algebra , together with its
invariant subspaces, known as ideals.
The Killing form of .
Both notions are related: the definition of the Killing form involves the
adjoint representation and the adjoint representation is skew-symmetric with
respect to the Killing form.
The idea is to proceed in a similar way to Theorem 2.1.40 and try to
decompose with the adjoint representation into irreducible, pairwise -
orthogonal pieces. This works out particularly well for a type of Lie algebra
known as a semisimple Lie algebra. The next step is to classify the pieces
where the adjoint representation is irreducible. These are called simple Lie
algebras. We will discuss the classification for the simple Lie algebras coming
from compact Lie groups, which appear in physics as gauge groups.

2.4.1 Simple and Semisimple Lie Algebras in General

Definition 2.4.1 Let be a Lie algebra. For subsets we define
as the set of all finite sums of elements of the form [X, Y ] with .

Definition 2.4.2 Let be a Lie algebra.

1. An ideal in is a vector subspace such that . Equivalently,

2. The center of is defined as

3. The commutator of is defined as .

The following is easy to check.

Lemma 2.4.3 For any Lie algebra the commutator is an ideal and the center
is an abelian ideal.

Proposition 2.4.4 The kernel of the adjoint representation of a Lie algebra

is the center . The adjoint representation is faithful if and only if .

Proof We have ad X ≡ 0 if and only if .□

This implies Ado’s Theorem 1.5.25 for Lie algebras with trivial center.

Definition 2.4.5 Let be a Lie algebra.

1. The Lie algebra is called simple if is non-abelian and has no non-

trivial ideals (different from 0 and ).

2. The Lie algebra is called semisimple if has no non-zero abelian ideals.

Simple Lie algebras are sometimes defined equivalently as follows:

Lemma 2.4.6 A Lie algebra is simple if and only if has dimension at least
two and has no non-trivial ideals.

Proof If is non-abelian, then it has dimension at least two. On the other hand,
if is abelian and has dimension at least two, then has non-trivial (abelian)
ideals. □
It is clear that every simple Lie algebra is semisimple.

Lemma 2.4.7 If is simple, then .

Proof The commutator is an ideal, hence equal to or 0. The second

possibility is excluded, because is not abelian. □

We can characterize simple Lie algebras as follows:

Proposition 2.4.8 (Criterion for Simplicity) A Lie algebra is simple if and

only if is non-abelian and the adjoint representation of is irreducible.

Proof The claim follows from the definition of an ideal. □

We can also characterize semisimple Lie algebras (we only prove one direction
following [83]; the proof of the converse, which would take us too far afield, can
be found in [77, 83]):

Theorem 2.4.9 (Cartan’s Criterion for Semisimplicity) A Lie algebra is

semisimple if and only if the Killing form is non-degenerate.

Proof We only prove that the Killing form is degenerate if the Lie algebra is
not semisimple. Let be a non-zero abelian ideal in . We choose a
complementary vector space with

Let and be arbitrary elements. Then

Under the splitting , the endomorphisms ad X and ad Y thus have the form

It follows that
and

Remark 2.4.10 In general, the Killing form of a semisimple Lie algebra is

indefinite, i.e. pseudo-Euclidean.

Assuming Cartan’s Criterion we can prove the following.

Theorem 2.4.11 (Structure of Semisimple Lie Algebras) If a Lie algebra

is semisimple, then is the direct sum

of ideals , each of which is a simple Lie algebra, and which are pairwise
orthogonal with respect to the Killing form.

Proof We ultimately would like to apply Theorem 2.1.40 and decompose the
adjoint representation on into irreducible summands, orthogonal with respect
to the Killing form. There is one problem which requires some work: the Killing
form is non-degenerate, but not (positive or negative) definite. Therefore
it is not immediately clear that orthogonal complements of invariant subspaces
lead to a direct sum decomposition.
Let be an ideal in and

the orthogonal complement with respect to the Killing form B. Then is

also an ideal in , because

by Proposition 2.3.6. Furthermore, is an abelian ideal in : it is

clear that the intersection of two ideals is an ideal and

This implies that is abelian, because B is non-degenerate. Since is

semisimple, it follows that .
This implies
and the restriction of the Killing form to and (which is just the Killing
form on these Lie algebras) is non-degenerate. We can continue splitting the
(finite-dimensional) Lie algebra in this fashion until we arrive at irreducible,
non-abelian (simple) ideals. □

Remark 2.4.12 In addition to semisimple and abelian Lie algebras there are
other classes of Lie algebras, like solvable and nilpotent Lie algebras, which we
have not discussed in detail.

2.4.2 Compact Lie Algebras

We are particularly interested in compact Lie algebras, including compact simple
and compact semisimple Lie algebras.

Definition 2.4.13 A real Lie algebra is called compact if it is the Lie algebra
of some compact Lie group.

Remark 2.4.14 Even if is compact, there could exist non-compact Lie groups
whose Lie algebra is also . For example, the abelian Lie algebra is the Lie
algebra of the compact Lie group U(1) = S 1 and of the non-compact Lie group
.

Example 2.4.15 Note that the abelian Lie algebra , for n ≥

1, is compact, but neither simple nor semisimple.

Theorem 2.4.16 (Killing Form of Compact Lie Algebras) Suppose is a

compact real Lie algebra. Then the Killing form is negative semidefinite:
We have

Proof We follow the proof in [14]. Since is the Lie algebra of a compact Lie
group G, according to Theorem 2.2.3 there exists a positive definite scalar
product ⋅ , ⋅ on which is Ad G -invariant. Let e 1, …, e n be an orthonormal
basis for with respect to this scalar product. We get
for the associated norm | | ⋅ | |. This implies

Equality holds if and only if ad X ≡ 0 on , i.e. .□

Remark 2.4.17 Note as an aside that the notion of a bilinear, symmetric form
being (semi-)definite is only meaningful on real and not on complex vector
spaces.

Corollary 2.4.18 Let be a compact Lie algebra with trivial center, .

Then the Killing form is negative definite.

Proof This follows from Theorem 2.4.16. □

Remark 2.4.19 The following converse to Corollary 2.4.18 can be proved (see
[77]): if the Killing form of a real Lie algebra is negative definite, then it is
compact with trivial center. In particular, every Lie subalgebra of a compact Lie
algebra is compact.

Corollary 2.4.20 Let be a compact Lie algebra. Then the Killing form is
negative definite if and only if is semisimple.

Proof One direction follows from Corollary 2.4.18, because semisimple Lie
algebras have trivial center. The other direction follows from Theorem 2.4.9. □

Theorem 2.4.21 (Decomposition of Compact Lie Algebras) Let be a

compact Lie algebra with center . Then there exists an ideal in such that

The ideal is a compact semisimple Lie algebra with negative definite Killing
form.

Proof Choose a positive definite scalar product ⋅ , ⋅ on which is Ad G -

invariant. Let be the orthogonal complement
with respect to this scalar product. Then is an ideal, because

It is clear that

By Theorem 2.4.16 the Killing form is negative definite on , which is thus

compact by Remark 2.4.19 and semisimple by Theorem 2.4.9. □

Corollary 2.4.22 (Structure of Compact Lie Algebras) Let be a

compact Lie algebra . Then is a direct sum of ideals

where the are compact simple Lie algebras.

Proof This follows from Theorem 2.4.11. The Lie algebras are compact by
Remark 2.4.19. □

Using considerable effort it is possible to classify simple Lie algebras, one of the
great achievements of 19th and 20th century mathematics. The result for
compact simple Lie algebras is the following (see [83] for a proof):

Theorem 2.4.23 (Killing–Cartan Classification of Compact Simple Lie

Algebras) Every compact simple Lie algebra is isomorphic to precisely
one of the following Lie algebras:

1. for n ≥ 1.

2. for n ≥ 2.

3. for n ≥ 3.

4. for n ≥ 4.

5. An exceptional Lie algebra of type G2, F4, E6, E7, E8.

 The families in the first four cases are also called A n , B n , C n , D n in this
order.

Remark 2.4.24 The lower index n in the series A n , B n , C n , D n as well as in

the exceptional cases G2, F4, E6, E7, E8 is the rank of the corresponding compact
Lie group, i.e. the dimension of a maximal torus subgroup embedded in the Lie
group.

Remark 2.4.25 The reason for the restrictions on n in the first four cases of the
classical Lie algebras is to avoid counting Lie algebras twice, because we have
the following isomorphisms (we only proved the first isomorphism in Sect. 1.5.
5):

There is also the abelian Lie algebra

and the semisimple Lie algebra

cf. Exercise 1.9.21.

The basic building blocks of all compact Lie algebras are thus
abelian Lie algebras
the four families of classical compact non-abelian Lie algebras
five exceptional compact Lie algebras.
In some sense, most compact Lie algebras are therefore classical or direct
sums of classical Lie algebras.
It is sometimes convenient to know that we can choose for a compact
semisimple Lie algebra a basis in such a way that the structure constants (see
Definition 1.4.17) have a nice form. Let be a compact semisimple Lie algebra.
According to Corollary 2.4.20 the Killing form is negative definite. Let T 1,
…, T n be an orthonormal basis of with respect to the Killing form:
Proposition 2.4.26 The structure constants f abc for a -orthonormal basis
{T a } of a semisimple Lie algebra are totally antisymmetric:

Proof This is Exercise 2.7.11. □

2.4.3 Compact Lie Groups

We briefly discuss the structure of compact Lie groups .

Definition 2.4.27 A connected Lie group G is called simple (or semisimple )

if its Lie algebra is simple (or semisimple).

Corollary 2.4.28 If G is simple, then Ad G is an irreducible representation.

Proof The claim follows from Proposition 2.4.8 because an Ad G -invariant

subspace in is also -invariant. □

A proof of the following theorem can be found in [77].

Theorem 2.4.29 (Structure of Compact Lie Groups) Let G be a compact

connected Lie group. Then G is a finite quotient of a product of the form

where the G i are compact simple Lie groups.

Compact simple Lie groups and the abelian Lie group U(1) are therefore the
building blocks of all compact connected Lie groups.

2.5 Ad-Invariant Scalar Products on Compact Lie

Groups
We know from Theorem 2.2.3 that compact Lie algebras admit scalar products
that are invariant under the adjoint action. Such scalar products are important in
gauge theory: they are necessary ingredients to construct the gauge-invariant
Yang–Mills action and are related to the notion of coupling constants. We
discuss, in particular, how to fix an Ad-invariant scalar product and how many
different ones exist on a given compact Lie algebra.
 We first consider Ad-invariant scalar products on compact simple Lie
algebras. We need the following variant of a famous theorem of Schur.

Theorem 2.5.1 (Schur’s Lemma for Scalar Products) Let ρ: G → GL(V ) be

an irreducible representation of a Lie group G on a real vector space V and ⋅ ,
⋅ 1 , ⋅ , ⋅ 2 two G-invariant symmetric bilinear forms on V, so that ⋅ , ⋅ 2 is
positive definite. Then there exists a real number such that

Remark 2.5.2 The assumption that the group representation is irreducible is

important.

Proof We follow the proof in [153]. Let L: V → V be the unique linear map
defined by (using non-degeneracy of the second scalar product)

We have

hence L is self-adjoint with respect to the second scalar product. We can split V
into the eigenspaces of L which are orthogonal with respect to the second scalar
product. Since both bilinear forms are G-invariant we have

We conclude that ρ(g) ∘ L = L ∘ρ(g) for all g G and thus the eigenspaces of L
are G-invariant. Since the representation ρ is irreducible, V itself must be an
eigenspace and hence L = a ⋅ Id V . This implies the claim. □

Theorem 2.5.3 (Ad-Invariant Scalar Products on Compact Simple Lie

Algebras) Let G be a compact simple Lie group. Then there exists up to a
positive factor a unique Ad-invariant positive definite scalar product on the Lie
algebra . The negative of the Killing form is an example of such an Ad-
invariant positive definite scalar product.

Proof Existence follows from 2.2.3. Uniqueness follows from Corollary 2.4.28
and Theorem 2.5.1. The claim about the Killing form follows from
Corollary 2.4.18. □

Let T = U(1) × … × U(1) denote an n-dimensional torus and a positive

definite scalar product on its Lie algebra

Since the adjoint representation of an abelian Lie group is trivial, any inner
product on an abelian Lie algebra is Ad-invariant. With respect to the standard
Euclidean scalar product on , the scalar product is determined by a
positive definite symmetric matrix.

Theorem 2.5.4 (Ad-Invariant Scalar Products on General Compact Lie

Algebras) Let G be a compact connected Lie group of the form

where the G i are compact simple Lie groups. Let be an Ad G -invariant

positive definite scalar product on the Lie algebra of G. Then is the
orthogonal direct sum of:

1. a positive definite scalar product on the center ;

2. -invariant positive definite scalar products on the Lie algebras

.
Conversely, the direct sum of any positive definite scalar product on the
abelian Lie algebra and any -invariant positive definite scalar products
on the simple Lie algebras is an Ad G -invariant positive definite scalar
product on .

Proof Let be an Ad G -invariant positive definite scalar product on the Lie

algebra . We have to show that it decomposes as an orthogonal direct sum of
scalar products on the summands. For any fixed i = 1, …, s we can write G = G i
× H with a compact Lie group H. Fix an arbitrary and let
Then f is a linear 1-form on and its kernel is a vector subspace of codimension
zero or one. Let g G i and . Then by Lemma 2.1.50

This implies that the kernel of f is -invariant. Since the adjoint

representation of G i is irreducible by Corollary 2.4.28 and since , the
kernel of f cannot have codimension 1. Therefore f must vanish identically.
This proves that the scalar product on decomposes as an orthogonal
direct sum of scalar products on and on . The scalar product
is Ad G -invariant, hence is -invariant and is Ad H -
invariant. We continue to split the scalar product on until the remaining
Lie algebra is the center.
Conversely, if is a scalar product on and are
-invariant scalar products on , then the orthogonal direct sum

is Ad G -invariant by Lemma 2.1.50. □

In the situation of Theorem 2.5.4 the Ad G -invariant scalar product on is

determined by certain constants:

1. The scalar product is determined by a positive definite symmetric

matrix with respect to the standard Euclidean scalar product on .

2. The scalar products are determined by positive constants relative to

some fixed -invariant positive definite scalar product on the simple Lie
algebras (like the negative of the Killing form).

Definition 2.5.5 The constants that determine an Ad G -invariant positive

 definite scalar product on the compact Lie algebra are called coupling
constants in physics.

Example 2.5.6

1. In the Standard Model, where G = SU(3) × SU(2) × U(1) , there are three
coupling constants, one for each factor.

2. In GUTs with a simple gauge group, like G = SU(5) or G = Spin(10) ,

there is only a single coupling constant.

2.6 Homotopy Groups of Lie Groups

In this section we collect some results (without proofs) on the homotopy groups
of compact Lie groups. The following fact is elementary and can be found in
textbooks on topology:

Proposition 2.6.1 (Fundamental Group of Topological Groups) The

fundamental group π 1(G) of any connected topological group G is abelian.

Regarding the order of the fundamental group of Lie groups it can be shown that
(for a proof, see [24, Sect. V.7]):

Theorem 2.6.2 (Fundamental Group of Compact Semisimple Lie Groups)

Let G be a compact connected Lie group. Then π 1(G) is finite if and only if G is
semisimple. In particular, every compact simple Lie group has a finite
fundamental group.

The only-if direction follows from Theorem 2.4.29. As an example, it is possible

to calculate the fundamental group of the classical Lie groups (see, for example,
[129]).

Proposition 2.6.3 (Fundamental Groups of Classical Compact Groups)

The fundamental groups of the classical compact linear groups are:
1. Special orthogonal groups:

2. Unitary groups (for all n ≥ 1):

3. Special unitary and symplectic groups (for all n ≥ 1):

We have the following result on the second homotopy group (for a proof, see
again [24, Sect. V.7]):

Theorem 2.6.4 (Second Homotopy Group of Compact Lie Groups) Let G

be a compact connected Lie group. Then π 2(G) = 0.

The next theorem on the third homotopy group was proved by M.R. Bott using
Morse theory [19]:

Theorem 2.6.5 (Third Homotopy Group of Compact Lie Groups) Let G be

a compact connected Lie group. Then π 3(G) is free abelian, i.e. isomorphic to
for some integer r. If G is compact, connected and simple, then .

Combining Theorem 2.6.2 and Theorem 2.6.5 we get a topological criterion to

decide whether a compact Lie group is simple:

Corollary 2.6.6 (Topological Criterion for Simplicity) Let G be a compact

connected Lie group. Then G is simple if and only if π 1(G) is finite and
.

2.7 Exercises for Chap. 2

2.7.1 Verify that the dual representations on V defined in Definition 2.1.23 and
Definition 2.1.27 are indeed representations of the Lie group G and the Lie
algebra .
2.7.2 Let

Prove the following equivalent description of :

2.7.3

1. Let denote the fundamental representation of and the

complex conjugate representation. Show that there exists a matrix
such that

Conclude that W and are isomorphic as -representations.

2. Let denote the representation of with winding number k ≠ 0.

Prove that V k and are not isomorphic as -representations.

3. Let denote the fundamental representation of the real Lie algebra

and the complex conjugate representation. Prove that V and are
not isomorphic as -representations.

4. Does the complex conjugate representation make sense for complex

representations of complex Lie algebras, like the complex Lie algebra
?

Remark It can be shown that the fundamental representation of for every

n ≥ 3 is not isomorphic to its complex conjugate. The only other compact simple
Lie algebras which have complex representations not isomorphic to their
conjugate are for every n ≥ 1 (Weyl spinor representations) and E 6 (a
27-dimensional representation), see [104]. This is one of the reasons why Lie
groups such as SU(5), Spin(10) or E 6 appear as gauge groups of Grand Unified
Theories ; see Sect. 8.5.3.

2.7.4 Determine the charges of the basis element in:

1. the fundamental representation of on ;

2. the representation of on via the isomorphism and the

complex fundamental representation of .
2.7.5

1. Consider the Lie group SU(2) with the fundamental representation on .

Each of the basis vectors τ 1, τ 2, τ 3 of from Example 1.5.32 generates
a one-parameter subgroup isomorphic to U(1). Determine the explicit
branching rule for the fundamental representation on under restriction to
these circle subgroups, i.e. determine the corresponding decomposition of
into invariant complex subspaces together with the winding numbers of
the induced representations.

2. Do the same exercise with the complex representation of SU(2) on via

the universal covering SU(2) → SO(3) and the complex fundamental
representation of SO(3).

3. Do the same exercise for the Lie group SU(3) with the fundamental
representation on and the circle subgroups generated by the basis vectors
v 1, …, v 8 of , where with the Gell-Mann matrices λ a from
Example 1.5.33 (cf. Exercise 1.9.26).
2.7.6 Consider the embedding

from Exercise 1.9.10. Let be the complex fundamental representation

of SO(2n). Determine the branching rule of the representation V under restriction
to the subgroup U(n) SO(2n). It may be helpful to first consider the case n =
1.
2.7.7 Let G be a Lie group. The center of G is defined as

Suppose that G is connected.

1. Prove that the center Z(G) is the kernel of the adjoint representation Ad G .
Conclude that Z(G) is an embedded Lie subgroup in G with Lie algebra
given by the center of .

2. Prove that is abelian if and only if G is abelian.

3. Prove that Ad G is trivial if and only if G is abelian. Conclude that the left-
invariant and right-invariant vector fields on a connected Lie group G
coincide if and only if G is abelian.
2.7.8 Consider the Lie algebra isomorphism of with from
Exercise 1.9.14.

1. Determine the symmetric bilinear form on corresponding under this

isomorphism to the Killing form .

2. Interpret the high school formula

where ⋅ denotes the scalar product, in light of the first part of this exercise.
2.7.9

1. Let be Lie algebras and a Lie algebra homomorphism. Suppose

that is simple. Show that ϕ is either injective or the trivial homomorphism.
In particular, every representation of a simple Lie algebra is either faithful or
trivial.

2. Show that every complex 1-dimensional representation of a semisimple Lie

algebra is trivial.
3. Show that every homomorphism from a connected semisimple Lie group to
U(1) is trivial. Find a non-trivial homomorphism from U(n) to U(1).
2.7.10 Let be a real Lie algebra. The complexification of is the complex
Lie algebra

with the Lie bracket from extended -bilinearly. Show that if is (semi-
)simple, then is (semi-)simple.

Remark The following converses can be shown: If is semisimple, then is

semisimple (this uses Theorem 2.4.9) and if is compact simple, then is
simple (see [83]).

2.7.11 Prove Proposition 2.4.26: the structure constants f abc for a -

orthonormal basis {T a } of a semisimple Lie algebra are totally antisymmetric:

2.7.12 Let τ 1, τ 2, τ 3 be the basis of the Lie algebra from Example 1.5.
32. Fix an arbitrary, positive real number g > 0 and let

Define a unique positive definite scalar product ⋅ , ⋅ g with associated norm

| | ⋅ | | g on so that β 1, β 2, β 3 form an orthonormal basis. Determine the
relation between det(X) and the norm | | X | | g for . Show that the scalar
product ⋅ , ⋅ g is AdSU(2)-invariant.
2.7.13 Consider the Lie algebra .

1. Calculate the Killing form directly from the definition and determine
the constant g so that , where ⋅ , ⋅ g is the scalar product from
Exercise 2.7.12.

2. Fix an arbitrary, positive, real number λ > 0 and set

where tr denotes the trace and ⋅ the matrix product. Show that − F λ is a
 negative definite AdSU(2)-invariant scalar product on . Determine the
constant λ so that .
2.7.14 Consider the Lie algebra with Killing form . Show that
there exists a constant such that

where tr denotes the trace of the matrix and ⋅ the matrix product. Determine
this constant λ. Is the Killing form definite? or non-degenerate?
2.7.15 Let .

1. Show that the Killing form of the Lie algebra can be calculated as

A suitable basis for to evaluate the trace on the left-hand side is

given by the elementary matrices E ij with a 1 at the intersection of the i-th
row and j-th column and zeros elsewhere.

2. Let be an ideal in a Lie algebra . Prove that for all

3. Show that the Killing form of the Lie algebra is equal to

Compare with Exercise 2.7.14.

2.7.16

1. Let be a real Lie algebra and its complexification as in Exercise 2.7.10.

Under the canonical inclusion as the real part show that for all

2. Explain the difference between the results for the Killing form in
Exercise 2.7.13 and Exercise 2.7.14, given the isomorphism of complex Lie
algebras
 from Exercise 1.9.18.

3. Show that every complex matrix A can be written uniquely as A = B + iC

with B, C skew-Hermitian. Conclude that

4. Show that the Killing forms of the Lie algebras and can be
calculated as

Compare with Exercise 2.7.13.

2.7.17 Consider the basis of given by the elements iλ a , where λ a are
the Gell-Mann matrices from Example 1.5.33, with a = 1, …, 8. Show that
these basis vectors are orthogonal with respect to the Killing form and
determine for all a.
2.7.18

1. The rank of a compact Lie group G is the maximal dimension of an

embedded torus subgroup T G. Prove that the rank of a product G × H of
compact Lie groups G and H is the sum of the ranks of G and H (you can
assume without proof that a connected abelian Lie group is a torus).

2. Classify compact semisimple Lie algebras of rank r = 1, 2, 3, 4, assuming

Theorem 2.4.23.

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]
3.
 Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
 1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)
34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]
51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
 [MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]
84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
 Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]
115.
 Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
129.
[MATH]
130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
 [MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions. Chin.
Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_3
Chapter 3 Group Actions
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

There are different ways in which Lie groups can act as transformation or
symmetry groups on geometric objects. One possibility, that we discussed in
Chap. 2, is the representation of Lie groups on vector spaces. A second
possibility, studied in this chapter, is Lie group actions on manifolds. Both
concepts are related: A representation is a linear action of the group where the
manifold is a vector space. Conversely, an action on a manifold can be thought
of as a non-linear representation of the group. More precisely, a linear
representation of a group corresponds to a homomorphism into the general linear
group of a vector space. A group action then corresponds to a homomorphism of
the group into the diffeomorphism group of a manifold.
Even though we are most interested in Lie group actions on manifolds, it is
useful to consider more general types of actions: actions of groups on sets and
actions of topological groups on topological spaces. We will also introduce
several standard notions related to group actions, like orbits and isotropy groups.
In the smooth case, if a Lie group G acts on a manifold M, then there is an
induced infinitesimal action of the Lie algebra , defining so-called fundamental
vector fields on M. This map can be understood as the induced Lie algebra
homomorphism from the Lie algebra of G to the Lie algebra of the
diffeomorphism group Diff(M).
In the case of smooth actions of a Lie group G on a manifold M, the
interesting question arises under which conditions the quotient space M⁄G again
admits the structure of a smooth manifold. The main (and rather difficult) result
that we prove in this context is Godement’s Theorem, which gives a necessary
and sufficient condition that quotient spaces under general equivalence relations
are smooth manifolds. The smooth structure on the quotient space is defined
using so-called slices for the equivalence classes.
 It turns out that the quotient space of a Lie group action admits the structure
of a smooth manifold in particular in the following cases:
A compact Lie group G acting smoothly and freely on a manifold M.
A closed subgroup H of a Lie group G acting on G by right (or left)
translations.
Both cases can be used to construct new and interesting smooth manifolds.
In the second case, if the closed subgroup H acts on the right on G, then there is
an additional left action of G on the quotient manifold G⁄H. This action is
transitive and G⁄H is an example of a homogeneous space. We will study
homogeneous spaces in detail in all of the three cases of group actions on sets,
topological spaces and manifolds and prove that any homogeneous space is of
the form G⁄H.
We finally apply the theory of group actions to construct the exceptional
compact simple Lie group G2, which plays an important part in M-theory, a
conjectured theory of quantum gravity in 11 dimensions, and derive some of its
properties.
General references for this chapter are [14, 24] and [142].

3.1 Transformation Groups

In this section we define group actions and study their basic properties. Since
many statements in this section are quite elementary, we designate some of the
proofs as exercises.
Before we begin with the formal definitions, let us consider some basic
examples to get a bird’s eye view of group actions. The simplest example is
perhaps the canonical left action of the general linear group GL(V ) on a vector
space V, given by the map

(3.1)
A representation of a group G on V then corresponds to a group
homomorphism

defining a linear action of G on V.

We would like to extend this idea to other types of actions. Suppose that
M is a set and S(M) the symmetric group of all bijections M → M; or
 M is a topological space and Homeo(M) the homeomorphism group of
M; or
M is a manifold and Diff(M) the diffeomorphism group of M.

Replacing V by M and GL(V ) by S(M) (Homeo(M), Diff(M)) in Eq. (3.1)

we get canonical actions of these automorphism groups on M. Actions of a
group G on M are then given by homomorphisms ϕ of G into these groups
and thus correspond to non-linear representations of G on M (which in the
case for Homeo(M) and Diff(M) should in some sense be continuous and
smooth).

In each of these cases, the images of the group G under the

homomorphisms ϕ define subgroups of GL(V ), S(M), Homeo(M) and
Diff(M) that are usually easier to handle than the full automorphism groups
themselves (which in the case of the diffeomorphism group, for example, are
infinite-dimensional if dimM ≥ 1).

An explicit example of a Lie group action on a manifold is the famous Hopf

action of S 1 = U(1) on S 3 defined by the map

where S 3 is the unit sphere in and S 1 the unit circle in (this is an

example of a right action). It is clear that the map is well-defined, i.e. it
preserves the 3-sphere, and it is smooth. The map also has the following
properties:

1. (v, w) ⋅ (λ ⋅ μ) = ((v, w) ⋅ λ) ⋅ μ

2. (v, w) ⋅ 1 = (v, w)
for all (v, w) S 3 and λ, μ S 1. We shall see that these are the defining
properties of group actions, ensuring that we obtain a homomorphism into the
diffeomorphism group. In the case of the Hopf action we can think of it as a
homomorphism
The Hopf action will also be an important example in subsequent chapters.
We shall later study properties of this and other actions. For example, we can
fix a point (v 0, w 0) S 3 and consider its orbit under the action:

In this case, the orbit map is injective for all (v 0, w 0) S 3 and the Hopf
action is therefore called free.

3.2 Definition and First Properties of Group Actions

We now come to the formal definition of group actions.

Definition 3.2.1 A left action of a group G on a set M is a map

satisfying the following properties:

1. (g ⋅ h) ⋅ p = g ⋅ (h ⋅ p) for all p M and g, h G.

2. e ⋅ p = p for all p M.
The group G is called a transformation group of M.

We can think of a group action as moving a point p M around in M as we vary

the group element g G. This is very similar to the concept of a representation
of a group on a vector space, where a vector is moved around as we vary the
group element.
If G is a topological group, M a topological space and Φ continuous, then Φ
is called a continuous left action . Similarly, if G is a Lie group, M a smooth
manifold and Φ is smooth, then Φ is called a smooth left action . Here G × M
carries the canonical product structure as a topological space or smooth
manifold.
Similarly right actions of a group G on a set M are defined as a map
satisfying the following properties:

1. p ⋅ (g ⋅ h) = ( p ⋅ g) ⋅ h for all p M and g, h G.

2. p ⋅ e = p for all p M.
There is, of course, also the notion of a continuous or smooth right action
(most of the following statements hold for both left and right actions). We can
turn every left action into a right action (and vice versa):

Proposition 3.2.2 Let

be a left action of a group G on a set M. Then

defines a right action of G on M.

Proof This is Exercise 3.12.1. □

A group action Φ is a map with two entries: a group element g G and a point
p M. It is useful to consider the maps that we obtain if we fix one of the
entries and let only the other one vary.

Definition 3.2.3 Let Φ: G × M → M be a left action. For g G we define the

left translation by

Similarly, for a right action Φ: M × G → M and g G we define the right

translation by

For p M the orbit map is given by

for a left action and

for a right action.

It is clear that for a continuous (smooth) left action the left translations l g for all
g G and the orbit maps ϕ p for all p M are continuous (smooth) maps. The
reason is that in the smooth case the map l g is given by the composition of
smooth maps

and ϕ p is given by the composition

The continuous case and the case of right actions follow similarly.
We could define left translations as above for any map Φ: G × M → M even
if Φ does not satisfy a priori the axioms of a left action. It is easy to see that
group actions are then characterized by the fact that all left translations l g for g
G are bijections of M and

is a group homomorphism. In the case of a continuous (smooth) left action,

the left translations define a group homomorphism

and

respectively, into the group of homeomorphisms (diffeomorphisms) of M.

Note that, as we said before, a continuous (smooth) group action is more than
just a group homomorphism into the homeomorphism (diffeomorphism) group,
because the group homomorphism has to be in addition continuous (smooth) in
the argument g G (one could make this precise by defining a topology or
smooth structure on the homeomorphism and diffeomorphism groups, which in
general are infinite-dimensional).
Here are some additional concepts for group actions (we define them in the
general case for group actions on sets, but they apply verbatim for continuous
and smooth group actions).
Definition 3.2.4 Let Φ be a left action of a group G on a set M.

1. The orbit of G through a point p M is

The orbit is the image of the orbit map (see Fig. 3.1).

2. The fixed point set of a group element g G is the set

3. The isotropy group or stabilizer of a point p M is

In physics, isotropy groups are also called little groups . It is an easy

exercise to show that the isotropy group G p is indeed a subgroup of G for all
p M.

Fig. 3.1 Orbit of a group action

There are analogous definitions for right actions.

Remark 3.2.5 We shall see later in Corollary 3.8.10 that for a smooth action of
a Lie group G on a manifold M the orbit through every point p M is an
(immersed or embedded) submanifold of M.

Lemma 3.2.6 (Two Orbits Are Either Disjoint or Identical) Let Φ be an

action of a group G on a set M and p M an arbitrary point. If , then
. Hence the orbits of two points in M are either disjoint or identical.
This means that orbits which intersect in one point are already identical.

Proof Suppose Φ is a left action. Then q is of the form q = g ⋅ p for some g

G. We get

because the right translation R g : G → G is a bijection. □

Remark 3.2.7 We can also phrase this differently: The relation

for p, q M defines an equivalence relation on M and the orbits of G are

precisely the equivalence classes. M is therefore the disjoint union of the orbits
of G.

Definition 3.2.8 Let Φ be an action of a group G on a set M. Then the

following subset of the powerset of M

is called the space of orbits or the quotient space of the action.

Note that the subsets of M become elements (points) in M⁄G. If we think of

the subset as a point in M⁄G, we also denote it by [p] or . The map

is called the canonical projection . If x M⁄G, then a point p M with [p] = x

is called a representative of x.
Concerning isotropy groups we can say the following.

Proposition 3.2.9 (Isotropy Groups Are (Closed) Subgroups) Let Φ be an

action of a group G on M and let p M be any point. If the group action is
continuous on a Hausdorff space M or smooth on a manifold M, then the
stabilizer G p is a closed subgroup of G. In particular, in the smooth case the
stabilizer G p is an embedded Lie subgroup of G by Cartan’s Theorem 1.1.44.

Proof This is an exercise. □

Suppose ϕ: G → S(M) is the group homomorphism induced from a group action.

Then
In particular, for continuous actions on Hausdorff spaces or smooth actions on
manifolds, the normal subgroup ker ϕ is closed in G.
We want to compare the isotropy groups of points on the same G-orbit.
Suppose p, q are points in M on the same G-orbit. It is easy to check that there
exists an element g G such that

In particular, the isotropy groups G p and G q are isomorphic.

In the case of a smooth action of a Lie group G we call the Lie algebra of
the stabilizer G p of a point p M the isotropy subalgebra . The following
description of the isotropy subalgebra is useful in applications.

Proposition 3.2.10 (The Isotropy Subalgebra and the Orbit Map) Let Φ be
a smooth action of a Lie group G on a manifold M. Fix a point p M and let ϕ
p denote the orbit map

as before. Then the kernel of the differential

is equal to the isotropy subalgebra .

Proof We assume that the action is on the left, the case of right actions follows
similarly. If , then exp(tX) G p for all and therefore

This implies that X is in the kernel of the differential D e ϕ p . Conversely,

suppose that X is in the kernel of D e ϕ p . Then

This implies

Therefore, the curve exp(tX) ⋅ p is constant and equal to exp(0) ⋅ p = p. This

implies that exp(tX) G p for all and thus by Corollary 1.8.11. □
Definition 3.2.11 Let Φ be an action of a group G on a set M. We distinguish
three cases, depending on whether the orbit map is surjective, injective or
bijective for every p M.

1. The action is called transitive if the orbit map is surjective for every p
M. In other words, M consists of only one orbit, for every p M.
We then call M a homogeneous space for G.

2. The action is called free if the orbit map is injective for every p M.

3. The action is called simply transitive if it is both transitive and free, i.e. if
the orbit map for every p M is a bijection from G onto M.

We leave it as an exercise to show the following properties of G-actions on M:

1. The orbit map is surjective for one p M if and only if it is surjective for
all p M.

2. The action is transitive if and only if M⁄G consists of precisely one point.

3. The orbit map of a point p M is injective if and only if the isotropy group
of p is trivial, G p = {e}.

4. The action is free if and only if g ⋅ p ≠ p for all p M, g ≠ e G. Hence

the action is free if and only if all points in M have trivial isotropy group or,
equivalently, all group elements g ≠ e have empty fixed point set.
As a consequence of Proposition 3.2.10 we see:

Corollary 3.2.12 (Orbit Maps of Smooth Free Actions) If Φ is a smooth

free action of a Lie group G on a manifold M, then the orbit maps ϕ p : G → M
are injective immersions for every point p M. If G is compact, then the orbit
maps are embeddings.

In the case of a free action of a compact Lie group G each orbit is therefore an
embedded submanifold diffeomorphic to G.
Definition 3.2.13 An action Φ is called faithful or effective if the induced
homomorphism ϕ: G → S(M) is injective.

It is not difficult to see that if one point in M has trivial isotropy group, then the
action is faithful. We can always make a group action faithful by passing to the
induced action of the quotient group G⁄ker ϕ.
It is sometimes important to compare actions of a group G on two sets M and
N. In particular, we would like to have a notion of isomorphism of group actions.

Definition 3.2.14 Let Φ: G × M → M and Ψ: G × N → N be left actions of a

group G on sets M and N. Then a G -equivariant map f: M → N is a map such
that

If G is a topological (Lie) group and the actions continuous (smooth), we

demand in addition that f is continuous (smooth). A G-equivariant bijection
(homeomorphism, diffeomorphism) is called an isomorphism of G-actions.
There are analogous definitions in the case of right actions.

3.3 Examples of Group Actions

We discuss some common examples of group actions, in particular, smooth
actions of Lie groups on manifolds.
It is quite easy to define group actions of discrete abelian groups on
manifolds: Any diffeomorphism f: M → M defines a smooth group action

If f happens to be periodic, f n = Id M for some integer n, then this defines a

smooth group action of the cyclic group :

If f 1, …, f m are pairwise commuting diffeomorphisms of M, then

is a smooth group action.

An example of a simply transitive action is to take M = G for an arbitrary
group G and let G act on itself by left (and right) translations:
 Another type of action that is easy to define is given by group
representations. Let ρ: G → GL(V ) be a representation of a Lie group G on a
real (or complex) vector space V. Then

is a smooth left action on the manifold V (which is diffeomorphic to a

Euclidean space). Such an action is called linear . We can also define a smooth
right action by

Note that it is important to take the inverse of ρ(g)−1, otherwise the first
property of a right action is in general not satisfied (see Exercise 3.12.1).
In both cases, the orbit of 0 V consists only of one point,

and thus the isotropy group of 0 is all of G,

For a non-zero vector v ≠ 0 the isotropy group in general will be a proper

subgroup of G,

This is the basic mathematical idea behind symmetry breaking (from the
full group G to the subgroup G v ), one of the centrepieces of the Standard Model
that we discuss in Chap. 8
For a linear representation, the homomorphism induced by the action has
image in GL(V )

The action is faithful if and only if the representation is faithful.

Suppose in addition that and the representation ρ is orthogonal. Then
ρ has image in O(n) and the action maps the unit sphere S n−1 in around the
origin to itself. We therefore get a smooth left action

Similarly, if and the representation ρ is unitary, so that it has image in

U(n), then the action preserves the unit sphere S 2n−1 in . We get a smooth left
action

Finally, assume that and the representation ρ is quaternionic

unitary, by which we mean that ρ has image in Sp(n). Then the action preserves
the standard symplectic scalar product (see Definition 1.2.9) and induces a
smooth left action on the unit sphere S 4n−1 in :

In each case we can similarly define right actions, using the inverses ρ(g)−1.
These actions on spheres are again called linear.
An important special case of this construction is the following:

Definition 3.3.1 Consider the groups , and of non-zero real, complex

and quaternionic numbers. We define for the following linear right
actions by scalar multiplication:

These actions are free and induce the following free linear right actions of the
groups of real, complex and quaternionic numbers of unit norm on unit spheres:

These actions are called Hopf actions . The most famous example is the action
of S 1 on S 3 that we already considered at the beginning of Sect. 3.1.

Note that , and . We shall see later in Example 3.7.34

that the quotient spaces under these free actions are smooth manifolds

of dimension n, 2n, 4n, called real , complex and quaternionic projective

space .
 We consider another example of linear actions on spheres.

Theorem 3.3.2 (Linear Transitive Actions of Classical Groups) The

defining (fundamental) representations of O(n), SO(n), U(n), SU(n) and Sp(n)
define the following linear transitive actions on spheres with associated isotropy
groups of the vector e 1 :

1. O(n)-action on S n−1 with isotropy group

Similarly, there is an

2. SO(n)-action on S n−1 with isotropy group isomorphic to SO(n − 1).

3. U(n)-action on S 2n−1 with isotropy group isomorphic to U(n − 1).

4. SU(n)-action on S 2n−1 with isotropy group isomorphic to SU(n − 1).

5. Sp(n)-action on S 4n−1 with isotropy group isomorphic to Sp(n − 1).

For the defining (fundamental) representations of and
define the following linear transitive actions with associated isotropy
groups of the vector e 1 :

6. -action on with isotropy group

7. -action on with isotropy group

8. -action on with isotropy group

Proof This is an exercise. The case of uses the following lemma. □

Lemma 3.3.3 For and the following equation holds:

Proof This is an exercise. □

We saw above that from a representation of a group on a vector space, we

sometimes get group actions on other manifolds, in particular on spheres. We
now show that from smooth actions on manifolds we also get representations on
certain vector spaces.
Let Φ: G × M → M be a smooth (left) action of a Lie group G on a manifold
M. Let p M be a point and G p its isotropy subgroup. By Proposition 3.2.9 the
isotropy group G p is an embedded Lie subgroup of G. The differential of the left
translation l g is a map

for all g G p . This is an isomorphism with inverse .

Theorem 3.3.4 (Isotropy Representation) The map

is a representation of the isotropy group G p on T p M, called the isotropy

representation .

Proof We follow [142]. For g, h G we calculate

where we used the chain rule. Hence ρ p is a group homomorphism. We want to

show that ρ p is smooth. Let v T p M be arbitrary and fixed. Then the map ρ p
(⋅ )v is the composition of smooth maps
given by

It follows that ρ p is a smooth homomorphism, hence a representation. □

We get an analogous isotropy representation for right actions using the

differential of right translations. Here is an almost trivial example of this
construction.

Example 3.3.5 Let G be a Lie group, ρ a G-representation on a vector space V

and Φ: G × V → V the induced linear action. Then the isotropy group of 0 V
is all of G,

and the isotropy representation on can be identified with ρ itself

because the action is linear.

Here is a more interesting example:

Example 3.3.6 Every Lie group G acts on itself on the left by conjugation:

The isotropy group of e G is the full group G,

and the isotropy representation on is the adjoint representation

The adjoint representation can thus be seen as a special case of the general
construction of isotropy representations.

3.4 Fundamental Vector Fields

Suppose a Lie group G acts smoothly on a manifold M. We want to discuss a
construction that defines for every vector in the Lie algebra a certain vector
field on M. These vector fields correspond to an infinitesimal action of on M
(the construction only works for smooth Lie group actions on manifolds).
We can think of this from an abstract point of view as follows: if f: G → H
 is a Lie group homomorphism, then we saw in Sect. 1.5.3 that there is an
induced Lie algebra homomorphism

Suppose now that the Lie group G acts smoothly on a manifold M. We

know that this action corresponds to a homomorphism

where ϕ is in a certain sense smooth. We can ask whether there is again an

induced homomorphism on the level of Lie algebras.

We first have to determine the Lie algebra of the diffeomorphism group

Diff(M): note that if Y is a vector field on M, then its flow generates a 1-
parameter family of diffeomorphisms of M. If we think of the flow of Y as an
exponential map applied to Y, it is clear that the Lie algebra of Diff(M)
consists of the Lie algebra of vector fields on M with the standard
commutator (this is plausible even if we do not formally define Diff(M) as an
infinite-dimensional Lie group). Given a Lie group action of G on M we
therefore look for an induced Lie algebra homomorphism

For example, in the case of the Hopf action

it follows from the definition below that the induced homomorphism

is given by

with . Here ϕ (ix) is indeed a tangent vector field on S 3.

Definition 3.4.1 Let G be a Lie group and M a manifold. Suppose that M × G

→ M is a right action. For we define the associated fundamental vector
field on M by

If we denote by ϕ p the orbit map for the right action,

then

Similarly, suppose that G × M → M is a left action. Then we define the

fundamental vector field by

for p M. If we denote by ϕ p ′ the following orbit map for the left action,

then

The minus sign in the definition of the fundamental vector field for left actions
has a reason that will become clear in Proposition 3.4.4.
The formula for the fundamental vector fields has the following
interpretation: recall that vectors X in the Lie algebra define one-parameter
subgroups, given by exp(tX) with . The action of such a subgroup on a point
p M defines a curve in M and the fundamental vector field in p is given as the
velocity vector of this curve at t = 0 (up to the sign in the case of left actions).

Example 3.4.2 Let ρ: G → GL(V ) be a representation of a Lie group G on a

vector space V. The representation defines a left action

Let be the induced representation of the Lie algebra. For ,

the fundamental vector field is then given by

Here are some properties of fundamental vector fields.

Proposition 3.4.3 (Fundamental Vector Fields of Free Actions) Let G be a

Lie group acting on a smooth manifold M. If the action is free, then the map
is injective.

Proof This follows from Proposition 3.2.10. □

Proposition 3.4.4 (Fundamental Vector Fields Define Lie Algebra

Homomorphism) Let G be a Lie group acting on a manifold M on the right
or left. The map

that associates to a Lie algebra element the corresponding fundamental vector

field on M is a Lie algebra homomorphism , i.e. it is an -linear map such that

In particular, the set of all fundamental vector fields is a Lie subalgebra of the
Lie algebra of all vector fields on M.

Proof We prove the claim if G acts on the left on M. The proof for right actions
follows similarly. Fix a point p M and let ϕ p ′ denote the following orbit map

The second definition of ,

shows that the map

is linear.
We want to show that the left-invariant vector field and are ϕ
′ -related. For this we have to show that
p

for all a G. We have, since X is a left-invariant vector field on G,

The claim now follows from Proposition A.1.49. □

Remark 3.4.5 The reason why we defined in Definition 3.4.1 the fundamental
vector field for left actions with a minus sign in exp(−tX) is so that

holds for all . If we defined the fundamental vector field for left actions
with exp(tX) instead (this is sometimes done in the literature), then we would get
a minus sign here:

because on the left-hand side we have to change the sign once and on the right-
hand side twice.

It is sometimes useful to know how fundamental vector fields behave under right
or left translations on the manifold. It will turn out that even though fundamental
vector fields are defined using the group action, they are in general not invariant
under the action.

Proposition 3.4.6 (Action of Right and Left Translations on Fundamental

Vector Fields) Suppose a Lie group G acts on a manifold M. Let and g
G.

1. If G acts on the right on M, then

where

2. If G acts on the left on M, then

where
Proof We prove the statement for right actions, the statement for left actions
follows similarly. At a point p M we calculate

with the orbit map

On the other hand

This implies the claim by the second definition of the fundamental vector field.
□

Corollary 3.4.7 (Translations of Fundamental Vector Fields Are

Fundamental) For a right (left) action of a Lie group G on a manifold M the
right (left) translations of fundamental vector fields are again fundamental
vector fields. If the Lie group G is abelian, then the fundamental vector fields are
invariant under all right (left) translations.

3.5 The Maurer–Cartan Form and the Differential

of a Smooth Group Action
3.5.1 Vector Space-Valued Forms
Recall from Definition A.2.3 that a k-form on a real vector space V is defined as
an alternating multilinear map

The vector space of all k-forms on V is denoted by Λ k V .

Suppose W is another real vector space. Then we define a k -form on V with
values in W as an alternating, multilinear map

The vector space of all k-forms on V with values in W can be identified with
the tensor product Λ k V W.
 Similarly we defined k-forms on a smooth manifold as alternating -
multilinear maps

and we defined Ω k (M) as the set of all k-forms on M; see Definition A.2.12.
We now define

as the set of all smooth maps from M into the vector space W (the vector
space W has a canonical structure of a manifold, so that smooth maps into W are
defined). A k -form on M with values in W is then an alternating -
multilinear map

The set of all k-forms on M with values in W can be identified with

. One also calls the forms in Ω k (M, W) twisted with W .

Remark 3.5.1 Note that there is no canonical wedge product of forms on a

vector space or a manifold with values in a vector space W, because there is in
general no canonical product W × W → W (an exception is forms with values in
, where there is indeed a canonical wedge product).

3.5.2 The Maurer–Cartan Form

The following notion of a vector space-valued form on a Lie group is useful for
studying group actions and principal bundles. Let G be a Lie group with Lie
algebra .

Definition 3.5.2 The Maurer–Cartan form is the 1-form on G

with values in defined by

for all g G and v T g G. The Maurer–Cartan form is also called the

canonical form or structure form.

The Maurer–Cartan form thus associates to a tangent vector v at the point g G

the unique left-invariant vector field X on G whose value at g is X g = v
(equivalently, the generating vector of this vector field at e G).

Proposition 3.5.3 (Invariance of Maurer–Cartan Form Under

Translations) The Maurer–Cartan form has the following invariance
properties under left and right translations:

for all g G.

Proof We calculate for all h G and v T h G:

The statement for L g follows similarly. □

3.5.3 The Differential of a Smooth Group Action

Recall that a smooth (right) action of a Lie group is a map Φ: M × G → M
satisfying certain axioms. It is sometimes useful to determine the differential of
this map in a given point (x, g) M × G. The formula for this differential
involves the Maurer–Cartan form.

Proposition 3.5.4 (The Differential of a Smooth Group Action) Let G be a

Lie group acting smoothly on the right on a manifold M,

Then under the canonical identification

the differential of the map Φ is given by

where r g denotes right translation and μ G denotes the Maurer–Cartan form.

Proof Let ϕ x : G → M denote the orbit map

Let x(t) be a curve in M tangent to X and g(t) a curve in G tangent to Y. Then

Let denote the left-invariant vector field corresponding to Y. Then y = μ G
(Y ). In the proof of Proposition 3.4.4 we saw that

(we proved the statement for left actions, but the corresponding statement also
holds for right actions). This proves the claim. □

3.6 Left or Right Actions?

In general there is no difference whether we assume that a group action is a left
or right action. However, when we discuss homogeneous spaces in Sect. 3.8,
there will be two different actions at the same time, which have to be
compatible. We therefore make the following conventions:
If we are interested in quotient spaces M⁄G, we take the G-action on M to be
a right action. In particular, if H G is a subgroup and we want to
consider G⁄H, then H acts on G on the right. When we consider principal
bundles in Chap. 4, we will take the G-action on the principal bundle to be a
right action as well. For example, the Hopf actions introduced in
Definition 3.3.1 are right actions whose quotient spaces are the projective
spaces.
If we are interested in homogeneous spaces, i.e. spaces M with a transitive
group action, we will take the G-action on M to be a left action. For
example, the linear transitive actions on spheres introduced in
Theorem 3.3.2 are left actions.
Usually we are not interested in the quotient space of a transitive group
action, because it consists only of a single point, so that both cases do not
overlap. Occasionally one encounters situations in the literature where we have a
right G-action on M with quotient space M⁄G and a non-transitive left K-action
on M⁄G. Then it makes sense to consider the quotient space K∖M⁄G under the left
K-action (we will not consider such quotients in the following).

3.7 Quotient Spaces

 An important objective in the study of group actions is to understand the
quotient space of a given action. In this section we are specifically interested
in the following question: Suppose that G is a Lie group acting smoothly on a
manifold M. Under which circumstances does the quotient set M⁄G have the
structure of a smooth manifold?

This question has many applications, because it is possible to construct

new and interesting manifolds as quotients of this form (like projective spaces
and lens spaces, to name only two examples). For instance, in the case of the
Hopf action

which is a free action, it can be shown that the quotient space S 3⁄U(1) is a
smooth manifold diffeomorphic to .

It is useful to study the question of quotients in greater generality: we first

consider quotients of manifolds (and topological spaces) under arbitrary
equivalence relations and later the case of the equivalence relation defined by
group actions. We follow [130] for smooth manifolds and the excellent
exposition in [139] in the general case. An additional reference is [89].

3.7.1 Quotient Spaces Under Equivalence Relations on

Topological Spaces
Suppose X is a set and ∼ an equivalence relation on X. We can describe ∼
equivalently by a subset R X × X so that

The equivalence class of an element x X is the subset

As subsets of X, equivalence classes of two elements x, x ′ X are either disjoint

or identical. We denote by X⁄R the space of equivalence classes, called the
quotient space

We have the canonical projection

We now specialize to the case when X is a topological space. Then we define on
X⁄R the usual quotient topology by setting U X⁄R open if and only if π −1(U)
X is open. It is easy to check that this indeed defines a topology on X⁄R. The
canonical projection π: X → X⁄R is continuous. The following is well-known:

Lemma 3.7.1 A map f: X⁄R → Y from a quotient space to another topological

space is continuous if and only if f ∘π is continuous:

We are first interested in the following question: under which conditions is the
quotient space X⁄R Hausdorff? The answer is given by the following lemma.

Lemma 3.7.2 (Hausdorff Property of Quotient Spaces Under Equivalence

Relations) Let X be a topological space.

1. If X⁄R is Hausdorff, then R X × X is closed.

2. If π: X → X⁄R is open and R X × X is closed, then X⁄R is Hausdorff.

Remark 3.7.3 Note that we do not need to assume that X is Hausdorff.

Proof We use in the proof the following standard fact from point set topology:
a topological space Y is Hausdorff if and only if the diagonal

is a closed subset in Y × Y. In the following, we denote by Δ the diagonal in the

space X⁄R × X⁄R.

1. The map

is continuous. Since X⁄R is Hausdorff, the diagonal Δ is closed, hence the

preimage (π ×π)−1(Δ) is closed. We have
 Hence R = (π ×π)−1(Δ) is closed in X × X.

2. The map π ×π is open and (X × X)∖R is open, hence its image in X⁄R × X⁄R is
open. We have

It follows that Δ is closed and X⁄R is Hausdorff.

□

3.7.2 Quotient Spaces Under Equivalence Relations on

Manifolds
We now consider the case of an equivalence relation R on a smooth manifold M
and we would like to determine when the quotient space M⁄R is a smooth
manifold. It is useful to demand that the smooth structure has the additional
property that π: M → M⁄R is a submersion. Consider the following lemma.

Lemma 3.7.4 (Surjective Submersions Admit Local Sections) Let p: M →

N be a surjective submersion between smooth manifolds. Then p admits smooth
local sections , i.e. for each x N there exists an open neighbourhood U N
of x and a smooth map s: U → M such that p ∘ s = Id U .

Proof This follows from the normal form theorem for submersions (see
Theorem A.1.28), because locally submersions are projections. □

The following lemma is very useful in applications.

Lemma 3.7.5 (Smoothness of Maps Out of the Target Space of a

Surjective Submersion) Let p: M → N be a surjective submersion. Then a
map f: N → Q is smooth if and only if f ∘ p: M → Q is smooth. Moreover, f is
a submersion if and only if f ∘ p is a submersion and f is surjective if and only
if f ∘ p is surjective.

Proof If f is smooth, then f ∘ p is smooth. Conversely, assume that f ∘ p is

smooth. Let x N and U N an open neighbourhood of x with a smooth
section s: U → M for p. On U we have p ∘ s = Id U , hence

Thus f is smooth on U and therefore on all of N.

The claim about submersions and surjectivity is clear, because p and its
differential are surjective. □

Corollary 3.7.6 Let M be a manifold and p: M → N a surjective map to a set

N. Then N admits at most one structure of a smooth manifold so that p is a
submersion.

Proof Suppose N 1 and N 2 are structures of smooth manifolds on N so that p is

a submersion in both cases. By Lemma 3.7.5 the identity map Id N : N 1 → N 2 is
a diffeomorphism.

Corollary 3.7.7 (Uniqueness of Smooth Manifold Structure on Quotient

Spaces) Let M be a smooth manifold and R an equivalence relation on M.
Then there exists at most one smooth structure on M⁄R so that π: M → M⁄R is a
submersion.

Remark 3.7.8 Lemma 3.7.5 and Corollary 3.7.7 are the reasons why the
smooth structure on M⁄R should have the property that π: M → M⁄R is a
submersion.

We assume from now on that M is a smooth manifold and R an equivalence

relation on M. We first derive a necessary condition for M⁄R to be a smooth
manifold such that π is a submersion.

Lemma 3.7.9 Let M⁄R have the structure of a smooth manifold so that π: M →
M⁄R is a surjective submersion. Then R is a closed embedded submanifold of M
× M and the restrictions of the projections

are surjective submersions, for i = 1, 2.

Proof The graph of the projection

is a closed embedded submanifold of M × M⁄R and

is a submersion. Therefore F −1(Γ) is a closed embedded submanifold of M × M.

We have

This shows that R is a closed embedded submanifold of M × M.

The map F | R : R → Γ is a surjective submersion. The projection pr1 | Γ : Γ →
M is also a surjective submersion, because

It follows that

is a smooth surjective submersion. This map is equal to pr1 | R . The claim for
pr2 | R follows by symmetry of the equivalence relation. □

It is a non-trivial fact that the converse also holds.

Theorem 3.7.10 (Godement’s Theorem on the Manifold Structure of

Quotient Spaces) Let R be an equivalence relation on a manifold M.
Suppose that R is a closed embedded submanifold of M × M and pr1 | R : R →
M a surjective submersion. Then M⁄R has a unique structure of a smooth
manifold such that the canonical projection π: M → M⁄R is a surjective
submersion .

The proof of Godement’s Theorem, which is not easy and quite technical, is
deferred to Sect. 3.11. We first want to derive some consequences of it.

3.7.3 Quotient Spaces Under Continuous Group Actions

We begin more generally by considering the case of a topological group G acting
continuously on the right on a topological space X. The map defining the action
is
We would like to determine under which conditions the quotient space X⁄G is
Hausdorff (we do not need to assume that X itself is Hausdorff).

Lemma 3.7.11 The canonical projection π: X → X⁄G is open.

Proof Let U be an open subset of x. We have to show that π −1(π(U)) is open in

X. However,

and each of the sets U ⋅ g is open, because right translations are

homeomorphisms. □

Corollary 3.7.12 (Hausdorff Property of Quotient Spaces Under Continuous

Group Actions) The quotient space X⁄G is Hausdorff if and only if the map

has closed image.

Proof According to Lemma 3.7.2 and since π: X → X⁄G is open, the space X⁄G
is Hausdorff if and only if the equivalence relation R X × X is closed. We have

This shows that R is equal to the image of the map Ψ. □

Let G be a topological group and H G a subgroup with the subspace topology.

Then H acts continuously on the right on G via right translations

We get a topological quotient space G⁄H.

Corollary 3.7.13 (Hausdorff Property of G⁄H) Let G be a topological group

and H G a subgroup. Then G⁄H is Hausdorff if and only if H is a closed set in
the topology of G.

Proof According to Corollary 3.7.12 we have to show that the image of

is closed if and only if H is closed. Consider the map

This map is a homeomorphism and we have Ψ = T | G×H . Hence the image of Ψ

is closed in G × G if and only if G × H is closed in G × G. This happens if and
only if H is closed in G. □

As an aside we note the following result, which is useful in applications such as

Example 3.8.11 (we follow [34] and [142] in the proof).

Proposition 3.7.14 (Connectedness of G and G⁄H) Let G be a topological

group and H G a closed subgroup. Suppose that H is connected. Then G⁄H is
connected if and only if G is connected.

Proof If G is connected, then G⁄H is connected, because the canonical

projection π: G → G⁄H is surjective and continuous.
Conversely, suppose that G⁄H is connected and

where U, V are open non-empty subsets of G. We have to show that U and V

cannot be disjoint.
By Lemma 3.7.11 the sets π(U) and π(V ) are open and non-empty in G⁄H
with

Since G⁄H is connected there exists an element

Because of G = U V we get

By construction gH ∩ U, gH ∩ V are open and non-empty in gH. Since gH is

connected, the claim follows. □

3.7.4 Proper Group Actions

We consider some topological notions that are useful in applications to group
actions.

Definition 3.7.15 A topological space X is called locally compact if every

point in X has a compact neighbourhood.
Lemma 3.7.16 Let X be a locally compact Hausdorff space. Then a subset A
X is closed if and only if the intersection of A with any compact subset of X is
compact.

Proof If A is closed, then the intersection with any compact subset of X is

compact. Conversely, assume that A ∩ K is compact for every compact subset K
X. Let x X∖A. Since X is locally compact, there exists an open
neighbourhood U X of x contained in a compact subset K X. By
assumption, C = A ∩ K is compact, hence closed in X, since X is Hausdorff.
Then U∖C = U ∩ (X∖C) is an open neighbourhood of x contained in X∖A. This
implies the claim. □

Definition 3.7.17 A continuous map f: X → Y between topological spaces is

called proper if the preimage f −1(K) of every compact subset K Y is compact
in X.

Lemma 3.7.18 Let f: X → Y be a continuous proper map between topological

spaces, where Y is locally compact Hausdorff. Then f is a closed map.

Proof Let A X be a closed set. By Lemma 3.7.16 we have to show that f(A)
∩ K is compact for every compact subset K Y. However,

Since f is proper, the set f −1(K) is compact and thus A ∩ f −1(K) and
are compact. This implies the claim. □

Lemma 3.7.19 Let f: X → Y be a closed continuous map between topological

spaces such that f −1(y) is compact for all y Y. Then f is proper.

Proof The proof is left as an exercise. □

We consider the following type of group actions.

Definition 3.7.20 A continuous action of a topological group G on a

topological space X is called proper if the map

is proper.
Corollary 3.7.21 (Map Ψ Is Closed If Action Is Proper) Let X × G → X be
a continuous, proper action of a topological group G on a topological space X,
where X is locally compact Hausdorff. Then the map

is closed. In particular, X⁄G is Hausdorff.

Proof This follows from Lemma 3.7.18 and Corollary 3.7.12. □

Here is a general situation in which group actions are proper.

Proposition 3.7.22 (Actions of Compact Topological Groups Are Proper)

Let X × G → X be a continuous action of a topological group G on a Hausdorff
space X. Suppose that G is compact. Then the action is proper.

Proof Let K X × X be a compact subset. Then

is a compact subset of X. If Ψ(x, g) = (x, xg) K, then x L, hence

However, Ψ −1(K) is closed in X × G and L × G is compact, hence Ψ −1(K) is

compact. □

Corollary 3.7.23 Let X × G → X be a continuous action of a topological

group G on a locally compact Hausdorff space X. Suppose that G is compact.
Then X⁄G is Hausdorff.

3.7.5 Quotient Spaces Under Smooth Group Actions

We have now arrived at the central topic in this section: to determine under
which conditions the quotient of a smooth action of a Lie group on a smooth
manifold is again a smooth manifold.

Definition 3.7.24 We call a smooth right action of a Lie group G on a manifold

M principal if the action is free and the map

is closed.
Theorem 3.7.25 (Manifold Structure on Quotient Spaces Under Principal
Actions of Lie Groups) Suppose that Φ is a principal right action of the Lie
group G on the manifold M. Then M⁄G has a unique structure of a smooth
manifold such that π: M → M⁄G is a submersion.

Proof Since the action of G on M is free, the map Ψ is injective. We want to

show that Ψ is an immersion: by Proposition 3.5.4 the differential of Ψ is given
by

If D (x, g)(X, Y ) = (0, 0), then X = 0 and . From Proposition 3.4.3 we

get μ G (Y ) = 0, hence Y = 0. This proves that the differential of Ψ is injective.
Since Ψ is a closed injective map, it is a homeomorphism onto its image R
and thus an embedding. Hence R is a closed embedded submanifold of M × M.
According to Theorem 3.7.10 it remains to show that pr1 | R : R → M is a
submersion. However,

is just pr1: M × G → M and thus a submersion. This implies the claim. □

Corollary 3.7.26 (The Differential of the Projection π: M → M⁄G) Suppose

that Φ is a principal right action of the Lie group G on the manifold M. Then the
dimension of the quotient manifold M⁄G is given by

In particular, the kernel of the differential

at a point p M is equal to the tangent space of the G-orbit through p.

Proof The claim about the dimension of M⁄G follows from the proof of
Theorem 3.7.10. The second claim then follows from Corollary 3.2.12. □

Remark 3.7.27 For a free action of a Lie group G on a manifold M the

preimage Ψ −1( p, q) of every point ( p, q) M × M is either empty or consists of
a single element (and is thus a compact set). Lemma 3.7.19 implies that principal
Lie group actions on manifolds are proper. Together with Corollary 3.7.21 we
conclude that principal actions are equivalent to free proper Lie group actions on
manifolds.
We formulate this as follows:

Corollary 3.7.28 (Free Proper Actions of Lie Groups Are Equivalent to

Principal Actions) Suppose that M × G → M is a smooth free action of a Lie
group G on a manifold M. Then the action is principal if and only if it is proper.

Proposition 3.7.22 then implies:

Corollary 3.7.29 (Free Actions of Compact Lie Groups Are Principal)

Suppose that M × G → M is a smooth free action of a compact Lie group G on a
manifold M. Then the action is principal.

We get the following corollary, which is very useful in applications.

Corollary 3.7.30 (Quotients Under Free Actions of Compact Lie

Groups) Let G be a compact Lie group acting smoothly and freely on a
manifold M. Then M⁄G has a unique structure of a smooth manifold such that
π: M → M⁄G is a submersion.

Proof This follows from Theorem 3.7.25 and Corollary 3.7.29. □

Remark 3.7.31 (Fundamental Groups) If in the situation of Corollary 3.7.30

the manifolds M and G are connected, it follows from Exercise 3.12.6 that π
maps the fundamental group of M surjectively onto the fundamental group of
M⁄G . In particular, if M is simply connected, then M⁄G is simply connected.

Example 3.7.32 (Quotients Under Free Actions of Finite Groups) Finite groups
with the discrete topology are compact. Hence if a finite group G acts freely and
smoothly on a manifold M, then the quotient M⁄G is a smooth manifold such that
the canonical projection is a submersion.

Example 3.7.33 (Lens Spaces) Let p > 0 be an integer and α = e 2πi⁄p S 1 the
corresponding root of unity. Let q ≠ 0 be an integer coprime to p. We consider
the following smooth action of on the unit sphere :

This action is free: if z 1 ≠ 0 and z 1 α k = z 1, then [k] = 0. If z 2 ≠ 0 and z 2 α kq =

z 2, then again [k] = 0, because q is coprime to p. According to Example 3.7.32
the quotient

under this action is a smooth 3-dimensional manifold. These manifolds are

called lens spaces .

Example 3.7.34 (Projective Spaces Are Smooth Manifolds) The projective

spaces

are quotients of manifolds under smooth free actions of compact Lie groups and
therefore smooth manifolds such that the canonical projections are submersions .

Let G be a Lie group and H G a closed subgroup. According to Cartan’s

Theorem 1.1.44 the subgroup H is an embedded Lie subgroup of G. There is a
smooth right action of H on G by right translations

Corollary 3.7.35 (Manifold Structure on G⁄H) Let G be a Lie group and

H G a closed subgroup. Then the right action of H on G is principal and
G⁄H has a unique structure of a smooth manifold such that π: G → G⁄H is a
submersion.

Proof It is clear that the orbit maps

are injective, hence the action is free. According to Theorem 3.7.25 it remains to
show that the map

is closed. As in the proof of Corollary 3.7.13 we consider the map

This map is a diffeomorphism with Ψ = T | G×H . If A G × H is closed, then A
is closed in G × G, since H is closed in G. This implies that Ψ(A) = T(A) is
closed in G × G. □

Corollary 3.7.36 Let G be a Lie group and H G a closed subgroup. Then

the dimension of the quotient manifold G⁄H is given by

In particular, the kernel of the differential

is equal to the Lie algebra of H.

Proof This follows from Corollary 3.7.26. □

3.8 Homogeneous Spaces

Recall that a set M together with a transitive action of a group G is called a
homogeneous space . An example is the transitive action of SO(n) on the sphere
S n−1 with isotropy group isomorphic to SO(n − 1), cf. Theorem 3.3.2. In this
section we study the structure of homogeneous spaces for actions of groups,
topological groups and Lie groups. We are most interested in the case of Lie
group actions, but the other two cases are useful as a warm-up. We will show
that every homogeneous space is, up to isomorphism of group actions, of the
form G⁄H, where H is a suitable subgroup of G.

3.8.1 Groups and Homogeneous Spaces

Let G be any group and H G a subgroup. Then H acts on the right on G. We
get a quotient space G⁄H of orbits, also called left cosets.

Definition 3.8.1 We define a map

Note that G acts on the left on the set of left cosets G⁄H.

We want to show that Φ is indeed a group action.

Proposition 3.8.2 (Φ Is a Transitive Left Action of G on G⁄H)

1. The map Φ is a well-defined, transitive group action of G on the set G⁄H.

2. The isotropy group of [e] G⁄H is equal to H. Therefore the isotropy group
of any point in G⁄H is isomorphic to H.

Proof This is an easy exercise. □

We now consider an arbitrary transitive group action of G on a set M. We want

to show that up to an equivariant bijection this group action is of the form above.

Proposition 3.8.3 (Structure of Transitive Group Actions on Sets) Let G ×

M → M be a transitive left action of a group G on a set M. Fix an arbitrary
point p M and let G p denote the isotropy group of p. Then G p G is a
subgroup and

is a well-defined G-equivariant bijection.

Proof Another easy exercise. □

This implies that every homogeneous G-space is of the form G⁄H for some
subgroup H G (not only as a set, but as the space of an action). We will show
in the following subsections that this result essentially still holds in the
continuous and smooth category.

3.8.2 Topological Groups and Homogeneous Spaces

Let G be a topological group and H G a subgroup with the subspace topology.
Consider the quotient space G⁄H with the subspace topology. According to
Proposition 3.8.2 we get a transitive group action

Proposition 3.8.4 (Φ Is a Continuous Action for Topological Groups)

Suppose G is a topological group and H G a subgroup. Then the transitive
group action
is continuous.

Proof Multiplication in G followed by projection onto G⁄H is continuous,

This implies, by the definition of the quotient topology on G⁄H, that the group
action

is continuous. □

According to Corollary 3.7.13 the space G⁄H is Hausdorff if and only if H is a

closed subset in G. We now consider the case of an arbitrary transitive
continuous group action.

Proposition 3.8.5 (Structure of Transitive Continuous Group Actions on

Topological Spaces) Let G × M → M be a transitive continuous left action of
a topological group G on a Hausdorff space M. Fix an arbitrary point p M
and let G p denote the isotropy group of p. Then G p G is a closed subgroup
and

is a well-defined continuous G-equivariant bijection between Hausdorff spaces.

If G is compact, then f is a homeomorphism.

Proof The isotropy group G p is closed in G by Proposition 3.2.9 (here we need

that M is Hausdorff). It is clear by Proposition 3.8.3 that f is a well-defined
equivariant bijection. It is also clear from the definition of the quotient topology
that f is continuous. The final statement follows, because a continuous bijection
from a compact space to a Hausdorff space is a homeomorphism. □

Remark 3.8.6 In general, if G is non-compact, the map f is not a

homeomorphism.

3.8.3 Lie Groups and Homogeneous Spaces

We now come to the case that we are most interested in: G is a Lie group and H
G a closed subgroup. By Corollary 3.7.35 the quotient G⁄H has a unique
structure of a smooth manifold such that π: G → G⁄H is a submersion. According
to Proposition 3.8.2 we get a transitive group action

Proposition 3.8.7 (Φ Is a Smooth Action for Lie Groups) Suppose G is a

Lie group and H G a closed subgroup. Then the transitive group action

is smooth.

Proof Multiplication in G followed by projection onto G⁄H is smooth,

By Lemma 3.7.5 the map

is smooth. □

We can now determine the structure of smooth manifolds that are homogeneous
under the action of a Lie group.

Theorem 3.8.8 (Structure of Transitive Smooth Group Actions on

Manifolds) Let G × M → M be a transitive smooth left action of a Lie group
G on a manifold M. Fix an arbitrary point p M and let G p denote the isotropy
group of p. Then G p G is a closed subgroup and

is a well-defined G-equivariant diffeomorphism between manifolds.

Proof It follows from Proposition 3.8.5 that f is well-defined, continuous,

bijective and G-equivariant. By Corollary 3.7.35 the quotient space G⁄G p is a
smooth manifold. It remains to show that f is smooth and a diffeomorphism.
The map f is smooth by Lemma 3.7.5, because the orbit map

is smooth. To show that f is a diffeomorphism it suffices to show that the

differential of f is an isomorphism at every point of G⁄G p . By G-equivariance of
f we have
Since left translations are diffeomorphisms of G⁄G p and M, the differential of
f is an isomorphism at every point of G⁄G p if and only if it is an isomorphism at
[e].
We first show that the differential of f is injective at [e]: let U G⁄G p be an
open neighbourhood of [e] and s: U → G a local section with π ∘ s = Id U , where
π: G → G⁄G p is the canonical projection. Without loss of generality s([e]) = e.
Then f = ϕ p ∘ s and

We also have

This shows that D [e] s is injective and its image is a complementary

subspace to the kernel of D e π, which is the Lie algebra of G p according to
Corollary 3.7.36. The kernel of D e ϕ p is also equal to according to
Proposition 3.2.10. This implies that the differential D [e] f is injective.
To show that D [e] f is surjective it suffices to show by G-equivariance that D
[a] f is surjective at some point [a] G⁄G p . This follows from the next lemma.
□

Lemma 3.8.9 Let f: X → Y be a surjective smooth map between manifolds.

Then there exists a point x X such that D x f is surjective.

Proof According to Sard’s Theorem A.1.27 there exists a regular value y Y

of f. Since f is surjective, there exists an x X with f(x) = y. Then x is a regular
point f, i.e. the differential D x f is surjective. □

Along the way we have shown the following more general result.

Corollary 3.8.10 (The Orbit Map Induces an Injective Immersion of G⁄G p

into M) Let G × M → M be a smooth left action of a Lie group G on a
manifold M, not necessarily transitive. Fix a point p M. Then

is an injective immersion of the manifold G⁄G p into M whose image is the orbit
of p. In particular, if the Lie group G is compact, then the orbit is an
embedded submanifold of M, diffeomorphic to G⁄G p .

Example 3.8.11 In Theorem 3.3.2 we saw that the standard representation of

O(n) on induces a transitive action of O(n) on the unit sphere S n−1 with
isotropy group of a point e 1 S n−1 isomorphic to the subgroup O(n − 1).
Theorem 3.8.8 then implies that the orbit map descends to a diffeomorphism

In a similar way we get diffeomorphisms

We also get diffeomorphisms

Note that the group structure on and is

not the direct product structure.

We can now prove Theorem 1.2.22 on the connected components of the classical
linear groups (the idea for this proof is from [34] and [142]).

Proof Let G be a Lie group and H G a closed connected subgroup.

According to Proposition 3.7.14 the quotient manifold G⁄H is connected if and
only if G is connected. We apply this inductively to the homogeneous spaces in
Example 3.8.11. We do the case of SO(n) explicitly, the other cases are left as an
exercise. It is clear that SO(1) = {1} is connected. Since S n−1 is connected for all
n ≥ 2, the diffeomorphism

shows that SO(n) is connected for all n ≥ 2. □

The following fact is sometimes useful:

Corollary 3.8.12 (Smooth Structure on Sets with a Transitive Lie Group

Action) Suppose that M is a set and G × M → M a transitive left action of a
Lie group G on M with closed isotropy group G p , for some p M. Then

is a bijection. The set M can be given a unique structure of a smooth manifold,

so that f becomes a diffeomorphism. If G is compact, then M is compact.

We conclude that in this situation we get the manifold structure on M for free,
without the (sometimes difficult) task of defining a topology and an atlas of
smoothly compatible charts for M.

3.9 Stiefel and Grassmann Manifolds

We discuss two examples of compact homogeneous spaces where the manifold
structure is defined by Corollary 3.8.12.

Example 3.9.1 (Stiefel Manifolds) Let and consider positive

integers k ≤ n. The Stiefel manifold is defined as the set of ordered k-
tuples of orthonormal vectors in with respect to the standard Euclidean
(Hermitian, symplectic) scalar product on ( , ) from Definition 1.2.9:

We consider the case in detail. The group O(n) acts on the set via

Since we can complete the vectors v 1, …, v k to an orthonormal basis of and

O(n) acts transitively on orthonormal bases, it follows that the action of O(n) on
is also transitive. The isotropy group of the point

is equal to

This holds, because if C O(n) satisfies C ⋅ p = p, then C is of the form

and CC T = E implies AA T = E and BA T = 0, hence A O(n − k) and B = 0. It

follows that the real Stiefel manifold admits the structure of a compact manifold
given by

In particular, has dimension

Similarly, it can be shown that

It follows that the complex and quaternionic Stiefel manifolds are connected
for all k ≤ n. For real Stiefel manifolds and k < n this follows from
Exercise 3.12.12.

Example 3.9.2 (Grassmann Manifolds) Let and consider non-

negative integers k ≤ n. The Grassmann manifold or Grassmannian is
defined as the set of k-dimensional vector subspaces in :

We consider the case . The group O(n) acts on the set via

This action is transitive, since we can choose a basis for U and the action of O(n)
on is transitive. The isotropy group of

is equal to

It follows that the real Grassmannian admits the structure of a compact

manifold given by

Note that there is a diffeomorphism

The dimension of is equal to

Similarly, it can be shown that

There are diffeomorphisms

for .

3.10 The Exceptional Lie Group G2

In this section we discuss the compact simple exceptional Lie group G2 . In
particular, we want to show that G2 has dimension 14. This is a nice application
of homogeneous spaces and Stiefel manifolds. We follow the paper [26] by
Robert Bryant.
Besides being mathematically interesting, the Lie group G2 plays an
important role in M-theory , a conjectured supersymmetric theory of quantum
gravity in 11 dimensions, which is related to the superstring theories in
dimension 10. If M-theory is a realistic theory of nature, with 4-dimensional
spacetime, 7 of the 11 dimensions have to be very small (compactified). The
vacuum or background of the theory is thus of the form , where is
Minkowski spacetime and K is a compact Riemannian 7-manifold. Moreover, for
the background to be a solution of the supergravity equations of motion,
preserving one supersymmetry in dimension 4 (the most interesting case from a
phenomenological point of view), the Riemannian metric on the 7-dimensional
compact manifold K has to have holonomy group equal to G2 (assuming that the
flux is set to zero). The first compact examples of Riemannian manifolds with
holonomy equal to G2 were constructed by Dominic Joyce.
A Riemannian metric has holonomy group G2 precisely if the 7-manifold
admits a certain type of 3-form that is parallel with respect to the Levi-Civita
connection. We will introduce the linear model of the 3-form on a vector space
of dimension seven and define G2 as its isotropy group.
3.10.1 Definition of the 3-Form ϕ and the Lie Group G2
We need some preparations: Let with the standard Euclidean scalar
product ⋅ , ⋅ and standard orthonormal basis {e j }. Let {ω i } denote the dual
basis of V , defined by

We use a shorthand notation for wedge products of the ω i . For example,

Definition 3.10.1 We define a 3-form ϕ Λ3V by:

Remark 3.10.2 The peculiar form of ϕ will be justified in Exercise 3.12.15.

