Nuclear and Particle Physics

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
Nuclear and Particle Physics

B. R. Martin
Department of Physics and Astronomy, University College London
Copyright # 2006 John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester,
West Sussex PO19 8SQ, England

Telephone (+44) 1243 779777

Email (for orders and customer service enquiries): [email protected]

Visit our Home Page on www.wileyeurope.com or www.wiley.com

All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or
otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a
licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1T 4LP, UK,
without the permission in writing of the Publisher. Requests to the Publisher should be addressed to the
Permissions Department, John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex
PO19 8SQ, England, or emailed to [email protected], or faxed to (+44) 1243 770620.
Designations used by companies to distinguish their products are often claimed as trademarks. All brand
names and product names used in this book are trade names, service marks, trademarks or registered
trademarks of their respective owners. The Publisher is not associated with any product or vendor
mentioned in this book.
This publication is designed to provide accurate and authoritative information in regard to the subject
matter covered. It is sold on the understanding that the Publisher is not engaged in rendering
professional services. If professional advice or other expert assistance is required, the services of a
competent professional should be sought.

Other Wiley Editorial Offices

John Wiley & Sons Inc., 111 River Street, Hoboken, NJ 07030, USA
Jossey-Bass, 989 Market Street, San Francisco, CA 94103-1741, USA
Wiley-VCH Verlag GmbH, Boschstr. 12, D-69469 Weinheim, Germany
John Wiley & Sons Australia Ltd, 42 McDougall Street, Milton, Queensland 4064, Australia

John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop # 02-01, Jin Xing Distripark, Singapore 129809
John Wiley & Sons Canada Ltd, 22 Worcester Road, Etobicoke, Ontario, Canada M9W 1L1
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not
be available in electronic books.
Library of Congress Cataloging-in-Publication Data
Martin, B. R. (Brian Robert)
Nuclear and particle physics/B. R. Martin.
p. cm.
ISBN-13: 978-0-470-01999-3 (HB)
ISBN-10: 0-470-01999-9 (HB)
ISBN-13: 978-0-470-02532-1 (pbk.)
ISBN-10: 0-470-02532-8 (pbk.)
1. Nuclear physics–Textbooks. 2. Particle physics–Textbooks. I. Title.
QC776.M34 2006
539.70 2–dc22
2005036437
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
ISBN-13 978-0 470 01999 9 (HB) ISBN-10 0 470 01999 9 (HB)
978-0 470 02532 8 (PB) 0 470 02532 8 (PB)
Typeset in 10.5/12.5pt Times by Thomson Press (India) Limited, New Delhi
Printed and bound in Great Britain by Antony Rowe Ltd., Chippenham, Wiltshire
This book is printed on acid-free paper responsibly manufactured from sustainable forestry
in which at least two trees are planted for each one used for paper production.
To Claire
Contents

Preface xi

Notes xiii

Physical Constants and Conversion Factors xv

1 Basic Concepts 1
1.1 History 1
1.1.1 The origins of nuclear physics 1
1.1.2 The emergence of particle physics: the standard model and hadrons 4
1.2 Relativity and antiparticles 7
1.3 Symmetries and conservation laws 9
1.3.1 Parity 10
1.3.2 Charge conjugation 12
1.4 Interactions and Feynman diagrams 13
1.4.1 Interactions 13
1.4.2 Feynman diagrams 15
1.5 Particle exchange: forces and potentials 17
1.5.1 Range of forces 17
1.5.2 The Yukawa potential 19
1.6 Observable quantities: cross sections and decay rates 20
1.6.1 Amplitudes 21
1.6.2 Cross-sections 23
1.6.3 Unstable states 27
1.7 Units: length, mass and energy 29
Problems 30

2 Nuclear Phenomenology 33
2.1 Mass spectroscopy and binding energies 33
2.2 Nuclear shapes and sizes 37
2.2.1 Charge distribution 37
2.2.2 Matter distribution 42
2.3 Nuclear instability 45
2.4 Radioactive decay 47
2.5 Semi-empirical mass formula: the liquid drop model 50
2.6 -decay phenomenology 55
2.6.1 Odd-mass nuclei 55
2.6.2 Even-mass nuclei 58
viii CONTENTS

2.7 Fission 59
2.8 -decays 62
2.9 Nuclear reactions 62
Problems 67

3 Particle Phenomenology 71
3.1 Leptons 71
3.1.1 Lepton multiplets and lepton numbers 71
3.1.2 Neutrinos 74
3.1.3 Neutrino mixing and oscillations 76
3.1.4 Neutrino masses 79
3.1.5 Universal lepton interactions – the number of neutrinos 84
3.2 Quarks 86
3.2.1 Evidence for quarks 86
3.2.2 Quark generations and quark numbers 89
3.3 Hadrons 92
3.3.1 Flavour independence and charge multiplets 92
3.3.2 Quark model spectroscopy 96
3.3.3 Hadron masses and magnetic moments 102
Problems 108

4 Experimental Methods 111

4.1 Overview 111
4.2 Accelerators and beams 113
4.2.1 DC accelerators 113
4.2.2 AC accelerators 115
4.2.3 Neutral and unstable particle beams 122
4.3 Particle interactions with matter 123
4.3.1 Short-range interactions with nuclei 123
4.3.2 Ionization energy losses 125
4.3.3 Radiation energy losses 128
4.3.4 Interactions of photons in matter 129
4.4 Particle detectors 131
4.4.1 Gas detectors 131
4.4.2 Scintillation counters 137
4.4.3 Semiconductor detectors 138
4.4.4 Particle identification 139
4.4.5 Calorimeters 142
4.5 Layered detectors 145
Problems 148

5 Quark Dynamics: the Strong Interaction 151

5.1 Colour 151
5.2 Quantum chromodynamics (QCD) 153
5.3 Heavy quark bound states 156
5.4 The strong coupling constant and asymptotic freedom 160
5.5 Jets and gluons 164
5.6 Colour counting 166
 CONTENTS ix

5.7 Deep inelastic scattering and nucleon structure 168

Problems 177

6 Electroweak Interactions 181

6.1 Charged and neutral currents 181
6.2 Symmetries of the weak interaction 182
6.3 Spin structure of the weak interactions 186
6.3.1 Neutrinos 187
6.3.2 Particles with mass: chirality 189
6.4 W and Z0 bosons 192
6.5 Weak interactions of hadrons 194
6.5.1 Semileptonic decays 194
6.5.2 Neutrino scattering 198
6.6 Neutral meson decays 201
6.6.1 CP violation 202
6.6.2 Flavour oscillations 206
6.7 Neutral currents and the unified theory 208
Problems 213

7 Models and Theories of Nuclear Physics 217

7.1 The nucleon – nucleon potential 217
7.2 Fermi gas model 220
7.3 Shell model 223
7.3.1 Shell structure of atoms 223
7.3.2 Nuclear magic numbers 225
7.3.3 Spins, parities and magnetic dipole moments 228
7.3.4 Excited states 230
7.4 Non-spherical nuclei 232
7.4.1 Electric quadrupole moments 232
7.4.2 Collective model 236
7.5 Summary of nuclear structure models 236
7.6 -decay 238
7.7 -decay 242
7.7.1 Fermi theory 242
7.7.2 Electron momentum distribution 244
7.7.3 Kurie plots and the neutrino mass 246
7.8 -emission and internal conversion 248
7.8.1 Selection rules 248
7.8.2 Transition rates 250
Problems 252

8 Applications of Nuclear Physics 255

8.1 Fission 255
8.1.1 Induced fission – fissile materials 255
8.1.2 Fission chain reactions 258
8.1.3 Nuclear power reactors 260
8.2 Fusion 266
8.2.1 Coulomb barrier 266
x CONTENTS

8.2.2 Stellar fusion 267

8.2.3 Fusion reaction rates 270
8.2.4 Fusion reactors 273
8.3 Biomedical applications 278
8.3.1 Biological effects of radiation: radiation therapy 278
8.3.2 Medical imaging using radiation 282
8.3.3 Magnetic resonance imaging 290
Problems 294

9 Outstanding Questions and Future Prospects 297

9.1 Particle physics 297
9.1.1 The Higgs boson 297
9.1.2 Grand unification 300
9.1.3 Supersymmetry 304
9.1.4 Particle astrophysics 307
9.2 Nuclear physics 315
9.2.1 The structure of hadrons and nuclei 316
9.2.2 Quark–gluon plasma, astrophysics and cosmology 320
9.2.3 Symmetries and the standard model 323
9.2.4 Nuclear medicine 324
9.2.5 Power production and nuclear waste 326

Appendix A: Some Results in Quantum Mechanics 331

A.1 Barrier penetration 331
A.2 Density of states 333
A.3 Perturbation theory and the Second Golden Rule 335

Appendix B: Relativistic Kinematics 339

B.1 Lorentz transformations and four-vectors 339
B.2 Frames of reference 341
B.3 Invariants 344
Problems 345

Appendix C: Rutherford Scattering 349

C.1 Classical physics 349
C.2 Quantum mechanics 352
Problems 354

Appendix D: Solutions to Problems 355

References 393

Bibliography 397

Index 401
Preface

It is common practice to teach nuclear physics and particle physics together in an

introductory course and it is for such a course that this book has been written. The
material is presented so that different selections can be made for a short course of
about 25–30 lectures depending on the lecturer’s preferences and the students’
backgrounds. On the latter, students should have taken a first course in quantum
physics, covering the traditional topics in non-relativistic quantum mechanics and
atomic physics. A few lectures on relativistic kinematics would also be useful, but
this is not essential as the necessary background is given in appendix B and is only
used in a few places in the book. I have not tried to be rigorous, or present proofs
of all the statements in the text. Rather, I have taken the view that it is more
important that students see an overview of the subject which for many – possibly
the majority – will be the only time they study nuclear and particle physics. For
future specialists, the details will form part of more advanced courses. Never-
theless, space restrictions have still meant that it has been necessary to make a
choice of topics covered and doubtless other, equally valid, choices could have
been made. This is particularly true in Chapter 8, which deals with applications of
nuclear physics, where I have chosen just three major areas to discuss. Nuclear and
particle physics have been, and still are, very important parts of the entire subject
of physics and its practitioners have won an impressive number of Nobel Prizes.
For historical interest, I have noted in the footnotes many of the awards for work
related to the field.
Some parts of the book dealing with particle physics owe much to a previous book,
Particle Physics, written with Graham Shaw of Manchester University, and I am
grateful to him and the publisher, John Wiley and Sons, for permission to adapt some
of that material for use here. I also thank Colin Wilkin for comments on all the chapters
of the book, David Miller and Peter Hobson for comments on Chapter 4 and Bob
Speller for comments on the medical physics section of Chapter 8. If errors or
misunderstandings still remain (and any such are of course due to me alone) I would be
grateful to hear about them. I have set up a website (www.hep.ucl.ac.uk/brm/
npbook.html) where I will post any corrections and comments.