Other choices, however, are possible and lead to equivalent descriptions of G2.

The group acts on the column vector space V on the left via the standard
representation. There is an induced representation on Λ k V defined by
(cf. Definition 2.1.23):

We think of this representation as a left action of on Λ k V .

Definition 3.10.3 We define as the isotropy group of the 3-form

ϕ:

This is a closed embedded Lie subgroup of .

3.10.2 G2 as a Compact Subgroup of SO(7)

Definition 3.10.4 For x V we denote by (contraction of ϕ with x) the 2-
form on V defined by

The following map is very useful in the study of the Lie group G2.
Definition 3.10.5 We set

Here are some properties of the map b.

Proposition 3.10.6 The map b is symmetric and bilinear. It is G 2 -

equivariant, i.e. we have

A calculation shows that

where vol = ω 1234567 is the standard volume form of V.

Proof It is clear that b is symmetric and bilinear. For g G2 and x, y, z V

we calculate

Therefore

and

The final property can be proved by a (tedious) direct calculation using the
explicit form of ϕ. Because of symmetry and bilinearity of b it suffices to show
that

We have
We then calculate all 28 wedge products of the form

with i ≤ j. For example,

The claim then follows from these calculations. □

Corollary 3.10.7 (G2 Is a Compact Subgroup of SO(7)) The following

identity holds

and

In particular, G2 preserves the standard scalar product and orientation on V and

is thus a compact embedded Lie subgroup of SO(7).

Proof For any we have

hence

By Proposition 3.10.6 this implies for all g G2

Therefore
Consider the matrix g T g. We have

hence

Calculating the determinant on both sides we get

hence

and

We get

and with detg = 1 it follows that G2 is a subgroup of SO(7). Since G2 is a closed

subgroup and SO(7) is compact, it follows that G2 is compact. □

3.10.3 An SU(2)-Subgroup of G2
Definition 3.10.8 Let P: V × V → V be the map defined by

Proposition 3.10.9 The map P is antisymmetric, bilinear and G2 -equivariant.

We have P(e 1, e 2) = e 3.

Proof The first two properties are clear. The third property follows because the
standard scalar product on V is G 2-invariant and ϕ is G 2-invariant. The final
claim follows immediately from the definition of ϕ. □

Consider the action

This is the restriction of the standard action of O(7) on the Stiefel manifold
 .

Definition 3.10.10 Let H G2 denote the isotropy group of the point

under this action.

Since P is G 2-equivariant and P(e 1, e 2) = e 3 we have He 3 = e 3. Therefore H is

the subgroup of G2 defined by

and the action of H restricts to an action on the orthogonal complement

Lemma 3.10.11 The Lie group H is isomorphic to the subgroup of SO(4),

acting on W and fixing the 2-forms

Proof This follows, because H G2 and G2 fixes the 3-form ϕ. □

Proposition 3.10.12 (H Is Isomorphic to SU(2)) The Lie group H is

isomorphic to the subgroup of SO(4), acting on W and fixing the complex
structure

and the complex volume form

Hence H is isomorphic to SU(2).

Proof Since

an element g SO(4) fixes ρ if and only if it fixes both β 2 and β 3. For any
vector v W we have

where denotes the vector dual to the 1-form with respect to the standard
scalar product on W. It follows that g SO(4) fixes J if and only if it fixes β 1.
□

3.10.4 The Dimension of G2

Corollary 3.10.13 (Upper Bound on the Dimension of G2) The action of G2
on the Stiefel manifold induces an injective immersion of G2⁄SU(2) into
. In particular,

with equality if and only if the action of G2 on is transitive.

Proof The first claim follows from Corollary 3.8.10. We have

according to the calculation in Example 3.9.1. Since dimSU(2) = 3 and the map

has injective differential, the second claim follows. The third claim follows since
in the case of equality the map f is a submersion, hence has open image, and the
image is closed, since G2 is compact (it can be shown that is connected,
cf. Exercise 3.12.12). □

Lemma 3.10.14 (Lower Bound on the Dimension of G2) The action of

on Λ 3 V induces an injective immersion

Hence dimG 2 ≥ 14, with equality if and only if the map h has open image.

Proof The first claim again follows from Corollary 3.8.10. The second claim
follows from

□
Collecting our results, we get the following theorem:

Theorem 3.10.15 (G2 Has Dimension 14) The Lie group G2 has dimension
14. It acts transitively on the Stiefel manifold . In particular, the standard
representation of G2 on is irreducible. Moreover, the -orbit of ϕ
in Λ 3 V is open.

Remark 3.10.16 (G2 Is a Simple Lie Group) A calculation of the homotopy

groups of G2, using the fibration

shows that

Hence G2 is connected, simply connected and simple, cf. Corollary 2.6.6. The
details of this calculation can be found in [26].

3.11 Godement’s Theorem on the Manifold Structure

of Quotient Spaces
In this section we want to prove Godement’s Theorem 3.7.10 . We continue to
follow [130] and [139]. Let R be an equivalence relation on a manifold M.
Suppose that R is a closed embedded submanifold of M × M and pr1 | R : R → M
a surjective submersion. By symmetry of equivalence relations it follows that pr2
| R : R → M is also a surjective submersion. We endow M⁄R with the quotient
topology.

3.11.1 Preliminary Facts

We want to prove two preliminary facts: we first show that the quotient M⁄R is
Hausdorff.

Lemma 3.11.1 (The Quotient Space Is Hausdorff) The canonical projection

π: M → M⁄R is open and M⁄R is Hausdorff.
Proof Suppose U M is open. We claim that

This holds because x π −1(π(U)) if and only if there exists a y U such that
(x, y) R. Since pr1 | R is a submersion and (M × U) ∩ R is open in R, the set π
−1(π(U)) is an open subset of M, hence π(U) is an open subset of M⁄R by the
definition of the quotient topology. This proves that π is an open map. The claim
about the Hausdorff property follows from Lemma 3.7.2, because R is by
assumption a closed subset of M × M. □

We denote the equivalence class of a point x M under the equivalence relation

R by [x]. We want to show that equivalence classes are embedded submanifolds
of M.

Lemma 3.11.2 (Equivalence Classes Are Embedded Submanifolds of M)

Every equivalence of R is a closed embedded submanifold of M of dimension
dimR − dimM.

Proof We can write

because

Since pr2 | R : R → M is a submersion, the subset K = (pr2 | R )−1({x}) is an

embedded submanifold of R of dimension dimR − dimM. However, K is
contained in M ×{ x} on which pr1 is a diffeomorphism onto M. Therefore [x] =
pr1(K) is an embedded submanifold of M of dimension dimR − dimM. □

3.11.2 The Slice Theorem

Our task is to show that the quotient space M⁄R has the structure of a smooth
manifold. To define charts for M⁄R we construct so-called slices for the
equivalence relation on open neighbourhoods for any point of M. In a second
step we will then construct slices for saturated open neighbourhoods, which are
the main tools needed to define the manifold structure on M⁄R.

Definition 3.11.3 Let U M be an open neighbourhood. Then a slice for the

intersection of the equivalence classes of R with U is a closed embedded
submanifold S U together with a surjective submersion q: U → S such that for
every x U the set [x] ∩ U intersects S precisely in the single point q(x).

Theorem 3.11.4 (Slice Theorem) Every point in M has an open

neighbourhood U M with a slice (S, q) for the intersection of the equivalence
classes of R with U.

To prove the theorem fix a M and let S ′ be any submanifold of M through a

of dimension dimM − dim[a] and transverse to the submanifold [a]. This means
that

We will show that we can find an open neighbourhood U of a in M such that S =

S ′ ∩ U is a slice.

Lemma 3.11.5 Consider

Then Z is a submanifold of R through (a, a) of dimension dimZ = dimM and pr1 |

Z : Z → M is a local diffeomorphism around (a, a).

Proof Since pr2 | R is a submersion, it is clear that Z is a submanifold of R with

Hence dimZ = dimM. We have

Since a S ′ and a ∼ a, it follows that (a, a) Z.

It remains to show that the differential of pr1 | Z in (a, a) is an isomorphism
onto T a M. We consider the following submanifolds of Z through (a, a):

The tangent spaces to these submanifolds are given by

These tangent spaces have zero intersection and their dimensions are dim[a]
and dimS ′ = dimM − dim[a] = dimZ − dim[a]. Hence

The image of T (a, a) Z under the differential of pr1 | Z is

hence the differential of pr1 | Z is surjective and thus an isomorphism. □

Note that

is a submersion. By Lemma 3.11.5 we can choose open neighbourhoods O and

U ′ of a M such that

is a diffeomorphism. Let s denote the inverse of this diffeomorphism and

Then q is a submersion of U ′ onto an open subset of S ′ ∩ O.

Our aim is to shrink U ′ to U so that S = S ′ ∩ U is a slice together with the
restriction of q. Note that

In particular, U ′ O.

Lemma 3.11.6 Let x S ′ ∩ U ′ . Then s(x) = (x, x) and q(x) = x. In particular,

if y U ′ and q(y) U ′ , then q(q(y)) = q(y).

Proof We have , hence . Thus

Moreover,

since s is the inverse of pr1 | Z∩(O×O). Since pr1 | Z∩(O×O) is injective, this implies
s(x) = (x, x) and thus q(x) = x.
Finally, if y U ′ and q(y) U ′ , then x = q(y) S ′ ∩ U ′ and the claim
follows. □

Lemma 3.11.7 Let

Then U and S together with the restriction of q to U satisfy the requirements of

Theorem 3.11.4 .

Proof Clearly U is an open neighbourhood of a in M, because a U ′ and a

S ′ , hence q(a) = a U ′ ∩ O by Lemma 3.11.6. We also have S U. Suppose
x U. Then x U ′ and q(x) U ′ ∩ O. Thus q(q(x)) = q(x) U ′ ∩ O and
therefore q(x) U by definition of U. But also q(x) S ′ by definition of q,
hence q(x) S. Therefore the restriction of q to U defines a submersion

If x S, then x S ′ ∩ U ′ and q(x) = x by Lemma 3.11.6. This implies that q is

surjective.
Finally, suppose that x U and y [x] ∩ S. Then

because U ′ O. Thus

hence y = q(x). This proves the final requirement for the slice (S, q). □

Definition 3.11.8 If V M is a subset, then we denote the restriction of R to V

by R V . As a subset of M × M we have R V = (V × V ) ∩ R. We denote by π V : V
→ V⁄R V the canonical projection.

Corollary 3.11.9 (Slice for Open Subset Defines Local Manifold Structure on
Quotient) Every point in M has an open neighbourhood U M such that U⁄R
U has the structure of a smooth manifold and π U : U → U⁄R U is a surjective
submersion.

Proof Let U M be an open subset with a slice (S, q). Then the map q: U → S
induces a bijection

We give U⁄R U the structure of a smooth manifold such that is a

diffeomorphism. Then is a surjective submersion. □

3.11.3 Slices for Saturated Neighbourhoods and Proof of

Godement’s Theorem
Definition 3.11.10 A subset V M is called saturated if

Equivalently, V is a union of equivalence classes. If U is an arbitrary subset of

M, then V = π −1(π(U)) is saturated.

We want to show that every point of M is contained in a saturated open

neighbourhood with a slice. This is the main fact that we need to prove that M⁄R
has the structure of a smooth manifold.
Corollary 3.11.11 (Slices for Saturated Open Subsets) Let U M be an
open subset with a slice (S, q) and V the saturated open subset V = π −1(π(U)).
Then there exists a surjective submersion q ′ : V → S so that (S, q ′ ) is a slice for
V.

Proof It is clear that U V. Let j: U ↪ V be the inclusion. We claim that there

is a well-defined map

and that this map is a bijection. The map is well-defined, because if x, y U are
equivalent, then they are equivalent in V. The map is also injective. Finally, the
map is surjective, because if x V, then there exists a y U with (x, y) R.
Using the bijection from the proof of Corollary 3.11.9, we get a
well-defined map q ′ : V → S:

The map q ′ has the following property: for x V, there exists a y U such
that [x] = [ y], i.e.

Then

This implies, since S U,

Hence [x] intersects S precisely in the point q ′ (x).

Since U V, the map q ′ is surjective. It remains to show that q ′ is a
submersion. We claim that there is a commutative diagram

where the arrows on the left, right and top are submersions. The arrow on the
right is a submersion, because (S, q) is a slice and the arrows on the top and on
the left are submersions, because pr1 | R , pr2 | R : R → M are submersions. To
show that the diagram is commutative, let (x, y) (M × U) ∩ R. Then x ∼ y and
x V. The statement then is

which we showed above. Lemma 3.7.5 then proves that q ′ is a submersion. □

Corollary 3.11.12 (Slice for Open Saturated Subset Defines Local Manifold
Structure on Quotient) Let V M be an open subset with a slice (S, q ′ ).
Then V⁄R V has the structure of a smooth manifold so that π V : V → V⁄R V is a
surjective submersion.

We can now finish the proof of Godement’s Theorem 3.7.10.

Proof We have shown that there exists a covering of M by open saturated sets
V i so that the open sets have the structure of a smooth
manifold with

being surjective submersions. Suppose V i ∩ V j ≠ . By Lemma 3.11.13 below

we have to show that the manifold structures on W i ∩ W j induced from W i and
W j are the same, i.e. the identity map between the open subsets W i ∩ W j W i
and W i ∩ W j W j is a diffeomorphism. Since V i and V j are saturated, we have

The manifold structure induced from V i and V j on V i ∩ V j are the same. Since π
is for each of these structures a submersion from V i ∩ V j onto W i ∩ W j , it
follows from Corollary 3.7.7 that the induced manifold structures on W i ∩ W j
are the same. It is then also clear that

is a surjective submersion. □

We used (a slight generalization of) the following lemma, whose proof is clear:

Lemma 3.11.13 Let X be a topological space, W 1, W 2 X open and

homeomorphisms onto open subsets U 1, U 2 of . Define the unique smooth
structure on W i such that ϕ i becomes a diffeomorphism, for i = 1, 2. Then the
change of charts

is a diffeomorphism if and only if

is a diffeomorphism.

3.12 Exercises for Chap. 3

3.12.1. Prove Proposition 3.2.2. Find an example of a left action

so that

does not define a right action of G on M.

3.12.2. Let M be a Hausdorff space with a continuous left action of a
topological group G. For a subset K G consider the fixed point set

Prove the following:

1. If K = {k} contains only one element, then M K is a closed subset of M.

2. M K is a closed subset of M for arbitrary subsets K G.

3.12.3. The Lie group G = SU(2) × U(1) acts on via

where Av denotes multiplication of the matrix A SU(2) with the column

vector . Let

where , v 0 ≠ 0.
1. Determine the isotropy subalgebra and the isotropy subgroup G p . Which
standard Lie group is G p isomorphic to?

2. Determine the orbit of p under the action of G. Which standard manifold

is diffeomorphic to?

In the electroweak gauge theory the Higgs field takes values in . The
vector p is known as a vacuum vector. The isotropy group G p is called the
unbroken subgroup.
3.12.4. We consider S 3 with the Hopf action:

We identify

Let s denote the stereographic projection of S 3 through the point (0, 1) S

3:

Let γ i : S 1 → S 3, for i = 1, 2, 3, denote the orbit maps of the points

on S 3 under the Hopf action. Consider the images

of these curves under the stereographic projection. Determine and sketch σ 1,

σ 2, σ 3 (for σ 2 it may be helpful to rotate the coordinate system, so that σ 2 lies
in a coordinate plane.) Show that σ 1 and σ 2 are circles and σ 3 is a line. The
circle σ 1 spans a flat disk in . Show that σ 2 intersects this disk transversely in
one point. This means that σ 1, σ 2 and hence γ 1, γ 2 are linked.

Remark It is possible to show that all orbits of the Hopf action on S 3 are linked
pairwise.

3.12.5. The aim of this exercise is to verify two propositions on fundamental

vector fields in a special case with a direct calculation. The standard
representation of the Lie group SU(2) on induces a left-action

We fix the vectors

1. Determine the fundamental vector fields on and show by direct

calculation that

without using Proposition 3.4.4.

2. Let

Calculate directly and compare with , where Z = Ad A τ 1, without

using Proposition 3.4.6.
3.12.6 (From [ 23 ]). Let G be a compact Lie group acting smoothly and
freely on a manifold M. Let π: M → M⁄G be the canonical projection.

1. Prove that for every smooth curve γ: I → M⁄G, defined on an interval I, there
exists a smooth lift with .

2. Suppose that M is connected and at least one of the orbits of G on M is

connected (e.g. G is connected). Prove that π maps the fundamental group
of M surjectively onto the fundamental group of M⁄G. In particular, if M is
simply connected, then M⁄G is simply connected.
3.12.7. Let G and H be topological groups and M and N topological spaces.
Suppose that G acts continuously on the right on M and H acts continuously on
the right on N. Let ϕ: G → H be a group homomorphism. Suppose that f: M → N
is ϕ -equivariant, i.e.

Prove the following:

1. If f is continuous, then f induces a continuous map f ϕ : M⁄G → N⁄H.

2. If ϕ is an isomorphism and f a homeomorphism, then f ϕ is a

homeomorphism.
3.12.8. Use Exercise 3.12.7 to prove the following facts about lens spaces:

1. There exists a homeomorphism L( p, q) → L( p, −q).

2. If , then there exists a homeomorphism L( p, q) → L( p, r).

Remark According to a theorem of Reidemeister there exists a

homeomorphism between lens spaces L( p, q 1) and L( p, q 2) only in these two
cases, their combination, or in the trivial case q 1 = q 2.

3.12.9.

1. Show that can be covered by two charts diffeomorphic to and that

is diffeomorphic to S 2.

2. Prove that is diffeomorphic to S 4.

3.12.10. Consider complex projective space . Show that there

is a transitive left action of SU(n + 1) on with isotropy group isomorphic to
U(n). Deduce that there is a diffeomorphism

3.12.11. Prove that there is a diffeomorphism .

3.12.12. Show that for k < n the real and complex Stiefel manifolds can be
written as homogeneous spaces
Deduce that for k < n the real Stiefel manifolds are connected.
3.12.13. Consider the half-plane

1. Show that the map

for

is well-defined and defines a left-action of on H.

2. Prove that this action is transitive and that the action defines a
diffeomorphism between H and .
3.12.14 (From [ 57 ]). According to Exercise 1.9.10 the group SO(2n) has a
subgroup isomorphic to U(n). We would like to identify the homogeneous space
SO(2n)⁄U(n).

1. Let

Show that the subgroup

of SO(2n) is isomorphic to U(n) (compare with Exercise 1.9.10).

2. Consider the set

This is the set of almost complex structures on , compatible with the

scalar product and the orientation. The group SO(2n) acts on by
conjugation
 Prove that this action is transitive.

3. Conclude that .

Remark It can be shown that and .

3.12.15. Let with standard scalar product ⋅ , ⋅ and let P: V × V → V

denote the antisymmetric, bilinear G2-equivariant map from Definition 3.10.8.

1. Let x, y V be arbitrary vectors. Show that there exists an element g G2

such that (at the same time)

with real coefficients x 1, y 1, y 2.

2. Use the first part of this exercise to prove the identity

3. Let

and define an -bilinear multiplication ⋅ on by

for all x, y V. Let (⋅ , ⋅ ) denote the scalar product on so that e 0, e 1, e 2,

…, e 7 are orthonormal, with associated norm | | ⋅ | |. Prove that

Hence is a real normed division algebra of dimension 8, known as the

octonions .
4. Prove that

5. For

with and x V define the conjugate

Show that

This implies that every non-zero octonion has a multiplicative inverse.

6. Calculate (e 1 ⋅ e 2) ⋅ e 4 and e 1 ⋅ (e 2 ⋅ e 4) and show that the octonions are

not associative.
3.12.16 (From [ 27 ]). We continue with the notation from Exercise 3.12.15.

1. Use the first part of Exercise 3.12.15 to prove the identity

2. Let x V be an arbitrary vector of norm 1 and V x the orthogonal

complement of in V. Then V x is a real 6-dimensional vector subspace of
V. Prove that multiplication of octonions defines a linear map

with J x 2 = −Id, i.e. a complex structure on V x .

3. Let S 6 be the unit sphere in V. Show that the restriction of the action of
SO(7) on S 6 to the subgroup G2 is transitive with isotropy group isomorphic
to SU(3). Conclude that S 6 can be realized as a homogeneous space

Remark Since the rank of the Lie group G2 is 2, it does not contain Lie
subgroups isomorphic to SU(n) for n ≥ 4.

3.12.17 (From [ 73 ]). We continue with the notation from Exercise 3.12.15. Our
aim is to show that G2 contains a certain Lie subgroup isomorphic to SO(4).

1. Consider on the representation of Sp(1) × Sp(1) given by

Show that this representation defines an embedding of

into SO(7) (compare with Exercises 1.9.20 and 1.9.21).

2. Identify V with via the embeddings

and

Prove that the embedding SO(4) ↪ SO(7) = SO(V ) above has image in G2
(for example, by showing that the Lie algebra maps to the Lie
algebra of G2).
3.12.18 (From [ 73 ]). We continue with the notation from Exercises 3.12.15
and 3.12.17. A 3-dimensional oriented real vector subspace U V is called
associative if the restriction ϕ | U is positive, i.e. a volume form, where ϕ denotes
the 3-form from the definition of the Lie group G2. Let G(ϕ) denote the set of all
associative subspaces of V.

1. Show that the action of G2 on V induces an action of G2 on G(ϕ).

2. Let U V be an associative subspace and x, y U orthonormal. Prove that

the vectors x, y, x ⋅ y span U. Show that the action of G2 on G(ϕ) is
transitive.

3. Show that the isotropy group H of U 0 = span(e 1, e 2, e 3) G(ϕ) contains

the subgroup SO(4) G2 from Exercise 3.12.17.
4. Let h H. Show that there exists an element k SO(4) such that

with g 2(1) = 1. Show that for we can write with multiplication of

octonions

Conclude that g 2 = Id, hence H = SO(4) and

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246
12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]
27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
 [MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
 Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)
76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/
108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)
123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
 [MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions. Chin.
Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_4
Chapter 4 Fibre Bundles
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

What is gauge theory? It is not an overstatement to say that gauge theory is

ultimately the theory of principal bundles and associated vector bundles.
Besides full gauge theories, it also proves beneficial in certain situations to study
the theory only involving principal bundles, sometimes called Yang–Mills theory
. In physics, an example of a full gauge theory would be quantum
chromodynamics (QCD) , the theory of quarks , gluons and their interactions,
while pure Yang–Mills theory would be a theory only of gluons, also called
gluodynamics . Even such a simplified theory is very interesting – the Clay
Millennium Prize Problem [37] on the mass gap , for instance, is a problem
concerning the spectrum of glueballs in pure quantum Yang–Mills theory.
With the background knowledge of Lie groups, Lie algebras, representations
and group actions, we will now study fibre bundles in general and more
specifically principal bundles, vector bundles and associated bundles, which
together form the core or the “stage” of gauge theories.
Fibre bundles can be thought of as twisted, non-trivial products between a
base manifold and a fibre manifold. Principal and vector bundles are fibre
bundles whose fibres are, respectively, Lie groups and vector spaces, so that the
bundle admits a special type of bundle atlas, preserving some of the additional
structure of the fibres.
The fundamental geometric object in a gauge theory is a principal bundle
over spacetime with structure group given by the gauge group. The fibres of a
principal bundle are sometimes thought of as an internal space at every
spacetime point, not belonging to spacetime itself. The gauge group acts at every
spacetime point on the internal space in a simply transitive way. Connections on
principal bundles, that we discuss in Chap. 5, correspond to gauge fields, whose
particle excitations in the associated quantum field theory are the gauge bosons
that transmit interactions. Matter fields in the Standard Model, like quarks and
leptons, or scalar fields, like the Higgs field, correspond to sections of vector
bundles associated to the principal bundle (and twisted by spinor bundles in the
case of fermions). The ultimate reason for the interaction between matter fields
and gauge fields is that both are related to the same principal bundle.
Fibre bundles are indispensable in gauge theory and physics in the situation
where spacetime, the base manifold, has a non-trivial topology. This happens, for
example, in string theory where spacetime is typically assumed to be a product
of Minkowski spacetime with a compact Riemannian manifold K. It also
happens if we compactify (Euclidean) spacetime to the 4-sphere S 4. In these
situations, fields on spacetime often cannot be described simply by a map to a
fixed vector space, but rather as sections of a non-trivial vector bundle.
Even in the case where the fibre bundles are trivial, for example, in the case
of principal bundles and vector bundles over contractible manifolds like ,
there is still a small, but important difference between a trivial fibre bundle and
the choice of an actual trivialization. We will see that this is similar to the
difference in special relativity between Minkowski spacetime and the choice of
an inertial system.
Fibre bundles are not only important in physics, but for a variety of reasons
also in differential geometry and differential topology: many non-trivial
manifolds can be constructed as (total spaces of) fibre bundles and numerous
structures on manifolds, such as vector fields, differential forms and metrics, are
defined using bundles. Mathematically, we are especially interested in the
construction of non-trivial fibre bundles (trivial bundles are just globally
products). We discuss the following methods that (potentially) yield non-trivial
bundles :
Mapping tori (Example 4.1.5) and the clutching construction (Sect. 4.6)
yield fibre bundles over the circle S 1 and higher-dimensional spheres S n .
Principal group actions define principal bundles (Sect. 4.2.2; specific
examples are the famous Hopf fibrations and principal bundles over
homogeneous spaces).
Actions of the structure group G of a principal bundle P → M on another
manifold F (the general fibre) yield associated fibre bundles (Sect. 4.7) over
M. In particular, all vector bundles can be obtained in this way.
The tangent bundle TM and frame bundle Fr(M) of smooth manifolds M are
 specific examples of vector and principal bundles.
In general, every fibre bundle can be constructed using a cocycle of
transition functions (Exercise 4.8.9).
This chapter, like the previous one, contains many definitions and concepts. I
hope that there are sufficiently many examples to illustrate the definitions and
balance the exposition. References for this chapter for fibre bundles in general
are [14, 84, 133] and [136] as well as [5, 25, 39, 74] and [78] for vector bundles
in particular.

4.1 General Fibre Bundles

4.1.1 Definition of Fibre Bundles
Before we begin with the definition of fibre bundles, we consider two very
general notions: suppose π: E → M is a surjective differentiable map between
smooth manifolds (occasionally we will consider the following notions even in
the case of a surjective map π: E → M between sets).

Definition 4.1.1

1. Let x M be an arbitrary point. The (non-empty) subset

is called the fibre of π over x.

2. For a subset U M we set

We can think of E U as the part of E “above” U. It is clear that E U is the

union of all fibres E x , where x U.

3. A differentiable map s: M → E such that

is called a (global) section of π. A differentiable map s: U → E, defined on

some open subset U M, satisfying

is called a local section.

Note that a differentiable map s: U → E is a (local) section of π: E → U if and
only if s(x) E x for all x U.
For a general surjective map, the fibres E x and E y over points x ≠ y M can
be very complicated and different, in particular, they may not be embedded
submanifolds of E and even when they are, they may not be diffeomorphic. The
simplest example where these properties do hold is a product E = M × F with π
given by the projection onto the first factor.
Fibre bundles are an important generalization of products E = M × F and
can be understood as twisted products. The fibres of a fibre bundle are still
embedded submanifolds and are all diffeomorphic. However, the fibration in
general is only locally trivial, i.e. locally a product, and not globally. We shall
see later in Corollary 4.2.9 and Corollary 4.5.12 that if the topology of M is
trivial (i.e. M is contractible), then certain types of fibre bundles over M (like
principal and vector bundles) are always globally trivial. If M has a non-trivial
topology (for example, if M is a sphere S n ), this may not be the case.
Consider, for instance, the Hopf action of S 1 = U(1) on S 3, introduced in
Definition 3.3.1. This is a free action, i.e. the orbit of every point in S 3 is an
embedded S 1 and the quotient space S 3⁄U(1) of this action is the smooth
manifold .

However, it is clear (e.g. by considering fundamental groups) that S 3

cannot be diffeomorphic to S 2 × S 1. We will see in Example 4.2.14 that S 3
really is the total space of a non-trivial S 1-bundle over S 2. We denote this
bundle by

or

This is the celebrated Hopf fibration . The total space S 3 is simply

connected even though the fibres S 1 are not. This is possible, because the
fibre bundle is globally non-trivial.

General fibre bundles are defined as follows.

Definition 4.1.2 Let E, F, M be manifolds and π: E → M a surjective
differentiable map. Then (E, π, M; F) is called a fibre bundle (or locally trivial
fibration or locally trivial bundle ) if the following holds: For every x M
there exists an open neighbourhood U M around x such that π restricted to E U
can be trivialized, i.e. there exists a diffeomorphism

such that

hence the following diagram commutes:

We also write

or

to denote a fibre bundle. We call

E the total space
M the base manifold
F the general fibre
π the projection
(U, ϕ U ) a local trivialization or bundle chart .
See Fig. 4.1.
Fig. 4.1 Fibre bundle

Remark 4.1.3 The classic references [133] and [81] use the term fibre bundle in
a more restrictive sense; see Remark 4.1.15.

It is easy to see, using a local trivialization (U, ϕ U ), that the fibre

is an embedded submanifold of the total space E for every x M and the map ϕ
Ux defined by

is a diffeomorphism between the fibre over x U and the general fibre.

Note that in a local trivialization the map

is a diffeomorphism and

is a submersion (its differential is everywhere surjective). This implies that

the projection π: E → M of a fibre bundle is always a submersion. The Regular
Value Theorem A.1.32 then shows again that the fibres E x = π −1(x) are
embedded submanifolds of E.

Example 4.1.4 (Trivial Bundle) Let M and F be arbitrary smooth manifolds and
E = M × F. Then π = pr1 defines a fibre bundle

This bundle is called trivial .

Example 4.1.5 (Mapping Torus) We discuss an example where the idea of a
fibre bundle as a “twisted product” becomes very apparent. Let F be a manifold
and ϕ: F → F a diffeomorphism. We construct a fibre bundle E ϕ as follows:
Take

modulo the equivalence relation defined by

The quotient is a fibre bundle over the circle S 1 with general

fibre F:

The bundle E ϕ is called the mapping torus with general fibre F and
monodromy ϕ. See Remark 4.6.4 for more details. We can think of the bundle E
ϕ as being obtained by gluing the two ends of F × [0, 1] together using the
diffeomorphism ϕ.
If ϕ is the identity, then the mapping torus is a trivial bundle, but if ϕ is not
the identity, the mapping torus may be non-trivial. For example, for the fibre F =
S 1 we can do the construction with ϕ the identity of S 1, in which case E ϕ is
diffeomorphic to the torus T 2, and with ϕ the reflection on , in
which case E ϕ is diffeomorphic to the Klein bottle. Since the Klein bottle is not
diffeomorphic to T 2, the second example is a non-trivial S 1-bundle over S 1.
The clutching construction that we discuss in Sect. 4.6 is a generalization of
the mapping torus construction which yields fibre bundles

over spheres of arbitrary dimension.

4.1.2 Bundle Maps

Definition 4.1.6 Let and be fibre bundles over the
manifold M. A bundle map or bundle morphism of these bundles is a smooth
map H: E → E′ such that
i.e. such that the following diagram commutes:

A bundle isomorphism is a bundle map which is a diffeomorphism. If such an

isomorphism exists, we write .

Remark 4.1.7 Note that a morphism H: E → E′ maps a point in the fibre of E

over x M to a point in the fibre of E′ over the same point x. A bundle map
therefore covers the identity of M. We could consider more general bundle maps
between bundles over different manifolds M and N that cover a given smooth
map f: M → N.

It is clear that a bundle isomorphism induces a diffeomorphism between the

fibres of E and E′ over any x M.

Definition 4.1.8 Fibre bundles isomorphic to a trivial bundle as in

Example 4.1.4 are also called trivial .

It is more difficult to construct non-trivial fibre bundles . The mapping tori

defined in Example 4.1.5 are for many choices of (F, ϕ) non-trivial bundles. We
will discuss other examples of (potentially) non-trivial bundles in Sect. 4.2.2 and
Sect. 4.6.

Remark 4.1.9 Let be a fibre bundle. The existence of a local

trivialization over U M then means that the restricted bundle

is isomorphic to the trivial bundle

This in hindsight justifies why fibre bundles are called locally trivial.

Isomorphic bundles have diffeomorphic general fibres. The converse is not true
in general: There may exist non-isomorphic bundles whose general fibres are
diffeomorphic. In particular, as we shall see later in detail, there exist bundles
not (globally) isomorphic to a trivial bundle.
We can characterize trivial bundles as follows:

Proposition 4.1.10 (Trivial Bundles and Projections onto the General

Fibre) Let be a fibre bundle. Then the bundle is isomorphic to a
trivial bundle if and only if there exists a smooth map τ: E → F such that the
restrictions

are diffeomorphisms for all x M.

Proof If the bundle is trivial

we can set τ = pr2.

Conversely, assume that a map τ: E → F exists which restricts to a
diffeomorphism on each fibre. Consider the map

Then H is a smooth with

The map H is bijective, because it maps E x bijectively onto F.

We have to show that H is a diffeomorphism. We claim that the differential
of H is an isomorphism for every point p E. Since the dimensions of E and M
× F agree (this follows from the existence of local trivializations for E), it
suffices to show that the differential is surjective for every p E. The details
are left as an exercise. □

4.1.3 Bundle Atlases

Definition 4.1.11 A bundle atlas for a fibre bundle

is an open covering {U i } i I of M together with bundle charts

Definition 4.1.12 Let {(U i , ϕ i )} i I be a bundle atlas for a fibre bundle

. If U i ∩ U j ≠ , we define the transition functions by
The transition functions are diffeomorphisms. These maps have a special
structure, because they preserve fibres: For every x U i ∩ U j we get a
diffeomorphism

The maps

into the group of diffeomorphisms of F are also called transition functions. See
Fig. 4.2.

Fig. 4.2 Transition functions

Lemma 4.1.13 (Cocycle Conditions) The transition functions {ϕ ji } i, j I

satisfy the following equations:

The third equation is called the cocycle condition .

Proof Follows immediately from the definitions. □

Remark 4.1.14 Exercise 4.8.9 shows that a bundle can be (re-)constructed from
its transition functions using a suitable quotient space. The three properties of
Lemma 4.1.13 ensure the existence of a certain equivalence relation, used in the
construction of this quotient space.

A bundle atlas is very similar to an atlas of charts for a manifold. One difference
is that in the case of charts for a manifold we demand that the images of the
charts are open sets in a Euclidean space . In the case of charts for a bundle
the images are of the form U × F. In both cases the transition functions are
smooth. In the case of a bundle atlas, the transition functions have an additional
special structure, because they preserve fibres.
We can compare the definitions of general manifolds and general fibre
bundles as in Table 4.1.
Table 4.1 Comparison between notions for manifolds and fibre bundles
Manifold Fibre bundle
Coordinate chart Bundle chart
Coordinate transformation Transition functions
Atlas Bundle atlas
Trivial manifold with only one chart: Trivial bundle with only one bundle chart: M × F

Non-trivial manifold needs at least two Non-trivial bundle needs at least two bundle charts (like a non-
charts (like S n ) trivial bundle over S n )

Remark 4.1.15 Some references, such as [133] and [81], use the term fibre
bundle more restrictively. If the topological definition in these books is
transferred to a smooth setting, the definition amounts to assuming that the
transition functions of a bundle atlas are smooth maps to a Lie group G, acting
smoothly as a transformation group on the fibre F, instead of maps to the full
diffeomorphism group Diff(F) of the fibre:

Equivalently, a fibre bundle is with this definition always an associated bundle in

the sense of Remark 4.7.8.

4.1.4 Pullback Bundle

We want to show that we can pull back a bundle via a map between the base
manifolds. Suppose

is a fibre bundle.

Lemma 4.1.16 (Restriction of Bundle over Submanifold in the Base) If W

M is an embedded submanifold, then the restriction

is a fibre bundle.

Proof Let {(U i , ϕ i )} i I be a bundle atlas for the fibre bundle F → E → M

with bundle charts

Then the sets V i = U i ∩ W form an open covering of W and

are bundle charts for the restriction of E over W. □

Suppose f: N → M is a differentiable map from some manifold N to M. We set

and

Theorem 4.1.17 (Pullback Bundles) The map π N with

is a fibre bundle over N, called the pullback of E under f.

Proof We have an obvious fibre bundle

 The graph

is an embedded submanifold of N × M. Therefore the restriction

is a fibre bundle by Lemma 4.1.16. Note that

Hence as a set

which defines a smooth structure on f E, and we have a fibre bundle

There exists a diffeomorphism

We can define a bundle over N using the projection τ ∘π N×M :

But

hence

This shows that

is a fibre bundle. □

Remark 4.1.18 Note that the pullback bundle f E has the same general fibre
F as the bundle E. The fibre of f E over a point x N is canonically
diffeomorphic to the fibre of E over f(x) M via the map

Remark 4.1.19 It is not difficult to show that the pull-back of a trivial bundle is
always trivial. The pull-back of a non-trivial bundle may be non-trivial or trivial,
depending on the situation; see Exercise 4.8.2.

4.1.5 Sections of Bundles

We want to study sections of fibre bundles. This is particularly simple in the case
of trivial bundles.

Definition 4.1.20 Let

be a fibre bundle. We denote the set of smooth global sections s: M → E by Γ(E)

and the set of smooth local sections s: U → E, for U M open, by Γ(U, E).

It is easy to see that for a trivial bundle E there is a 1-to-1 correspondence

between sections of E and maps from the base manifold M to the general fibre F.
This implies:

Corollary 4.1.21 (Existence of (Local) Sections)

1. Every trivial fibre bundle has smooth global sections (for example, under
the above correspondence, we could take constant maps from the base M to
the fibre F) .

2. Every fibre bundle has smooth local sections, since every fibre bundle is
locally trivial .

Note that non-trivial fibre bundles can, but do not need to have smooth global
sections (for example, vector bundles, to be discussed later, always have global
sections, but principal bundles in general do not). In particular, for a non-trivial
bundle, a map from the base manifold to the general fibre usually does not
define a section.

4.2 Principal Fibre Bundles

Principal fibre bundles are a combination of the concepts of fibre bundles and
group actions: they are fibre bundles which also have a Lie group action so that
both structures are compatible in a certain sense. Principal bundles together with
so-called connections play an important role in gauge theory. Generally
speaking, principal bundles are the primary place where Lie groups appear in
gauge theories (Lie groups also appear as global symmetry groups, like the
flavour or chiral symmetry in QCD; see Sect. 9.1).

4.2.1 Definition of Principal Bundles

We consider again the Hopf action of S 1 on S 3, introduced in Definition 3.3.

1, with quotient space equal to . If we accept for the moment that S 3
is the total space of an S 1-bundle over S 2,

(we will prove this in Example 4.2.14), then we can say the following: there
is an action of the Lie group S 1 on the total space S 3 of the bundle which
preserves the fibres and is simply transitive on them. In addition we will show
that there is a special type of bundle atlas for the Hopf fibration which is
compatible with this S 1-action.

This leads us to the following definition:

Definition 4.2.1 Let

be a fibre bundle with general fibre a Lie group G and a smooth action P × G →
P on the right. Then P is called a principal G -bundle if:

1. The action of G preserves the fibres of π and is simply transitive on

 them, i.e. the action restricts to

and the orbit map

is a bijection, for all x M, p P x.

2. There exists a bundle atlas of G -equivariant bundle charts

, satisfying

where on the right-hand side G acts on (x, a) U i × G via

We also call such an atlas a principal bundle atlas for P and the charts in a
principal bundle atlas principal bundle charts .
The group G is called the structure group of the principal bundle P.

There are two features that distinguish a principal bundle P → M from a

standard fibre bundle whose general fibre is a Lie group G:

1. there exists a right G-action on P, simply transitive on each fibre P x , for

x M;

2. the bundle P has a principal bundle atlas.

If P → M is a principal G-bundle and g G, then we denote as before by r g

the right translation

The fibre P x is a submanifold of the total space P for every x M and the
orbit map

is an embedding for all p P x , according to Corollary 3.8.10, because the

stabilizer G p = {e} is trivial.

Example 4.2.2 The trivial bundle

has the canonical structure of a principal G-bundle with G-action

and the principal bundle atlas consisting of only one bundle chart

Example 4.2.3 If is a principal bundle and f: N → M a smooth

map, then the pullback f P has the canonical structure of a principal G-bundle
over N (this is Exercise 4.8.4).

Transition functions for a principal bundle atlas have a special form:

Proposition 4.2.4 (Transition Functions of Principal Bundles) Let P → M

be a principal G-bundle and {(U i , ϕ i )} i I a principal bundle atlas for P. Then
the transition functions take values in the subgroup G of Diff(G),

where an element g G acts as a diffeomorphism on G through left

multiplication,

Proof For x U i ∩ U j we have a diffeomorphism

We set

Then by equivariance of the bundle charts

This implies the claim. □

The following criterion sometimes simplifies the task of showing that a group
action on a manifold P defines a principal bundle (we follow [14, Theorem 2.4]).

Theorem 4.2.5 (Principal Bundles Defined via Local Sections) Let G be

a Lie group and π: P → M a smooth surjective map between manifolds with a
smooth action P × G → P on the right. Then P is a principal G-bundle if and
only if the following holds:

1. The action of G preserves the fibres of π and is simply transitive on them.

2. There exists an open covering {U i } i I of M together with local

sections s i : U i → P of the map π.

Remark 4.2.6 Recall that we defined in Sect. 4.1.1 the notion of a section for
any smooth surjective map, not only for fibre bundles.

Proof Suppose that π: P → M is a principal bundle. Choose a principal bundle

atlas {(U i , ϕ i )} for P with

Then the following maps are local sections

where e G is the neutral element.

Conversely, suppose that an open covering {U i } i I with sections s i : U i
→ P is given. According to the following lemma these sections define charts in a
principal bundle atlas for P. □

Lemma 4.2.7 Let G be a Lie group and π: P → M a smooth surjective map

between manifolds with a smooth action P × G → P on the right. Suppose that
the action of G preserves the fibres of π and is simply transitive on them. Let s: U
→ P be a local section for π. Then

is a G-equivariant diffeomorphism.
Proof Let s: U → P be a local section of the surjective map π: P → M. We have
to show that

is a G-equivariant diffeomorphism. It is clear that t is smooth, because the local

section s is smooth and the G-action on P is smooth. The map t is also G-
equivariant by the definition of group actions and it is bijective: the reason is that
the map

is bijective for every fixed x U, since the G-action on P is simply transitive on

the fibres. The set P U = π −1(U) is an open subset of P. Since t is smooth and
surjective, Sard’s Theorem A.1.27 implies that

It remains to show that the differential of t is injective in each point (x, g) U×

G. Then t is a diffeomorphism.
The differential

is given according to Proposition 3.5.4 by

We set

The map

is an isomorphism, because the action of G is simply transitive on the fibre P

x , cf. Corollary 3.2.12. We consider the map

Note that s′ is also a local section of P over U, since

The chain rule shows that

This implies that D x s′ is injective and the image of D x s′ intersected with T s
′(x) P x ker D s′(x) π is zero. We conclude that D (x, g) t is injective. □

A proof of the following theorem can be found in [81, Chap. 4, Corollary 10.3].

Theorem 4.2.8 (Principal Bundles and Homotopy Equivalences) Let f: M

→ N be a smooth homotopy equivalence between manifolds and G a Lie group.
Then the pullback f is a bijection between isomorphism classes of principal G-
bundles over N and principal G-bundles over M.

In particular we get:

Corollary 4.2.9 (Principal Bundles over Contractible Manifolds Are

Trivial) If M is a contractible manifold and G a Lie group, then every
principal G-bundle over M is trivial . This holds, in particular, if for
some n.

4.2.2 Principal Bundles Defined by Principal Group

Actions
Recall from Definition 3.7.24 that a smooth right action of a Lie group G on a
manifold P is called principal if the action is free and the map

is closed. We want to show as an application of Theorem 4.2.5 that principal Lie

group actions define principal bundles.

Theorem 4.2.10 (Principal Lie Group Actions Define Principal Bundles)

Let Φ be a principal right action of a Lie group G on a manifold P. Then P⁄G is a
smooth manifold and

is a principal bundle with structure group G.

Proof According to Theorem 3.7.25 the topological space P⁄G has the unique
structure of a smooth manifold so that π: P → P⁄G is a submersion. In particular,
by Lemma 3.7.4, the projection π admits local sections
The claim then follows from Theorem 4.2.5. □

Corollary 4.2.11 (Free Actions by Compact Lie Groups Define Principal

Bundles) Let G be a compact Lie group acting freely on a smooth manifold P.
Then P⁄G is a smooth manifold and

is a principal G-bundle.

Proof This follows from Corollary 3.7.29. □

We can also prove the following converse to Theorem 4.2.10.

Theorem 4.2.12 (Principal Bundles Define Principal Actions) Let

be a principal G-bundle. Then the right action of G on P is principal.

Proof The G-action on P is free by the definition of principal bundles. If G is

compact, then the claim follows from Corollary 3.7.29. In the general case,
consider the map

We have to show that Ψ is closed.

Let A P × G be a closed subset and a sequence
converging to ( p, q) P × P. There exist uniquely determined g i G such
that q i = p i ⋅ g i , where ( p i , g i ) A and

We want to show that the sequence converges in G.

Let π( p) = x and U M be an open neighbourhood of x with a principal
bundle chart

There exists an integer N such that for all i ≥ N the p i are contained in P U .
Then we can write
with certain x i U and h i , h G. Since q i → q and x i → x, it follows that

for some h′ G. Since h i → h and h i g i → h′, it follows that the sequence

converges in G to

The set A is closed, hence ( p, g) A. We have q = p ⋅ g and we conclude

that ( p, q) is in Ψ(A). □

Corollary 4.2.13 Principal bundles with structure group G correspond

precisely to principal G-actions.

Example 4.2.14 (Hopf Fibration) Let

be a sphere of odd dimension. Consider the Lie group (unit

circle). It acts on the sphere S 2n+1 via

This is the Hopf action from Definition 3.3.1. The quotient S 2n+1⁄U(1) of this
action can be identified with the complex projective space . Corollary 4.2.11
implies that

is a principal S 1-bundle, called the Hopf fibration or Hopf bundle.

To give an alternative proof of this statement, we can also apply
Theorem 4.2.5 directly (we follow [14, Example 2.7]). We have to find an open
covering of together with local sections (the first condition in the theorem is
clearly satisfied, because the action of S 1 on S 2n+1 is free). We set
and define for i = 0, …, n

The subset U i is open in , since π is an open map by Lemma 3.7.11. We

also set

and

These are well-defined local sections for the canonical projection π:

since s i ([w]) is a complex multiple of w. Therefore we see again that

is a principal fibre bundle.
It is clear (considering fundamental groups, for example) that S 2n+1 is not
diffeomorphic to . The Hopf fibration is thus an example of a non-trivial
(principal) fibre bundle .
Similar arguments for the standard action of the Lie group on
lead to a Hopf fibration

over the quaternionic projective space (there is also a principal -

bundle over real projective space ). Special cases of this construction
are the Hopf fibrations (see Exercise 3.12.9)

and
We consider another class of examples of principal bundles. Let G be a Lie
group and H G a closed subgroup, acting smoothly on G by right translations:

According to Corollary 3.7.35 there is a (unique) smooth structure on the

quotient space G⁄H, so that π: G → G⁄H is a submersion.

Theorem 4.2.15 (The Canonical Principal Bundles over Homogeneous

Spaces) If G is a Lie group and H G a closed subgroup, then

is a principal bundle with structure group H.

Proof This follows from Theorem 4.2.10. We can also verify the conditions of
Theorem 4.2.5 directly. The first condition is clearly satisfied, because the action
of H on G is free. By Lemma 3.7.4 there exist smooth local sections

for the canonical projection π: G → G⁄H, where the open subsets U i G⁄H
cover G⁄H. This proves the claim. □

Example 4.2.16 (Principal Bundles over Homogeneous Spheres) From

Example 3.8.11 we get the following principal bundles over spheres:
In particular, we get the following principal sphere bundles over spheres:

From the examples in Sect. 3.9 we also get principal bundles over the Stiefel and
Grassmann manifolds, such as

and

and similarly for the complex and quaternionic Stiefel and Grassmann
manifolds. According to the results in Sect. 3.10.4 there is a principal bundle
and according to Exercise 3.12.16 there is a principal bundle

4.2.3 Bundle Morphisms, Reductions of the Structure

Group and Gauges
We define homomorphisms of principal bundles as follows:

Definition 4.2.17 Suppose and are principal bundles

over the same base manifold M and f: G → G′ is a Lie group homomorphism.
Then a bundle morphism between P and P′ is an f -equivariant smooth bundle
map H: P → P′, i.e.

and

Given the principal G′-bundle P′ and the homomorphism f: G → G′, the principal
G-bundle P together with the bundle morphism H: P → P′ is also known as an f -
reduction of P′.
If f: G → G′ is an embedding, then H is called a G -reduction of P′ and the
image of H is called a principal G -subbundle of P′. If G = G′, f = Id G and H is
a G-equivariant bundle isomorphism, then H is called a bundle isomorphism.
A principal G-bundle isomorphic to the trivial bundle in Example 4.2.2 is
also called trivial.

As before in the case of general bundles we could consider morphisms between

principal bundles over different base manifolds M and N that cover a smooth
map from M to N.
The following notion is especially relevant in gauge theory.

Definition 4.2.18 Let π: P → M be a principal bundle. A global gauge for

the principal bundle is a global section s: M → P. Similarly, a local gauge is a
 local section s: U → P defined on an open subset U M.

Any local gauge defines a local trivialization of a principal bundle:

Theorem 4.2.19 (Gauges Correspond to Trivializations) Let

be a principal G-bundle and s: U → P a local gauge. Then

is a G-equivariant diffeomorphism. In particular, if s: M → G is a global

gauge, then the principal bundle is trivial, with trivialization given by the inverse
of t:

Proof This follows from Lemma 4.2.7. □

Remark 4.2.20 Note that for this construction to work we need the G-action on
P. The result would not hold if we just had a fibre bundle with fibre G.

Remark 4.2.21 Theorem 4.2.19 has the following interpretation, see Table 4.2:
A local gauge defines a local trivialization of a principal G-bundle, i.e. an
identification . A choice of local gauge thus corresponds to the
choice of a local coordinate system for a principal bundle in the fibre direction.
This can be compared, in special relativity , to the choice of an inertial system
for Minkowski spacetime M, which defines an identification .

Table 4.2 Comparison between notions for special relativity and gauge theory

Manifold Trivialization Transformations

and invariance
Special Spacetime M via inertial system Lorentz
relativity
Gauge theory Principal bundle P → via choice of Gauge
M
gauge

Of course, different choices of gauges are possible, leading to different

trivializations of the principal bundle, just as different choices of inertial systems
lead to different identifications of spacetime with . The idea of gauge theory is
that physics should be independent of the choice of gauge. This can be compared
to the theory of relativity which says that physics is independent of the choice of
inertial system.
Note that, if we consider principal bundles over Minkowski spacetimes , it
does not matter for this discussion that principal bundles over Euclidean spaces
are always trivial by Corollary 4.2.9. What matters is the independence of the
actual choice of trivialization, i.e. the choice of (global) gauge. Even on a trivial
principal bundle there are non-trivial gauge transformations. This is very similar
to special relativity, where spacetime is trivial, i.e. isometric to with a
Minkowski metric , but what matters is the independence of the actual
trivialization, i.e. the choice of inertial system. Transformations between inertial
systems are called Lorentz transformations , transformations between (local)
gauges are called gauge transformations .

4.3 Formal Bundle Atlases

We briefly return to the case of general fibre bundles. We are sometimes in the
following situation: We have a manifold M, a set E and a surjective map π: E →
M. However, we do not a priori have a topology or the structure of a smooth
manifold on E. Under which circumstances can we define such structures, so that
π: E → M becomes a smooth fibre bundle?

Example 4.3.1 Let M be a smooth manifold of dimension n. The tangent space

T p M is an n-dimensional vector space for all p M. Let TM be the disjoint
union

with the obvious projection π: TM → M. How do we define the structure of a

smooth manifold on the set TM, such that TM becomes a fibre bundle over M,
with fibres given by T p M? We can also define for each tangent space T p M the
dual vector space T p M or the exterior algebra Λ k T p M. How do we
construct smooth fibre bundles that have these vector spaces as fibres?

The following notion is useful in this context (we follow [14, Sect. 2.1]).
Definition 4.3.2 Let M and F be manifolds, E a set and π: E → M a surjective
map.

1. Suppose U M is open and

is a bijection with

Then we call (U, ϕ U ) a formal bundle chart for E.

2. A family {(U i , ϕ i )} i I of formal bundle charts, where {U i } i I is an

open covering of M, is called a formal bundle atlas for E.

3. We call the charts in a formal bundle atlas {(U i , ϕ i )} i I smoothly

compatible if all transition functions

for U i ∩ U j ≠ , are smooth maps (i.e. diffeomorphisms).

We then have:

Theorem 4.3.3 (Formal Bundle Atlases Define Fibre Bundles) Let M and F
be manifolds, E a set and π: E → M a surjective map. Suppose that {(U i , ϕ i )} i
I is a formal bundle atlas for E of smoothly compatible charts. Then there
exists a unique topology and a unique structure of a smooth manifold on E such
that

is a smooth fibre bundle with smooth bundle atlas {(U i , ϕ i )} i I.

The proof consists of several steps. We first define a topology on E: consider the
bijections

We define a subset O E to be open if and only if

is open in U i × F for all i I.

Lemma 4.3.4 (The Topology on E Defined by a Formal Bundle Atlas) This
defines a topology on E which is Hausdorff and has a countable base. It is the
unique topology on E such that all formal bundle charts are
homeomorphisms.

Proof We first show that this defines a topology on E: it is clear that and E
are open. It is also easy to see that arbitrary unions and finite intersections of
open sets are open.
By definition the maps ϕ i are open. Suppose that and ϕ i (O) is open.
Then for all j I

It follows that O is open in E and that is a homeomorphism.

Since M and F are Hausdorff, it is not difficult to show that the topology on
E is Hausdorff, by considering for arbitrary points p, q E first the case π( p) ≠
π(q) with π( p) U i , π(q) U j and then the case π( p) = π(q) U i .
To show that the topology on E has a countable base we choose a countable
base {V j } j J for the topology of M and a countable base {W k } k K for the
topology of F. Without loss of generality we can assume that the family {U i } i
I is countable, without changing the topology of E. Let O E be an arbitrary
open set and p O a point. Then for some i and there exist j J
and k K such that

This shows that the countable family

of open sets of E forms a base.

The uniqueness statement for the topology of E is clear. □

We can now finish the proof of Theorem 4.3.3.

Proof To define a smooth structure on E, we first define the smooth structure

on such that the homeomorphism

is a diffeomorphism. Then this defines a smooth structure on E, because the

transition functions
are diffeomorphisms. This is the unique smooth structure on E so that π: E → M
is a smooth fibre bundle with general fibre F and {(U i , ϕ i )} i I is a smooth
bundle atlas. □

4.4 Frame Bundles

We want to apply Theorem 4.3.3 to define so-called frame bundles. Let M be a
smooth, n-dimensional manifold. For a point p M we define the set of all
bases of T p M

and define the disjoint union

There is a natural projection π: FrGL(M) → M and an action

given by

Theorem 4.4.1 (Frame Bundles) The projection π and the action of

define the structure of a principal -bundle

This bundle is called the frame bundle of the manifold M.

Proof We defined FrGL(M) so far only as a set. It is clear that the action of
preserves the fibres of π and is simply transitive on them. Let (U i , ψ i )
be a local manifold chart for M,

Then
is a local section for π. We have

We define the inverse of a formal bundle chart by

The transition functions are

with

These maps are smooth, because the transition functions ψ j ∘ψ i −1 are smooth.
This shows that the maps ϕ i are smoothly compatible formal bundle charts and
by Theorem 4.3.3 there exists a manifold structure on FrGL(M) such that π
becomes a fibre bundle with general fibre .
The -action is smooth (by considering the action in the bundle
charts) and the (inverse) bundle charts ϕ i −1 are -equivariant:

Therefore π: FrGL(M) → M is a principal -bundle over M. □

Remark 4.4.2 (Orthogonal Frame Bundles) If (M, g) is an n-dimensional

Riemannian manifold, we can define a principal O(n)-bundle

whose fibre over p M consists of the set of orthonormal bases in T p M. If M is

in addition oriented, then there is also a principal SO(n)-bundle

defined using oriented orthonormal bases. There are similar constructions of

orthonormal frame bundles for pseudo-Riemannian manifolds.

Remark 4.4.3 A frame, i.e. a basis of a tangent space to a manifold, is in

physics often called a vielbein , in particular in the case of an orthonormal frame
to a Lorentz manifold (the word “vielbein” is German and means “many-leg”. It
is a generalization of the word tetrad in the 4-dimensional case.)

Definition 4.4.4 Let G be a Lie group. A principal G-subbundle of the frame

bundle FrGL(M) of a smooth manifold M, i.e. a G-reduction of the frame bundle,
is called a G -structure on M.

In particular, a Riemannian metric on M n defines an O(n)-structure and, together

with an orientation, an SO(n)-structure on M.

4.5 Vector Bundles

We consider another class of fibre bundles, called vector bundles, that are
ubiquitous in differential geometry and gauge theory. The prototype of a vector
bundle is the tangent bundle TM of a smooth manifold M. Moreover, in physics,
matter fields in gauge theories are described classically by sections of vector
bundles. In addition to [14] we follow in this section [25, 74] and [78].

4.5.1 Definitions and Basic Concepts

Let be the field or .

Definition 4.5.1 A fibre bundle

is called a (real or complex) vector bundle of rank m if:

1. The general fibre V and every fibre E x , for x M, are m-dimensional

vector spaces over .

2. There exists a bundle atlas {(U i , ϕ i )} i I for E such that the induced maps

are vector space isomorphisms for all x U i . We call such an atlas a

vector bundle atlas for E and the charts in a vector bundle atlas vector
bundle charts . See Fig. 4.3.
Fig. 4.3 Vector bundle

A vector bundle of rank 1 is called a line bundle .

There are two features that distinguish a vector bundle E → M from a

standard fibre bundle whose general fibre is a vector space V:

1. the vector space structure on each fibre E x , for x M;

2. the bundle E has a vector bundle atlas.

The vector space structure on each fibre implies that we can add any two
sections of a vector bundle E and multiply sections with a scalar or a smooth
function on the base manifold M with values in .

Example 4.5.2 The simplest example of a vector bundle is the trivial bundle
, often denoted by . It has the canonical vector space structure on each
fibre , for p M, and the vector bundle atlas consisting of only one
bundle chart

Here is a more interesting example:

Example 4.5.3 (The Tangent Bundle of a Smooth Manifold) We want to show

that the tangent bundle of a smooth manifold is canonically a smooth real vector
bundle. Let M be a smooth manifold of dimension n. We define the set
with the canonical projection π: TM → M. We claim that TM has the structure of
a smooth real vector bundle of rank n over M: First, the general fibre and
each fibre T p M are n-dimensional real vector spaces. If

is a local manifold chart for M, then

is a formal bundle chart for TM. These formal bundle charts are smoothly
compatible, because

is a smooth map. By Theorem 4.3.3, π: TM → M has the structure of a smooth

fibre bundle with general fibre diffeomorphic to . Since the bundle charts (U i
, Ψ i ) are linear isomorphisms on each fibre, the bundle TM is a vector bundle of
rank n.

Remark 4.5.4 Note that sections of TM are the same as vector fields on M:

Transition functions for a vector bundle atlas have a special form:

Proposition 4.5.5 (Transition Functions of Vector Bundles) Let E → M be a

-vector bundle of rank m and {(U i , ϕ i )} i I a vector bundle atlas for E.
Then the transition functions take values in the subgroup of ,

The following definition applies only to real vector bundles.

Definition 4.5.6 A real vector bundle E → M of rank m is called orientable if

it admits a vector bundle atlas {(U i , ϕ i )} i I such that all transition functions
map to
where denotes the subgroup of invertible matrices with positive
determinant.

Clearly, if E → M is a complex vector bundle of rank m, then forgetting the

complex structure it defines an underlying real rank 2m vector bundle .
The bundle is always orientable, because the identification as real
vector spaces induces an embedding by Exercise 1.9.10.
There is a notion of a homomorphism between vector bundles over the same
manifold.

Definition 4.5.7 Let and be vector bundles over M

over the same field .

1. A smooth bundle map L: E → F, satisfying π F ∘ L = π E , is called a vector

bundle homomorphism if the restriction to a fibre

is a linear map for all x M. A vector bundle homomorphism which is

injective (surjective) on each fibre is called a vector bundle
monomorphism (epimorphism) . If L: E → E is a homomorphism, then L
is also called a vector bundle endomorphism .

2. A vector bundle isomorphism is a vector bundle homomorphism which is

a diffeomorphism of the total spaces and an isomorphism on each fibre. A
vector bundle is called trivial if it is isomorphic to the trivial bundle. If the
tangent bundle TM of a manifold M is trivial, then M is called parallelizable
.

Remark 4.5.8 According to Exercise 4.8.5, a vector bundle homomorphism

which is an isomorphism on each fibre is a vector bundle isomorphism.

As before, we could consider vector bundle homomorphisms between vector

bundles over different base manifolds M and N that cover a smooth map from M
to N. It is not difficult to prove with Remark 4.5.8 that a vector bundle E → M of
rank m is trivial if and only if it has m global sections
such that v 1(x), …, v m (x) form a basis of the fibre E x , for all x M.

Remark 4.5.9 (Sections of Vector Bundles) Note that (contrary to principal fibre
bundles) vector bundles always admit global sections : the section that is equal
to zero everywhere on M is a trivial example (the fibres of a vector bundle are
vector spaces, so there is a canonical element, namely 0. The fibres of a principal
bundle are only diffeomorphic to a Lie group, so the neutral element e is not a
canonical element in a fibre.) However, in the case of a vector bundle it is not
clear that there are sections without zeros, and even if this is the case, it is not
clear that there are m sections which form a basis in each fibre.

Example 4.5.10 (Parallelizable Spheres) We want to show that the spheres S 0,

S 1, S 3 and S 7 are parallelizable. This is trivial for S 0, which consists only of
two points. We consider S 1 as the unit sphere in . For x S 1, the vector
is orthogonal to x with respect to the standard Euclidean scalar product:

where is the standard Hermitian scalar product on . This implies that

is a trivialization of TS 1.
Similarly, we can consider S 3 as the unit sphere in . Then

is a trivialization of TS 3.
Finally, we consider S 7 as the unit sphere in the octonions . The octonions
are spanned by e 0, e 1, …, e 7, where

The map

is a trivialization of TS 7.
It is a deep theorem due to J.F. Adams [2] that S 0, S 1, S 3 and S 7 are the
only spheres which are parallelizable. This is related to the fact that division
algebras exist only in dimension 1, 2, 4 and 8. A proof using K-theory can be
found in [74]. See also Exercise 6.13.8.

A proof of the following theorem can be found in [5, 74] and [81].

Theorem 4.5.11 (Vector Bundles and Homotopy Equivalences) Let f: M →

N be a smooth homotopy equivalence between manifolds. Then the pullback f
is a bijection between isomorphism classes of vector bundles over N and vector
bundles over M of the same rank and over the same field .

In particular, we get:

Corollary 4.5.12 (Vector Bundles over Contractible Manifolds Are

Trivial) If M is a contractible manifold, then every vector bundle over M is
trivial . This holds, in particular, if for some n.

4.5.2 Linear Algebra Constructions for Vector Bundles

A useful fact is that we can construct new vector bundles from given ones by
applying linear algebra constructions fibrewise: suppose E, F are vector bundles
over M over the same field . Then there exist canonically defined vector
bundles

over M. If there also exists a complex conjugate vector bundle . The

fibres of these vector bundles are given by

and similarly in the other cases. This follows from Theorem 4.3.3, because local
vector bundle charts for E and F can be combined to yield smoothly compatible
formal vector bundle charts for the set E F, defining the structure of a smooth
vector bundle on E F → M. Similarly in the other cases.
Purely linear algebraic constructions, such as the direct sum and tensor
product of vector spaces, extend to smooth vector bundles and yield new
vector bundles with canonically defined smooth bundle structures.

Example 4.5.13 Consider the tangent bundle TM → M. Then there exist

canonically associated vector bundles T M and Λ k T M over M. Sections of
Λ kT M are k-forms on M:

More generally, for a vector bundle E → M, sections of the bundle Λ k T M

E are k-forms on M with values in E, i.e. elements of Ω k (M, E): If ω Ω k
(M, E), then at a point x M

is multilinear and alternating. This generalizes the notion of vector space-valued

forms in Sect. 3.5.1 to forms which have values in a vector bundle. One
sometimes calls Λ k T M E the bundle of k -forms over M twisted with E .

We want to define the concept of vector subbundle (following [25]):

Definition 4.5.14 Let π: E → M be a -vector bundle of rank m. A subset F

E is called a vector subbundle of rank k if each point p M has an open
neighbourhood U together with a vector bundle chart (U, ϕ) of E such that

where is the vector subspace . It follows that F is an embedded

submanifold of E and π | F : F → M has the canonical structure of a -vector
bundle of rank k over M.

Example 4.5.15 (Normal Bundle of Spheres) For n ≠ 0, 1, 3, 7 the sphere S n

does not have a trivial tangent bundle according to Adams’ Theorem mentioned
in Example 4.5.10. However, the normal bundle ν(S n ) of S n in is trivial
for any n ≥ 0: The normal bundle is defined as

with projection onto the first factor. It is clear that the normal bundle is a real
line bundle. The following map is a trivialization of ν(S n ):

Note that

We conclude that the sum of a non-trivial vector bundle (the tangent bundle to
the sphere) and a trivial vector bundle (the normal bundle) can be trivial. One
says that the tangent bundle of the sphere S n is stably trivial : It becomes trivial
after taking the direct sum with a trivial bundle (here a trivial line bundle). Both
TS n and ν(S n ) are vector subbundles of the trivial bundle . Note that this
also means that a trivial vector bundle can have non-trivial subbundles.

Definition 4.5.16 Let E → M be a -vector bundle over M. A (Euclidean or

Hermitian) bundle metric is a metric on each fibre E x that varies smoothly
with x M. More precisely, it is a section

or

which defines in each point x M a non-degenerate symmetric ( ) or

Hermitian ( ) form

Proposition 4.5.17 (Existence of Bundle Metrics) Every -vector bundle

over a manifold M admits a positive definite bundle metric.

Proof This follows by a partition of unity argument, because a convex

combination of positive definite metrics on a vector space is still a positive
definite metric. □

For associated vector bundles we will give a more explicit construction of

bundle metrics in Proposition 4.7.12.

Example 4.5.18 The tangent bundle TM of any submanifold has a

bundle metric induced from the standard Euclidean scalar product on . In
particular, TS n−1 has a canonical bundle metric.

Proposition 4.5.19 (Orthogonal Complement of a Vector Subbundle) Let E

→ M be a vector bundle with a positive definite bundle metric and F E a
vector subbundle. Then the orthogonal complement F is a vector subbundle of
E and F F is isomorphic to E.

Proof This is Exercise 4.8.16. □

The following is clear.

Proposition 4.5.20 (Transition Functions of Vector Bundles with a Metric)

Let E → M be a -vector bundle of rank m with a positive definite bundle
metric. Choosing local trivializations given by orthonormal bases it follows that
there exists a vector bundle atlas {(U i , ϕ i )} i I with transition functions of the
form

or

If the bundle is real and orientable, then we can find a vector bundle atlas
such that

If E → M is a real (complex) vector bundle of rank m with a positive definite

bundle metric, then the set consisting of all vectors of length 1 in each fibre
forms the unit sphere bundle S(E) → M, which is a smooth S m−1-bundle (S 2m
−1-bundle) over M.

4.6 The Clutching Construction

We want to describe a construction that yields (all) vector bundles over spheres S
n . The idea of this so-called clutching construction is to glue together trivial

vector bundles over the northern and southern hemisphere of S n along the
equator (we follow [74]).
Let S n be the unit sphere in , where n ≥ 1. We define the north and south
pole

and the open sets

Both U + and U − are diffeomorphic to via the stereographic projection;

cf. Example A.1.8. Let f be any smooth map
where we think of S n−1 S n as the equator of S n and . Such a map
is called a clutching function. We write

Let p denote the following retraction of the intersection U + ∩ U − onto the

equator:

Here

is the Euclidean norm. Note that this map is well-defined, because x ≠ 0 on U

+ ∩ U −. We use the retraction to extend the clutching function to a smooth map
on U + ∩ U −:

Definition 4.6.1 Let be the quotient set of the disjoint union

by identifying

with

Theorem 4.6.2 (Vector Bundle over a Sphere Defined by a Clutching

Function) Via the projection

the set E f has a canonical structure of a -vector bundle of rank k over the
sphere S n :
Proof (See also Exercise 4.8.9.) Note that the map π is well-defined on the
quotient set E f and surjective onto S n . Let σ denote the quotient map

The map σ decomposes into injective maps σ ± on . For

define

Then

The maps

are well-defined formal bundle charts. We want to show that these formal
bundle charts are smoothly compatible: We calculate

which is a smooth map. It follows from Theorem 4.3.3 that π: E f → S n has

the structure of a fibre bundle. Since the bundle charts ϕ +, ϕ − are linear
isomorphisms on each fibre, it follows that π: E f → S n is a vector bundle with
general fibre .□

The following can be shown, see [5] or [74]:

Theorem 4.6.3 (Vector Bundles over Spheres and Homotopy Classes of

Clutching Functions)

1. Every complex vector bundle over S n of rank k is isomorphic to a bundle E f

for a certain clutching function f: S n−1 → U(k), unique up to homotopy.
2. Similarly, every orientable real vector bundle over S n of rank k is
isomorphic to a bundle E f for a certain clutching function f: S n−1 → SO(k),
unique up to homotopy.

Every vector bundle over a sphere can be constructed using a clutching function,
because by Corollary 4.5.12 every vector bundle over S n is trivial over U + and
U −. Given an arbitrary -vector bundle E → S n of rank k we obtain an
associated clutching function as follows:
Let E ± denote the restrictions of E to U ±. Choose vector bundle
trivializations

Consider

This map is a linear isomorphism on each fibre and defines the clutching
function

The clutching functions in Theorem 4.6.3 can be taken to have image in U(k)
or SO(k), because there are deformation retractions

Complex and real orientable vector bundles over S n are therefore essentially
classified by the homotopy groups π n−1(U(k)) and π n−1(SO(k)).

Remark 4.6.4 (Clutching Construction for Arbitrary Fibres) Let F be a smooth

manifold and

a “smooth” map, again called a clutching function (the precise formulation in the
general case is not completely trivial, because we did not define a smooth
structure on the diffeomorphism group Diff(F)). A similar construction to the
one above yields a fibre bundle
over S n with general fibre F. The mapping torus construction in Example 4.1.5
can be seen as a special case of the clutching construction for n = 1: If ϕ: F → F
is the monodromy of the mapping torus, we choose

Example 4.6.5 (Exotic 7-Spheres) Another very nice application appears in

Milnor’s classic paper [94], where certain exotic7-spheres (homeomorphic
but not diffeomorphic to S 7) are defined as S 3-bundles over S 4, using the
clutching construction with clutching function

given by

Here u, v S 3 = Sp(1) and . For certain values of the integers h and j,

the S 3-bundle over S 4 determined by the clutching function f hj is an exotic 7-
sphere.
Milnor’s paper started the field known as differential topology and led to
an extensive investigation of exotic spheres of arbitrary dimension. The study
of the smooth topology of general 4-manifolds, using Donaldson theory and
later Seiberg–Witten theory , also belongs to the field of differential topology.

4.7 Associated Vector Bundles

In Chap. 2 we studied the theory of Lie group representations. We now want
to combine this theory with the theory of principal bundles from the present
chapter.

We said before that principal bundles are the place where Lie groups
appear in gauge theories. Associated vector bundles, which we discuss in this
 section, are precisely the place where representations on vector spaces are
built into gauge theories. We can summarize this in the following diagram:

The third row will be explained in Chap. 5.

For example, in the Standard Model, one generation of fermions is described

by associated complex vector bundles of rank 8 for left-handed fermions and
rank 7 for right-handed fermions, associated to representations of the gauge
group SU(3) × SU(2) × U(1). Taking particles and antiparticles together we get
two associated complex vector bundles of rank 15 (right-handed and left-
handed) which are related by complex conjugation. The complete fermionic
content of the Standard Model is described by the direct sum of three copies of
these vector bundles (a complex vector bundle of rank 90), corresponding to the
three generations. These constructions will be described in detail in Sect. 8.5.1.

4.7.1 Basic Concepts

As an introduction, consider again the Hopf fibration

We are interested in the complex representations of S 1 on with winding

number :

Our aim is to define an associated bundle

This will be a complex line bundle over S 2

 whose transition functions are given by the transition functions of the Hopf
fibration composed with the group homomorphism ρ k .

The general definition is the following: Let denote the field or . We

then associate to each principal G-bundle

and each representation

of the structure group G on a -vector space V of dimension k a vector

bundle E → M with fibres isomorphic to V.

Lemma 4.7.1 Let P be a principal G-bundle and ρ a representation of the Lie

group G on a -vector space V. Then the map

defines a free principal right action of the Lie group G on the product manifold
P × V. In particular, the quotient space E = (P × V )⁄G is a smooth manifold such
that the projection P × V → E is a submersion.

Proof It is clear that

is a right action, which is free since the action of G on P is free. If G is compact,

then the claim follows from Corollary 3.7.29. In the general case, the action is
principal by an argument very similar to the one in the proof of Theorem 4.2.12.
□

Theorem 4.7.2 (Associated Vector Bundle Constructed as a Quotient) Let

P be a principal G-bundle and ρ a representation of the Lie group G on a -
vector space V. Then the quotient space E = (P × V )⁄G has the structure of a -
vector bundle over M, with projection
and fibres

isomorphic to V. The vector space structure on the fibre E x over x M is

defined by

where π P ( p) = x.

Proof It is clear that π E is well-defined and that V is isomorphic to the fibres E

x via v ↦ [ p x , v] with a fixed p x P x . We need to find a vector bundle atlas
for E. Let (U, ϕ U ) be a bundle chart for the principal bundle P:

We set

Since P × V → E is a submersion, the map ψ U is smooth. It is a diffeomorphism

with smooth inverse

Its restriction to each fibre E x is a linear isomorphism to the vector space V.

Thus ψ U defines a chart in a vector bundle atlas for E. □

Definition 4.7.3 The vector bundle

is called the vector bundle associated to the principal bundle P and the
representation ρ on V:

The group G (or its image ρ(G) GL(V )) is known as the structure group
of E.
Remark 4.7.4 Note that in the definition of the vector space structure on the
fibres E x ,

we have to choose both representatives with the same point p in the fibre of P
over x M.

Example 4.7.5 For every principal G-bundle P → M and every vector space V,
the vector bundle associated to the trivial homomorphism

is a trivial vector bundle. See Exercise 4.8.20.

It is useful in applications to have a suitable description of local sections of an

associated vector bundle.

Proposition 4.7.6 (Local Sections of Associated Vector Bundles) Let P be a

principal bundle and E = P × ρ V an associated vector bundle. Let s: U → P be a
local gauge. Then there is a 1-to-1 relation between smooth sections τ: U → E
and smooth maps f: U → V, given by

In particular, the local gauge defines a preferred isomorphism between V and

every fibre E x over x U.

Proof If f: U → V is a smooth map, then

is smooth and hence τ: U → E is smooth. The map τ is a section, because

Conversely, let τ: U → E be a smooth section. Since E x = (P x × V )⁄G and the

action of G on P x is simply transitive, there is a unique f(x) V such that

We have to show that f: U → V is smooth: Define a bundle chart ϕ U of the

principal bundle using the section s:
Then with the notation in the proof of Theorem 4.7.2 we have β U (s(x)) = e
and

Since ψ U and τ are smooth, it follows that f is smooth. □

Matter fields in physics are described by smooth sections of vector bundles E

associated to principal bundles P via representations of the gauge group G on
a vector space V (in the case of fermions the associated bundle E is twisted in
addition with a spinor bundle S, i.e. the bundle is S E). It follows that,
given a local gauge of the gauge bundle P, the section in E corresponds to a
unique local map from spacetime into the vector space V.

In particular, since principal bundles on are trivial by Corollary 4.2.9,

we can describe matter fields on a spacetime diffeomorphic to by unique
maps from into a vector space, once a global gauge for the principal
bundle has been chosen. A (local) trivialization of the gauge bundle thus
determines a unique (local) trivialization of all associated vector bundles.

Definition 4.7.7 Let E = P × ρ V be an associated vector bundle. If the

representation

is non-trivial, then the sections of E are called charged .

This term will be explained in more detail in Sect. 5.9.

Remark 4.7.8 (Associated Fibre Bundles with Arbitrary Fibres) Given a

principal bundle P → M with structure group G, a manifold F and a smooth left
action

a similar construction using the quotient

under the G-action

yields an associated fibre bundle

with structure group given by the image of G in the diffeomorphism group

Diff(F), determined by the action Ψ.

Example 4.7.9 (Flat Bundles) Here is an example of the construction in

Remark 4.7.8. Let M and F be manifolds and

a group homomorphism. This defines an action of the (discrete) group π 1(M) on

F. The universal covering

can be considered as a principal bundle with discrete structure group π 1(M). The
associated fibre bundle

is called the flat bundle with holonomy ψ. In the case of M = S 1 this yields
again the mapping torus from Example 4.1.5. More generally, for M = T n , a
collection of n pairwise commuting diffeomorphisms

defines a flat bundle

4.7.2 Adapted Bundle Atlases for Associated Vector

Bundles
We discuss a specific type of bundle atlas for associated vector bundles. Let P →
M be a principal G-bundle and E = P × ρ V an associated vector bundle, where
ρ: G → GL(V ) is a representation. We choose a principal bundle atlas {(U i , ϕ i
)} i I for P, determined by local gauges s i : U i → P.

Definition 4.7.10 The principal bundle atlas for P defines an adapted bundle
atlas for E with local trivializations

whose inverses are given by

Proposition 4.7.11 (Adapted Bundle Atlases and the Structure Group)

Suppose the transition functions of the principal bundle charts for P are given by

Then the transition functions for the adapted bundle atlas for E are

The transition functions of E thus have image in the subgroup ρ(G) GL(V ),
where G is the structure group of P.

Proof We have

This implies

Therefore

4.7.3 Bundle Metrics on Associated Vector Bundles

It is often important to consider bundle metrics on an associated vector bundle.
We can construct such metrics as follows: let π P : P → M be a principal bundle
with structure group G, ρ: G → GL(V ) a representation and E → M the
associated vector bundle E = P × ρ V.

Proposition 4.7.12 (Bundle Metrics on Associated Vector Bundles from G-

Invariant Scalar Products) Suppose that ⋅ , ⋅ V is a G-invariant scalar
product on V. Then the bundle metric ⋅ , ⋅ E on the associated vector bundle E
given by

for arbitrary p P x , is well-defined.

Proof This is an easy calculation choosing two different representatives for the
vectors in the fibre E x . □

4.7.4 Examples
Example 4.7.13 (From Vector Bundles to Principal Bundles and Back) We
claim that every vector bundle has the structure of an associated vector bundle
for some principal bundle. We first consider the tangent bundle TM: Let M be an
n-dimensional smooth manifold and consider the frame bundle

Let

be the standard representation, given by matrix multiplication from the left on

column vectors. Then there exists an isomorphism of vector bundles

An isomorphism is given by

It is easy to check that the map H is well-defined and a bundle isomorphism.

Choosing a Riemannian metric g on M we can define the orthonormal frame
bundle FrO(M). Using the standard representation ρ O of O(n) on we get
another vector bundle isomorphism

Similarly, every real vector bundle E of rank n is associated to a principal

-bundle (and a principal O(n)-bundle), defined using frames in the fibres
of E. If E is orientable, it is associated to a principal SO(n)-bundle. Similar
statements hold for complex vector bundles with principal - and U(n)-
bundles.

We get:

Proposition 4.7.14 Let E → M be a real or complex vector bundle. Then E is

associated to some principal O(n)- or U(n)-bundle P → M.

In particular, the vector bundles over spheres that we defined in Sect. 4.6 using
the clutching construction are associated vector bundles. Note that the structure
as an associated vector bundle is not unique: as we saw above in the case of the
frame bundle, the same vector bundle can be associated to principal bundles with
different Lie groups.
We can use our constructions of principal bundles over spheres, projective
spaces, and Stiefel and Grassmann manifolds to define associated vector bundles
over those manifolds.

Example 4.7.15 Recall the principal bundle

from Example 4.2.16. Then any representation of SO(n − 1) on a real or complex

vector space defines an associated vector bundle over the sphere S n−1. A similar
construction works for any of the other principal bundles over spheres given in
Example 4.2.16. Alternatively, these bundles can also be realized (up to
isomorphism) by the clutching construction.
The construction also applies to the principal bundles over Stiefel and
Grassmann manifolds, like

and

These examples can be generalized: start with any smooth homogeneous

space G⁄H and consider the canonical principal bundle
according to Theorem 4.2.15. Then representations of H define associated
vector bundles over G⁄H, known as homogeneous vector bundles.

Example 4.7.16 Let

be the Hopf fibration . We want to study complex line bundles associated to this
principal S 1-bundle. For consider the homomorphism

of winding number k. Then the associated bundle

is a complex line bundle. The bundle γ 0 is trivial and γ k is isomorphic to

and

See Exercise 4.8.21.

Similarly, using representations of we can define vector bundles
associated to the quaternionic Hopf fibration

Example 4.7.17 (Adjoint Bundle) An important general example of an

associated vector bundle is the following: let

be a principal bundle with structure group G. Consider the adjoint

 representation

Then the associated vector bundle

is called the adjoint bundle . Its general fibre is isomorphic to the vector
space underlying the Lie algebra :

4.8 Exercises for Chap. 4

4.8.1 The Möbius strip can be defined as the submanifold

The projection π: M → S 1 is defined as π = pr1 | M .

1. Show that π: M → S 1 is a fibre bundle with general fibre [−1, 1] (here we

consider a small generalization of the notion of a fibre bundle to manifolds
with boundary).

2. Prove that the boundary ∂M is connected and that the bundle π is not trivial.

3. Prove that the image of any smooth section s: S 1 → M intersects the zero
section z: S 1 → M, z(α) = (α, 0).
Hint: Note that the map S 1 → S 1, e iϕ ↦ e iϕ⁄2 is not well-defined.
4.8.2 Let π: M → S 1 denote the Möbius strip from Exercise 4.8.1 and
consider the map f n : S 1 → S 1, f n (z) = z n for .

1. Show that the pull-back bundle f n M is isomorphic to the bundle M n → S

1 defined by

(with projection onto the first factor).

2. Determine those for which f n M is trivial and those for which it is
non-trivial.
4.8.3 (Fibre Sum) Suppose that F → E → M and F′ → E′ → M′ are two
fibre bundles over n-dimensional manifolds M and M′. Let D and D′ be
embedded open n-discs in M and M′ together with trivializations F × D and F′ ×
D′ of the fibrations over D and D′. We assume that F and F′ are diffeomorphic
and choose a diffeomorphism

We write D and D′ minus the centre 0 as S n−1 × (0, 1) and fix a

diffeomorphism r from (0, 1) to (0, 1) which reverses orientation. Let τ: S n−1 →
S n−1 be the diffeomorphism which reverses the sign of one of the coordinates on
. Consider the diffeomorphism

The fibre sum E# ψ E′ is defined by gluing together the manifolds M∖F and
M′∖F′ along the diffeomorphism ψ. Prove that E# ψ E′ is a smooth fibre bundle
over the connected sum M#M′ with general fibre F.
4.8.4 Let be a principal bundle and f: N → M a smooth map
between manifolds. Prove that the pullback f P has the canonical structure of a
principal G-bundle over N.
4.8.5

1. Let and be fibre bundles and H: E → E′ a bundle

morphism. Suppose that H maps every fibre of E diffeomorphically onto a
fibre of E′. Show that H is a diffeomorphism and hence a bundle
isomorphism.

2. Let and be principal bundles and f: G → G′ a Lie

group isomorphism. Show that every f-equivariant bundle morphism H: P →
P′ is a diffeomorphism.
4.8.6 (From [ 14 ]) We consider the Hopf bundle
The total space S 3 of this bundle admits two different S 1-actions: The
standard action

and the reversed action

Both actions endow the same fibre bundle with the structure of
a principal bundle. Prove that these principal bundles are not isomorphic as
principal bundles.
4.8.7 Recall the definition of lens spaces from Example 3.7.33. Show that the
lens space L( p, 1) is the total space of a principal fibre bundle over S 2 with
structure group S 1.
4.8.8 Show that there is a canonical free O(k)-action on the Stiefel manifold
and that this defines a principal O(k)-bundle

4.8.9 We want to discuss another way to construct fibre bundles. Let M, F be

smooth manifolds and {U i } i I an open covering of M together with
diffeomorphisms

whenever U i ∩ U j ≠ , satisfying

We also write ϕ ji (x) = ϕ ji (x, −) for x U i ∩ U j . Let be the disjoint

union

1. Show that
 defines an equivalence relation on if and only if the ϕ ji satisfy the three
conditions of Lemma 4.1.13.

2. Show that if the ϕ ji satisfy the conditions of Lemma 4.1.13, then the
quotient set

has the canonical structure of a smooth fibre bundle over M with general
fibre F and transition functions ϕ ji .
4.8.10 Prove that the principal bundle

from Example 4.2.16 is isomorphic to the frame bundle FrSO(S n−1).

4.8.11 Prove that a subset F E is a subbundle of the vector bundle E if and
only if F is the image of a vector bundle monomorphism to E.
4.8.12 Prove that

with projection onto the first factor, defines a -vector bundle over the
Grassmann manifold of rank k. This bundle is called the tautological
vector bundle . Particular examples, for k = 1, are the tautological line bundles
over and .
4.8.13 We denote by L → S 1 the infinite Möbius strip, defined by

It follows from Exercise 4.8.1 that this is a non-trivial, real line bundle over
the circle. Prove that the real vector bundle L L → S 1 is trivial.
4.8.14 Let L → S 1 be the infinite Möbius strip.

1. Show that under the diffeomorphism the infinite Möbius strip is

isomorphic to the tautological line bundle over .

2. Prove that the tautological line bundle over is non-trivial for all n ≥ 1.
 4.8.15 Let E → M be a real vector bundle of rank m. Show that E is
orientable if and only if Λ m E is trivial.
4.8.16 Let E → M be a -vector bundle of rank m with a positive definite
(Euclidean or Hermitian) bundle metric. Suppose that F E is a vector
subbundle. Prove that the orthogonal complement F is a vector subbundle of E
and that F F is isomorphic to E.
4.8.17 Determine the clutching function of the tangent bundle TS 2 → S 2
geometrically as follows:

1. Draw two disks in the plane and label them N and S. Draw on the boundary
circle of disk N four points a, b, c, d counter-clockwise with 90∘ between
consecutive points. Draw on the boundary circle of disk S corresponding
points a, b, c, d, such that the disks under identification of the boundary
circles yield a sphere S 2.

2. Draw in the center of disk N an orthonormal basis and label the vectors 1
and 2. Parallel transport this basis to the points a, b, c, d. Take these bases
and draw the matching bases on the S side in the points a, b, c, d. Call these
bases I.

3. Take the basis at the point a on disk S and parallel transport it to the center
of disk S. Then parallel transport this basis from the center to the points
b, c, d. Call these bases II.

4. Determine how bases I twist against bases II and thus determine the
clutching function, i.e. the degree of the map

To fix the sign of the degree, you probably need at least one more point at
45∘ between a and b, for example.
What do you get if you do something similar for TS 3 → S 3 by realizing S 3
as two solid cubes identified along their six faces?
4.8.18 Determine the clutching function of the tautological complex line
bundle . The total space of the line bundle is
and is covered by

4.8.19 Determine the clutching functions in the sense of Remark 4.6.4 for the
Hopf fibrations

and

4.8.20 Let

be a principal G-bundle and ρ: G → GL(V ), ρ i : G → GL(V i ), for i = 1, 2,

representations. Let

be the associated vector bundles. Show that the dual bundle E , the direct
sum E 1 E 2 and the tensor product E 1 E 2 are isomorphic to vector
bundles associated to P. Determine the corresponding representations of G and
the vector bundle isomorphisms. Show that the vector bundle associated to the
trivial representation is trivial.
4.8.21 (From [ 14 ]) Let

be the complex line bundle defined in Example 4.7.16.

1. Prove that γ 0 is trivial and γ 1 is isomorphic to the tautological line bundle.

2. Prove that for all and

 4.8.22 Let E → M be a complex vector bundle of rank n ≥ 2. Show that E is
associated to a principal SU(n)-bundle over M if and only if Λ n E is a trivial
complex line bundle.
4.8.23 Let E = P × ρ V be an associated vector bundle and α a section of the
adjoint bundle Ad(P). Prove that α defines a canonical endomorphism of the
vector bundle E.
4.8.24 Let M = G⁄H be a smooth homogeneous space and consider the
canonical principal H-bundle

Suppose that ρ: H → GL(V ) is a representation with associated

homogeneous vector bundle

1. Prove that there exists a canonical smooth left action of the Lie group G on
the total space E. Show that this action maps fibres of E by linear
isomorphisms onto fibres of E and that any given fibre of E can be mapped
by a group element onto any other fibre.

2. Identify the space Γ(E) of sections of the vector bundle E over the manifold
M with a suitable vector subspace Map H (G, V ) of the vector space
Map(G, V ).

Remark The representation of G on Γ(E), induced by this construction from the

representation of the closed subgroup H on V, is denoted by Ind H G (V ).

4.8.25 (From [ 30 ]) Let M = G⁄H be a smooth homogeneous space and consider

the canonical principal H-bundle

Prove that the tangent bundle TM is isomorphic to the homogeneous vector

bundle over M, defined by the representation ρ of H on the vector space ,
given by

where Ad denotes the adjoint representation of G.

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

13.
 Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
 [MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)
45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)
61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]
78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]
94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
 www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]
125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

140.
141.
 van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions. Chin.
Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_5
Chapter 5 Connections and Curvature
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

We said in the introduction to Chap. 4 that principal bundles and associated

vector bundles are the stage for gauge theories. From a mathematical and
physical point of view it is very important that we can define on principal
bundles certain fields, known as connection 1-forms. At least locally (after a
choice of local gauge) we can interpret connection 1-forms as fields on
spacetime (the base manifold) with values in the Lie algebra of the gauge group.
These fields are often called gauge fields and correspond in the associated
quantum field theory to gauge bosons. Every connection 1-form A defines a
curvature 2-form F which can be identified with the field strength of the gauge
field. Connection and curvature can be seen as generalizations of the classical
potential A μ and field strength F μν in electromagnetism, which is a U(1)-gauge
theory, to possibly non-abelian gauge groups.
Pure gauge theory, also known as Yang–Mills theory , involves only the
gauge field A and its curvature F. Additional matter fields, like fermions or
scalars, can be introduced using associated vector bundles. The crucial point is
that connections (the gauge fields) define a covariant derivative on these
associated vector bundles, leading to a coupling between gauge fields and matter
fields (if the matter fields are charged, i.e. the vector bundles are associated to a
non-trivial representation of the gauge group). In a gauge-invariant Lagrangian
this results in terms of order higher than two in the matter and gauge fields,
which are interpreted as interactions between the corresponding particles.
In non-abelian gauge theories, like quantum chromodynamics (QCD) , there
are also terms in the Lagrangian of order higher than two in the gauge fields
themselves, coming from a quadratic term in the curvature that appears in the
Yang–Mills Lagrangian . This implies a direct interaction between gauge bosons
(the gluons in QCD) that does not occur in abelian gauge theories like quantum
electrodynamics (QED) . The difficulties that are still present nowadays in trying
to understand the quantum version of non-abelian gauge theories, like quantum
chromodynamics, can ultimately be traced back to this interaction between
gauge bosons.
Although in this book we are mainly interested in applications of gauge
theories to physics, gauge theories are also very influential in pure mathematics,
for example, the Donaldson and Seiberg–Witten theories of 4-manifolds and
Chern–Simons theory of 3-manifolds (see Exercise 5.15.16 for an introduction to
the Chern–Simons action).
References for this chapter are [14, 39] and [84].

5.1 Distributions and Connections

Definition 5.1.1 A distribution on a manifold M is a vector subbundle of the
tangent bundle TM.

This notion of distributions is not related to the concept of distributions in

analysis. Connections on principal bundles P, sometimes also called
Ehresmann connections, are defined as certain distributions on the total space
of the principal bundle.

5.1.1 The Vertical Tangent Bundle

We first want to show that on the total space of every principal bundle there is a
canonical vertical bundle. Let

be a principal G-bundle. For a point x M we have the fibre

over x, which is an embedded submanifold of P. Let p P x be a point in the

fibre.

Definition 5.1.2 The vertical tangent space V p of the total space P in the
point p is the tangent space T p (P x ) to the fibre.

Proposition 5.1.3 (Vertical Tangent Bundle) For all p P the vertical

tangent space has the following properties:
1. V p = ker D p π.

2. The map

where is the fundamental vector field associated to , determined by

the G-action on P, is a vector space isomorphism between and V p .

3. The set of all vertical tangent spaces V p for p P forms a smooth

distribution on P, called the vertical tangent bundle V. Its rank is equal to
the dimension of G. The distribution V is globally trivial as a vector bundle,
with trivialization given by

4. The vertical tangent bundle is right-invariant , i.e.

Proof

1. We have

because we can write Y as the tangent vector to a curve in P x , which maps

under π to the constant point x M. Hence

Since π: P → M is a submersion, it follows from the Regular Value

Theorem A.1.32 that

This implies the claim.

2. It is clear that this map has image in V p and considering dimensions it
suffices to show that the map is injective. This follows from Proposition 3.4.
3.

3. This is clear by 2.

4. This follows because according to Proposition 3.4.6 , with

.
□

5.1.2 Ehresmann Connections

Let P → M be a principal G-bundle.

Definition 5.1.4 A horizontal tangent space in p P is a subspace H p of T p

P complementary to the vertical tangent space V p , so that

Note that horizontal tangent spaces are not defined uniquely (if the dimensions
of G and M are positive).

The following should be clear:

Proposition 5.1.5 Let H p be a horizontal tangent space at p P, π( p) = x.

Then

is a vector space isomorphism.

Definition 5.1.6 Let H be a distribution on P consisting of horizontal tangent

spaces. Then H is called an Ehresmann connection or a connection on P if it is
right-invariant, i.e.

The distribution H is also called horizontal tangent bundle given by the

connection.

Right-invariance of an Ehresmann connection means that along a fibre P x the

horizontal subspaces are mutually “parallel” (with respect to right translation
along the fibre). In particular, all H p along a fibre P x are determined by fixing a
single for some p 0 P x , since the G-action is transitive on the fibres of P.
Right-invariance of a connection can also be seen as a symmetry property: The
right action of the gauge group G on P induces a natural right action on TP and
Ehresmann connections are invariant under this action.

Example 5.1.7 (Connections on the Trivial Bundle) Let

be the trivial principal G-bundle. Then the vertical subspaces are given by

We can choose

It is clear that this defines a horizontal subspace complementary to the vertical

subspace. Furthermore, the collection of all of these horizontal subspaces are
right-invariant and hence define a connection on the trivial bundle, called the
canonical flat connection .

It can also be shown that every non-trivial principal bundle has a connection (see
Exercise 5.15.1 for a proof in the case of compact structure groups).
Notice that connections are not unique (if dimM, dimG ≥ 1), not even in the
case of trivial principal bundles (all connections that appear in the Standard
Model over Minkowski spacetime, for example, are defined on trivial principal
bundles).

5.2 Connection 1-Forms

In this section we study an equivalent description of connections using
differential forms.

5.2.1 Basic Definitions

Recall that we defined in Sect. 3.5.1 the notion of differential forms on a
manifold with values in a vector space. We now need this notion to define so-
called connection 1-forms.
 Definition 5.2.1 A connection 1-form or connection on a principal G-
bundle π: P → M is a 1-form on the total space P with the
following properties:

1. for all g G.

2. for all , where is the fundamental vector field

associated to X.

A connection 1-form is also called a gauge field on P.

At a point p P, a connection 1-form is thus a linear map

Recall that is a linear isomorphism of onto itself. This shows that the
composition is well-defined and again an element of .
We want to show that the notion of connection 1-forms is completely
equivalent to the notion of Ehresmann connections on a principal bundle as
defined in Sect. 5.1.2.

Theorem 5.2.2 (Correspondence Between Ehresmann Connections and

Connection 1-Forms) There is a bijective correspondence between
Ehresmann connections on a principal G-bundle π: P → M and connection 1-
forms:

1. Let H be an Ehresmann connection on P. Then

for p P, and Y p H p , defines a connection 1-form A on P.

2. Let be a connection 1-form on P. Then

defines an Ehresmann connection H on P.

Proof

1. We have to verify the conditions defining a connection 1-form. It is clear

that

We want to calculate r g A. We have shown in Proposition 3.4.6 that

, where . Note that r g Y p is horizontal if Y p is
horizontal by the definition of Ehresmann connections. Therefore

This implies the claim.

2. We have to verify that H is a horizontal right-invariant distribution on P. We

first show that H is a subbundle of TP: using a basis {T a } for the Lie
algebra we can write A = ∑ a A a T a , where A a Ω 1(P) are real-valued
1-forms. Since for all , it follows that . In
particular, the 1-forms A a are linearly independent in each point p P. Let
g be a Riemannian metric on P and Z a the vector fields g-dual to the 1-
forms A a . The {Z a } are linearly independent and span a subbundle ζ of TP
of rank . It follows that H is the g-orthogonal complement of ζ in TP
and hence a distribution.
To verify that H is horizontal, we first show that H p ∩ V p = {0}: Let Y
ker A p ∩ V p . Then Y is equal to a fundamental vector, hence for
some . But then

hence Y = 0.
 Furthermore, the 1-form A p is surjective onto , hence

Thus T p P = ker A p V p and H p is horizontal. To check that H is

right-invariant, let Y H p . Then

This shows r g Y H p⋅ g and hence the claim.

□

Example 5.2.3 Let G be a Lie group and H G a closed subgroup. By

Theorem 4.2.15

is an H-principal bundle. Suppose there exists a vector subspace such that

The homogeneous space is then called reductive . In this situation we can define
a canonical connection 1-form A on the bundle G → G⁄H. See Exercise 5.15.6
for details.

We describe another explicit example of a connection 1-form in the following

subsection.

5.2.2 A Connection 1-Form on the Hopf Bundle S 3 → S

2
In this subsection we follow reference [14, Example 3.3]. We consider the Hopf
bundle

and think of S 1 as the unit circle in with Lie algebra . If , then

exp(Y ) = e iy S 1. We also think of S 3 as the unit sphere in with tangent
spaces
We define 1-forms by

Proposition 5.2.4 (Connection 1-Form on the Hopf Bundle) The 1-form A

on S 3 , given by

has values in and is a connection 1-form for the Hopf bundle.

Proof It is clear that A has values in , since . We check the defining

properties of connection 1-forms. Since S 1 is abelian, we have for all
g S 1. We therefore have to show that

We fix a tangent vector , given by the velocity vector

of a suitable curve in S 3. Then

Hence

where we used that .

We also have to show that

for all Y in the Lie algebra of S 1. Let Y = iy, with . Then the associated
fundamental vector field is given by
This implies

since (z 0, z 1) S 3. □

See Exercise 7.9.9 for a generalization of this construction to the Hopf bundle S
7 → S 4 with structure group SU(2).

5.3 Gauge Transformations

Let π: P → M be a principal G-bundle.

Definition 5.3.1 A (global) gauge transformation is a bundle

automorphism of P, i.e. a diffeomorphism f: P → P which preserves the fibres
of P and is G-equivariant:

1. π ∘ f = π.

2. f( p ⋅ g) = f( p) ⋅ g for all p P and g G.

Under composition of diffeomorphisms the set of all gauge transformations
forms a group that we denote by or Aut(P). A local gauge transformation
is a bundle automorphism on the principal G-bundle π: P U → U, where U M
is an open set.

We sometimes prefer to call gauge transformations in this sense bundle

automorphisms and leave the name gauge transformations to physical gauge
transformations that we introduce later. Notice that whether a bundle
automorphism f is global or local is not related to the question of whether f is
constant or non-constant in some sense. We will later call gauge transformations
rigid if they are constant in a specific way.
Depending on the context one sometimes calls G or the gauge
group of the principal bundle P. The group is infinite-dimensional if both
the dimensions of M and G are at least 1.
 The group Aut(P) of bundle automorphisms is one of the places in
differential geometry where an infinite-dimensional group appears naturally.
Gauge theories, which are field theories invariant under all gauge
transformations, in this regard have the huge symmetry group Aut(P). The
diffeomorphism group Diff(M) of spacetime M plays a comparable role in
general relativity .

5.3.1 Bundle Automorphisms as G-Valued Maps on P

We would like to give another, equivalent description of bundle automorphisms
(we follow [14, Sect. 3.5]).

Definition 5.3.2 We denote by C ∞ (P, G) G the following set of maps from P to

G:

where is conjugation by g −1. This set is a group under pointwise

multiplication:

The neutral element is given by the constant map on P with value e G.

Proposition 5.3.3 (Correspondence Between Bundle Automorphisms and G-

Valued Maps) The map

with σ f defined by

is a well-defined group isomorphism. We can therefore identify the group of

bundle automorphisms with .

Proof Since f( p) is in the same fibre as p, there exists a unique g G such

that f( p) = p ⋅ g. This g we call σ f ( p).
We first have to show that σ f is an element of : It is not difficult to
show that σ f is a smooth map from P to G. We have
This implies that

and thus .
The inverse of the map above is given by

with f σ defined by

We only have to show that f σ is a bundle automorphism. It is clear that f σ ( p)

is in the same fibre as p, hence f σ is a bundle map. It is easy to check that f σ is
G-equivariant and that , hence f σ is a diffeomorphism. Thus .
Finally, we can check that σ f′∘f = σ f′ ⋅ σ f , hence the map defines a group
isomorphism between and .□

In the special case when the structure group G is abelian we have a simpler
description.

Proposition 5.3.4 (Bundle Automorphisms for Abelian Structure Groups)

If the Lie group G is abelian, then there is a group isomorphism

where denotes the set of smooth maps from M to G (a group under

pointwise multiplication) and σ τ is defined by

where π: P → M is the projection.

Proof This is Exercise 5.15.2. □

Corollary 5.3.5 For principal T n -bundles P → M there is an isomorphism of

the group of bundle automorphisms with the group of smooth maps from M
to T n .
5.3.2 Physical Gauge Transformations
In physics, gauge transformations are often defined as maps on the base
manifold M to the structure group G, even for non-abelian Lie groups G. We
discuss the relation of this notion to our definition of gauge transformations as
bundle automorphisms.

Definition 5.3.6 Let π: P → M be a principal G-bundle. A physical gauge

transformation is a smooth map τ: U → G, defined on an open subset U M.
The set of all physical gauge transformations on U forms a group with
pointwise multiplication. A rigid physical gauge transformation is a constant
map τ: U → G. The rigid physical gauge transformations form a group
isomorphic to G.

Proposition 5.3.7 (Physical Gauge Transformations and Bundle

Automorphisms) Let s: U → P be a local section. Then s defines a group
isomorphism

The inverse of this map is given by

where

Proof The proof is left as an exercise. □

The upshot is that after a choice of local gauge s on U we can identify local
bundle automorphisms on the principal G-bundle P U → U with physical
gauge transformations on U.

5.3.3 The Action of Bundle Automorphisms on

Associated Vector Bundles
Bundle automorphisms on a principal bundle have the important property that
they act on every associated vector bundle. Let π P : P → M be a principal G-
bundle and π E : E = P × ρ V → M an associated vector bundle.

Theorem 5.3.8 (Action of Bundle Automorphisms on Associated Bundles)

The group of bundle automorphisms of the principal bundle acts on the
associated vector bundle through bundle isomorphisms via

Proof We only have to show that the action is well-defined: If ,

then p′ = p ⋅ g and v′ = ρ(g)−1 v for some g G, so that

We can also describe this action in the language of physics:

Theorem 5.3.9 (Action of Physical Gauge Transformations on Associated

Bundles) Let s: U → P be a local gauge and Φ: U → E a local section. We
write the section with respect to the local gauge as

where ϕ: U → V is a smooth map. Suppose f is a local bundle automorphism of P

over U and τ f : U → G the associated physical gauge transformation. Then

Proof This is a simple calculation. □

As a consequence the action of a local bundle automorphism on a local section

Φ of E is given by the action of the physical gauge transformation on the vector-
valued map ϕ. In physics one writes the action of a physical gauge
transformations τ: U → G on a field ϕ: U → V as

The more general notion of bundle automorphism above has the advantage that it
also works for non-trivial principal bundles and associated vector bundles,
independent of the choice of (local) gauge.

Remark 5.3.10 There is a simple, but profound, difference between gauge

theories and general relativity (Edward Witten [150] attributes this insight to
Bryce DeWitt). In gauge theories the group of symmetries, the gauge group
, acts through bundle automorphisms, i.e. it preserves all points on the base
manifold M. This is related to the fact that gauge theories describe local
interactions (the interactions occur in single spacetime points). In general
relativity, however, the group of symmetries, the diffeomorphism group Diff(M)
, acts by moving points around in M. If the diffeomorphism invariance holds in
quantum gravity on the level of Green’s functions (correlators) , then they must
be constant, in striking contrast to the behaviour of Green’s functions in Poincaré
invariant quantum field theories .
It is nowadays thought that gravity cannot be described by a local quantum
field theory of point particles and that a theory of quantum gravity must be
fundamentally non-local. This leads to alternatives such as string theory , where
the graviton and other particles are no longer 0-dimensional point particles, but
1-dimensional strings.

5.4 Local Connection 1-Forms and Gauge

Transformations
Let π: P → M be a principal G-bundle and a connection 1-form. It is
very useful to consider the following notion.

Definition 5.4.1 Let s: U → P be a local gauge of the principal bundle on

an open subset U M. Then we define the local connection 1-form (or
local gauge field ) , determined by s, by

The local connection 1-form is thus defined on an open subset in the base
manifold M and can be considered as a “field on spacetime” in the usual
sense. If we have a manifold chart on U and {∂ μ } μ = 1, …, n are the local
coordinate basis vector fields on U, we set

We can also choose in addition a basis {e a } of the Lie algebra and then
expand
 The real-valued fields and the corresponding real-valued 1-
forms A s a Ω 1(U) are called (local) gauge boson fields .

A principal bundle can have many local gauges and it is interesting to

determine how the local connection 1-forms transform as we change the local
gauge. Let s i : U i → P and s j : U j → P be local gauges with U i ∩ U j ≠ .
Recall from the proof of Proposition 4.7.11 that

where

is the smooth transition function between the associated local trivializations.

We can consider g ji as a physical gauge transformation between the local gauges
s i and s j .
We have local connection 1-forms

We want to calculate the relation between A i and A j . Recall that the

Maurer–Cartan form was defined as

for v T g G. We set

Then we have:

Theorem 5.4.2 (Transformation of Local Gauge Fields Under Changes of

Gauge) With the notation above, the local connection 1-forms transform as

on U i ∩ U j . If is a matrix Lie group, then

where ⋅ denotes matrix multiplication, g ji −1 denotes the inverse in G and dg ji is

the differential of each component of the function . In
particular, if G is abelian, then

Proof Let x U i ∩ U j and Z T x M. We set

With the group action

we calculate by Proposition 3.5.4 and the chain rule

Therefore, by the defining properties of a connection 1-form A,

To prove the second claim recall from Proposition 2.1.48 that for a matrix Lie
group

for all g G and , and μ G (v) = g −1 ⋅ v for v T g G, hence

Remark 5.4.3 In physics one considers connection 1-forms usually only in the
local sense as -valued 1-forms A i on open subset U i of M together with the
transformation rule given by Theorem 5.4.2. The mathematical concept of
connections on principal bundles clarifies the invariant geometric object behind
this transformation principle.

A very similar argument implies the following global statement:

Theorem 5.4.4 (Transformation of Connections Under Bundle
Automorphisms) Let P → M be a principal bundle and a
connection 1-form on P. Suppose that is a global bundle automorphism.
Then f A is a connection 1-form on P and

Proof This is Exercise 5.15.3. □

Theorem 5.4.4 corresponds to the “active” point of view for gauge

transformations (symmetries are related to the behaviour under certain bundle
automorphisms), while Theorem 5.4.2 corresponds to the “passive” point of
view (symmetries are implicit in the behaviour under coordinate
transformations).

5.5 Curvature
5.5.1 Curvature 2-Forms
Let π: P → M be a principal G-bundle and a connection 1-form on P.
Let H be the associated horizontal vector bundle, defined as the kernel of A. We
have

and set

for the projection onto the horizontal vector bundle.

Definition 5.5.1 The 2-form , defined by

is called the curvature 2-form or curvature of the connection A. We sometimes

write F A to emphasize the dependence on A.

Here are some simple properties of the curvature.

Proposition 5.5.2 The following identities hold:

1. for all g G.
2. for all , where denotes insertion of the vector field .

Proof

1. Note that

Hence

on T p P, since both sides evaluated on X = X h + X v T p P, where X h is

horizontal and X v is vertical, are equal to r g (X h ). We now calculate for
vectors X, Y T p P:

2. This is clear, because .

5.5.2 The Structure Equation

Definition 5.5.3 Let P be a manifold and a Lie algebra. For and
we define by

where the commutators on the right are the commutators in the Lie algebra . In
the literature one also finds the notation η ϕ or [η ϕ] for [η, ϕ].

If we expand in a vector space basis {T a } for the Lie algebra

with η a , ϕ a standard real-valued k- and l-forms, then the definition is equivalent
to

Most of the time we need the definition only for 1-forms , where we
have

and

We can now state the following important formula for the curvature 2-form.

Theorem 5.5.4 (Structure Equation) The curvature form F of a

connection form A satisfies

We need the following lemma:

Lemma 5.5.5 Let be a fundamental vector field and Y a horizontal

vector field on P. Then the commutator [X, Y ] is horizontal.

Proof The flow of X is given by ϕ t = r exp(tV ). This implies by Theorem A.1.46

since Y p⋅ exp(tV ) H p⋅ exp(tV ) and ϕ −t preserves the horizontal tangent bundle.

□

We can now prove Theorem 5.5.4.

Proof We check the formula by inserting X, Y T p P on both sides of the

equation, where we distinguish the following three cases:
1. Both X and Y are vertical: Then X and Y are fundamental vectors,

for certain elements . We get

On the other hand we have

The differential dA of a 1-form A is given according to Proposition A.2.22

by

where we extend the vectors X and Y to vector fields in a neighbourhood of

p. If we choose the extension by the fundamental vector fields and ,
then

since V and W are constant maps from P to and we used that

according to Proposition 3.4.4. This implies the claim.

2. Both X and Y are horizontal: Then

and

This implies the claim.

3. X is vertical and Y is horizontal: Then for some . We have

and

Furthermore,
 since is horizontal by Lemma 5.5.5. This implies the claim.

The structure equation is very useful when we want to calculate the curvature of
a given connection.

5.5.3 The Bianchi Identity

Let F be the curvature 2-form of a connection A. Then dF is a 3-form on P with
values in the Lie algebra . We want to consider the situation where we insert in
all three arguments of dF a vector in the horizontal subbundle H defined by A.

Theorem 5.5.6 (Bianchi Identity (First Form)) The differential dF of the

curvature 2-form vanishes on H × H × H.

Proof We use the following formula for the differential of a 2-form η on P, see
Proposition A.2.22:

for all vector fields X, Y, Z on P. By the structure equation we have

so that

We set . Then

for all X, Y, Z T p P. We have if V is a horizontal vector field,

since

for an arbitrary vector field W on P. This implies the claim, because we can
assume that X, Y, Z are horizontal in the neighbourhood of p P. □
5.6 Local Curvature 2-Forms
Let A be a connection 1-form on the principal bundle P and s: U → P a local
section (local gauge), defined on an open subset U M. We then defined the
local connection 1-form (or local gauge field) by

Similarly we define:

Definition 5.6.1 The local curvature 2-form (or local field strength )
, determined by s, is defined by

If we have a manifold chart on U and {∂ μ } are local coordinate basis vector

fields on U, we set

If we choose in addition a basis {e a } of the Lie algebra , we can expand

the local field strength as

Proposition 5.6.2 (Local Structure Equation) The local field strength can be
calculated as

and

If the structure group G is abelian, then F s = dA s and

Proof We calculate
Here we used that

which is easy to verify. This implies the first formula. The second formula
follows from

Here we used that [∂ μ , ∂ ν ] = 0, because the basis vector fields {∂ μ } come from
a chart on U. □

In physics, the quadratic term [A μ , A ν ] in the expression for F μν (leading to

cubic and quartic terms in the Yang–Mills Lagrangian , see Definition 7.3.1
and the corresponding local formula in Eq. (7.1)) is interpreted as a direct
interaction between gauge bosons described by the gauge field A μ . The
quadratic term in the curvature is only present if the gauge group G is non-
abelian , like G = SU(3) in quantum chromodynamics (QCD) , but not if G is
abelian , like G = U(1) in quantum electrodynamics (QED) .

This explains why gluons , the gauge bosons of QCD, interact directly
with each other, while photons , the gauge bosons of QED, do not. It is also
the reason for phenomena in QCD such as colour confinement (at low
energies) and asymptotic freedom (at high energies).

We would like to determine how the local field strength transforms under
local gauge transformations . Let s i : U i → P and s j : U j → P be local gauges
with U i ∩ U j ≠ and associated local curvature 2-forms F i , F j . The local
gauge transformation

is defined by

We then have:
Theorem 5.6.3 The local curvature 2-forms transform as

on U i ∩ U j . If G is a matrix Lie group, then

Proof Recall from the proof of Theorem 5.4.2 that

for a vector V T x M. Hence we get for V, W T x M, since F vanishes if a

vertical vector is inserted:

Corollary 5.6.4 If G is abelian , then F s is independent of the choice of local

gauge and hence determines a well-defined, global, closed 2-form .

Proof In the abelian case the curvature defines a global form on M by

Theorem 5.6.3. It remains to check that F M is closed. In a local gauge s we have
according to the local structure equation

Since G is abelian, we have [A s , A s ] = 0. We conclude that

Remark 5.6.5 Note one important point about this corollary: Locally we have F
s = dA s if G is abelian, hence F s is locally exact. However, the 2-form F M in
general is not globally exact, because A s does not define a global 1-form on M
(there is a change of A s under changes of local gauge even if the structure group
is abelian, see Theorem 5.4.2).

5.6.1 The Curvature 2-Form of the Connection on the

Hopf Bundle S 3 → S 2
We consider the connection 1-form A on the Hopf bundle from Sect. 5.2.2 and
continue to use the same notation (we follow [14, Example 3.10]).

Proposition 5.6.6 The curvature of the connection 1-form A on the Hopf

bundle is given by

Proof Since S 1 is abelian, we have [A, A] = 0, hence F A = dA. The claim

follows once we have shown that

since then also . This is clear from the definition of α i and : for
example, if

with curves γ 0, γ 1, then

According to Corollary 5.6.4 the curvature F A determines a well-defined, global,

closed 2-form on S 2, where

for any local gauge s: U → S 3 on an open subset U S 2. We want to determine

the 2-form . Consider the open subset

together with the chart map

We consider a 2-form on , defined by

Proposition 5.6.7 The 2-form is given on U 1 S 2 by

Proof It suffices to show that

because then

for all local gauges s: U 1 → S 3. To prove the formula note that

This implies for (z 0, z 1) S3

We have , hence

This implies

Therefore

Proposition 5.6.8 For the connection on the Hopf bundle the following
equation holds:
Proof Since

(i.e. is the one-point compactification of ) we can calculate

The form extends to a well-defined 2-form on all of , which is equal to

. The 2-form

is known as the Fubini–Study form of . It is related to the standard volume

form ω std on S 2 of area 4π by

Remark 5.6.9 We can define for any principal S 1-bundle P → M over a

manifold M the first Chern class or Euler class as

This is a real cohomology class in H dR 2(M). It turns out that this class does not
depend on the choice of connection 1-form on P (even though the 2-form F M
does). In the case of the Hopf bundle we have
5.7 Generalized Electric and Magnetic Fields on
Minkowski Spacetime of Dimension 4
For the following notion from physics see, for example, [100]. Suppose π: P →
M is a principal G-bundle and M is with Minkowski metric η of signature (+,
−, −, −) (a similar construction works locally on any four-dimensional Lorentz
manifold) . Let x 0, x 1, x 2, x 3 be global coordinates in an inertial frame with
coordinate vector fields satisfying

We also write for the coordinates

Suppose A is a connection 1-form on P with curvature F. Let s: M → P be a

global gauge and

as above. We write

and we have the local structure equation

The 4 × 4-matrix (F μν ) comes from a 2-form on M and is skew-symmetric.

Definition 5.7.1 The generalized electric and magnetic field , determined by

the connection, the choice of gauge and the inertial frame, are the -valued
functions

defined by

Equivalently,
where ε ijk is totally antisymmetric with ε 123 = 1. We could expand the
generalized electric and magnetic fields further in a basis for the Lie algebra .

For quantum electrodynamics (QED) with G = U(1) these are the standard real-
valued electric and magnetic fields (after choosing a basis for ). In this
situation the electric and magnetic fields do not depend on the choice of gauge
according to Corollary 5.6.4, because G is abelian (the gauge field A s does
depend on the choice of gauge).
For G = SU(n), in particular G = SU(3) corresponding to quantum
chromodynamics (QCD) , these -valued fields are also called chromo-electric
and chromo-magnetic fields (or colour-electric and colour-magnetic fields).
They describe the field strength of the gluon field corresponding to the
connection 1-form A μ .

5.8 Parallel Transport

Connections define an important concept: parallel transport in principal and
associated vector bundles. The notion of parallel transport also leads to the
concept of covariant derivative on associated vector bundles.
Let π: P → M be a principal G-bundle with a connection A. We want to lift
curves in M to horizontal curves in P, which are defined in the following way
(by a curve we always mean in this and the following sections a smooth curve).

Definition 5.8.1 A curve γ : I → P is called a horizontal lift of a curve γ: I

→ M, defined on an interval I, if:

1. π ∘γ =γ

2. the velocity vectors are horizontal, i.e. elements of , for all t I.

The following theorem says that a horizontal lift of a curve in the base manifold
always exists and is unique once the starting point has been given.

Theorem 5.8.2 (Existence and Uniqueness of Horizontal Lifts of Curves)

Let γ: [a, b] → M be a curve with γ(a) = x. Let p be a point in the fibre P x . Then
there exists a unique horizontal lift γ p of γ with γ p (a) = p.

Proof Since P is locally trivial, there exists some lift δ of γ with δ(a) = p (one
could also argue that the pullback of the bundle P under the map γ is trivial,
because [a, b] is contractible). We want to find a map g: [a, b] → G such that

is horizontal. We will determine g(t) as the solution of a differential equation.

The curve γ (t) will be horizontal if

We can calculate with Proposition 3.5.4:

Hence

We conclude that g(t) has to be the solution of the differential equation

with g(0) = e. This is the integral curve in the Lie group G through e of the
time-dependent right-invariant vector field on G, corresponding to the Lie
algebra element . Such an integral curve on the interval [a, b] exists
by Theorem 1.7.18. An explicit solution for g(t) in the case of a linear Lie group
G can also be found in Proposition 5.10.4. □

Definition 5.8.3 Let γ: [a, b] → M be a curve in M. The map

is called parallel transport in the principal bundle P along γ with respect to

the connection A. See Fig. 5.1.
Fig. 5.1 Parallel transport

Theorem 5.8.4 (Properties of Parallel Transport) Let P be a principal

bundle with connection A.

1. Parallel transport Π γ A is a smooth map between the fibres P γ(a) and P γ(b)
and does not depend on the parametrization of the curve γ.

2. Let γ be a curve in M from x to y and γ′ a curve from y to z. Denote the

concatenation by γ γ′, where γ comes first. Then

3. If γ − denotes the curve γ traversed backwards, then

In particular, parallel transport is a diffeomorphism between the fibres.

4. Parallel transport is G-equivariant: The following identity holds:

Proof Properties 1–3 follow from the theory of ordinary differential equations.
We only prove 4: let γ be a curve from x to y in M and p P x . Let γ p be the
horizontal lift of γ to p. For g G consider the curve r g ∘γ p . This curve starts
at p ⋅ g and projects to γ. Furthermore, it is horizontal, because r g maps
horizontal vectors to horizontal vectors by the definition of connections. It
follows that r g ∘γ p is equal to γ p⋅ g . We get

Since parallel transport does not depend on the parametrization of the curve γ,
we will often assume that γ is defined on the interval [0, 1].

5.9 The Covariant Derivative on Associated Vector

Bundles
So far we have considered connections on principal bundles. Associated vector
bundles play an important role in gauge theory, because matter fields are sections
of such bundles. It turns out that connections on principal bundles define so-
called covariant derivatives on all associated vector bundles (this will explain
the third row in the diagram at the beginning of Sect. 4.7). These covariant
derivatives appear in physics, in particular, in the Lagrangians and field
equations defining gauge theories.
We first want to define the notion of parallel transport in associated vector
bundles. Let P → M be a principal G-bundle with a connection A, ρ: G → GL(V
) a representation on a -vector space V ( ) and E = P × ρ V the
associated vector bundle.

Theorem 5.9.1 For a curve γ: [0, 1] → M the map

is well-defined and a linear isomorphism. This map is called parallel transport

in the associated vector bundle E along the curve γ with respect to the
connection A.

Proof We first show that Π γ E, A is well-defined, independent of the choice of

representative [ p, v]. Suppose that
 Then there exists an element g G such that

Part 4 of Theorem 5.8.4 then implies

Hence the map Π γ E, A is well-defined. It is then also clear that Π γ E, A is a

linear isomorphism. □

Let Φ be a section of E, x M a point and X T x M a tangent vector. We want

to define a covariant derivative as follows: choose an arbitrary curve γ: (−ε, ε)
→ M with

For each t (−ε, ε) parallel transport the vector Φ(γ(t)) E γ(t) back to E x
along γ. Then take the derivative in t = 0 of the resulting curve in the fibre E x ,
giving an element in E x . More formally, we set

Here γ t denotes the restriction of the curve γ starting at time 0 and ending at time
t, for t (−ε, ε).
We want to prove the following formula.

Theorem 5.9.2 Let s: U → P be a local gauge, A s = s A and ϕ: U → V the

map with Φ = [s, ϕ]. Then the vector D(Φ, γ, x, A) E x is given by

Proof We have

Let q(t) be the uniquely determined smooth curve in the fibre P x such that

Write
with a uniquely determined smooth curve g(t) in G. Then

For t = 0 we have

hence

This implies

It follows that

It remains to calculate . We have

and

Since the curve is horizontal with respect to A, we get

However,

hence

by the definition of connection 1-form. It follows that

and thus the claim. □

The theorem implies that D(Φ, γ, x, A) depends only on the tangent vector X and
not on the curve γ itself. We can therefore set:

Definition 5.9.3 Let Φ be a section of an associated vector bundle E and

 a vector field on M. Then the covariant derivative AΦ is the
X
section of E defined by

where γ is any curve through x and tangent to X x . The covariant derivative

is a map

The fact that A Φ is a smooth 1-form in Ω 1(M, E) for every Φ Γ(E) is

clear from the local formula.
We often write in a local gauge s: U → P, with Φ = [s, ϕ], the covariant
derivative as

where

i.e.

Here are some properties of the covariant derivative.

Proposition 5.9.4 (Properties of Covariant Derivative) The map A is -

linear in both entries and satisfies

for all smooth functions and the Leibniz rule

for all smooth functions .

Proof -linearity of A and function linearity in X is clear. Let be a

smooth function. Then
Here we used the product rule for functions to the vector space V multiplied by
functions to the scalars . □

Remark 5.9.5 If {∂ μ } are local basis vector fields on U, we get

where

In physics the covariant derivative is typically written in this form and

acts on functions ϕ on U with values in the vector space V, determined by
sections Φ in E and the local gauge s. In mathematics the covariant
derivative acts directly on the sections of the vector bundle. We denote
both operators by A (it will be clear from the context which operator is
meant).

From a physics point of view it is important that the second summand

A μ ϕ in the covariant derivative is non-linear (quadratic) in the fields A μ
and ϕ. This non-linearity, called minimal coupling , leads to non-quadratic
terms in the Lagrangian (see Definition 7.5.5 and Definition 7.6.2 as well
as the local formulas in Eqs. (7.3) and (7.4)), which are interpreted as an
interaction between gauge bosons described by A μ and the particles
described by the field ϕ.

Notice the crucial role played by the representation ρ: It is not only

needed to define the associated vector bundle E, but also to define the
covariant derivative A . The gauge field A can act on maps with values in
V (or sections of E) only if V carries a representation ρ of the gauge group
G. If the representation

is non-trivial and hence the coupling between the gauge field A μ and
 the field ϕ is (potentially) non-trivial, then the particles corresponding to ϕ
are called charged (charged particles are affected by the gauge field). In
Chaps. 8 and 9 we will discuss in some detail the representations that
appear in the description of matter particles in the Standard Model and in
Grand Unified Theories.

Figure 5.2 shows the Feynman diagrams for the cubic and quartic
terms which appear in the Klein–Gordon Lagrangian in Eq. (7.3),
representing the interaction between a gauge field A and a charged scalar
field described locally by a map ϕ with values in V.

Fig. 5.2 Feynman diagrams for interaction between gauge field and charged scalar

Remark 5.9.6 In physics the covariant derivative is often defined (without

referring to parallel transport) by the local formula

where

One then has to show that this definition is independent of the choice of local
gauge: Suppose s′: U → P is another local gauge. Then there exists a smooth
physical gauge transformation g: U → G such that

We have

with
Furthermore,

where

It follows that

A lengthy calculation (if done in this abstract setting) then shows that

is equal to

It is often important to consider covariant derivatives compatible with a bundle

metric on E. The natural bundle metrics constructed in Proposition 4.7.12 are
compatible with covariant derivatives.

Proposition 5.9.7 (Natural Bundle Metrics Are Compatible with Covariant

Derivatives) Let ⋅ , ⋅ V be a G-invariant scalar product on the vector space
V and ⋅ , ⋅ E the induced bundle metric on the associated vector bundle E = P ×
ρ V. Then the covariant derivative associated to a connection A is compatible
with the bundle metric in the sense that

for all sections Φ, Φ′ of E and all vector fields X on M.

Proof Since the scalar product on V is G-invariant, the map ρ induced by the
representation satisfies

for all and ϕ, ϕ′ V; see Proposition 2.1.37. This implies:

□
5.10 Parallel Transport and Path-Ordered Exponentials
We derive in this section a formula that is used in physics to calculate the
parallel transport on principal bundles. The following arguments are outlined in
[103]. Recall that for the proof of Theorem 5.8.2 concerning the existence of a
horizontal lift γ of a curve γ: [0, 1] → M, where

we had to solve the differential equation

with g(0) = e, where δ is some lift of γ and g: [0, 1] → G is a map with

There is a nice way to write the solution g(t) explicitly, at least if G is a matrix
Lie group and γ is contained in an open set over which the principal bundle is
trivial.
Suppose that the curve γ is contained in an open set U M, so that P U is
trivial over U. Let s: U → P be a local gauge with s(γ(0)) = p. We can choose

We then have to solve

Suppose that is a linear group, i.e. a closed Lie subgroup. Then

the differential equation can be written as

We write this as

where is a smooth map determined by γ(t), independent of g.

5.10.1 Path-Ordered Exponentials

Let G be a linear group with Lie algebra .

Definition 5.10.1 For a smooth map we define for all t [0, 1]

the following matrices in :

where in the definition of P n ( f, t)

The following is easy to show:

Lemma 5.10.2 For all n ≥ 2 the integral

where

evaluates to

Considering

with respect to a matrix norm on it follows that:

Proposition 5.10.3 The series

converges for every t [0, 1] and defines a smooth map

We write

and call this the path-ordered exponential of the function f.

Path-ordered exponentials are useful, because they define solutions to the

ordinary differential equation we are interested in.
Proposition 5.10.4 (Path-Ordered Exponential Defines Solution of ODE)
Consider a smooth map and define a smooth map

by

Then

In particular, g is a map

Proof It is clear that g(0) = I n . Calculating the derivative with respect to t we

get

This implies the first claim.

To prove the claim that g takes values in G, note that

defines a time-dependent vector field X on , which is right-

invariant in the sense that

The smooth map

is an integral curve of the vector field X through the unit matrix I n . Let

be the integral curve through I n of the restriction of the right-invariant vector

field X to G, given by Theorem 1.7.18 (the vector field X is tangent to G,
because f takes values in the Lie algebra ). Then uniqueness of the solution to
ordinary differential equations shows that g ≡ h, hence g takes values in G. □

5.10.2 Explicit Formula for Parallel Transport

Returning to the situation before Sect. 5.10.1, we can write the curve γ(t) in a
chart on U with coordinates x μ as γ(t) = x μ (t). Then

The solution to this differential equation is

where γ t denotes the restriction of the curve γ to [0, t]. In particular,

We therefore get:

Theorem 5.10.5 (Parallel Transport Expressed with Path-Ordered

Exponential) Let P → M be a principal bundle with matrix structure group
G. Suppose that γ: [0, 1] → M is a curve inside an open set U M over which P
U is trivial. Let p P γ(0) be a point and s: U → P a local gauge, such that
s(γ(0)) = p. Suppose s(γ(1)) = q. Then the parallel transport of p can be written
as

5.11 Holonomy and Wilson Loops

We saw that the induced parallel transport on associated vector bundles can be
used to define covariant derivatives. We want to explain another concept where
parallel transport on associated vector bundles appears in physics (we follow the
definition in [47]).
Suppose γ: [0, 1] → M is a closed curve in M (a loop) with γ(0) = γ(1) = x.
Then parallel transport Π γ E, A is a linear isomorphism of the fibre E x to itself.

Definition 5.11.1 We call the isomorphism Π γ E, A of E x the holonomy Hol γ, x

E (A) of the loop γ in the basepoint x with respect to the connection A.

We can express the holonomy using path-ordered exponentials.

Proposition 5.11.2 (Holonomy Expressed with Path-Ordered Exponential)
Suppose that G is a matrix Lie group and the loop γ is contained in an open set
U M over which P is trivial and s: U → P is a local gauge. Then s determines
an isomorphism

and with respect to this isomorphism

Proof This is Exercise 5.15.8. □

We want to understand how the holonomy changes if we choose a different base

point on the curve γ.

Lemma 5.11.3 Let y be another point on the closed curve γ and σ the part of γ
from x to y. Then

where we abbreviate Π σ = Π σ E, A .

Proof Let σ′ be the remaining part of γ from y to x. Then γ = σ σ′ and

This implies the claim. □

Therefore the following map is well-defined.

Definition 5.11.4 The Wilson operator or Wilson loop is the map W γ E that
associates to a connection A and a loop γ the number

where tr denotes trace, x is any point on γ, and the second formula holds if G is a
matrix Lie group and γ is inside a trivializing open set U for P.

Proposition 5.11.5 (Wilson Loops Are Gauge Invariant) The Wilson loop is
invariant under all bundle automorphisms of P:
Proof This is Exercise 5.15.9. □

In quantum field theory, the gauge field A μ is a function on spacetime with

values in the operators on the Hilbert state space V (if we ignore for the
moment questions of whether this operator is well-defined and issues of
regularization). The formula

shows that the Wilson loop W γ E (A) is a gauge invariant operator on this
Hilbert space.

5.12 The Exterior Covariant Derivative

In Sect. 5.9 we defined a covariant derivative

We can think of this map as a generalization of the differential

In fact, the differential d can be identified with the covariant derivative on the
trivial line bundle over M induced from the trivial connection. The differential d
can be uniquely extended in the standard way to an exterior derivative

by demanding that ddf = 0 for all functions and

for all α Ω k (M) and β Ω l (M). This differential satisfies d ∘ d = 0 on all

forms, see Exercise 5.15.11. Because of this property, the de Rham cohomology

is well-defined for all k.

We want to show that we can extend the covariant derivative in a similar way
to an exterior covariant derivative
This exterior covariant derivative, however, in general does not satisfy d A ∘
d A = 0. The non-vanishing of d A ∘ d A is precisely measured by the curvature of
A, see Exercise 5.15.12.
The constructions in this section work for general covariant derivatives on
vector bundles, which are defined as follows.

Definition 5.12.1 Let E → M be a -vector bundle. Then a covariant

derivative on E is a -linear map

such that

for all smooth functions and the Leibniz rule

holds for all smooth functions and sections e Γ(E).

We need the following wedge product.

Definition 5.12.2 There is a well-defined wedge product

between standard differential forms (with values in ) and differential forms

with values in E.

To explain this definition, we only have to see that the standard definition of the
wedge product works in this case. The standard definition involves the sum over
products of the two differential forms after we inserted a permutation of the
vectors (cf. Definition A.2.5). In the standard case we get the product between
two scalars in , while here we get the product between a scalar in and a
vector in E, which is still well defined.
To define the exterior covariant derivative, let ω be an element of Ω k (M, E).
We choose a local basis e 1, …e r of E over an open set U M. Then ω can be
written as
with uniquely defined k-forms ω i Ω k (U) (with values in ).

Definition 5.12.3 Let be a covariant derivative on a vector bundle E. Then

we define the exterior covariant derivative or covariant differential

by

If = A is the covariant derivative on an associated vector bundle determined

by a connection A on a principal bundle, we write d A = d .

Lemma 5.12.4 The definition of d is independent of the choice of local basis

{e i } for E.

Proof Let {e i ′ } be another local basis of E over U. Then there exist unique
functions with

The matrix C with entries C ji is invertible. Let C −1 be the inverse matrix with
entries C lj −1 and define

Then

We calculate

But
This implies the claim. □

The first part of the next proposition follows immediately from the definition by
considering a local basis {e i } for E. The second part is clear.

Proposition 5.12.5 The exterior covariant derivative d satisfies

for all ω, ω′ Ω k (M, E), σ Ω k (M) and e Γ(E). Furthermore, we have on

Γ(E) = Ω 0(M, E)

so that the exterior covariant derivative d is an extension of the covariant

derivative .

We want to show the following formula:

Proposition 5.12.6 (Leibniz Formula for Exterior Covariant Derivative)

The exterior covariant derivative d satisfies

for all σ Ω k (M) and ω Ω l (M, E).

Note that this reduces in the case of a 0-form ω with values in E to the second
formula in Proposition 5.12.5, because d on Ω 0(M, E) = Γ(E) is equal to .

Proof We write

with a local basis {e i } of E over U and ω i Ω l (U). Then

and
□

Remark 5.12.7 Contrary to the case of the standard exterior derivative d, it can
be shown that d in general has square

The non-vanishing of d 2 is related to the curvature F of the covariant

derivative (see Exercise 5.15.12).

We finally want to derive a local formula for the exterior covariant derivative d A
in the case of an associated vector bundle. Let P → M be a principal G-bundle,
ρ: G → GL(V ) a representation and E = P × ρ V the associated vector bundle. Let
A be a connection 1-form on P.

Definition 5.12.8 We define the wedge product

by expanding ω = ∑ i = 1 n ω i v i in an arbitrary basis {v i } for V and setting

This is independent of the choice of basis {v i } for V.

Let s: U → P be a local gauge. With respect to the local gauge a form σ Ωl

(M, E) defines a form σ s Ω l (M, V ). We get:

Theorem 5.12.9 With respect to a local gauge s: U → P we can write

for all ω Ω k (M, E).

Proof Choose a basis v 1, …, v n for V. This determines a local frame e 1, …, e n

for E by setting e i = [s, v i ]. If we expand a form σ Ω l (M, E) as
then

We write

and calculate

which implies

5.13 Forms with Values in Ad(P)

Recall that connections are 1-forms on the total space of a principal bundle P
with values in the Lie algebra . We now want to show that the difference
between two connections can be understood as a field on the base manifold M
with values in the vector bundle Ad(P). We then get a better understanding of
why gauge bosons in physics are said to transform under the adjoint
representation.
Let π P : P → M be a principal G-bundle. We then have the vector space
of k-forms on P with values in the Lie algebra . We want to consider a
certain vector subspace of this vector space (we follow [14, Chap. 3]).

Definition 5.13.1 Let be a k-form on P with values in the Lie

algebra . We call ω

1. horizontal if for all p P

 whenever at least one of the vectors X i T p P is vertical.

2. of type Ad if

for all g G.
We denote the set of horizontal k-forms of type Ad on P with values in by

which is clearly a real vector space (usually infinite-dimensional).

This notion is useful for the following reason.

Proposition 5.13.2 Let P → M be a principal G-bundle.

1. Suppose are connection 1-forms on P. Then

Moreover, if ω is an arbitrary element in , then A + ω is a

connection on P.

2. The curvature F of a connection A on P is an element of .

Proof This follows immediately from the defining properties of connections

and the curvature. □

Corollary 5.13.3 The set of connection 1-forms on P is an affine space over

the vector space . A base point is given by any connection 1-form
on P.

It is sometimes useful to have a different description of the vector space of

horizontal k-forms of type Ad on P. Recall that we defined in Example 4.7.17
the adjoint bundle

which is the real vector bundle associated to the principal bundle P via the
adjoint representation .
Theorem 5.13.4 The vector space is canonically isomorphic to the
vector space Ω k (M, Ad(P)).

Proof We define a map

as follows: Let be an element of . Then we define by

where
x M and p P are arbitrary with π P ( p) = x.
Xi T x M and Y i T p P are arbitrary with π P (Y i ) = X i .
We first show that ω is well-defined. For fixed p P the definition is
independent of the choice of vectors Y i : If Y i ′ are a different set of vectors with
π P (Y i ′ ) = X i , then

hence Y i ′ − Y i is vertical. Since is horizontal, we get

We now show independence of the choice of p in the fibre P x : Let p′ be

another point in P with π P ( p′) = x. Then p′ = p ⋅ g −1 for some g G. Let Y 1,
…, Y k be vectors in T p′ P. We calculate

where we set Z i = r g Y i . We have

This proves independence of the choice of p.

We prove that the form ω is smooth: Let s: U → P be a local gauge and X 1,
…, X k smooth vector fields on U. Then
Hence ω Ω k (M, Ad(P)).
This shows that the map Λ is well-defined and it is clearly linear. It remains
to show that Λ is bijective: Let ω Ω k (M, Ad(P)) and define by

Then and .□

As a consequence, we get the following statement about connection 1-forms and

curvature 2-forms.

Corollary 5.13.5 (Connections, Curvature and Forms with Values in

Ad(P)) Let P → M be a principal G-bundle.

1. The difference of two connection 1-forms on P can be identified with an

element of Ω 1(M, Ad(P)). The set of all connections on P is an affine space
over Ω 1(M, Ad(P)).

2. The curvature F A of a connection A on P can be identified with an element

F M A of Ω 2(M, Ad(P)).

The notation F M A generalizes the notation in Corollary 5.6.4, because for an

abelian structure group G the adjoint bundle Ad(P) is trivial and F M A has values
in .

In quantum field theory, particles in general are described as excitations

of a given vacuum field. In the case of a gauge field the vacuum field is a
certain specific connection 1-form A 0 on the principal bundle (the form A 0 ≡
0 is not a connection). Strictly speaking, gauge bosons , the excitations of the
gauge field , should then be described classically by the difference A − A 0,
where A is some other connection 1-form, and not by the field A itself. By
Corollary 5.13.5 this difference can be identified with a 1-form on spacetime
M with values in Ad(P). In physics this fact is expressed by saying that gauge
bosons, the differences A μ − A μ 0, are fields on spacetime that transform in
the adjoint representation of G under gauge transformations.
5.14 A Second and Third Version of the Bianchi
Identity
Let P → M be a principal G-bundle and A a connection 1-form with curvature F
A . According to Exercise 5.15.14 we can state the Bianchi identity in the

following equivalent form:

Theorem 5.14.1 (Bianchi Identity (Second Form)) The curvature F A Ω2

satisfies

for any connection A on P.

In Sect. 5.13 we saw that the curvature F A can be identified with an element F M
A in Ω 2(M, Ad(P)). On the other hand the connection A defines an exterior

covariant derivative d A on the associated bundle Ad(P):

We can then write the Bianchi identity in a third equivalent form:

Theorem 5.14.2 (Bianchi Identity (Third Form)) The curvature F M A Ω

2 (M, Ad(P)) satisfies

for any connection A on P.

Proof This is an immediate consequence of Theorem 5.12.9 and

Theorem 5.14.1. □

5.15 Exercises for Chap. 5

5.15.1. Let G be a compact Lie group.

1. Suppose P × G → P is a right-action of G on a manifold P. Prove that P has

a G-invariant Riemannian metric.

2. Prove that every principal G-bundle π: P → M has a connection.

 5.15.2. Suppose that π: P → M is a principal G-bundle where the Lie group
G is abelian. Show that the following map is a group isomorphism

where denotes the set of smooth maps from M to G (a group under

pointwise multiplication) and σ τ is defined by

5.15.3. Let P → M be a principal bundle and a connection 1-form

on P. Suppose that is a global bundle automorphism. Prove that f A is
a connection 1-form on P and

5.15.4. Let P → M be a principal G-bundle with a connection 1-form

and curvature 2-form . Let X and Y be horizontal vector
fields on P with respect to the Ehresmann connection H defined by A.

1. Show that F(X, Y ) = −A([X, Y ]).

2. Prove that the curvature F vanishes identically if and only if the distribution
H is integrable , i.e. [X, Y ] is a horizontal vector field for all horizontal
vector fields X, Y on P.

3. Suppose that M is connected and simply connected (π 1(M) = 1) and the

curvature F vanishes identically. Prove that P is trivial and there exists a
global gauge s: M → P such that A s = s A ≡ 0.
5.15.5. Let G be a Lie group. Then G acts by right-multiplication on the right
of G:

Since the action is simply transitive, it follows that this defines a principal G-
bundle over the manifold consisting of one point:
1. Show that the Maurer–Cartan form is a connection on this
principal bundle and that it is the only one.

2. Determine the curvature of the connection μ G . What is the interpretation of

the structure equation?
5.15.6. Let G be a Lie group and H G a closed subgroup. By Theorem 4.2.
15

is an H-principal bundle. We assume that there exists a vector subspace

such that

i.e. the homogeneous space is reductive.

1. Consider . Prove that A is a connection 1-form on G →

G⁄H.

2. Show that the vertical and horizontal subspaces (defined by the connection
A) at a point g G are given by and .

3. Prove that the curvature of the connection A is given by

(the commutator on the right is taken in ).

5.15.7. Recall from Exercise 4.8.7 that the Hopf right action of S 1 on S 3
induces a right action of on and the lens space L( p, 1)
thus has the structure of a principal circle bundle over S 2:

Prove that the connection A on the Hopf bundle defined in Sect. 5.2.2
induces a connection A′ on L( p, 1) → S 2. Determine the relation between the
global curvature 2-form of this connection and the curvature 2-form of
the Hopf connection, as well as the integral

5.15.8. Suppose that P → M is a principal G-bundle with a matrix Lie group

G, E an associated vector bundle, A a connection on P and γ a loop in M.
Suppose that the loop γ is inside an open set U M over which P is trivial and
s: U → P is a local gauge. Then s determines an isomorphism

Prove that with respect to this isomorphism

5.15.9. Let P → M be a principal G-bundle with a connection A and a

bundle automorphism. Suppose γ: [0, 1] → M is a curve in M.

1. Show that parallel transport with respect to the connection f A is given by

2. Let E → M be a vector bundle associated to P and suppose that γ is a closed

curve in M. Show that the Wilson loop is invariant under bundle
automorphisms of P:

5.15.10. Let

be the Hopf bundle with the connection A defined in Sect. 5.2.2. Let σ denote
the equator in , starting and ending at the point .

1. Show that σ can be parametrized as

 for a suitable identification of S 2 with .

2. Determine the horizontal lift σ : [0, 2π] → S 3 of σ with respect to the

connection A, starting at .

3. Let γ k → S 2 be the complex line bundle associated to the Hopf bundle via
the representation

as in Example 4.7.16. Determine the Wilson loop .

5.15.11. Define the differential

by demanding that
is the standard differential of functions

ddf = 0 for all

d(α β) = dα β + (−1) k α dβ for all α Ω k (M) and β Ω l (M).

Prove that ddω = 0 for all ω Ω k (M) and all k.
5.15.12. Let E → M be a vector bundle with a covariant derivative . We
define the curvature of by

where and Φ Γ(E).

1. Show that F (X, Y )Φ is function linear in each argument X, Y, Φ. The

curvature thus defines an element

where End(E) denotes the endomorphism bundle of E over M, whose fibre

End(E) x over x M is given by End(E x ).
2. Show that, as 2-forms with value in E,

for all sections Φ Γ(E).

3. Define a wedge product

by writing ω Ω l (M, E) with a local frame {e i } for E over U as ω = ∑ i =

nω e i with ω i Ω l (U) and setting for α Ω k (M, End(E))
1 i

(this definition is independent of choices). Prove that

for all ω Ω l (M, E).

5.15.13. Let P → M be a principal G-bundle and E = P × ρ V → M an
associated vector bundle. We fix a connection A on P with induced covariant
derivative A on E. Let F be the curvature of A defined in Exercise 5.15.12.
Suppose that s: U → P is a local gauge and write a section Φ of E locally as Φ =
[s, ϕ] with v: U → V. Show that

where F s = s F and F is the curvature of A on P.

5.15.14. Suppose that P is a manifold and a Lie algebra. Consider forms
, and . Prove the following identities:

Derive as an application the following second form of the Bianchi identity

where F A is the curvature of a connection 1-form A on a principal bundle P

→ M.
5.15.15. This exercise is a preparation for Exercise 5.15.16. Suppose that P is
a manifold and a Lie algebra with a scalar product ⋅ , ⋅ which is skew-
symmetric with respect to the adjoint representation ad. Let and
. We define a real-valued form ω, τ Ω k+l (P) by

Consider also a 1-form . Prove the following identities:

5.15.16 (From [ 52 ]). We use the notation from Exercise 5.15.15. Suppose
that P → M is a principal G-bundle over a manifold M and ⋅ , ⋅ an Ad-invariant
scalar product on the Lie algebra . Let A be a connection 1-form on P with
curvature F. We define the Chern–Simons form α(A) Ω 3(P) by

1. Prove that dα = F, F .

2. Let be a bundle automorphism with induced map σ f : P → G and

set ϕ = σ f μ G . Prove that the Chern–Simons form changes under bundle
automorphisms as

3. Show that the form

is closed.

4. Suppose that M is a closed oriented 3-manifold, P → M a trivial G-bundle

and τ G represents an integral class in . If s: M → P is a global
gauge, define the Chern–Simons action by
 Prove that modulo the number is independent of
the choice of global gauge s.

Remark Notice that the Chern–Simons action is purely topological, i.e. does
not depend on the choice of a metric on M. This leads to the concept of
topological quantum field theories (TQFT). Similar Chern–Simons terms appear
in many places in physics, for example, in the actions of supergravity and in the
actions for D-branes in string theory.

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum:
Bull. Am. Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
 doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory.
W. A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives.
Eur. Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]
26.
Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
 [MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]
42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
 and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]
58.
 Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories.
Phys. Rev. Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry.
Phys. Rev. D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles.
Phys. Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009
75.
 Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]
91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles.
Bull. Am. Math. Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings.
Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/
107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html
122.
 Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics.
Phys. Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)
[MATH]

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)
138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions.
Chin. Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_6
Chapter 6 Spinors
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

In Chap. 5 we studied gauge fields that mediate interactions between particles in

gauge theories. Gauge fields correspond to gauge bosons (spin 1 particles) and
are described by 1-forms or, dually, vector fields. In physics, of course, there
also exist matter particles, like electrons, quarks and neutrinos. These particles
are fermions (spin particles) and are described byspinor fields (spinors). Like
vector fields or tensor fields, spinors have a specific transformation behaviour
under rotations. However, spinors do not transform directly under the orthogonal
group, but under a certain double covering, called the(orthochronous) spin
group. In the case of Minkowski spacetime, rotations correspond to Lorentz
transformations. The corresponding spin group is the Lorentz spin group.
In many mathematical expositions the discussion of spinors is restricted to
the Riemannian case, because in most situations, manifolds in differential
geometry carry a Riemannian metric. The pseudo-Riemannian case, like the case
of Minkowski spacetime, is discussed less often, even though it is very important
for physics (a notable exception is the thorough discussion in Helga Baum’s
book [13]). Since we are ultimately interested in applications of differential
geometry and gauge theory to physics, it seems worthwhile to study orthogonal
groups, Clifford algebras, spin groups and spinors from a mathematical point of
view also in the Lorentzian and general pseudo-Riemannian case.
The discussion in the present chapter is far more extensive than strictly
necessary to understand particle physics and the Standard Model in 4-
dimensional flat spacetime, but may be useful for further studies leading to
theories such as supersymmetry, supergravity and superstrings in higher
dimensions. We sometimes only sketch the arguments that are more or less
standard and can be found in other references, and focus instead on topics which
are perhaps less well-known in the mathematics literature, like Majorana spinors
and scalar products. Even though spinors are elementary objects, some of their
properties (like the periodicity modulo 8, real and quaternionic structures, or
bilinear and Hermitian scalar products) are not at all obvious, already on the
level of linear algebra, and do not have a direct analogue in the bosonic world of
vectors and tensors.
General references for this chapter are [13, 15, 20, 28, 35, 55, 88] and [93].

6.1 The Pseudo-Orthogonal Group O(s, t) of Indefinite

Scalar Products
We begin by describing the structure of the pseudo-orthogonal group for a
general indefinite scalar product. This includes the particular case of the Lorentz
group of Minkowski spacetime. In Sect. 6.5 we will then introduce the spin
group, which is a certain double covering of a (pseudo-)orthogonal group.
In this chapter denotes the field or . Let V be a finite-dimensional -
vector space.

Definition 6.1.1 A symmetric bilinear form is a map

which is symmetric and -bilinear. The form Q is called non-degenerate if for

each v ≠ 0 V there exists a vector w V with Q(v, w) ≠ 0.

Notice that in the complex case we also considercomplex bilinear and not
Hermitian forms.

Example 6.1.2 We denote by the vector space with standard basis e 1,

…, e s+t and the standard symmetric bilinear form η defined by

The bilinear form η has signature (s, t), also written as

and is non-degenerate. We set n = s + t.

The space
 is Euclidean space , and

and are the two versions of Minkowski spacetime (both versions

are used in physics).
In general, if both s and t are non-zero, we call η a pseudo- or semi-
Euclidean scalar product.
A set of vectors v 1, …, v s , w 1, …, w t in with

is called an orthonormal basis . Occasionally we write for

called the (η)-norm squared . We usually try to avoid this notation, because
it suggests that this norm is non-negative, which may not be the case.

Example 6.1.3 On the vector space we consider the non-degenerate

standard symmetric complex bilinear form q, given in the standard basis e 1,
…, e d by

The following is a well-known result from linear algebra.

Proposition 6.1.4 Every non-degenerate symmetric bilinear form Q on an -

or -vector space V is isomorphic to precisely one of these examples. In
particular, every non-degenerate symmetric bilinear form on an -vector space
V has a well-defined signature (s, t).

Remark 6.1.5 Since we can multiply a vector in a -vector space by i and the
form Q is complex bilinear, we can change Q(v, v) = +1 to Q(iv, iv) = −1. This
explains why there is no signature for symmetric bilinear forms on complex
vector spaces.

Definition 6.1.6 Let V be a -vector space with a non-degenerate symmetric

bilinear form Q. Then the (pseudo-)orthogonal group of (V, Q) is defined as the
automorphism group of Q:

In the following we will only consider the case of the standard symmetric
bilinear form η on of signature (s, t), where we write

The groups O(1, t) and O(s, 1) are called Lorentz groups.

If we write a matrix in the block form

(6.1)
with

then A O(s, t) if and only if

Taking the determinant on both sides of this equation shows that:

Lemma 6.1.7 Matrices A O(s, t) satisfy detA = ±1.

If we write a matrix in the form

with , then A O(s, t) if and only if

i.e. the vectors v 1, …, v s , w 1, …, w t form an orthonormal basis for ,

generalizing a well-known property of orthogonal matrices.

Lemma 6.1.8 There exists a canonical isomorphism .

Proof This is Exercise 6.13.1. □

We note the following facts:

Proposition 6.1.9 (Properties of Pseudo-Orthogonal Groups)

1. The group O(s, t) is a linear Lie group.

2. If both s, t ≠ 0, then O(s, t) is not compact.

3. Let denote the Lie algebra of O(s, t). Then as complex Lie algebras

In particular,

Proof

1. A similar argument to the one in Theorem 1.2.17 shows that the group
O(s, t) is a closed subgroup of , hence a linear Lie group by
Cartan’s Theorem 1.1.44.

2. This is Exercise 6.13.2.

3. This follows for a reason similar to the one explained in Remark 6.1.5.
□

It is also clear that , the standard orthogonal group. This

group has two connected components and the connected component of the
identity is SO(n), the group of orthogonal matrices of determinant + 1.
We would like to determine the number of connected components of the
pseudo-orthogonal group O(s, t) for general values of s and t. Let V + and V − be
the vector subspaces of defined by
Then η is positive definite on V + and negative definite on V −. Let π: V → V
+ denote the projection along V −.
Suppose that W V is any maximally η-positive definite vector subspace of
dimension s. Then

is an isomorphism, because any non-zero element in the kernel must have

negative η-norm squared and hence cannot be an element of W.