Brian R. Martin
January 2006
Notes

References
References are referred to in the text in the form Ab95, where Ab is the start of the
first author’s surname and 1995 is the year of publication. A list of references with
full publication details is given at the end of the book.

Data
Data for particle physics may be obtained from the biannual publications of the
Particle Data Group (PDG) and the 2004 edition of the PDG definitive Review
of Particle Properties is given in Ei04. The PDG Review is also available at
http://pdg.lbl.gov and this site contains links to other sites where compilations of
particle data may be found. Nuclear physics data are available from a number of
sources. Examples are: the combined Isotopes Project of the Lawrence Berkeley
Laboratory, USA, and the Lund University Nuclear Data WWW Service, Sweden
(http://ie.lbl.gov/toi.html), the National Nuclear Data Center (NNDC) based
at Brookhaven National Laboratory, USA (http://www.nndc.bnl.gov), and
the Nuclear Data Centre of the Japan Atomic Energy Research Institute
(http://www.nndc.tokai.jaeri.go.jp). All three sites have extensive links to other
data compilations. It is important that students have some familiarity with these
data compilations.

Problems
Problems are provided for all chapters and appendices except Chapter 9 and
Appendices A and D. They are an integral part of the text. The problems are
mainly numerical and require values of physical constants that are given in a table
following these notes. A few also require input data that may be found in the
references given above. Solutions to all the problems are given in Appendix D.
xiv NOTES

Illustrations
Some illustrations in the text have been adapted from, or are loosely based on,
diagrams that have been published elsewhere. In a few cases they have been
reproduced exactly as previously published. In all cases this is stated in the
captions. I acknowledge, with thanks, permission to use such illustrations from the
relevant copyright holders.
Physical Constants
and Conversion Factors

Quantity Symbol Value

Speed of light in vacuum c 2:998
108 ms1

Planck’s constant h 4:136
1024 GeV s
h  h=2
 6:582
1025 GeV s

hc 1:973
1016 GeV m
hcÞ2
ð 3:894
1031 GeV2 m2
electron charge (magnitude) e 1:602
1019 C
Avogadro’s number NA 6:022
1026 kg-mole1
Boltzmann’s constant kB 8:617
1011 MeV K1
electron mass me 0:511 MeV=c2
proton mass mp 0:9383 GeV=c2
neutron mass mn 0:9396 GeV=c2
W boson mass MW 80:43 GeV=c2
Z boson mass MZ 91:19 GeV=c2
atomic mass unit 1
u  ð12 mass12 C atomÞ 931:494 MeV=c2
Bohr magneton B  e
h=2me 5:788
1011 MeV T1
Nuclear magneton N  e
h=2mp 3:152
1014 MeV T1
gravitational constant GN hcðGeV=c2 Þ2
6:709
1039 
fine structure constant   e2 =4"0 
hc 7:297
103 ¼ 1=137:04
Fermi coupling constant hcÞ3
GF =ð 1:166
105 GeV2
strong coupling constant s ðMZ c2 Þ 0.119

1 eV ¼ 1:602
1019 J 1 eV=c2 ¼ 1:783
1036 kg

15
1 fermi ¼ 1 fm  10 m 1 barn ¼ 1 b  1028 m2
1 Tesla ¼ 1T ¼ 0:561
1030 MeV=c2 C1 s1 1 year ¼ 3:1536
107 s
1
Basic Concepts

1.1 History
Although this book will not follow a strictly historical development, to ‘set the
scene’ this first chapter will start with a brief review of the most important
discoveries that led to the separation of nuclear physics from atomic physics as a
subject in its own right and later work that in its turn led to the emergence of
particle physics from nuclear physics.1

1.1.1 The origins of nuclear physics

Nuclear physics as a subject distinct from atomic physics could be said to date
from 1896, the year that Henri Becquerel observed that photographic plates were
being fogged by an unknown radiation emanating from uranium ores. He had
accidentally discovered radioactivity: the fact that some nuclei are unstable and
spontaneously decay. In the years that followed, the phenomenon was extensively
investigated, notably by the husband and wife team of Pierre and Marie Curie and
by Ernest Rutherford and his collaborators,2 and it was established that there were
three distinct types of radiation involved: these were named (by Rutherford) -, -
and -rays. We know now that -rays are bound states of two protons and two
neutrons (we will see later that they are the nuclei of helium atoms), -rays are
electrons and -rays are photons, the quanta of electromagnetic radiation, but the
historical names are still commonly used.
1
An interesting account of the early period, with descriptions of the personalities involved, is given in Se80.
An overview of the later period is given in Chapter 1 of Gr87.
2
The 1903 Nobel Prize in Physics was awarded jointly to Becquerel for his discovery and to Pierre and Marie
Curie for their subsequent research into radioactivity. Rutherford had to wait until 1908, when he was
awarded the Nobel Prize in Chemistry for his ‘investigations into the disintegration of the elements and the
chemistry of radioactive substances’.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
2 CH1 BASIC CONCEPTS

At about the same time as Becquerel’s discovery, J. J. Thomson was extending

the work of Perrin and others on the radiation that had been observed to occur
when an electric field was established between electrodes in an evacuated glass
tube, and in 1897 he was the first to definitively establish the nature of these
‘cathode rays’. We now know the emanation consists of free electrons, (the name
‘electron’ had been coined in 1894 by Stoney) denoted e
(the superscript denotes
the electric charge) and Thomson measured their mass and charge.3 The view of the
atom at that time was that it consisted of two components, with positive and negative
electric charges, the latter now being the electrons. Thomson suggested a model
where the electrons were embedded and free to move in a region of positive charge
filling the entire volume of the atom – the so-called ‘plum pudding model’.
This model could account for the stability of atoms, but could not account for
the discrete wavelengths observed in the spectra of light emitted from excited
atoms. Neither could it explain the results of a classic series of experiments
performed in 1911 at the suggestion of Rutherford by his collaborators, Geiger and
Marsden. These consisted of scattering
-particles by very thin gold foils. In the
Thomson model, most of the
-particles would pass through the foil, with only a
few suffering deflections through small angles. Rutherford suggested they should
look for large-angle scattering and to their surprise they found that some particles
were indeed scattered through very large angles, even greater than 90 . Rutherford
showed that this behaviour was not due to multiple small-angle deflections, but
could only be the result of the
-particles encountering a very small positively
charged central nucleus. (The reason for these two different behaviours is
discussed in Appendix C.)
To explain the results of these experiments Rutherford formulated a ‘planetary’
model, where the atom was likened to a planetary system, with the electrons (the
‘planets’) occupying discrete orbits about a central positively charged nucleus (the
‘Sun’). Because photons of a definite energy would be emitted when electrons
moved from one orbit to another, this model could explain the discrete nature of
the observed electromagnetic spectra when excited atoms decayed. In the simplest
case of hydrogen, the nucleus is a single proton (p) with electric charge þe, where e
is the magnitude of the charge on the electron4, orbited by a single electron.
Heavier atoms were considered to have nuclei consisting of several protons. This
view persisted for a long time and was supported by the fact that the masses of
many naturally occurring elements are integer multiples of a unit that is about
1 per cent smaller than the mass of the hydrogen atom. Examples are carbon and
nitrogen, with masses of 12.0 and 14.0 in these units. However, it could not explain
why not all atoms obeyed this rule. For example, chlorine has a mass of 35.5 in these

3
J. J. Thomson received the 1906 Nobel Prize in Physics for his discovery. A year earlier, Philipp von Lenard
had received the Physics Prize for his work on cathode rays.
4
Why the charge on the proton should have exactly the same magnitude as that on the electron is a very long-
standing puzzle, the solution to which is suggested by some as yet unproven, but widely believed, theories of
particle physics that will be discussed briefly in Chapter 9.
 HISTORY 3

units. At about the same time, the concept of isotopism (a name coined by Soddy)
was conceived. Isotopes are atoms whose nuclei have different masses, but the same
charge. Naturally occurring elements were postulated to consist of a mixture of
different isotopes, giving rise to the observed masses.5
The explanation of isotopes had to wait 20 years until a classic discovery by
Chadwick in 1932. His work followed earlier experiments by Irène Curie (the
daughter of Pierre and Marie Curie) and her husband Frédéric Joliot.6 They had
observed that neutral radiation was emitted when
-particles bombarded beryllium
and later work had studied the energy of protons emitted when paraffin was
exposed to this neutral radiation. Chadwick refined and extended these experi-
ments and demonstrated that they implied the existence of an electrically neutral
particle of approximately the same mass as the proton. He had discovered the
neutron (n) and in so doing had produced almost the final ingredient for under-
standing nuclei.7
There remained the problem of reconciling the planetary model with the
observation of stable atoms. In classical physics, the electrons in the planetary
model would be constantly accelerating and would therefore lose energy by
radiation, leading to the collapse of the atom. This problem was solved by Bohr
in 1913. He applied the newly emerging quantum theory and the result was the
now well-known Bohr model of the atom. Refined modern versions of this model,
including relativistic effects described by the Dirac equation (the relativistic
analogue of the Schrödinger equation that applies to electrons), are capable of
explaining the phenomena of atomic physics. Later workers, including Heisenberg,
another of the founders of quantum theory,8 applied quantum mechanics to the
nucleus, now viewed as a collection of neutrons and protons, collectively called
nucleons. In this case, however, the force binding the nucleus is not the
electromagnetic force that holds electrons in their orbits, but is a short-range9
force whose magnitude is independent of the type of nucleon, proton or neutron
(i.e. charge-independent). This binding interaction is called the strong nuclear
force.
These ideas still form the essential framework of our understanding of the
nucleus today, where nuclei are bound states of nucleons held together by a strong
charge-independent short-range force. Nevertheless, there is still no single theory
that is capable of explaining all the data of nuclear physics and we shall see that
different models are used to interpret different classes of phenomena.