Definition 6.1.10 Fix an orientation on the vector subspace V +. Then there

exists for every maximally positive definite vector subspace W V a unique
orientation so that the isomorphism π | W is orientation preserving.

Suppose A O(s, t). Since A preserves η, it maps (maximally) positive definite

subspaces of V to (maximally) positive definite subspaces.

Definition 6.1.11 We define the time-orientability of A O(s, t) to be + 1 or

− 1 depending on whether

preserves or does not preserve orientation, where the orientation on A(V +) is

chosen via the projection π as above.

Remark 6.1.12 It might seem more natural to call this notion space-
orientability, because often a positive definite subspace of V is called spacelike.
We could then define time-orientability as the corresponding notion with V +
replaced by V −. However, as mentioned above, depending on the convention, 4-
dimensional Minkowski spacetime in quantum field theory can have signature
(+, −, −, −), so that time caries the plus sign. Furthermore, it follows from
Lemma 6.1.18 that if detA = 1 (the only situation in which we are going to
consider time-orientability) time-orientability and space-orientability are
equivalent. Since the term time-orientability is much more common in physics,
we will continue to use it.

Lemma 6.1.13 Let A O(s, t) and W V an arbitrary maximally positive

definite subspace. Then A has time-orientability + 1 if and only if
preserves orientation, with orientation on W and A(W) chosen via the projection
π as above.

Proof The proof is a deformation argument. Suppose that A O(s, t) and W

V is a maximally positive definite subspace. Since π | W : W → V + is an
isomorphism, we can find a unique basis w 1, …, w s for W of the form

where v i − are elements of V −. The fact that W is positive definite means that

where

By construction, α, α < 0.
For τ [0, 1] consider the following subspace of V:

It is easy to see that W τ is maximally positive definite for all τ [0, 1] and
W 0 = V +, W 1 = W. Let

and consider the following commutative diagram:

We know that π restricted to W τ and A(W τ ) is an orientation preserving

isomorphism to V + for all τ [0, 1], because these subspaces are maximally
positive definite. Since A τ is an isomorphism, it follows that B τ is an
isomorphism for all τ [0, 1], i.e. an element of GL(V +). Moreover, A has
time-orientability + 1 if and only if B 0 has positive determinant. Since the
determinant is continuous and B τ is a continuous curve in GL(V +), it follows
that B 0 has positive determinant if and only if B 1 has positive determinant. This
proves the claim. □
The construction of the family W τ in the proof shows the following:

Lemma 6.1.14 The set of all maximally positive definite subspaces W V

forms a contractible subset of the Grassmannian Gr s (V ).

Lemma 6.1.13 implies:

Proposition 6.1.15 If A, B O(s, t) both have time-orientability + 1, then the

same holds for AB and A −1.

Definition 6.1.16 We call

respectively,

the proper or special pseudo-orthogonal group,

the orthochronous pseudo-orthogonal group, and

the proper orthochronous pseudo-orthogonal group .

These subsets are indeed subgroups of O(s, t): this is clear for SO(s, t)
and follows for O+(s, t) from Proposition 6.1.15.

In particular, for s = 1 or t = 1 these groups are called

the proper Lorentz group ,

the orthochronous Lorentz group, and

the proper orthochronous Lorentz group .

If we write a matrix A O(s, t) as in Eq. (6.1) in the block form

then

The definition of time-orientability also works for O(n), in which case the time-
orientability of A is equal to the determinant of A and

If η is indefinite, then the three corresponding groups are not identical.

A similar argument to the proof of Theorem 1.2.22 in Sect. 3.8.3 shows (see
Exercise 6.13.3):

Proposition 6.1.17 (Connected Component of the Identity of Pseudo-

Orthogonal Groups) The subgroup SO+(s, t) is the connected component of
the identity in O(s, t).

If both s and t are non-zero, this implies that O(s, t) has precisely four connected
components.
Replacing in Definition 6.1.11 V + by V − we can define a new time-
orientability using an orientation on V −. Since the connected component of the
identity has to be the same whether we define the time-orientability via
orientations on V + or V −, we conclude that:

Lemma 6.1.18 The subgroup SO+(s, t) can also be characterized by

This also follows from the isomorphism in Lemma 6.1.8.

Remark 6.1.19 It can be shown that as a smooth manifold SO+(s, t) is

diffeomorphic to (if s, t ≥ 1), because SO(s) × SO(t) is the
maximal compact subgroup of SO+(s, t) (cf. [13, 83]). This also determines the
fundamental group of SO+(s, t).

Example 6.1.20 For applications concerning the Standard Model, the most
important of these Lie groups is the proper orthochronous Lorentz group
of 4-dimensional Minkowski spacetime. This is a connected,
non-compact Lie group of dimension 6. As a smooth manifold it is
diffeomorphic to and has fundamental group .

6.2 Clifford Algebras

We saw in Sect. 6.1 that a vector space (V, Q) with a (non-degenerate) symmetric
bilinear form defines a canonical Lie group O(V, Q) and there is an associated
Lie algebra . It is less obvious that the vector space (V, Q) also defines
another canonical algebraic object, an associative algebra Cl(V, Q), called the
Clifford algebra of (V, Q).

Definition 6.2.1 Let . An associative -algebra with unit element 1

is a -vector space A of finite dimension together with a bilinear, associative
product

and an element 1 A such that 1 ⋅ a = a = a ⋅ 1 for all a A. In particular, the

product on A is distributive and associative, but in general not commutative. The
direct sum of associative algebras A, B with unit element is defined as the vector
space A B with the product

The tensor product of associative algebras A, B with unit element is defined as

the vector space with the product

Definition 6.2.2 A homomorphism between -algebras A, B with unit

elements is a -linear map ϕ: A → B such that ϕ(1) = 1 and

An isomorphism is a bijective homomorphism. An automorphism is an

isomorphism ϕ: A → A. A representation of A on a vector space V is a
homomorphism ϕ: A → End(V ) into the endomorphism algebra of V. A
representation is called faithful if the homomorphism ϕ: A → End(V ) is
injective.

Definition 6.2.3 If C is an associative algebra and a, b C, we define the

commutator and anticommutator by

Definition 6.2.4 Let V be a -vector space with a symmetric bilinear form

Q. A Clifford algebra of (V, Q) is a pair (Cl(V, Q), γ), where

1. Cl(V, Q) is an associative -algebra with unit element 1.

2. Clifford relation: γ: V → Cl(V, Q) is a linear map with

3. Universal property: If A is some other associative -algebra with

unit element 1 and δ: V → A a -linear map with

then there exists a unique algebra homomorphism ϕ: Cl(V, Q) → A

such that the following diagram commutes:

Remark 6.2.5 We can think of the linear map γ as a linear square root of the
symmetric bilinear form − Q: in the definition of Clifford algebras, it suffices to
demand that
because, considering this equation for vectors v, w, v + w, the equation

follows. The element γ(v) in the Clifford algebra associated to a vector v V is

thus a “square root” of − Q(v, v) ⋅ 1, depending linearly on v.

Remark 6.2.6 Clifford algebras arose in physics in the work of P.A.M. Dirac,
who tried to find a “square root” of the Laplacian Δ, i.e. a differential operator D
on a pseudo-Euclidean vector space of dimension n = s + t such that

where η i = η(e i , e i ) in an orthonormal basis e 1, …, e s+t .

If we expand formally

it follows that

This implies that the symbols γ(e i ) have to satisfy

i.e. the relation of the Clifford algebra. We conclude that if we have a

representation of the Clifford algebra of on a vector space Δ, then we can
define a Dirac operator D, a square root of the Laplacian, for maps on with
values in Δ.

Remark 6.2.7 The Clifford relation has another important consequence,

that we will discuss in detail in Sect. 6.5: if v V is a vector with Q(v, v) =
±1, then for all x V

We will see in Corollary 6.2.18 that the linear map γ is always injective and
we can identify V with its image γ(V ). This implies that the vectors v V
with Q(v, v) = ±1 act on V as reflections in the hyperplane v and thus
 arbitrary products of such vectors act as pseudo-orthogonal transformations.
Conversely, demanding that ±γ(v) ⋅ γ(x) ⋅ γ(v) is the reflection in v almost
inevitably leads to the Clifford relation.
Note that v and − v define the same reflection in v . Since the
expression ±γ(v) ⋅ γ(x) ⋅ γ(v) depends quadratically on v, we can think of
γ(v) as a “square root” of the reflection in v . This construction, together
with the Cartan–Dieudonné Theorem on reflections, will enable us to
define the spin groups, certain double coverings of the pseudo-orthogonal
groups, essentially using square roots of pseudo-orthogonal
transformations.

6.2.1 Existence and Uniqueness of Clifford Algebras

Our first aim is to prove the existence and uniqueness of Clifford algebras.

Theorem 6.2.8 (Existence of Clifford Algebras) For every finite-

dimensional -vector space V with a symmetric bilinear form Q there exists a
Clifford algebra (Cl(V, Q), γ).

Proof We denote by T(V ) the tensor algebra of V:

Let I(Q) denote the two-sided ideal in T(V ) generated by the set

The tensor algebra T(V ) is determined by V alone (and infinite-dimensional),

whereas the ideal I(Q) depends also on the symmetric bilinear form Q.
We set

Then Cl(V, Q) is an associative algebra with product

If Q ≡ 0, it follows that the Clifford algebra is the exterior algebra Λ V. In

general, let

and
denote the canonical embedding and projection. Then

is linear and satisfies the identity

This implies by polarization

Since the vector space V generates T(V ) by taking tensor products, it follows
that the image γ(V ) generates Cl(V, Q) multiplicatively.
By definition of the tensor algebra, every linear map

to an associative algebra A extends to an algebra homomorphism

If δ satisfies the identity

then I(Q) is a subset of ker Δ and the map Δ descends to a homomorphism

with ϕ ∘γ = δ. Given A and δ, the homomorphism ϕ with ϕ ∘γ = δ is uniquely

determined, because γ(V ) generates Cl(V, Q) multiplicatively and ϕ is fixed on
the image γ(V ). □

Corollary 6.2.9 (Uniqueness of Clifford Algebras) If the associative

algebras (Cl(V, Q), γ) and are both Clifford algebras for the same
vector space (V, Q), then there exists a unique algebra isomorphism

so that f ∘γ = γ ′ .

Proof This follows from the universal property of Clifford algebras. □

From the proof of Theorem 6.2.8 we see:

Corollary 6.2.10 The image of the vector space V under γ generates Cl(V, Q)
multiplicatively.

Corollary 6.2.11 If Q ≡ 0, then there exists an algebra isomorphism

where γ is given by the standard embedding of V into Λ V.

Corollary 6.2.12 Suppose that and e 1, …, e n is an orthonormal

basis for (V, Q). Then the set of elements of the form

where 1 ≤ i 1 < i 2 < … < i k ≤ n, 0 ≤ k ≤ n, and the empty product for k = 0 is

equal to 1, span Cl(V, Q) as a vector space. This implies that

Proof The elements of γ(V ) generate Cl(V, Q) multiplicatively, hence the

collection of all products of the basis vectors {γ(e i )} span Cl(V, Q) as a vector
space. If in such a product the same vector γ(e i ) appears twice, then we can
cancel it (after possibly permuting the vectors) because of the Clifford relation. □

Example 6.2.13 The simplest non-trivial example of a Clifford algebra is the

Clifford algebra over a 1-dimensional vector space . Let e be a non-zero
element of . Then is 2-dimensional and is spanned as a -vector
space by {1, γ(e)} with

Definition 6.2.14 Let T 0(V ) and T 1(V ) denote the vector subspaces of
elements of the tensor algebra of even and odd degree. We set

and call these vector subspaces the even and odd part of the Clifford algebra .
Since

it follows that

and Cl(V, Q) has the structure of a -graded associative algebra or

superalgebra :

In particular, Cl0(V, Q) is a subalgebra of Cl(V, Q) and is spanned by products v

1⋯v 2k of an even number of vectors v i V.

6.2.2 Clifford Algebras and Exterior Algebras

We saw in Corollary 6.2.11 that for Q ≡ 0 the Clifford algebra Cl(V, Q) is
isomorphic to the exterior algebra Λ V with the wedge product. In this section
we will show that for an arbitrary symmetric bilinear form Q the Clifford algebra
Cl(V, Q) is still (canonically) isomorphic to the exterior algebra as a vector
space. The multiplication in Cl(V, Q) can thus be thought of as a deformation
(depending on Q) of the wedge product on Λ V (we follow the exposition in
[20]).
Let V be a finite-dimensional -vector space with a symmetric bilinear form
Q.

Definition 6.2.15 For v V and σ Λ k V there is a unique (k − 1)-form

, called the contraction of v and σ, with the following properties:

1. if σ V, then ;

2. for all σ Λ k V, ω Λ l V we have

The proof is left as an exercise.

Theorem 6.2.16 (The Clifford Algebra Is Isomorphic to the Exterior

Algebra as a Vector Space) There exists a canonical isomorphism of vector
spaces

In any orthonormal basis e 1, …, e n of (V, Q) this isomorphism is given by

In particular, the dimension of the Clifford algebra is

where .

Proof We define a linear map

by

We have

By the universal property of Clifford algebras, the linear map δ extends to an

algebra homomorphism

Consider the linear map

If e 1, …, e n is an orthonormal basis for (V, Q), then

In particular, f is surjective. Corollary 6.2.12 then implies that f is an

isomorphism of vector spaces. □

Remark 6.2.17 The linear map

is called the symbol map in [15], its inverse f −1 the quantization map.

Corollary 6.2.18 (Linear Map γ Is an Embedding) Let (Cl(V, Q), γ) be a

Clifford algebra. Then the linear map

is injective and we can therefore identify V with its image under γ.

Remark 6.2.19 Since we now know the dimension of Clifford algebras, the
universal property can be used to find isomorphisms of the Clifford algebra
Cl(V, Q) to other associative algebras A as follows:
Find a linear map δ: V → A that satisfies

It then induces an algebra homomorphism ϕ: Cl(V, Q) → A.

 Let e 1, …, e n be an orthonormal basis of V. Show that the products of the
images δ(e i ) span A. Then ϕ is surjective.
Suppose that the algebras Cl(V, Q) and A have the same dimensions. Then ϕ
is an algebra isomorphism.
We will use this strategy several times in the following sections, for instance,
in Lemma 6.3.2, Lemma 6.3.3 and Lemma 6.3.20.

Here is an application of Remark 6.2.19.

Lemma 6.2.20 The linear map

induces an algebra automorphism

with α 2 = 1. The subspace Cl j (V, Q) is equal to the (−1) j -eigenspace of α. In

particular,

6.3 The Clifford Algebras for the Standard Symmetric

Bilinear Forms
Definition 6.3.1

1. For the standard vector space we denote the Clifford algebra

by Cl(s, t). This is a real associative algebra. For (s, t) = (n, 0) we also denote
the Clifford algebra by Cl(n).

2. For the standard vector space we denote the Clifford algebra

by (note that q is the standard non-degenerate complex bilinear form
and not a Hermitian form). This is a complex associative algebra.

Lemma 6.3.2 (Complexified Clifford Algebra) There exists an isomorphism

of complex associative algebras
Complex representations of Cl(s, t) are equivalent to complex representations of
.

Proof The complex linear map

satisfies

The claim follows from Remark 6.2.19. □

Lemma 6.3.3 (Even Complex Clifford Algebra) For any n ≥ 1

Proof Let e n be the n-th standard basis vector of and define

Then δ satisfies

The claim follows from Remark 6.2.19 and the dimension formula in
Lemma 6.2.20. □

6.3.1 Gamma Matrices

The Clifford algebra Cl(V, Q) is generated multiplicatively by the subspace γ(V ).
It is therefore important to know the elements γ(e i ) for a basis e 1, …, e n of V.

Definition 6.3.4 Let with standard basis e 1, …, e n .

Suppose

is a representation of Cl(V, Q) on a -vector space . Then we define

 for a = 1, …, n the mathematical gamma matrices by

and the physical gamma matrices by

The anticommutators are given by

where I N denotes the N × N-unit matrix. We also set

6.3.2 The Chirality Operator in Even Dimensions

For an even-dimensional real vector space, a choice of orientation defines an
important element in the Clifford algebra.
Suppose that n = s + t = 2k is even. A chirality element for Cl(s, t) is a
Clifford element of the form

where e 1, …, e n is an oriented orthonormal basis of and λ is a complex

constant, determined below.

Lemma 6.3.5 A chirality element ω does not depend on the choice of oriented
orthonormal basis e 1, …, e n .

Proof Suppose that

with a matrix A SO(s, t). Under the vector space isomorphism

from Theorem 6.2.16 we have

However
Hence e 1 ′ ⋯e n ′ = e 1⋯e n . □

Lemma 6.3.6 Every chirality element ω satisfies

for all 1 ≤ a, b ≤ n.

Proof The first equation follows from

since n is even. The second equation is a consequence of the first. □

Lemma 6.3.7 If λ 2 = (−1) k+t , then the chirality element satisfies

Proof

where we used that

Different choices for λ with λ 2 = (−1) k+t are possible and several different
choices appear in the literature. The simplest choice is probably

We will use in the following this choice up to a sign (because of certain

conventions for Weyl spinors, described later) and set:
Definition 6.3.8 The chirality element in even dimension n = s + t = 2k for
Cl(s, t) is defined by

If k + t is even, then ω is an element of the real Clifford algebra Cl(s, t).

If γ a are mathematical gamma matrices in a complex representation of
Cl(s, t), then the mathematical chirality operator is defined by

If Γ a are physical gamma matrices, then the physical chirality operator is

defined by

We have γ n+1 = (−1) k Γ n+1. The chirality operators do not depend on the choice
of oriented orthonormal basis for .

Remark 6.3.9 By analogy, we can define a chirality element for the complex
Clifford algebra by

where e 1, …, e 2k is an orthonormal basis of . Taking an orthonormal basis

of and the corresponding orthonormal basis

of , we see that the chirality elements for the real and complex Clifford
algebra coincide under the isomorphism .

6.3.3 Raising Indices of Gamma Matrices

Let Γ 1, …, Γ n be physical gamma matrices . We set

and similarly for the mathematical γ-matrices (in the first equation there is an
implicit sum over c; this is an instance of the Einstein summation convention).
These matrices satisfy by Lemma 6.3.6

6.3.4 Examples of Clifford Algebras in Low Dimensions

In the following examples we use the Pauli matrices

It is easy to check that they satisfy the identities

where in the second equation j + 1 and j + 2 are taken .

Example 6.3.10 The Clifford algebra Cl(1, 0) is spanned as a real vector space
by elements 1, γ(e 1) with

It follows that Cl(1, 0) is isomorphic as a real algebra to the 2-dimensional

algebra with γ(e 1) = i.

Example 6.3.11 The Clifford algebra Cl(0, 1) is spanned as a real vector space
by elements 1, γ(e 1) with

It follows that Cl(0, 1) is isomorphic as a real algebra to the 2-dimensional

algebra with multiplication (a, b) ⋅ (a ′ , b ′ ) = (aa ′ , bb ′ ), unit element 1 =
(1, 1) and γ(e 1) = (1, −1).

Example 6.3.12 The algebra is spanned as a vector space by 1 = (1, 1)

and γ(e 1) = (i, −i). It follows that

Example 6.3.13 The matrices

satisfy

Therefore γ 1, γ 2 are mathematical gamma matrices for Cl(2, 0) and generate a

faithful representation of this Clifford algebra on . We can identify γ(e i ) with
γ i . The Clifford algebra is 4-dimensional and is spanned as a real vector space
by {1, γ 1, γ 2, γ 1 γ 2}.

Example 6.3.14 The matrices

satisfy

Therefore γ 1, γ 2 are mathematical gamma matrices for Cl(1, 1) and generate a

faithful representation of this Clifford algebra on . Again, the Clifford algebra
is spanned as a real vector space by {1, γ 1, γ 2, γ 1 γ 2}.

Example 6.3.15 Note that the matrices

span as a complex vector space. It follows that

Example 6.3.16 The matrices

satisfy
They are mathematical gamma matrices for Cl(4, 0), i.e. for the Clifford algebra
of Euclidean space of dimension 4. Another choice for the same Clifford algebra
is

The γ a generate a faithful representation of Cl(4, 0) on . We identify γ(e i )

with γ i . As a real vector space Cl(4, 0) has dimension 16 and is spanned by

Example 6.3.17 The matrices

satisfy

They are physical gamma matrices for Cl(1, 3), i.e. for the Clifford algebra of
Minkowski spacetime with signature (+, −, −, −), in the so-called Weyl
representation or chiral representation . The associated mathematical gamma
matrices γ a = iΓ a satisfy

The γ a generate a faithful representation of Cl(1, 3) on . As a real vector space

Cl(1, 3) has dimension 16 and is spanned by the matrices corresponding to the
ones above.

Example 6.3.18 Let Γ a and γ a = iΓ a be the physical and mathematical gamma

matrices for Cl(1, 3) considered in Example 6.3.17. If we set

then these are physical and mathematical gamma matrices for the Clifford
algebra Cl(3, 1) of Minkowski spacetime with signature (−, +, +, +).

Example 6.3.19 A similar argument to the one in Example 6.3.15 shows that

6.3.5 The Structure of the Standard Clifford Algebras

Lemma 6.3.20 (Complex Clifford Algebras Are Periodic) The complex
Clifford algebras satisfy the periodicity

Here denotes the standard tensor product of associative algebras.

Proof We follow [20]. Write as and decompose an element of

accordingly as (x, y). Let e 1, e 2 be the standard basis of and ω = −ie 1 e
2 the corresponding chirality element. Define the linear map

Then

because ω 2 = 1 and {ω, y} = 0. The first isomorphism then follows by

Remark 6.2.19. The second isomorphism follows from Example 6.3.15. □

Together with Example 6.3.12, Example 6.3.15 and Lemma 6.3.3 this implies
the following structure theorem.
Theorem 6.3.21 (Structure Theorem for Complex Clifford Algebras) As
complex algebras the complex Clifford algebra and its even part are given by
Table 6.1 .
Table 6.1 Complex Clifford algebras
n N

Even 2 n⁄2

Odd 2(n−1)⁄2

Example 6.3.22 In dimension n = 4 we have

Without proof we mention the following theorem (see [28, 40, 49]).

Theorem 6.3.23 (Structure Theorem of Real Clifford Algebras) The

structure of the real Clifford algebras Cl(s, t) and Cl0(s, t) is given by Tables 6.2
and 6.3 , where we set ρ = s − t and n = s + t and all endomorphism algebras are
understood as real algebras.
Table 6.2 Real Clifford algebras

Cl(s, t) N

0 2 n⁄2

1 2(n−1)⁄2

2 2(n−2)⁄2

3 2(n−3)⁄2

4 2(n−2)⁄2

5 2(n−1)⁄2

6 2 n⁄2
7 2(n−1)⁄2

Table 6.3 Even part of real Clifford algebras

Cl0(s, t) N

0 2(n−2)⁄2

1 2(n−1)⁄2

2 2(n−2)⁄2

3 2(n−3)⁄2

4 2(n−4)⁄2

5 2(n−3)⁄2

6 2(n−2)⁄2

7 2(n−1)⁄2

Example 6.3.24 For Minkowski spacetime in dimension 4 we have

6.4 The Spinor Representation

Definition 6.4.1 The vector space of (Dirac) spinors is , where N is
given by the values in Table 6.1. The (Dirac) spinor representation of the
complex Clifford algebra

defined by the structure theorem of the complex Clifford algebras , is given

by Table 6.4. There are induced complex spinor representations of Cl(s, t).
Table 6.4 Complex spinor representation

n Representation
Even

Odd

Notice that the space Δ n of Dirac spinors is a complex vector space, whose
dimension is always even and grows exponentially with n.

Definition 6.4.2 The bilinear map

is called mathematical Clifford multiplication of a spinor with a vector.

Similarly, physical Clifford multiplication is given by (−i) times mathematical
Clifford multiplication. More generally, via the isomorphism of vector spaces

from Theorem 6.2.16, followed by the complex spinor representation, we can

define a (mathematical) Clifford multiplication of spinors with forms .

Example 6.4.3 Suppose that the restriction of the Dirac spinor representation to
is given by physical gamma matrices Γ a and mathematical
gamma matrices γ a = iΓ a in . Then physical Clifford multiplication
of a basis vector with a spinor is equal to

whereas mathematical Clifford multiplication is equal to

For the following result, recall that according to Lemma 6.3.3 there is an
isomorphism
 Corollary 6.4.4 (Induced Spinor Representation on the Even Clifford
Algebra) Consider the restriction of the Dirac spinor representation to the
even subalgebra .

If n is odd, then the induced representation is irreducible:

If n is even, then the induced representation splits into left-handed

(positive) and right-handed (negative) Weyl spinors :

See Table 6.5 .

Table 6.5 Weyl spinor representations

n Induced representation Spinor space N

Even 2 n⁄2

Odd 2(n−1)⁄2

If n is odd we identify here with the first summand in

so that the restriction of the spinor representation is non-trivial.

The result in the even-dimensional case can be clarified as follows.

Proposition 6.4.5 (Weyl Spinors and Chirality in Even Dimensions) Let n

= 2k be even, Δ n the Dirac spinor representation and Γ n+1 the chirality
operator.

1. Δ n ± can be identified with the (±1)-eigenspaces of Γ n+1 on Δ n .

2. The induced representation of maps Δ n ± to itself, while elements in

(such as vectors in ) map Δ n ± to Δ n . It follows that
Proof We can split the space into the (±1)-eigenspaces of Γ n+1,
because Γ n+1 2 = 1. We call these spaces Δ n ±. We have

by Lemma 6.3.6. This shows that Δ n ± are invariant under the representation of
and also implies the second claim in 2. The final claim then follows from
.□

Remark 6.4.6 Note that the definition of positive and negative Weyl spinors
depends on the sign of the chirality operator, which can be chosen arbitrarily.
Moreover, in the literature sometimes positive Weyl spinors are called right-
handed and negative Weyl spinors left-handed. We continue to use our
conventions.

6.5 The Spin Groups

In this section we want to discuss the spin groups, which are certain subgroups
embedded in the Clifford algebra. Recall how we defined linear Lie groups in
Chap. 1: for we started with the associative endomorphism algebra

and considered the group of invertible elements

which is an open subset of . The linear Lie groups were then defined as
closed subgroups of the Lie groups .
To define spin groups we will follow a similar approach, where we replace
the endomorphism algebra by the Clifford algebra Cl(s, t). It turns out
that spin groups are certain double coverings of (pseudo-)orthogonal groups. In
particular, the Lie algebra given by the Clifford algebra with the canonical
commutator of associative algebras

contains as a Lie subalgebra the Lie algebra of the pseudo-orthogonal group

SO+(s, t). We follow in this section [13] and [55].
6.5.1 The Pin and Spin Groups
Definition 6.5.1 The group of invertible elements in the Clifford algebra
Cl(s, t) is defined by

There is an analogous definition for .

Lemma 6.5.2 The group Cl×(s, t) is an open subset of Cl(s, t) and therefore a
Lie group.

Proof Let n = s + t. According to Lemma 6.3.2

Decomposing into y = u + iv with u, v Cl(s, t) it follows that

However, Theorem 6.3.21 on the structure of complex Clifford algebras implies

that is an open subset of for all n, because the general linear group
is open in . This implies that the intersection is
open in Cl(s, t). □

Definition 6.5.3 We define the following subsets of :

Here η is the standard symmetric bilinear form of signature (s, t).

Proposition 6.5.4 The following subsets of Cl(s, t) form subgroups of the

group Cl×(s, t):

We endow these subsets with the subset topology from the vector space Cl(s, t).
We also set
We call the group Pin(s, t) the pin group , the group Spin(s, t) the spin group and
the group Spin+(s, t) the orthochronous spin group .

Proof The proof is not difficult and left as an exercise. Note that in the
definition of Spin+(s, t) both numbers 2p and 2q are even. □

Remark 6.5.5 In the literature the group Spin+(s, t) is also often simply called
the spin group. We added the adjective orthochronous to distinguish it from the
spin group Spin(s, t).

Definition 6.5.6 For an element u = v 1 v 2⋯v r Pin(s, t) we set

Hence

Definition 6.5.7 We identify the vector space in the canonical way via
the embedding γ with a vector subspace of Cl(s, t) and consider the following
map:

We prove in Lemma 6.5.10 below that this map is well-defined and yields a
continuous homomorphism

Remark 6.5.8 Notice that the definition of the map R is very similar to the
definition of the maps

and

from Exercise 1.9.20 and Exercise 1.9.21.

Remark 6.5.9 It will follow from Theorem 6.5.13 that

Hence the degrees and of an element u Pin(s, t) measure whether

λ(u) preserves orientation or time-orientation.

Lemma 6.5.10

1. The map R is well-defined, i.e. it has image in the subspace of the

Clifford algebra.

2. For any vector v S ± s, t the map

is the reflection in the hyperplane .

3. The map R yields a continuous homomorphism

Proof Fix a vector v R s, t with η(v, v) = ±1. Then v −1 = v and

We get
hence and R v is reflection in v . Since

we conclude that R(u, x) is indeed an element of for all u Pin(s, t) and

. Since reflections are elements of the orthogonal group, it follows that λ
is a continuous homomorphism to the orthogonal group. □

We see that vectors with η(v, v) = ±1 act as reflections on and a

general element a of the pin group acts as a composition of such reflections,
hence as an element of the pseudo-orthogonal group. Note that the vectors ± v
define the same reflections, hence elements ± a of the pin group define the
same element of the pseudo-orthogonal group. As we will see in
Theorem 6.5.13, this implies that the pin group is a double covering of the
pseudo-orthogonal group.

We need the following algebraic theorem.

Theorem 6.5.11 (Cartan–Dieudonné Theorem) Every element of O(s, t) can

be written as a composition of at most 2(s + t) reflections in hyperplanes v i
with vectors v i S ± s, t .

Proof The idea is to set for an element A O(s, t)

with the standard basis e 1, …, e s+t . We can then find a composition R of

reflections which map the orthonormal basis a 1, …, a s+t to e 1, …, e s+t ,
implying that A = R: we first find a composition R 1 of reflections that maps

as follows:

1. If a 1 = e 1, then R 1 is the identity.

2. If the norm of a 1 − e 1 is non-zero, then R 1 is the reflection in the

hyperplane (a 1 − e 1) .
 If the norm of a 1 − e 1 is zero, then R 1 is the composition of the reflection
3.
in the hyperplane (a 1 + e 1) followed by the reflection in the hyperplane e
1 .

4. In the second and third case we normalize the vectors a 1 − e 1 and a 1 + e 1

to norm ± 1.
Since R 1(a 1) = e 1, it follows that R 1 maps

To prove the claim we then proceed by induction on s + t. □

Theorem 6.5.12 (Special Orthogonal Groups and Reflections) Let R

O(s, t) be a composition of reflections in hyperplanes v i with vectors v i S±
s, t .

1. R is an element of SO(s, t) if and only if the number of vectors v i is even.

2. R is an element of SO+(s, t) if and only if both the number of vectors v i S

s, t and the number of vectors v S − s, t are even.
+ i

Proof Notice that for any vector v S ± s, t we can split into maximally
positive and negative definite subspaces W + W − so that, with respect to a
suitable basis,
This implies that an even number of reflections in hyperplanes v with η(v, v) =
±1 is in SO(s, t), i.e. has determinant 1. If in addition the number of reflections in
hyperplanes v with η(v, v) = +1 is even, then the map is in SO+(s, t). See [13,
Theorem 1.5] for more details. □

Theorem 6.5.13 (Relation Between the Pin and Spin Groups and the
Orthogonal Groups) Consider the homomorphism

1. The homomorphism λ is open and surjective with kernel equal to { ± 1}.

2. The preimages under λ of the subgroups SO(s, t) and SO+(s, t) are equal to
Spin(s, t) and Spin+(s, t), which are therefore open subgroups of Pin(s, t).

3. The homomorphism λ restricts to surjective homomorphisms

 with kernel equal to { ± 1}.

4. As a topological space the orthochronous spin group Spin+(s, t) is connected

if s ≥ 2 or t ≥ 2.

Proof The statement that λ is surjective and open follows from Theorem 6.5.11
and its proof. We show that the kernel of λ is equal to { ± 1}: Suppose that

Then , since R u has to be composed of an even number of reflections.

It follows that

and therefore
(6.2)
Expanding u in the standard basis for the Clifford algebra, suppose that

with k ≥ 1 and . Applying Eq. (6.2) with i = i 2k , it follows that a = 0. This

implies that . Since u Pin(s, t), we conclude that u = ±1.
Statements 2. and 3. then follow from Proposition 6.5.12.
We finally show that Spin+(s, t) is connected if s ≥ 2 or t ≥ 2. Since SO+(s, t)
is connected and the kernel of λ is { ± 1}, it suffices to show that every u
Spin+(s, t) can be connected to − u by a continuous curve in Spin+(s, t). Suppose
that s ≥ 2 and consider the curve

where . It is easy to check that this is indeed a curve in Spin+(s, t) with

There is a similar argument in the case t ≥ 2. □

Since the homomorphism λ is continuous, open and surjective and has kernel
equal to { ± 1}, it follows that:

Corollary 6.5.14 We can define a unique Lie group structure on the groups
so that λ becomes a smooth double covering of Lie groups.

Remark 6.5.15 There is a more natural way to define a smooth structure on the
pin and spin groups: Using a different definition of the pin group, which can be
found in the classic paper [7] of Atiyah, Bott and Shapiro, it is possible to show
that Pin(s, t) is a closed subset of Cl×(s, t) and therefore by Cartan’s Theorem 1.1.
44 an embedded Lie subgroup. Theorem 6.5.13 then implies that Spin(s, t) and
Spin+(s, t) are also embedded Lie subgroups of Cl×(s, t). The Lie group structure
on these groups as closed subgroups of Cl×(s, t) coincides with the one from
Corollary 6.5.14.

Corollary 6.5.16 For all n ≥ 3 the homomorphisms

are the universal coverings.

Proof This follows, because according to Proposition 2.6.3 and Remark 6.1.19
for n ≥ 3 the Lie groups SO(n), SO+(n, 1) and SO+(1, n) have fundamental group
.□

Example 6.5.17 Exercises 1.9.20 and 1.9.21 imply that there exist
isomorphisms

Similarly, it can be shown that

See Sect. 6.8.2 and Exercise 6.13.17.

6.5.2 The Spinor Representation of the Orthochronous

Spin Group
Definition 6.5.18 We denote by
the spinor representation induced by restriction of the spinor representation of
the even Clifford algebra Cl0(s, t).

Proposition 6.5.19 The spinor representation is compatible with Clifford

multiplication in the following way:

for all g Spin+(s, t), and ψ Δ n.

Proof Let

denote the spinor representation. Then

6.5.3 The Lie Algebra of the Spin Group

In this subsection we determine the Lie algebra of Spin+(s, t) and calculate the
differential of the covering homomorphism λ from the orthochronous spin group
to the proper orthochronous orthogonal group.
The Lie group Cl×(s, t) is an open subset of Cl(s, t), hence its Lie algebra
is canonically isomorphic as a vector space to Cl(s, t) with commutator

Since Pin(s, t) and Spin+(s, t) are Lie subgroups of Cl×(s, t), it follows that the
Lie algebras and are Lie subalgebras of .

Definition 6.5.20 Let e 1, …, e s+t be an orthonormal basis of . We define

Lemma 6.5.21 The vector space M(s, t) is a Lie subalgebra of of

dimension
Proof This is Exercise 6.13.9. □

Proposition 6.5.22 (Description of the Lie Algebra of the Spin Group)

For all s, t ≥ 0 the following identity holds:

Proof We follow an idea in [99]. To show that e i e j for i < j is an element of

the Lie algebra of the spin group, it suffices to find a curve in Spin+(s, t) through
the unit element 1 with velocity vector e i e j . We set η i = η(e i , e i ) = | | e i | | η 2.

1. Suppose that i ≠ j and η i = η j . Then the curve

through 1 is contained in Spin+(s, t), because we can write

and

2. In the case i ≠ j and η i = −η j a similar argument shows that the curve

through 1 is contained in Spin+(s, t).

Taking the derivative in τ = 0 of both curves, it follows that

This implies that

However, the dimensions of both vector spaces agree and we conclude that
.□

Let E ij be the elementary (s + t) × (s + t)-matrix with a 1 at the intersection of

the i-th row and the j-th column and zeros elsewhere. We define the matrix
Corollary 6.5.23 (Differential of the Covering Homomorphism λ) The
differential of the homomorphism

is given by

with

for all . This implies the following explicit formula:

We can recover any from its image λ (z) via

Proof The formula λ (z)x = [z, x] follows from

The remaining formulas follow by direct calculation. □

The isomorphism λ between and will appear in the

formula for the spin covariant derivative in Sect. 6.10.2 and also, in a rather
unexpected way, in the discussion of the Grand Unified Theory with gauge
group Spin(10) in Sect. 9.5.5.

6.6 Majorana Spinors

In Sect. 6.6 and Sect. 6.7 we discuss two concepts – Majorana spinors and spin
invariant scalar products – that are less well-known in the mathematical
literature, but play quite an important role in physics. Majorana spinors are
related to so-called real or quaternionic structures for the spinor representation.
It turns out that in every dimension the spinor representation admits a real or
quaternionic structure, which in even dimensions may or may not be compatible
with the decomposition into Weyl spinors .

Definition 6.6.1 Let V be a complex vector space with a representation of a

Lie group G.

1. A real structure on V is a complex antilinear G-equivariant map σ: V → V

with σ 2 = Id.

2. A quaternionic structure on V is a complex antilinear G-equivariant map

J: V → V with J 2 = −Id.

Proposition 6.6.2 Let σ be a real structure on the complex G-representation

space V and V σ the real subspace

Then

hence we can decompose elements in V into a “real” and “imaginary” part. The
complex representation on V induces real representations of G on V σ and iV σ ,
which are both isomorphic (as real representations).

Proof We set

It is clear that V σ and V ′σ are real subspaces of V and

is a decomposition of a vector v V into vectors in V σ and V ′σ . Hence we can

decompose V = V σ V ′σ . Since σ is complex antilinear, we have

This proves V ′σ = iV σ . If g G and v V σ , then

because σ is G-equivariant. This shows that g ⋅ v V σ , hence G preserves V σ ,

and similarly V ′σ . It follows that the representation of G restricts to
representations on these real vector spaces. The map

is G-equivariant, because the G-representation on V is complex, and thus defines

a real isomorphism between the G-representations on V σ and V ′σ . □

Proposition 6.6.3 Let J be a quaternionic structure on the complex G-

representation space V and

Then I, J and K = IJ define the structure of a G-equivariant quaternionic vector

space on V.

Proof It is clear that I is G-equivariant, because the G-representation on V is

complex, and J is G-equivariant by definition. Hence K is also G-equivariant.
Moreover, I 2 = −1 and J 2 = −1. Since J is complex antilinear, we have JI = −IJ.
This implies that I, J, K satisfy the quaternionic identities. □

Definition 6.6.4 Let κ: Spin+(r, s) → GL(Δ) with Δ = Δ n be the complex spinor

representation.

1. The spinor representation Δ is called Majorana if it admits a real structure

σ. In this case there exists in Δ a real subspace

of half dimension

and the complex spinor representation on Δ induces a real representation of

Spin+(s, t) on Δ σ . Elements of Δ σ are called Majorana spinors. We can
decompose any ψ Δ uniquely as

with Majorana spinors ϕ 1, ϕ 2. Spinors α 1, …, α k which form a real basis

for Δ σ satisfy

2. The spinor representation Δ is called symplectic Majorana if it admits a

quaternionic structure J. In this case Δ has a Spin+(r, s)-equivariant structure
I, J, K = IJ of a quaternionic vector space. Elements of Δ are called
symplectic Majorana spinors. We can find a complex basis χ 1, λ 1, …, χ k , λ
 k ofΔ so that the spinors χ 1, …, χ k form a quaternionic basis of Δ and J is
given by

hence they are related by the standard symplectic matrix

This explains the name symplectic Majorana spinors.

Remark 6.6.5 In the physics literature (see, for instance, [54]) one writes for a
spinor ψ Δ

if σ is a real structure and

if J is a quaternionic structure for the spinor representation Δ. The spinor ψ C is

called the charge conjugate . Choosing a complex basis for the spinor space Δ,
we can identify . Real and quaternionic structures then correspond to a
matrix B with
(6.3)
where denotes the complex conjugate, B satisfies B B = I N for a real and B
B = −I N for a quaternionic structure (with the unit matrix I N ), and Spin+(s, t)-
equivariance corresponds to

where κ: Spin+(s, t) → GL(Δ) denotes the spinor representation.

Recall that in even dimensions the complex spinor representation Δ splits into
the complex Weyl spinors Δ = Δ + Δ −. We denote by π +: Δ → Δ + the
projection along Δ −.

Definition 6.6.6 Let κ: Spin+(r, s) → GL(Δ) with Δ = Δ n be the complex spinor

representation over an even-dimensional vector space.

1. The spinor representation Δ is called Majorana–Weyl if it admits a real

 structure σ that commutes with π +. In this case σ induces on both Weyl
spinor spaces Δ ± a real structure. The elements of

are called left-handed and right-handed Majorana–Weyl spinors.

2. The spinor representation Δ is called symplectic Majorana–Weyl if it

admits a quaternionic structure J that commutes with π +. In this case I, J, K
= IJ induce on both Weyl spinor spaces Δ ± the structure of a quaternionic
vector space. Elements of Δ ± are called left-handed and right-handed
symplectic Majorana–Weyl spinors.

It is known precisely in which dimensions and signatures these types of spinors

exist: For the standard scalar product η on of signature (s, t) we set again ρ =
s − t and n = s + t. We then have, without proof, Table 6.6 (see [40] and compare
with Theorem 6.3.23).
Table 6.6 Majorana spinors

Type of spinors Minimal real spinor representation

0 Majorana–Weyl Δ ± σ of dimension 2(n−2)⁄2

1 Majorana, not Weyl Δ σ of dimension 2(n−1)⁄2
2 Majorana and Weyl, but not Majorana–Weyl Δ σ and Δ of dimension 2 n⁄2
±
3 Symplectic Majorana, not Weyl Δ of dimension 2(n+1)⁄2
4 Symplectic Majorana–Weyl Δ ± of dimension 2 n⁄2
5 Symplectic Majorana, not Weyl Δ of dimension 2(n+1)⁄2
6 Majorana and Weyl, but not Majorana–Weyl Δ σ and Δ of dimension 2 n⁄2
±
7 Majorana, not Weyl Δ σ of dimension 2(n−1)⁄2

Example 6.6.7 For Minkowski spacetime of dimension n we have n

= ρ + 2. We see that in Minkowski spacetime of dimension 4 there exist both
Majorana and Weyl spinors of real dimension 4, but not Majorana–Weyl
spinors. Note that this implies that Majorana spinors in this dimension always
 have components both in Δ + and Δ −.
In Minkowski spacetime of dimension 2 and 10 (relevant in superstring
theory) there exist Majorana–Weyl spinors of real dimension 1 and 16,
respectively, and in Minkowski spacetime of dimension 11 (relevant in M-
theory) there exist Majorana spinors of real dimension 32.

6.7 Spin Invariant Scalar Products

In this section we will study complex bilinear and Hermitian scalar products on
the spinor space Δ, which are invariant under the action of the spin group. It
turns out that the combination of both types of scalar products is related to the
existence of real and quaternionic structures and thus to Majorana spinors.
Hermitian scalar products are particularly important, because we need them in
Chap. 7 to define Lorentz invariant Lagrangians involving spinors.

6.7.1 Majorana Forms

Definition 6.7.1 Let Δ = Δ n be the complex spinor representation of Cl(s, t).
We fix constants μ, ν = ±1 and consider complex bilinear, non-degenerate forms

with the following properties:

1. (X ⋅ ψ, ϕ) = μ(ψ, X ⋅ ϕ) for all and all ψ, ϕ Δ.

2. (ψ, ϕ) = ν(ϕ, ψ) for all ψ, ϕ Δ.

We call such a form a Majorana form .

Lemma 6.7.2 For a Majorana form let {χ α } be a complex basis of Δ and C

the matrix with entries

If we expand
and identify ψ, ϕ with the column vectors with entries ψ α , ϕ α , then

Furthermore, property 1. and 2. in Definition 6.7.1 are equivalent to

1. γ a T = μCγ a C −1 for all a = 1, …, s + t.

2. C T = νC.
The first equation also holds with the physical Clifford matrices Γ a instead
of the mathematical gamma matrices γ a .

Proof This is left as an exercise. □

Definition 6.7.3 The matrix C is called the charge conjugation matrix (this
convention is a bit confusing, because the matrix C rather than B from Sect. 6.6
is called the charge conjugation matrix).

Lemma 6.7.4 Every Majorana form is invariant under the action of Spin+
(s, t).

Proof Property 1. in Definition 6.7.1 implies

Since every element g Spin+(s, t) is of the form

with η(X i , X i ) = 1, η(Y j , Y j ) = −1 it follows that

Not all combinations of μ and ν are possible. Table 6.7 lists without proof the
combinations that are allowed, depending on the dimension n = s + t (see [40,
54, 140]). In even dimensions n there are always two possibilities.
Table 6.7 Signs in Majorana forms

μ ν
0 −1 +1
0 +1 +1
1 +1 +1
2 +1 +1
2 −1 −1
3 −1 −1
4 −1 −1
4 +1 −1
5 +1 −1
6 +1 −1
6 −1 +1
7 −1 +1

Example 6.7.5 In dimension 4 a charge conjugation matrix is necessarily

antisymmetric

and can satisfy either

or

for all a = 1, 2, 3, 4.

Definition 6.7.6 The Majorana conjugate of a spinor ψ Δ with respect

to a Majorana form is defined by

In a basis {χ α } for Δ where ψ = ∑ α ψ α χ α we have with respect to the dual

basis of Δ

In the literature (for example, [54]) the Majorana conjugate is often denoted by a
bar.
Majorana forms have very interesting applications in neutrino physics,
because they can be used to define a Majorana mass term for neutrinos; see
Sect. 7.8 and Sect. 9.2.4.
Remark 6.7.7 In our discussion so far we assumed that the components of
spinors in expressions like

are commuting complex numbers. In quantum field theory, spinors become

fields of operators on spacetime acting on a Hilbert space. In the classical limit
→ 0 these operators are anticommuting . If we treat spinors as anticommuting,
then we have to introduce another minus sign in property 2. in Definition 6.7.1:

with ν still given by Table 6.7 (property 1. stays the same). This has the
paradoxical consequence that symmetric (antisymmetric) Majorana forms
become antisymmetric (symmetric). For instance, in the situation of
Example 6.7.5, a Majorana form in dimension 4 with anticommuting spinors is
symmetric. We will always use commuting spinors except where stated
otherwise.

6.7.2 Dirac Forms

Definition 6.7.8 Let Δ = Δ n be the complex spinor representation of Cl(s, t).
We fix a constant δ = ±1 and consider non-degenerate -bilinear forms

with the following properties, where denotes complex conjugation:

1. X ⋅ ψ, ϕ = δ ψ, X ⋅ ϕ for all and all ψ, ϕ Δ.

2. ψ, ϕ = ϕ, ψ for all ψ, ϕ Δ.

3. ψ, cϕ = c ψ, ϕ = c ψ, ϕ for all ψ, ϕ Δ and all .

We call such a form a Dirac form (we do not assume that the form is
positive definite).

Lemma 6.7.9 For a Dirac form let {χ α } be a complex basis of Δ and A the
matrix with entries

If we expand
and also denote by ψ, ϕ the column vectors with entries ψ α , ϕ α , then

Furthermore, property 1. and 2. in Definition 6.7.8 are equivalent to

1. γ a † = δAγ a A −1 for all a = 1, …, s + t.

2. A † = A.
There is an equivalent equation to the first one with physical Clifford
matrices Γ a :

1. Γ a † = −δAΓ a A −1 for all a = 1, …, s + t.

Proof This is left as an exercise. □

Lemma 6.7.10 Every Dirac form is invariant under the action of Spin+(s, t).

Proof This is left as an exercise. □

Definition 6.7.11 We call a complex representation of the Clifford algebra

Cl(s, t) basis unitary if all gamma matrices γ a are unitary, or equivalently if all
Γ a are unitary.

It can be shown that the spinor representations of Clifford algebras can be

chosen basis unitary, see Exercise 6.13.5. We will choose in all examples that we
discuss a Hermitian scalar product on the spinor space such that the spinor
representation is basis unitary. Then the identity

implies

Remark 6.7.12 If a representation of the Clifford algebra is basis unitary, then

all matrices
are unitary. Hence for t = 0 the induced representations of the Lie groups Pin(n)
and Spin(n) are unitary. This is not the case if t ≥ 1.

The following result can be found in [49].

Proposition 6.7.13 (Standard Expressions for Dirac Forms) For a basis

unitary spinor representation of Cl(s, t) consider one of the two options for the
matrix A in Table 6.8 , where (we always choose ε = 1 or ε = i). In each
case A satisfies the properties of Lemma 6.7.9 and defines a Dirac form for the
spinor representation. The matrix A is unitary for both options.
Table 6.8 Two options for matrix A defining Dirac forms

δ A ε
(−1) t+1 εΓ s+1⋯Γ s+t ε = (−1) t(t+1)⁄2ε

(−1) s εΓ1⋯Γ s ε = (−1) s(s−1)⁄2ε

Proof We only prove the first case, the second one follows similarly. We
calculate for a = 1, …, s

and for a = s + 1, …, s + t

This verifies property 1. in Definition 6.7.8. To check that A is Hermitian we

calculate

It is also easy to check that A is unitary. □

Remark 6.7.14 If t = 0, the first choice yields A = I with δ = −1, ε = 1. In this

case all physical gamma matrices are Hermitian (and unitary by assumption),
while the mathematical gamma matrices are skew-Hermitian.
Definition 6.7.15 The Dirac conjugate of a spinor ψ Δ with respect to a
Dirac form is defined by

In a basis {χ α } for Δ where ψ = ∑ α ψ α χ α we have with respect to the dual

basis of Δ

We write the Dirac scalar product of spinors ψ, ϕ as

Dirac forms are used in the Standard Model to define a Dirac mass term in
the Lagrangian for all fermions (except possibly the neutrinos) and, together
with the Dirac operator, the kinetic term and the interaction term; see Sect.
7.6.

Remark 6.7.16 If we treat spinors as anticommuting , then contrary to the

situation in Remark 6.7.7, we still have

The explanation is that ψ, ψ has to be real for all ψ Δ. More generally,

complex anticommuting Grassmann numbers α, β satisfy (αβ) = β α .

Relation Between Invariant Forms and Majorana Spinors

We follow in this subsection [140]. Let Δ = Δ n be the complex spinor
representation of Cl(s, t). We fix a Majorana form (⋅ , ⋅ ) and a Dirac form ⋅ , ⋅ ,
described in a basis {χ α } for Δ by matrices C, A with

and

Lemma 6.7.17 There exists a unique complex antilinear map

such that

In the basis {χ α } the map τ is given by

with matrix B equal to

Proof The existence of a unique map τ follows, because the Majorana form is
non-degenerate. The map τ is complex antilinear, because the Dirac form is
complex antilinear in the first entry. A simple calculation shows that B is the
correct matrix. Note that

Remark 6.7.18 In the literature (see [140]) one sometimes finds the definition
B T = CA −1. We continue to use our definition.

Lemma 6.7.19 The map τ satisfies

Equivalently, the matrix B satisfies

Proof We calculate:

This implies the claim, because the Dirac form is non-degenerate. The formula
for B follows from this or by direct calculation. □

Theorem 6.7.20 (Scalar Products and Real and Quaternionic Structures)

Suppose that the spinor representation Δ is basis unitary, the matrix C is unitary
and (A, δ) is one of the two choices in Proposition 6.7.13 . Then B = CA is
unitary.
1. In the first case, if A = εΓ s+1 ⋯Γ s+t , then the matrix B satisfies

2. In the second case, if A = εΓ 1 ⋯Γ s , then the matrix B satisfies

3. In both choices for A the matrix B B is + I if and only if

where ρ = s − t. In these situations B defines a real structure for the spinor

representation of the orthochronous spin group.

4. The matrix B B is − I if and only if

In these situations B defines a quaternionic structure for the spinor

representation of the orthochronous spin group.

Proof

1. We calculate
2. This follows similarly.

3. and 4. The remaining claims follow by considering Table 6.7. Note that the
value of (−1) t(t−1)⁄2 depends only on the value of , and similarly for s.
□

Remark 6.7.21 This theorem explains why Majorana spinors exist if

and for the right choice of μ also if (see Sect. 6.6).

Corollary 6.7.22 For a spinor ψ Δ the Majorana conjugate is equal to the

Dirac conjugate, , i.e.

if and only if τ(ψ) = ψ or, equivalently, Bψ = ψ . This happens for a non-zero

spinor ψ if and only if B defines a real structure and ψ is a Majorana spinor.

This corollary finally explains the relation between Majorana spinors and
invariant forms for the spinor representations.

6.8 Explicit Formulas for Minkowski Spacetime

of Dimension 4
We collect some explicit formulas concerning Clifford algebras and spinors for
the case of 4-dimensional Minkowski spacetime.

6.8.1 The Lorentz Clifford Algebra

In Minkowski spacetime of dimension 4 and signature (+, −, −, −) (usually used
in quantum field theory) there exist both Weyl and Majorana spinors, but not
Majorana–Weyl spinors.
Recall from Example 6.3.17 that the matrices

are physical gamma matrices for Cl(1, 3). This is the so-called Weyl
representation of the Clifford algebra. It is a basis unitary representation. The
physical chirality operator is given by

Hence in our convention the first two components of a Dirac 4-spinor

correspond to a left-handed Weyl spinor and the last two components to a right-
handed Weyl spinor (this is the standard convention in quantum field theory and
the Standard Model) .
For the matrix A defining the Dirac form we use the second choice in
Proposition 6.7.13. Then

with δ = −1, ε = 1. This implies

and the Dirac conjugate of a spinor ψ is given by

with Hermitian scalar product

satisfying

If we write a Dirac spinor as

then

Note that this Hermitian scalar product is not positive definite. In fact the
subspaces of left-handed and right-handed Weyl spinors are each null.
For the matrix C defining the Majorana form we choose

Then
hence μ = ν = −1. The matrix C is unitary and

The matrix C corresponds to the matrix ε in Sect. 2.1.3. We have

and the charge conjugate of a Dirac spinor is given by

This implies (as expected from Theorem 6.7.20)

and B defines a real structure on the spinor space Δ. Majorana spinors ψ are
characterized by

If we write

then Majorana spinors are precisely those of the form

6.8.2 The Orthochronous Lorentz Spin Group

Our aim in this subsection is to prove that the orthochronous spin group
Spin+(1, 3) of 4-dimensional Minkowski spacetime is isomorphic to the 6-
dimensional real Lie group . We identify with the Hermitian 2 × 2-
complex matrices via the real vector space isomorphism

Here σ 1, σ 2, σ 3 are the Pauli matrices and σ 0 = I 2. The following is easy to

verify:

Lemma 6.8.1 Under this identification, η(x, x) = detX for all .

We can then prove the following (see [146, Appendix A]).

Theorem 6.8.2 (Explicit Description of the Orthochronous Lorentz Spin
Group) Under the identification above, the following map is well-defined

This maps yields a Lie group homomorphism

which is surjective and has kernel {I, −I}. Since is simply connected, it
follows that .

Proof This is Exercise 6.13.17. Compare with Exercises 1.9.20 and 1.9.21. □

6.9 Spin Structures and Spinor Bundles

In this section we discuss the notion of spin structure that is needed to define
spinors globally on manifolds. If the tangent bundle TM of a manifold M is a
trivial vector bundle, then spin structures always exist. If TM is non-trivial, then
there may exist a topological obstruction to the existence of a spin structure,
measured by the second Stiefel–Whitney class w 2(M). If a spin structure for a
pseudo-Riemannian manifold (M, g) exists, then we can define an associated
complex vector bundle, called the spinor bundle. The sections of this bundle are
called spinor fields or spinors on the manifold M.

6.9.1 Spin Structures

Definition 6.9.1 Let M be a smooth manifold. A pseudo-Riemannian metric
g of signature (s, t), where

is a section g Γ(T M T M) that defines at each point x M a non-

degenerate, symmetric bilinear form

of signature (s, t).

Let M be a smooth manifold with a pseudo-Riemannian metric g of signature

(s, t). The frame bundle then has the structure of a principal O(s, t)-bundle. For
the following definitions recall the notion of a G-reduction of a principal bundle
from Definition 4.2.17.
Definition 6.9.2

1. The pseudo-Riemannian manifold (M, g) is called orientable if the frame

bundle can be reduced to a principal SO(s, t)-bundle under the embedding
SO(s, t) O(s, t).

2. The pseudo-Riemannian manifold (M, g) is called time-orientable if the

frame bundle can be reduced to a principal O+(s, t)-bundle under the
embedding O+(s, t) O(s, t).

3. The pseudo-Riemannian manifold (M, g) is called orientable and time-

orientable if the frame bundle can be reduced to a principal SO+(s, t)-
bundle under the embedding SO+(s, t) O(s, t).
If such reductions are chosen, we call the pseudo-Riemannian manifold
(M, g) oriented , time-oriented or oriented and time-oriented.

An orientation of M is just an orientation of the tangent bundle TM. A time-

orientation has the following interpretation: using some homotopy theory and
Lemma 6.1.14 it is not difficult to see that the tangent bundle TM admits
maximally g-positive definite vector subbundles W → M and that any two such
subbundles are homotopic. A time-orientation of (M, g) then corresponds to a
choice of orientation of such a maximally g-positive definite vector subbundle
W. On an orientable pseudo-Riemannian manifold there are precisely two
different orientations and similarly for time-orientations on time-orientable
pseudo-Riemannian manifolds.
Suppose that (M, g) is oriented and time-oriented. We denote the SO+(s, t)-
frame bundle by

Recall the double covering

from Sect. 6.5.1.

Definition 6.9.3 A spin structure on M is a Spin+(s, t)-principal bundle

with a double covering

such that the following diagram commutes:

Here the horizontal arrow in the top and bottom line are the right actions of the
structure groups on the principal bundles.
According to Definition 4.2.17, a spin structure is thus a λ -equivariant
bundle morphism Λ: Spin+(M) → SO+(M), i.e. a λ -reduction of SO+(M).

Definition 6.9.4 Two spin structures

are called isomorphic if there exists a Spin+(s, t)-equivariant bundle

isomorphism

such that the following diagram commutes:

Remark 6.9.5 Note that a spin structure is more than just a double covering of
the frame bundle which fibrewise looks like the covering λ. We demand in
addition that the actions of the structure groups on both principal bundles are
compatible. This additional structure is needed, for example, in the proof of
Proposition 6.9.13 to define Clifford multiplication on the level of bundles and
in Proposition 6.10.7 to define a connection 1-form on Spin+(M) associated to
the Levi-Civita connection on TM.

Remark 6.9.6 We can define in the same way spin structures for any principal
SO+(s, t)-bundle, not only for the tangent bundle of a pseudo-Riemannian
manifold.
If the tangent bundle TM of the manifold M is trivial, then SO+(M) is trivial and
a spin structure is easy to define. In general, if TM is non-trivial, there may be a
topological obstruction to defining a spin structure. If the topology of M is non-
trivial there may also be several non-isomorphic spin structures. The precise
statement is the following.

Theorem 6.9.7 (Existence and Uniqueness of Spin Structures)

1. The frame bundle SO+(M) admits a spin structure if and only if the second
Stiefel–Whitney class of M vanishes, w 2(M) = 0.

2. If SO+(M) admits a spin structure, then there is a bijection between the set of
isomorphism classes of spin structures on M and the cohomology group
.

Proof The details of the proof would take us too far afield, because we have
not discussed characteristic classes like the Stiefel–Whitney classes. A proof can
be found in [55] and [88] in the Riemannian case and in [13] in the general
pseudo-Riemannian case. Note that a manifold M is orientable if and only if w
1(M) = 0 and the number of different orientations is given by the number of
elements of , hence the existence of a spin structure can be understood
as the existence of a “higher orientation” for M. □

We call a manifold M spin if w 2(M) = 0. The following manifolds can be shown

to be spin:
All manifolds M with trivial tangent bundle TM, in particular, all Euclidean
spaces and tori T n .

All spheres S n .
All orientable 2-dimensional manifolds.
Complex projective spaces are spin if and only if n is odd.
If M and N are spin manifolds, then so is M × N.
The cohomology group vanishes, for example, if π 1(M) is trivial.
In particular we get:
Corollary 6.9.8 The manifold admits for all s, t ≥ 0 a unique spin
structure.

Note that if the tangent bundle of M is trivial, then there always exists (after a
choice of trivialization of TM) a canonical (trivial) spin structure, but there exist
additional (non-trivial) ones if . This happens, for instance, in the
case of the torus T n , where

has 2 n elements.

Remark 6.9.9 If we think of the structure group of a principal bundle as a

symmetry group, then the existence of a reduction of the structure group means
that the bundle admits a more fundamental or hidden symmetry group. In
particular, the frame bundle SO+(M) of a pseudo-Riemannian manifold M admits
a spin structure if and only if it has a more fundamental underlying symmetry,
given by the orthochronous spin group Spin+(s, t).
We can recover all tensor bundles and the tangent bundle itself from the
principal bundle Spin+(M) as associated vector bundles. However, we also get
additional vector bundles that cannot be associated to the SO+(M)-frame bundle.
In general, the spin structure Λ: Spin+(M) → SO+(M) is not unique: several non-
isomorphic spin structures may induce the same frame bundle SO+(M).

We briefly want to discuss how a section of the bundle SO+(M) determines

sections of Spin+(M).

Definition 6.9.10 We call a local section e = (e 1, …, e n ) of SO+(M) an n -bein

or vielbein (and tetrad in dimension n = 4).

Lemma 6.9.11 Suppose we have chosen a spin structure on M. Then for every
vielbein e on a contractible open set U M there exist precisely two local
sections ε ± of Spin+(M) over U such that Λ ∘ε ± = e.

Proof The image of e is a contractible open subset U′ of SO+(M) diffeomorphic

to U and thus
is a trivial two-sheeted covering, admitting precisely two sections

Then ε ± = s ±∘ e have the claimed properties. See Fig. 6.1. □

Fig. 6.1 Local frames for Spin+(M)

6.9.2 Spinor Bundles

Definition 6.9.12 Let Spin+(M) → M be a spin structure on M and

the spinor representation. Then the (Dirac) spinor bundle is the associated
complex vector bundle

over M. Sections of S are called spinor fields or spinors . Note that the
spinor bundle may depend on the choice of spin structure.

Proposition 6.9.13 Let S → M be the spinor bundle associated to a spin

structure.

1. There exists a well-defined bilinear Clifford multiplication

on the level of bundles, restricting to a map T p M × S p → S p in every point

p M. This map also induces a well-defined Clifford multiplication of
forms with spinors.
2. If the dimension n of M is even, then S splits as a direct sum of complex
Weyl spinor bundles S = S + S − , defined by

In this case, Clifford multiplication with a vector maps S ± to S .

Proof

1. Let ρ SO be the standard representation of SO+(s, t) on . Then the tangent

bundle TM is isomorphic to the associated vector bundle

Suppose ε is an element of Spin+(M). Then Λ(ε) is an element of SO+(M)

and we can define Clifford multiplication TM × S → S as follows:

Here x ⋅ ψ is the standard Clifford multiplication between elements

and ψ Δ. A direct argument using Definition 6.9.3 of spin structures,
Definition 4.7.3 of associated vector bundles and Proposition 6.5.19 on the
compatibility of the spinor representation with Clifford multiplication shows
that the map TM × S → S is well-defined.

2. This follows, because by Lemma 6.3.5 the chirality element

is well-defined, independent of the choice of vielbein, and Clifford

multiplication of the chirality element with a spinor is well-defined by
part 1. of the proposition.
□

It is sometimes convenient to describe spinors locally: Let e be a local

vielbein and ε ± the associated local sections of the spin structure principal
bundle on a contractible, open subset U in M. If Ψ: U → S is a local section of
the spinor bundle, then there exist two maps
 such that Ψ can be written as

We have ψ ± = −ψ . We choose one of the two local sections ε and the

associated map ψ, so that

If we define a formula locally using ε and ψ, we always have to check that the
result is independent of this choice. Generally speaking, this will always be
the case if the expression is linear in ψ. For example, physical Clifford
multiplication with a basis vector can be expressed as

The right-hand side is indeed independent of the choice of ε, since Γ a ψ is

linear in ψ.

6.9.3 Structures on Spinor Bundles

In Sect. 6.6 and Sect. 6.7 we considered several structures on the spinor
representation space Δ:

1. Real structures σ and quaternionic structures J.

2. Majorana forms (⋅ , ⋅ ) and Dirac forms ⋅ , ⋅ .

It is clear that these structures extend fibrewise to globally well-defined,
smooth structures on the spinor bundle S → M: this follows, because by
definition real and quaternionic structures commute with the spinor
representation of the orthochronous spin group and Majorana and Dirac forms
are invariant under the orthochronous spin group, cf. Lemma 6.7.4 and
Lemma 6.7.10. We can therefore define the following structures:

1. Real bundle automorphisms σ S and quaternionic bundle automorphisms J S

of the spinor bundle S (depending on the dimension and signature of the
manifold) together with (symplectic) Majorana or (symplectic) Majorana–
Weyl sections of S.

2. Majorana bundle metrics (⋅ , ⋅ ) S and Dirac bundle metrics ⋅ , ⋅ S on S.

6.10 The Spin Covariant Derivative
Given a spin structure on a pseudo-Riemannian manifold and the spinor bundle
S, we would like to have a covariant derivative on S so that we can define field
equations involving derivatives of spinors. It turns out that the standard Levi-
Civita connection on the tangent bundle, determined by the pseudo-Riemannian
metric, defines a unique compatible covariant derivative on S, called the spin
covariant derivative.
Let (M, g) be an oriented and time-oriented pseudo-Riemannian manifold of
signature (s, t). We assume that M is spin and a spin structure has been chosen.
Let denote the Levi-Civita connection on TM associated to the metric g.

6.10.1 Spin Connection

Definition 6.10.1 In a local vielbein e = (e 1, …, e n ) on an open set U M we
can write

with certain uniquely determined real-valued 1-forms ω ab on U (and the

Einstein convention is understood here and in the following formulas). We also
set

Remark 6.10.2 In the physics literature the forms ω ab are often defined with
the opposite sign. We continue to use our definition.

Lemma 6.10.3 The 1-forms ω ab are antisymmetric in the indices a, b:

Proof This is Exercise 6.13.19. □

Definition 6.10.4 The anholonomy coefficients Ω ab c of a local vielbein e are

defined by

With respect to local coordinates x μ in a local chart on U M one also defines

in the physics literature real functions E a μ on U by
Lemma 6.10.5 The 1-forms ω ab are determined by the anholonomy
coefficients as follows:

where Ω abc = Ω ab dη
dc .

Proof This is Exercise 6.13.20. □

The Levi-Civita connection induces a connection 1-form

on the frame bundle SO+(M). The Levi-Civita connection is the associated

covariant derivative on the tangent bundle

This means that if Y = ∑ a y a e a is the local expansion of a vector field Y on M,

then

where A SO e = e A SO is the local connection 1-form. This implies:

Lemma 6.10.6 The local connection 1-form A SO e is given by

Consider the Lie group homomorphism

Since this map is a covering, it induces an isomorphism

Proposition 6.10.7 Let

be the covering map given by the spin structure. Then

is a connection 1-form on the principal bundle Spin+(M) → M, called the

spin connection .

Proof We have to verify the properties of connection 1-forms.

1. Let g Spin+(s, t) and Y TSpin+(M). Then by the defining properties of

a spin structure

2. Let X spin+(s, t) and ε Spin+(M). Then

Note that the proof of this proposition crucially needs the compatibility of the
structure group actions in Definition 6.9.3 of spin structures.

6.10.2 Spin Covariant Derivative

Definition 6.10.8 We call the associated covariant derivative on the spinor
bundle S defined by the spin connection A Spin the spin covariant derivative ,
again denoted by .

The spin covariant derivative is completely determined by the Levi-Civita

connection, once a spin structure has been chosen. We want to derive a local
formula for the spin covariant derivative. Suppose that U is contractible and ε ±
the associated local trivializations of the principal bundle Spin+(M) on U. We
choose one of those trivializations, called ε. Then we get associated local
connection 1-forms

We write a section Ψ of the spinor bundle S as

with ψ: U → Δ. Then the spin covariant derivative on the spinor bundle can be
written locally as

where

and A Spin ε (X) acts through the representation κ of on Δ induced

from the spinor representation κ.

Proposition 6.10.9 (Explicit Formula for the Spin Covariant Derivative)

The spin covariant derivative can be calculated by the explicit formula

We need the following lemma:

Lemma 6.10.10 Write the components of a matrix as

Then the map

where κ is the spinor representation, is given by

Proof This follows from the last formula in Corollary 6.5.23. □

We can now prove Proposition 6.10.9.

Proof We want to prove that

where κ is the homomorphism induced by the spinor representation. We have

The claim then follows immediately from Lemma 6.10.6 and Lemma 6.10.10. □
Remark 6.10.11 As mentioned above, in physics the 1-forms ω ab are
sometimes defined with the opposite sign,

Then Xψ has to be defined by

Lemma 6.10.12 If the dimension n of the manifold M is even, then the spin
covariant derivative preserves the splitting of the spinor bundle S into the Weyl
spinor bundles S + and S − . This means that

Proof This follows from .□

The spin covariant derivative has the following property:

Theorem 6.10.13 The spin covariant derivative is compatible with the Levi-
Civita connection in the following way : for all vector fields and
spinors Ψ Γ(S) the identity

holds.

Proof This is Exercise 6.13.21. □

The following result is often useful:

Theorem 6.10.14 Let (⋅ , ⋅ ) S be a Majorana bundle metric and ⋅ , ⋅ S a

Dirac bundle metric on the spinor bundle S, defined by Majorana and Dirac
forms for the spinor representation. Then these metrics are compatible with the
spin covariant derivative in the following way:

for all vector fields and spinors Ψ, Φ Γ(S).

Proof This follows from Proposition 5.9.7, because both types of forms on the
spinor space Δ are invariant under Spin+(s, t). □
6.10.3 Dirac Operator
We can now define the Dirac operator.

Definition 6.10.15 The Dirac operator D: Γ(S) → Γ(S) on the spinor bundle S
is defined by

(with mathematical Clifford multiplication and the Einstein summation

convention). In a given vielbein e where Ψ = [ε, ψ] we can write

where

In the physics literature the Dirac operator is often denoted by .

It is easy to check that the definition of the Dirac operator is independent of the
local vielbein e: This can be checked directly or by noticing that the Dirac
operator D is the composition of the maps

where η: T M → TM is the isomorphism induced by the pseudo-Riemannian

metric on M and γ is Clifford multiplication. The Dirac operator is obviously a
first-order differential operator on the spinor bundle S.
In even dimensions n, preserves the splitting S = S + S −, while Clifford
multiplication with a vector interchanges S + and S −. This implies:

Corollary 6.10.16 If the dimension n of the manifold M is even, then the Dirac
operator D maps

i.e. it takes sections of S + to sections of S − and vice versa.

In any dimension n, suppose that ⋅ , ⋅ Sa Dirac bundle metric with δ = −1, i.e.

On the complex vector space Γ 0(S) of sections of S with compact support in M

we define an L 2 -scalar product of spinors

by

Here dvol g is the volume element associated to the pseudo-Riemannian metric g

and the orientation of M (cf. Sect. 7.2.1). We need to restrict to sections with
compact support, because otherwise the integral may not be finite.
We can then prove the following:

Theorem 6.10.17 (Dirac Operator Is Formally Self-Adjoint) Let M be a

manifold without boundary. If the Dirac form satisfies δ = −1, then the Dirac
operator D: Γ 0(S) → Γ 0(S) is formally self-adjoint, i.e.

for all spinors Φ, Ψ Γ 0(S).

A proof (of a more general theorem) can be found in Exercise 7.9.12.

6.11 Twisted Spinor Bundles

Let P → M be a principal G-bundle and ρ: G → GL(V ) a complex
representation with associated vector bundle E = P × ρ V. Let S → M be the
spinor bundle associated to a spin structure on M.

Definition 6.11.1 We call the bundle S E a twisted spinor bundle or gauge

multiplet spinor bundle.

Let s: U → P be a local gauge. Then we can write a local section τ of the

associated vector bundle E as

with a map v: U → V. We identify V with using a basis v 1, …, v r for V. This

defines a local frame τ 1, …, τ r for E, given by τ i = [s, v i ].
Let e be a local vielbein for TM and ε: U → Spin+(M) a corresponding local
trivialization.

Lemma 6.11.2 Any section Ψ of the twisted spinor bundle can be written
locally as
with sections Ψ i of S. Equivalently,

where ψ is a multiplet of the form

and ψ i are maps from U to Δ. This decomposition is unique, once local sections
of P and Spin+(M) as well as a basis for V have been chosen.

We want to consider covariant derivatives on twisted spinor bundles. Let A be a

connection 1-form on the principal G-bundle P.

Definition 6.11.3 The twisted spin covariant derivative A on the twisted

spinor bundle S E is defined by

where

Here ψ is a map on U to Δ V and the matrices Γ bc act on the Δ-part of ψ,

i.e. on each of the spinor components ψ i separately, while ρ A s (X) acts on the
V -part of ψ, i.e. mixes the components of the multiplet ψ.

It is easy to check that the definition of A does not depend on the choice of ε
and s. We also get a Dirac operator on twisted spinor bundles.

Definition 6.11.4 The Dirac operator

on a twisted spinor bundle S E is defined by

Equivalently, D A is the composition of the maps

Locally we have

with

In the physics literature the Dirac operator is sometimes denoted by .

Suppose again that ⋅ , ⋅ S is a Dirac bundle metric with δ = −1, ⋅ , ⋅ E is a

Hermitian bundle metric on E and ⋅ , ⋅ S E the induced bundle metric on S
E. On the complex vector space Γ 0(S E) of sections of S E with compact
support in M we define an L 2 -scalar product of twisted spinors

by

We then get the following analogue of Theorem 6.10.17.

Theorem 6.11.5 (Twisted Dirac Operator Is Formally Self-Adjoint) Let M

be a manifold without boundary. If the Dirac form satisfies δ = −1, then the
twisted Dirac operator D: Γ 0(S E) → Γ 0(S E) is formally self-adjoint, i.e.

This is proved in Exercise 7.9.12.

6.12 Twisted Chiral Spinors

Suppose the dimension n of the manifold M is even. Then the spinor bundle
splits into Weyl spinor bundles S = S + S −. Let P → M be a principal G-
bundle and

two (possibly distinct) representations of G on complex vector spaces V ±. Let

be the associated vector bundles.

Definition 6.12.1 We call

a twisted chiral spinor bundle . We also consider the twisted bundle

We can write a section of the twisted chiral spinor bundle as

where

Suppose A is a connection 1-form on P.

Definition 6.12.2 The twisted chiral spin covariant derivative A on the

twisted chiral spinor bundle (S E)+ is defined by

where

We can again define a Dirac operator.

Definition 6.12.3 The Dirac operator

is locally given by

We can decompose the Dirac operator D A into

We again denote the Dirac operator also by .

Remark 6.12.4 Similar to Remark 6.2.6 for the classical Dirac operator it is
possible to calculate the square of (twisted) Dirac operators over general
manifolds. This square is again given by the Laplacian plus certain correction
terms that depend on the curvature of the vector bundles. The formula for the
square of the Dirac operator is known as the Lichnerowicz–Weitzenböck
formula (see [15, 88] and [115] for more details).
 Remark 6.12.5 It can be proved that on a closed (compact without
boundary) Riemannian manifold the twisted chiral Dirac operator

as a linear map between infinite-dimensional vector spaces, has finite-

dimensional kernel and cokernel. This includes as a special case a twisted
Dirac operator of the form

The index of a twisted chiral Dirac operator is the integer

The famous Atiyah–Singer Index Theorem gives a formula for this index
in terms of characteristic classes of the vector bundles TM and E ±. See [15,
88] and [115] for detailed expositions of the index theorem ([15] considers
specifically the case of a twisted chiral spinor bundle and calls the vector
bundle E = E + E − a superbundle).

Remark 6.12.6 The notion of twisted chiral spinor bundles seems like an
unnecessary complication from a physics point of view. However, as we shall
see in Chap. 8, they are crucial when describing the weak interactions in the
Standard Model. In fact, in the Standard Model where

the vector spaces V + and V − even have different dimensions: V + has

dimension 24 and V − has dimension 21. As G-representations these spaces
decompose into irreducible representations of dimensions

This is related to the fact that the weak interaction in the Standard Model is
not invariant under parity inversion that exchanges left-handed with right-
handed fermions. See Sect. 8.5 for more details.

Remark 6.12.7 It is sometimes useful to form the fibre product of the

principal Spin+(s, t)-bundle

and the principal G-bundle

defined by

This is a principal bundle for the group Spin+(s, t) × G. Twisted spinor bundles S
E are then vector bundles associated to this principal bundle via a
representation of the group Spin+(s, t) × G (the same is true for vector bundles of
the form T E where T is associated to any Spin+(s, t)-representation, for
example, the scalar or vector representation). The group Spin+(s, t) × G can thus
be considered as the full symmetry group of the gauge theory.

6.13 Exercises for Chap. 6

6.13.1 Find a canonical isomorphism between O(s, t) and O(t, s) for all s, t.
6.13.2 Show that if both s, t ≠ 0, then O(s, t) is not compact.
6.13.3 Prove that the subgroup SO+(s, t) is the connected component of the
identity in O(s, t).
6.13.4

1. The general form of a matrix A O(1, 1) depends on one real parameter a

≥ 1 and three signs δ, ε, γ. Determine this general form of A.

2. Determine the general form of a matrix A SO+(1, 1).

6.13.5 Consider the Clifford algebra Cl(s, t) with s + t = n. The set of all
products of the basis vectors forms a subgroup of the
multiplicative group of Cl(s, t) of order 2 n . Prove that every complex (real)
representation of Cl(s, t) admits a Hermitian (Euclidean) scalar product such that
all gamma matrices are unitary (orthogonal).
6.13.6 Use Table 6.2 to show that for all

This result is called Bott periodicity.

6.13.7 Consider the Clifford algebras Cl(n) for .

1. For define an integer ρ′(m) inductively by

 and

Use Table 6.2 and Bott periodicity to prove that admits a representation
of Cl(ρ′(m)) for all .

2. Write an integer as n = (2a + 1)2 m with , i.e. 2 m is the largest

power of 2 dividing n. Define an integer ρ(n) by

Prove that admits a representation of Cl(ρ(n) − 1) for all .

6.13.8 Let n ≥ 1 be an integer and consider the integer ρ(n) from
Exercise 6.13.7. Prove that the unit sphere admits a set of ρ(n) − 1
orthonormal tangent vector fields. Determine those n for which this construction
gives an orthonormal trivialization of the tangent bundle TS n−1.

Remark According to a theorem of Adams [3] the sphere S n−1 admits no more
than ρ(n) − 1 linearly independent tangent vector fields.

6.13.9 Let e 1, …, e s+t be an orthonormal basis of and define

Prove that the vector space M(s, t) is a Lie subalgebra of with the
commutator

Show that M(s, t) has dimension .

6.13.10

1. Let n ≥ 3. Show that to every connected and simply connected Lie subgroup
H SO(n) we can associate a canonical Lie subgroup in Spin(n)
isomorphic to H.

2. Find embeddings of the Lie groups SU(m) (for m ≥ 2), Sp(k) (for k ≥ 1) and
G2 into suitable spin groups Spin(n).
Remark According to Exercise 3.12.6 these embeddings define simply
connected homogeneous spaces .

6.13.11 Consider the spinor representation of Cl(1, 1) on , defined using

the mathematical gamma matrices from Example 6.3.14. Find an explicit map

which is complex antilinear, commutes with γ k for k = 1, 2, preserves both Weyl

spinor spaces Δ + and Δ − and satisfies σ 2 = Id Δ . This proves that the spinor
representation of Cl(1, 1) is Majorana–Weyl. Determine the subspaces of left-
handed and right-handed Majorana–Weyl spinors.
6.13.12 Consider the spinor representation of Cl(4, 0) on , defined
using the mathematical gamma matrices

from Example 6.3.16.

1. Find an explicit map

which is complex antilinear, commutes with γ k for all k = 0, 1, 2, 3,

preserves both Weyl spinor spaces Δ + and Δ − and satisfies J 2 = −Id Δ .

2. Show that I = i, J and K = IJ turn the Weyl spinor spaces Δ ± into

quaternionic vector spaces. The spinor representation of Cl(4, 0) is therefore
symplectic Majorana–Weyl.
6.13.13 Find mathematical gamma matrices for the spinor representation of
Cl(3, 0) and show by an explicit calculation that it is symplectic Majorana.
6.13.14 Consider the spinor representation of Cl(1, 1) on , defined using
the mathematical gamma matrices from Example 6.3.14.

1. Find explicit unitary charge conjugation matrices C for both cases in the
table in Sect. 6.7.1.
2. Show that the representation is basis unitary and determine the matrices A
for both choices in Proposition 6.7.13.

3. Calculate all four combinations for the matrix B = CA and show that in each
case B B = I 2, hence B defines a real structure. Determine the Majorana
spinors for each choice of B.
6.13.15 Consider the spinor representation of Cl(4, 0) on , defined using
the mathematical gamma matrices

from Example 6.3.16.

1. Find explicit unitary charge conjugation matrices C for both cases in the
table in Sect. 6.7.1.

2. Show that the representation is basis unitary and determine the two choices
for the matrix A from Proposition 6.7.13.

3. Calculate all four combinations for the matrix B = CA and show that in each
case B B = −I 2, hence B defines a quaternionic structure.
6.13.16 Consider the spinor representation of Cl(1, 3) as in Sect. 6.8.

1. Determine a charge conjugation matrix C ′ such that

2. Calculate the matrix B = C ′ A with A = Γ 0 and show that B B = −I, hence

B defines a quaternionic structure.

3. Determine the matrix A ′ corresponding to the first choice in

Proposition 6.7.13.
4. Calculate the matrices B = CA ′ (with C as in Sect. 6.8) and B = C ′ A ′ and
determine whether they define a real or quaternionic structure.
6.13.17 Consider the identification from Sect. 6.8.2 and set

1. Prove that this map is well-defined and yields a homomorphism

of Lie groups.

2. Show that ψ is surjective and has kernel {I, −I}.

6.13.18 Do a similar construction to the one in Exercise 6.13.17 to show that
there exists a surjective Lie group homomorphism

with kernel {I, −I}.

Remark The 3-dimensional Lie group is isomorphic to Spin+(1, 2)

(note that SO+(1, 2) has fundamental group ).

6.13.19 Show that the 1-forms ω ab defined by the Levi-Civita connection with
respect to a vielbein are antisymmetric in the indices a, b:

6.13.20 Prove that the 1-forms ω ab are determined by the anholonomy

coefficients Ω ab c as follows:

where Ω abc = Ω ab d η dc .
6.13.21 Show that the spin covariant derivative is compatible with the Levi-
Civita connection in the following way: For all vector fields and
spinors Ψ Γ(S) the identity

holds.
References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum:
Bull. Am. Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
 [MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory.
W. A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives.
Eur. Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]
31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)
47.
 Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories.
Phys. Rev. Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)
63.
 Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry.
Phys. Rev. D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles.
Phys. Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)
80.
Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
 Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles.
Bull. Am. Math. Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
 New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings.
Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/
112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics.
Phys. Lett. B 388, 753–760 (1996)
[MathSciNet]
127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)
[MATH]

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
 Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions.
Chin. Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/
Part II
The Standard Model of Elementary
Particle Physics
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_7
Chapter 7 The Classical Lagrangians of
Gauge Theories
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

If we consider, from an abstract point of view, a field theory involving several

types of fields on spacetime (scalar fields, gauge fields, spinors, etc.), then the
Lagrangian of the field theory is the formula that contains the dynamics and all
interactions between these fields. In classical field theory, the equations of
motion, i.e. the field equations, that govern the evolution of the fields over time,
are derived from the Lagrangian. In quantum field theory, the Lagrangian
(through the action , the integral of the Lagrangian over spacetime) enters the
formula for path integrals that are used to calculate correlators and scattering
amplitudes for elementary particles .
Given that the structure of the common Lagrangians is quite simple, it is
truly remarkable that the enormous complexity and intricacy of quantum field
theories are already contained in the Lagrangians. The Lagrangians can be
considered the fundamental cornerstones of field theories.
Lagrangians can be categorized depending on which types of fields and
interactions they involve: there are Lagrangians for free fields, Lagrangians for a
single interacting field and Lagrangians for several interacting fields. As a
general rule, Lagrangians which are harmonic , i.e. quadratic in the fields,
correspond to free theories, while Lagrangians which contain anharmonic terms
of order three or higher in the fields lead in the quantum field theory to the
creation and annihilation of particles and thus to interactions. Interactions
between fields (in particular, in the case of weakly interacting, perturbative
quantum field theories) are depicted using Feynman diagrams . Interacting
quantum field theories are usually very complicated and in many cases
(including the Standard Model) not fully understood.
 There are a priori countless Lagrangians that one could consider for a given
set of fields. The Lagrangians that are important in physics are mainly restricted
by three principles:

1. Existence of symmetries.

2. The quantum field theory should be renormalizable.

3. The quantum field theory should be free of gauge anomalies.

We will briefly discuss how these principles restrict the possible Lagrangians
and then study the Lagrangians that appear in the Standard Model of elementary
particles. These Lagrangians are called:
the Yang–Mills Lagrangian
the Klein–Gordon and Higgs Lagrangian
the Dirac Lagrangian
Yukawa coupling (itself not a complete dynamic Lagrangian)
The Lagrangians in the Standard Model are all Lorentz invariant and gauge
invariant. Lorentz invariance here means invariance under local Lorentz
transformations of the spacetime manifold, acting on each tangent space. This
implies that for fixed values of the fields the Lagrangian is a scalar function on
spacetime. Lorentz invariance for a field theory involving spinors always means
invariance under the orthochronous Lorentz spin group .
There are numerous books and articles on field theory and the Standard
Model. In the present and the following chapter on the Standard Model we
mainly rely on the following references:
The book [16] by David Bleecker is one of the best mathematical
treatments of symmetry breaking. Our discussion of symmetry breaking and
the Higgs mechanism in Chap. 8 draws heavily from it.
The article [9] by John C. Baez and John Huerta is an excellent
mathematical exposition of the representations of the Standard Model and
Grand Unified Theories. Our notation for the representations of the
Standard Model in Chap. 8 mainly follows this reference.
The book [100] by Ulrich Mosel is a concise summary of the Standard
Model with a very good exposition of the Lagrangians and symmetry
 breaking. The explicit Lagrangians for the Standard Model that we derive in
Chap. 8 mainly follow the notation in Mosel’s book.
The book [137] by Mark Thomson is an excellent modern and readable
treatment of particle physics with many interesting details and explanations
concerning experimental and theoretical aspects.
The topic of the book [22] by Gustavo Castelo Branco, Luís Lavoura and
João Paulo Silva is CP violation, but it also has a very clear description of
the Standard Model and its Lagrangians.
The book [62] by Carlo Giunti and Chung W. Kim focuses mainly on
neutrino physics, but also contains in the first chapters a concise and
modern description of the Standard Model, including details about the
Lagrangians and quark mixing.
Our main references for results from quantum field theory are the book
[125] by Matthew Schwartz and the books [143–145] by Steven Weinberg.
The website [105] of the Particle Data Group contains many up-to-date
experimental values for elementary particles as well as some succinct
theoretical discussions.
A great source for the history and development of the Standard Model is the
book [79] by Lillian Hoddeson, Laurie Brown, Michael Riordan and Max
Dresden (editors).1 Historical remarks can also be found on the official
website [117] for the Nobel Prize in Physics. In particular, [118–120] and
[121] contain very readable background material for the Nobel Prizes in
Physics 2004, 2008, 2013 and 2015. A short history of gauge theory in
physics and mathematics can be found in the book review [92] for
Bleecker’s book.
Further references are the books [33, 71] and [113] (on the complete
Standard Model), [112] (on the electroweak theory), [42] (on QCD), [124, 132]
(on QFT in general) and [39, 41, 101, 102, 114, 122, 123] (on the mathematics
of the Standard Model) as well as the lecture notes [141].

7.1 Restrictions on the Set of Lagrangians

The Lagrangians that occur in physics are restricted from the infinite set of
possible Lagrangians by certain principles that we want to discuss in this section.
7.1.1 Existence of Symmetries
The Lagrangian (or the action) of a field theory should be invariant under certain
transformations of the fields, i.e. under certain symmetry groups. Particular
examples are:
Lorentz symmetry
gauge symmetry
conformal symmetry
supersymmetry
We have to distinguish two meanings of symmetries in field theories. Here
we think of the primary meaning: the Lagrangian for the fields and thus the laws
of physics, not the field configurations or their initial values themselves, are
invariant under symmetry transformations.
The secondary meaning of symmetry (invariance of the actual field
configuration) is also sometimes of significance in physics. For example, the
action of general relativity for the spacetime metric is invariant under the full
orientation preserving diffeomorphism group of the spacetime manifold. A
specific metric, however, is invariant only under a much smaller symmetry
group, the isometry group of this metric (which could just consist of a single
element, the identity map).
Similarly the actions of supergravity theories are invariant under all local
supersymmetries, but a specific supersymmetric configuration is invariant under
a much smaller group of supersymmetries (a generic field configuration is not
supersymmetric at all).
A third example is spontaneously broken gauge theories , which we consider
in detail in Chap. 8. In this case the Lagrangian is invariant under gauge
transformations with values in a Lie group G, but due to the existence of the
Higgs condensate, the vacuum configuration is invariant only under gauge
transformations with values in a subgroup H G.
The existence of gauge symmetries is particularly important: it can be
shown that a quantum field theory involving massless spin 1 bosons can be
consistent (i.e. unitary, see Sect. 7.1.3) only if it is gauge invariant [125, 143].
This is the reason why we demand Lagrangians involving vector fields (or 1-
forms) to be invariant under gauge transformations.

7.1.2 The Quantum Field Theory Should Be

Renormalizable
The quantum field theory associated to the Lagrangian should be renormalizable
to yield in the end (after renormalization of the parameters, such as coupling
constants and masses, cf. Sect. B.2.8) finite results that can be compared with
experiments and used to adjust the free parameters of the theory. A simple
calculation of the mass dimension of summands in the Lagrangian determines
which terms have a chance to yield renormalizable theories.
For example, let

be a renormalizable Lagrangian, where ϕ 1, …, ϕ n denote certain fields on

spacetime (not necessarily scalars). Suppose that is Lorentz invariant and, say,
gauge invariant (for instance, could be the Yang–Mills Lagrangian or the
Klein–Gordon Lagrangian). Then for all natural numbers k the k-th power
will also be Lorentz invariant and gauge invariant. However, in almost all cases,
for k ≥ 2, the Lagrangian will be non-renormalizable , because it has the
wrong mass dimension.
Demanding that the quantum field theory is renormalizable thus greatly
restricts the possible terms that can appear in Lagrangians. Calculating mass
dimensions (power counting), it can be shown that in 4-dimensional spacetime
the only renormalizable and gauge invariant Lagrangians are sums of the
Lagrangians that we discuss in this chapter2 (where in the case of the Higgs
Lagrangian for a scalar field ϕ the potential has to be a polynomial in ϕ of degree
less than or equal to 4). See [143, Sect. 12.3] for details.
This is very satisfying, because it means, from the point of view of quantum
field theory in 4-dimensional spacetime, that there will be no additional types of
interactions. It also turns out that all of the allowed Lagrangians actually appear
in the Standard Model (in a certain specific form, i.e. with a specific gauge
group G, specific charged fermions, etc.). The restriction of renormalizability
does not hold for effective Lagrangians, i.e. Lagrangians that are only used for
calculations at low energies.

7.1.3 The Quantum Field Theory Should Be Free of

Gauge Anomalies
Symmetries of the classical field theory, like gauge symmetries, do not
necessarily hold in the quantum field theory. The reason is that the measure
involved in the definition of path integrals may not be invariant under the
symmetry. If this happens, the symmetry is called anomalous.
In quantum theory, we demand that the Hilbert space of the system does not
contain both vectors of positive norm and negative norm (states of negative
norm are called ghost states ). This property is sometimes called unitarity (a
vector space with a positive definite Hermitian scalar product is also known as a
unitary vector space).3 If unitarity does not hold, i.e. there exist both states of
positive and negative norm in the Hilbert space, then the scalar product does not
have a probability interpretation (see Exercise 7.9.1), violating a fundamental
axiom of quantum theory.
It is possible to show that in 4-dimensional Minkowski spacetime, anomalies
of gauge symmetries imply that the quantum theory violates unitarity (this is
related to the fact that the Lorentz metric is indefinite and that the scalar product
on the Hilbert space of the quantum field theory must be Poincaré invariant; see
[125, Chap. 8] for details). It follows that the quantum theory has to be free of
gauge anomalies. In practice, this restricts the possible representations and
charges of the fermions: the contributions of the fermions in the theory to the
gauge anomaly depend on both the gauge groups and fermion representations
and have to cancel each other. The Standard Model is anomaly free, see Sect. 8.
5.8. For more details on anomalies, see [125, Chap. 30].
One therefore has to be careful: even if a gauge theory is well-defined on
the classical level, this may not be true for the associated quantum theory. In
particular, vanishing of gauge anomalies has to be checked for every theory
beyond the Standard Model, like Grand Unified Theories or supersymmetric
extensions.

7.1.4 The Lagrangian of the Standard Model

Our aim in this chapter is to understand each term in the following
Lagrangian, which is essentially the Lagrangian of the Standard Model and
could be called the Yang–Mills–Dirac–Higgs–Yukawa Lagrangian:
7.2 The Hodge Star and the Codifferential
Throughout this chapter, (M, g) is an n-dimensional oriented pseudo-Riemannian
manifold. In physics, M is spacetime and g usually has Lorentzian signature. In
mathematics, M is an arbitrary manifold and g is often taken to be Riemannian.
We first want to understand the Yang–Mills Lagrangian for a
connection A on a principal bundle P → M and derive the associated equation of
motion, called the Yang–Mills equation. This equation is most easily stated using
the codifferential, whose definition involves the Hodge star operator. The metric
g on the manifold M enters the Yang–Mills equation precisely through the Hodge
star. In this section we discuss as a mathematical preparation the Hodge star
operator, the codifferential and some related concepts. We follow the exposition
in [14].

7.2.1 Scalar Products on Forms and the Hodge Star

Operator
The metric g together with the orientation of the manifold M define a canonical
volume form dvol g on M: If e 1, …, e n is an oriented, orthonormal basis of T p
M, then dvol g is characterized by

Lemma 7.2.1 If (U, ϕ) is an oriented chart for M with local coordinates x μ ,

then

where

is the absolute value of the determinant of the matrix with entries

Proof This is Exercise 7.9.2. □

We denote by g μν the entries of the matrix inverse to the matrix with entries g μν .
We can raise indices of tensors in the standard way using g μν . For example,

where the Einstein summation convention is understood.

 The semi-Riemannian metric g on M defines bundle metrics on the vector
bundles of k-forms Λ k T M for all k. This yields scalar products between
sections of these bundles that we can write explicitly as follows:

Definition 7.2.2 For we define the scalar product of forms

as follows: for real-valued k-forms on M we set

where

in a local chart (U, ϕ) of M and the second and third sum extend over all k-tuples
μ 1 …μ k . For complex-valued k-forms on M we set

These scalar products are well-defined, independent of the choice of local chart.
The associated norm is given in both cases by

Remark 7.2.3 On a pseudo-Riemannian manifold the norm is in general not

positive definite. In particular, | ω |2 = 0 does not imply ω = 0. For this reason we
usually try to avoid the notation | ω |2.

Definition 7.2.4 The Hodge star operator

is the linear map defined for real-valued forms by

and for complex-valued forms by

Choosing a local frame, it can be shown that this uniquely defines .

Remark 7.2.5 This definition of the Hodge star operator for pseudo-
Riemannian manifolds does not necessarily coincide with the definition
sometimes found in the literature. Baum [14], for instance, uses the definition

We continue to use our definition.

Suppose e 1, …, e n is an oriented, orthonormal basis of tangent vectors with

Let α 1, …, α n be the dual basis of 1-forms with α i (e j ) = δ j i . Then

and we have:

Lemma 7.2.6 The Hodge star operator is given by

In this formula there is on the right-hand side no summation over indices, {m

k+1, …, m n } is a complementary set to {m 1, …, m k } and ε is totally
antisymmetric with

In particular,

Definition 7.2.7 Let denote the differential forms with compact

support on M. Then we define the L 2 -scalar product of forms

by

We have to restrict the L 2-scalar product to forms with compact support,

because otherwise the integral may not be finite.
We can generalize these constructions to twisted differential forms. Suppose
that E → M is a -vector bundle with bundle metric ⋅ , ⋅ E . Together with the
semi-Riemannian metric g we then get induced bundle metrics on the vector
bundle Λ k T M E of twisted k-forms for all k. More explicitly we can write:
Definition 7.2.8 We define the scalar product of twisted forms

as follows: choose a local frame e 1, …, e r for E over U M and expand k-

forms F, G twisted with E as

with . Then we set

This scalar product is independent of the choice of local frame {e i }. We can

also define a Hodge star operator on twisted forms

by

and an L 2 -scalar product of twisted forms

by

7.2.2 The Codifferential

Let (M, g) be an oriented semi-Riemannian manifold of dimension n and
signature (s, t). We have the usual exterior differential

on forms.

Definition 7.2.9 We define the codifferential

by
The codifferential has the following interesting property:

Theorem 7.2.10 (Codifferential on Forms Is Formal Adjoint of

Differential) Let M be a manifold without boundary. Then the codifferential d
is the formal adjoint of the differential d with respect to the L 2 -scalar product
on forms with compact support, i.e.

for all ω Ω 0 k (M), η Ω 0 k+1(M).

Proof We calculate the difference

with respect to the (pointwise) scalar product of forms. According to

Exercise 7.9.3

is given by

We have

This implies the claim by Stokes’ Theorem A.2.24. □

We want to generalize the definition of the codifferential to twisted forms. Let E

→ M be a -vector bundle with a scalar product ⋅ , ⋅ E and a compatible
covariant derivative . In Sect. 5.12 we defined the associated exterior covariant
derivative (or covariant differential)

Definition 7.2.11 We define the covariant codifferential

by

We then get the following analogue of Theorem 7.2.10.

Theorem 7.2.12 (Covariant Codifferential on Twisted Forms Is Formal
Adjoint of Covariant Differential) Let M be a manifold without boundary.
Then the covariant codifferential d is the formal adjoint of the exterior
covariant differential d with respect to the L 2 -scalar product on forms with
compact support, i.e.

for all ω Ω 0 k (M, E), η Ω 0 k+1(M, E).

Proof We follow the proof in [14]. Since d and d are linear, it suffices to
prove the statement for forms ω, η of the form

Then

and

In particular, with Exercise 7.9.3,

We introduce a scalar product

by setting

and extending linearly.

For the difference of the pointwise scalar products we then get

In the final step we used that is compatible with the scalar product on E.
The claim now follows by Stokes’ Theorem A.2.24. □

7.3 The Yang–Mills Lagrangian

In this section we define the Yang–Mills Lagrangian and derive the associated
Yang–Mills equation. We fix the following data:
an n-dimensional oriented pseudo-Riemannian manifold (M, g)
a principal G-bundle P → M with compact structure group G of dimension
r
an Ad-invariant positive definite scalar product on , determined by
certain coupling constants, as in Sect. 2.5
a -orthonormal vector space basis T 1, …, T r for .
The Ad-invariant scalar product on determines a bundle metric on
the associated real vector bundle that we denote by ⋅ , ⋅ Ad(P).

7.3.1 The Yang–Mills Lagrangian

Let A be a connection 1-form on the principal bundle P with curvature 2-form
. According to Corollary 5.13.5 the curvature defines a twisted 2-
form

Definition 7.3.1 The Yang–Mills Lagrangian is defined by

For a fixed connection A, the Yang–Mills Lagrangian is a global smooth

function

Theorem 7.3.2 The Yang–Mills Lagrangian is gauge invariant , i.e.

for all bundle automorphisms and all connections A on P.

Proof Theorem 5.4.4 implies that the curvature form transforms as

Let f⋅ denote the action of f on the adjoint bundle, given by Theorem 5.3.8. Then

Since the scalar product is Ad-invariant, it follows that ⋅ , ⋅ Ad(P) is

invariant under the action of f −1. This implies the claim. □

We want to find a formula for the Yang–Mills Lagrangian in local coordinates

and in a local gauge. Let s: U → P be a local gauge. Then the local field strength
is given by

The scalar product on the Lie algebra defines a scalar product

As before we set in a chart with coordinates x μ

We can expand

and

where F s Aa Ω 2(U) are real-valued differential forms, are real-

valued smooth functions on U and we sum over the indices a.
We can then write the Yang–Mills Lagrangian locally as

where we sum over all μ, ν. The local field strength is given by

and

with the structure constants defined by

Lemma 7.3.3 The structure constants of the Lie algebra with respect to a
-orthonormal basis {T a } satisfy

and

for all indices a, b, c. In particular,

Proof The first claim is clear, because the Lie bracket is antisymmetric. The
second claim follows, because the T a are an orthonormal basis and the scalar
product on is Ad-invariant: we have

which implies the claim. □

We can therefore also write the structure equation for the curvature as

This implies the following explicit formula for the Yang–Mills Lagrangian:

(7.1)

The term in the second line is quadratic in the gauge field. It describes free
(non-interacting) gauge bosons and is the only term if the group G is abelian .
The terms in the third and fourth line are cubic and quartic in the gauge field
and describe a direct interaction between the gauge bosons in non-abelian
gauge theories. In the case of QCD these terms are called 3-gluon vertex and
4-gluon vertex . Figure 7.1 shows the Feynman diagrams for these vertices.
 Fig. 7.1 Interaction vertices for non-abelian gauge bosons

Remark 7.3.4 In physics, the quantum field theory for a gauge field A
determined by the Yang–Mills Lagrangian, without any additional matter fields,
is known as pure Yang–Mills theory or gluodynamics . For non-abelian Lie
groups, the quantum version of pure Yang–Mills theory predicts particles, known
as glueballs , which only consist of gauge bosons (gluons in QCD). The Clay
Millennium Prize Problem [37] on the mass gap is to prove that the masses of
glueballs in a quantum pure Yang–Mills theory on with compact simple
gauge group G are bounded from below by a positive (non-zero) number.

Remark 7.3.5 The term “gauge invariance” was invented by Hermann Weyl in
1929 for the U(1) gauge theory of electromagnetism. Gauge theory for non-
abelian structure groups G was first developed by Chen Ning Yang (Nobel Prize
in Physics 1957 ) and Robert L. Mills (for G = SU(2)) in the 1950s.

7.3.2 The Yang–Mills Equation

We assume now that
the semi-Riemannian manifold (M, g) is closed, i.e. compact and without
boundary.

Definition 7.3.6 Let denote the space of all connection 1-forms A on the
principal bundle P. This is by the discussion in Sect. 5.13 a (usually infinite-
dimensional) affine space over the vector space
with isomorphism given by the map Λ. For we set

Definition 7.3.7 The Yang–Mills action for a principal G-bundle P → M is

the smooth map

defined by

The integral is well-defined, because M is compact.

Definition 7.3.8 We call a connection A on the principal bundle P a critical

point of the Yang–Mills action if

for all variations

For a connection A on P we denote by d A the associated covariant differential

and by d A the covariant codifferential. We want to prove:

Theorem 7.3.9 A connection A on a principal bundle P → M is a critical

point of the Yang–Mills action if and only if A satisfies the Yang–Mills
equation

i.e.

Proof We follow the proof in [14]. According to the structure equation in

Theorem 5.5.4 we can calculate

This implies
We get with Theorem 7.2.12

Since the scalar product on the Lie algebra is non-degenerate, the L 2-scalar
product on Ω 1(M, Ad(P)) is non-degenerate. It follows that A is a critical point
of the Yang–Mills Lagrangian if and only if d A F M A = 0. □

In a local gauge s: U → P the Yang–Mills equation can be written as

Remark 7.3.10 Recall that any connection A on the principal bundle P has to
satisfy the Bianchi identity , which can be written according to Theorem 5.14.2
as

Atiyah and Bott [6] have noted that the curvature F M A of a connection A that
satisfies in addition to the Bianchi identity the Yang–Mills equation d A F M A
= 0 can thus be considered as a harmonic form (in a non-linear sense if G is
non-abelian) in Ω 2(M, Ad(P)) (compare with Exercise 7.9.5). The Yang–Mills
equation is a second-order partial differential equation for the connection A.

Remark 7.3.11 Note that the Yang–Mills equation depends through the Hodge
star operator on the pseudo-Riemannian metric g on M. If the equation holds for
one metric, it does not necessarily hold for another metric.

Example 7.3.12 (Maxwell’s Equations) In the case when G = U(1), the

local curvature forms F s are independent of the choice of local gauge s and
define a global 2-form , see Corollary 5.6.4. The Bianchi
identity and Yang–Mills equation are then given by

These are Maxwell’s equations for a source-free electromagnetic field (on a

 general n-dimensional oriented pseudo-Riemannian manifold). On
Minkowski spacetime of dimension 4 we can use the construction in Sect. 5.7
to write Maxwell’s equations in terms of the electric and magnetic field .
Maxwell’s equations generalize to any abelian Lie group G. Note that
in this case both the Bianchi identity and the Yang–Mills equation are
linear. For a non-abelian structure group these equations are non-linear
(and therefore much harder to solve).

We could study the Yang–Mills equation on any of the examples of principal

bundles that we defined in Chap. 4, in particular, on the Hopf fibrations over
projective spaces or on the canonical principal bundles over homogeneous
spaces, once (pseudo-)Riemannian metrics on the base manifolds have been
defined.

Definition 7.3.13 We call a connection A on a principal bundle a Yang–Mills

connection if it satisfies the Yang–Mills equation.

Since the Yang–Mills equations do not depend on the choice of local gauge, the
gauge group of the principal bundle P → M acts on the space of Yang–
Mills connections. We can therefore set:

Definition 7.3.14 The Yang–Mills moduli space of a principal bundle P → M

over a pseudo-Riemannian manifold (M, g) is the space of Yang–Mills
connections A modulo the gauge group .

The moduli space is usually the quotient of an infinite-dimensional space by the

action of an infinite-dimensional group. It is therefore non-trivial to define, for
example, a smooth structure on the moduli space.

Example 7.3.15 (Instantons) Let P → M be a principal G-bundle over an

oriented Riemannian 4-manifold (M, g). In this case the Hodge star operator
satisfies = 1 on 2-forms on M. We consider connections A on P with
curvature F M A Ω 2(M, Ad(P)) such that either

or
 Connections that satisfy these identities are called self-dual and anti-self-
dual instantons , respectively (see Exercise 7.9.3 for the notion of self-
duality).
Since any connection A satisfies the Bianchi identity, instantons
automatically satisfy the Yang–Mills equation. The instanton equations are
examples of BPS equations , i.e. special first order equations (here for the
gauge field A) whose solutions are (often) automatically solutions of the
second order field equations (here the Yang–Mills equations). BPS
equations appear in many other parts of physics, for example, in the theory
of magnetic monopoles (Bogomolny equations) or in supergravity (Killing
spinor equations).

The instanton equations are preserved under the action of the gauge
group and we can define instanton moduli spaces . These moduli
spaces, especially for structure groups G = SU(2) and G = SO(3), are the
cornerstone of Donaldson theory , which revolutionized the
understanding of smooth 4-manifolds in the 1980s.

7.3.3 Massive Gauge Bosons

The Yang–Mills Lagrangian

describes massless gauge bosons . Arguments from physics show that gauge
bosons of mass m are described by adding (in a local gauge) a term of the form

(7.2)
to the Yang–Mills Lagrangian. We could try to write this Lagrangian in an
invariant form as above, such as

however the gauge field A does not define an element A M Ω 1(M, Ad(P))
(only the difference of two gauge fields is such a twisted form). This indicates
that the Lagrangian in Eq. (7.2) is not well-defined, independent of local gauge.
It is also easy to see directly that local gauge transformations g: U → G, which
are not constant, in general do not leave the Lagrangian in Eq. (7.2) invariant.

Remark 7.3.16 One of the main features of the Higgs mechanism , discussed in
Chap. 8, is that it allows us to introduce a non-zero mass for gauge bosons with a
gauge invariant Lagrangian. Introducing a mass for gauge bosons is necessary to
describe the weak interaction as a gauge theory, because experiments show that
the W- and Z-gauge bosons of the weak interaction have a non-zero mass.

7.4 Mathematical and Physical Conventions

for Gauge Theories
In mathematics and physics slightly different conventions are used for scalar
products, coupling constants and covariant derivatives. We want to compare
these conventions in this section. We fix the following data:
a compact Lie group G which is either simple or U(1) (the conventions
below can be generalized to any compact Lie group)
an Ad G -invariant positive definite scalar product on the Lie algebra
(if is simple we can take the negative of the Killing form and for
we can take any positive definite scalar product)
a -orthonormal basis S 1, …, S r of the Lie algebra
a real coupling constant g > 0.

1. In mathematics we choose the scalar product

with orthonormal basis

We expand the gauge field and curvature as

The covariant derivative (after a choice of local gauge) is

 The local curvature is

The Yang–Mills Lagrangian is

2. In physics we choose the Hermitian scalar product on associated to

and the orthonormal basis

of . We expand the gauge field and curvature as

There are two different sign conventions for the covariant derivative :

The local curvature is

The Yang–Mills Lagrangian is

3. The correspondence between the mathematical and physical conventions is

given by setting

If the representation of the Lie group G on a vector space V is unitary, then

the field A μ will act as a skew-Hermitian operator and B μ will act as a
Hermitian operator. We have μ A = μ B and
 Most of the time we shall use the mathematical convention and indicate
when we use the physical convention.

Remark 7.4.1 Note one interesting point that can be seen most clearly in the
physical convention: The coupling constant g appearing in the covariant
derivative (describing the coupling of the gauge field to other fields, as we will
see below) is the same as the coupling constant appearing in front of the term [B
μ , B ν ] in the curvature G μν , describing the coupling between the gauge bosons
in non-abelian gauge theories.

7.5 The Klein–Gordon and Higgs Lagrangians

So far we have considered pure gauge theories that involve only a gauge field
(connection) A. In physics, however, we are also interested in matter fields that
couple to the gauge field. We first consider the case of scalar fields, like the
Higgs field. We again fix an oriented pseudo-Riemannian manifold (M, g).

7.5.1 The Pure Scalar Field

Definition 7.5.1 A complex scalar field is a smooth map

A multiplet of complex scalar fields is a smooth map

for some r > 1.

We consider the standard Hermitian scalar product

on . If ϕ is a multiplet of scalar field with values in , then the differential dϕ

is an element

There is an induced Hermitian scalar product on the vector space-valued 1-forms

.
Definition 7.5.2 The free Klein–Gordon Lagrangian for a multiplet of
complex scalar fields of mass m is defined by

For a given field ϕ the free Klein–Gordon Lagrangian defines a smooth map

The expression dϕ, dϕ is called the kinetic term and the expression − m 2 ϕ, ϕ
is called the Klein–Gordon mass term .

In local coordinates on M the kinetic term is given by

It is also useful to consider a more general situation.

Definition 7.5.3 Let be a smooth function, called a potential. Then

the Higgs Lagrangian for a multiplet of complex scalar fields ϕ with potential V
is defined by

where V (ϕ) denotes V ( ϕ, ϕ ) (of course it suffices to define the potential on

). The Higgs field in the Standard Model, which we study in Chap. 8, is a
multiplet of complex scalars described by a similar Lagrangian.

The potential V, if it contains terms of order higher than two in the field ϕ,
describes a direct interaction between particles of the field ϕ. In the Standard
Model, for instance, the potential V of the Higgs field is a quadratic polynomial
in ϕ † ϕ, hence of order four in ϕ.

7.5.2 The Scalar Field Coupled to a Gauge Field

We now consider the case of a scalar field ϕ coupled to a gauge field A. We fix
the following data:
an n-dimensional oriented pseudo-Riemannian manifold (M, g)
a principal G-bundle P → M with compact structure group G of dimension
r
a complex representation ρ: G → GL(W) with associated complex vector
bundle E = P × ρ W → M
 a G-invariant Hermitian scalar product ⋅ , ⋅ W on W with associated bundle
metric ⋅ , ⋅ E on the vector bundle E.
We then define:

Definition 7.5.4 If the dimension of W is one, then a smooth section of E is

called a complex scalar field and if the dimension of W is greater than one, then
a smooth section of E is called a multiplet of complex scalar fields (or simply a
scalar field) and the vector space W is called a multiplet space.

With the exterior covariant derivative

and the scalar product ⋅ , ⋅ E on Ω 1(M, E) we set:

Definition 7.5.5 The Klein–Gordon Lagrangian for a multiplet of complex

scalar fields Φ Γ(E) of mass m coupled to a gauge field A is defined by

For given fields Φ and A the Klein–Gordon Lagrangian is a smooth function

The associated action S KG [Φ, A] is the integral over the Klein–Gordon

Lagrangian (on a closed manifold M).

In local coordinates on M we can write the kinetic term as

It is sometimes useful to have an even more explicit local formula for the Klein–
Gordon Lagrangian: Choosing a local gauge s: U → P, we can write

where ϕ: U → W is a smooth function. The covariant derivative is given by

The term A μ ϕ is called the minimal coupling (we suppress in the notation the
induced representation ρ of the Lie algebra on W). We identify W with
and the scalar product on W with the standard Hermitian product

on . Since the representation of G on W is unitary and the gauge field A μ has

values in , this implies that A μ acts through skew-Hermitian matrices on :

In a local gauge s for the principal bundle, the Klein–Gordon Lagrangian can
then be written as

(7.3)

The two terms in the first line, which are quadratic in the field ϕ with values in
, are the Klein–Gordon Lagrangian for a free multiplet of complex scalar
fields of mass m, consisting of the kinetic term and the mass term .
The terms in the second and third line are cubic and quartic in the fields ϕ
and A μ . These interaction terms describe an interaction (or coupling) between
the gauge field and the multiplet of scalar fields and thus an indirect interaction
between particles of the scalar field, mediated by the gauge bosons (see the
Feynman diagrams after Remark 5.9.5 for a depiction of the interaction between
a scalar field and a gauge field).
We see here (and later in the case of the Dirac Lagrangian for fermions)
that in gauge theories where G does not act diagonally on the multiplet vector
space , the action of the gauge group leads to two related kinds of
mixing:
The representation of the gauge group G on W, defining the associated
bundle E, mixes different components of the multiplet, i.e. different
components are gauge equivalent. In other words, the identification of a
section of E with a map to V and the splitting into components depends
on the choice of gauge.
Via the induced representation of the Lie algebra on W, the gauge field
A pairs different components of the multiplet in the interaction vertices.

This has important consequences for the Standard Model, where different
particles like the up and down quark or the electron and electron neutrino
form SU(2) × U(1)-doublets.

Definition 7.5.6 Sections Φ of an associated vector bundle E = P × ρ V with

non-trivial are called charged scalars . It follows that charged scalars have a
non-trivial coupling to the gauge field A.

Theorem 7.5.7 The Klein–Gordon Lagrangian of a multiplet of complex

scalar fields, coupled to a gauge field, is gauge invariant :

for all bundle automorphisms .

We need the following lemma.

Lemma 7.5.8 Let be a bundle automorphism. Then

Proof This follows from a calculation in local coordinates for d A Φ(X) = X A

Φ with a vector field X. For a more invariant argument, note that by the
definition of covariant derivatives using parallel transport in Sect. 5.9 it suffices
to show that

This follows from Exercise 5.15.9. □

We can now prove Theorem 7.5.7.

Proof The kinetic term d A Φ, d A Φ E and the mass term − m 2 Φ, Φ E are

both separately invariant under gauge transformations, because the scalar
product ⋅ , ⋅ W on the vector space W is G-invariant, hence ⋅ , ⋅ E is invariant
under the action of f −1. □

In the Klein–Gordon Lagrangian for a scalar field the gauge field A is non-
dynamic, i.e. does not appear with derivatives, and is just a fixed background
field. The total Lagrangian that describes the dynamics of the scalar field, the
gauge field and their interactions is the Yang–Mills–Klein–Gordon
Lagrangian

We can also consider the case of a scalar field with a potential coupled to a
gauge field.

Definition 7.5.9 The Higgs Lagrangian for a multiplet of complex scalar

fields coupled to a gauge field is defined by

where V (Φ) is a gauge invariant potential. We only consider the case where

with a function .

This Lagrangian describes an interaction between particles of the scalar field

and particles of the gauge field and in addition a direct interaction between
the particles of the scalar field (if the potential V contains terms of order three
or higher in Φ).

A similar argument to the one in Theorem 7.5.7 shows:

Theorem 7.5.10 The Higgs Lagrangian of a multiplet of complex scalar fields

with potential V and coupled to a gauge field is gauge invariant:

for all bundle automorphisms .

The sum of the Higgs and Yang–Mills Lagrangians is called the Yang–Mills–
Higgs Lagrangian

Remark 7.5.11 It is sometimes useful to consider real scalar fields Φ, which

are sections in vector bundles E associated to real orthogonal representations of
the Lie group G. The Klein–Gordon Lagrangian for a real scalar field of mass m
coupled to a gauge field is

There is an analogous generalization to real scalar fields with a potential V.

7.6 The Dirac Lagrangian
Fermions are described classically by spinor fields on spacetime. In this section
we define a Lagrangian for fermions. We fix the following data:
an n-dimensional oriented and time-oriented pseudo-Riemannian spin
manifold (M, g) of signature (s, t)
a spin structure Spin+(M) together with complex spinor bundle S → M
a Dirac form ⋅ , ⋅ (not necessarily positive definite) on the Dirac spinor
space Δ = Δ n with associated Dirac bundle metric ⋅ , ⋅ S . We abbreviate
Ψ, Φ S by .
We can then define the Dirac Lagrangian for a free spinor field.

Definition 7.6.1 The Dirac Lagrangian for a free spinor field Ψ Γ(S) of
mass m is defined by

where D: Γ(S) → Γ(S) denotes the Dirac operator. The expression is

called the kinetic term and is called the Dirac mass term .

Taking the real part in the kinetic term is necessary, because the Lagrangian has
to be real. If the Dirac form ⋅ , ⋅ S has δ = −1, then the calculation in
Exercise 7.9.12 implies that

for some (n − 1)-form α on M depending on the spinor Ψ Γ(S). As a

consequence the kinetic term of the Dirac Lagrangian satisfies

This implies by Stokes’ Theorem A.2.24 that the action defined by Ψ, DΨ S and
its real part are the same if the manifold M has no boundary and Ψ has compact
support.

7.6.1 The Fermion Field Coupled to a Gauge Field

Similar to a scalar field, a spinor can be coupled to a gauge field. This
construction is very important, because it defines the interaction between matter
particles (fermions) and gauge bosons in gauge theories (for example, the
interaction between electrons and photons in QED or the interaction between
quarks and gluons in QCD ). We fix in addition to the data above the following
data:
a principal G-bundle P → M with compact structure group G of dimension
r
a complex representation ρ: G → GL(V ) with associated complex vector
bundle E = P × ρ V → M
a G-invariant Hermitian scalar product ⋅ , ⋅ V on V with associated bundle
metric ⋅ , ⋅ E on the vector bundle E. Together with the Dirac form on the
spinor bundle S we get a Hermitian scalar product ⋅ , ⋅ S E on the twisted
spinor bundle S E. We again abbreviate Ψ, Φ S E by .
Choosing a local gauge s: U → P and an orthonormal basis v 1, …, v s for V,
the twisted spinors Ψ, Φ correspond to multiplets

where Ψ i and Φ j are sections of the spinor bundle S over U. The scalar
product on S E can then be written as

Definition 7.6.2 The Dirac Lagrangian for a twisted spinor field Ψ Γ(S
E) of mass m coupled to a gauge field A on the principal bundle P is
defined by

where D A : Γ(S E) → Γ(S E) denotes the twisted Dirac operator. The

associated action S D [Ψ, A] is the integral over the Dirac Lagrangian (on a
closed manifold M).
Choosing in addition to the local gauge for P and the orthonormal basis
 for V a local vielbein e for the tangent bundle TM with associated local
trivialization ε of Spin+(M), we can write the Dirac Lagrangian as

(7.4)
where ψ is a map with values in Δ V, ψ j are maps with value in Δ, and
Γ p are physical gamma matrices. Here the two terms in the first line are the
Dirac Lagrangian for a free multiplet of fermions, consisting of the kinetic
term

a coupling between the spinor field and the metric g via ω pqr , and the
Dirac mass term. The term in the second line, which is cubic in the fields, is
the interaction term that describes an interaction between the fermions and
the gauge field and thus an indirect interaction between the fermions (see the
Feynman diagram in Fig. 7.2 for the interaction between a fermion ψ and a
gauge field A p ).

Fig. 7.2 Interaction vertex for fermion and gauge field

The gauge field A p with values in the Lie algebra acts on the V part of ψ
through the induced representation (suppressed in the notation). Since the
gauge field A acts by skew-Hermitian matrices, the interaction term is
automatically real and we can drop the symbol Re.

Definition 7.6.3 Sections Ψ of a twisted spinor bundle S E, where E is

associated to a representation ρ of the gauge group G on a vector space V with

non-trivial, are called charged fermions . It follows that charged fermions have
a non-trivial coupling to the gauge field A.

Theorem 7.6.4 The Dirac Lagrangian for a twisted spinor field is gauge
invariant:

for all bundle automorphisms .

Proof This is Exercise 7.9.11. □

Example 7.6.5 For the strong interaction (QCD) we have G = SU(3),

and there are six multiplets Ψ f , called quarks , for the flavours f = u, d, c, s, t, b.
The three components of every multiplet are called colours . The interaction
term involving the gauge field A μ with values in (corresponding to the
eight gluons) mixes different colours of a quark of a given flavour, but does not
mix different flavours (different flavours of quarks are only mixed by the weak
interaction ). The Lagrangian for QCD can thus be written as

where the sum runs over the six different flavours f.

Remark 7.6.6 Considering the mass term in the Dirac Lagrangian, it is

clear that all components of the multiplet Ψ have the same mass m. We could try
to generalize this and introduce different mass terms for different components of
the multiplet. However, if the components with different masses are related by
the action of a group element g G, then gauge invariance of the Lagrangian
will be lost.
This could be a problem for the Standard Model, because we want to
combine particles with very different masses, like the electron and electron
neutrino, into SU(2) × U(1)-doublets and at the same time keep the Lagrangian
gauge invariant. It turns out that the situation in the Standard Model is even
more difficult, because left-handed and right-handed fermions transform in
different representations of SU(2) × U(1), so that a gauge invariant Dirac mass
term is not defined, even if all components of the multiplet had the same mass.
See Sect. 7.6.2 for more details. As we will discuss in Chap. 8, these problems
can be solved by introducing a Higgs field .

We can again make both the spinor multiplet Ψ and the connection 1-form A
dynamic by considering the Yang–Mills–Dirac Lagrangian

7.6.2 Lagrangians for Chiral Fermions

In this subsection we consider the case of an oriented and time-oriented
Lorentzian spin manifold M of even dimension n with metric of signature (1, n −
1) or (n − 1, 1) (the most interesting case for the Standard Model is Minkowski
spacetime of dimension n = 4). The Dirac bundle metric ⋅ , ⋅ S on the spinor
bundle S → M has a special property: both choices of the matrix A in
Proposition 6.7.13 consist of a product of an odd number of gamma matrices.
Hence if we decompose spinors Ψ, Φ Γ(S) into left-handed (positive) and
right-handed (negative) components we get

(7.5)

In particular, we observe the following:

Proposition 7.6.7 (Dirac Bundle Metrics for Spinors on Lorentz

Manifolds) On even-dimensional oriented and time-oriented Lorentzian
spin manifolds, for both choices of the matrix A in Proposition 6.7.13 , the
Dirac bundle metric is null on the subbundles S L and S R and pairs left-
handed with right-handed spinors. In particular, the decomposition S = S L
S R of the spinor bundle into left-handed and right-handed Weyl spinors is not
orthogonal with respect to the Dirac bundle metric.

Remark 7.6.8 This is different from the situation on even-dimensional

Riemannian spin manifolds, where we can take A = I (the identity matrix) so
that left-handed and right-handed Weyl spinors are orthogonal.
Formula (7.5) also holds if the spinors are sections of a twisted spinor bundle

We get:

Proposition 7.6.9 On even-dimensional oriented and time-oriented Lorentzian

manifolds, for both choices of the matrix A in Proposition 6.7.13 , the (gauge
invariant) Dirac Lagrangian for twisted spinors can be written as

In the second line all three Hermitian scalar products are taken in S E.

We want to generalize this discussion to the case of a twisted chiral spinor

bundle . We consider a twisted chiral spinor bundle over a Lorentzian spin
manifold of even dimension:

Here E L and E R are complex vector bundles associated to representations

We fix G-invariant Hermitian scalar products on V L and V R which define

Hermitian bundle metrics and .
We can then define a massless Dirac Lagrangian as before:

In the second line the first scalar product is taken in S E L and the second
scalar product in S E R (the Dirac operator only acts on the S-component and
does not change the E-component). It is not difficult to check that this
Lagrangian is gauge invariant.
However, if we now also want to define a Dirac mass term as before, we run
into a problem that can ultimately be traced back to Proposition 7.6.7: the natural
mass term

is so far not defined: it pairs a spinor with an E L -component and a spinor

with an E R -component, but the Hermitian bundle metrics are only defined if
both spinors have the same type of E-component.
 We could try to introduce a Dirac mass term in this situation as follows:

Definition 7.6.10 Let V R and V L be unitary representations of a Lie group G.

Then a mass pairing is a G-invariant form

which is complex antilinear in the first argument and complex linear in the
second.

A mass pairing κ defines a form on the level of bundles

that can then be used to define a gauge invariant Dirac mass term for chiral
twisted spinors. However, the following theorem shows that in many cases a
mass pairing vanishes identically:

Theorem 7.6.11 (Triviality of Mass Pairings) Suppose that V L and V R are

irreducible, unitary, non-isomorphic representations of G. Then every mass
pairing κ is identically zero.

Proof We can identify , the dual of the complex conjugate of V L , with

The induced G-representation on this space is defined by

for g G, v L V L . The map

where is the G-invariant Hermitian form on V L , defines a complex linear

G-equivariant isomorphism.
Suppose a mass pairing κ ≠ 0 exists. Then

is a complex linear G-equivariant map. Combining both maps we get a

complex linear G-equivariant map

which is non-zero, because κ ≠ 0. By Schur’s Lemma this map has to be an

isomorphism of the representations V R and V L (because the kernel and image of
the map are G-invariant), contradicting our assumption. □

It is known from experiments that a realistic theory of particle physics has to

involve twisted chiral fermions with a non-zero mass, because the weak
interaction is not invariant under parity inversion (see Sect. 8.5). Together
with Remark 7.3.16 and Remark 7.6.6 it follows that there are three situations
in which it is not clear how to define mass terms and at the same time keep
the Lagrangian gauge invariant:

non-zero masses for gauge bosons

different masses for fermions in the same gauge multiplet

non-zero masses for twisted chiral fermions.

We shall see in Chap. 8 that the introduction of the Higgs field allows a
very elegant solution of these problems: using the Higgs field we can define a
fully gauge invariant Lagrangian that contains certain interaction terms
between the gauge bosons and the Higgs field and the fermions and the Higgs
field. In a specific type of gauge, called a unitary gauge, these interaction
terms take the form of mass terms for the gauge bosons and fermions.

7.7 Yukawa Couplings

In this section we discuss Yukawa couplings which are used in the Standard
Model to define a mass for twisted chiral fermions. Yukawa couplings are
certain trilinear forms involving two twisted chiral spinors and one scalar field.
The idea is that the G-representation on the scalar field precisely cancels the
difference between the representations on the twisted chiral spinors so that the
whole trilinear expression is gauge invariant. We consider the case of an oriented
and time-oriented Lorentzian spin manifold (M, g) of dimension n with signature
(1, n − 1) or (n − 1, 1) together with a principal G-bundle P → M.
Definition 7.7.1 Suppose that V L , V R , W are unitary representation spaces of
the compact Lie group G. Then we define a Yukawa form as a map

which is invariant under the action of G, complex antilinear in V L , real linear in

W and complex linear in V R .

Suppose τ is a Yukawa form. We then define:

Definition 7.7.2 For a real constant g Y the G-invariant scalar

is called a Yukawa coupling (the constant g Y is also called a Yukawa coupling

and sometimes already appears in the definition of the Yukawa form τ). It
defines a gauge invariant Lagrangian for which we use the shorthand notation

where the Yukawa form τ is implicit,

and E L , F, E R are the complex vector bundles associated to the principal bundle
P via the G-representations V L , W, V R .

We will discuss in Chap. 8 how Yukawa coupling between two twisted chiral
fermions and the Higgs field leads to masses for the fermions. The
Lagrangian of the Standard Model is then essentially the sum of all the
Lagrangians that we discussed in this chapter, i.e. the following Yang–Mills–
Dirac–Higgs–Yukawa Lagrangian :

Remark 7.7.3 In the discussions in this chapter, the pseudo-Riemannian metric

g on the manifold M has been considered as a fixed background. Classically we
can add a Lagrangian for the metric (like the Einstein–Hilbert Lagrangian) to
make g dynamic. However, this approach does not yield a well-defined quantum
field theory. Since we are mainly interested in the Standard Model, which is
defined on flat Minkowski spacetime of dimension 4, we will not discuss aspects
of quantum gravity.

7.8 Dirac and Majorana Mass Terms

So far we have considered Dirac mass terms for spinor fields. For a spinor Ψ
Γ(S) such a mass term is given by

where ⋅ , ⋅ is a Dirac form on the spinor space. We want to discuss a second

type of mass term that is important in neutrino physics.

Definition 7.8.1 Let (⋅ , ⋅ ) denote a Majorana form on the spinor space Δ as in

Sect. 6.7.1. Then

is called a Majorana mass term .

It is clear that both Dirac and Majorana mass terms are invariant under the action
of the spin group. We want to compare these forms in the case of Minkowski
spacetime of dimension 4. Recall from Sect. 6.8 that the Dirac form is defined by
the matrix

and the Majorana form is defined by the matrix

If we decompose a Dirac spinor Ψ into left-handed and right-handed Weyl

spinors

then the Dirac mass term is given by

and the Majorana mass term is given by

Here we used the notation for the Majorana conjugate from Definition
6.7.6. The important consequence is that the Dirac mass term is zero for spinors
which have only one Weyl component ψ L or ψ R , while the Majorana mass term
may be non-zero in this case.
We briefly want to discuss the extension of these Lorentz invariant mass
terms to Lorentz and gauge invariant mass terms for charged fermions,
i.e. sections of twisted spinor bundles S E. In the case of the Dirac mass term
we saw in Sect. 7.6.2 that such an extension is always possible if both left-
handed and right-handed Weyl spinor bundles are twisted with the same
associated vector bundle E, using the Hermitian scalar product on S E,
coming from the Dirac form on S and a Hermitian scalar product on E.
In the case of the Majorana mass term there is now a problem, because the
complex bilinear Majorana form on S usually does not combine with the
Hermitian scalar product on E. If E is the associated bundle P × ρ V, then we
need a G-invariant complex bilinear form on the vector space V. However, even
in simple situations such an invariant bilinear form does not exist:

Lemma 7.8.2 Let

be the complex representation ofU(1) on of winding number k. Suppose that B

is aU(1)-invariant complex bilinear form on . Then B ≡ 0.

Proof We have

It follows that B(z, z) = 0 for all , hence B ≡ 0. □

This indicates that there is no straightforward extension of the Majorana mass

term to charged fermions .

7.9 Exercises for Chap. 7

7.9.1 (From [ 125 ]) Let H be a Hilbert space with a bilinear form

satisfying

where denotes complex conjugation. We say that the bilinear form has a
probability interpretation if the following holds: for all vectors ϕ, ψ H with

the following inequality holds:

1. Suppose that the bilinear form is positive definite. Prove that the bilinear
form has a probability interpretation.

2. Suppose that there exist vectors ϕ, ψ in H such that

Prove that the bilinear form does not have a probability interpretation.
7.9.2 Let (M, g) be an oriented pseudo-Riemannian manifold and (U, ϕ) an
oriented chart for M with local coordinates x μ . Prove that the volume form dvol
g is given by

where

is the absolute value of the determinant of the matrix with entries

7.9.3 Let (M, g) be an n-dimensional oriented pseudo-Riemannian manifold

of signature (s, t) and the Hodge star operator.

1. Prove that

is given by
2. Determine the even dimensions n = 2k where = 1 on Ω k (M) if (M, g)
is Riemannian or Lorentzian. In these dimensions we can define self-dual
and anti-self-dual k-forms ω, satisfying ω = ω and ω = −ω,
respectively.
7.9.4 Let (M, g) be an n-dimensional oriented pseudo-Riemannian manifold
of signature (s, t) and the Hodge star operator.

1. Let denote the Levi-Civita connection of g and suppose that α Ω 1(M)

is a 1-form. Prove that if α is parallel ( α = 0), then α is closed (dα = 0).

2. Let η Ω 1(M) be a 1-form, p M a point and e 1, …, e n a local oriented

g-orthonormal frame of the tangent bundle in an open neighbourhood of p
with ( e i )( p) = 0 for all i. Let η i = η(e i ) and η i = g ii η i (no summation).
Prove that at the point p

7.9.5 Let (M, g) be a closed (compact without boundary) n-dimensional

oriented pseudo-Riemannian manifold of signature (s, t). The Laplace operator
on k-forms is defined by

where d is the codifferential from Definition 7.2.9. A form ω is called

harmonic if Δω = 0. Suppose that (M, g) is Riemannian.

1. Prove that

2. Prove that

7.9.6 Let (M 4, g) be a pseudo-Riemannian 4-manifold with a principal

bundle P → M. Prove that the Yang–Mills action S Y M [A] is invariant under a
conformal change of the metric g:
where is an arbitrary smooth function on M.
7.9.7

1. Prove that the connection A from Sect. 5.2.2 on the Hopf bundle S 3 → S 2
with structure group U(1) satisfies the Yang–Mills equation (i.e. Maxwell’s
equations) if S 2 has the standard round Riemannian metric.

2. Prove that the Yang–Mills moduli space for the Hopf bundle S 3 → S 2 over
the round sphere S 2 consists of a single point.
7.9.8 Let be Minkowski spacetime with the flat Minkowski metric
η. Let P → M be a trivial principal G-bundle with a global gauge s: M → P. For
a connection A decompose the curvature F = F A as in Sect. 5.7 into generalized
electric and magnetic fields E and B with values in the Lie algebra .

1. Express the Bianchi identity and the Yang–Mills equation in terms of E, B

and A.

2. Express the instanton equations F = F and F = −F in terms of E and B.

7.9.9

1. On the Hopf bundle S 7 → S 4 with structure group SU(2) define in analogy

to the construction in Sect. 5.2.2 an explicit connection 1-form
using quaternions.

2. Prove that A is an anti-self-dual instanton for the standard round Riemannian

metric on S 4.
7.9.10 Let (M, g) be a closed oriented pseudo-Riemannian manifold, P → M
a principal G-bundle with compact structure group G and E → M an associated
vector bundle with Hermitian bundle metric ⋅ , ⋅ E . We fix an Ad-invariant
positive definite scalar product on the Lie algebra and consider the Yang–
Mills–Higgs Lagrangian
We are looking for critical points of the associated action S Y MH under
variations of Φ and A.

1. Prove that variation of the field Φ leads to the field equation

(7.6)
where V ′ is the derivative of and V ′(Φ) = V ′( Φ, Φ E ).

2. Show that elements α M Ω 1(M, Ad(P)), Φ Γ(E) define a canonical

twisted 1-form α M ⋅ Φ Ω 1(M, E).

3. Prove that there exists a unique twisted 1-form

such that

for all α M Ω 1(M, Ad(P)).

4. Show that variation of the connection A leads to the field equation

(7.7)
Equations (7.6) and (7.7) are called Yang–Mills–Higgs equations .
7.9.11 Prove the statement in Theorem 7.6.4 concerning the gauge invariance
of the Dirac Lagrangian.
7.9.12

1. Under the assumptions of Theorem 6.11.5, define a 1-form by

and prove that

(Exercise 7.9.4 could be helpful).

2. Prove Theorem 6.11.5.

3. Discuss what can be said in the case δ = +1 and the implications for the
Dirac Lagrangian.
7.9.13 Let (M, g) be an n-dimensional closed oriented and time-oriented
pseudo-Riemannian spin manifold, S → M a spinor bundle with Dirac bundle
metric ⋅ , ⋅ S with δ = −1, P → M a principal G-bundle with compact structure
group G and E → M an associated vector bundle with Hermitian bundle metric
⋅ , ⋅ E . We fix an Ad-invariant positive definite scalar product on the Lie
algebra and consider the Yang–Mills–Dirac Lagrangian

We are looking for critical points of the associated action S Y MD under

variations of Ψ and A.

1. Prove that variation of the spinor Ψ leads to the Dirac equation

2. Show that α M Ω 1(M, Ad(P)) and Ψ Γ(S E) define via Clifford

multiplication a canonical section α M ⋅ Ψ Γ(S E).

3. Prove that there exists a unique twisted 1-form

such that

for all α M Ω 1(M, Ad(P)).

4. Show that variation of the connection A leads to the field equation

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
 oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum:
Bull. Am. Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]
17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory.
W. A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives.
Eur. Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson
32.
Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
 York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
 [MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories.
Phys. Rev. Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry.
Phys. Rev. D 2, 1285–1292 (1970)
65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles.
Phys. Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]
82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles.
Bull. Am. Math. Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]
98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings.
Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]
114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics.
Phys. Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)
[MATH]
128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]
145.
 Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions.
Chin. Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/

Footnotes
1 I thank Anthony Britto for pointing out this reference.

2
An exception, that we do not discuss in this book, is the topological theta term , that

appears in some modifications of QCD and in supersymmetric gauge theories.

3 There is another concept of unitarity (unitarity of the S-matrix, i.e. of time evolution) that we do not
consider here.
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_8
Chapter 8 The Higgs Mechanism and the
Standard Model
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

In this chapter we finally apply the formalism of mathematical gauge theory to

physics. We first discuss gauge theories in which the gauge symmetry is
spontaneously broken, leading to the existence of one or several Higgs bosons.
We also study the Standard Model of elementary particles in some detail,
including the particle content and the representations of the gauge group, the
Higgs mechanism of mass generation and the explicit Lagrangians containing
the interactions between all known elementary particles. For the Higgs
mechanism we first consider the case of a general Lie group and Higgs vector
space (with possibly several Higgs bosons) as well as the specific case of the
Standard Model (with a single Higgs boson).
The predictions of the Standard Model, which was developed in the 1960s
and 1970s, have been tested and verified with enormous accuracy, especially
using different types of particle colliders. Typical colliders involve two
accelerated and collimated beams of particles brought to collision, for example,
of protons-protons, protons-antiprotons, electrons-positrons, protons-electrons or
protons-positrons. The final particles in the Standard Model, which were
postulated by the theory and then observed in experiments, were the top quark
(1995 at Fermilab ), the tau neutrino (2000 at Fermilab ) and the Higgs boson
(2012 at CERN ).
It is known that the Standard Model is not a complete theory of particle
physics. For example, the neutrinos in the Standard Model are massless, but
experiments show that they have a small non-zero mass. Furthermore,
observations of galaxies indicate that there is another form of matter in the
universe, called dark matter, that only interacts with Standard Model particles
through gravity and perhaps the weak force. This type of matter cannot be
explained with the Standard Model. Finally, the gravitational interaction is not
included in the Standard Model. The reason is that on the fundamental level the
Standard Model is a quantum field theory and this kind of quantum theory most
probably cannot be used to describe gravity. To find a quantum theory of gravity,
and a unification with the other forces described by the Standard Model, is the
primary aim of Theories of Everything, like string theory.
Even though the Standard Model is not complete, it is extremely well
verified and will be a touchstone for any theory that tries to go beyond it.
Beyond its significance for particle physics, the Standard Model has a
remarkable mathematical structure that develops from basic principles (specific
gauge groups, representations and Lagrangians) through a sequence of
mathematical steps into a complex theory with rich and often unexpected
properties. Our aim in this chapter is to understand to some degree the
fundamental principles of the theory and how they generate the complexity of
particle physics.
The references for this chapter are the same as those given at the beginning
of Chap. 7

8.1 The Higgs Field and Symmetry Breaking

8.1.1 The Yang–Mills–Higgs Lagrangian
We fix the following data:
an n-dimensional oriented pseudo-Riemannian manifold (M, g)
a principal G-bundle P → M with compact structure group G of dimension
r
a complex representation ρ: G → GL(W) with associated complex vector
bundle E = P × ρ W → M (sometimes the vector space W and the
representation ρ are real)
a G-invariant Hermitian scalar product ⋅ , ⋅ W on W with associated bundle
metric ⋅ , ⋅ E on the vector bundle E. We denote by ⋅ , ⋅ W = Re ⋅ , ⋅ W
the associated positive definite Euclidean scalar product on the real vector
space underlying W.

Definition 8.1.1 We call the vector space W the Higgs vector space , the
associated vector bundle E = P × ρ W the Higgs bundle and a section Φ of E the
Higgs field . If the induced representation ρ of the Lie algebra is non-trivial,
then the Higgs field is a charged scalar .

We can assume that and the G-invariant scalar product on W is given by

the standard Hermitian product

We consider a potential, also called the Higgs potential ,

The potential appears in the Higgs Lagrangian for the Higgs field Φ:

where

The combined Lagrangian for the Higgs field and the gauge field is then the
Yang–Mills–Higgs Lagrangian

8.1.2 Spontaneously Broken Gauge Theories

Definition 8.1.2 A vacuum configuration or vacuum for the Yang–Mills–
Higgs Lagrangian is a pair (Φ 0, A 0) comprising a Higgs field and a connection
such that:

1. A 0 is a flat connection, .

2. Φ 0 is covariantly constant, .

3. The value of Φ 0 is at every point of M a minimum of the potential V.

Remark 8.1.3 It can be shown that vacuum configurations in this sense

correspond to the minima of the energy determined by the Yang–Mills–Higgs
Lagrangian and can thus be considered “stable”. It follows from Exercise 7.9.10
that vacuum configurations are solutions of the classical equations of motion.

The structure of vacua of the Yang–Mills–Higgs Lagrangian for bundles E over

general manifolds M can be quite complicated, due to the existence of non-trivial
flat connections on the principal bundle P. For our purposes of understanding the
Standard Model it suffices to restrict from now on to the following case:
the manifold M is connected and simply connected.
Then Exercise 5.15.4 implies that a flat connection A 0 can only exist on
trivial principal bundles P. Hence we also assume that
the principal G-bundle P is trivial.
This also implies that the Higgs bundle E is trivial. We define:

Definition 8.1.4 A vacuum vector is an element w 0 W which is a

minimum of the real-valued function

on W. The set of vacuum vectors in the Higgs vector space W is called the space
of vacua or the vacuum manifold for V.

We then get:

Proposition 8.1.5 Suppose that the manifold M is connected and simply

connected and the principal bundle P is trivial. Let (Φ 0, A 0) be a vacuum
configuration. Then there exists a global gauge s 0: M → P, called the vacuum
gauge , such that
(8.1)
and
(8.2)
where w 0 W is a constant vacuum vector. Conversely, for an arbitrary fixed
global gauge s 0 of the principal bundle P, every vacuum vector w 0 determines a
unique vacuum configuration (Φ 0, A 0) of the form in Eqs. (8.1) and (8.2).

Proof The first statement follows from Exercise 5.15.4. The second statement
then follows, because with respect to the global gauge s, the covariant derivative
on E is just the standard derivative on vector-valued functions from M to W.
Hence Φ 0 is covariantly constant if and only if it is constant as a map to W. □

We collect certain assumptions that hold throughout Sect. 8.1 and Sect. 8.2. We
fix from now on:
 a global vacuum gauge s 0: M → P
a vacuum vector w 0 W
the associated vacuum configuration (Φ 0, A 0).
Recall that we have a unitary representation of the Lie group G on W.

Definition 8.1.6 The unbroken subgroup of the vacuum configuration is the

isotropy group of the vacuum vector w 0 W:

The group H is a closed Lie subgroup of G according to Proposition 3.2.9. Since

G was assumed compact, H is compact as well. We call the gauge theory
spontaneously broken if H is a proper subgroup of G, i.e. .

We will assume from now on that the gauge theory is spontaneously

broken.
Note that the unbroken subgroup H is only well-defined for a constant
vacuum vector w 0 and that its explicit embedding in G depends on the choice of
w 0.
We will assume from now on that the Higgs potential V (w) has a minimum,
but not at w = 0, so that w 0 ≠ 0. We call the spontaneous process (usually
assumed to have happened moments after the Big Bang) where the Higgs
field acquires from the unstable value w = 0 the value w 0 ≠ 0
(spontaneous) symmetry breaking .
Mathematically the basic idea of symmetry breaking is that the isotropy
group of the vector w 0 ≠ 0 is smaller than the isotropy group G = G 0 of
the vector 0, hence some of the original symmetry has disappeared, is hidden or
broken.

Definition 8.1.7 We call the nowhere vanishing field Φ 0 the Higgs

condensate . The Higgs condensate is a non-zero background field in which all
other fields and the corresponding elementary particles propagate. The Higgs
field is the only classical field in the Standard Model with a non-zero value in
vacuum.

Remark 8.1.8 In quantum field theory, fields become fields of operators on the
Hilbert space of the system and do not have classical values. In particular, the
Higgs condensate is thought of as the vacuum expectation value (vev)

of an operator field Φ 0, where | Ω is the vacuum state. For our purposes it

suffices to treat the Higgs condensate as a classical section Φ 0 of the vector
bundle E.

In gauge theories we demand that the Lagrangian and hence the laws of
physics are invariant under all gauge transformations for the structure group
G. This is still true in a spontaneously broken gauge theory. However, in this
case the Higgs condensate is only invariant under gauge transformations with
values in the smaller subgroup H (compared to a gauge theory where the
vacuum value is zero and hence invariant under all gauge transformations).

As we will see later, the non-zero value of the Higgs condensate after
symmetry breaking and the coupling of the fields (gauge fields and matter
fields) to the Higgs field are precisely the reasons why some elementary
particles (gauge bosons and fermions) have a non-vanishing mass.

Example 8.1.9 We consider the case of the electroweak interaction in the

Standard Model. The manifold M is 4-dimensional flat Minkowski spacetime
with Minkowski metric η. We set

where the indices L and Y denote weak and hypercharge , and consider the Higgs
vector space

with the standard Hermitian scalar product and the unitary representation

Here n Y is a certain non-zero natural number that will be fixed to n Y = 3 in

Remark 8.3.1.
The potential of the Higgs field is
 (8.3)
with certain constants μ, λ > 0. It is clear that the Higgs potential is
invariant under the action of G. The choice of Higgs potential is restricted by
the conditions that

it is G-invariant

it is a polynomial of order at most four in w, so that the interaction

defined by it is renormalizable

it has a minimum, but not in w = 0.

Exercise 8.11.1 shows that the Higgs potential then must have the form V
(w) in Eq. (8.3).

A vector w 0 W is a vacuum vector for the Higgs potential V if and only if

We sometimes set

It follows that the vacuum manifold is a 3-sphere in around the origin of

radius | | w 0 | |. All vacua are gauge equivalent under constant gauge
transformations, i.e. the group G acts transitively on the vacuum manifold .
The unbroken subgroup H is isomorphic to U(1), but its embedding into G
depends on the precise choice of w 0. We choose

Then H is the diagonal subgroup

(8.4)
Note that H is not the second factor of G. The index Q stands for
electromagnetic (we will explain this later in Sect. 8.3).
If we reduce to and S 3 to S 1, then the Higgs potential V has the form
of a Mexican hat ; see Fig. 8.1. The values of the parameters μ and λ realized in
nature have to be determined from experiments. It follows from the experimental
values collected in Sect. 8.3.4 that

Fig. 8.1 The Higgs potential V (w) = −μ | w |2 + λ | w |4 on reduced to

8.1.3 The Hessian of the Higgs Potential

Let w 0 be a vacuum vector and s 0 a global vacuum gauge of the principal
bundle P. We write the Higgs field as

with a globally defined map

also called Higgs field. The Higgs condensate is the constant field with value w
0. We consider Higgs fields ϕ whose values are in the vicinity of the Higgs
condensate:
where Δϕ is called the shifted Higgs field . We want to derive an approximation
to V (ϕ) for small values of Δϕ using the Taylor expansion of the potential V up
to terms of second order in Δϕ.

Remark 8.1.10 Note that if V is a polynomial in ϕ, like in the electroweak

theory, then the full Taylor expansion of V around w 0 is a polynomial in Δϕ,
i.e. has only finitely many terms. In the Standard Model this will allow us to
determine the cubic and quartic self-interactions of the Higgs boson in
Theorem 8.7.2.

The Higgs vector space W (more precisely, the tangent space to W at the point w
0) can be split orthogonally as

where is the orbit of the gauge group G through the vacuum vector w 0,

and is the orthogonal complement with respect to the positive definite

scalar product

It follows from Corollary 3.8.10, since the Lie group G is compact, that the orbit
is an embedded submanifold of W, diffeomorphic to the quotient space G⁄H,
where H is the unbroken subgroup.

Definition 8.1.11 We set d for the dimension of . This is equal to

the codimension of the Lie subgroup H in G.

Let Hess(V ) denote the Hessian of the potential V. The Hessian is a symmetric
linear map

on the tangent space to W at a point w and can be defined on any Riemannian

manifold as

where in our situation X is a vector tangent to W and denotes the (flat) Levi-
Civita connection induced by the positive definite inner product ⋅ , ⋅ W on W.
This linear map is symmetric in the sense that
As a matrix the Hessian is given by the symmetric matrix of second derivatives

in standard coordinates x 1, …, x 2n on the real vector space underlying W.

We want to diagonalize this symmetric linear map in the vacuum point w 0.
Since V has a minimum along the whole orbit of G through w 0 (because the
potential is invariant under the unitary action of G), it follows that grad V
vanishes along the orbit and hence

for vectors tangent to the orbit. This implies:

Lemma 8.1.12 The HessianHess(V ) preserves the orthogonal splitting of

into the vector space tangent to the orbit and its orthogonal complement.

Proof We have just argued that

The symmetry of the Hessian then implies for every and

that

and thus

We can therefore find a diagonalization of the Hessian adapted to the splitting of

the tangent space into the vector space tangent to the orbit and its
orthogonal complement.

Proposition 8.1.13 (Orthonormal Eigenbasis for Hessian of the Higgs

Potential) There exist real orthonormal bases
e 1, …, e d of

f 1, …, f 2n−d of
 consisting of eigenvectors of the HessianHess(V ) in w 0 , where
the e i have eigenvalue 0 and
the f j have non-negative eigenvalues (because w 0 is a local minimum). We
set for the eigenvalue of f j (with ).

8.1.4 The Nambu–Goldstone and Higgs Bosons

Definition 8.1.14 We expand the shifted Higgs field in the
orthonormal eigenbasis of the Hessian from Proposition 8.1.13:

where ξ i and η j are real scalar fields on the spacetime manifold M. The ξ i
are called Nambu–Goldstone bosons , the η j are called Higgs bosons .

It follows that

the number d of Nambu–Goldstone bosons is equal to the dimension

of G⁄H

the number 2n − d of Higgs bosons is equal to the real dimension of

the Higgs vector space W minus the dimension of G⁄H.

It is important to distinguish between the Higgs field and the Higgs bosons :
the Nambu–Goldstone bosons correspond to perturbations of the Higgs field ϕ
along the orbit of G through the vacuum vector w 0, while the Higgs bosons
correspond to perturbations orthogonal to the orbit. As particles in the associated
quantum field theory the Nambu–Goldstone and Higgs bosons are thus minimal
excitations of the Higgs condensate.

Theorem 8.1.15 (Taylor Expansion of Higgs Potential) Up to second order

in the shifted Higgs field, we have by the Taylor expansion around the vacuum
vector w 0
Proof The Taylor formula up to second order in Δϕ is

The claim now follows, because w 0 is a minimum of V, hence grad V (w 0) = 0,

and e i and f j are an orthonormal basis of eigenvectors of the Hessian with
eigenvalues 0 and .□

We conclude that the potential V is up to second order (and up to the irrelevant

constant V (w 0)) the sum over the standard Klein–Gordon mass terms for real
scalar fields η j of mass . We shall see in Sect. 8.2.3 that the Lagrangian
for the Higgs field ϕ expressed in terms of the shifted Higgs field contains
Klein–Gordon summands of the form

and

plus other terms, some of which we shall determine later. This discussion
implies:

Corollary 8.1.16 The Nambu–Goldstone bosons are real scalar fields of mass
zero and the Higgs bosons are real scalar fields of mass .

The terms of higher than quadratic order in the Taylor expansion of the potential
V around w 0 can be interpreted as interactions between Higgs bosons (see
Theorem 8.7.2 for the case of the Standard Model).

Example 8.1.17 We continue with Example 8.1.9. In the case of the

electroweak theory we have G = SU(2) L × U(1) Y and H = U(1) Q , embedded in
G as a diagonal subgroup, not as the second factor. Therefore we have three
Nambu–Goldstone bosons ξ 1, ξ 2, ξ 3 and one Higgs boson η (also denoted by
H, not to be confused with the unbroken gauge group), because W has complex
dimension 2.
 The Higgs potential is of the form

We can check that the mass of the Higgs boson is : We choose as

before the vacuum vector

Then is the real span of the vectors

and is the real span of the vector

The Higgs field can be decomposed as

with real scalar fields ξ j and H. In the standard coordinates x 1 + ix 2, x 3 + ix

4 for we have

Then

and

In the basis for

we get for the Hessian

This implies 2m H 2 = 4μ, hence the mass of the Higgs boson is .
Note that the Hessian in the given basis is diagonal and the eigenvalues vanish in
the direction along the orbit, while the eigenvalue along the direction of
the vector f orthogonal to the orbit is positive, as expected.
The mass of the Higgs boson is thus determined by the quadratic
self-coupling of the Higgs field or, if we write

by the absolute value of the Higgs condensate and the quartic self-coupling.