5
Frederick Soddy was awarded the 1921 Nobel Prize in Chemistry for his work on isotopes.
6
Irène Curie and Frédéric Joliot received the 1935 Nobel Prize in Chemistry for ‘synthesizing new
radioactive elements’.
7
James Chadwick received the 1935 Nobel Prize in Physics for his discovery of the neutron.
8
Werner Heisenberg received the 1932 Nobel Prize in Physics for his contributions to the creation of
quantum mechanics and the idea of isospin symmetry, which we will discuss in Chapter 3.
9
The concept of range will be discussed in more detail in Section 1.5.1, but for the present it may be taken as
the effective distance beyond which the force is insignificant.
4 CH1 BASIC CONCEPTS

1.1.2 The emergence of particle physics: the standard model

and hadrons

By the early 1930s, the 19th century view of atoms as indivisible elementary
particles had been replaced and a larger group of physically smaller entities now
enjoyed this status: electrons, protons and neutrons. To these we must add two
electrically neutral particles: the photon () and the neutrino (). The photon was
postulated by Planck in 1900 to explain black-body radiation, where the classical
description of electromagnetic radiation led to results incompatible with experi-
ments.10 The neutrino was postulated by Fermi in 1930 to explain the apparent
non-conservation of energy observed in the decay products of some unstable nuclei
where -rays are emitted, the so-called -decays. Prior to Fermi’s suggestion,
-decay had been viewed as a parent nucleus decaying to a daughter nucleus and
an electron. As this would be a two-body decay, it would imply that the electron
would have a unique momentum, whereas experiments showed that the electron
actually had a momentum spectrum. Fermi’s hypothesis of a third particle (the
neutrino) in the final state solved this problem, as well as a problem with angular
momentum conservation, which was apparently also violated if the decay was two-
body. The -decay data implied that the neutrino mass was very small and was
compatible with the neutrino being massless.11 It took more than 25 years before
Fermi’s hypothesis was confirmed by Reines and Cowan in a classic experiment in
1956 that detected free neutrinos from -decay.12
The 1950s also saw technological developments that enabled high-energy
beams of particles to be produced in laboratories. As a consequence, a wide
range of controlled scattering experiments could be performed and the greater
use of computers meant that sophisticated analysis techniques could be devel-
oped to handle the huge quantities of data that were being produced. By the
1960s this had resulted in the discovery of a very large number of unstable
particles with very short lifetimes and there was an urgent need for a theory that
could make sense of all these states. This emerged in the mid 1960s in the form
of the so-called quark model, first suggested by Murray Gell-Mann and
independently and simultaneously by George Zweig, who postulated that the
new particles were bound states of three families of more fundamental physical
particles.

10
X-rays had already been observed by Röntgen in 1895 (for which he received the first Nobel Prize in
Physics in 1901) and -rays were seen by Villard in 1900, but it was Planck who first made the startling
suggestion that electromagnetic energy was quantized. For this he was awarded the 1918 Nobel Prize in
Physics. Many years later, he said that his hypothesis was an ‘act of desperation’ as he had exhausted all
other possibilities.
11
However, in Section 3.1.4 we will discuss recent evidence that neutrinos have very small, but non-zero,
masses.
12
A description of this experiment is given in Chapter 12 of Tr75. Frederick Reines shared the 1995 Nobel
Prize in Physics for his work in neutrino physics and particularly for the detection of the electron neutrino.
 HISTORY 5

Gell-Mann called these quarks (q).13 Because no free quarks were detected
experimentally, there was initially considerable scepticism for this view. We now
know that there is a fundamental reason why quarks cannot be observed as free
particles (it will be discussed in Chapter 5), but at the time many physicists looked
upon quarks as a convenient mathematical description, rather than physical
particles.14 However, evidence for the existence of quarks as real particles came
in the 1960s from a series of experiments analogous to those of Rutherford and his
co-workers, where high-energy beams of electrons and neutrinos were scattered
from nucleons. (These experiments will also be discussed in Chapter 5.) Analysis
of the angular distributions of the scattered particles showed that the nucleons were
themselves bound states of three point-like charged entities, with properties
consistent with those hypothesized in the quark model. One of these properties
was unexpected and unusual: quarks have fractional electric charges, in practice

13 e and þ 23 e. This is essentially the picture today, where elementary particles
are now considered to be a small number of physical entities, including quarks, the
electron, neutrinos, the photon and a few others we shall meet, but no longer
nucleons.
The best theory of elementary particles we have at present is called, rather
prosaically, the standard model. This aims to explain all the phenomena of particle
physics, except those due to gravity, in terms of the properties and interactions of a
small number of elementary (or fundamental) particles, which are now defined as
being point-like, without internal structure or excited states. Particle physics thus
differs from nuclear physics in having a single theory to interpret its data.
An elementary particle is characterized by, amongst other things, its mass, its
electric charge and its spin. The latter is a permanent angular momentum
possessed by all particles in quantum theory, even when they are at rest. Spin
has no classical analogue and is not to be confused with the use of the same word
in classical physics, where it usually refers to the (orbital) angular momentum of
extended objects. The maximum value of the spin angular momentum about any
axis is s
hð
h  h=2Þ, where h is Planck’s constant and s is the spin quantum
number, or spin for short. It has a fixed value for particles of any given type (for
example s ¼ 12 for electrons) and general quantum mechanical principles restrict
the possible values of s to be 0, 12, 1, 32, . . .. Particles with half-integer spin are
called fermions and those with integer spin are called bosons. There are three
families of elementary particles in the standard model: two spin-12 families of
fermions called leptons and quarks; and one family of spin-1 bosons. In addition,

13
Gell-Mann received the 1969 Nobel Prize in Physics for ‘contributions and discoveries concerning the
classification of elementary particles and their interactions’. For the origin of the word ‘quark’, he cited the
now famous quotation ‘Three quarks for Muster Mark’ from James Joyce’s book Finnegans Wake. Zweig
had suggested the name ‘aces’, which with hindsight might have been more appropriate, as later experiments
revealed that there were four and not three families of quarks.
14
This was history repeating itself. In the early days of the atomic model many very distinguished scientists
were reluctant to accept that atoms existed, because they could not be ‘seen’ in a conventional sense.
6 CH1 BASIC CONCEPTS

at least one other spin-0 particle, called the Higgs boson, is postulated to explain
the origin of mass within the theory.15
The most familiar elementary particle is the electron, which we know is bound
in atoms by the electromagnetic interaction, one of the four forces of nature.16 One
test of the elementarity of the electron is the size of its magnetic moment. A
charged particle with spin necessarily has an intrinsic magnetic moment l. It can
be shown from the Dirac equation that a point-like spin-12 particle of charge q and
mass m has a magnetic moment l ¼ ðq=mÞS, where S is its spin vector, and hence l
has magnitude  ¼ q
h=2m. The magnetic moment of the electron very accurately
obeys this relation, confirming that electrons are elementary.
The electron is a member of the family of leptons. Another is the neutrino,
which was mentioned earlier as a decay product in -decays. Strictly speaking, this
particle should be called the electron neutrino, written e , because it is always
produced in association with an electron (the reason for this is discussed in
Section 3.1.1). The force responsible for -decay is an example of a second
fundamental force, the weak interaction. Finally, there is the third force, the
(fundamental) strong interaction, which, for example, binds quarks in nucleons.
The strong nuclear force mentioned in Section 1.1.1 is not the same as this
fundamental strong interaction, but is a consequence of it. The relation between
the two will be discussed in more detail later.
The standard model also specifies the origin of these three forces. In classical
physics the electromagnetic interaction is propagated by electromagnetic waves,
which are continuously emitted and absorbed. While this is an adequate descrip-
tion at long distances, at short distances the quantum nature of the interaction must
be taken into account. In quantum theory, the interaction is transmitted discon-
tinuously by the exchange of photons, which are members of the family of
fundamental spin-1 bosons of the standard model. Photons are referred to as the
gauge bosons, or ‘force carriers’, of the electromagnetic interaction. The use of the
word ‘gauge’ refers to the fact that the electromagnetic interaction possesses a
fundamental symmetry called gauge invariance. For example, Maxwell’s equa-
tions of classical electromagnetism are invariant under a specific phase transfor-
mation of the electromagnetic fields – the gauge transformation.17 This property is
common to all the three interactions of nature we will be discussing and has
profound consequences, but we will not need its details in this book. The weak and
strong interactions are also mediated by the exchange of spin-1 gauge bosons. For
the weak interaction these are the W þ , W
and Z 0 bosons (again the superscripts
denote the electric charges) with masses about 80–90 times the mass of the proton.