8.1.5 Unitary Gauge and the Nambu–Goldstone Bosons

Recall that the principal bundle P → M is trivial and we fixed a global gauge s
0: M → P, called the vacuum gauge. The Higgs field Φ is given by

with a smooth map ϕ: M → W.

We now consider gauge transformations on P. With respect to the global
gauge s 0 they are given by physical gauge transformations

that act on ϕ by

where the G-representation on W is implicit. Note that, if we think of the

manifold M as spacetime, physical gauge transformations τ in general are time-
dependent.

Definition 8.1.18 For a given Higgs field ϕ, we call a smooth, physical gauge
transformation τ: M → G a unitary gauge with respect to a vacuum vector w 0 if
all Nambu–Goldstone bosons of the transformed field τ ⋅ ϕ with respect to the
vacuum vector w 0 vanish identically on M. We then say that the transformed
Higgs field ϕ′ = τ ⋅ ϕ is in unitary gauge (with respect to the vacuum vector w 0).

An equivalent condition is:

Lemma 8.1.19 A Higgs field ϕ is in unitary gauge with respect to the vacuum
vector w 0 if the shifted Higgs field

is at every point x M orthogonal to the tangent space to the orbit

.

The physical intuition behind unitary gauges is that Nambu–Goldstone bosons

are not physical particles, but can be gauged away, in contrast to the Higgs
bosons. Some details about the general existence of unitary gauges can be found
in [16]. We only need the following statement.

Theorem 8.1.20 (Existence of Unitary Gauges in the Electroweak Theory)

Consider the electroweak gauge theory as in Example 8.1.9 with gauge group G
= SU(2) L × U(1) Y and Higgs field of the form

where . Assume that ϕ 2(x) ≠ 0 for all . Then there

exists a physical gauge transformation τ: M → G such that

where is a real-valued function. The transformed Higgs field τ ⋅ ϕ is

then in unitary gauge with respect to the vacuum vector

Proof We just sketch the proof and leave the details to Exercise 8.11.2. We
write

where all functions r, θ, s, μ are real-valued and s(x) ≠ 0 for all x M by

assumption. Since , the functions θ and μ exist as globally well-
defined single-valued functions (otherwise we would have to work directly with
S 1-valued functions instead of e iθ and e iμ ).
We can first find an explicit smooth SU(2) gauge transformation to get ϕ into
the form

A suitable U(1) smooth gauge transformation then yields ϕ in the form

Setting ψ = s′, the claim follows. □

8.2 Mass Generation for Gauge Bosons

The purpose of this section is to describe how masses of gauge bosons are
generated through the Higgs field.

8.2.1 Broken and Unbroken Gauge Bosons

We fix in addition to the data above the following data:
an Ad-invariant positive definite scalar product on the Lie algebra ,
determined by certain coupling constants with respect to an arbitrary fixed
scalar product.
Let H G denote the stabilizer subgroup of the vacuum vector w 0 W.
We denote by the Lie algebra of H and by its orthogonal complement
with respect to the scalar product . We denote the dimension of as
before by d.

Definition 8.2.1 We define the positive semi-definite, bilinear symmetric mass

form m on by

Here A ⋅ w 0 denotes the induced representation ρ of the Lie algebra on the

Higgs vector space W.

According to Proposition 3.2.10 the kernel of the map

is equal to the Lie algebra of the isotropy group. Hence A ⋅ w 0 = 0 for all
, while the map A ↦ A ⋅ w 0 is injective on the orthogonal complement . This
implies that m(A, ⋅ ) ≡ 0 if and the restriction of m onto the orthogonal
complement is positive definite. Since m is a symmetric form, we can
diagonalize it and get:

Proposition 8.2.2 We can find -orthonormal bases

α 1, …, α d of the subspace , called broken generators , and

α d+1, …, α r of the subspace , called unbroken generators

such that the symmetric mass form m is diagonal in this basis. We can write

with M a > 0 for the broken generators and M a = 0 for the unbroken
generators.

If is a gauge field, then we can decompose

where A μ a are the broken and unbroken gauge bosons , respectively. The
numbers M a are called the masses of the gauge bosons .

Remark 8.2.3 Note two interesting facts: the masses M a are proportional to the
norm | | w 0 | | of the value of the Higgs condensate and they also depend on the
choice of the scalar product on (and thus on the coupling constants ) via
the choice of orthonormal basis {α a }.

8.2.2 The Combined Lagrangian

For a gauge field A μ we consider as before (in the mathematical convention) the
covariant derivative

and the curvature (or field strength)

We can also express the field strength in our chosen orthonormal basis of broken
and unbroken generators for the Lie algebra :

The total Lagrangian for the Higgs field ϕ and the gauge field A μ is:

There is a summation over Lie algebra indices a in the last term.

We write ϕ = w 0 + Δϕ with the shifted Higgs field Δϕ as before. We would
like to determine the terms up to second order in the fields Δϕ and A μ in the
Lagrangian , i.e. the terms corresponding to “free” fields. Higher-order
terms contain interactions between these fields.

Proposition 8.2.4 Up to terms of second order the Lagrangian is given

by

Proof We have

However, A μ ⋅ Δϕ is already quadratic in the fields and can be ignored. This

implies the terms in the first line of the equation. The second line is clear (where
V (ϕ) should be taken up to second order in Δϕ). We have also ignored terms of
order 3 and 4 in the gauge field A μ , appearing in the Yang–Mills Lagrangian for
non-abelian gauge theories. □

8.2.3 Simplifying the Lagrangian

We want to simplify each of the summands in the Lagrangian in
Proposition 8.2.4. Recall that according to Proposition 4.7.6 we can identify
sections of associated vector bundles with vector space-valued maps once we
have chosen a gauge for the principal bundle. So far we have not specified this
choice of gauge in the formula in Proposition 8.2.4. We now make the following
assumption:
 The Higgs field ϕ is in unitary gauge with respect to the vacuum vector w 0.
By this we mean that all Nambu–Goldstone bosons of ϕ vanish on all of the
manifold M identically.

Remark 8.2.5 We have not discussed the existence of unitary gauges in

general. However, according to Theorem 8.1.20 such gauges exist in the
electroweak theory if the second component ϕ 2 of the Higgs field ϕ is
everywhere on spacetime M = ℝ 4 non-zero. This is the case, in particular, if the
value ϕ(x) of the Higgs field is everywhere on spacetime in an appropriate
neighbourhood of the vacuum vector in the vector space W. It follows that after
symmetry breaking we can assume in the electroweak theory that the Higgs field
is in unitary gauge, at least if its fluctuations around the vacuum value are not
too large.

Lemma 8.2.6 If ϕ is in unitary gauge, then

Proof Note that A μ ⋅ w 0 is tangential to the orbit of w 0 under G, while the

shifted Higgs field Δϕ and therefore its derivative is everywhere on M
orthogonal to the orbit, by the assumption of unitary gauge. This implies the
claim. □

Lemma 8.2.7 If ϕ is in unitary gauge, then

where η j are the Higgs bosons.

Proof Follows immediately, because under our assumption that the Nambu–
Goldstone bosons vanish we have

Lemma 8.2.8 Up to second order in the Higgs bosons we have

where is the mass of the j-th Higgs boson.

Proof This formula can be found in Theorem 8.1.15. □

Lemma 8.2.9 We have

where M a are the masses of the broken gauge bosons defined above.

Proof We have

for the bilinear mass form m on . This implies the claim by our choice of basis
α a. □

We now collect all terms and get:

Theorem 8.2.10 (Mass Generation for Gauge Bosons) If the Higgs field ϕ is
in unitary gauge after symmetry breaking , then the Lagrangian , up to
terms of second order in the shifted Higgs field and the gauge field, is given by

Here we removed the irrelevant constant V (w 0). This Lagrangian has the
following interpretation:
The two terms in the first line are the Klein–Gordon Lagrangian for 2n − d
real scalar Higgs bosons η j of mass .
The two terms in the second line are the Lagrangian for d broken, massive
gauge bosons A μ 1, …, A μ d of mass M a .
The term in the third line is the Lagrangian for r − d unbroken, massless
gauge bosons A μ d+1, …, A μ r .
 This is the celebrated Brout–Englert–Higgs mechanism of creating in a
gauge invariant way masses for gauge bosons. The exact Lagrangian
contains terms of order higher than two that describe interactions between Higgs
bosons, between gauge bosons (in the non-abelian case) and between Higgs
bosons and gauge bosons. For the Standard Model these terms can be found in
Theorem 8.7.2, Corollary 8.7.5 and Theorem 8.7.6.

Remark 8.2.11 Notice why we get the mass term

for the broken gauge bosons: in the Yang–Mills–Higgs Lagrangian we have

the term

involving the covariant derivative of the Higgs field, which describes the
coupling between the Higgs field and the gauge field. We have then written

which results in the term

This term is non-zero in general, because the value w 0 of the Higgs

condensate is non-zero, and leads to the mass term for the broken gauge
bosons. We see that the ultimate reason for the masses of the broken gauge
bosons is the coupling of the gauge field to the Higgs field via the covariant
derivative (the Higgs field is a charged scalar) and the non-zero value w 0 of
the Higgs condensate after symmetry breaking. This also explains why the
masses of the gauge bosons depend on both the coupling constants and the
vacuum value | | w 0 | |.

Remark 8.2.12 The mechanism of spontaneous symmetry breaking was

developed by Philip W. Anderson (Nobel Prize in Physics 1977) , Yoichiro
Nambu (Nobel Prize in Physics 2008 ) and Jeffrey Goldstone in the early 1960s.
It was extended to Yang–Mills theory independently by Robert Brout and
François Englert [48], Peter W. Higgs [76], and Gerald S. Guralnik, Carl
R. Hagen and Thomas W.B. Kibble [67] in 1964 (Nobel Prize in Physics 2013
for Englert and Higgs). The three papers on symmetry breaking from 1964 all
appeared in vol. 13 of the Physical Review Letters. The theory was confirmed in
July 2012 by the discovery of the Higgs boson at the Large Hadron Collider
(LHC) at CERN in Geneva.

8.3 Massive Gauge Bosons in the SU(2) × U(1)-Theory

of the Electroweak Interaction
The weak interaction, which describes particle decays such as the decay of a
muon into an electron, an electron antineutrino and a muon neutrino

was originally described by the so-called 4-Fermi interaction with Lagrangian

(G F is a coupling constant) and associated Feynman diagram Fig. 8.2. This

Lagrangian with a direct interaction between four fermions did not appear in
Chap. 7 and, in fact, it is non-renormalizable .

Fig. 8.2 Muon decay with 4-Fermi interaction

In the Standard Model, the weak interaction together with the

electromagnetic interaction are described by a gauge theory with gauge group
SU(2) L × U(1) Y . This theory is called the electroweak theory . The Feynman
diagram of the 4-Fermi interaction is replaced by the Feynman diagram in
Fig. 8.3, involving a virtual W −-gauge boson. The interactions in this diagram
are renormalizable .
Fig. 8.3 Muon decay in electroweak theory

In this section we want to study the Higgs mechanism of mass generation in

the special case of the electroweak interaction and discuss the associated
massive gauge bosons W ± and Z 0.

8.3.1 The Lie Algebra

We continue with Example 8.1.9 and Example 8.1.17. Our gauge group is G =
SU(2) L × U(1) Y and the manifold M is 4-dimensional flat Minkowski spacetime.
We choose the Ad-invariant scalar product on the Lie algebra
in such a way that the following vectors form an orthonormal
basis:

where σ l are the Pauli matrices

and the positive real numbers g w and g′ are the coupling constants corresponding
to SU(2) L and U(1) Y . The non-zero natural number n Y is a normalization
constant.

Remark 8.3.1 We will fix from now on n Y = 3 (this is the convention used, for
example, by [9] and [137]). Other references (such as [125]) use the convention
n Y = 6. We continue to use the traditional convention n Y = 3.
Lemma 8.3.2 The restriction of the scalar product to is given by

where is the Killing form of . In particular, the scalar product isAd-

invariant .

Proof This is Exercise 8.11.3. Compare with Exercises 2.7.13 and 2.7.16. □

The Higgs vector space is . Recall the unitary representation of the gauge
group G on W from Example 8.1.9. The basis elements act on a vector

as

The vacuum vector is given by

where μ and λ are the parameters of the Higgs field potential that can be found in
Example 8.1.9.

8.3.2 The Gauge Bosons

A direct calculation shows that the mass form

is given in the basis β a by

If we define a new orthonormal basis

then the bilinear form m becomes diagonal:

We see that the subalgebra of the stabilizer group is given by

Indeed, α 4 acts on the vacuum vector as

Note that with respect to the standard unnormalized basis

the vector α 4 is just given by

hence proportional to β 3 ′ + β 4 ′ as we expect from Example 8.1.9. We shall see

in Sect. 8.3.5 that the factor

can be identified with the elementary electric charge.

The subspace of broken generators is given by

From the diagonal mass form we can read off the masses of the gauge
bosons:
There are three massive gauge bosons : two gauge bosons, corresponding
to α 1, α 2, of mass
 and one gauge boson, corresponding to α 3, of mass

As mentioned in Remark 8.2.3 we see precisely how the masses of the

gauge bosons depend on the vacuum value | | w 0 | | of the Higgs field and
the coupling constants.
We also have one massless gauge boson , corresponding to α 4.

8.3.3 The Physics Notation

In physics the following notation is used: We set

where is the Weinberg angle or weak mixing angle . Then

Hence the basis (α 3, α 4) is rotated by an angle θ W with respect to (β 3, β 4)

(clockwise in our situation). The Weinberg angle describes the direction of the
unbroken generator α 4 of U(1) Q with respect to the generators β 3 and β 4 of
SU(2) L × U(1) Y .
We can then decompose our gauge field

as

where
are the -bosons of mass

Z μ 0 = cosθ W A μ 3 − sinθ W A μ 4 is the -boson of mass

 γ μ = sinθ W A μ 3 + cosθ W A μ 4 is the massless photon (not to be confused
with a mathematical gamma matrix).
The masses of the W- and Z-bosons are related to the Weinberg angle via

We shall see in Sect. 8.5.5 that W ± have electric charge ± 1 whereas Z 0 and γ
have electric charge 0. In general, gauge bosons are also called vector bosons ,
in particular, the W- and Z-bosons. The gauge field A μ 3 is sometimes denoted by
W μ 0 and the gauge field A μ 4 by B μ .

8.3.4 Experimental Values

We discuss some experimental values (our reference is [106, 108]). The Fermi
constant

which describes the effective strength of the weak interaction, can be determined
from muon decay. Its value is

This gives for

the value

The weak mixing angle and the masses of the W ± and Z bosons are

This implies

and with the Fermi constant

We will see in Sect. 8.3.5 that the electric coupling constant is given by

In particular,

so the weakness of the weak interaction compared to electromagnetism comes

mainly from the large mass of the W- and Z-bosons. We write with
fine-structure constant α given by

The final value that was determined experimentally is the mass of the
Higgs boson:

The various constants are not all independent, i.e. if we know some of the
values, then we can predict others with the theory. Note that the values are not
fully consistent with the formulas above, because there are higher-order
corrections to these formulas coming from the associated quantum field theory.
To be precise we have to indicate, for instance, the energy scale at which the
coupling constants have been measured; see Sect. 9.4.

Remark 8.3.3 The theory of the electroweak interaction , the unification of

electromagnetism and the weak nuclear force as an SU(2) × U(1) gauge theory
together with spontaneous symmetry breaking, was developed during the 1960s
by Sheldon Glashow, Abdus Salam and Steven Weinberg (Nobel Prize in
Physics 1979 ). The discovery of the W- and Z-bosons at the Super Proton
Synchrotron (SPS) at CERN was announced in 1983 (Nobel Prize in Physics
1984 for Carlo Rubbia and Simon Van der Meer). The quantum field theory of
the electroweak interaction was shown to be renormalizable by Gerardus ’t
Hooft and Martinus J.G. Veltman (Nobel Prize in Physics 1999 ).

8.3.5 Charges
We continue to consider the case of the electroweak interaction with gauge
group G = SU(2) L × U(1) Y . Suppose V is a complex vector space of dimension
m with a unitary representation of G. The generators β l of act as
and the generator β 4 of acts as

where and are certain Hermitian operators on V.

Definition 8.3.4 The eigenvalues of are called weak isospin and of weak
hypercharge .

Since and commute in every representation of G, we can find an

orthonormal basis of V of common eigenvectors for both operators. If we
identify V with via this basis, then both and act as diagonal matrices,
whose entries are the charges of the multiplet component fields corresponding to
the basis vectors for V.

Remark 8.3.5 The eigenvalues of the weak isospin operator and thus the
weak isospin charges are determined by the weights of the representation. The
weights are elements in the dual space of the Cartan subalgebra in ,
which is spanned by iσ 3. See, for example, [153] for more details.

If we set

(the third equation is known as the Gell-Mann–Nishijima formula ), then a

general gauge field acts on the multiplets with values in V as

(8.5)

It follows that the elementary electric charge is given by

For example, on the Higgs field

with values in the charge operators act as

We can then directly read off the charges of the components of ϕ. In order to
indicate the electric charge of the Higgs field , it is often written as

(the factor is a convention).

Remark 8.3.6 If the convention n Y = 6 instead of our n Y = 3 is used, then the

Gell-Mann–Nishijima formula becomes

8.4 The SU(3)-Theory of the Strong Interaction (QCD)

Quantum chromodynamics (QCD) , the theory of the strong interaction , is a
gauge theory with gauge group SU(3) C , where C stands for colour (chroma is
the Ancient Greek word for colour). There are eight gauge bosons for this gauge
group, called gluons .
The matter particles (fermions) in QCD are called quarks. We will see later
that there are six different types of quarks, called quark flavours . In nature,
quarks are only observed in bound states called hadrons : baryons (like
nucleons , i.e. the proton and neutron ), that consist of three quarks (or
antiquarks) and have half-integer spin, and mesons , that consist of a quark-
antiquark pair and have integer spin. The lifetime of the heaviest quark, the top
quark , is too short to form hadrons. All other quarks can form hadrons. There is
evidence that exotic hadrons such as tetraquarks , consisting of two quarks and
two antiquarks, and pentaquarks , consisting of four quarks (antiquarks) and one
antiquark (quark), exist. New hadrons and hadron resonances (excited states) are
still being discovered.
 The gluons virtually “glue” the quarks together into hadrons. The exact
reason why quarks appear only in colour neutral hadrons and not as isolated
particles, known as colour confinement , is not fully understood. Basic
properties of these bound states, like their masses, so far cannot be derived
theoretically, but only through experiments or numerical simulations (lattice
QCD) . Hadrons like the proton and neutron are therefore surprisingly complex
objects.
Even though protons and neutrons are colour neutral, there is a residual
strong interaction between them (and other hadrons). For the nucleons this
interaction is known as the strong nuclear force and leads to the formation of
atomic nuclei. As a residual interaction it is comparable to the electromagnetic-
chemical interaction between neutral atoms (atoms can be thought of as
electromagnetic bound states), leading to the formation of molecules.
The quarks (valence quarks ) that constitute a hadron, together with virtual
gluons and sea quarks , which are virtual quark-antiquark pairs produced by
virtual gluons, are collectively known as partons .
The decomposition of a hadron into partons (more precisely, the fraction of
the hadron momentum carried by each type of parton) is complicated and is
described by the parton distribution function (PDF) (or parton density
function) of the hadron: the PDF at a given energy scale Q for a given type a of
parton (quark flavour or gluon) is a probability density function f a (x, Q). The
probability for finding a parton of type a carrying a fraction x a [x, x + dx] of
the longitudinal momentum of the hadron is f a (x, Q)dx (see [42]). The PDF of
the proton is very important for the correct interpretation of scattering
experiments at colliders like the LHC . At present PDFs have to be determined
experimentally and cannot be calculated from first principles, because they
involve non-perturbative QCD [137].
The full gauge group of the Standard Model is therefore

The group SU(3) C acts trivially on the Higgs vector space and thus
leaves the vacuum vector w 0 invariant. It follows that the gluons are unbroken,
massless gauge bosons . The full unbroken gauge group of the Standard Model,
the isotropy group of w 0, is

The full symmetry group of the Standard Model in the sense of Remark 6.12.
7 is therefore the group
or, equivalently,

depending on the signature chosen for Minkowski spacetime.

A very detailed discussion of QCD including theoretical and experimental
aspects, especially concerning perturbative QCD , can be found in the book [42].

8.4.1 Basis for

Recall from Example 2.1.49 that can be described as the vector space of
skew-Hermitian tracefree matrices X, with adjoint representation given by

We can identify complex 3 × 3-matrices X with elements of

The standard Hermitian scalar product on is invariant under the fundamental

representation of SU(3). This implies that

as SU(3)-representations, hence

For the fundamental representation of SU(3) C we choose the basis

(8.6)

and write a general element of as

We call with this basis colour space . A basis of is given by the
corresponding vectors

(8.7)

Hence under the canonical antilinear isomorphism , every colour gets

mapped to its anticolour.
It follows that a complex basis of is given by the following nine
matrices

where

Since is equal to the vector space of skew-Hermitian, tracefree

elements of , we get with the Gell-Mann matrices λ a from Example 1.5.
33:

Proposition 8.4.1 A real basis of is given by the matrices ,

which can be written as
Lemma 8.4.2 The basis vectors ν a are orthonormal with respect to the
positive definite scalar product

where is the Killing form of .

Proof This is Exercise 8.4.2. Compare with Exercises 2.7.16 and 2.7.17. □

It is sometimes convenient to consider the complexification of the Lie algebra

:

Proposition 8.4.3 A complex basis of is given by the tracefree

matrices

These matrices are orthonormal with respect to the Hermitian scalar product
defined by the complexification of .

Definition 8.4.4 The scalar product on is defined by

where g s is the strong coupling constant . In particular, the scalar product is
Ad-invariant . Orthonormal bases for and (with the
complexification of the scalar product) are given by the vectors

and

The choices of basis and scalar product are for QCD less standardized than for
the electroweak interaction. We can expand the gluon gauge field G μ with values
in in either of the bases {ν a }, {μ a }:

(8.8)

8.5 The Particle Content of the Standard Model

8.5.1 Fermions
We want to add fermions, i.e. matter particles, to the Standard Model. In general,
charged fermions , which couple to the gauge fields, are described by twisted
chiral spinors, i.e. sections of twisted chiral spinor bundles
(8.9)
where S L is the left-handed and S R the right-handed Weyl spinor bundle over 4-
dimensional flat Minkowski spacetime M. The bundles F L and F R are
associated vector bundles defined by complex unitary representations V L and V
R of the gauge group G. We now want to describe the representations of G in the
Standard Model , where

The complex vector spaces V L and V R have dimensions

As representations of G they decompose into orthogonal sums

of G-subrepresentations V L i , V R i , i = 1, 2, 3, called the three generations or

families . The generations have dimensions

The left-handed generations V L i for i = 1, 2, 3 are all isomorphic as G-

representations and the same is true for the right-handed generations V R i . Each
generation again decomposes into orthogonal sums

of G-subrepresentations Q L i , Q R i and L L i , L R i , called quark sectors and

lepton sectors . They have dimensions

Again the left-handed (right-handed) quark sectors are all isomorphic and the
same is true for the left-handed (right-handed) lepton sectors across generations.
We denote by the fundamental representation of SU(3) C and by the
fundamental representation of SU(2) L , both with the standard invariant
Hermitian scalar product . For both Lie groups we denote by the trivial 1-
dimensional representation . We also denote by the representation of U(1) Y
where the generator acts as

We note the following:

Lemma 8.5.1 An element α U(1) Y acts on by

This representation is well-defined for all weak hypercharges y which are integer
multiples of . The representation has winding number 3y.

For instance, the representation has winding number 4. Table 8.1 describes
the G-representations of the quark and lepton sectors. As G-representations the
vector spaces V L and V R thus decompose into irreducible representations of
dimensions

Interestingly, we see that the four simplest representations of the Lie

group SU(3) × SU(2)

all appear in the Standard Model.

We now define bases for these representations, see Table 8.2. The basis for
the quark sectors Q L i and Q R i are obtained as the tensor product of the basis
vectors r, g, b from Eq. (8.6) and the basis vectors for the SU(2) L × U(1) Y -
representations. For both quarks and leptons the bases for the SU(2) L × U(1) Y -
representations are defined so that they consist of simultaneous eigenvectors for
both charge operators and . We list these basis elements for the first
generation together with their weak isospin , weak hypercharge and electric
charge (recall that ).

Table 8.1 Fermion sectors

Sector Representation Physics notation Complex dimension

QLi (3,2) 1⁄3 6

QRi (3,1) 4⁄3 (3,1) −2⁄3 6

LLi (1,2) −1 2

LRi (1,1) −2 1

The bold integers denote representations of certain dimensions. The usage of

bold face seems to be standard in the physics literature; see also Definition 2.1.
18
Table 8.2 Fermion representations
Sector SU(2) L × U(1) Y representation Basis vectors Particle Charges

QL1 uL

dL′

QR1 uR 0

dR′ 0

LL1 ν eL −1 0

eL −1 −1

LR1 1 eR 0 −2 −1

Remark 8.5.2 In references that use the convention n Y = 6 instead of our n Y =

3, the value of the hypercharge is one half of the value of our hypercharge Y.

It follows that the left-handed quarks and leptons of each generation form
isospin doublets (isodoublets) , while the right-handed quarks and leptons are
isospin singlets (isosinglets) . The quarks are at the same time colour triplets ,
while the leptons are colour singlets.
In the fourth column in Table 8.2 we state the corresponding names for the
particles of specific weak isospin. We write
where ψ L , ψ R are arbitrary maps to the Weyl spinor spaces Δ L , Δ R and

is a general element of the colour space. Here

is read as a “red left-handed down quark” and

is read as a “green right-handed up quark”. Similarly

are the leptons. A general map to

can then be written as

and a general map to

as

For the second generation we make the following replacements:

(8.10)

For the third generation we make the following replacements:

(8.11)

The representations and charges stay the same. Table 8.3 lists the names of
these particles. The different types u, d, c, s, t, b of quarks are called quark
flavours and the different types e, ν e , μ, ν μ , τ, ν τ are called lepton flavours .
The prime ′ on the down-type quarks will be explained in Sect. 8.8.2. The
electron, muon and tau are collectively known as the charged leptons . Note
that in the Standard Model there are no right-handed neutrinos .
The remarkable exact repetition (except for the masses) of the first
generation in two more generations cannot be explained in the Standard
Model. In some theories beyond the Standard Model a right-handed neutrino
singlet is added to each generation, making the leptons very similar to the
quarks concerning the structure of SU(2) representations (the weak
hypercharges and thus the electric charges are different); see Sect. 9.2.1.

Table 8.3 Names of the fermions

Type First generation Second generation Third generation

Quark u Up c Charm t Top (Truth)
d′ Down s′ Strange b′ Bottom (Beauty)
Lepton e Electron μ Muon τ Tau
ν e Electron neutrino ν μ Muon neutrino ν τ Tau neutrino

Remark 8.5.3 The quark model was developed by Murray Gell-Mann (Nobel
Prize in Physics 1969 ) and independently by George Zweig in 1964, originally
containing only the up, down and strange flavour.
The Yang–Mills SU(3) gauge theory of the strong interaction, containing
coloured quarks and a colour octet of gluons, was proposed in 1973 by Harald
Fritzsch, Murray Gell-Mann and Heinrich Leutwyler. The bottom and top quark
were postulated by Makoto Kobayashi and Toshihide Maskawa in 1972 (Nobel
Prize in Physics 2008 ). The last quark in the three generations, the top quark,
was experimentally observed for the first time in 1995 at the Collider Detector at
Fermilab (CDF) .

Remark 8.5.4 The fact that the SU(2) L × U(1) Y -representations for left-handed
and right-handed fermions are different implies that they interact differently with
the W- and Z-bosons and thus the weak interaction is not invariant under
inversion of parity (handedness) . This was first predicted theoretically in 1956
by Tsung-Dao Lee and Chen Ning Yang (Nobel Prize in Physics 1957 ) and
verified experimentally by Chien-Shiung Wu in 1957.

8.5.2 Antiparticles
Every fermion has an antiparticle . Antiparticles are sections of the complex
conjugate bundle

We note the following useful fact:

Lemma 8.5.5 There are complex linear isomorphisms of Lorentz spin

representations

where the first isomorphism is given by the Majorana form (⋅ , ⋅ ) and the second
isomorphism by the Dirac form ⋅ , ⋅ .

In fact, these isomorphisms are given by the map τ from Lemma 6.7.17

where

for the unitary matrices C and A from Sect. 6.8 and we used the notation for the
charge conjugate from Eq. (6.3).
We can compare this with the constructions in Sect. 2.1.3, where we
essentially defined Δ R as . We now understand that this isomorphism comes
from the Dirac form ⋅ , ⋅ . The second isomorphisms in Lemma 8.5.5 are given
by the matrix ε, corresponding to the matrix C defining the Majorana form (⋅ , ⋅ ).
If we set

and extend this notation to the representations Q L , Q R , L L , L R and the

bundles F L , F R , we get from Lemma 8.5.5 that

(8.12)
Then each generation of antiparticles is described by the representations in
Table 8.4. Under the complex antilinear isomorphisms

we map
and similarly for the second and third generation. It is clear that charge
conjugation ψ ↦ ψ C = B −1 ψ is an involution,

since the matrix B defines a real structure on the spinor space.

Table 8.4 Antifermion sectors
Sector Representation Physics notation Complex dimension
Q R iC 6

Q L iC 6

L R iC 2

L L iC (1,1) 2 1

The bold integers denote representations of certain dimensions. The usage of

bold face seems to be standard in the physics literature; see also Definition 2.1.
18

We then get the SU(2) L × U(1) Y -representations in Table 8.5. There are
corresponding representations for the second and third generation. The antiquark

for example, is read as “antiblue left-handed top antiquark” . The antiparticle

of the electron is called a positron . All other antiparticles are named with the
prefix “anti”.
Table 8.5 Antifermion representations
Sector SU(2) L × U(1) Y representation Basis vectors Particle Charges
Q R 1C uRC

d R ′C

Q L 1C uLC 0

d L ′C 0

L R 1C ν eR C 1 0

eRC 1 1

L L 1C 1 eLC 0 2 1

Table 8.6 Left-handed and right-handed particles and antiparticles

Left-handed fermions and antifermions

Right-handed fermions and antifermions

8.5.3 Chirality of the Standard Model

It is sometimes useful to separate fermions not into particles and antiparticles,
but into left-handed and right-handed particles and antiparticles as in Table 8.6.
Each generation of left-handed particles and antiparticles is described by the
representation

and each generation of right-handed particles and antiparticles is described

by the representation
These representations of G = SU(3) C × SU(2) L × U(1) Y both have dimension
15, but are not isomorphic as complex representations. This follows from

even though

(see Exercise 2.7.3 and the remark following it). Hence we can say that the
Standard Model is a chiral gauge theory in the following sense:

Definition 8.5.6 A gauge theory with fermions and gauge group G is called
chiral if the G-representation for the right-handed particles and antiparticles is
not complex linearly isomorphic to the G-representation for the left-handed
particles and antiparticles.

Note that in any case

This implies that a gauge theory is chiral if the representation V L i V L iC for

the left-handed particles and antiparticles is not isomorphic to its complex
conjugate representation. It follows that every gauge theory that aims at
describing realistic physics has to have a gauge group that admits a complex
representation not isomorphic to its complex conjugate. Such representations are
sometimes called complex (in a different sense!). This is an important restriction
on the possible gauge groups of Grand Unified Theories (GUTs) .
The complete left-handed representations for fermions and antifermions
of one generation is (in the physics notation)
(8.13)
Every realistic theory beyond the Standard Model should recover this
representation. The complete right-handed representation is the complex
conjugate of the left-handed representation.

8.5.4 Higgs Field

We saw above that the Higgs bundle is the vector bundle
(8.14)
associated to the principal bundle P via a unitary representation on W. Here
denotes the trivial complex line bundle associated to the trivial (scalar)
representation of the Lorentz spin group. The representation W is given by
Table 8.7. A basis with corresponding charges can be found in Table 8.8. The
vector f corresponding to the Higgs boson is an element of given by

The Higgs boson therefore also has

Table 8.7 Higgs sector

Higgs vector space Representation Physics notation Complex dimension

W (1,2) 1 2

The bold integers denote representations of certain dimensions. The usage of

bold face seems to be standard in the physics literature; see also Definition 2.1.
18
Table 8.8 Higgs field representation

Sector SU(2) L × U(1) Y representation Basis vectors Particle Charges

W ϕ+ 1 1

ϕ0 1 0

8.5.5 Gauge Fields

We finally want to summarize the representations of the gauge fields in the
Standard Model. The gauge theory of the Standard Model is defined by a
(trivial) principal G-bundle P over 4-dimensional Minkowski spacetime for the
Lie group
A connection A on the principal bundle P is a 1-form with values in the Lie
algebra

We decompose A accordingly into gauge fields

which we call the gluon gauge field , the weak gauge field and the
hypercharge gauge field (G is the standard notation for the gluon gauge field,
not to be confused with the Lie group G).
We know from Sect. 5.13 that the difference ΔA of an arbitrary gauge field A
minus a fixed reference gauge field A 0 can be thought of as a 1-form on
spacetime M with values in the vector bundle associated to P via the
adjoint representation of G on the Lie algebra . Hence ΔA is a section of the
twisted vector bundle

We write

The adjoint representation of the Standard Model gauge group G on the Lie
algebra splits into three orthogonal subrepresentations

called the gluon sector , the weak sector and the hypercharge sector . Hence
the bundle decomposes into a direct sum of twisted bundles

(8.15)
Let denote the adjoint representation of SU(3) C , the adjoint
representation of SU(2) L and the trivial representation . Then the
corresponding G-representations defining the adjoint bundles are given by
Table 8.9. In the electroweak sector , it can be shown that the
orthonormal basis

that we have chosen in Sect. 8.3.3 consists of simultaneous eigenvectors for

the charge operators and (here we use the complexified adjoint
representation). The charges are summarized in Table 8.10 (see Exercise 8.11.6).
Note that the W-bosons have both a non-zero weak isospin and electric charge.
The basis vectors

are orthonormal with respect to the Hermitian scalar product on

defined by the complexification of the positive definite real scalar product on

, for which α 1, α 2, α 3, α 4 form an orthonormal basis.

Table 8.9 Gauge sectors

Gauge sector Representation Physics notation Real dimension

(8,1) 0 8

(1,3) 0 3

(1,1) 0 1

The bold integers denote representations of certain dimensions. The usage of

bold face seems to be standard in the physics literature; see also Definition 2.1.
18
Table 8.10 Representation of electroweak gauge bosons

Gauge sector SU(2) L × U(1) Y representation Basis vectors Boson Charges

W+ 1 0 1

W− −1 0 −1

α3 Z0 0 0 0
α4 γ 0 0 0

8.5.6 The Total Particle Content of the Standard Model

If we like, we could now define the total particle content of the Standard Model
as sections of the direct sum of the bundles in Eqs. (8.9), (8.12), (8.14)
and (8.15) (with complexified adjoint representation).

8.5.7 Hypercharges: Constraints from Group Theory

The specific assignments of the values of the weak hypercharge Y, which
determine the representations of the group U(1) Y in the Standard Model, are not
arbitrary, but have a certain pattern that we explain in this subsection and the
following. These assignments have important consequences. For example, it is
well-known that the sum of the electric charge of the proton and the electron is
zero: the proton consists of two up valence quarks and one down valence quark
and has electric charge

while the electron has electric charge − 1. This equality, fundamental to the
existence of neutral atoms, holds even though in the Standard Model the electric
charges for the quarks are a priori independent of the electric charges for the
leptons.
We begin by describing purely group theoretic constraints on the
hypercharges in this section and quantum constraints in the following.
Remarkably, it turns out that the Standard Model can only define a consistent
quantum theory if, in particular, the sum of the electric charge of the proton and
the electron is zero.

A -Subgroup of the Standard Model Group

We first note the following (see [9]):

Theorem 8.5.7 ( -Subgroup of the Standard Model Group) The subgroup

of elements of the form

acts trivially on the representations V L , V R and V L C , V R C . Here I 2 and I 3

denote the unit matrices.
Conversely, suppose the subgroup of the Standard Model group is
given and each of the representations

is invariant under K. Then in each case the hypercharge y is related to the

hypercharge y SM in the Standard Model by
Proof This is Exercise 8.11.7. □

The first part of this theorem implies:

Corollary 8.5.8 The representations V L , V R and V L C , V R C of the Standard

Model group G descend to representations of

As we saw in Exercise 1.9.12, there is a natural embedding

Corollary 8.5.8 is one of the reasons why a SU(5) theory of Grand Unification
(GUT) is possible (for more details, see Sect. 9.5).

Charge Quantization
The Lie algebra of the Standard Model is

Concerning charges in general, we need to distinguish between semisimple Lie

algebras and abelian Lie algebras. For example, using representation theory it is
possible to show that in any representation of the Lie algebra the weak
isospin must have values which are integer multiples of . The possible charges
are thus quantized . To make this plausible, note that the commutation relation

implies that a non-trivial representation

does not yield a representation after a rescaling λ ⋅ ϕ, with . Similarly

the charges for any semisimple Lie algebra are quantized (the charges are related
to the discrete weight lattice ).
On the other hand, if we consider the abelian Lie algebra , then
representations

yield representations after arbitrary rescalings λ ⋅ ϕ with . Hence all

values of weak hypercharge (even irrational ones) are possible and the charges
are not quantized. They are only quantized if the representations of come
from representations of the compact circle U(1) Y (for a suitable, fixed
circumference).
This shows that on the level of Lie algebras, there is no reason why the
values of the weak hypercharge and electric charge for all particles in the
Standard Model should be quantized and, in particular, be multiples of . The
quantization, however, would follow naturally for algebraic reasons if the Lie
algebra of the Standard Model is a Lie subalgebra of some larger (Grand
Unified) compact simple (or semisimple) Lie algebra .

8.5.8 Hypercharges: Constraints from Vanishing of

Anomalies
If we take into account the quantum field theory defined by the Standard Model,
there are additional constraints that restrict the assignments of weak
hypercharges. These constraints result from demanding that the Standard Model
is free of gauge anomalies , i.e. that all gauge symmetries of the classical theory
still persist in the quantum theory. This means that the derivative (4-divergence)
of certain Green’s functions (correlators ) has to vanish.

Fig. 8.4 Chiral anomaly

It can be shown that in 4-dimensional spacetime the only possible non-zero

contribution to the 4-divergence of these Green’s functions comes from triangle
Feynman diagrams of the form in Fig. 8.4, with three external gauge bosons and
one fermion loop (Feynman diagrams involving fermions and gauge bosons will
be explained in more detail in Sect. 8.6. The appearance of loop diagrams
indicates that the anomalies are indeed a quantum effect.) We denote the various
hypercharges in the Standard Model (for one generation) as follows:
For the Standard Model with gauge group

the anomalies depend on which of the factors of G the three gauge bosons in
the triangle diagram belong to. We denote the anomalies accordingly by U(1) Y 3,
SU(3) C 2U(1) Y , etc.
Setting the gauge anomaly, also called the chiral anomaly , calculated from
each of the triangle diagrams to zero, leads to constraints on the hypercharges
that are summarized in Table 8.11 (the table is from [125, Sect. 30.4]; grav
denotes a graviton).1 All other anomalies, like the ones associated to SU(3) C 3 or
SU(3) C U(1) Y 2, vanish automatically.

Table 8.11 Constraints on hypercharges from anomaly cancellations

Anomaly Constraint
U(1) Y 3 (2Y L 3 − Y e 3 − Y ν 3) + 3(2Y Q 3 − Y u 3 − Y d 3) = 0

SU(3) C 2U(1) Y 2Y Q − Y u − Y d = 0

SU(2) L 2U(1) Y Y L + 3Y Q = 0

grav2U(1) Y (2Y L − Y e − Y ν ) + 3(2Y Q − Y u − Y d ) = 0

Without trying to explain the calculation of these constraints in detail, the

following is apparent from the table:

1. The assignments of hypercharges in the Standard Model

satisfy all constraints. Hence the Standard Model is free of gauge

 anomalies.

2. The first and third constraint are only satisfied if contributions from both
quarks and leptons are taken together, i.e. these contributions to the triangle
diagrams have to cancel each other.

3. Up to an overall factor the equations on the hypercharges strongly constrain

their possible values. In particular, according to an argument in [125,
Sect. 30.4], if Y ν = 0, then up to an overall factor the hypercharges must
have precisely the values in the Standard Model (or satisfy Y Q = Y L = Y e =
0, which does not occur in nature).

It follows that in a realistic situation, the assignments of hypercharges in

the Standard Model are completely fixed (up to an overall factor) by the
vanishing of gauge anomalies. In particular, vanishing of gauge anomalies
implies that the hypercharge (and thus the electric charge) are quantized and
that the proton charge plus the electron charge is zero. As mentioned above,
in Grand Unified Theories, electric charge can be quantized automatically for
purely group theoretic reasons, without invoking vanishing of gauge
anomalies.

8.6 Interactions Between Fermions and Gauge Bosons

The interaction between fermions and gauge bosons comes from the following
(massless) Dirac Lagrangian 2

with Dirac operator given by

in a global Lorentz frame for flat 4-dimensional Minkowski spacetime M. The

cubic term responsible for the interaction is the interaction vertex

where we have chosen a global trivialization ε of Spin+(M) corresponding to the

Lorentz frame. The scalar products are taken in Δ V L and Δ V R . Recall
that the interaction vertex is automatically real.
 As in Sect. 8.5 we can decompose the gauge field

corresponding to the summands and the orthonormal bases for the Lie
algebra

and we can decompose the maps

in the bases for the quark and lepton sectors .

The interaction vertex decomposes into the electroweak and strong
interaction vertex :

where involves only the gauge field W + B and only the gauge
field G.

8.6.1 The Electroweak Interaction Vertex

We first discuss the electroweak weak interaction vertex.

Lemma 8.6.1 In the representation ofSU(2) L × U(1) Y the gauge field

W + B acts as

In the representation ofSU(2) L × U(1) Y the gauge field W + B acts as

Here we have set e = g w sinθ W for the elementary electric charge.

Proof This is Exercise 8.11.8. □

Using this lemma we get the following explicit formula (with Dirac conjugate
according to Sect. 6.8):
Theorem 8.6.2 (Electroweak Interaction Vertex) The electroweak
interaction vertex for the leptons and quarks is given by
(8.16)
where for the leptons we have

(8.17)

and for the quarks we have (with the standard Hermitian scalar product over
the components in colour space implicit)

(8.18)

For the second and third generation we make the replacements in Eqs. (8.10)
and (8.11). We have set e = g W sinθ W for the elementary electric charge.

Proof This is Exercise 8.11.9. □

These Lagrangians are sometimes written as (e.g. [62])

with the currents (for the first generation)

and

The expressions for j W, l μ and j W, q μ hold for both commuting and

anticommuting spinors . For anticommuting spinors we calculate

where we used Γ μ† = Γ 0 Γ μ Γ 0. Similarly for j W, q μ .

Definition 8.6.3 The interactions between fermions involving W-bosons are

called charged current interactions and interactions involving Z-bosons
neutral current interactions .

Remark 8.6.4 Note that the W-bosons pair different flavours of particles
(neutrinos with electrons, up quarks with down quarks) with different electric
charges and different weak isospin, because they act off-diagonally on .
The sum of charges (weak isospin , weak hypercharge , electric charge ) at
each vertex is conserved. Interactions involving W-bosons are the only
vertices in the Standard Model that change flavour. The Z-boson and the
photon γ on the other hand act diagonally and only pair particles of the same
flavour (hence with the same charges). The charged W-bosons also couple
only to left-handed fermions, while the Z-boson and the photon γ couple to
both left-handed and right-handed fermions. Contrary to the Z-boson, the
photon γ pairs both left-handed and right-handed fermions in exactly the same
way, i.e. electromagnetism is invariant under parity inversion . The photon
does not couple to the neutrino , because its electric charge is zero.

We summarize the electroweak interaction vertices in the corresponding

Feynman diagrams in Figs. 8.5, 8.6, 8.7, 8.8 (note that, for instance, in the left
diagram in Fig. 8.6, ν e L only pairs with ν e L , e L with e L , and e R with e R , etc.).
In the associated quantum field theory these diagrams also describe the
interactions involving antifermions . Each diagram can be interpreted as the
following interactions between (possibly virtual) particles :
A fermion or antifermion radiates off (or emits) a gauge boson (possibly
changing flavour).
A fermion or antifermion absorbs a gauge boson (possibly changing
flavour).
A fermion and antifermion annihilate in a gauge boson.
A gauge boson decays into a fermion and antifermion.

Fig. 8.5 Interaction vertex: leptons and W ±

Fig. 8.6 Interaction vertex: leptons and Z 0, γ

Fig. 8.7 Interaction vertex: quarks and W ±

Fig. 8.8 Interaction vertex: quarks and Z 0, γ

The full Feynman diagram of a process is a combination of such diagrams.

An example is the Feynman diagram in Fig. 8.9 that explains β-decay of a
neutron (consisting of two down and one up quark) into a proton (consisting of
two up and one down quark), an electron and an electron antineutrino, via a
virtual W −-boson:
Fig. 8.9 β-decay of neutron

Neutrinos are produced by similar reactions in enormous amounts in the

process of nucleosynthesis inside stars like the sun (solar neutrinos ).

Remark 8.6.5 The interaction vertices in Theorem 8.6.2 are exactly the same
for all three generations of leptons and quarks. For example, on a fundamental
level the only difference in the Standard Model between the electron, muon and
tau are their different masses (which lead to different lifetimes, etc.). This is
known as lepton flavour universality . For quarks, the corresponding statement
is true for the weak eigenstates considered above, but not for the so-called mass
eigenstates because of quark mixing , to be discussed in Sect. 8.8.2.

8.6.2 The Strong Interaction Vertex

We briefly want to state a formula for the strong interaction vertex .
Fermions appear in two representations of SU(3) C : the trivial representation
and the fundamental representation .
If we expand the gluon field as in Eq. (8.8),

we get:

Lemma 8.6.6 In the representation ofSU(3) C the gluon field G acts as

where the matrices λ a act in the standard way as endomorphisms of by
multiplication from the left. In the representation ofSU(3) C the gluon field G
acts as

It follows that the strong interaction is restricted to quarks and antiquarks and
does not affect leptons . Let q L f denote the left-handed quarks and q R f the right-
handed quarks for flavours f = u, d′, c, s′, t, b′. We can think of q L f as a map on
spacetime with values in and similarly q R f as a map with values in
, where is the colour space spanned by the colour vectors r, g, b. We
write

where q L fr , q L fg , q L fb are ordinary left-handed Weyl spinors, i.e. maps with

values in Δ L , corresponding to the three different colours (analogously for the
right-handed quarks). We then have:

Theorem 8.6.7 (Strong Interaction Vertex) The strong interaction vertex for
the quarks is given by

(8.19)

There are implicit standard Hermitian scalar products in the colour space
.

Note that the strong interaction does not pair quarks of different flavours like
 the weak interaction . If we expand the gluon field alternatively in the basis
{μ a },

we see that the gluons

pair quarks of different colours, because they act off-diagonally on the colour
space, while

act diagonally and pair quarks of the same colour. The first type of gluons can
thus be roughly compared to the W-bosons in the electroweak theory, while
the second type of gluons corresponds to the Z-boson and photon γ.

See the Feynman diagrams in Figs. 8.10 and 8.11 for a generic strong
interaction vertex and two examples of more specific ones. These diagrams can
be interpreted as in Remark 8.6.4 (radiation/emission , absorption , annihilation ,
decay ). The sum of colour charges at each vertex is conserved if the colour
charge of gluons is defined suitably. Gluons can mediate interactions between
quarks of different flavours like in the Feynman diagram in Fig. 8.12.

Fig. 8.10 Generic quark-gluon interaction vertex

Fig. 8.11 Specific quark-gluon interaction vertices

Fig. 8.12 Strong interaction between quarks of different flavours

8.6.3 The Dirac Lagrangian for Fermions

The complete Dirac Lagrangian for the fermions can now be written as
(8.20)
where the kinetic term is

(8.21)

and
with the electroweak interaction vertex from Eq. (8.16) and the strong
interaction vertex from Eq. (8.19).

8.7 Interactions Between Higgs Bosons and Gauge

Bosons
We want to determine the Yang–Mills–Higgs Lagrangian

for the electroweak interaction.

8.7.1 The Higgs Lagrangian

We first calculate the Higgs Lagrangian (the first two summands in the Yang–
Mills–Higgs Lagrangian). We assume that we have chosen a unitary gauge so
that the Higgs field is of the form

where is the Higgs boson and . The Higgs potential is

given by

and the mass of the Higgs boson is . Let A = W + B be the electroweak

gauge field. The mass of the W-bosons is equal to

and the mass of the Z-boson is equal to

Lemma 8.7.1 The covariant derivative Aϕ is given by

μ
The potential V (ϕ) is (up to an irrelevant constant)

Proof This is Exercise 8.11.11. □

We get:

Theorem 8.7.2 (Electroweak Higgs Lagrangian) After symmetry breaking

the Higgs Lagrangian is given in unitary gauge by

(8.22)

Proof This is Exercise 8.11.12. □

These terms have the following interpretation:

The term in the first line is the Lagrangian for a free real scalar Higgs boson
of mass m H .
The term in the second line describes the interaction between Higgs bosons.
The term in the third line is the mass term for the W ±-bosons.
The term in the fourth line describes the residual interaction between the W
±-bosons and the Higgs boson.

The term in the fifth line is the mass term for the Z 0-boson.
The term in the sixth line describes the residual interaction between the Z 0-
boson and the Higgs boson.
Fig. 8.13 Interaction vertices: Higgs boson

Fig. 8.14 Interaction vertices: electroweak gauge bosons and Higgs boson

The free terms in this Lagrangian have already been determined in

Theorem 8.2.10. We can also write

using the definition

from Sect. 8.3.3. The couplings of the W- and Z-bosons to the Higgs boson
are proportional to their masses squared, hence quite strong. Note that the photon
does not couple to the Higgs boson. Figures 8.13 and 8.14 show Feynman
diagrams for the interactions in Theorem 8.7.2. The diagram on the left in
Fig. 8.14, for example, can be interpreted in one of the following ways:
Higgs–Strahlung (vector bosons radiate off or emit a Higgs boson): W ± →
W ± H and Z 0 → Z 0 H
Absorption of a Higgs boson : W ± H → W ± and Z 0 H → Z 0
 Vector boson fusion : W ± W → H and Z 0 Z 0 → H
Vector boson decay : H → W ± W and H → Z 0 Z 0.
The electric charge at each vertex is conserved.

8.7.2 The Yang–Mills Lagrangian

We now calculate the Yang–Mills Lagrangian . We first determine the
commutators of the basis vectors α 1, α 2, α 3, α 4 for in Sect. 8.3.2.

Lemma 8.7.3 In the complexification of the Lie algebra we have

Proof This is the first part of Exercise 8.11.13. □

This implies for the curvature of the electroweak gauge field W + B:

Theorem 8.7.4 The curvature of the electroweak gauge field

is given by
Proof This is the second part of Exercise 8.11.13. □

Let H be a 2-form on Minkowski spacetime M with values in the complex

numbers. We write

Then we get:

Corollary 8.7.5 (Electroweak Yang–Mills Lagrangian) The Yang–Mills

Lagrangian for the electroweak gauge field W + B is given by

(8.23)

More explicitly we can write

where (compare with [112])

The Yang–Mills Lagrangian contains in addition to the quadratic kinetic terms
various cubic and quartic couplings between the gauge bosons. The
corresponding Feynman diagrams are depicted in Figs. 8.15 and 8.16 (electric
charge is conserved at each vertex). We can do a similar calculation for the gluon
gauge field G. We expand

and define structure constants for the Lie algebra in the Gell-Mann basis by

Then the curvature of the gauge field G is

Fig. 8.15 Electroweak gauge bosons: 3-boson interaction vertex

Fig. 8.16 Electroweak gauge bosons: 4-boson interaction vertices

Theorem 8.7.6 (Gluon Yang–Mills Lagrangian) The Yang–Mills

Lagrangian for the gluon gauge field G is given by

(8.24)

More explicitly we get (as in Eq. (7.1)):

where a sum over Lie algebra indices is implicit. The Feynman diagrams are in
Fig. 8.17. The total Yang–Mills Lagrangian for both the electroweak and the
gluon field is then
 (8.25)

Fig. 8.17 3- and 4-gluon interaction vertices

8.8 Mass Generation for Fermions in the Standard Model

So far the fermions (leptons and quarks) in the Standard Model were assumed
massless. In this section we define Yukawa couplings between the fermions and
the Higgs field which lead to Dirac mass terms for the fermions (except for the
neutrinos) after symmetry breaking . The masses of the fermions are
proportional to the Yukawa coupling constants and the absolute value v of the
Higgs condensate.
The Yukawa couplings also lead to a residual interaction vertex between two
fermions of the same flavour and the Higgs boson. This means that fermions can
interact by emitting and absorbing virtual Higgs bosons, hence the Higgs boson
can be interpreted as the mediating particle of a new type of interaction, different
from gauge interactions (in Sect. 8.7.1 we saw that the Higgs boson can also
mediate an interaction between the weak gauge bosons).

8.8.1 Yukawa Couplings for Leptons

We begin with the leptons. Recall that as complex vector spaces the lepton and
Higgs sector are equal to

for each generation. We set i = e, μ, τ to denote the lepton generations and fix
Yukawa couplings g i .
Lemma 8.8.1 For each generation i = e, μ, τ the map

is anSU(2) L × U(1) Y -invariant Yukawa form .

Proof This is the first part of Exercise 8.11.14. □

In physics the following notation is used: We write for the first generation

and similarly for generation 2 and 3 with (ν e , e) replaced by (ν μ , μ) and (ν τ , τ),

respectively. Then

Suppose that the Higgs field ϕ is in unitary gauge,

We define the lepton masses

for i = e, μ, τ. Then we get:

Theorem 8.8.2 (Yukawa Coupling for Leptons) After symmetry breaking the
Yukawa Lagrangian for the three lepton generations associated to the Yukawa
form in Lemma 8.8.1 is given in unitary gauge by

(8.26)

Proof This is the second part of Exercise 8.11.14. □

The three terms in the second line are the Dirac mass terms for the electron ,
muon and tau . The terms in the third line are residual interactions between these
leptons and the Higgs boson (see the Feynman diagram in Fig. 8.18). The
coupling of the leptons to the Higgs boson is proportional to their mass. Note
that the neutrinos do not appear in this Lagrangian. In particular, their mass is
zero in the Standard Model.

Fig. 8.18 Lepton-Higgs interaction vertex

8.8.2 Yukawa Couplings for Quarks and Quark Mixing

Across Generations
We consider the following SU(2) L × U(1) Y representation spaces:

(8.27)
and

(8.28)
The quark representation sectors Q L and Q R are obtained from these
representations by tensoring with the fundamental representation of SU(3) C .
It turns out that the Yukawa couplings for the quarks are only diagonal in
flavour space in another basis than the one we used so far. We write

for the quarks that correspond to the standard basis elements for the
irreducible summands in Eqs. (8.27) and (8.28) and indicate the left-handed basis
by an index L and the right-handed basis by an index R. We also write for the
left-handed isodoublets
Let

be the Higgs field. We set

In the physics literature, ϕ c is often denoted by .

Lemma 8.8.3 The field ϕ c satisfies

Hence ϕ c has the same weak isospin as ϕ and weak hypercharge Y = −1.

Proof This is Exercise 8.11.15. □

Lemma 8.8.4 For arbitrary complex matrices Y u and Y d the expression

is anSU(3) C × SU(2) L × U(1) Y -invariant Yukawa form (the second line is an

abbreviation). The Hermitian scalar product over the components in colour
space in the terms on the right is implicit.

Proof This is the first part of Exercise 8.11.16. □

We can find pairs of unitary matrices V L u , V R u and V L d , V R d that diagonalize

the matrices Y u and Y d :
where all entries of the diagonal matrices are real and positive. We define new
quarks

Since we transformed up-type and down-type quarks with the same matrices V L
u and V u , these quarks define a new splitting of the representation spaces into
R
direct summands as in Eqs. (8.27) and (8.28). The quarks u, d′, c, s′, t, b′ can be
identified with the quarks we considered before in Sect. 8.5.
We also define new quarks

Definition 8.8.5 We call the basis vectors corresponding to the quarks

u, d, c, s, t, b mass eigenstates and the basis corresponding to u, d′, c, s′, t, b′
weak eigenstates or current eigenstates (for reasons that become apparent
below). We also define the quark masses

for each flavour i = u, d, c, s, t, b.

We then get:

Theorem 8.8.6 (Yukawa Coupling for Quarks) After symmetry breaking the
Yukawa Lagrangian for the three quark generations associated to the Yukawa
form in Lemma 8.8.4 is given in unitary gauge and in the mass eigenstate basis
(u, d, c, s, t, b) by

(8.29)

Here the Hermitian scalar products over the colour space are implicit.

Proof This is the second part of Exercise 8.11.16. □

The six terms in the first and second line are the Dirac mass terms for the quarks.
The terms in the third and fourth line are residual interactions between the
quarks and the Higgs boson (see the Feynman diagram in Fig. 8.19). The
couplings of the quarks to the Higgs boson is again proportional to their mass.
The top quark is the heaviest fermion and thus has the strongest coupling to the
Higgs boson.

Fig. 8.19 Quark-Higgs interaction vertex

Remark 8.8.7 The process that leads to the strongest production of Higgs
bosons at the hadron collider LHC is the so-called gluon fusion process [60,
147] with a virtual top quark loop depicted in Fig. 8.20. There are corresponding
processes with other quark flavours, which are, however, much weaker, because
the Higgs boson couples most strongly to the top quark.
Fig. 8.20 Gluon-fusion production of Higgs boson

Definition 8.8.8 The matrix

is called the Cabibbo–Kobayashi–Maskawa (CKM) matrix . The CKM

matrix describes the physical effects of left-handed quark mixing across
generations from the mass eigenstate basis to the weak eigenstate basis.

Since the matrices V L, R d V L, R u† are unitary, we can write the Dirac Lagrangian
for the strong interaction of quarks

either with the weak eigenstates basis u, d′, c, s′, t, b′ or the mass eigenstate basis
u, d, c, s, t, b (the sums in both cases are identical). The terms in the weak
interaction vertex , however, only have the form in Theorem 8.6.2 if they
are written in the weak eigenstate basis, otherwise up-type and down-type quarks
from different generations are paired with the W-bosons.
To see this explicitly consider the following charged current term in Eq.
(8.18):

(8.30)
with the quark current for the α-th generation

Using the mass eigenstate basis we can write the current equivalently as
 (8.31)
It follows that the interactions with the W-bosons can connect quarks from
different generations if the CKM matrix is not diagonal.

Fig. 8.21 Kaon decay via up quark

The Feynman diagram in Fig. 8.21 depicts a typical experimentally observed

process where W-bosons pair quarks of different generations: the decay of a
neutral kaon (consisting of a down quark and a strange antiquark) into a muon
and antimuon:

The W +-boson pairs the up quark from the first generation and the strange
quark from the second generation. This process was observed before the
existence of a fourth quark (the charm quark) was known. The decay rate of K 0
into μ and μ C could not be explained with this process alone. This led
S.L. Glashow, J. Iliopoulos and L. Maiani in 1970 to the postulation [64] of the
charm quark with the process in Fig. 8.22: here the W −-boson pairs the down
quark from the first generation with the charm quark from the second generation.
These diagrams and much more details can be found in Thomson’s book [137].
Fig. 8.22 Kaon decay via charm quark

Summarizing we see that the charged current vertices with W-bosons are
the only vertices in the Standard Model that

connect different flavours of quarks and leptons (flavour changing

vertex)

can connect different generations of quarks (generation changing

vertex) .

Remark 8.8.9 Similarly in general gauge theories the most interesting

gauge bosons are those that act off-diagonally in the fermion representations,
because they can connect different types of particles (in Grand Unified
Theories this leads, for example, to the prediction of proton decay ; see Sect.
9.5 ).

8.8.3 Experimental Values for the CKM Matrix and

Fermion Masses
We discuss the experimental values for the CKM matrix (following [100] and
[110]).
The CKM quark-mixing matrix V CKM U(3) has a priori nine real
parameters, because this is the dimension of the Lie group U(3). We can multiply
all quark flavours d, s, b and d′, s′, b′ by an arbitrary phase in U(1) without
changing the physics. Any collection of five out of these six real parameters
change the matrix V CKM by multiplying rows and columns with a phase, so that
4 real parameters remain. These parameters are three mixing angles
and the KM phase δ [0, 2π). The KM phase is also called
the CP violating phase for reasons that will become clear in Sect. 9.3.
It is possible to show that up to rephasings we can always write

where

The matrix V CKM is also often written in the following form:

Here the entry V ud , for instance, connects in Eq. (8.31) the down quark to
the up quark. The absolute values of the entries of this matrix determined by
experiments are approximately [110]

(8.32)

The current experimental values for the masses of the quarks and leptons
(excluding neutrinos) in absolute value and relative to the mass of the lightest
fermion (the electron) can be found in Table 8.12 (see [106]).
Table 8.12 Fermion masses
First generation Second generation Third generation
m u = 2. 15 ± 0. 15 MeV ≈ 4. 2m m c = 1. 28 ± 0. 025 GeV ≈ 2500m e m t ≈ 173 GeV ≈ 340000m e
e
m d = 4. 70 ± 0. 20 MeV ≈ 9. 2m m s = 93. 5 ± 2 MeV ≈ 180m e m b = 4. 18 ± 0. 03 GeV ≈
e 8200m e
m e = 0. 5109989461±0. m μ = 105. 6583745±0. 0000024 m τ = 1. 77686±0. 00012 GeV ≈
0000000031 MeV MeV ≈ 210m e 3500m e
8.8.4 The Yukawa Lagrangian for Fermions
The complete Yukawa Lagrangian for the fermions is the sum
(8.33)
with the Yukawa Lagrangian for the leptons from Eq. (8.26) and for the
quarks from Eq. (8.29).

8.9 The Complete Lagrangian of the Standard Model

After symmetry breaking the complete Lagrangian of the Standard Model in
unitary gauge is the sum of all the boxed Lagrangians above, i.e. the sum of the
Lagrangians in Eqs. (8.20), (8.22), (8.33) and (8.25):

The Standard Model discussed so far has the following 18 parameters

that have to be determined by experiments:

The coupling constants g s , g w , g′ for the gauge group SU(3) C ×

SU(2) L × U(1) Y (equivalently, the strong and electric coupling
constants g s and e, and the Weinberg angle θ W ).

The parameters μ, λ of the Higgs potential (equivalently, the absolute

value v of the Higgs condensate and the mass m H of the Higgs boson
).

Three Yukawa couplings for the leptons and six Yukawa couplings for
the quarks (equivalently, the masses of the leptons and the quarks).

Three quark mixing angles θ 12, θ 13, θ 23 and the KM phase δ that
determine the CKM matrix.

8.10 Lepton and Baryon Numbers

We briefly discuss lepton numbers and baryon numbers.

Definition 8.10.1 We make the following definitions:

the electron lepton number L e is the number of electrons and electron
neutrinos minus the number of their antiparticles.
the muon lepton number L μ and the tau lepton number L τ are defined
similarly.
the total lepton number is L = L e + L μ + L τ .

Inspecting all interaction vertices in the Feynman diagrams above it follows that:

Corollary 8.10.2 (Lepton Number Conservation) The interactions in the

Standard Model conserve all lepton numbers L e , L μ , L τ , L separately.

For example, an electron can change into an electron neutrino by emitting a W −-

boson or an electron and an electron antineutrino can annihilate in a W −-boson.
In each process the electron lepton number is conserved.
In contrast to the Standard Model, it turns out that the neutrinos in nature
have small non-zero masses. This experimental observation is related to the
phenomenon of neutrino oscillations (see Sect. 9.2.2), which implies that the
lepton numbers L e , L μ , L τ are in fact not conserved separately, but only the
total lepton number L is conserved. This leads to the question of whether
charged lepton flavour violation (CLFV) might also occur in nature,
i.e. processes such as

or

which are not possible in the Standard Model and where some or all of the
lepton numbers L e , L μ , L τ are not conserved for the charged leptons e, μ, τ.
CLFV is predicted by certain theories beyond the Standard Model. In addition,
there may be processes involving lepton number violation (LNV) where the total
lepton number L is not conserved. For more details, see [75].

Definition 8.10.3 We define the baryon number B to be one third of the

number of quarks minus the number of antiquarks:
Again, inspecting all interaction vertices in the Feynman diagrams above it
follows that:

Corollary 8.10.4 (Baryon Number Conservation) The interactions in the

Standard Model conserve the baryon number B.

Because of quark mixing we do not split the baryon number into different
generations. Baryons consisting of three quarks, like the proton and neutron,
have baryon number + 1. Mesons consisting of quark-antiquark pairs have
baryon number 0.
Some theories beyond the Standard Model, for instance, GUTs with proton
decay (see Sect. 9.5.8), predict that the baryon number B is not conserved
(baryon number violation) .

Remark 8.10.5 Most matter particles observed in nature, leptons and hadrons ,
are unstable and decay , usually with very short lifetimes . Decays are due to the
weak, electromagnetic and strong interaction, with decays via the weak
interaction having the longest lifetime and decays via the strong interaction
having the shortest lifetime. Decays are related to the creation and annihilation
of particles and can be described by the Feynman diagrams discussed above.
In the Standard Model the only stable leptons are
the electron e and
the neutrinos ν e , ν μ , ν τ (as mentioned above there exist, however,
oscillations between neutrinos)
and the only stable hadron is
the proton p,
together with the corresponding antiparticles (isolated neutrons n are
unstable with a lifetime of 880 seconds [106], but they can become stable when
bound in a nucleus).
Particles that are created in particle colliders and that decay via the weak
interaction with lifetimes of 10−8 seconds or longer can travel many meters
before they decay, due to the relativistic effect of time dilation, and can thus be
found in particle collectors. Particles that decay via the strong or electromagnetic
interaction are usually too short-lived to be detected themselves [42].
8.11 Exercises for Chap. 8
8.11.1 Consider

with the representation

Let be a function which is invariant under the action of SU(2) × U(1).

1. Prove that there exists a function such that V (w) = f( | | w | | ),

where | | w | |2 = w † w.

2. Suppose that V is a polynomial of order at most 4 in the components of w

and that V has a minimum, but not in w = 0. Prove that V (w) is of the form

with certain constants μ, λ > 0.

8.11.2 Work out the details in the proof of Theorem 8.1.20 on the existence
of unitary gauges in the electroweak theory.
8.11.3 Prove Lemma 8.3.2 on the scalar product on .
8.11.4 Prove Lemma 8.4.2 on the scalar product on .
8.11.5 (From [ 141 ]) Suppose that the value of the Fermi constant G F 0 and
the electric charge e are known, but the value of the coupling constants g w , g′
and the Weinberg angle θ W are unknown. Determine lower bounds on the
masses of the W- and Z-bosons assuming the electroweak gauge theory.
8.11.6 Derive the weak isospin, weak hypercharge and electric charge of the
gauge bosons W ±, Z 0 and γ in Table 8.10.
8.11.7 Prove Theorem 8.5.7 on the subgroup in the Standard Model group
SU(3) C × SU(2) L × U(1) Y .
8.11.8 Derive the formulas in Lemma 8.6.1 concerning the action of the
electroweak gauge field W + B in the representations and of
SU(2) L × U(1) Y .
 8.11.9 Prove the formulas for the electroweak interaction vertices of leptons
and quarks in Theorem 8.6.2.
8.11.10 Can an electron-positron collider produce hadrons? Which particles
can a proton-proton collider produce?
8.11.11 Prove the formulas for the covariant derivative of the Higgs field and
for the Higgs potential in Lemma 8.7.1.
8.11.12 Prove the formula for the electroweak Higgs Lagrangian in
Theorem 8.7.2.
8.11.13 Prove Lemma 8.7.3 and Proposition 8.7.4 on the curvature of the
electroweak gauge field.
8.11.14 Prove Lemma 8.8.1 and Theorem 8.8.2 on the Yukawa Lagrangian
for leptons.
8.11.15 Prove Lemma 8.8.3 on the action of on ϕ c .
8.11.16 Prove Lemma 8.8.4 and Theorem 8.8.6 on the Yukawa Lagrangian
for quarks.
8.11.17 The Georgi–Glashow SO(3) model is described by a Yang–Mills–
Higgs Lagrangian over 4-dimensional Minkowski spacetime M with gauge
group G = SO(3), Higgs vector space with the fundamental
representation of SO(3) and standard scalar product, and Higgs potential

for positive constants μ, λ. Discuss the Higgs mechanism for this theory,
determine the masses of the gauge bosons and the explicit form of the Yang–
Mills–Higgs Lagrangian in unitary gauge.

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
5.
 [MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
 Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)
36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
 [MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
 177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
 Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]
101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]
117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]
132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions. Chin.
 Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/

Footnotes
1 This is one of the few places in this book where we cite a result from quantum field theory . The
equations hold independent of the choice of normalization of hypercharge.

2 Masses for fermions will be introduced in Sect. 8.8.

© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_9
Chapter 9 Modern Developments and
Topics Beyond the Standard Model
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

This chapter contains some advanced concepts in particle physics as well as

modern developments that aim at going beyond the Standard Model. The topics
range from well-established phenomena to more hypothetical theories that
predict, for example, the existence of new particles and interactions that so far
have not been observed. Most sections in this chapter can be read independently
of one another and are only loosely interconnected.
Rather than trying to give a detailed account of these subjects, the intention
is to enable the reader to study the extensive research literature him- or herself.
Each section is accompanied by some suggestions for further reading. Again,
these small guides to the literature do not try to be complete.

9.1 Flavour and Chiral Symmetry

In this section we discuss flavour symmetry and chiral symmetry breaking in
QCD (see [36] for more details).
We fix the following data:
an oriented and time-oriented Lorentzian spin manifold M of even
dimension n with metric g of signature (1, n − 1) or (n − 1, 1)
a spin structure Spin+(M) together with complex spinor bundle S = S L S
R→ M

a Dirac form ⋅ ⋅ on the spinor space Δ = Δ L Δ R defined by a matrix A

as in Proposition 6.7.13 together with associated Dirac bundle metric ⋅ ⋅ S
 a principal G-bundle P → M with compact structure group G of dimension
r
an Ad-invariant positive definite scalar product on the Lie algebra ,
together with the induced bundle metric ⋅ ⋅ Ad(P) on the associated vector
bundle Ad(P)
a complex representation ρ: G → GL(V ) with associated complex vector
bundle E = P × ρ V → M
a G-invariant Hermitian scalar product ⋅ ⋅ V on V with associated bundle
metric ⋅ ⋅ E on the vector bundle E. Together with the Dirac form on the
spinor bundle S we get a Hermitian scalar product ⋅ ⋅ S E on the twisted
spinor bundle S E. We abbreviate Ψ, Φ S E by .
We consider N f copies E f , with f = 1, …, N f , of the associated vector bundle
E and form the direct sum

The different copies of E are called flavours . The Hermitian scalar product
on F is the direct sum of the scalar products on each copy E f . The associated
Yang–Mills–Dirac Lagrangian is given by

where Ψ f is a twisted spinor with values in S E f and m f is the mass of the

flavour with index f.

Example 9.1.1 For , G = SU(3) with the fundamental representation on

and six flavours f = u, d, c, s, t, b the Lagrangian is the Lagrangian
of QCD for quarks of masses m f . We can consider the Lagrangian as a
generalization of QCD to an arbitrary compact gauge group G and an arbitrary
number of quark flavours. For example, one sometimes considers versions of
QCD with gauge group G = SU(N c ), corresponding to N c colours.

The Lagrangian is gauge invariant under gauge transformations with

values in G. However, depending on the quark masses m f there are additional
global symmetries , i.e. symmetries which are constant over spacetime. We only
consider the Dirac part

of the Lagrangian , because the global symmetries we consider leave the

gauge field A and hence the Yang–Mills Lagrangian invariant.

Lemma 9.1.2 Suppose that all flavours have the same mass m f = m, for f = 1,
…, N f . Let

be a (constant, independent of the point in spacetime M) unitary matrix. Then

the Dirac Lagrangian is invariant under the flavour symmetry

The flavour symmetry acts through bundle automorphisms of S F, mixing the

flavour components.

Recall that the Dirac mass term is equal to

The first term in the Dirac Lagrangian is the real part of

Lemma 9.1.3 Suppose that all flavours have vanishing mass m f = 0, for f = 1,
…, N f . Let

be a pair of (constant) unitary matrices. Then the Dirac Lagrangian is

invariant under the chiral symmetry

The chiral symmetry acts through bundle automorphisms of S L F and S R

F, mixing the flavours of right-handed and left-handed components separately.
Example 9.1.4 The chiral symmetry is already interesting for a single massless
fermion, N f = 1. In this case the chiral symmetry group is

It is sometimes useful to consider the following subgroups of the chiral

symmetry group

which are called the subgroups of vector and axial symmetries. The action of
the axial symmetry on a spinor Ψ = Ψ L + Ψ R is often written in the physics
literature as

where Γ n+1 is the physical chirality operator which is + 1 on Ψ L and − 1 on Ψ R .

As we saw in Lemma 9.1.2 the vector symmetry still exists in the case when the
mass m is non-zero.

Example 9.1.5 In QCD the quarks with flavours u, d, s are much lighter than
the quarks with flavours c, t, b. If we set m u = m d = m s = 0 (this is called the
chiral limit of QCD), then QCD should have the chiral symmetry U(3) L × U(3)
R.

The chiral symmetry group U(N f ) L × U(N f ) R has the following subgroups:

For the quantum field theory associated to QCD in the chiral limit it is known
that the full chiral symmetry group U(3) L × U(3) R of the classical field theory
breaks in two steps to the flavour symmetry group U(3) V :

The reason for the first breaking is that the axial symmetry U(1) A does not hold
in the quantum theory (because of the axial anomaly); note that the discrete
subgroup is still contained in . The second
breaking is called chiral symmetry breaking , which happens because the
vacuum state of QCD is not invariant under the full symmetry group
(this is an example of spontaneous symmetry breaking of a
global symmetry).
Chiral symmetry breaking is not fully understood theoretically, because it
happens at low energies (similar to quark confinement ), where the strong
coupling constant is large and perturbation theory is not valid; see Sect. 9.4.
However, it can be studied in numerical simulations using lattice QCD and
analytically in supersymmetric generalizations of QCD, which are better
understood non-perturbatively (see [127, 128]).
The unbroken symmetries of the quantum version of QCD, in addition to
gauge symmetry, are therefore
the abelian symmetry U(1) V (related to baryon number conservation)
the (special) flavour symmetry SU(3) V .
Baryon number conservation is an exact symmetry of QCD, independent of
quark masses. Flavour symmetry, however, in real world QCD, where the quark
masses m u , m d , m s are not precisely the same, is only an approximate
symmetry, which can nevertheless still be observed in the spectrum of hadrons
(mesons and baryons ) that are composed of quarks with flavours u, d, s.
Historically, flavour symmetry SU(3) V preceded the development of QCD with
gauge symmetry SU(3) C (the dimension 8 of the flavour symmetry group is
related to the concept of the eightfold way , developed by Murray Gell-Mann in
the 1960s).
To summarize, in QCD the Lie group SU(3) appears in two completely
different places:

as the colour gauge group SU(3) C . Gauge symmetry in QCD is a

local and exact symmetry.

as the flavour symmetry group SU(3) V . Flavour symmetry in QCD is

a global and with non-identical quark masses only approximate
symmetry .

Remark 9.1.6 Non-perturbative QCD is an active and difficult area of

research with many open questions. Understanding non-perturbative QCD in
particular would mean to understand the structure of hadrons theoretically.
It is not known at present, to mention only one example, how to calculate the
mass of the proton and neutron (which are well-known from experiments)
analytically from first principles using a “formula” (there are numerical
calculations from first principles using lattice QCD which are accurate to within
a few percents). The proton consists of two up and one down valence quark and
the neutron of two down and one up valence quarks. The masses of the proton
and neutron are [106]

However, the sum of the quark rest masses for the proton and neutron are
approximately 9 MeV and 11. 6 MeV, i.e. they only amount to roughly 1% of the
proton and neutron mass. This leads to the remarkable conclusion that 99% of
the mass of the visible matter in the universe comes from the binding energy of
the gluon field inside the proton and neutron (and hence does not have its origin
in the Higgs mechanism ) [11, 137].

9.1.1 Further Reading

Reference [21] is an extensive review of strongly coupled QCD. The notes [36]
contain a discussion of chiral symmetry breaking. References [44] and [18] are
examples of precise calculations of hadron masses using lattice gauge theory.

9.2 Massive Neutrinos

In the Standard Model, neutrinos are assumed massless. According to
experiments, however, neutrinos show oscillations between different flavours ,
for example, a free electron neutrino can change in-flight into a muon neutrino
or tau neutrino and back (first observed at the Super-Kamiokande detector in
1998; Nobel Prize in Physics 2015 for Takaaki Kajita and Arthur B. McDonald).
This can only be explained if neutrinos have different and hence (at least two of
them) non-zero masses.
In this section we discuss how mass terms for neutrinos can be added to the
Standard Model as well as the phenomenon of neutrino oscillations. The
neutrino masses are extremely small compared to the other fermions in the
Standard Model and an obvious question is why neutrinos are so light. The
famous seesaw mechanism , that we discuss in Sect. 9.2.5, is one natural
explanation (we follow references [62] and [137] throughout this section).

9.2.1 Dirac Mass Terms

To define Dirac mass terms for neutrinos we have to postulate the existence of
new particles, right-handed neutrinos ν iR for each generation i = e, μ, τ. Like
right-handed electrons, muons and taus, right-handed neutrinos are singlets in
the trivial representation of SU(3) C × SU(2) L . However, because they have zero
electric charge, this implies that their weak hypercharge Y is also zero. Hence
right-handed neutrinos, if they exist, are sterile , i.e. live in the trivial singlet
representation of the full Standard Model gauge group

and do not interact with other particles via gauge bosons. They can only interact
with other particles via the Higgs boson (if their mass is generated by the Higgs
mechanism) and gravity. For this reason, sterile neutrinos are one of the
candidates to explain dark matter in the universe.
Table 9.1 Lepton representations including right-handed neutrinos

Sector SU(2) L × U(1) Y representation Basis vectors Particle Charges

LL1 ν eL −1 0

eL −1 −1

LR1 1 ν eR 0 0 0

1 eR 0 −2 −1

If we add right-handed neutrinos to the Standard Model, the particle content

of the lepton sector becomes formally very similar to the quark sector , see
Table 9.1 : there are one left-handed isodoublet and two right-handed isosinglets
for each generation. The Dirac mass generation for neutrinos using Yukawa
couplings is a straightforward generalization of the mass generation for quarks in
Sect. 8.8.2: we would like to determine the most general gauge invariant Yukawa
form that generates masses for both electron-type particles and neutrinos. We
consider the following SU(2) L × U(1) Y representation spaces:
 (9.1)
and

(9.2)
We write

for the irreducible summands in Eqs. (9.1) and (9.2) and indicate the left-
handed basis by an index L and the right-handed basis by an index R. We also
write for the left-handed isodoublets

In complete analogy to Lemma 8.8.4 we have:

Lemma 9.2.1 For arbitrary complex matrices Y e and Y ν the expression

is an SU(2) L × U(1) Y -invariant Yukawa form (the second line is an

abbreviation).

We can find pairs of unitary matrices U L e , U R e and U L ν , U R ν that diagonalize

the matrices Y e and Y ν :

where all entries of the diagonal matrices are real and positive. We define new
leptons
These leptons define a new splitting of the representation spaces into direct
summands as in Eqs. (9.1) and (9.2).
We also define new neutrinos

Definition 9.2.2 We call the basis vectors corresponding to the leptons e, ν 1,

μ, ν 2, τ, ν 3 mass eigenstates and the basis corresponding to e, ν e , μ, ν μ , τ, ν τ
weak eigenstates .
Suppose that the Higgs field ϕ is in unitary gauge ,

and define the lepton masses

for each flavour i = e, ν 1, μ, ν 2, τ, ν 3.

We then get:

Theorem 9.2.3 (Yukawa Coupling Including Dirac Mass Terms for

Neutrinos) After symmetry breaking the Yukawa Lagrangian for the three
lepton generations associated to the Yukawa form in Lemma 9.2.1 , including
right-handed neutrinos, is given in unitary gauge and in the mass eigenstate
basis (e, ν 1, μ, ν 2, τ, ν 3) by
 (9.3)

Fig. 9.1 Neutrino-Higgs interaction vertex

The only difference to the Yukawa Lagrangian in Theorem 8.8.2 is that the
Yukawa Lagrangian in Theorem 9.2.3 contains (potentially non-zero) masses for
the neutrinos and interactions between the neutrinos and the Higgs boson,
depicted in the Feynman diagram in Fig. 9.1. We also see that there is in general
a non-trivial neutrino mixing from the mass eigenstate basis to the weak
eigenstate basis.

Definition 9.2.4 The unitary matrix

is called the Pontecorvo–Maki–Nakagawa–Sakata (PMNS) matrix . The

PMNS matrix plays the same role for the mixing of left-handed neutrinos
across generations as the CKM matrix for the left-handed down quarks.

The PMNS matrix is often written in the following form:

This matrix appears in the flavour changing electroweak interaction vertices

involving W ±-bosons if we want to write them in the mass eigenstate basis
instead of the weak eigenstate basis: consider the following charged current term
in Eq. (8.17):

with the Hermitian conjugate of the lepton current

Using the mass eigenstate basis we can write the current equivalently as

(9.4)
The interaction terms involving Z-bosons and photons γ are the same in both the
mass eigenstate basis and the weak eigenstate basis. Similarly to our discussion
in the case of quarks it follows that if neutrinos have a Dirac mass, then
interaction vertices involving W-bosons can connect leptons from different
generations. As a result the lepton numbers L e , L μ , L τ are not conserved
separately, only the total lepton number L is invariant.
If only a Dirac mass term is present for neutrinos, it is unclear why the
neutrino masses (or the corresponding Yukawa couplings) are so much
smaller than for the other fermions. If we add to the Lagrangian a Majorana
mass term for the right-handed neutrinos, the seesaw mechanism described in
Sect. 9.2.5 allows both: a Dirac mass of similar size to the masses of the other
fermions and a neutrino of very small mass. In addition, the theory then
predicts another neutrino of very large mass (that so far has not been
observed).

9.2.2 Neutrino Oscillations

In this section we discuss how different masses of neutrinos lead to the
phenomenon of neutrino oscillations . We just describe the basic idea. Consider
the following elementary lemma.

Lemma 9.2.5 Let E j and p j denote the energies and 3-momenta of relativistic
particles with rest mass m j , where j = 1, 2. Suppose that

Then m 1 = m 2.

Proof Energy and 3-momentum determine the rest mass according to the
formula m 2 = E 2 − p 2 (where c = 1). This implies the claim. □

To simplify the discussion we only consider two neutrino generations, say the
electron neutrino and the muon neutrino (we could similarly analyse the case of
three generations). Let | ν j for j = 1, 2 denote the mass eigenstates of neutrinos
with mass m j . For a freely moving neutrino in one of the mass eigenstates, the
spacetime dependence of the state is given by

Clearly, the probability of finding the j-th neutrino at the spacetime point (x, t) in
the state | ν j is 1:

On the other hand consider, for example, the electron neutrino ν e . As a state it
decomposes as

where U is the neutrino mixing matrix corresponding to the PMNS matrix for
three generations. We have

and

since the matrix U is unitary.

The spacetime dependence of the electron neutrino state is given by

The probability of finding the state at the spacetime point (x, t) in the state | ν
e is
Proposition 9.2.6 Suppose that U ej ≠ 0 for j = 1, 2. If

then m 1 = m 2.

Proof Since the matrix U is unitary we have

which is equal to 1 for all only if E 1 = E 2 and p 1 = p 2. The claim

then follows by Lemma 9.2.5. □

This means that if the masses m 1 and m 2 are not the same and U ej ≠ 0 for j = 1,
2, then there will be spacetime points (x, t) where

But this implies that there is a non-zero probability of finding | ν e (x, t) in the
state | ν μ :

These spacetime points are related to the values of the phases and occur
periodically. This is the mechanism of neutrino oscillations : there is a non-zero
probability of finding a neutrino, which is in a weak eigenstate at one spacetime
point (0, 0), in another weak eigenstate at another spacetime point (x, t).
For more details, see the excellent discussion in Thomson’s book [137]. It is
possible to show that under certain assumptions the converse to
Proposition 9.2.6 holds: if neutrino oscillations occur, then the masses of the
neutrinos cannot be the same. These neutrino oscillations are indeed observed in
experiments. A detailed calculation shows that using neutrino oscillations only
the differences

can be determined experimentally, but not the absolute values of m i for i =

1, 2, 3.

9.2.3 Experimental Values for the PMNS Matrix and

Neutrino Masses
We briefly discuss experimental values for neutrino masses and the PMNS
matrix (following [107] and [137, Sect. 13.9]). Since there exists a large number
of neutrinos in the universe, neutrino masses potentially have cosmological
implications. Recent cosmological measurements set a rough upper limit of

Direct measurements of the averaged electron (anti)neutrino mass , defined

by the square root of

(in particular, for neutrinos from the supernova SN 1987A and β-decay of
Tritium3H) give similar upper bounds [43]. From neutrino oscillation
experiments the following is known:

The sign of Δm 32 2 is not known, hence it is not clear whether m 3 > m 2 > m 1
(normal mass hierarchy) or m 3 < m 1 < m 2 (inverted mass hierarchy) .
We can write the PMNS matrix in a similar way to the CKM matrix using
neutrino mixing angles:

where

The following values are known:

The phase δ is unknown. This leads to the following approximate absolute

values for the entries of the PMNS matrix:
We conclude that the PMNS matrix is clearly much less diagonal than the
CKM matrix in Eq. (8.32).

9.2.4 Majorana Mass Terms

Majorana mass terms allow us to introduce a non-zero mass for neutrinos
without postulating the existence of right-handed neutrinos. We begin by
considering only the first generation. We want to define a Majorana mass term

where

denotes the Majorana conjugate . This expression is not gauge invariant, because
the left-handed electron neutrino comes in an isodoublet with the left-handed
electron:

with Majorana conjugate

We need a generalization of the Higgs mechanism : consider the Higgs doublet

Lemma 9.2.7 For all matrices A SU(2) the following identity holds:

Hence the expression

is a left-handed Weyl spinor, invariant under SU(2) L × U(1) Y . Its Majorana

conjugate is .

Proof This follows from Lemma 8.8.3, because for A SU(2). □

Using the pure uncharged section ϕ T σ 2 L eL of the left-handed spinor bundle S L
, we can now form a Majorana mass term:

Proposition 9.2.8 For constants and the expression

is Lorentz and SU(2) L × U(1) Y invariant.

Suppose that the Higgs field ϕ is in unitary gauge,

and define the neutrino mass

Corollary 9.2.9 (Majorana Mass Terms for Neutrinos) After symmetry

breaking the Lagrangian is given in unitary gauge by

The term in the first line is the Majorana mass term for the electron neutrino, the
terms in the second line are hypothetical interactions between the electron
neutrino and the Higgs boson.
The coupling between two scalars and two spinors in the Lagrangian
(or between two spinors and the square of the Higgs boson after symmetry
breaking) does not appear in Chap. 7 and is non-renormalizable . We
conclude that a gauge invariant Majorana mass term for left-handed neutrinos
can only be added to the Lagrangian using non-renormalizable interactions.
The Lagrangian is understood as an effective Lagrangian that hides new
interactions beyond the Standard Model. This can be compared to the non-
renormalizable 4-Fermi Lagrangian for the weak force that was hiding gauge
interactions (see the beginning of Sect. 8.3).

If is dimensionless, then must have dimension of mass so that has

the dimension of a Lagrangian. The constant is interpreted as a new (large)
energy scale. If we assume that and v ∼ 102 GeV, as in the Standard
Model, as well as GeV (below a typical Grand Unification scale), then
eV. This is in agreement with experiments.

Remark 9.2.10 Note that for right-handed sterile neutrinos ν eR , the Majorana
mass term

is already gauge invariant and thus well-defined without the need to introduce
the Higgs field. We will combine this term with the Dirac mass term in
Sect. 9.2.5.

9.2.5 Dirac–Majorana Mass Terms and the Seesaw

Mechanism
We again make the assumption from Sect. 9.2.1 that there exist right-handed
sterile neutrinos in each of the three lepton generations. We saw that a Dirac
mass term can then be generated in a gauge invariant way via the Higgs
mechanism. Furthermore, in Remark 9.2.10 we saw that a gauge invariant
Majorana mass term for the right-handed neutrinos can be added to the
Lagrangian (we ignore the Majorana mass term for the left-handed neutrinos that
can only be added in a gauge invariant and renormalizable way by introducing
new interactions).
Let m D be the Dirac mass and M the Majorana mass of the right-handed
neutrino (we only consider one generation). Then the Dirac mass term is

and the Majorana mass term is

Recall from Sect. 6.7 and Sect. 6.8 the definitions

where for 4-dimensional Minkowski spacetime with signature (+, −, −, −)

It follows that

Remark 9.2.11 We assume from now for the rest of this section that
components of spinors anticommute .

Lemma 9.2.12 The Majorana mass term can be written as

Proof We calculate

In the last step we used that (αβ) = −α β for Grassmann numbers α, β. □

Lemma 9.2.13 The Dirac mass term can be written as

The second line follows from the first, because spinor components anticommute.

Proof We calculate as in Lemma 9.2.12

Proposition 9.2.14 The sum of the Dirac and Majorana mass term can be
written as
The basis ν L , ν R C is not a mass eigenbasis for m D ≠ 0, because the matrix is not
diagonal. The following is easy to verify:

Lemma 9.2.15 The matrix

has real eigenvalues

If M m D > 0, then one eigenvalue m + is positive and the other eigenvalue −

m − is negative, where

Proposition 9.2.16 (Diagonalization of Dirac–Majorana Mass Terms) Let

R be an orthogonal matrix such that

Define the unitary matrix

and set

Then the Dirac–Majorana mass term can be written as

We introduce the spinors

Then
 hence ν ± are Majorana spinors. Moreover,

We get:

Corollary 9.2.17 With the Majorana spinors ν ± the Dirac–Majorana mass

term can be written as the Dirac mass term

The relation between m + and m − for fixed m D is called a seesaw mechanism

: if m + is large, then m − is small and if m − is small, then m + is large. The
seesaw mechanism allows a Dirac mass m D (generated by the Higgs
mechanism ) to be of the same size as for the other fermions of the Standard
Model (around 1 GeV) and to have at the same time a small neutrino mass m
− (around 0. 01 eV). The theory then predicts another neutrino which has a
very large mass m + (around 1011 GeV).

Considering Dirac–Majorana mass terms for all three generations there is

again a phenomenon of neutrino mixing; see [62] for details.

9.2.6 Further Reading

The book [137] by M. Thomson contains many details on neutrino experiments
as well as theoretical discussions of neutrino mixing and neutrino oscillations.
The book [62] by C. Giunti and C.W. Kim is a comprehensive exposition of
neutrino physics, including astrophysical and cosmological implications.

9.3 C, P and CP Violation

In addition to Lorentz symmetries, gauge symmetries and global continuous
symmetries (like flavour symmetry), Lagrangians (or actions) can also have
discrete symmetries . Charge conjugation C and parity inversion P, together
with time reversal T, are examples of such symmetries. If the action is not
invariant under the composition CP of P followed by C, the theory is said to
violate CP. It is an important observation that the action of the Standard Model,
because of quark mixing and the weak interaction of quarks, indeed violates CP.
 Violation of CP in general is linked to the asymmetry between the amount of
matter and antimatter in the universe. However, the CP violation coming from
the weak interaction of quarks alone does not suffice to explain the observed
degree of this asymmetry. There are two other instances in the Standard Model
where CP violation can possibly occur: in the interaction of neutrinos, related to
neutrino mixing , and perhaps in the strong interaction of quarks if an additional
CP violating term is added to the Lagrangian (although this is almost certainly
ruled out by experiments, leading to the so-called strong CP problem , i.e. to
explain why the strong interaction preserves CP).
We only discuss CP violation coming from quark mixing and the weak
interaction. Throughout this section we follow the excellent exposition in the
book [22], where more details can be found.

9.3.1 The CKM Matrix and the Jarlskog Invariant

We consider the following form of the CKM matrix from Sect. 8.8.3:

To understand CP violation in the Standard Model, it will be important to know

in which sense the CKM matrix is real or complex. We can change the entries of
the CKM matrix without changing the physical content by multiplying any row
or column by a phase in U(1). This corresponds to changing the entries by
(9.5)
where α i , α j [0, 2π) are arbitrary angles. Only expressions which are
invariant under these rephasings are physically relevant. For example, the term

is invariant (the experimental values of these norms can be found in Eq. (8.32)).
Another invariant term is

which changes under the rephasings in Eq. (9.5) by the factor

Definition 9.3.1 The Jarlskog invariant is

Proposition 9.3.2 Suppose that all entries of the CKM matrix V CKM are non-
zero. Then the Jarlskog invariant vanishes, J = 0, if and only if V CKM can be
brought by the rephasings in Eq. (9.5) into real form (i.e. all entries of the matrix
become real).

Proof It is possible to show that we can always bring the CKM matrix with
rephasings into the form

where

We get

Since all entries of the CKM matrix are non-zero, the factor in front of sinδ is
non-zero. Hence J = 0 if and only if e iδ = ±1, which happens if and only if the
CKM matrix can be brought into real form. □

According to [110] the experimental value for the Jarlskog invariant is

This is a small but non-zero number, and implies:

Corollary 9.3.3 The CKM matrix for quark mixing realized in nature cannot
be brought into real form by the rephasings in Eq. (9.5).

Remark 9.3.4 A similar discussion applies to the PMNS matrix describing

neutrino mixing. However, as mentioned in Sect. 9.2.3, the analogue of the
Jarlskog invariant (or the phase δ) is currently not known for the PMNS matrix.
It is likely that δ ≠ 0, π so that the PMNS matrix cannot be brought into real form
by rephasings, but so far this cannot be excluded completely [1].

Remark 9.3.5 Exercise 9.7.1 shows that any unitary 2 × 2-matrix can be
brought into real form by rephasings. This is the reason that led Makoto
Kobayashi and Toshihide Maskawa in 1972 (Nobel Prize in Physics 2008 ) to
postulate the existence of a third quark generation.

9.3.2 C and P Transformations

There is some freedom how to define charge conjugation C and parity inversion
P in field theories. For the following notation it is useful to think in this section
of fields as quantum fields, i.e. fields on spacetime with values in the operators
on a Hilbert space. Then C and P are unitary operators on this Hilbert space and
a field ϕ transforms as

We consider 4-dimensional Minkowski spacetime of signature (+, −, −, −) with

the conventions for spinors from Sect. 6.8. Note that according to these
conventions

We first define parity inversion P. On spacetime, P acts by

This implies for partial derivatives

On a complex scalar field ϕ(x) we define

which implies

Here α p is an arbitrary real constant. On a spinor field ψ(x) we define

which implies for

Here β p is an arbitrary real constant.

Proposition 9.3.6 (Parity Inversion and Invariance of Actions) Define

parity inversion P on a gauge field A μ by

Then the Klein–Gordon Lagrangian

the Higgs Lagrangian

the Dirac Lagrangian

and the Yang–Mills Lagrangian

transform as

In particular, the Klein–Gordon, Higgs, Dirac and Yang–Mills actions (the

spacetime integrals over the Lagrangians) are invariant under parity inversion.

Proof This is Exercise 9.7.2. □

We now define charge conjugation C. On a complex scalar field ϕ(x) we define

which implies

Here α c is an arbitrary real constant. Let ψ be a spinor field. Recall from Sect. 6.
8 that

We set
which implies

Here β c is an arbitrary real constant.

Proposition 9.3.7 (Charge Conjugation and Invariance of Actions) Define

charge conjugation C on a gauge field A μ by

where A μ = −A μ † is the matrix-valued gauge field in a given unitary

representation of the compact gauge group G. Then the Klein–Gordon
Lagrangian , the Higgs Lagrangian , the Dirac Lagrangian and the
Yang–Mills Lagrangian transform as

This implies that the Klein–Gordon, Higgs, Dirac and Yang–Mills actions are
invariant under charge conjugation.

Remark 9.3.8 In the case of the Dirac Lagrangian we assume in this

proposition that the field operators ψ and ψ † anticommute .

Proof We do the calculation for the Dirac Lagrangian and leave the remaining
cases to Exercise 9.7.3. The Dirac mass term transforms as

In the step from the third to the fourth line we used that ψ T and ψ
anticommute. The term involving the Dirac operator transforms as
From the second to the third line we used that Γ 2 Γ μ Γ 2 = Γ μ , from the third
to the fourth line that ψ T and ψ anticommute and from the fourth to the fifth
line that Γ μ† Γ 0 = Γ 0 Γ μ and A μ † = −A μ . □

Definition 9.3.9 The transformation CP is defined by first applying P and

then applying C.

From the formulas above we see that the transformation CP is given on

complex scalar fields ϕ by

and on spinors ψ by

where α cp , β cp are arbitrary real numbers.

Corollary 9.3.10 Define the CP transformation on a gauge field A μ by

Then the Klein–Gordon, Higgs, Dirac and Yang–Mills actions are CP invariant.

We shall see in the following section that if we introduce in addition to these

Lagrangians a Yukawa Lagrangian , then the complete action may no longer be
CP invariant.

9.3.3 CP Violation in the Standard Model

We want to prove the following theorem.

Theorem 9.3.11 (CP Invariance and the Jarlskog Invariant) The action of
the Standard Model (with vanishing neutrino masses) is CP invariant if and only
if the Jarlskog invariant J of the CKM quark mixing matrix vanishes.

Proof We follow the argument in [22] and use anticommuting spinors

throughout the proof. We first assume that the action of the Standard Model is
CP invariant and consider the Lagrangian of the Standard Model without fixing a
particular gauge like the unitary gauge. We write the Higgs field as

and set ϕ − = ϕ + . We only need to consider the following parts of the complete
Lagrangian:

From the term in the Higgs Lagrangian the part

From the Dirac Lagrangian the following charged current part of the
electroweak interaction vertex with quarks in the mass eigenstate basis:

with

and

as in Eqs. (8.31) and (8.30). Here we use the index α to denote up-type
quarks and k to denote down-type quarks. The matrix V is the CKM matrix.
The Yukawa form for the quarks in Lemma 8.8.4 before symmetry breaking
is in the mass eigenstate basis

where V is the CKM matrix. The Yukawa Lagrangian for the quarks before
symmetry breaking can then be written as
 This is equal to

(9.6)

We now argue as follows (without using ): under the CP transformation

the field ϕ + transforms as

for some real number ξ W (we do not write the change of argument x ↦ x p ).
This implies

The Lagrangian is then CP invariant only if we transform

The first line of the Lagrangian in Eq. (9.6) is invariant under the
transformations

where ξ α , ξ k are arbitrary real numbers (and spinor components are

anticommuting). Consider the term

in . Its CP transform is
 Comparing with the last term in we see that CP invariance of the
Lagrangian implies that

This implies

Hence Q uscb is real and the Jarlskog invariant J = 0.

We can argue similarly with the Lagrangian : consider the term

Its CP transform is

Comparing this with the second term in we conclude again that

hence J = 0.
Conversely, it can be shown that if J = 0, then all terms in the Standard
Model Lagrangian are CP invariant. □

Corollary 9.3.12 Since in nature J ≈ 3. 04 ⋅ 10−5 ≠ 0, the action of the

Standard Model is not CP invariant.

A similar discussion can be done with the PMNS matrix describing neutrino
mixing . If the corresponding Jarlskog invariant is non-zero, then neutrino
mixing also leads to CP violation .

9.3.4 Further Reading

The book [22] by G.C. Branco, L. Lavoura and J.P. Silva is a very good source
for numerous details on CP violation. The paper [66] by W. Grimus and
M.N. Rebelo is a mathematical reference for CP violation in general gauge
theories.

9.4 Vacuum Polarization and Running Coupling

Constants
So far we have treated the parameters of the Standard Model, in particular, the
coupling constants of the gauge interactions and the masses of the particles, as
constants. This is only true at the classical or tree level , i.e. to zeroth order in
the Planck constant . If quantum corrections in higher order of are taken
into account, which can be calculated using loop diagrams and renormalization ,
it turns out that the masses and the coupling constants depend on an energy
scale μ with respect to a fixed renormalization scheme. In particular, the
coupling constants g become functions g(μ). These functions are known as
running coupling constants . In this section we depart from the usual course
and discuss (without proofs) this quantum effect, as a preparation for Sect. 9.5 on
Grand Unification.
The quantum corrections to the coupling constants are interpreted as vacuum
polarization . Corresponding loop diagrams are shown in Fig. 9.2, involving a
fermion loop, a gauge loop with two 3-boson vertices and a gauge loop with a 4-
boson vertex (these diagrams and more details can be found in [125, 137]). For
an abelian gauge theory with no direct interaction between gauge bosons, there
is only the diagram on the left with a fermion loop.

Fig. 9.2 Vacuum polarization diagrams

The dependence on the energy scale is given by the following differential

equation
where β is the so-called beta function . It can be proved that to lowest order
(one-loop) approximation the beta function is of the form

for a certain constant b. We get

which can be integrated with respect to an arbitrary reference energy μ 0 = M.

This yields:

Proposition 9.4.1 (Running Coupling Constants) The energy dependence of

the coupling constant g is given in one-loop approximation by

where b is a constant and M an arbitrary reference energy. If is small, then

We see that the sign of b determines whether g(μ) increases or decreases as the
energy scale μ increases:
if the constant b is positive, then g(μ) increases as the energy scale μ
increases
if b is negative, then g(μ) decreases as the energy scale μ increases.
Our aim is to give a formula for b depending on the gauge group G and the
fermion representations.
It will turn out that the vacuum polarization diagram involving the
fermion loop gives a positive contribution to b: at long distances (low
energies) the vacuum polarization due to the fermion loop shields the charges
and reduces the effective coupling .

On the other hand, the vacuum polarization diagrams involving gauge

boson loops in non-abelian gauge theories give a negative contribution to b:
at long distances the interactions between the gauge bosons increase the
 effective coupling. As a consequence, if the gauge group is non-abelian and
there are not too many fermions in the theory, the constant b will be negative,
hence the coupling g(μ) will be strong at low energies and weak at high
energies.

9.4.1 Casimir Operators

In this subsection we follow [125, Sect. 25.1]. Suppose that G is a semisimple
Lie group with Lie algebra of dimension n. According to Cartan’s Criterion
2.4.9 the Killing form is non-degenerate. Let {T a } a = 1 n be an arbitrary
basis of and

The n × n-matrix B is invertible and we denote the entries of the inverse matrix
by B ab .

Definition 9.4.2 Let R: G → GL(V ) be a representation of G. Then we define

the Casimir operator on V by

If R = Ad G is the adjoint representation, then we write .

It can be shown that the Casimir operator is independent of the choice of basis
{T a }.

Lemma 9.4.3 The Casimir operator commutes with all generators R

(T a ) for a = 1, …, n.

Definition 9.4.4 Suppose that R is an irreducible complex representation.

Since the Casimir operator commutes with all generators, it follows by Schur’s
Lemma that

for some complex number C 2(R), called the Casimir invariant of the
representation. If G is simple and R the complexified adjoint representation
(which is irreducible), we write C 2(R) = C 2(G).
Lemma 9.4.5 Let G = SU(N) and be the fundamental
(defining) representation. Then

For more mathematical details on the Casimir operator see, for example, [83,
Sect. V.4].

9.4.2 Running Coupling for Gauge Theories with

Fermions
Suppose we have an abelian gauge theory with Lie group G = U(1). We fix the
scalar product on such that the vector has length 1⁄g 2, where g is the
coupling constant . We consider N f massless Weyl fermions ψ f in
representations

of U(1), where f = 1, …, N f and q f are integers (winding numbers). We cite the

following theorem from quantum field theory:

Theorem 9.4.6 (Vacuum Polarization in Abelian Gauge Theories)

Consider the gauge group G = U(1). Taking into account the massless fermion
one-loop contribution to the vacuum polarization , the constant b is given by

References for this formula are [125, Sect. 16.3.3] (for one Dirac fermion of
charge q = 1) and [80, equation (4.21)] (for arbitrary Weyl fermions).

Example 9.4.7 For QED with gauge group U(1) Q and one Dirac electron we
get

In general, we see that every fermion in abelian gauge theories gives a positive
contribution to b.
 Since b Q is positive in QED, the electric coupling constant e(μ) increases
with increasing energy scale μ.

We can also calculate the vacuum polarization in the case of non-abelian gauge
theories: suppose we have a gauge theory with simple Lie group G = SU(N c ) of
dimension n = N c 2 − 1. We fix the Ad-invariant scalar product on by

where g is the coupling constant. We consider N f massless Weyl fermions ψ j in

representations R f of G, where f = 1, …, N f . We again cite without proof a
theorem from quantum field theory:

Theorem 9.4.8 (Vacuum Polarization in Non-abelian Gauge Theories)

Consider the gauge group G = SU(N c ). Taking into account the gauge boson
and massless fermion one-loop contributions to the vacuum polarization , the
constant b is given by

where

and d(R) is the dimension of the representation R. For G = SU(N c ) with n = N c

2 − 1 we have

and

for the fundamental representation R and

for the trivial representation R.

References for this formula (which holds for an arbitrary simple gauge group G
with a suitably normalized Ad-invariant scalar product on the Lie algebra) are
[125, equation (26.93)], [144, equation (17.5.41)], [85, Part I, equation (2.2)]
(with Dirac fermions) and [80, equation (4.29)] (for Weyl fermions). The
formula shows that the non-abelian group G itself (the vacuum polarization due
to gauge boson loops) gives a negative contribution to b, while the vacuum
polarization due to the fermion loops give a positive contribution.
The formulas in the following examples for the Standard Model can also be
found in [85, Part I, equation (2.4)].

Example 9.4.9 In QCD we have G = SU(3) C and 6 flavours of quarks , i.e. 12

Weyl fermions in the fundamental representation. We get in the limit of massless
quarks

Since b C is negative in QCD, the strong coupling constant g s (μ) gets

smaller as the energy scale μ increases. This is known as asymptotic
freedom and was discovered by David J. Gross, H. David Politzer and Frank
Wilczek (Nobel Prize in Physics 2004 ).

Asymptotic freedom is not only theoretically interesting, but also

practically very useful, because it shows that QCD at high energies is
amenable to perturbation theory . For low energies the strong coupling
constant g s (μ) is large, hence QCD becomes non-perturbative .

Example 9.4.10 In the weak interaction described by the group G = SU(2) L we

have in each generation three quark doublets of colours red, green and blue and
one lepton doublet, giving in total 12 doublets of left-handed Weyl fermions in
the fundamental representation. We also have right-handed quarks and fermions
in the trivial representation that have T(R) = 0. In the limit of massless fermions,

Example 9.4.11 In the hypercharge interaction described by the group G = U(1)

Y and coupling constant g′ we get with correctly normalized coupling constant

and charges
in the limit of massless fermions in three generations of the representations

the coefficient

9.4.3 Experimental Values for Coupling Constants

We discuss some experimental values of coupling constants (following [108,
109]). At the energy scale

where m Z is the mass of the Z 0-boson, the electric fine-structure constant, the
strong fine-structure constant and the Weinberg angle are given by (with respect
to the minimal subtraction renormalization scheme )

The coupling constants g are related to the fine-structure constants by .

Together with

the coupling constants corresponding to the factors of the Standard Model gauge
group SU(3) C × SU(2) L × U(1) Y are seen to be:

Since

it is conceivable that there is some high energy μ m Z where the coupling

constants g s , g w , g′ become equal (to make this argument precise one has to
rescale g′ suitably; see Sect. 9.5.3). This leads to the idea of Grand Unification
that we discuss in Sect. 9.5.
9.4.4 Further Reading
The books [80] by T.J. Hollowood, [125] by M.D. Schwartz and [144] by
S. Weinberg discuss vacuum polarization. The section by A. Masiero in the book
[85] and Sect. 5.5 in the book [96] by R.N. Mohapatra also have short accounts
of running coupling constants in connection with Grand Unification.

9.5 Grand Unified Theories

The gauge group of the Standard Model is the compact Lie group

The idea of Grand Unification is to unify all forces described by the Standard
Model into a simple Lie group with only a single coupling constant . Even
though the coupling constants of the strong, weak and electromagnetic
interaction are different at energies around 100 GeV, the existence of such a
unified theory is not impossible from the outset, because of the quantum effect
of running coupling discussed in Sect. 9.4 (this was first realized by H. Georgi,
H.R. Quinn and S. Weinberg in 1974 [61]). Strictly speaking, the electroweak
interaction, described by the gauge group SU(2) L × U(1) Y , is not a unification
of the weak and electromagnetic interaction in this sense, because it still
involves two coupling constants.
After some general remarks, we want to study in this section the Grand
Unified Theories described by the simple Lie groups SU(5) and Spin(10). We
follow the mathematical reference [9] and the physics references [85] and [96]
throughout this section.

9.5.1 Group Theoretic Preliminaries

Definition 9.5.1 We call a Lie group G a possible Grand Unification group if
it has the following properties:
G is simple, so that it has only one coupling constant, or G is a product of
several copies of the same simple group, where the coupling constant for
each factor is set the same (by a discrete symmetry).
The Lie group G contains (a finite quotient of) the Standard Model group
SU(3) × SU(2) × U(1).
G admits complex representations not isomorphic to their complex
conjugates; see Sect. 8.5.3.
We would like to find all possible Grand Unification groups. It is more suitable
to classify these groups according to rank (the maximal dimension of an
embedded torus subgroup) than according to dimension. The Standard Model
group SU(3) × SU(2) × U(1) has rank 4. A possible Grand Unification group
therefore must have rank at least 4. We restrict to the three simplest cases where
G has rank 4, 5 or 6.
Using the Killing–Cartan Theorem 2.4.23 on the classification of compact
simple Lie algebras we can list all simple (and simply connected) compact Lie
groups of rank less than or equal to 6:
rank 1: SU(2)
rank 2: SU(3), Spin(5), G2
rank 3: SU(4), Spin(7), Sp(3)
rank 4: SU(5), Spin(8), Spin(9), Sp(4), F4
rank 5: SU(6), Spin(10), Spin(11), Sp(5)
rank 6: SU(7), Spin(12), Spin(13), Sp(6), E6.
The semisimple Lie groups of rank 4 with a single coupling constant are
therefore (see the first part of Exercise 2.7.18)

(an exponent k denotes the product of k copies of the group). The semisimple
Lie groups of rank 5 with a single coupling constant are

The semisimple Lie groups of rank 6 with a single coupling constant are

According to [104, Sect. 5.4] the only compact, simply connected, simple
Lie groups which have representations not isomorphic to their complex
conjugates are
This implies:

Proposition 9.5.2 The only possible Grand Unification groups of rank less
than or equal to 6 are:
rank 4: SU(3)2 and SU(5)
rank 5: SU(6) and Spin(10)
rank 6: SU(3)3 , SU(4)2 , SU(7) and E6.

Here are some references where actual Grand Unified Theories using these
groups (except SU(3)2 and SU(4)2) have been constructed:
rank 4:
– SU(5): the Georgi–Glashow theory [59] with gauge group SU(5)
from 1974 was the first Grand Unified Theory based on a simple Lie
group. In the same paper the Lie group SU(3)2 is ruled out for
physical reasons, leaving SU(5) as the only GUT group of rank 4.
rank 5:
– Spin(10): the SO(10) theory, as it is called in physics, was first
developed by H. Georgi [58] and H. Fritzsch and P. Minkowski [56]
in 1975.
– SU(6): there is a theory of A. Hartanto and L.T. Handoko [72] from
2005.
rank 6:
– E6: a Grand Unified Theory with this gauge group arises naturally in
heterotic string theory and was first developed by F. Gürsey,
P. Ramond and P. Sikivie [68] in 1976.
– SU(3)3: there is a theory called Trinification with this gauge group
proposed by A. de Rújula, H. Georgi and S.L. Glashow [63] in 1984.
– SU(7): a Grand Unified Theory based on this Lie group was studied
by K. Yamamoto [151] in 1981.
 Our aim is to discuss in some detail the Grand Unified Theories
corresponding to the simple Lie groups SU(5) and Spin(10) (we only consider
Grand Unified Theories defined on 4-dimensional Minkowski spacetime).

9.5.2 Embeddings of the Standard Model Gauge Group

into the Simple Lie Groups SU(5) and Spin(10)
We describe how the gauge group of the Standard Model (actually a quotient
of it) can be embedded into the simple Lie groups SU(5) and Spin(10) (we
follow [9]).

Definition 9.5.3 For integers m, n ≥ 1 we set

The Lie group S(U(m) × U(n)) is naturally a subgroup of SU(m + n) under the
embedding

We set

Proposition 9.5.4 (Embedding of into SU(5)) The Lie group

homomorphism

induces an injective Lie group embedding

with image

Here is the subgroup

Proof It is clear that f is a Lie group homomorphism to U(5). The image is

contained in SU(5), since

We show that f is surjective onto S(U(3) × U(2)): let

There exists a complex number α U(1) such that Aα 2 SU(3). Consider the
matrix Bα −3. Then

Hence Bα −3 SU(2) and

We show that the kernel of f is K: suppose

Then

Since g SU(3) and h SU(2) we have

This proves the claim. □

Proposition 9.5.5 For all n ≥ 2 there exists a canonical Lie group embedding

Proof According to Exercise 1.9.10 there exists an embedding

and thus
Since SU(n) is simply connected and Spin(2n) → SO(2n) is a covering map, this
embedding can be lifted to an embedding

Together with Proposition 9.5.4 we get:

Corollary 9.5.6 (Cascade of Grand Unification Groups) There exists a

sequence of Lie group embeddings

Under the first embedding gets identified with S(U(3) × U(2)).

Together with an embedding of Spin(10) into the exceptional simple Lie

group E6 (see [8, 46]), these groups form the famous cascade of simple Grand
Unification groups down to the quotient of the Standard Model group:

In the physics literature (e.g. [126]) the Lie group Spin(10) is sometimes
called E5 and the Lie group SU(5) is called E4.

Remark 9.5.7 The Lie group Spin(10) actually contains the larger compact
embedded Lie group

In the physics literature on Grand Unification the group is often denoted

by SU(5) × U(1). See Exercise 9.7.4 for more details.

9.5.3 Normalized Hypercharge and Unification of

Coupling Constants
We saw in Sect. 9.4 that, due to the quantum effect of running couplings , there is
a possibility that the coupling constants of the strong and electroweak
interactions become the same at some high energy. We discuss this in more detail
(we follow [85] and [96]).
In this subsection we consider a simple Grand Unification group G. Under
the embedding of into G, all basis vectors of G SM must be normalized,
because they belong to an orthonormal basis of the Lie algebra . This means
that with respect to the Killing form of , the basis vectors

should have the same length, where λ > 0 is a normalization constant to be

determined. We first consider the case G = SU(5) with the embedding given by
Proposition 9.5.4. Under this embedding

According to Exercise 2.7.16 the Killing form of is given by

hence

A similar argument shows that the basis of already has the correct
normalization. This implies:

Proposition 9.5.8 The correctly normalized coupling constants of G SM under

the embedding of into SU(5) are given by

For a general simple Grand Unification gauge group G we can argue as follows:
let ρ denote the representation of G SM on one generation of left-handed
fermions and antifermions (see Eq. (8.13)):

The vectors β 3 ′ , β 4 ′ act under ρ as

The symmetric bilinear form

is Ad G -invariant, hence by Theorem 2.5.1 it must be a multiple of the Killing

form. We have

Setting

we conclude again that

From Sect. 9.4 the correctly normalized running coupling constants are given by

With the approximate values

at m Z ≈ 91 GeV we get (see also Fig. 9.3)

We see that the running coupling constants do not exactly match at the same
energy, but as a first approximation this is roughly true. In supersymmetric
Grand Unified Theories (where more particles take part in loop diagrams) the
running coupling constants become equal to a much better approximation.

Fig. 9.3 Approximate unification of running fine-structure constants α = g 2⁄(4π)

At an energy where g w = g″, we get for the Weinberg angle θ W the

prediction

hence

The experimental value for the Weinberg angle at the much lower energy m Z
is

The difference could be explained with a running coupling effect,

i.e. quantum corrections.

9.5.4 The Fermions in the SU(5) Grand Unified Theory

We discuss two specific representations of SU(5) and understand how they
decompose under restriction to the subgroup (we follow [9] and [96]).

Lemma 9.5.9 Consider the complex conjugate of the fundamental

representation of SU(5). Under the restriction to the subgroup this
representation has the branching rule

Proof Consider the Lie group homomorphism

and decompose the column vectors in into the first three components and
the last two components . On the complex conjugate of the fundamental
representation an element (g, h, α) G SM acts on the first three components
as

and on the last two components as

According to Lemma 8.5.1 the representation

has y = 2⁄3 and the representation

has y = −1. Therefore the representation splits as

Note that for the fundamental SU(2) representation by Exercise 2.7.3.

Lemma 9.5.10 Consider the representation of SU(5). Under the

restriction to the subgroup this representation has the branching rule

Proof We identify with the complex 5 × 5 antisymmetric matrices X.

Then M SU(5) acts on X by

We write X as

with

We consider the Lie group homomorphism

Then (g, h, α) acts on X by

We can directly read off the hypercharges:

It is clear that the SU(3) × SU(2) representation on B is (3,2) and on the 1-

dimensional space C it can only be (1,1). Exercise 9.7.5 shows that the
representation on A is .□

Comparing with Eq. (8.13) we have:

Theorem 9.5.11 (Fermions in SU(5) Grand Unification) The

representation of SU(5) under restriction to the subgroup has
the branching rule

The representation on the right is isomorphic to the representation V L i VL

iC of one full generation of left-handed fermions and antifermions. The

explicit identification is given by

The particle fields which appear in these formulas are the current eigenstates
that correspond to the weak eigenstates in the Standard Model.

Remark 9.5.12 We can add a left-handed sterile antineutrino in the trivial

singlet representation of SU(5). Then the left-handed fermions and
antifermions form a 16-dimensional representation

All fermions and antifermions (left-handed and right-handed) together then form
the 32-dimensional representation

It is very remarkable that the very simple, decomposable SU(5)-representation

thus suffices to describe all fermions. This discussion and more details can
be found in [9].
9.5.5 The Fermions in the Spin(10) Grand Unified
Theory
The idea of the Grand Unified Theory with Lie group Spin(10) is to look for a
representation of Spin(10) that under restriction to the subgroup SU(5)
decomposes into the representation . As we saw in Sect. 9.5.4 this
representation then decomposes under restriction to into the Standard
Model representation.
The most natural representation of the Lie group Spin(10) is the spinor
representation. The Weyl spinor representations of Spin(10) have dimension 16.
We want to show that these representations can accommodate the 16-
dimensional representations of one full generation of fermions and antifermions,
including a sterile neutrino, mentioned in Remark 9.5.12 (we follow in this
subsection [9]).
We consider more generally for an arbitrary integer n ≥ 2 the Lie group
Spin(2n). Recall from Sect. 6.4 and Sect. 6.5 that this spin group has a Dirac
spinor representation on which is reducible and decomposes into the
Weyl spinor representations

where Δ ± each have half dimension. Note that

as complex vector spaces. In fact, we can realize the spinor representation on

Δ explicitly on :

Lemma 9.5.13 (Dirac Spinor Representation of Spin(2n) on ) For

let

and

Then the complex Dirac spinor representation of on is induced

by the following Clifford multiplication of vectors
on the exterior algebra :

Proof The original proof is due to [7]. It is easy to check that δ satisfies

By the universal property of Clifford algebras there is an induced non-trivial

algebra homomorphism

Both algebras are isomorphic to . The kernel of ϕ is a two-sided ideal in

and hence, by a general property of endomorphism algebras,

either 0 or . Since ϕ is non-trivial, it follows that ϕ is injective and hence
an isomorphism. □

Remark 9.5.14 It can be shown that the subspaces of Weyl spinors correspond
to the forms of even and odd degree, where the precise association depends on
the integer n; see Exercise 9.7.6.

There is also a natural representation of SU(n) on induced by the

fundamental representation on . Consider the Lie group embedding

from Proposition 9.5.5.

Theorem 9.5.15 The Dirac spinor representation Δ of Spin(2n) has under

restriction to the subgroup SU(n) the following branching rule:

In particular, if the Dirac spinor representation of Spin(2n) on (which

preserves only the parity of the degree) is restricted to SU(n), it preserves the
integral degree of all forms.

Proof We need to find an explicit embedding of SU(n) into Spin(2n). It suffices

to do this on the level of Lie algebras. According to Exercise 1.9.10 the
embedding
can be realized by

where A 1, A 2 are real n × n matrices with

Let e 1, …, e 2n be the standard basis of and let E rs denote the elementary 2n

× 2n-matrix with a 1 at the intersection of the r-th row and s-th column and zeros
elsewhere. We set

Applied to a standard basis vector e j we get

Then we can expand

This follows because the sums of the ε-matrices in this equation look like
(indices on 1 and − 1 are matrix indices of the entry)
 Recall from Sect. 6.5.3 that the Lie algebra of Spin(2n) is given by

and the isomorphism

maps

It follows that

To apply Lemma 9.5.13 we take for r ≤ n the basis vector e r in the first -
summand and e r+n in the second summand. Then a calculation shows that for a
1-form and 1 ≤ r < s ≤ n

hence
where we used that A 1 is skew-symmetric.
Similarly

This implies

since A 2 is symmetric and trA 2 = 0. We conclude that

On the left we have the restriction of the spinor representation to SU(n) and
on the right the fundamental representation of SU(n). So far we have proved the
claim on , but it is possible to conclude from that the full claim on forms of
arbitrary degree. The details are left as Exercise 9.7.7. □

Theorem 9.5.15 and Remark 9.5.14 imply:

Corollary 9.5.16 (Fermions in Spin(10) Grand Unification) The Dirac

spinor representation 32 of Spin(10) has under restriction to the subgroup
SU(5) the branching rule

The Weyl spinor representations 16 and of Spin(10) have under restriction

to SU(5) the branching rules

From Remark 9.5.12 it follows that the Weyl spinor representation 16 of

Spin(10) can accommodate one full left-handed generation of the Standard
Model and the Weyl spinor representation one full right-handed
generation, including sterile neutrinos . We get for the first left-handed
generation a decomposition of the form
 Note that the spinor representations of Spin(10) force us to include a sterile
fermion, which can be interpreted as a sterile neutrino. In the case of SU(5)
we still had the choice whether we want to add a trivial 1-dimensional
representation 1. In some sense [111] the unification provided by the Spin(10)
theory is more complete than for the SU(5) theory, because all (left-handed)
fermions of one generation belong to the single, irreducible representation 16.

Remark 9.5.17 See the remark after Exercise 9.7.9 for the U(1) charges under
restriction of the spinor representation of Spin(10) to the subgroup .

9.5.6 The Fermions in the E6 Grand Unified Theory

We briefly discuss (without proofs) the Grand Unification group E6 (see [87] for
more details). The Lie group E6 does not have a non-trivial 15- or 16-
dimensional representation, but it has a 27-dimensional representation 27. Under
restriction to the subgroup Spin(10) this representation and its complex
conjugate have the branching rules

Here 16 is the spinor representation and 10 the vector representation of Spin(10).

As we saw in Sect. 9.5.5, the 16 can accommodate one left-handed generation of
the Standard Model, including a sterile neutrino. Similarly the can
accommodate one right-handed generation. The representations 10, and 1
correspond to new fermions that so far have not been detected.

9.5.7 Gauge Anomalies

It can be proved that the GUTs constructed above with gauge groups SU(5) and
Spin(10) (and E6) are anomaly free. See [85] for details.

9.5.8 The Gauge Bosons in the SU(5) Grand Unified

Theory
We want to study the gauge bosons in the SU(5) Grand Unified Theory (we
follow [85]). We first discuss the embedding of the Lie algebra into .
According to the calculation in Sect. 9.5.3 a correctly normalized basis for is
given by

Here λ a are the Gell-Mann matrices and σ b the Pauli matrices . This implies:

Lemma 9.5.18 The embedding

is given by mapping

to

We conclude the following:

Theorem 9.5.19 (The SU(5) Gauge Field) The SU(5) gauge field A μ with
values in is given by

Here g SU(5) is the coupling constant of SU(5) and X j μ , Y j μ are new gauge
bosons, corresponding to new forces not present in the Standard Model. We thus
get 6 complex (12 real) additional gauge bosons. This is clear, because dimSU(5)
= 24 and dimG SM = 12.
From Remark 8.8.9 we expect that the off-diagonal X- and Y -bosons could
lead to interesting effects. To understand this in more detail we calculate the
relevant part of the Dirac Lagrangian (the same calculation, up to a different
choice of signs, can be found in [85]).

Theorem 9.5.20 (X- and Y -Boson Interaction Vertex) Consider the left-
handed interaction term in the Dirac Lagrangian , given by

The part of this term involving X- and Y -bosons can be calculated for the first
fermion generation as

with the currents

In each current the first two terms come from the representation 10 and the third
term from the representation . In the first summand there is a sum over indices
j, k. Quark indices 1, 2, 3 correspond to r, g, b. There are corresponding terms
for the second and third generation.

Feynman diagrams for these interactions are depicted in Figs. 9.4, 9.5, 9.6, 9.7 .

Fig. 9.4 X-boson: quark-lepton vertices

Fig. 9.5 X-boson: quark vertices

Fig. 9.6 Y -boson: quark-lepton vertices

Fig. 9.7 Y -boson: quark vertices

Fig. 9.8 X-boson mediated proton decay

In the interaction vertices in Figs. 9.4 and 9.6, an X- or Y -boson pairs a

lepton and a quark, hence the X- and Y -bosons are examples of so-called
(vector) leptoquarks (there are theories beyond the Standard Model that also
predict scalar leptoquarks). The corresponding processes can convert a quark
into a lepton or vice versa, hence baryon and lepton numbers are not conserved
(if the X- and Y -bosons are not assigned both a baryon and lepton number).
In particular, the following process is possible (see Fig. 9.8): two up quarks
annihilate in an X-boson, that then decays into a positron and a down antiquark.
This leads to one of the most famous predictions of Grand Unified Theories:
proton decay:1
(9.7)
The GUT scale of ∼ 1015 GeV is not directly accessible in experiments and
totally out of the reach of present day colliders. However, proton decay as a
signature of Grand Unification may in principle be observable, it is just
extremely rare. This is related to the extremely heavy mass of the X-boson,
making the production of a virtual X-boson via uu → X very unlikely.
The predicted lifetime of the proton in the SU(5) GUT is 1030 − 1031 years.
Experiments show that the lifetime of the proton for the decay into a pion and
positron in Eq. (9.7) is longer than 8. 2 ⋅ 1033 years [106], hence the minimal
version of the SU(5) GUT is already ruled out. In supersymmetric Grand Unified
Theories the lifetime of the proton becomes longer and can reach, depending on
the model, 1034 − 1036 years or more (see [29] for a nice overview).
A similar discussion applies to the gauge bosons in the Spin(10) theory.
Since dimSpin(10) = 45, we get 33 additional gauge bosons compared to the
Standard Model; see [85] for more details.

9.5.9 Symmetry Breaking and the Higgs Mechanism

in the SU(5) Grand Unified Theory
We briefly discuss symmetry breaking in the SU(5) Grand Unified Theory
(following [85] and [96]). In the Standard Model, the gauge group

is broken to

where according to Eq. (8.4) the unbroken electromagnetic group U(1) Q is the
subgroup

The discrete subgroup of elements of the form

is actually a subgroup of G CQ (since α −3 = α 3) and we get an embedding

The embedding of into SU(5) is induced by the homomorphism

(9.8)

which is the restriction of f: G SM → SU(5) from Proposition 9.5.4 if we

identify β U(1) Q with

Symmetry breaking in the SU(5) GUT occurs in two steps:

The first step

can be realized by a Higgs field ϕ with values in

carrying the adjoint representation of SU(5). To understand this, consider a
Higgs condensate of the form

where a, b, V ′ are real numbers. Since , we must have

Setting V = aV ′ this implies that

(9.9)

with a real constant V.

Proposition 9.5.21 (Symmetry Breaking from SU(5) to ) The

isotropy group of the vector in Eq. (9.9) (with V ≠ 0) under the
adjoint representation is the subgroup

Proof This is Exercise 9.7.10. □

The second step

can be realized by a Higgs field H with values in

carrying the fundamental representation of SU(5). Consider a Higgs condensate

of the form
 (9.10)

where ρ is a real number. We then have:

Proposition 9.5.22 (Symmetry Breaking from to ) The

isotropy group of the vector in Eq. (9.10) (with ρ ≠ 0)
under the fundamental representation is the subgroup

Proof This is immediate from Eq. (9.8). □

In the model discussed so far, the potential for the Higgs fields is the sum V (ϕ) +
V (H) of potentials for both fields. For phenomenological reasons it turns out that
crossterms of the form V (ϕ, H) have to be included. The minimum ϕ 0 then takes
the form

and thus also breaks SU(2) L .

With the correct Higgs fields determined, it is then possible to calculate the
masses of all broken gauge bosons. Finally, suitable Yukawa couplings can be
introduced to give masses to the fermions and there is a generalized version of
fermion mixing. A similar discussion applies to symmetry breaking in the
Spin(10) GUT. Details for these constructions and their physical consequences
can be found in [85] and [96].

9.5.10 Further Reading

The article [9] of J. Baez and J. Huerta from 2010 is a mathematical introduction
to the representations of the Standard Model and Grand Unified Theories. The
review articles [87] by P. Langacker and [131] by R. Slansky summarize the
state of the art in Grand Unification in 1981. The book [85] by C. Kounnas et
al. from 1984 has a very readable account of the Standard Model and Grand
Unification and the book [96] by R.N. Mohapatra from 2003 contains details on
Grand Unification and supersymmetry. Shorter expositions of GUTs can be
found in the paper [148] by E. Witten from 2002 and the gauge theory books
[10] by D. Bailin and A. Love (1993) and [32] by M. Chaichian and N.F. Nelipa
(1984).

9.6 A Short Introduction to the Minimal Supersymmetric

Standard Model (MSSM)
In this section we give a very brief introduction to supersymmetry. In addition to
Lorentz invariance and gauge symmetry, Lagrangians (or actions) of field
theories can be supersymmetric. The Standard Model itself is not
supersymmetric, but can be extended to a (Minimal) Supersymmetric Standard
Model by adding superpartners to the known particles. Some of these
superpartners are potential dark matter candidates. Supersymmetry also offers a
solution to the so-called hierarchy problem concerning the unnatural smallness
of the Higgs boson mass. Furthermore, Grand Unification becomes more
realistic (for example, the predicted decay rate of the proton and the unification
of coupling constants) under the assumption of supersymmetry.
These are some of the reasons to believe that supersymmetry is realized (as a
broken symmetry) in nature. In addition to yielding extensions of the Standard
Model, supersymmetric field theories are also theoretically interesting, because
they are often better understood non-perturbatively (at strong coupling) than
non-supersymmetric field theories.

9.6.1 Graded Lie Algebras and the Supersymmetry

Algebra
In this subsection we define graded Lie algebras and the supersymmetry algebra
(we follow [53]).

Definition 9.6.1 A graded Lie algebra is an algebra, i.e. a vector space L = L

0 L 1 with a bilinear product

that has the following properties for all x i L i,x j L j:

1. Grading:
2. Supersymmetry:

3. Super Jacobi identity:

The vector subspace L 0 is called the even part and L 1 the odd part.

It follows that the product ⋅ is

antisymmetric on L 0 × L 0 and maps to L 0 (written as [⋅ ⋅ ])
symmetric on L 1 × L 1 and maps to L 0 (written as {⋅ ⋅ })
antisymmetric on L 1 × L 0 and L 0 × L 1 and maps to L 1 (written as [⋅ ⋅ ]).
The notation [⋅ ⋅ ] and {⋅ ⋅ } is just a different notation for the product ⋅ on
the algebra L. On two general elements in L, which have components in both L 0
and L 1, the product is neither symmetric nor antisymmetric and sometimes
written as {[⋅ ⋅ ]}. The following is easy to show:

Proposition 9.6.2 (Characterization of Graded Lie Algebras) Let (L, {[⋅

⋅ ]}) be a graded Lie algebra.

1. The vector subspace L 0 with the product [⋅ ⋅ ] is a Lie algebra.

2. The map

with

is a representation of the Lie algebra (L 0, [⋅ ⋅ ]) on the vector space L 1.

3. The map

is a vector space-valued symmetric bilinear form.

 Conversely, a Lie algebra (L 0, [⋅ ⋅ ]) together with a representation ϕ of L 0
on a vector space L 1 and a symmetric bilinear form {⋅ ⋅ } on L 1 with values in L
0 define a graded Lie algebra if the following identities are satisfied:

(9.11)
and
(9.12)
Proof This is Exercise 9.7.11. □

The supersymmetry algebra that we now discuss is a special graded Lie algebra.
We first define the Poincaré algebra : let (V, η) be Minkowski spacetime
of dimension n + 1 with Minkowski metric η. Let denote
the Lie algebra of the proper orthochronous Lorentz group SO+(V ) . Elements of
are denoted in physics by M. They correspond to infinitesimal spacetime
rotations.

Definition 9.6.3 The Poincaré algebra is a real Lie algebra with underlying
vector space

also denoted by . Elements of the subspace V are denoted in physics

by P. The commutator on is a semi-direct product:

1. The commutator on the subspace is the standard one.

2. The commutator on the subspace V vanishes:

3. Consider the vector representation

The commutator between elements of and V maps to V and is defined

by
It is easy to check that is indeed a Lie algebra (in the standard sense). In
physics the commutators are often written with respect to a basis for the Lie
algebra. The elements of the Poincaré algebra correspond to infinitesimal
spacetime rotations and translations.
The Poincaré algebra can be extended to the super-Poincaré algebra by
adding a spinor representation.

Definition 9.6.4 The N = 1 super-Poincaré algebra or supersymmetry

algebra is a graded real Lie algebra with underlying vector space

also denoted by . Here S is the dual of a real spinor

representation space of minimal dimension of . Elements of S
are denoted in physics by Q. The multiplication {[⋅ ⋅ ]} on is in
some sense a graded semi-direct product defined as follows:

1. The multiplication on the even subspace is the standard

commutator in the Poincaré algebra.

2. Consider the dual spinor representation

The multiplication between and the odd subspace S is defined

by

3. The multiplication between V and S vanishes:

4. It can be shown that there exists a symmetric bilinear form

that is Lorentz equivariant, i.e. satisfies the identity

 (9.13)
for all . The multiplication on S maps to
and is defined by

It is easy to check using Proposition 9.6.2 that is a graded Lie algebra .

Equation (9.11) is satisfied, because V acts trivially on S , and Eq. (9.12)
corresponds to Eq. 9.13. Note that in the super-Poincaré algebra the product

only maps to the subspace V of . Again, in the physics literature the

multiplication on is usually written with respect to a basis. The elements
are called (infinitesimal) supersymmetries. The
supersymmetry algebra thus consists of infinitesimal spacetime
rotations, translations and supersymmetries.
The super-Poincaré algebra for N ≥ 2 supersymmetries is obtained similarly
from the Poincaré algebra by adding N copies of S . In this situation there is
more flexibility in the definition of the symmetric bilinear form Γ leading to the
concept of central charges . We will restrict to the case N = 1.

9.6.2 Supersymmetric Field Theories

Field theories on Minkowski spacetime are usually assumed to be Poincaré
invariant. This means that there is a representation of the Poincaré algebra
on the fields and the action of the theory is invariant under this algebra.
Supersymmetric field theories on Minkowski spacetime are field theories
where the fields have a representation of the super-Poincaré algebra
leaving the action invariant. We only consider the case of rigid
supersymmetries where the spinors Q S generating the supersymmetries
are constant (parallel with respect to the canonical spin covariant derivative) on
Minkowski spacetime (the infinitesimal rotations and translations in
are constant as well). Just as the infinitesimal isometries in the Poincaré algebra
are defined by certain vector fields on spacetime, infinitesimal supersymmetries
are defined by (constant) spinors.
Theories with an action invariant under arbitrary supersymmetries (that are
local , i.e. not necessarily constant) are called supergravities . Since the
anticommutator of two supersymmetries is basically a vector (translation),
theories which are invariant under spacetime-dependent supersymmetries are
invariant under the action of all vector fields and hence under all infinitesimal
diffeomorphisms. Supergravities are therefore automatically diffeomorphism
invariant (at least under diffeomorphisms that can be connected to the identity),
i.e. theories of gravity.
This means that spinors can appear in field theories in two different ways:
(twisted) spinors describe matter particles, i.e. (charged) fermions
pure (untwisted) spinors can be infinitesimal generators of supersymmetries
.
This can be compared to vector fields (or 1-forms) in field theories:
vector fields with values in the adjoint bundle describe gauge bosons
pure vector fields (sections of the tangent bundle) describe infinitesimal
diffeomorphisms or isometries of the manifold.
Since field theories like the Standard Model involve charged fermions, it
seems from this point of view natural to extend the bosonic symmetries of
spacetime (isometries or diffeomorphisms) by fermionic symmetries (rigid or
local supersymmetries). According to the Haag–Łopuszański–Sohnius Theorem
the only consistent way to do so is by using the super-Poincaré algebra.
One of the interesting things about supersymmetric field theories is that the
particle content and the interactions that can occur in the Lagrangian are very
restricted. Representations of the Poincaré algebra are given by fields of a
certain spin. The elements of the Poincaré algebra map fields of one spin to
fields of the same spin. Supersymmetries, on the other hand, map a field of one
spin to a field of another spin. This means that fields of different spin are
combined in supersymmetry multiplets and the elements of the super-Poincaré
algebra map a multiplet of one type to a multiplet of the same type.
Supersymmetries do not preserve the spin of a field, but they do preserve the
type of the supersymmetry multiplet. Only complete supersymmetry multiplets,
not just some of their components, can be added to supersymmetric field
theories.
For N = 1 supersymmetry in D = 4 Minkowski spacetime there are only
two types of supersymmetry multiplets with fields of spin at most 1 (i.e. that do
not contain a graviton):
vector multiplet (or gauge multiplet ) consisting of a vector (spin 1) and a
Majorana spinor (spin )

chiral multiplet consisting of a Weyl spinor (spin ) and a complex scalar

(spin 0).
Recall that the particles in the Standard Model are of three types:
vectors (gauge bosons)
fermions (quarks and leptons)
scalars (Higgs field and Higgs boson).
It turns out that the particles in the Standard Model (vectors–fermions and
fermions–scalars) do not themselves combine to form supersymmetry multiplets.
For example, both fields in a vector multiplet have to be associated to the same
representation of the gauge group (i.e. to the adjoint representation). Similarly,
both fields in a chiral multiplet have to be associated to the same gauge
representation.
If we want to extend the Standard Model to a Minimal N = 1
Supersymmetric Standard Model (MSSM) , it follows that we have to add
for each particle in the Standard Model a superpartner to form either a
vector multiplet or a chiral multiplet (see Table 9.2):

for the gauge bosons (W-bosons , B-boson , Z-boson , photon , gluons

) we add Majorana fermions called gauginos (winos , bino , zino ,
photino , gluinos ) to form vector multiplets ;

for the left-handed and right-handed fermions (quarks and leptons)

we add complex scalars (squarks , sleptons ) to form chiral multiplets
;

for the scalars (Higgs field) we add Weyl fermions (Higgsinos ) to

form chiral multiplets . In addition to the Higgs isodoublet from the
Standard Model with Y = 1, now denoted by ϕ u , we need another
Higgs isodoublet with Y = −1, denoted by ϕ d , to ensure cancellation
 of gauge anomalies in the MSSM.

Using these multiplets it is possible to write down Lagrangians such that the
actions are Poincaré invariant and invariant under supersymmetries generated by
spinors Q S . The symmetries together satisfy the relations of the super-
Poincaré algebra (sometimes only on-shell , i.e. if the fields satisfy the equations
of motion).
Table 9.2 Particle content of MSSM (fermions and sfermions repeat in 3 generations)
Standard Model particles Superpartners
Name Field Spin Name Field Spin
Quarks Squarks 0

uR,dR , 0

Leptons Sleptons 0

eR 0

Higgs 0 Higgsinos

Gluons G 1 Gluinos

W-bosons W ± 1 Winos

W0 1

B-boson B 1 Bino

Z-boson Z0 1 Zino

Photon γ 1 Photino
 Supersymmetry predicts that all particles in a supersymmetry multiplet have
the same mass. Since the superpartners of the particles in the Standard Model so
far have not been observed, supersymmetry (if it exists) must be broken in
nature and the superpartners must be heavier than the known particles.
The Lagrangian of the MSSM itself is very restricted by demanding
supersymmetry. If we want to introduce supersymmetry breaking, however, the
Lagrangian becomes much more complicated and involves many additional
terms. The precise mechanism of supersymmetry breaking is still under
discussion.
Supersymmetry can be combined with Grand Unification and yields, for
instance, N = 1 supersymmetric extensions of the SU(5) and Spin(10) Grand
Unified Theories . Supersymmetric GUTs can also be derived naturally from
superstring and M-theory .

9.6.3 Further Reading

The book [53] by D.S. Freed contains a mathematical discussion of
supersymmetry. The classical and quantum theory of supersymmetric field
theories are covered in the lecture notes [4] by P.C. Argyres. The paper [91] by
S.P. Martin is an extensive and readable introduction to supersymmetry and the
MSSM. The book [96] by R.N. Mohapatra discusses, among other things, the
MSSM and supersymmetric Grand Unification.

9.7 Exercises for Chap. 9

9.7.1 Prove that any unitary 2 × 2-matrix V can be brought into real form by the
rephasings in Eq. (9.5).
9.7.2 Prove Proposition 9.3.6 on the invariance of Lagrangians under parity
inversion.
9.7.3 Prove the remaining cases of Proposition 9.3.7 on the invariance of
Lagrangians under charge conjugation.
9.7.4 (from [ 55 ]) We consider the Lie group U(n) and the homomorphism

On the subgroup SU(2n) the homomorphism σ is the identity. Let ι: U(n) ↪

SO(2n) be the standard embedding. Suppose that n is odd.

1. Use covering theory to prove that the composition ι ∘σ lifts to a

 homomorphism

2. Determine the kernel of σ and the kernel of ϕ. Prove that there is a sequence
of Lie group embeddings

9.7.5 Prove the following statement in Lemma 9.5.10: the representation of

SU(3) on 3 × 3-antisymmetric matrices A, given by

is isomorphic to .
9.7.6 Recall the identification of the spinor representation Δ of Spin(2n) with
a representation on from Lemma 9.5.13. Determine in which dimensions n
the left-handed (positive) Weyl spinor space Δ + corresponds under this
identification to the subspace or of forms of even and odd degree.
9.7.7 Prove the statement in Theorem 9.5.15 in the remaining case for forms
of arbitrary degree in (compare with [9]).
9.7.8 Suppose that n is odd. Consider the embedding

from Exercise 9.7.4 and the Dirac spinor representation of

Spin(2n). Let X be the element

Under the homomorphism σ this element maps to

which is then embedded into as in the proof of Theorem 9.5.15. We

want to determine how this element acts on some of the summands of .
Prove that X acts
on by multiplication with − ni

on by multiplication with − (n − 2)i and

 on by multiplication with − (n − 4)i.

9.7.9 Let n ≥ 2 be an arbitrary integer and consider the homomorphism

from Exercise 9.7.4. Recall from Exercise 2.7.6 that the complex
fundamental representation decomposes under restriction to U(n) into
, where W is the complex fundamental representation of U(n).
Let X be the element in from Exercise 9.7.8 and

Show that X acts on W by multiplication with 2i and on by multiplication

with − 2i.

Remark In the case n = 5, taking U(1)-charges with respect to the basis vector
X, it follows that the left-handed Weyl spinor and the vector representation of
Spin(10) decompose under restriction to as

Compare with [131, Table 43].

9.7.10 Prove Proposition 9.5.21 on the isotropy group of under the

adjoint representation.
9.7.11 Prove Proposition 9.6.2 on graded Lie algebras.

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html
5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
 [MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]
20.
Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
 Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)
36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
 [MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
 177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
 Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]
101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]
117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
131. [MathSciNet]
132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149.
 Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions. Chin.
Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/

Footnotes
1 This is a decay of isolated protons, not to be confused with the weak β +-decay that can only occur for
protons bound in certain nuclei.
Part III
Appendix
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_10

Appendix A: Background on Differentiable

Manifolds
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

From a mathematical point of view, gauge theories are described by a spacetime

M together with certain fibre bundles (principal bundles, associated vector
bundles, spinor bundles) over M. Spacetime and fibre bundles are assumed to
have the structure of differentiable manifolds. Differentiable manifolds in turn
are certain topological spaces that essentially have the property of being locally
Euclidean, i.e. locally look like an open set in some , and that have a
differentiable structure, so that we can define differentiable maps (and their
derivatives), vector fields, differential forms, etc. on them.
We briefly sketch the definitions of these concepts. More details can be
found in any textbook on differentiable manifolds or differential geometry, like
[84] and [142].

A.1 Manifolds
A.1.1 Topological Manifolds
Topological manifolds are topological spaces with certain additional structures.
They are a first step towards differentiable manifolds, which are the main spaces
that we will consider in this book.

Definition A.1.1 An n -dimensional topological manifold , also called a

topological n -manifold, is a topological space M such that:

1. M is locally Euclidean , i.e. locally homeomorphic to . This means that

around every point p M there exists an open neighbourhood U M that
is homeomorphic to some open set (both open sets with the subspace
topology).

2. M is Hausdorff.

3. M has a countable basis for its topology.

The local homeomorphisms (and sometimes the subsets
U) are called charts or local coordinate systems for M. Axiom (a) says that we
can cover the whole manifold M by charts. Note that the dimension n is assumed
to be the same over the whole manifold. Axiom (c) is of a technical nature and
usually can be neglected for our purposes. We often denote an n-manifold by M n
.

Example A.1.2 The simplest topological n-manifold is itself. We can

cover M by one chart , given by the identity.

Example A.1.3 Another example of a topological n-manifold is the n -sphere

M = S n for n ≥ 0. We define

Here | | x | | denotes the Euclidean norm. We endow S n with the subspace

topology of . It follows that S n is Hausdorff, compact and has a countable
basis for its topology.
We thus only have to cover S n by charts that define local homeomorphisms
to . A very useful choice are two charts given by stereographic projection .
We think of as the hyperplane {x n+1 = 0} in . We then project a point x
in U N = S n ∖{N}, where N is the north pole

along the line through N and x onto the hyperplane . It is easy to check
that this defines a map
Similarly projection through the south pole

defines a map on U S = S n ∖{S}, given by

We can check that ϕ N and ϕ S are bijective, continuous and have continuous
inverses. Therefore they are homeomorphisms. They define two charts that cover
S n and hence the n-sphere is shown to be a topological manifold.

A.1.2 Differentiable Structures and Atlases

Suppose we have two topological manifolds M and N and a continuous map f: M
→ N between them. We want to define what it means that f is differentiable. To
do so we first have to define a differentiable (or smooth) structure on both
manifolds.

Definition A.1.4 Let M be a topological n-manifold. Suppose (U, ϕ) and (V, ψ)

are two charts of M. We call these charts compatible if the change of
coordinates (or coordinate transformation ), given by the map

is a smooth diffeomorphism between open subsets of , i.e. the

homeomorphism ψ ∘ϕ −1 and its inverse are infinitely differentiable.

Definition A.1.5 Let be a set of charts that cover M. We call an atlas if

any two charts in are compatible. We call a maximal atlas (or
differentiable structure ) if the following holds: Any chart of M that is
compatible with all charts in belongs to . It can be checked that any given
atlas for M is contained in a unique maximal atlas.

Definition A.1.6 A topological manifold M together with a maximal atlas is

called a differentiable (or smooth) manifold .

Example A.1.7 The topological manifold is a differentiable manifold: We

have one chart , where is the identity. Since we only have a
single chart, there are no non-trivial changes of coordinates. Therefore
forms an atlas that induces a unique differentiable structure on
(the standard differentiable structure).

Example A.1.8 Recall that we defined on the n-sphere S n two charts (U N , ϕ N )

and (U S , ϕ S ). We want to show that these two charts are compatible and hence
form an atlas . This atlas is contained in a unique maximal atlas that defines a
differentiable structure on the n-sphere (the standard structure).
We first have to calculate the inverse of the chart mappings: We have

and

Since

we get

with

A similar calculation shows that

Since these maps are infinitely differentiable, it follows that the charts (U N ,
ϕ N ) and (U S , ϕ S ) are compatible and define a smooth structure on the n-sphere
S n.

Remark A.1.9 In certain dimensions n there exist exotic spheres, which are
differentiable structures on the topological manifold S n not diffeomorphic to the
standard structure. The first examples have been described by Milnor and
Kervaire.
Remark A.1.10 From now we consider only smooth manifolds.

Example A.1.11 It is possible to extend the definition of smooth manifolds to

include manifolds M with boundary ∂M. We usually consider only manifolds
without boundary, even though most concepts in this book also make sense for
manifolds with boundary.

Definition A.1.12 A manifold M is called closed if it is compact and without

boundary.

Definition A.1.13 A manifold M is called oriented if it has an atlas of

charts {(U i , ϕ i )} such that the differential (represented by the
Jacobi matrix) of any change of coordinates has positive determinant at each
point.

A.1.3 Differentiable Mappings

We can now define the notion of a differentiable map between differentiable
manifolds.

Definition A.1.14 Let M m and N n be differentiable manifolds and f: M → N a

continuous map. Let p M be a point and (V, ψ) a chart of N around f( p). Since
f is continuous, there exists a chart (U, ϕ) around p such that f(U) V. We call f
differentiable at p if the map

is infinitely differentiable (in the usual sense) at ϕ( p) as a map between open

subsets of .

Remark A.1.15 The property of a map f being differentiable at a point p does

not depend on the choice of charts, precisely because all changes of coordinates
are diffeomorphisms: if f is differentiable at p for one pair of charts, then it is
also differentiable for all other pairs.

Definition A.1.16 We call a continuous map f: M → N differentiable if it is

differentiable at every p M. We call f a diffeomorphism if it is a
homeomorphism such that f and f −1 are differentiable.

Remark A.1.17 All differentiable maps between manifolds in the following will
be infinitely differentiable (smooth), also called .

Example A.1.18 It is a nice exercise to show that the involution

is a diffeomorphism.

A.1.4 Products of Manifolds

Let M m and N n be differentiable manifolds. Then the Cartesian product X m+n =
M m × N n canonically has the structure of a differentiable manifold of dimension
m + n. We have to define charts for X: Let (U, ϕ) and (V, ψ) be local charts for M
and N. Then (U × V, ϕ ×ψ) is a local chart for X, where

It can easily be checked that with this definition the changes of coordinates are
smooth.

A.1.5 Tangent Space

Suppose M n is a differentiable manifold and p M is a point. An important
notion is that of the tangent space T p M of the manifold at the point p. This is
something that only exists on smooth manifolds and not on topological
manifolds.
How can we define such a tangent space? To get some intuition, we can first
consider the case of a submanifold of some Euclidean space. The
standard definition is that the tangent space in p M is the subspace of
consisting of all tangent vectors to differentiable curves through p:

The problem with general manifolds is that they are a priori not embedded in
any surrounding space, so this notion of tangent vector does not work. However,
what we can do, is that instead of taking the tangent vectors in the surrounding
space, we take the full set of curves through p in the manifold M and define on
this set an equivalence relation that identifies two of them, α and β, if in a chart
they have the same tangent vector in p:
To be equivalent in this sense does not depend on the choice of charts: If we
choose another chart around p, then the tangent vectors in the
charts ϕ and ψ are related by a linear map, the differential D ϕ( p)(ψ ∘ϕ −1) of the
change of coordinates. Since the tangent vectors of α and β in chart ϕ are
identical, they will thus still be identical in chart ψ. With this equivalence
relation we can therefore set:

Definition A.1.19 The tangent space of a smooth manifold M n at a point p

M is defined by

For the equivalence class of the curve γ in M we write

and call this a tangent vector .

Proposition A.1.20 At any point p M n the tangent space T p M has the

structure of a real n-dimensional vector space.

Proof Let be a chart around p. We set

It can be shown that this is a bijection. We define the vector space structure on T
p M so that this map becomes a vector space isomorphism. This structure does
not depend on the choice of chart: If is another chart around p, then
the following diagram is commutative, where D ϕ( p)(ψ ∘ϕ −1) is a vector space
isomorphism:

Hence the identity between T p M and T p M defined with the respective vector
space structures is a vector space isomorphism. □

Definition A.1.21 The set

is called the tangent bundle of M.

In Sect. 4.5 it is shown that the tangent bundle is an example of a vector bundle
over M with fibres T p M.

A.1.6 Differential of a Smooth Map

Let f: M → N be a smooth map between differentiable manifolds. With the
tangent space at hand, we can now define the differential of f.

Definition A.1.22 The differential D p f of the map f at a point p M is

defined by

Equivalently,

The differential is a well-defined (independent of choice of representatives for

[γ]) linear map between the tangent spaces.

For a vector X T p M we sometimes write

The differential satisfies the so-called chain rule.

Proposition A.1.23 The following chain rule holds for the differential: If f: X
→ Y and g: Y → Z are differentiable maps, then g ∘ f is differentiable and at any
point p X

Corollary A.1.24 The differential D p f of a diffeomorphism f: M → N is at

every point p M a linear isomorphism of tangent spaces.

Definition A.1.25 Let f: M → N be a differentiable map between manifolds.