15
In the theory without the Higgs boson, all elementary particles are predicted to have zero mass, in obvious
contradiction with experiment. A solution to this problem involving the Higgs boson will be discussed briefly
in Chapter 9.
16
Gravity is so weak that it can be neglected in nuclear and particle physics at presently accessible energies.
Because of this, we will often refer in practice to the three forces of nature.
17
See, for example, Appendix C.2 of Ma97.
 RELATIVITY AND ANTIPARTICLES 7

For the strong interaction, the force carriers are called gluons. There are eight
gluons, all of which have zero mass and are electrically neutral.18
In addition to the elementary particles of the standard model, there are other
important particles we will be studying. These are the hadrons, the bound states of
quarks. Nucleons are examples of hadrons,19 but there are several hundred more,
not including nuclei, most of which are unstable and decay by one of the three
interactions. It was the abundance of these states that drove the search for a
simplifying theory that would give an explanation for their existence and led to the
quark model in the 1960s. The most common unstable example of a hadron is the
pion, which exists in three electrical charge states, written ðþ ; 0 ; 
Þ. Hadrons
are important because free quarks are unobservable in nature and so to deduce
their properties we are forced to study hadrons. An analogy would be if we had to
deduce the properties of nucleons by exclusively studying the properties of nuclei.
Since nucleons are bound states of quarks and nuclei are bound states of
nucleons, the properties of nuclei should, in principle, be deducible from the
properties of quarks and their interactions, i.e. from the standard model. In
practice, however, this is far beyond present calculational techniques and some-
times nuclear and particle physics are treated as two almost separate subjects.
However, there are many connections between them and in introductory treatments
it is still useful to present both subjects together.
The remaining sections of this chapter are devoted to introducing some of the
basic theoretical tools needed to describe the phenomena of both nuclear and
particle physics, starting with a key concept: antiparticles.

1.2 Relativity and Antiparticles

Elementary particle physics is also called high-energy physics. One reason for this
is that if we wish to produce new particles in a collision between two other
particles, then because of the relativistic mass–energy relation E ¼ mc2 , energies
are needed at least as great as the rest masses of the particles produced. The second
reason is that to explore the structure of a particle requires a probe whose
wavelength is smaller than the structure to be explored. By the de Broglie
relation ¼ h=p, this implies that the momentum p of the probing particle, and
hence its energy, must be large. For example, to explore the internal structure of
the proton using electrons requires wavelengths that are much smaller than the

18
Note that the word ‘electric’ has been used when talking about charge. This is because the weak and strong
interactions also have associated ‘charges’ which determine the strengths of the interactions, just as the
electric charge determines the strength of the electromagnetic interaction. This will be discussed in more
detail in later chapters.
19
The magnetic moments of the proton and neutron do not obey the prediction of the Dirac equation and this
is evidence that nucleons have structure and are not elementary. The proton magnetic moment was first
measured by Otto Stern using a molecular beam method that he developed and for this he received the 1943
Nobel Prize in Physics.
8 CH1 BASIC CONCEPTS

classical radius of the proton, which is roughly 10
15 m. This in turn requires
electron energies that are greater than 103 times the rest energy of the electron,
implying electron velocities very close to the speed of light. Hence any explanation
of the phenomena of elementary particle physics must take account of the
requirements of the theory of special relativity, in addition to those of quantum
theory. There are very few places in particle physics where a non-relativistic
treatment is adequate, whereas the need for a relativistic treatment is far less in
nuclear physics.
Constructing a quantum theory that is consistent with special relativity leads to
the conclusion that for every particle of nature, there must exist an associated
particle, called an antiparticle, with the same mass as the corresponding particle.
This important theoretical prediction was first made by Dirac and follows from the
solutions of the equation he first wrote down to describe relativistic electrons.20
The Dirac equation is of the form

@Cðx; tÞ
i
h ¼ Hðx; p
^ÞCðx; tÞ; ð1:1Þ
@t

^ ¼
i
hr is the usual quantum mechanical momentum operator and the

where p
Hamiltonian was postulated by Dirac to be

^ þ mc2 :
H ¼ ca p ð1:2Þ

The coefficients a and  are determined by the requirement that the solutions of
Equation (1.1) are also solutions of the Klein–Gordon equation 21

@ 2 Cðx; tÞ


h2 ¼

h2 c2 r2 Cðx; tÞ þ m2 c4 Cðx; tÞ: ð1:3Þ
@t2

This leads to the conclusion that a and  cannot be simple numbers; their simplest
forms are 4
4 matrices. Thus the solutions of the Dirac equation are four-
component wavefunctions (called spinors) with the form22
0 1
C1 ðx; tÞ
B C2 ðx; tÞ C
Cðx; tÞ ¼ B C
@ C3 ðx; tÞ A: ð1:4Þ
C4 ðx; tÞ

20
Paul Dirac shared the 1933 Nobel Prize in Physics with Erwin Schrödinger. The somewhat cryptic citation
stated ‘for the discovery of new productive forms of atomic theory’.
21
This is a relativistic equation, which is ‘derived’ by starting from the relativistic mass–energy relation
E2 ¼ p2 c2 þ m2 c4 and using the usual quantum mechanical operator substitutions, p ^ ¼
i
hr and
E ¼ i
h@=@t.
22
The details may be found in most quantum mechanics books, for example, pp. 475–477 of Sc68.
 SYMMETRIES AND CONSERVATION LAWS 9

The interpretation of Equation (1.4) is that the four components describe the two
spin states of a negatively charged electron with positive energy and the two spin
states of a corresponding particle having the same mass but with negative energy.
Two spin states arise because in quantum mechanics the projection in any direction
of the spin vector of a spin-12 particle can only result in one of the two values 12,
called ‘spin up’ and ‘spin down’, respectively. The two energy solutions arise from
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
the two solutions of the relativistic mass–energy relation E ¼  p2 c2 þ m2 c4 .
The latter states can be shown to behave in all respects as positively charged
electrons (called positrons), but with positive energy. The positron is referred to as
the antiparticle of the electron. The discovery of the positron by Anderson in 1933,
with all the predicted properties, was a spectacular verification of the Dirac
prediction.
Although Dirac originally made his prediction for electrons, the result is general
and is true whether the particle is an elementary particle or a hadron. If we denote
a particle by P, then the antiparticle is in general written with a bar over it, i.e. P
.
23
For example, the antiparticle of the proton is the antiproton
p, with negative
electric charge; and associated with every quark, q, is an antiquark,
q. However,
for some very common particles the bar is usually omitted. Thus, for example,
in the case of the positron eþ , the superscript denoting the charge makes
explicit the fact that the antiparticle has the opposite electric charge to that of
its associated particle. Electric charge is just one example of a quantum number
(spin is another) that characterizes a particle, whether it is elementary or composite
(i.e. a hadron).
Many quantum numbers differ in sign for particle and antiparticle, and electric
charge is an example of this. We will meet others later. When brought together,
particle–antiparticle pairs, each of mass m, can annihilate, releasing their com-
bined rest energy 2mc2 as photons or other particles. Finally, we note that there is
symmetry between particles and antiparticles, and it is a convention to call the
electron the particle and the positron its antiparticle. This reflects the fact that the
normal matter contains electrons rather than positrons.

1.3 Symmetries and Conservation Laws

Symmetries and the invariance properties of the underlying interactions play an
important role in physics. Some lead to conservation laws that are universal.
Familiar examples are translational invariance, leading to the conservation of
linear momentum; and rotational invariance, leading to conservation of angular
momentum. The latter plays an important role in nuclear and particle physics as it
leads to a scheme for the classification of states based, among other quantum

23
Carl Anderson shared the 1936 Nobel Prize in Physics for the discovery of the positron. The 1958 Prize
was awarded to Emilio Segrè and Owen Chamberlain for their discovery of the antiproton.
10 CH1 BASIC CONCEPTS

numbers, on their spins.24 Another very important invariance that we have briefly
mentioned is gauge invariance. This fundamental property of all three interactions
restricts the forms of the interactions in a profound way that initially is contra-
dicted by experiment. This is the prediction of zero masses for all elementary
particles, mentioned earlier. There are theoretical solutions to this problem whose
experimental verification (or otherwise) is probably the most eagerly awaited
result in particle physics today.
In nuclear and particle physics we need to consider additional symmetries of the
Hamiltonian and the conservation laws that follow and in the remainder of this
section we discuss two of the most important of these that we will need later –
parity and charge conjugation.

1.3.1 Parity

Parity was first introduced in the context of atomic physics by Eugene Wigner in
1927.25 It refers to the behaviour of a state under a spatial reflection, i.e. x !
x.
If we consider a single-particle state, represented for simplicity by a non-
relativistic wavefunction ðx; tÞ, then under the parity operator, P ^,

^ ðx; tÞ  P ð
x; tÞ:

P ð1:5Þ

Applying the operator again, gives

^ 2 ðx; tÞ ¼ PP
P ^ ð
x; tÞ ¼ P2 ðx; tÞ; ð1:6Þ

implying P ¼ 1. If the particle is an eigenfunction of linear momentum p, i.e.

ðx; tÞ  p ðx; tÞ ¼ exp½iðp x
EtÞ; ð1:7Þ

then

^
P p ðx; tÞ ¼P p ð
x; tÞ ¼P
p ðx; tÞ ð1:8Þ

and so a particle at rest, with p ¼ 0, is an eigenstate of parity. The eigenvalue

P ¼ 1 is called the intrinsic parity, or just the parity, of the state. By considering
a multiparticle state with a wavefunction that is the product of single-particle
wavefunctions, it is clear that parity is a multiplicative quantum number.
The strong and electromagnetic interactions, but not the weak interactions, are
invariant under parity, i.e. the Hamiltonian of the system remains unchanged under

24
These points are explored in more detail in, for example, Chapter 4 of Ma97.
25
Eugene Wigner shared the 1963 Nobel Prize in Physics, principally for his work on symmetries.
 SYMMETRIES AND CONSERVATION LAWS 11

a parity transformation on the position vectors of all particles in the system. Parity
is therefore conserved, by which we mean that the total parity quantum number
remains unchanged in the interaction. Compelling evidence for parity conservation
in the strong and electromagnetic interactions comes from the absence of
transitions between nuclear states and atomic states, respectively, that would
violate parity conservation. The evidence for non-conservation of parity in the
weak interaction will be discussed in detail in Chapter 6.
There is also a contribution to the total parity if the particle has an orbital
angular momentum l. In this case its wave function is a product of a radial part Rnl
and an angular part Ylm ð ; Þ:

lmn ðxÞ ¼ Rnl Ylm ð ; Þ; ð1:9Þ

where n and m are the principal and magnetic quantum numbers and Ylm ð ; Þ is a
spherical harmonic. It is straightforward to show from the relations between
Cartesian ðx; y; zÞ and spherical polar co-ordinates ðr; ; Þ, i.e.

x ¼ r sin cos ; y ¼ r sin sin ; z ¼ r cos ; ð1:10Þ

that the parity transformation x !
x implies

r ! r; ! 
;  !  þ ; ð1:11Þ

and from this it can be shown that

Ylm ð ; Þ ! Ylm ð
;  þ Þ ¼ ð
Þl Ylm ð ; Þ: ð1:12Þ

Equation (1.12) may easily be verified directly for specific cases; for example,
for the first three spherical harmonics,

12
1
1
1 3 2 3 2
Y00 ¼ ; Y10 ¼ cos ; Y11 ¼ sin ei : ð1:13Þ
4 4 8

Hence

^
P lmn ðxÞ ¼P lmn ð
xÞ ¼ Pð
Þl lmn ðxÞ; ð1:14Þ

i.e. lmn ðxÞ is an eigenstate of parity with eigenvalue Pð
1Þl .

An analysis of the Dirac Equation (1.1) for relativistic electrons, shows that it is
invariant under a parity transformation only if Pðeþ e
Þ ¼
1. This is a general
result for all fermion–antifermion pairs, so it is a convention to assign P ¼ þ1 to
all leptons and P ¼
1 to their antiparticles. We will see in Chapter 3 that in
strong interactions quarks can only be created as part of a quark–antiquark pair, so
the intrinsic parity of a single quark cannot be measured. For this reason, it is also
12 CH1 BASIC CONCEPTS

a convention to assign P ¼ þ1 to quarks. Since quarks are fermions, it follows

from the Dirac result that P ¼
1 for antiquarks. The intrinsic parities of hadrons
then follow from their structure in terms of quarks and the orbital angular
momentum between the constituent quarks, using Equation (1.14). This will be
explored in Chapter 3 as part of the discussion of the quark model.

1.3.2 Charge conjugation

Charge conjugation is the operation of changing a particle into its antiparticle.

Like parity, it gives rise to a multiplicative quantum number that is conserved in
strong and electromagnetic interactions, but violated in the weak interaction. In
discussing charge conjugation, we will need to distinguish between states such as
the photon  and the neutral pion 0 that do not have distinct antiparticles and
those such as the þ and the neutron, which do. Particles in the former class we
will collectively denote by a and those of the latter type will be denoted by b. It is
also convenient at this point to extend our notation for states. Thus we will
represent a state of type a having a wavefunction a by ja; a i and similarly for a
state of type b. Then under the charge conjugation operator, C ^,

^ ja;
C ai ¼ Ca ja; a i; ð1:15aÞ

and

^ jb;
C bi ¼ j

b;
b i; ð1:15bÞ

where Ca is a phase factor analogous to the phase factor in Equation (1.5).26

Applying the operator twice, in the same way as for parity, leads to Ca ¼ 1. From
^ with eigenvalues
Equation (1.15a), we see that states of type a are eigenstates of C
1, called their C-parities. States with distinct antiparticles can only form
^ as linear combinations.
eigenstates of C
As an example of the latter, consider a þ 
pair with orbital angular
momentum L between them. In this case

^ jþ 
; Li ¼ ð
1ÞL jþ 
; Li;

C ð1:16Þ

because interchanging the pions reverses their relative positions in the spatial
wavefunction. The same factor occurs for spin-12 fermion pairs f
f , but in addition
there are two other factors. The first is ð
1ÞSþ1 , where S is the total spin of the pair.

26
A phase factor could have been inserted in Equation (1.15b), but it is straightforward to show that the
relative phase of the two states b and
b cannot be measured and so a phase introduced in this way would have
no physical consequences.
 INTERACTIONS AND FEYNMAN DIAGRAMS 13

This follows directly from the structure of the spin wavefunctions:

9
"1 "2 Sz ¼ 1 =
p1ffiffi ð" # þ # " Þ Sz ¼ 0 S¼1 ð1:17aÞ
2 1 2 1 2
;
#1 #2 Sz ¼
1

and

p1ffiffi ð" #
#1 "2 Þ Sz ¼ 0 S ¼ 0; ð1:17bÞ

2 1 2

where "i ð#i Þ represents particle i having spin ‘up’ (‘down’) in the z-direction. A
second factor ð
1Þ arises whenever fermions and antifermions are interchanged.
This has its origins in quantum field theory.27 Combining these factors, finally
we have

^ j f
f ; J; L; Si ¼ ð
1ÞLþS j f
f ; J; L; Si;
C ð1:18Þ

for fermion–antifermion pairs having total, orbital and spin angular momentum
quantum numbers J, L and S, respectively.

1.4 Interactions and Feynman Diagrams

We now turn to a discussion of particle interactions and how they can be described
by the very useful pictorial methods of Feynman diagrams.

1.4.1 Interactions

Interactions involving elementary particles and/or hadrons are conveniently

summarized by ‘equations’ by analogy with chemical reactions, in which the
different particles are represented by symbols which usually – but not always –
have a superscript to denote their electric charge. In the interaction

e þ n ! e
þ p; ð1:19Þ

for example, an electron neutrino e collides with a neutron n to produce an

electron e
and a proton p, whereas the equation

e
þ p ! e
þ p ð1:20Þ

27
See, for example, pp. 249–250 of Go86.
14 CH1 BASIC CONCEPTS

represents an electron and proton interacting to give the same particles in the final
state, but in general travelling in different directions. In such equations, conserved
quantum numbers must have the same total values in initial and final states.
Particles may be transferred from initial to final states and vice versa, when they
become antiparticles. Thus the process


þ p ! 
þ p; ð1:21aÞ

also implies the reaction

p þ
p ! þ þ 
; ð1:22Þ

which is obtained by taking the proton from the final state to an antiproton in the
initial state and the negatively charged pion in the initial state to a positively
charged pion in the final state.
The interactions in Equations (1.20) and (1.21a), in which the particles remain
unchanged, are examples of elastic scattering, in contrast to the reactions in
Equations (1.19) and (1.22), where the final-state particles differ from those in the
initial state. Collisions between a given pair of initial particles do not always lead
to the same final state, but can lead to different final states with different
probabilities. For example, the collision of a negatively charged pion and a proton
can give rise to elastic scattering (Equation (1.21a)) and a variety of other
reactions, such as


þ p ! n þ 0 and 
þ p ! p þ 
þ 
þ þ ; ð1:21bÞ

depending on the initial energy. In particle physics it is common to refer (rather

imprecisely) to such interactions as ‘inelastic’ scattering.
Similar considerations apply to nuclear physics, but the term inelastic scattering
is reserved for the case where the final state is an excited state of the parent nucleus
A, that subsequently decays, for example via photon emission, i.e.

a þ A ! a þ A ; A ! A þ ; ð1:23Þ

where a is a projectile and A is an excited state of A. A useful shorthand notation

used in nuclear physics for the general reaction a þ A ! b þ B is Aða; bÞB. It is
usual in nuclear physics to further subdivide types of interactions according to the
underlying mechanism that produced them. We will return to this in Section 2.9, as
part of a more general discussion of nuclear reactions.
Finally, many particles are unstable and spontaneously decay to other, lighter
(i.e. having less mass) particles. An example of this is the free neutron (i.e. one not
bound in a nucleus), which decays by the -decay reaction

n ! p þ e
þ 
e ; ð1:24Þ
 INTERACTIONS AND FEYNMAN DIAGRAMS 15

with a mean lifetime of about 900 s.28 The same notation can also be used in
nuclear physics. For example, many nuclei decay via the -decay reaction. Thus,
denoting a nucleus with Z protons and N nucleons as (Z, N ), we have

ðZ; NÞ ! ðZ
1; NÞ þ eþ þ e : ð1:25Þ

This is also a weak interaction. This reaction is effectively the decay of a proton
bound in a nucleus. Although a free proton cannot decay by the -decay
p ! n þ eþ þ e because it violates energy conservation (the final-state particles
have greater total mass than the proton), a proton bound in a nucleus can decay
because of its binding energy. This will be explained in Chapter 2.

1.4.2 Feynman diagrams

The forces producing all the above interactions are due to the exchange of particles
and a convenient way of illustrating this is to use Feynman diagrams. There are
mathematical rules and techniques associated with these that enable them to be
used to calculate the quantum mechanical probabilities for given reactions to
occur, but in this book Feynman diagrams will only be used as a convenient very
useful pictorial description of reaction mechanisms.
We first illustrate them at the level of elementary particles for the case of
electromagnetic interactions, which arise from the emission and/or absorption of
photons. For example, the dominant interaction between two electrons is due to the
exchange of a single photon, which is emitted by one electron and absorbed by the
other. This mechanism, which gives rise to the familiar Coulomb interaction at
large distances, is illustrated in the Feynman diagram Figure 1.1(a).
In such diagrams, we will use the convention that particles in the initial state are
shown on the left and particles in the final state are shown on the right. (Some
authors take time to run along the y-axis.) Spin-12 fermions (such as the electron)