A point p M is called a regular point of f if the differential D p f is
surjective onto T f( p) N.
 A point q N is called a regular value of f if each point p in the preimage
f −1(q) M is a regular point.
The map f is called a submersion if every point p M is regular.
The map f is called an immersion if the differential D p f is injective at
every point p M.

Remark A.1.26 Every point of N that is not in the image f(M) is automatically a
regular value, because the condition is empty.

Theorem A.1.27 (Sard’s Theorem) For any differentiable map f: M → N

between smooth manifolds M and N the set of regular values is dense in N.

The following theorem shows that a map f has a certain normal form in a
neighbourhood of a regular point.

Theorem A.1.28 (Regular Point Theorem) Let p be a regular point of the

map f. Then there exist charts (U, ϕ) of M around p and (V, ψ) of N around f( p)
with
ϕ( p) = 0
ψ(f( p)) = 0
f(U) V
such that the map ψ ∘ f ∘ϕ −1 has the form

where dimM = n + k and dimN = n.

Remark A.1.29 The theorem says that in suitable charts the map f is given by
the standard projection of onto .

A.1.7 Immersed and Embedded Submanifolds

There are two notions of submanifolds which need to be distinguished.

Definition A.1.30 Let M be a smooth manifold.

1. An immersed submanifold of M is the image of an injective immersion f: N

 → M from a manifold N to M.

2. An embedded submanifold of M is the image of an injective immersion

f: N → M from a manifold N to M which is a homeomorphism onto its
image.
In both cases, the set f(N) is endowed with the topology and manifold
structure making f: N → f(N) a diffeomorphism. The difference between
embedded and immersed submanifolds f(N) M is whether the topology on
f(N) coincides with the subspace topology on f(N) inherited from M or not.

An embedded submanifold can be characterized equivalently as follows:

Proposition A.1.31 A subset K of an m-dimensional manifold M is an

embedded submanifold of dimension k if and only if around each point p K
there exists a chart (U, ϕ) of M such that

Such a chart is also called a submanifold chart or flattener for K.

The regular point theorem implies:

Theorem A.1.32 (Regular Value Theorem) Let q N be a regular value of

a smooth map f: M → N and L = f −1(q) the preimage of q. Then L is an
embedded submanifold of M of dimension

A.1.8 Vector Fields

Let M n be a smooth manifold. A vector field on M is a map X that assigns to
each point p M a tangent vector X p T p M in a smooth way. To make this
precise let be a chart. We set

for the tangent bundle of U and define the map

The map Dϕ is on each fibre { p} × T p M of TU an isomorphism onto

.

Definition A.1.33 A vector field X on M is a map X: M → TM such that:

1. X p = X( p) T p M for all p M.

2. The map X is differentiable in the following sense: For any chart (U, ϕ) the
lower horizontal map in the following diagram

is differentiable (this is just a standard vector field on ).

A particularly important set of vector fields is defined by a chart.

Definition A.1.34 Let (U, ϕ) be a chart for M. Then we define at every point p
U the following vectors:

where e 1, …, e n is the standard basis of . We also write

For a fixed index μ, as p varies, the vectors ∂ μ ( p) form a smooth vector field ∂ μ
on U. We call the vector fields ∂ μ basis vector fields or coordinate vector
fields on U.

Lemma A.1.35 At each point p U the vectors ∂ 1( p), …∂ n ( p) form a basis

for the tangent space T p M.

Proof This is clear, because is an isomorphism of vector

spaces. □

Proposition A.1.36 Every smooth vector field X on M can be written on U as

where are smooth real-valued functions on U, called the
components of X with respect to the basis {∂ μ }.

Remark A.1.37 The second equality in this proposition is an example of the so-
called Einstein summation convention .

A.1.9 Integral Curves

Let M be a smooth manifold and X a smooth vector field on M.

Definition A.1.38 A curve γ: I → M, where is an open interval around 0,

is called an integral curve for X through p M if

The theory of ordinary differential equations (ODEs) applied in a chart for M

shows that:

Theorem A.1.39 For every point q M there exists an interval I q around 0

and a unique curve γ q : I q → M which is an integral curve for X.

Using a theorem on the behaviour of solutions to ODEs under variation of the

initial condition we get:

Theorem A.1.40 For all p M there exists an open neighbourhood U of p in

M and an open interval I around 0 such that the integral curves γ q are defined
on I for all q U. The map

is differentiable and is called the local flow of X.

Theorem A.1.41 Let M be a closed manifold (compact and without

boundary). Then there exists a global flow of X which is a smooth map

The map
is a diffeomorphism for all .

A.1.10 The Commutator of Vector Fields

Let X be a smooth vector field on the manifold M.

Definition A.1.42 The Lie derivative L X is the map

defined by

for all and p M.

The Lie derivative L X is the directional derivative of a smooth function along the
vector field X: If γ is a curve through p such that , then

Proposition A.1.43 The Lie derivative is a derivation , i.e.

1. L X is -linear

2. L X satisfies the Leibniz rule :

Using the Lie derivative we can define the so-called commutator of vector fields.

Theorem A.1.44 Let X and Y be smooth vector fields on M. Then there exists
a unique vector field [X, Y ] on M, called the commutator of X and Y, such that
(A.1)
If in a local chart (U, ϕ) the vector fields are given by

then [X, Y ] is given by

(A.2)
Theorem A.1.45 The set of vector field together with the commutator is
an (infinite-dimensional) Lie algebra , i.e. for all we have:
antisymmetry:

-bilinearity:

Jacobi identity :

We can calculate the commutator [X, Y ] using the flow of X:

Theorem A.1.46 Let X and Y be smooth vector fields on M, ϕ t the flow of X

and p M a point. Then

Note that is a smooth curve in T p M.

A.1.11 Vector Fields Related by a Smooth Map

Definition A.1.47 Let M and N be smooth manifolds and ϕ: M → N a smooth
map. Suppose that X is a vector field on M and Y a vector field on N. Then Y is
said to be ϕ -related to X if

Lemma A.1.48 Let M and N be smooth manifolds, ϕ: M → N a smooth map.

Suppose that X and Y are vector fields on M and N and that Y is ϕ-related to X.
Then

Proposition A.1.49 Let M and N be smooth manifolds and ϕ: M → N a

smooth map. Suppose that X′ is ϕ-related to X and Y ′ is ϕ-related to Y. Then [X′,
Y ′] is ϕ-related to [X, Y ].

Definition A.1.50 If ϕ: M → N is a diffeomorphism and X is a smooth vector

field on M, then we define a smooth vector field ϕ X on N, called the
pushforward of X under ϕ, by

Note that ϕ X is the unique vector field on N that is ϕ-related to X.

Corollary A.1.51 If ϕ: M → N is a diffeomorphism, then

for all vector fields X and Y on M.

A.1.12 Distributions and Foliations

We consider some concepts related to distributions and foliations on manifolds
(we follow [142] where proofs and more details can be found). Let M be a
smooth manifold of dimension n.

Definition A.1.52 A distribution D of rank k on M is a collection of vector

subspaces D p T p M of dimension k for all p M which vary smoothly over
M, i.e. each p M has an open neighbourhood U M so that D | U is spanned
by k smooth vector fields X 1, …, X k on U.

An equivalent definition is that D is a subbundle of rank k of the tangent bundle

TM.

Definition A.1.53 A distribution is called involutive or integrable if for all

vector fields X, Y on M with X p , Y p D p for all p M, the vector field [X, Y ]
on M again satisfies [X, Y ] p D p for all p M.

Definition A.1.54 A foliation of rank k on M is a decomposition of M into

k-dimensional immersed submanifolds, called leaves, which locally have the
following structure: around each point p M there exists a coordinate
neighbourhood diffeomorphic to such that the leaves of the foliation
decompose into , with the leaves given by the affine subspaces
for all .

It is clear that the tangent spaces to the leaves of a foliation define a distribution.
In fact, we have:
Theorem A.1.55 (Frobenius Theorem) A distribution D defines a foliation
if and only if D is integrable.

The following statement is Theorem 1.62 in [142].

Theorem A.1.56 Let f: N → M be a smooth map between manifolds, a

foliation on M and H M a leaf of . Suppose that f has image in H. Then
f: N → H is smooth.

This theorem is clear if H is an embedded submanifold of M and only non-trivial

if H is an immersed submanifold.

A.2 Tensors and Forms

A.2.1 Tensors and Exterior Algebra of Vector Spaces
We recall some notions from linear algebra. Let V denote an n-dimensional real
vector space.

Definition A.2.1 We set

for the dual space of V. The dual space V is itself an n-dimensional real vector
space. We call the elements λ V 1-forms on V.

If {e μ } is a basis for V we get a dual basis {ω ν } for V defined by

where δ μ ν is the standard Kronecker delta. Just as we decompose any vector X

V in the basis {e μ } as

we can decompose any 1-form λ V as

(Note the Einstein summation convention in both cases.)

Definition A.2.2 A tensor of type (l, k) is a multilinear map

In particular, a (0, 1)-tensor is a 1-form and a (1, 0)-tensor is a vector. The set of
all (l, k)-tensors forms a vector space.

We are interested in a particular class of tensors on a vector space V.

Definition A.2.3 We call a (0, k)-tensor

a k-form on V if λ is alternating , i.e. totally antisymmetric:

for all insertions of vectors into λ, where only the vectors v and w are
interchanged. The set of k-forms on V forms a vector space denoted by Λ k V .

Remark A.2.4 It follows that for k-forms λ

and

whenever the vectors v 1, v 2, …, v k are linearly dependent. In particular, every k-

form on V vanishes identically if k is larger than the dimension of V.

Definition A.2.5 Let λ be a k-form and μ an l-form. Then the wedge product
of λ μ is the (k + l)-form defined by

Here S k+l denotes the set of permutations of {1, 2, …, k + l}. It can be checked
that λ μ is indeed a k + l-form.

Example A.2.6 Let α, β be 1-forms on V. Then

for all vectors X, Y V.

Lemma A.2.7 Let V be a vector space of dimension n and {ω ν } a basis for V

. Then the set of k-forms

is a basis for the vector space of k-forms.

A.2.2 Tensors and Differential Forms on Manifolds
Let M be an n-dimensional smooth manifold. We want to extend the notion of
tensors and forms on vector spaces to tensors and forms on M. One possibility is
to first define certain vector bundles and then tensors and forms as smooth
sections of these bundles. However, since we define vector bundles in Sect. 4.5,
we use here another, equivalent definition for tensors.

Remark A.2.8 In the following all functions and vector fields on M are smooth.

Definition A.2.9 We denote by the ring of all smooth functions

. We also denote by the set of all smooth vector fields on M. The
set is a real vector space and module over by point-wise
multiplication.

We can now define:

Definition A.2.10 A 1-form λ on the manifold M is a map

that is linear over , i.e.

for all vector fields and functions . We denote the set of all
1-forms on M by Ω 1(M), which is a real vector space and module over .

The following can be proved:

Proposition A.2.11 The value of λ(X)( p) for a 1-form λ and vector field X at a
point p M depends only on X p . Hence if Y is another vector field on M with Y
p = X p , then λ(X)( p) = λ(Y )( p).

A proof of this proposition can be found in [142, p. 64]. Similarly we set:

Definition A.2.12 A tensor T of type (l, k) on M is a map

that is -linear in each entry. A k -form or differential form ω on M is a

(0, k)-tensor

that is in addition alternating (totally antisymmetric). We denote the set of k-

forms on M by Ω k (M).

Remark A.2.13 An argument similar to the proof of Proposition A.2.11 shows

that tensors and k-forms on manifolds have well-defined values at every point p
M. We can therefore insert, for example, in a k-form ω Ω k (M) vectors X
1, …, X k in the tangent space T p M at any point p M and get a real number.
We can also speak unambiguously of the value of a tensor or form at a point.

Remark A.2.14 We can define the wedge product of forms as before by

replacing in the definition vectors by vector fields on the manifold. The wedge
product is then a map

A.2.3 Scalar Products and Metrics on Manifolds

We consider the following definition from linear algebra.

Definition A.2.15 A scalar product on the vector space V is a symmetric

non-degenerate (0, 2)-tensor g on V:

The scalar product g is called Euclidean if it is positive definite

and pseudo-Euclidean otherwise.

We can do the same construction on manifolds.

Definition A.2.16 A metric on a smooth manifold M is a (0, 2)-tensor g which

is a scalar product at each point p M. The metric is called Riemannian if the
scalar products g p are Euclidean and pseudo-Riemannian if the scalar products
g p are pseudo-Euclidean, for all p M.
It can be shown using partitions of unity that every smooth manifold admits a
Riemannian metric (but not necessarily a pseudo-Riemannian metric).

A.2.4 The Levi-Civita Connection

Let (M, g) be a pseudo-Riemannian manifold. The Levi-Civita connection is a
metric and torsion-free, covariant derivative on the tangent bundle of the
manifold, i.e. a map

with the following properties:

1. is -linear in both X and Y.

2. is -linear in X and satisfies

3. is metric, i.e.

4. is torsion-free, i.e.

The Levi-Civita connection can be calculated with the following Koszul

formula :

A.2.5 Coordinate Representations

We saw above that we can represent every vector field X on a chart
neighbourhood U by X | U = X μ ∂ μ , where X μ are certain functions on U, called
components. We want to decompose in a similar way tensors and forms on U. In
the physics literature tensors and forms are often given in terms of their
components in coordinate systems.
Definition A.2.17 Let U be a chart neighbourhood. We define the set of dual
1-forms dx μ , for μ = 1, …, n, by dx μ (∂ ν ) = δ ν μ at each point p U.

Proposition A.2.18 Let λ be a 1-form on M. Then we can decompose λ on U

as λ | U = λ μ dx μ for certain smooth functions λ μ on M. Similarly, we can
decompose a k-form ω as

with smooth functions .

Note that these functions, corresponding to the components, depend on the

choice of the chart (U, ϕ), while the objects themselves (vectors fields, k-forms)
are independent of charts.

A.2.6 The Pullback of Forms on Manifolds

Let ω Ω k (N) be a k-form on a manifold N and f: M → N a smooth map.

Definition A.2.19 The pullback of ω under f is the k-form f ω Ω k (M) on

M defined by

for all tangent vectors X 1, …, X k T p M and all p M.

Proposition A.2.20 The pullback defines a map f : Ω k (N) ⟶ Ω k (M). We

have

for all ω Ω k (N), η Ω l (N) and

for all smooth maps f: M → N, g: N → Q.

The second property follows from the chain rule for the differential of the map g
∘ f.

A.2.7 The Differential of Forms on Manifolds

The differential is a very important map on forms on a manifold that raises the
degree by one.
Theorem A.2.21 Let M be a smooth manifold. Then there is a unique map

for every k ≥ 0, called the differential or exterior derivative , that satisfies the
following properties:

1. d is -linear.

2. For a function and a vector field we have df(X)

= L X f.

3. d 2 = d ∘ d = 0: Ω k (M) → Ω k+2(M).

4. d satisfies the following Leibniz rule :

for all α Ω k (M), β Ω l (M).

The proof of this fundamental theorem can be found in any book on differential
geometry. Let (U, ϕ) be a local chart. If we assume that the differential d has
these properties, then it follows that the differential is given on functions f by

and on Ω k (M) by

The defining properties of the differential d imply for 1-forms and 2-forms:

Proposition A.2.22

1. Let α Ω 1(M) be a 1-form. Then

2. Let β Ω 2(M) be a 2-form. Then

The differential is natural under pullback:

Proposition A.2.23 If f: M → N is a smooth map and ω Ω k (N), then d(f

ω) = f dω.

Let M be a compact oriented n-dimensional manifold and σ Ω n (M) a form of

top degree. Then there is a well-defined integral

The integral can also be defined if M is non-compact and σ has compact support.

Theorem A.2.24 (Stokes’ Theorem)

1. Let M be a compact n-dimensional oriented manifold with boundary ∂M and

ω Ω n−1(M). Then (with a suitable orientation of the boundary)

2. Let M be an n-dimensional oriented manifold (not necessarily compact)

without boundary and ω Ω n−1(M) an (n − 1)-form with compact support.
Then

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]
4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)
19.
 Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
 [MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]
51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]
67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]
84.
 Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
 [MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
(2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
 Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]
130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]
147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions. Chin.
Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/
© Springer International Publishing AG 2017
Mark J.D. Hamilton, Mathematical Gauge Theory, Universitext, https://doi.org/10.1007/978-3-319-68439-
0_11

Appendix B: Background on Special

Relativity and Quantum Field Theory
Mark J. D. Hamilton1
(1) Department of Mathematics, Ludwig-Maximilian University of Munich,
Munich, Germany

B.1 Basics of Special Relativity

We very briefly recall some basic concepts from special relativity. A very good
introduction to the physics and mathematics of special relativity can be found in
[95], covering much more than we need.
Special relativity is formulated on Minkowski spacetime with a
pseudo-Riemannian metric known as Minkowski metric η given by (we use
units where the speed of light c = 1)

This choice of signs (+, −, −, −) is called the West Coast metric . Sometimes
the East Coast metric with signature (−, +, +, +) is used instead. The x μ are the
standard coordinates on , also written as

or x μ = (t, x). Coordinate systems (charts) on Minkowski spacetime

correspond to reference frames of moving observers. All inertial systems
(unaccelerated orthonormal reference frames, also called Lorentz frames ) with
the same origin (0, 0) are related by Lorentz transformations
 Here the Einstein summation convention is understood and Λ is a matrix
preserving the metric η

also written as

or

If the origins are different, then the coordinate transformations between

inertial systems are given by Poincaré transformations

where is a constant vector. Poincaré transformations are affine

transformations.
Since the Minkowski distance between two points is independent of the
chosen inertial frame, two points with distance zero (lightlike) in one frame have
the same distance zero in any other Lorentz frame, meaning that the speed of
light c = 1 is the same in any inertial frame.
The basis vectors of Lorentz frames transform as

so that the vector Δx μ e μ is invariant:

The same vector a on spacetime can be expressed in the frame e μ or the

frame e ρ ′ :

This implies that

Similarly, a 1-form ω can be expressed in the frame dx μ or the frame dx ′ρ :

This implies that

If a is a vector with components a μ in the frame e μ and we set

then a ν transforms as the components of a 1-form. Similarly, if we define the

matrix η μν as the inverse of the matrix η μν (in the Minkowski case this is the
same matrix) and if ω μ are the components of a 1-form in the frame dx μ , then

transforms as the components of a vector. This is the idea behind lowering

and raising indices, which can be extended to arbitrary tensors.
Let be the velocity vector of a particle of mass m > 0. Going to the rest
frame of the particle, the vector u has components u = (1, 0) = (1, 0, 0, 0), which
implies that η(u, u) = 1 independent of the chosen frame. In any frame we define
the 4-momentum

where E is the energy and p the 3-momentum of the particle in that frame.
Then η(u, u) = 1 implies

If we introduce again the speed of light c, then η(u, u) = c 2,

and

Relativistic theories of physics are theories formulated in Minkowski

spacetime whose laws are invariant under Poincaré transformations. This means
that the laws of physics are independent of where and when experiments are
performed (invariance under space and time translations), how the experiments
are oriented in space (invariance under rotations) and whether they are
performed in different inertial systems moving with constant velocity (invariance
under Lorentz boosts). For example, the principle of relativity claims that the
laws of physics are the same here and in the Andromeda galaxy, they are the
same now and in 1 million years and they are the same on board two spacecrafts
flying with arbitrary, constant velocities in different directions.

B.2 A Short Introduction to Quantum Field Theory

From classical gravity and electromagnetism we are used to thinking of matter as
particles and interactions as carried by fields. However, according to quantum
field theory, matter and interactions can both be described by particles and
fields. Quantum field theory can be thought of as a unification of the concepts of
classical fields and point particles and thus as a unification (in some sense) of
interactions and matter (supersymmetric quantum field theories are a unification
of both concepts in an even stronger sense). The remarkable consequences of
this approach are that forces between matter particles can be reduced to
couplings between different types of fields and that symmetry groups, such as
gauge symmetries, can act through representations on both interaction and
matter fields.
In the following sections we briefly want to discuss the basics of quantum
field theory and the relation between particles and fields. Our intention is to give
a short overview and interpretation, without any calculations or trying to be
mathematically rigorous. We also assume a basic familiarity with quantum
mechanics.

B.2.1 Quantum Field Theory and Quantum Mechanics

Quantum field theory (QFT) is a quantum theory, in some sense similar to
quantum mechanics (QM) :
A quantum system has a Hilbert space V with a Hermitian scalar product
⋅ | ⋅ . Elements of the vector space V are state vectors (states) | v (we
normalize these vectors to unit norm). We think of the state of the system as
being time-dependent | v, t (Schrödinger picture) . However, we can
equivalently think of the states as being time-independent and instead the
operators as being time-dependent (Heisenberg picture, usually preferred in
QFT) .
We cannot measure the state of a system directly, we can only measure the
value of observables , described by Hermitian operators A on V. If | v is an
eigenvector of A with eigenvalue a,

and we measure the observable A if the system is in the state | v , then the
value is the eigenvalue a. For an arbitrary state | w , the expectation value
of the observable A is related to w | A | w .
We are also interested in transition amplitudes between states, given by
scalar products w | v . The amplitudes determine transition probabilities
 (the probability that the system in the state | v is found after a measurement
in the state | w ) by taking the absolute value squared of this complex
number.
There is a Hermitian Hamiltonian operator H which determines the
evolution of states between times t 0 and t (by convention t 0 = 0): we define
the unitary operator

where is the Planck constant (note that the exponential of a skew-

Hermitian operator is unitary). Then time evolution of states is given by

One of our aims is to determine the time evolution operator U(t, 0). Ideally
we would like to diagonalize H, i.e. find an eigenbasis for H of states | n of
energy E n ,

because such states have a very simple time evolution:

where is just a complex number of absolute value 1. In general,

in an interacting theory, this will be practically impossible.
We can change from the Heisenberg picture to the Schrödinger picture and
vice versa as follows: the Schrödinger-type operator is the Heisenberg-type
operator taken at t 0 = 0:

The time evolution of the Heisenberg-type operator is then given by the

Hamiltonian:

So far everything should be familiar from QM. We now discuss what is

peculiar about QFT.

B.2.2 Free Quantum Field Theory on 0-Dimensional

Space
Suppose that space is 0-dimensional and consists only of a single point. A real-
valued field at this point is just a time dependent real number ϕ(t). The simplest
type of quadratic Lagrangian for this field is

This Lagrangian is known as the harmonic oscillator . The Euler–Lagrange

equation for this Lagrangian is the ordinary differential equation

The Hilbert space of the associated quantum theory can be described as follows:
let

The basis states corresponding to this direct sum decomposition are denoted by

The vector space is the Hilbert space of the harmonic oscillator.

The states | n are eigenvectors for the Hamiltonian H with energy E n
growing linearly with n. These states are interpreted as the discrete set of
different vibrational modes of the field at the point.
There is a Hermitian number operator N (an observable) whose
eigenvectors are | n with eigenvalue n:

B.2.3 Free Quantum Field Theory on d-Dimensional

Space
Free quantum field theories (and to a certain degree, weakly interacting,
perturbative quantum field theories) have an interpretation in terms of particles .

Canonical Quantization
We consider the case of field theories on d-dimensional Euclidean space (for
simplicity we assume d = 3). A real-valued field is now a real function ϕ(t, x)
depending on time t and the space coordinate x. The simplest type of quadratic
Lagrangian for this field is the Klein–Gordon Lagrangian .

(B.1)
The Euler–Lagrange equation for this Lagrangian is the linear wave equation
called the Klein–Gordon equation .
The Hilbert space of the associated quantum field theory can be described as
follows: let V 1 be the Hilbert space of a single free bosonic particle. It is
spanned by the basis states | p = | 1 p , where is the momentum of the
particle, related to its energy by m 2 = E 2 − p 2. In these states the particle is
totally delocalized in position space. States where the particle is localized both in
momentum and position space with a certain minimal width (given by
Heisenberg’s uncertainty principle) can be obtained as linear combinations of the
states | p , called wave packets .
A general construction in quantum theory implies that the Hilbert space of n
indistinguishable particles of the same type is given by

n-particle states are thus (linear combinations of) symmetrized tensor

products of 1-particle states. We then form the (bosonic) Fock space

that contains states with an arbitrary number of particles. It turns out that the
Fock space is a suitable Hilbert space for the quantum field theory described
by the Klein–Gordon Lagrangian (B.1).
A basis for the Fock space is given by states

where are vectors and is the number of particles of momentum p

i. The total number of particles in this state is

These states are eigenvectors of the Hamiltonian H with energy growing

linearly with the numbers and they again correspond to different
vibrational modes of the field.
The basis state | 0 of the 1-dimensional space , where all are
zero, is called the vacuum. The vacuum is the unique eigenstate of the
Hamiltonian of eigenvalue 0.
 For every vector p there is a Hermitian number operator N p with
eigenvectors

The number operator is an observable which returns the number of particles

of a given momentum in a given quantum state.
The classical field ϕ(x) becomes in the QFT a field of Schrödinger-type
operators , depending on the space point x (more precisely, an operator-
valued distribution), that all act on the same Hilbert space V. This field of
operators is called the quantum field . Together with the adjoint quantum
field it creates and annihilates particles in the point x, i.e. adds or
removes these particles from the state in the Hilbert space. In the
Heisenberg picture the field depends on the point (t, x) in spacetime.
Similarly, the Fock space V for a free fermionic field can be generated using
antisymmetrized tensor products:

In this case the numbers n p can only take the values 0 or 1.

These descriptions of the Fock spaces make it clear that the Hilbert space
in quantum field theory is infinite-dimensional in two ways: the one-particle
space V 1 is infinite-dimensional because the vector space of momenta p
has infinitely many elements (this is related to the fact that space is
continuous and infinitely extended). In addition there is the infinite direct sum
over the number of particles that we already encountered in the case
of quantum mechanics of the harmonic oscillator (for the harmonic oscillator
the vector space corresponding to V 1 is 1-dimensional).

B.2.4 Unitary Representation of the Poincaré Group

As one of the general axioms of QFT on 4-dimensional Minkowski spacetime
we assume that the Hilbert space of the quantum theory carries a unitary
representation of the universal covering group of the Poincaré
group. Note that the non-compact simple Lie group does not admit non-
trivial finite-dimensional unitary representations according to Theorem 2.1.44.
B.2.5 Interacting Quantum Field Theories
A typical question in QFT is to calculate scattering amplitudes : Suppose we
send in an (idealized) collider a total number of n particles with certain momenta
p i and want to determine the probability that we find after collision n′ particles
with certain momenta p j ′ . This process is governed by the laws of quantum
theory: in general we can only calculate a probability for the process or
transition to happen, we cannot predict the outcome completely, even if we know
the initial state exactly. Since the numbers n and n′ as well as the types of
particles and their momenta can be different, certain particles get created and
others annihilated in the scattering process .1
To describe interacting QFTs, like the ϕ 4-theory with Lagrangian

we make the following assumptions, known as the interaction picture:

The Hilbert space V of the interacting theory is the same Hilbert space as in
the free theory (the Fock space). The Schrödinger-type field operators
(operator-valued distributions) are also the same as in the free theory.
The Hamiltonian H can be calculated from the Lagrangian of the field
theory, expressed through the fields, that we collectively denote by ϕ. The
Hamiltonian of the interacting theory is of the form

where H 0 is the Hamiltonian of the free theory and H I is the interaction

part. Since the Hamiltonian H of the interacting theory is different from the
Hamiltonian H 0 of the free theory, we expect the vacuum state | Ω of the
interacting theory to be different from the vacuum state | 0 of the free
theory, even though both states (under our assumption) are elements of the
same Fock space V.
The time-dependence of the Heisenberg-type field operators will be
different in the free and interacting theories, because the Hamiltonians are
different. One considers two types of time-dependent quantum fields: the
Heisenberg picture field is given the time evolution according to the full
Hamiltonian H and the interaction picture field is given the time evolution
according to the free Hamiltonian H 0.
Scattering of particles in a collider can now be described as follows. We
assume that for time t → −∞ in the distant past and for time t′ → +∞ in the
distant future the particles are far apart and can be considered as free. We can
then think of the collections of particles that we send into and get out of the
collider as states in the Hilbert space V:

We want to calculate the scalar product

If we know these scalar products, we can calculate the transition amplitudes

between any two states, because the momentum states form a basis for the
Hilbert space. Since the time evolution operator U is defined via the Hamiltonian
H, it follows that the scalar product for different ingoing and outgoing states will
be non-zero only if the action of H on particle states creates and annihilates
certain particles.
More precisely, in a free field theory the Hamiltonian is quadratic in the
fields ϕ and can be diagonalized. The eigenbasis is just given by the particle
momentum states

which are fixed by the action of H up to multiplication by the eigenvalue (the

total energy E of the collection of particles). It follows that in a free field theory
particles do not get created or annihilated. The transition amplitude between
different particle momentum states is zero and all scattering processes are trivial:
we get the same particles with the same momenta out that we sent into the
collider. The vibrational modes of the field described by these states are
constant, independent of time.
In an interacting theory the Hamiltonian contains anharmonic terms ,
i.e. terms of order three or higher in the fields ϕ. Such Hamiltonians lead to non-
trivial creation and annihilation of particles and thus to non-trivial scattering
processes. Heuristically, the vibrational modes of the fields change with time
and, since the fields are coupled and the corresponding equations of motion are
non-linear wave equations, the vibrations of one field can start vibrations of
another field.
The description so far assumed that the Hilbert space and the action of the
Schrödinger-type field operators are the same for the interacting theory as for the
free theory, only the vacuum state and the Hamiltonian have changed. This
assumption is merely a first approximation and actually not consistent, according
to Haag’s Theorem : Schrödinger-type quantum fields for a free and an
interacting theory cannot be the same; see, for example, [116, p. 391]. To define
quantum field theories in a mathematically rigorous way is the aim of
constructive quantum field theory (CQFT) and algebraic quantum field theory
(AQFT) .

B.2.6 Path Integrals

Transition amplitudes can also be calculated using path integrals . Path integrals
for interacting theories, in particular, gauge theories, are more convenient than
canonical quantization. The path integral approach to quantum theory was
originally developed by P.A.M. Dirac and R.P. Feynman.
To understand the idea of path integrals, recall that standard integrals are
integrals of functions over finite-dimensional vector spaces (or finite-
dimensional manifolds). Path integrals are integrals of functions (also called
functionals) over infinite-dimensional vector spaces (or infinite-dimensional
manifolds). These vector spaces arise naturally as the spaces of all fields of a
certain type on spacetime, i.e. the space , where M is spacetime and W
is the vector space in which the field ϕ takes values (more generally, we could
consider the space of all sections of a vector bundle over M). Path integrals over
the infinite-dimensional vector space can be approximated by standard
integrals over a finite-dimensional vector space if spacetime is replaced by a
lattice with finite lattice spacing a > 0 (and finite extension).
The path integrals that appear in QFT are of the form

called Green’s functions or correlators . Here x 1, …, x n M are points in

spacetime, ϕ is the field, is the path integral measure on the space
and is the Lagrangian of the field theory. The number is a
normalization constant. It is clear that (say for a complex scalar field ϕ, taking
value in the complex numbers ) for fixed points x 1, …, x n the map

is a function (functional) on the vector space . The path integral

is the integral of this function over the infinite-dimensional space .

 Notice that the field ϕ here is the classical field, not the quantum field of
operators. The approach to QFT using path integrals is independent of the
approach using Hilbert spaces and quantum fields. However, it can be shown
that if one knows all the Green’s functions, then the Hilbert space together with
the quantum fields can be reconstructed (this is known as the Wightman
Reconstruction Theorem ).
In general, it is very difficult to calculate or even define these path integrals
precisely. Usually, this can only be done in the case of the free field with a
quadratic Lagrangian . Scattering amplitudes can be calculated from Green’s
functions using the LSZ reduction formula , named after H. Lehmann,
K. Symanzik and W. Zimmermann.

B.2.7 Series Expansions

The actual calculation of scattering amplitudes is a formidable task and often can
only be done approximately, using power series expansions. In general, the
Green’s functions are functions of the coupling constant(s) g and the Planck
constant :

There are mainly two types of series expansions:

Perturbation Theory
Perturbation theory works if the coupling constant g is small so that the full
Lagrangian is a small perturbation

of the free Lagrangian . The Green’s functions for are known and the
Green’s functions for can be calculated in a series expansion in orders of g, by
expanding the exponential

in a power series in orders of g and then interchanging the path integral and the
infinite sum (this step is mathematically not justified [50]). The terms in the
power series expansion are described by Feynman diagrams . With the order of
g increasing in each step by 1, the terms in the series expansion for a process
with fixed external lines (in-coming and out-going particles) are called leading
order (LO) , next-to-leading order (NLO) , next-to-next-to-leading order
(NNLO) , and so on.
In perturbation theory, the full interacting Lagrangian is treated as a small
perturbation of the free Lagrangian. Since the states of a free quantum field
have an interpretation in terms of particles, it makes sense to think of
perturbation theory as describing (weakly) interacting particles. Feynman
diagrams, that depict these interactions, are the hallmark of perturbative
quantum field theory.

Internal lines in Feynman diagrams represent intermediate virtual particles

which are off-shell, i.e. do not satisfy the mass energy relation m 2 = E 2 − p 2
(even though momentum and energy are conserved at each vertex). In contrast to
the in- and out-state particles, virtual particles are therefore not “real” particles
and cannot be detected (the photons that mediate interactions between electrons
are different from the photons that we can see or detect with cameras).
A problem of perturbation theory is that the perturbation expansion in orders
of the coupling constant g actually converges only for g = 0, i.e. the radius of
convergence of the power series is zero (there is an argument due to Freeman
Dyson that a QFT cannot be well-defined for negative values of the coupling
constant g, hence the expansion around g = 0 must have radius of convergence
equal to zero). This implies that the perturbation expansion only makes sense as
an asymptotic expansion : up to a certain optimal order of g the series
expansion approximates the Green’s function better and better, but then, adding
terms of higher order, the series expansion starts getting worse and eventually
diverges.

Semi-Classical Approximation
Semi-classical approximation can be used if the Planck constant is a
(relatively) small number. The power series for the Green’s functions is then an
expansion in orders of . The lowest term of order zero is the classical
contribution and terms in higher order of are quantum corrections . In
Feynman diagrams the classical contribution corresponds to tree diagrams ,
whereas quantum corrections correspond to loop diagrams . With respect to
path integrals the semi-classical approximation is an expansion around the
critical points of the Lagrangian, i.e. the classical solutions of the field equations
(for → 0 the path integral localizes at these classical solutions).

Non-Perturbative Quantum Field Theories

Note that we do not claim that the Green’s functions are analytic in g or . Most
smooth functions are, of course, not analytic, because analytic functions are
determined everywhere in their domain of definition by their values in an
arbitrarily small neighbourhood of the center of expansion. The series
expansions in QFT will therefore only be approximately accurate for small
values of g and and unusable if these parameters are large. In particular, if g is
large, the QFT is called non-perturbative .
The term non-perturbative is essentially a synonym for non-analytic.
Non-perturbative effects in QFT will typically become dominant if the
coupling constant g is large. The particle interpretation breaks down (at least
for the fundamental fields) for strongly interacting, non-perturbative QFTs.

B.2.8 Renormalization
The calculation of the contribution of Feynman diagrams with loops 2 involves
certain integrals that can diverge and lead to infinite Green’s functions. The idea
of renormalization is to absorb the infinities that occur in the Green’s functions
into the parameters (in particular, the masses and coupling constants), which
then become infinite themselves, while the Green’s functions become finite. For
this to work the parameters have to go in “the right way” to infinity, so that the
Green’s functions stay finite. More precisely, the parameters are no longer
constants, but certain functions of a cutoff, and go to infinity when the cutoff is
removed, whereas the Green’s functions remain finite.
Alternatively, renormalization can be understood as adding to the original
Lagrangian of the field theory a counterterm Lagrangian that cancels the
divergences of the Green’s functions. If we include terms (interactions) of the
same form as the counterterms in the original Lagrangian, then adding
counterterms is equivalent to a renormalization of parameters. A QFT is called
renormalizable if finitely many counterterms are needed to cancel the
divergences and non-renormalizable if infinitely many counterterms are
needed. Non-renormalizable theories contain infinitely many different types of
interactions and infinitely many parameters, but can still be useful (cf. [125,
Sect. 21.2.2]).
The process of renormalization can be explained with a classical analogy,
first observed by M. Abraham and H. Lorentz: the electric field of a charged
point particle is of the form
where α ≠ 0 is some constant and r is the radial vector. The energy density u
of the electric field is proportional to | E |2, hence of the form

with a constant β ≠ 0. It follows that the total energy of the field is

This integral, when extended all the way to 0, is infinite. It follows that a
charged point particle, like an electron, has an infinite energy in its electric field.
If this energy is added to the bare rest mass of the electron via E = mc 2,
corresponding to an electromagnetic mass, the total mass becomes infinite,
which seems like a contradiction.
The idea is to set the bare (unobservable) rest mass m B of the electron equal
to −∞, so that when we add the infinite energy due to the electric field the total
(observable) mass m becomes finite. We define a cutoff ε > 0 and set

which is finite for all ε > 0. This is called regularization of the divergent
integral. We also define

where m is the observed mass of the electron, known from experiments. This
is called renormalization of the mass. The bare mass is thus a function of the
cutoff ε and goes to −∞ if we let ε → 0. However, the total mass is now

which is constant and equal to the finite mass m for all ε > 0. We see that we
have hidden the infinity from the divergent integral I(0) in the renormalization of
the mass m B .
In general, the divergences encountered in QFTs can be traced to two aspects
of space: space is continuous (leading to UV divergences) and space is infinitely
extended (leading to IR divergences). Both aspects imply that QFTs, which
describe time-dependent fields defined on space, have to deal with systems with
infinitely many degrees of freedom, the crucial difference to QM. A QFT can be
regularized by introducing cutoffs: a UV cutoff essentially means to reduce
space to a lattice with finite lattice spacing a > 0 (corresponding to an upper
cutoff on the norm | p | of the momentum) and an IR cutoff means to consider
the theory in a finite volume V < ∞ of space (corresponding to a discrete set of
momenta). Both regularizations together reduce the QFT to a system with
finitely many degrees of freedom (in continuous time), essentially a version of
QM.

B.2.9 Further Reading

Introductory accounts of quantum field theory are [51] and [86]. Extensive
discussions from a physics point of view can be found, for example, in [124,
125, 132] and [143–145]. Mathematically rigorous discussions can be found in
the classic references [17, 69, 134] and in [38, 45, 82]. Perturbation theory, semi-
classical approximation and renormalization are very well and comprehensibly
explained from a mathematical point of view in the lecture notes [50].

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72, 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75, 603–632 (1962)
[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308, 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3, 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39, 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42, 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47, 483–552
(2010)
10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
(2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347(6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84, 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74, 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)
25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126, 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17, 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704, 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
[MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40, 101–129
(2001)
[MathSciNet][MATH]
41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013, Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322, 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6, 111–244 (1957)

47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113, 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93, 193–266
(1975)
 [MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32, 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33, 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)

63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2, 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56, 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281, 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13, 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E6. Phys. Lett. B 60,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71,
095013 (2005)
73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148, 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K-theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95, 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13, 508–509 (1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)

80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72, 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]
90. Leigh, R.G., Strassler, M.J.: Duality of Sp(2N c ) and SO(N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356, 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9, 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64, 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)

96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56, 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40, 100001
 (2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40, 100001 (2016). http://
www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

111. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 16. Grand unified theories. Chin. Phys. C 40, 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
120. range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html
121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388, 753–760 (1996)
[MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426, 19–52 (1994). Erratum: Nucl. Phys. B 430,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431, 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79, 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
 University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506, 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39, 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp(N) and SO(N) supersymmetric gauge theory in four dimensions. Chin.
Ann. Math. 24, 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251, 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/
Footnotes
1 Bound states , like atoms and hadrons, are described in QFTs using other methods.

2 Only loop diagrams require renormalization.

References
1. Abe, K. et al. (The T2K Collaboration): First combined analysis of neutrino and antineutrino
oscillations at T2K. arXiv:1701.00432

2. Adams, J.F.: On the non-existence of elements of Hopf invariant one. Ann. Math. 72 , 20–104 (1960)
[MathSciNet][MATH]

3. Adams, J.F.: Vector fields on spheres. Ann. Math. 75 , 603–632 (1962)

[MathSciNet][MATH]

4. Argyres, P.C.: An introduction to global supersymmetry. Lecture notes, Cornell University 2001.
Available at http://homepages.uc.edu/~argyrepc/cu661-gr-SUSY/index.html

5. Atiyah, M.F.: K-Theory. Notes by D.W. Anderson. W.A. Benjamin, New York/Amsterdam (1967)
[MATH]

6. Atiyah, M.F., Bott, R.: The Yang–Mills equations over Riemann surfaces. Phil. Trans. R. Soc. Lond. A
308 , 523–615 (1983)
[MathSciNet][MATH]

7. Atiyah, M.F., Bott, R., Shapiro, A.: Clifford modules. Topology 3 , 3–38 (1964)
[MathSciNet][MATH]

8. Baez, J.C.: The octonions. Bull. Am. Math. Soc. (N.S.) 39 , 145–205 (2002); Erratum: Bull. Am.
Math. Soc. (N.S.) 42 , 213 (2005)

9. Baez, J., Huerta, J.: The algebra of grand unified theories. Bull. Am. Math. Soc. (N.S.) 47 , 483–552
(2010)

10. Bailin, D., Love, A.: Introduction to Gauge Field Theory. Institute of Physics Publishing,
Bristol/Philadelphia (1993)
[MATH]

11. Ball. P.: Nuclear masses calculated from scratch. Nature, published online 20 November 2008.
doi:10.1038/news.2008.1246

12. Barut, A.O., Raczka, R.: Theory of Group Representations and Applications. Polish Scientific
Publishers, Warszawa (1980)
[MATH]

13. Baum, H.: Spin-Strukturen und Dirac-Operatoren über pseudoriemannschen Mannigfaltigkeiten.

Teubner Verlagsgesellschaft, Leipzig (1981)
[MATH]

14. Baum, H.: Eichfeldtheorie. Springer, Berlin/Heidelberg (2014)

[MATH]

15. Berline, N., Getzler, E., Vergne, M.: Heat Kernels and Dirac Operators. Springer, Berlin/Heidelberg
 (2004)
[MATH]

16. Bleecker, D.: Gauge Theory and Variational Principles. Addison-Wesley Publishing Company,
Reading, MA (1981)
[MATH]

17. Bogolubov, N.N., Logunov, A.A., Todorov, I.T.: Introduction to Axiomatic Quantum Field Theory. W.
A. Benjamin, Reading, MA (1975)

18. Borsanyi, Sz. et al: Ab initio calculation of the neutron-proton mass difference. Science 347 (6229),
1452–1455 (2015)

19. Bott, M.R.: An application of the Morse theory to the topology of Lie-groups, Bull. Soc. Math. France
84 , 251–281 (1956)
[MathSciNet][MATH]

20. Bourguignon, J.-P., Hijazi, O., Milhorat, J.-L., Moroianu, A., Moroianu, S.: A Spinorial Approach to
Riemannian and Conformal Geometry. European Mathematical Society, Zürich (2015)
[MATH]

21. Brambilla, N. et al.: QCD and strongly coupled gauge theories: challenges and perspectives. Eur.
Phys. J. C 74 , 2981 (2014)

22. Branco, G.C., Lavoura, L., Silva, J.P.: CP Violation. Oxford University Press, Oxford (1999)

23. Bredon, G.E.: Introduction to Compact Transformation Groups. Academic Press, New York/London
(1972)
[MATH]

24. Bröcker, T., tom Dieck, T.: Representations of Compact Lie Groups. Springer, Berlin/Heidelberg/New
York (2010)

25. Bröcker, T., Jänich, K.: Einführung in die Differentialtopologie. Springer, Berlin/Heidelberg/New
York (1990)
[MATH]

26. Bryant, R.L.: Metrics with exceptional holonomy. Ann. Math. 126 , 525–576 (1987)
[MathSciNet][MATH]

27. Bryant, R.L.: Submanifolds and special structures on the octonians. J. Differ. Geom. 17 , 185–232
(1982)
[MathSciNet][MATH]

28. Budinich, R., Trautman, A.: The Spinorial Chessboard. Springer, Berlin/Heidelberg (1988)
[MATH]

29. Bueno, A. et al.: Nucleon decay searches with large liquid Argon TPC detectors at shallow depths:
atmospheric neutrinos and cosmogenic backgrounds. JHEP 0704 , 041 (2007)

30. Čap, A., Slovák, J.: Parabolic Geometries I: Background and General Theory. American Mathematical
Society, Providence, RI (2009)
 [MATH]

31. CERN Press Release: CERN experiments observe particle consistent with long-sought Higgs boson.
Available at http://press.cern/press-releases/2012/07/cern-experiments-observe-particle-consistent-
long-sought-higgs-boson

32. Chaichian, M., Nelipa, N.F.: Introduction to Gauge Field Theories. Springer, Berlin/Heidelberg/New
York/Tokyo (1984)

33. Cheng, T.-P., Li, L.-F.: Gauge Theory of Elementary Particle Physics. Oxford University Press,
Oxford (1988)

34. Chevalley, C.: Theory of Lie Groups I. Princeton University Press, Princeton (1946)
[MATH]

35. Chevalley, C.: The Algebraic Theory of Spinors and Clifford Algebras. Collected Works, Vol. 2.
Springer, Berlin/Heidelberg (1997)

36. Chivukula, R.S.: The origin of mass in QCD. arXiv:hep-ph/0411198

37. Clay Mathematics Institute: Millenium problems. Yang–Mills and mass gap. Available at http://www.
claymath.org/millennium-problems/yang--mills-and-mass-gap

38. Costello, K.: Renormalization and Effective Field Theory. Mathematical Surveys and Monographs,
Vol. 170. American Mathematical Society, Providence, RI (2011)

39. Darling, R.W.R.: Differential Forms and Connections. Cambridge University Press, Cambridge (1994)
[MATH]

40. D’Auria, R., Ferrara, S., Lledó, M.A., Varadarajan, V.S.: Spinor algebras. J. Geom. Phys. 40 , 101–
129 (2001)
[MathSciNet][MATH]

41. Derdzinski, A.: Geometry of the Standard Model of Elementary Particles. Springer, Berlin/Heidelberg
(1992)
[MATH]

42. Dissertori, G., Knowles, I., Schmelling, M.: Quantum Chromodynamics. High Energy Experiments
and Theory. Oxford University Press, Oxford (2003)
[MATH]

43. Drexlin, G., Hannen, V., Mertens, S., Weinheimer, C.: Current direct neutrino mass experiments. Adv.
High Energy Phys. 2013 , Article ID 293986 (2013)

44. Dürr, S. et al.: Ab initio determination of light hadron masses. Science 322 , 1224–1227 (2008)

45. Duncan, A.: The Conceptual Framework of Quantum Field Theory. Oxford University Press, Oxford
(2013)
[MATH]

46. Dynkin, E.B.: Semisimple subalgebras of the semisimple Lie algebras. (Russian) Mat. Sbornik 30 ,
349–462 (1952); English translation: Am. Math. Soc. Transl. Ser. 2 6 , 111–244 (1957)
47. Elliott, C.: Gauge Theoretic Aspects of the Geometric Langlands Correspondence. Ph.D. Thesis,
Northwestern University (2016)

48. Englert, F., Brout R.: Broken symmetry and the mass of gauge vector mesons. Phys. Rev. Lett. 13 ,
321–323 (1964)
[MathSciNet]

49. Figueroa-O’Farrill, J.: Majorana Spinors. Lecture Notes. University of Edinburgh (2015)

50. Flory, M., Helling, R.C., Sluka, C.: How I learned to stop worrying and love QFT. arXiv:1201.2714
[math-ph]

51. Folland, G.B.: Quantum Field Theory. A Tourist Guide for Mathematicians. American Mathematical
Society, Providence, Rhodes Island (2008)
[MATH]

52. Freed, D.S.: Classical Chern–Simons theory, 1. Adv. Math. 113 , 237–303 (1995)
[MathSciNet][MATH]

53. Freed, D.S.: Five Lectures on Supersymmetry. American Mathematical Society, Providence, RI
(1999)
[MATH]

54. Freedman, D.Z., Van Proeyen, A.: Supergravity. Cambridge University Press, Cambridge (2012)
[MATH]

55. Friedrich, T.: Dirac Operators in Riemannian Geometry. American Mathematical Society, Providence,
RI (2000)
[MATH]

56. Fritzsch, H., Minkowski, P.: Unified interactions of leptons and hadrons. Ann. Phys. 93 , 193–266
(1975)
[MathSciNet]

57. Geiges, H.: An Introduction to Contact Topology. Cambridge University Press, Cambridge (2008)
[MATH]

58. Georgi, H.: The state of the art – gauge theories. In: Carlson, C.E. (ed.) Particles and Fields – 1974:
Proceedings of the Williamsburg Meeting of APS/DPF, pp. 575–582. AIP, New York (1975)

59. Georgi, H., Glashow, S.L.: Unity of all elementary-particle forces. Phys. Rev. Lett. 32 , 438–441
(1974)

60. Georgi, H.M., Glashow, S.L., Machacek, M.E., Nanopoulos, D.V.: Higgs Bosons from two-gluon
annihilation in proton-proton collisions. Phys. Rev. Lett. 40 692 (1978)

61. Georgi, H., Quinn, H.R., Weinberg, S.: Hierarchy of interactions in unified gauge theories. Phys. Rev.
Lett. 33 , 451–454 (1974)

62. Giunti, C., Kim, C.W.: Fundamentals of Neutrino Physics and Astrophysics. Oxford University Press,
Oxford (2007)
63. Glashow, S.L.: Trinification of all elementary particle forces. In: 5th Workshop on Grand Unification,
Providence, RI, April 12–14, 1984

64. Glashow, S.L., Iliopoulos, J., Maiani, L.: Weak interactions with lepton-hadron symmetry. Phys. Rev.
D 2 , 1285–1292 (1970)

65. Gleason, A.M.: Groups without small subgroups. Ann. Math. 56 , 193–212 (1952)
[MathSciNet][MATH]

66. Grimus, W., Rebelo, M.N.: Automorphisms in gauge theories and the definition of CP and P. Phys.
Rep. 281 , 239–308 (1997)
[MathSciNet]

67. Guralnik, G.S., Hagen, C.R., Kibble, T.W.B.: Global conservation laws and massless particles. Phys.
Rev. Lett. 13 , 585–587 (1964)

68. Gürsey, F., Ramond, P., Sikivie, P.: A universal gauge theory model based on E 6 . Phys. Lett. B 60 ,
177–180 (1976)

69. Haag, R.: Local Quantum Physics. Fields, Particles, Algebras. Springer, Berlin/ Heidelberg/New York
(1996)
[MATH]

70. Hall, B.C.: Lie Groups, Lie Algebras and Representations. An Elementary Introduction. Springer,
Cham Heidelberg/New York/Dordrecht/London (2016)

71. Halzen, F., Martin, A.D.: Quarks and Leptons. An Introductory Course in Modern Particle Physics.
Wiley, New York/Chichester/Brisbane/Toronto/Singapore (1984)

72. Hartanto, A., Handoko L.T.: Grand unified theory based on the SU(6) symmetry. Phys. Rev. D 71 ,
095013 (2005)

73. Harvey, R., Lawson, H.B.: Calibrated geometries. Acta Math. 148 , 47–157 (1982)
[MathSciNet][MATH]

74. Hatcher, A.: Vector bundles and K -theory. Version 2.1, May 2009

75. Heeck, J.: Interpretation of lepton flavor violation. Phys. Rev. D 95 , 015022 (2017)

76. Higgs, P.W.: Broken symmetries and the masses of gauge bosons. Phys. Rev. Lett. 13 , 508–509
(1964)
[MathSciNet]

77. Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, New York/
Dordrecht/Heidelberg/London (2012)
[MATH]

78. Hirsch, M.W.: Differential Topology. Springer, New York/Berlin/Heidelberg (1997)

79. Hoddeson, L., Brown, L., Riordan, M., Dresden, M. (ed.): The Rise of the Standard Model: Particle
Physics in the 1960s and 1970s. Cambridge University Press, Cambridge (1997)
80. Hollowood, T.J.: Renormalization Group and Fixed Points in Quantum Field Theory. Springer,
Heidelberg/New York/Dordrecht/London (2013)
[MATH]

81. Husemoller, D.: Fibre Bundles. Springer, New York (1994)

[MATH]

82. Klaczynski, L.: Haag’s Theorem in renormalisable quantum field theory. Ph.D. Thesis, Humboldt
Universität zu Berlin (2015)

83. Knapp, A.W.: Lie Groups Beyond an Introduction. Birkhäuser, Boston/Basel/Berlin (2002)
[MATH]

84. Kobayashi, S., Nomizu, K.: Foundations of Differential Geometry, Vol. I. Interscience Publishers,
New York/London (1963)
[MATH]

85. Kounnas, C., Masiero, A., Nanopoulos, D.V., Olive, K.A.: Grand Unification with and Without
Supersymmetry and Cosmological Implications. World Scientific, Singapore (1984)

86. Lancaster, T., Blundell, S. J.: Quantum Field Theory for the Gifted Amateur. Oxford University Press,
Oxford (2014)
[MATH]

87. Langacker, P.: Grand unified theories and proton decay. Phys. Rep. 72 , 185–385 (1981)

88. Lawson, H.B. Jr., Michelsohn, M.-L.: Spin Geometry. Princeton University Press, Princeton, NJ
(1989)
[MATH]

89. Lee, J.M.: Introduction to Smooth Manifolds. Springer, New York/Heidelberg/Dordrecht/ London
(2013)
[MATH]

90. Leigh, R.G., Strassler, M.J.: Duality of Sp (2 N c ) and SO ( N c ) supersymmetric gauge theories with
adjoint matter. Phys. Lett. B 356 , 492–499 (1995)
[MathSciNet]

91. Martin, S.P.: A supersymmetry primer. arXiv:hep-ph/9709356

92. Mayer, M.E.: Review: David D. Bleecker, Gauge theory and variational principles. Bull. Am. Math.
Soc. (N.S.) 9 , 83–92 (1983)

93. Meinrenken, E.: Clifford Algebras and Lie Theory. Springer, Berlin/Heidelberg (2013)
[MATH]

94. Milnor, J.: On manifolds homeomorphic to the 7-sphere. Ann. Math. 64 , 399–405 (1956)
[MathSciNet][MATH]

95. Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W. H. Freeman and Company, New York
(1973)
96. Mohapatra, R.N.: Unification and Supersymmetry. The Frontiers of Quark-Lepton Physics. Springer,
New York/Berlin/Heidelberg (2003)
[MATH]

97. Montgomery, D., Zippin, L.: Small subgroups of finite-dimensional groups. Ann. Math. 56 , 213–241
(1952)
[MathSciNet][MATH]

98. Moore, J.D.: Lectures on Seiberg–Witten Invariants. Springer, Berlin/Heidelberg/New York (2001)
[MATH]

99. Morgan, J.W.: The Seiberg–Witten Equations and Applications to the Topology of Smooth Four-
Manifolds. Princeton University Press, Princeton, NJ (1996)
[MATH]

100. Mosel, U.: Fields, Symmetries, and Quarks. Springer, Berlin/Heidelberg (1999)
[MATH]

101. Naber, G.L.: Topology, Geometry and Gauge Fields. Foundations. Springer, New York (2011)
[MATH]

102. Naber, G.L.: Topology, Geometry and Gauge Fields. Interactions. Springer, New York (2011)
[MATH]

103. Nakahara, M.: Geometry, Topology and Physics, 2nd edn. IOP Publishing Ltd, Bristol/ Philadelphia
(2003)
[MATH]

104. O’Raifeartaigh, L.: Group Structure of Gauge Theories. Cambridge University Press, Cambridge
(1986)
[MATH]

105. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Chin. Phys. C 40 ,
100001 (2016). http://www-pdg.lbl.gov/

106. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Chin.
Phys. C 40 , 100001 (2016). http://www-pdg.lbl.gov/

107. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Particle listings. Neutrino
mixing. Chin. Phys. C 40 , 100001 (2016). http://www-pdg.lbl.gov/

108. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 1. Physical constants. Chin. Phys. C 40 , 100001 (2016). http://www-pdg.lbl.gov/

109. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 10. Electroweak model and constraints on new physics. Chin. Phys. C 40 , 100001 (2016).
http://www-pdg.lbl.gov/

110. Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
plots. 12. The CKM quark-mixing matrix. Chin. Phys. C 40 , 100001 (2016). http://www-pdg.lbl.gov/
111.
Patrignani, C. et al. (Particle Data Group): 2016 Review of particle physics. Reviews, tables, and
 plots. 16. Grand unified theories. Chin. Phys. C 40 , 100001 (2016). http://www-pdg.lbl.gov/

112. Pich, A.: The Standard Model of electroweak interactions. arXiv:1201.0537 [hep-ph]

113. Quigg, C.: Gauge Theories of the Strong, Weak, and Electromagnetic Interactions. Westview Press,
Boulder, Colorado (1997)
[MATH]

114. Robinson, M.: Symmetry and the Standard Model. Mathematics and Particle Physics. Springer, New
York/Dordrecht/Heidelberg/London (2011)
[MATH]

115. Roe, J.: Elliptic Operators, Topology and Asymptotic Methods. Longman Scientific & Technical,
Harlow (1988)
[MATH]

116. Roman, P.: Introduction to Quantum Field Theory. Wiley, New York/London/Sydney/Toronto (1969)
[MATH]

117. Royal Swedish Academy of Sciences: The official web site of the nobel prize. https://www.nobelprize.
org/nobel_prizes/physics/

118. Royal Swedish Academy of Sciences: Asymptotic freedom and quantum chromodynamics: the key to
the understanding of the strong nuclear forces. Advanced information on the Nobel Prize in Physics, 5
October 2004. https://www.nobelprize.org/nobel_prizes/physics/laureates/2004/advanced.html

119. Royal Swedish Academy of Sciences: Class of Physics. Broken symmetries. Scientific Background on
the Nobel Prize in Physics 2008. https://www.nobelprize.org/nobel_prizes/physics/laureates/2008/
advanced.html

120. Royal Swedish Academy of Sciences: Class of Physics. The BEH-mechanism, interactions with short
range forces and scalar particles. Scientific Background on the Nobel Prize in Physics 2013. https://
www.nobelprize.org/nobel_prizes/physics/laureates/2013/advanced.html

121. Royal Swedish Academy of Sciences: Class of Physics. Neutrino oscillations. Scientific Background
on the Nobel Prize in Physics 2015. https://www.nobelprize.org/nobel_prizes/physics/laureates/2015/
advanced.html

122. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part I. Manifolds, Lie
Groups and Hamiltonian Systems. Springer Netherlands, Dordrecht (2013)

123. Rudolph, G., Schmidt, M.: Differential Geometry and Mathematical Physics. Part II. Fibre Bundles,
Topology and Gauge Fields. Springer Netherlands, Dordrecht (2017)

124. Ryder, L.H.: Quantum Field Theory. Cambridge University Press, Cambridge (1996)
[MATH]

125. Schwartz, M.D.: Quantum Field Theory and the Standard Model. Cambridge University Press,
Cambridge (2014)

126. Seiberg, N.: Five dimensional SUSY field theories, non-trivial fixed points and string dynamics. Phys.
Lett. B 388 , 753–760 (1996)
 [MathSciNet]

127. Seiberg, N., Witten, E.: Electric-magnetic duality, monopole condensation, and confinement in N = 2
supersymmetric Yang–Mills theory. Nucl. Phys. B 426 , 19–52 (1994). Erratum: Nucl. Phys. B 430 ,
485–486 (1994)

128. Seiberg, N., Witten, E.: Monopoles, duality and chiral symmetry breaking in N = 2 supersymmetric
QCD. Nucl. Phys. B 431 , 484–550 (1994)
[MathSciNet][MATH]

129. Sepanski, M.R.: Compact Lie Groups. Springer Science+Business Media LLC, New York (2007)
[MATH]

130. Serre, J.-P.: Lie Algebras and Lie Groups. 1964 Lectures given at Harvard University. Springer,
Berlin/Heidelberg (1992)

131. Slansky, R.: Group theory for unified model building. Phys. Rep. 79 , 1–128 (1981)
[MathSciNet]

132. Srednicki, M.: Quantum Field Theory. Cambridge University Press, Cambridge (2007)
[MATH]

133. Steenrod, N.: The Topology of Fibre Bundles. Princeton University Press, Princeton (1951)
[MATH]

134. Streater, R.F., Wightman, A.S.: PCT, Spin and Statistics, and All That. Princeton University Press,
Princeton, NJ (2000)
[MATH]

135. Tao, T.: Hilbert’s Fifth Problem and Related Topics. Graduate Studies in Mathematics, Vol. 153.
American Mathematical Society, Providence, RI (2014)

136. Taubes, C.H.: Differential Geometry. Bundles, Connections, Metrics and Curvature. Oxford
University Press, Oxford (2011)
[MATH]

137. Thomson, M.: Modern Particle Physics. Cambridge University Press, Cambridge (2013)

138. Vafa, C., Zwiebach, B.: N = 1 dualities of SO and USp gauge theories and T-duality of string theory.
Nucl. Phys. B 506 , 143–156 (1997)
[MathSciNet][MATH]

139. van den Ban, E.P.: Notes on quotients and group actions. Fall 2006. Universiteit Utrecht

140. Van Proeyen, A.: Tools for supersymmetry. arXiv:hep-th/9910030

141. van Vulpen, I.: The Standard Model Higgs boson. Part of the Lecture Particle Physics II, University of
Amsterdam Particle Physics Master 2013–2014

142. Warner, F.W.: Foundations of Differentiable Manifolds and Lie Groups. Springer, New York (2010)

143. Weinberg, S.: The Quantum Theory of Fields, Vol. I. Foundations. Cambridge University Press,
 Cambridge (1995)

144. Weinberg, S.: The Quantum Theory of Fields, Vol. II. Modern Applications. Cambridge University
Press, Cambridge (1996)
[MATH]

145. Weinberg, S.: The Quantum Theory of Fields, Vol. III. Supersymmetry. Cambridge University Press,
Cambridge (2005)
[MATH]

146. Wess, J., Bagger, J.: Supersymmetry and Supergravity, 2nd edn. Princeton University Press, Princeton,
NJ (1992)
[MATH]

147. Wilczek, F.: Decays of heavy vector mesons into Higgs particles. Phys. Rev. Lett. 39 , 1304 (1977)

148. Witten, E.: Quest for unification. arXiv:hep-ph/0207124

149. Witten, E.: Chiral ring of Sp ( N ) and SO ( N ) supersymmetric gauge theory in four dimensions.
Chin. Ann. Math. 24 , 403 (2003)
[MathSciNet][MATH]

150. Witten, E.: Newton lecture 2010: String theory and the universe. Available at http://www.iop.org/
resources/videos/lectures/page_44292.html Cited 20 Nov 2016

151. Yamamoto, K.: SU(7) Grand Unified Theory. Ph.D. thesis, Kyoto University (1981)

152. Zhang, F.: Quaternions and matrices of quaternions. Linear Algebra Appl. 251 , 21–57 (1997)
[MathSciNet][MATH]

153. Ziller, W.: Lie Groups. Representation theory and symmetric spaces. Lecture Notes, University of
Pennsylvania, Fall 2010. Available at https://www.math.upenn.edu/~wziller/

Index
4-Fermi interaction
absorption
action
Yang–Mills
algebra
anticommutator
associative
-graded
direct sum
homomorphism
isomorphism
 representation
representation, faithful
tensor product
Clifford
basis unitary representation
chiral representation
Clifford relation
complexified
dimension
even part
for standard symmetric bilinear forms
in low dimensions
odd part
periodicity complex mod2
periodicity real mod8
universal property
Weyl representation
commutator
division
Lie
abelian
bracket
Cartan subalgebra
center
commutator
compact
general linear group
graded
Heisenberg
ideal
linear groups
of a Lie group
orthogonal group
semisimple
simple
special linear groups
special orthogonal group
special unitary group
spin group
 structure constants
subalgebra
subalgebra, intersection
symplectic group
unitary group
normed
Poincaré
super-Poincaré
supersymmetry
tensor
amplitude
scattering
transition
angle
weak mixing
Weinberg
annihilation
anomaly
chiral
gauge
antimatter
antiparticle
asymptotic freedom
atlas
associated vector bundle
adapted
bundle
formal
transition function
transition function, cocycle condition
manifold
principal bundle
transition function
vector bundle
transition function
automorphism
group
inner
Lie algebra
 Lie group
principal bundle
axial symmetry
baryon
number
conservation
violation
beta function
beta-decay ( β -decay)
Bianchi identity
boson
BPS equation
bundle
G -structure
base manifold
clutching construction
fibre
fibre sum
flat
frame
isomorphism
locally trivial
Möbius strip
map
morphism
non-trivial
principal
morphism
over contractible manifold
reduction of the structure group
reduction under group homomorphism
projection
pullback
restriction over submanifold
section
global
local
of vector bundle
smoothly compatible charts
 total space
trivial
trivialization
local
unit sphere
vector
adjoint
associated
bundle metric
complex conjugate
endomorphism
epimorphism
homomorphism
horizontal tangent
isomorphism
line bundle
monomorphism
normal
orientable
over contractible manifold
stably trivial
subbundle
tangent bundle
tautological bundle
tautological line bundle
twisted forms
vertical tangent
Cabibbo–Kobayashi–Maskawa (CKM) matrix
KM phase
mixing angles
Casimir
invariant
operator
CERN
chain rule
charge
central
conjugation
conservation
 electric
elementary electric
operator
quantized
weak hypercharge
weak isospin
charged
fermion
particle
scalar
section
charged lepton flavour violation (CLFV)
chart
bundle
formal
manifold
principal bundle
submanifold
vector bundle
Chern class
Chern–Simons
action
theory
chiral
limit
symmetry
breaking
chirality
element
operator
mathematical
physical
classical contribution
Clay Millennium Prize Problem
codifferential
covariant
colour
confinement
space
conjugate
complex number
quaternion
conjugation
connection
1-form
local
canonical flat
Ehresmann
constant
Fermi
fine-structure
coordinate
change of coordinates
system
transformation
correlator
coupling constant
effective
electric
running
strong
covariant differential
CP
problem
strong
transformation
violation
current
charged current interaction
neutral current interaction
curvature
2-form
local 2-form
cutoff
dark matter
decay
derivative
covariant
 and interaction between particles
compatible with bundle metrics
exterior
directional
exterior
Lie
diffeomorphism
differentiable
map
structure
differential
of smooth map
on forms
differential form
twisted with a vector space
with values in a vector space
differential topology
Dirac operator
index
distribution
integrable
Donaldson theory
dual
basis
space
eigenstate
current
mass
weak
eightfold way
Einstein summation convention
electron
electroweak
interaction
theory
emission
energy
electric field
equivalence relation
Euclidean space
Euler class
exotic 7-spheres
expansion
asymptotic
series
exponential map
embedded Lie subgroup
linear group
determinant
torus
exterior covariant derivative
family
Fermilab
fermion
Feynman diagram
β -decay
electroweak interaction vertices
gauge bosons
gauge field and scalar field
interaction between electroweak gauge bosons
interaction between gluons
interaction between Higgs bosons and gauge bosons
interaction between leptons and Higgs boson
interaction between quarks and Higgs boson
interactions between X , Y -bosons and fermions
loop
neutral Kaon decay
strong interaction vertex
tree
vacuum polarization
fibre
field
chromo-electric
chromo-magnetic
complex scalar
electric
generalized electric
generalized magnetic
 magnetic
real scalar
strength
local
fixed point
set
flattener
flavour
changing vertex
lepton
neutrino
quark
symmetry
foliation
form
alternating
bilinear
conjugate symmetric
contraction
positive definite
sesquilinear
symmetric
with values in a vector space
gauge
boson
B
W
W0
X
Y
Z
broken
gluon
mass
massive
massless
off-diagonal action
photon
unbroken
 vector
weak
boson field
gluon
hypercharge
weak
field
local
global
group
Standard Model
local
sector
gluon
hypercharge
weak
theory
chiral
transformation
global
local
physical
rigid physical
unitary
gaugino
bino
gluino
photino
wino
zino
Gell-Mann–Nishijima formula
generation
changing vertex
generator
broken
unbroken
geodesic
ghost state
glueballs
gluodynamics
gluon fusion process
Grand Unification
possible group
Grand Unified Theory (GUT)
E6
SU(3) 3
SU(5)
SU(6)
SU(7)
Spin(10)
Grassmann manifold
Grassmannian
Green’s function
group
action
continuous
effective
equivariant map
faithful
free
infinitesimal
isomorphism
isotropy group
isotropy subalgebra
left
linear
on spheres
orbit
principal
right
simply transitive
smooth
stabilizer
transitive
diffeomorphism
Lie
G2
 abelian
classical
closed subgroup
compact
connected component of the identity
dimension
embedded subgroup
exceptional
general linear
Heisenberg
homotopy groups
immersed subgroup
linear
matrix
maximal torus
non-abelian
non-compact
orthogonal
pin
product
pseudo-orthogonal
pseudo-orthogonal, orthochronous
pseudo-orthogonal, proper
pseudo-orthogonal, proper orthochronous
rank
semisimple
simple
special linear
special orthogonal
special unitary
spin
spin, orthochronous
symplectic
uniqueness of smooth structure
unitary
little
topological
transformation
hadron
Hamiltonian operator
handedness
harmonic form
harmonic oscillator
Hausdorff space
Hessian
Higgs
boson
absorption
and vector boson decay
and vector boson fusion
mass
bundle
condensate
field
electric charge
self-coupling
shifted
Taylor expansion
mechanism
Brout–Englert–Higgs
potential
Strahlung
vector space
Higgsino
Hilbert space
Hodge star operator on forms
holonomy
homogeneous space
reductive
homomorphism
Lie algebra
induced
trivial
Lie group
SL(2, ) to SO + (1, 3)
SU(2) to SO(3)
SU(2) × SU(2) to SO(4)
 continuous
differential
image
kernel
trivial
Hopf
action
fibration
horizontal lift
immersion
inertial system
instanton
interaction
and covariant derivative
between fermions and gauge bosons
between gauge bosons
between Higgs bosons
between scalar field and gauge bosons
electromagnetic
electroweak
strong
vertex
electroweak
strong
weak
Jacobi identity
super
Jarlskog invariant
Killing form
Klein–Gordon equation
Koszul formula
Lagrangian
anharmonic term
Dirac
effective
Einstein–Hilbert
gauge invariance
harmonic term
Higgs
 kinetic term
Klein–Gordon
coupled to gauge field
free
Standard Model
Yang–Mills
electroweak gauge bosons
gluons
Yang–Mills–Dirac
Yang–Mills–Dirac–Higgs–Yukawa
Yang–Mills–Higgs
Yang–Mills–Klein–Gordon
Yukawa
Large Hadron Collider (LHC)
Leibniz rule
lens space
lepton
charged
flavour universality
mass
mass (experimental)
number
conservation
electron
electron
muon
tau
total
violation (LNV)
sector
leptoquark
Levi-Civita connection
Lichnerowicz–Weitzenböck formula
lifetime
locally Euclidean
Lorentz
frame
group
orthochronous
 proper
proper orthochronous
manifold
spin
orthochronous spin group
transformation
LSZ reduction formula
M-theory
manifold
closed
differentiable
oriented
product
smooth
topological
mapping torus
mass
bare
form
gap
generation
for fermions
for gauge boson
pairing
term
Dirac
Klein–Gordon
Majorana
mass hierarchy
inverted
normal
matrices
gamma
mathematical
physical
raising index
Gell-Mann
Pauli
matrix
 antisymmetric
determinant
Hermitian
skew-Hermitian
skew-symmetric
symmetric
trace
transpose
Maurer–Cartan form
Maxwell’s equations
meson
metric
Lie group
bi-invariant
left-invariant
right-invariant
pseudo-Riemannian
Riemannian
Mexican hat
minimal coupling
Minimal Supersymmetric Standard Model (MSSM)
Minkowski
metric
East Coast
West Coast
spacetime
moduli space
instanton
Yang–Mills
momentum
monodromy
multiplet
chiral
complex scalar
gauge
spinor
vector
muon
Nambu–Goldstone boson
neutrino
electron
mass
mixing
muon
oscillation
right-handed
solar
sterile
tau
neutron
Nobel Prize in Physics
1957
1969
1977
1979
1984
1999
2004
2008
2013
2015
non-perturbative
norm
complex number
of differential forms
of pseudo-Euclidean scalar product
quaternion
nucleon
number operator
field
harmonic oscillator
observable
octonions
on-shell
operator
lowering
raising
order
 leading (LO)
next-to-leading (NLO)
next-to-next-to-leading (NNLO)
orientation
time-
parallel transport
in associated vector bundle
in principal bundle
parallelizable
manifold
sphere
parameters of the Standard Model
parity
inversion
particle
annihilation
collider
creation
elementary
matter
point
virtual
parton
distribution function (PDF)
path integral
measure
path-ordered exponential
pentaquark
perturbation theory
perturbative
picture
Heisenberg
Schrödinger
Planck constant
Poincaré transformation
Pontecorvo–Maki–Nakagawa–Sakata (PMNS) matrix
positron
probability
projection
 canonical
stereographic
projective space
complex
quaternionic
real
proton
decay
lifetime
pullback
push forward
quantum chromodynamics (QCD)
lattice
quantum correction
quantum electrodynamics (QED)
quantum field
quantum field theory (QFT)
algebraic
constructive
quantum gravity
quantum mechanics (QM)
quark
antiquark
bottom
charm
down
mass
mass (experimental)
mixing
sea
sector
strange
top
up
valence
quaternionic matrix
adjoint
determinant
inverse
quaternions
imaginary
real
quotient space
fundamental group
smooth structure and submersion
radiation
regular
point
value
regularization
relativity
general theory
special theory
renormalizable
non-
renormalization
representation
adjoint
direct product
Lie algebra
Lie group
linear group
branching rule
decomposition
doublet
fermions in the Standard Model
induced
irreducible
isodoublet
isosinglet
isotropy
Lie algebra
equivariant map
faithful
integrability
intertwining map
isomorphism
linear algebra constructions
 morphism
restricted
skew-Hermitian
skew-symmetric
Lie group
SL(2, ) Weyl spinor
equivariant map
faithful
intertwining map
isomorphism
linear algebra constructions
morphism
orthogonal
quaternionic structure
real structure
restricted
unitary
linear group
defining
fundamental
standard
Poincaré algebra
reducible
singlet
super-Poincaré algebra
triplet
colour
trivial
unitary
weight lattice
scalar product
G -invariant
L 2 -scalar product of forms
L 2 -scalar product of spinors
L 2 -scalar product of twisted forms
L 2 -scalar product of twisted spinors
Ad-invariant
Euclidean
 Hermitian
of forms
of twisted forms
pseudo-Euclidean
semi-Euclidean
standard
scattering experiment
section
sector
lepton
quark
seesaw mechanism
Seiberg–Witten theory
self-duality
anti-self-dual
self-dual
semi-classical approximation
skew field
slepton
speed of light
sphere
spin
connection
covariant derivative
compatible with bundle metrics
compatible with Levi-Civita connection
twisted
twisted chiral
manifold
structure
spinor
anticommuting
bundle
Dirac
Dirac metric
Majorana metric
twisted
twisted chiral
Weyl
 charge conjugate
charge conjugation matrix
Clifford multiplication
form
mathematical
physical
Dirac
form
Dirac
Majorana
Majorana
Majorana conjugate
Majorana–Weyl
on manifold
representation
Dirac
real
Weyl
symplectic Majorana
symplectic Majorana–Weyl
twisted
Weyl
left-handed (positive)
right-handed (negative)
squark
Standard Model
state
bound
Stiefel manifold
Stiefel–Whitney class
structure equation
local
structure group
submanifold
embedded
immersed
submersion
and local sections
superalgebra
supergravity
superpartner
superstring theory
supersymmetry
breaking
field theory
generator
Grand Unification
local
rigid
symmetric bilinear form
signature
standard
orthonormal basis
symmetry
approximate
bosonic
breaking
conformal
discrete
fermionic
full group
gauge
global
Lorentz
tangent
space
horizontal
vertical
vector
tau
tensor
tetrad
tetraquarks
Theorem
Adams
Ado
Atiyah–Singer Index Theorem
Cartan’s Criterion for Semisimplicity
 Cartan’s Theorem on Closed Subgroups
Cartan–Dieudonné
Cayley
Frobenius
Godement
Haag’s
Haag–Łopusza´nski–Sohnius
Killing–Cartan Classification
Lie’s Third
Peter–Weyl
Regular Point
Regular Value
Schur’s Lemma
Wightman Reconstruction
time-orientability
translation
group action
left
right
Lie group
left
right
tree level
Trinification
unitarity
vacuum
configuration
expectation value (vev)
gauge
manifold
polarization
space of vacua
vector
vector field
basis
commutator
components
coordinate
flow
 global
local
fundamental
integral curve
left-invariant
flow
integral curve
vector space
complex conjugate
vielbein
volume form
canonical
right-invariant
standard
wave packet
wedge product
Wilson loop
winding number
Yang–Mills
connection
equation
theory
Yang–Mills–Higgs equations
Yukawa
coupling
form
leptons
quarks