Figure 1.1 One-photon exchange in (a) e
þ e
! e
þ e
and (b) eþ þ eþ ! eþ þ eþ

28
The reason that this decay involves an antineutrino rather than a neutrino will become clear in Chapter 3.
16 CH1 BASIC CONCEPTS

are drawn as solid lines and photons are drawn as wiggly lines. Arrowheads
pointing to the right indicate that the solid lines represent electrons. In the case of
photon exchange between two positrons, which is shown in Figure 1.1(b), the
arrowheads on the antiparticle (positron) lines are conventionally shown as
pointing to the left. In interpreting these diagrams, it is important to remember:
(1) that the direction of the arrows on fermion lines does not indicate the particle’s
direction of motion, but merely whether the fermions are particles or antiparticles,
and (2) that particles in the initial state are always to the left and particles in the
final state are always to the right.
A feature of the above diagrams is that they are constructed from combinations
of simple three-line vertices. This is characteristic of electromagnetic processes.
Each vertex has a line corresponding to a single photon being emitted or absorbed,
while one fermion line has the arrow pointing toward the vertex and the other away
from the vertex, guaranteeing charge conservation at the vertex, which is one of
the rules of Feynman diagrams.29 For example, a vertex like Figure 1.2 would
correspond to a process in which an electron emitted a photon and turned into a
positron. This would violate charge conservation and is therefore forbidden.

e–

e+

Figure 1.2 The forbidden vertex e
! eþ þ 

Feynman diagrams can also be used to describe the fundamental weak and
strong interactions. This is illustrated by Figure 1.3(a), which shows contributions
to the elastic weak scattering reaction e
þ e ! e
þ e due to the exchange of a
Z 0 , and by Figure 1.3(b), which shows the exchange of a gluon g (represented by a
coiled line) between two quarks, which is a strong interaction.
Feynman diagrams can also be drawn at the level of hadrons. As an illustration,
Figure 1.4 shows the exchange of a charged pion (shown as a dashed line) between
a proton and a neutron. We shall see later that this mechanism is a major
contribution to the strong nuclear force between a proton and a neutron.
We turn now to consider in more detail the relation between exchanged particles
and forces.

29
Compare Kirchhoff’s laws in electromagnetism.
 PARTICLE EXCHANGE: FORCES AND POTENTIALS 17

Figure 1.3 (a) Contributions of (a) Z 0 exchange to the elastic weak scattering reaction
e
þ e ! e
þ e , and (b) gluon exchange contribution to the strong interaction q þ q ! q þ q

Figure 1.4 Single-pion exchange in the reaction p þ n ! n þ p

1.5 Particle Exchange: Forces and Potentials

This section starts with a discussion of the important relationship between forces
and particle exchanges and then relates this to potentials. Although the idea of a
potential has its greatest use in non-relativistic physics, nevertheless it is useful to
illustrate concepts and is used in later sections as an intermediate step in relating
theoretical Feynman diagrams to measurable quantities. The results can easily be
extended to more general situations.

1.5.1 Range of forces

At each vertex of a Feynman diagram, charge is conserved by construction. We

will see later that, depending on the nature of the interaction (strong, weak or
electromagnetic), other quantum numbers are also conserved. However, it is easy
to show that energy and momentum cannot be conserved simultaneously.
Consider the general case of a reaction A þ B ! A þ B mediated by the
exchange of a particle X, as shown in Figure 1.5. In the rest frame of the incident
18 CH1 BASIC CONCEPTS

Figure 1.5 Exchange of a particle X in the reaction A þ B ! A þ B

particle A, the lower vertex represents the virtual process (‘virtual’ because X does
not appear as a real particle in the final state),

AðMA c2 ; 0Þ ! AðEA ; pA cÞ þ XðEX ;
pA cÞ; ð1:26Þ

where EA is the total energy of particle A and pA is its three-momentum.30 Thus, if

we denote by PA the four-vector for particle A,

PA ¼ ðEA =c; pA Þ ð1:27Þ

and

P2A ¼ EA2 =c2
p2A ¼ MA2 c2 : ð1:28Þ

Applying this to the diagram and imposing momentum conservation, gives

EA ¼ ðp2 c2 þ MA2 c4 Þ1=2 ; EX ¼ ðp2 c2 þ MX2 c4 Þ1=2 ; ð1:29Þ

where p ¼ jpj. The energy difference between the final and initial states is given by

E ¼ EX þ EA
MA c2 ! 2pc; p!1
! MX c2 ; p!0 ð1:30Þ

and thus E  MX c2 for all p, i.e. energy is not conserved. However, by the
Heisenberg uncertainty principle, such an energy violation is allowed, but only for
a time  

h=E, so we immediately obtain

r  R  h
=MX c ð1:31Þ

30
A resumé of relativistic kinematics is given in Appendix B.
 PARTICLE EXCHANGE: FORCES AND POTENTIALS 19

as the maximum distance over which X can propagate before being absorbed by
particle B. This maximum distance is called the range of the interaction and this
was the sense of the word used in Section 1.1.1.
The electromagnetic interaction has an infinite range, because the exchanged
particle is a massless photon. In contrast, the weak interaction is associated with
the exchange of very heavy particles – the W and Z bosons. These lead to ranges
that from Equation (1.31) are of the order of RW;Z  2
10
18 m. The funda-
mental strong interaction has infinite range because, like the photon, gluons have
zero mass. On the other hand, the strong nuclear force, as exemplified by Figure 1.4
for example, has a much shorter range of approximately (1–2)
10
15 m. We will
comment briefly on the relation between these two different manifestations of the
strong interaction in Section 7.1.

1.5.2 The Yukawa potential

In the limit that MA becomes large, we can regard B as being scattered by a static
potential of which A is the source. This potential will in general be spin dependent,
but its main features can be obtained by neglecting spin and considering X to be a
spin-0 boson, in which case it will obey the Klein–Gordon equation

@ 2 ðx; tÞ


h2 ¼

h2 c2 r2 ðx; tÞ þ MX2 c4 ðx; tÞ: ð1:32Þ
@t2

The static solution of this equation satisfies

MX2 c4
r2 ðxÞ ¼ ðxÞ; ð1:33Þ

h2

where ðxÞ is interpreted as a static potential. For MX ¼ 0 this equation is the same
as that obeyed by the electrostatic potential, and for a point charge
e interacting
with a point charge þe at the origin, the appropriate solution is the Coulomb potential

e2 1
VðrÞ ¼
eðrÞ ¼
; ð1:34Þ
4"0 r

where r ¼ jxj and "0 is the dielectric constant. The corresponding solution in the
case where MX2 6¼ 0 is easily verified by substitution to be

g2 e
r=R
VðrÞ ¼
; ð1:35Þ
4 r

where R is the range defined earlier and g, the so-called coupling constant, is a
parameter associated with each vertex of a Feynman diagram and represents the
20 CH1 BASIC CONCEPTS

basic strength of the interaction.31 For simplicity, we have assumed equal strengths
for the couplings of particle X to the particles A and B.
The form of VðrÞ in Equation (1.35) is called a Yukawa potential, after the physicist
who first introduced the idea of forces due to massive particle exchange in 1935.32
As MX ! 0, R ! 1 and the Coulomb potential is recovered from the Yukawa
potential, while for very large masses the interaction is approximately point-like
(zero range). It is conventional to introduce a dimensionless parameter
X by

g2

X ¼ ; ð1:36Þ
4
hc

that characterizes the strength of the interaction at short distances r  R. For the
electromagnetic interaction this is the fine structure constant

 e2 =4"0 h
c  1=137 ð1:37Þ

that governs the splittings of atom energy levels.33

The forces between hadrons are also generated by the exchange of particles. Thus,
in addition to the electromagnetic interaction between charged hadrons, all hadrons,
whether charged or neutral, experience a strong short-range interaction, which in the
case of two nucleons, for example, has a range of about 10
15 m, corresponding to
the exchange of a particle with an effective mass of about 17th the mass of the proton.
The dominant contribution to this force is the exchange of a single pion, as shown in
Figure 1.4. This nuclear strong interaction is a complicated effect that has its origins
in the fundamental strong interactions between the quark distributions within the two
hadrons. Two neutral atoms also experience an electromagnetic interaction (the van
der Waals force), which has its origins in the fundamental Coulomb forces, but is of
much shorter range. Although an analogous mechanism is not in fact responsible for
the nuclear strong interaction, it is a useful reminder that the force between two
distributions of particles can be much more complicated than the forces between
their components. We will return to this point when we discuss the nature of the
nuclear potential in more detail in Section 7.1.

1.6 Observable Quantities: Cross Sections and Decay Rates

We have mentioned earlier that Feynman diagrams can be turned into probabilities
for a process by a set of mathematical rules (the Feynman Rules) that can be

31
Although we call g a (point) coupling constant, in general it will have a dependence on the momentum
carried by the exchanged particle. We ignore this in what follows.
32
For this insight, Hideki Yukawa received the 1949 Nobel Prize in Physics.
33
Like g, the coupling
X will in general have a dependence on the momentum carried by particle X. In the
case of the electromagnetic interaction, this dependence is relatively weak.
 OBSERVABLE QUANTITIES: CROSS SECTIONS AND DECAY RATES 21

derived from the quantum theory of the underlying interaction.34 We will not
pursue this in detail in this book, but rather will show in principle their relation to
observables, i.e. things that can be measured, concentrating on the cases of two-
body scattering reactions and decays of unstable states.

1.6.1 Amplitudes

The intermediate step is the amplitude f, the modulus squared of which is directly
related to the probability of the process occurring. It is also called the invariant
amplitude because, as we shall show, it is directly related to observable quantities
and these have to be the same in all inertial frames of reference. To get some
qualitative idea of the structure of f, we will use non-relativistic quantum
mechanics and assume that the coupling constant g2 is small compared with
4
hc, so that the interaction is a small perturbation on the free particle solution,
which will be taken as plane waves.
If we expand the amplitude in a power series in g2 and keep just the first term
(i.e. lowest-order perturbation theory), then the amplitude for a particle in an initial
state with momentum qi to be scattered to a final state with momentum qf by a
potential VðxÞ is proportional to
ð
f ðq2 Þ ¼ d3 xVðxÞ exp½iq x=
h; ð1:38Þ

i.e. the Fourier transform of the potential, where q  qi
qf is the momentum

transfer.35
The integration may be done using polar co-ordinates. Taking q in the x-
direction, gives

q x ¼ jqjr cos ð1:39Þ

and

d3 x ¼ r2 sin d dr d; ð1:40Þ

where r ¼ jxj. For the Yukawa potential, the integral in Equation (1.38) gives

h2

g2

f ðq2 Þ ¼ : ð1:41Þ
q2 þ MX2 c2

34
In the case of the electromagnetic interaction, the theory is called Quantum Electrodynamics (QED) and is
spectacularly successful in explaining experimental results. Richard Feynman shared the 1965 Nobel Prize in
Physics with Sin-Itiro Tomonoga and Julian Schwinger for their work on formulating quantum electro-
dynamics. The Feynman Rules are discussed in an accessible way in Gr87.
35
See, for example, Chapter 11 of Ma92.
22 CH1 BASIC CONCEPTS

This amplitude corresponds to the exchange of a single particle, as shown for

example in Figures 1.3 and 1.4. The structure of the amplitude, which is quite
general, is a numerator proportional to the product of the couplings at the two
vertices (or equivalently
X in this case), and a denominator that depends on the
mass of the exchanged particle and its momentum transfer squared. The denomi-
nator is called the propagator for particle X. In a relativistic calculation, the term
q2 becomes q2 , where q is the four-momentum transfer.
Returning to the zero-range approximation, one area where it is used extensively
is in weak interactions, particularly applied to nuclear decay processes. In these
situations, MX ¼ MW;Z and f !
GF , where GF , the so-called Fermi coupling
constant, is given from Equation (1.36) by

GF 4
W
3
¼ ¼ 1:166
10
5 GeV
2 : ð1:42Þ
ð
hcÞ ðMW c2 Þ2

The numerical value is obtained from analyses of decay processes, including that
of the neutron and a heavier version of the electron called the muon, whose
properties will be discussed in Chapter 3.
All the above is for the exchange of a single particle. It is also possible to draw
more complicated Feynman diagrams that correspond to the exchange of more
than one particle. An example of such a diagram for elastic e
e
scattering, where
two photons are exchanged, is shown in Figure 1.6.

Figure 1.6 Two-photon exchange in the reaction e
þ e
! e
þ e

The number of vertices in any diagram is called the order n, and when the
amplitude associated p with
ffiffiffiffi n any given Feynman diagram is calculated, it always
contains a factor of ð
Þ . Since the probability is proportional to the square of
the modulus of the amplitude, the former will contain a factor
n. The probability
associated with the single-photon exchange diagrams of Figure 1.1 thus contain a
factor of
2 and the contribution from two-photon exchange is of the order of
4 .
 OBSERVABLE QUANTITIES: CROSS SECTIONS AND DECAY RATES 23

As
 1=137, the latter is usually very small compared with the contribution
from single-photon exchange.
This is a general feature of electromagnetic interactions: because the fine structure
constant is very small, in most cases only the lowest-order diagrams that contribute to
a given process need be taken into account, and more complicated higher-order
diagrams with more vertices can (to a good approximation) be ignored in many
applications.

1.6.2 Cross-sections

The next step is to relate the amplitude to measurables. For scattering reactions,
the appropriate observable is the cross-section. In a typical scattering experiment, a
beam of particles is allowed to hit a target and the rates of production of various
particles in the final state are counted.36 It is clear that the rates will be
proportional to: (a) the number N of particles in the target illuminated by the
beam, and (b) the rate per unit area at which beam particles cross a small surface
placed in the beam at rest with respect to the target and perpendicular to the beam
direction. The latter is called the flux and is given by

J ¼ nb vi ; ð1:43Þ

where nb is the number density of particles in the beam and vi their velocity in the
rest frame of the target. Hence the rate Wr at which a specific reaction r occurs in a
particular experiment can be written in the form

Wr ¼ JNr ; ð1:44aÞ

where r , the constant of proportionality, is called the cross-section for reaction r.

If the beam has a cross-sectional area S, its intensity is I ¼ JS and so an alternative
expression for the rate is

Wr ¼ Nr I=S ¼ Ir nt t; ð1:44bÞ

where nt is the number of target particles per unit volume and t is the thickness of
the target. If the target consists of an isotopic species of atomic mass MA (in atomic
mass units-defined in Section 1.7 below), then nt ¼ NA =MA , where  is the density
of the target and NA is Avogadro’s constant. Thus, Equation (1.44b) may be written

Wr ¼ Ir ðtÞNA =MA ; ð1:44cÞ

36
The practical aspects of experiments are discussed in Chapter 4.
24 CH1 BASIC CONCEPTS

where ðtÞ is a measure of the amount of material in the target, expressed in units
of mass per unit area. The form of Equation (1.44c) is particularly useful for the
case of thin targets, commonly used in experiments. In the above, the product JN is
called the luminosity L, i.e.

L  JN ð1:45Þ

and contains all the dependencies on the densities and geometries of the beam and
target. The cross-section is independent of these factors.
It can be seen from the above equations that the cross-section has the dimensions
of an area; the rate per target particle Jr at which the reaction occurs is equal to
the rate at which beam particles would hit a surface of area r , placed in the beam
at rest with respect to the target and perpendicular to the beam direction. Since the
area of such a surface is unchanged by a Lorentz transformation in the beam
direction, the cross-section is the same in all inertial frames of reference, i.e. it is a
Lorentz invariant.
The quantity r is better named the partial cross-section, because it is the cross-
section for a particular reaction r. The total cross-section  is defined by
X
 r : ð1:46Þ
r

Another useful quantity is the differential cross-section, dr ð ; Þ=d, which is

defined by

dr ð ; Þ
dWr  JN d; ð1:47Þ
d

where dWr is the measured rate for the particles to be emitted into an element of
solid angle d ¼ d cos d in the direction ð ; Þ, as shown in Figure 1.7.
The total cross-section is obtained by integrating the partial cross-section over
all angles, i.e.
ð 2 ð1
dr ð ; Þ
r ¼ d d cos : ð1:48Þ
0
1 d

The final step is to write these formulae in terms of the scattering amplitude f ðq2 Þ
appropriate for describing the scattering of a non-relativistic spinless particle from
a potential.
To do this it is convenient to consider a single beam particle interacting with a
single target particle and to confine the whole system in an arbitrary volume V
(which cancels in the final result). The incident flux is then given by

J ¼ nb vi ¼ vi =V ð1:49Þ
 OBSERVABLE QUANTITIES: CROSS SECTIONS AND DECAY RATES 25

Figure 1.7 Geometry of the differential cross-section: a beam of particles is incident along the
z-axis and collides with a stationary target at the origin; the differential cross-section is
proportional to the rate for particles to be scattered into a small solid angle d in the direction
ð ; Þ

and since the number of target particles is N ¼ 1, the differential rate is

vi dr ð ; Þ
dWr ¼ d: ð1:50Þ
V d

In quantum mechanics, provided the interaction is not too strong, the transition
rate for any process is given in perturbation theory by the Born approximation37

ð 2
2  3  

dWr ¼ dx f VðxÞ i  ðEf Þ: ð1:51Þ

h 

The term ðEf Þ is the density-of-states factor (see below) and we take the initial
and final state wavefunctions to be plane waves:

1 1
i ¼ pffiffiffiffi exp½iqi x=
h; f ¼ pffiffiffiffi exp½iqf x=
h; ð1:52Þ
V V

37
This equation is a form of the Second Golden Rule. It is discussed in Appendix A.
26 CH1 BASIC CONCEPTS

where the final momentum qf lies within a small solid angle d located in the
direction ð ; Þ (see Figure 1.7.). Then, by direct integration,
2
dWr ¼ j f ðq2 Þj2 ðEf Þ; ð1:53Þ

hV 2
where f ðq2 Þ is the scattering amplitude defined in Equation (1.38).
The density of states ðEf Þ that appears in Equation (1.51) is the number of
possible final states with energy lying between Ef and Ef þ dEf and is given by38
V q2f
ðEf Þ ¼ d: ð1:54Þ
ð2
hÞ3 vf
If we use this and Equation (1.53) in Equation (1.50), we have

d 1 q2
¼ 2 4 f j f ðq2 Þj2 : ð1:55Þ
d 4
h vi vf

Although this result has been derived in the laboratory system, because we have
taken a massive target it is also valid in the centre-of-mass system. For a finite
mass target it would be necessary to make a Lorentz transformation on Equation
(1.55). The expression is also true for the general two-body relativistic scattering
process a þ b ! c þ d.
All the above is for spinless particles, so finally we have to generalize Equation
(1.55) to include the effects of spin. Suppose the initial-state particles a and b, have
spins sa and sb and the final-state particles c and d have spins sc and sd . The total
numbers of spin substates available to the initial and final states are gi and gf ,
respectively, given by

gi ¼ ð2sa þ 1Þð2sb þ 1Þ and gf ¼ ð2sc þ 1Þ ð2sd þ 1Þ: ð1:56Þ

If the initial particles are unpolarized (which is the most common case in practice),
then we must average over all possible initial spin configurations (because each is
equally likely) and sum over the final configurations. Thus, Equation (1.55) becomes
d gf q2
¼ 2 4 f jMfi j2 ; ð1:57Þ
d 4
h vi vf
where
jMfi j2  j f ðq2 Þj2 ð1:58Þ

and the bar over the amplitude denotes a spin-average of the squared matrix
element.

38
The derivation is given in detail in Appendix A.
 OBSERVABLE QUANTITIES: CROSS SECTIONS AND DECAY RATES 27

1.6.3 Unstable states

In the case of an unstable state, the observable of interest is its lifetime at rest , or
equivalently its natural decay width, given by  ¼
h=, which is a measure of the
rate of the decay reaction. In general, an initial unstable state will decay to several
final states and in this case we define f as the partial width for channel f and
X
¼ f ð1:59Þ
f

as the total decay width, while

Bf  f = ð1:60Þ

is defined as the branching ratio for decay to channel f.

The energy distribution of an unstable state to a final state f has the Breit–Wigner
form
f
Nf ðWÞ / ; ð1:61Þ
ðW
MÞ2 c4 þ 2 =4

where M is the mass of the decaying state and W is the invariant mass of the decay
products.39 The Breit–Wigner formula is shown in Figure 1.8 and is the same
formula that describes the widths of atomic and nuclear spectral lines. (The overall
factor depends on the spins of the particles involved.) It is a symmetrical bell-
shaped curve with a maximum at W ¼ M and a full width  at half the maximum
height of the curve. It is proportional to the number of events with invariant mass W.
If an unstable state is produced in a scattering reaction, then the cross section for
that reaction will show an enhancement described by the same Breit–Wigner formula.
In this case we say we have produced a resonance state. In the vicinity of a resonance
of mass M, and width , the cross-section for the reaction i ! f has the form

i f
fi / ; ð1:62Þ
ðE
Mc2 Þ2 þ 2 =4

where E is the total energy of the system. Again, the form of the overall constant
will depend on the spins of the particles involved. Thus, for example, if the
resonance particle has spin j and the spins of the initial particles are s1 and s2 , then


h2 2j þ 1 i f
fi ¼ : ð1:63Þ
qi ð2s1 þ 1Þð2s2 þ 1Þ ðE
Mc2 Þ2 þ 2 =4
2

39
This form arises from a state that decays exponentially with time, although a proper proof of this is quite
lengthy. See, for example, Appendix B of Ma97.
28 CH1 BASIC CONCEPTS

Figure 1.8 The Breit--Wigner formula (Equation (1.61))

In practice there will also be kinematical and angular momentum effects that will
distort this formula from its perfectly symmetric shape.
An example of resonance formation in 
p interactions is given in Figure 1.9,
which shows the 
p total cross-section in the centre-of-mass energy range

Figure 1.9 Total cross-sections for 
p interactions (data from Ca68)

 UNITS: LENGTH, MASS AND ENERGY 29

1.2–2.4 GeV. (The units used in the plots will become clear after the next section.)
Two enhancements can be seen that are of the approximate Breit–Wigner resonance
form and there are two other maxima at higher energies. In principle, the mass and
width of a resonance may be obtained by using a Breit–Wigner formula and varying
M and  to fit the cross-section in the region of the enhancement. In practice, more
sophisticated methods are used that fit a wide range of data, including differential
cross-sections, simultaneously and also take account of non-resonant contributions
to the scattering. The widths obtained from such analyses are of the order of
100 MeV, with corresponding interaction times of order 10
23 s, which is consistent
with the time taken for a relativistic pion to transit the dimension of a proton.
Resonances are also a feature of interactions in nuclear physics and we will return to
this in Section 2.9 when we discuss nuclear reaction mechanisms.

1.7 Units: Length, Mass and Energy

Most branches of science introduce special units that are convenient for their own
purposes. Nuclear and particle physics are no exceptions. Distances tend to be measured
in femtometres (fm) or, equivalently fermis, with 1 fm  10
15 m. In these units, the
radius of the proton is about 0.8 fm. The range of the strong nuclear force between
protons and neutrons is of the order of 1–2 fm, while the range of the weak force is of the
order of 10
3 fm. For comparison, the radii of atoms are of the order of 105 fm. A
common unit for area is the barn (b) defined by 1 b ¼ 10
28 m2. For example, the total
cross-section for pp scattering (a strong interaction) is a few tens of millibarns (mb)
(compare also the 
p total cross-section in Figure 1.9), whereas the same quantity for
p scattering (a weak interaction) is a few tens of femtobarns (fb), depending on the
energies involved. Nuclear cross-sections are very much larger and increase approxi-
mately like A2=3 , where A is the total number of nucleons in the nucleus.
Energies are invariably specified in terms of the electron volt (eV) defined as the
energy required to raise the electric potential of an electron or proton by 1 V.
In SI units, 1 eV ¼ 1:6
10
19 J. The units 1 keV ¼ 103 eV, 1 MeV ¼ 106 eV,
1 GeV ¼ 109 eV and 1 TeV ¼ 1012 eV are also in general use. In terms of these
units, atomic ionization energies are typically a few eV, the energies needed to
bind nucleons in heavy nuclei are typically 7–8 MeV per particle, and the highest
particle energies produced in the laboratory are of the order of 1 TeV for protons.
Momenta are specified in eV=c, MeV=c, etc..
In order to create a new particle of mass M, an energy at least as great as its rest
energy Mc2 must be supplied. The rest energies of the electron and proton are
0.51 MeV and 0.94 GeV respectively, whereas the W and Z 0 bosons have rest
energies of 80 GeV and 91 GeV, respectively. Correspondingly, their masses are
conveniently measured in MeV=c2 or GeV=c2, so that, for example,

Me ¼ 0:51 MeV=c2 ; Mp ¼ 0:94 GeV=c2 ;

ð1:64Þ
MW ¼ 80:3 GeV=c2 ; MZ ¼ 91:2 GeV=c2
30 CH1 BASIC CONCEPTS

In SI units, 1 MeV=c2 ¼ 1:78

10
30 kg. In nuclear physics it is also common to
express masses in atomic mass units (u), where 1 u ¼ 1:661
10
27 kg ¼
931:5 MeV=c2 .
Although practical calculations are expressed in the above units, it is usual in
particle physics to make theoretical calculations in units chosen such that
h  h=2 ¼ 1 and c ¼ 1 (called natural units) and many books do this. However,


as this book is about both nuclear and particle physics, only practical units will be
used.
A table giving numerical values of fundamental and derived constants, together
with some useful conversion factors is given on page XV.

Problems
1.1 ‘Derive’ the Klein–Gordon equation using the information in Footnote 21 and verify
that Equation (1.35) is a static solution of the equation.
qffiffi
1.2 Verify that the spherical harmonic Y11 ¼ 38 sin ei is an eigenfunction of parity
with eigenvalue P ¼
1.

1.3 A proton and antiproton at rest in an S-state annihilate to produce 0 0 pairs. Show
that this reaction cannot be a strong interaction.

1.4 Suppose that an intrinsic C-parity factor is introduced into Equation (1.15b), which
then becomes C ^ jb; b i ¼ Cb j

b;
b i. Show that the eigenvalue corresponding to any
^ is independent of Cb , so that Cb cannot be measured.
eigenstate of C

1.5 Consider the reaction 
d ! nn, where d is a spin-1 S-wave bound state of a proton
and a neutron called the deuteron and the initial pion is at rest. Deduce the intrinsic
parity of the negative pion.

1.6 Write down equations in symbol form that describe the following interactions:

(a) elastic scattering of an electron antineutrino and a positron;

(b) inelastic production of a pair of neutral pions in proton–proton interactions;

(c) the annihilation of an antiproton with a neutron to produce three pions.

1.7 Draw a lowest-order Feynman diagram for the following processes:

(a) e  elastic scattering;

(b) n ! p þ e
þ 
e ;
 PROBLEMS 31

(c) eþ e
! eþ e
;

(d) a fourth-order diagram for the reaction  þ  ! eþ þ e
.

1.8 Calculate the energy–momentum transfer between two particles equivalent to a

distance of approach of (a) 1fm and (b) 10
3 fm. Assuming that the intrinsic
strengths of the fundamental weak and electromagnetic interactions are approxi-
mately equal, compare the relative sizes of the invariant (scattering) amplitudes for
weak and electromagnetic processes at these two energy–momentum transfers.

h2 ½q2 þ m2 c2 
1 is the amplitude

1.9 Verify by explicit integration that f ðqÞ ¼
g2

corresponding to the Yukawa potential

g2 e
r=R
VðrÞ ¼
;
4 r

where R ¼

h=mc is the range and r ¼ jxj.

1.10 Two beams of particles consisting of n bunches with Ni ði ¼ 1; 2Þ particles in each,

traverse circular paths and collide ‘head-on’. Show that in this case the general
expression Equation (1.45) for the luminosity reduces to L ¼ nN1 N2 f =A, where A is
the cross-sectional area of the beam and f is the frequency, i.e. f ¼ 1=T, where T is
the time taken for the particles to make one traversal of the ring.

1.11 A thin (density 1 mg cm
2 ) target of 24 Mg (MA ¼ 24:3 atomic mass units) is
bombarded with a 10 nA beam of alpha particles. A detector subtending a solid
angle of 2
10
3 sr, records 20 protons/s. If the scattering is isotropic, determine
the cross-section for the 24 Mgð
; pÞ reaction.
2
Nuclear Phenomenology

In this chapter we start to examine some of the things that can be learned from
experiments, beginning with basic facts about nuclei, including what can be deduced
about their shapes and sizes. Then we discuss the important topic of nuclear stability
and the phenomenology of the various ways that unstable nuclei decay to stable
states. Finally, we briefly review the classification of reactions in nuclear physics.
Before that we need to introduce some notation.

2.1 Mass Spectroscopy and Binding Energies

Nuclei are specified by:

Z – atomic number ¼ the number of protons,

N – neutron number ¼ the number of neutrons,

A – mass number ¼ the number of nucleons, so that A ¼ Z þ N.

We will also refer to A as the nucleon number. The charge on the nucleus is
þZe, where e is the absolute value of the electric charge on the electron. Nuclei
with combinations of these three numbers are also called nuclides and are written
A
Y or AZ Y, where Y is the chemical symbol for the element. Some other common
nomenclature is:

nuclides with the same mass number are called isobars,

nuclides with the same atomic number are called isotopes,

nuclides with the same neutron number are called isotones.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
34 CH2 NUCLEAR PHENOMENOLOGY

The concept of isotopes was introduced in Chapter 1. For example, stable isotopes
of carbon are 12 C and 13 C, and the unstable isotope used in dating ancient objects
(see later in this chapter) is 14 C; all three have Z ¼ 6.
Just as in the case of electrons in atoms, the forces that bind the nucleons in
nuclei contribute to the total mass of an atom M(Z, A) and in terms of the masses of
the proton Mp and neutron Mn

MðZ; AÞ < Z ðMp þ me Þ þ N Mn : ð2:1Þ

The mass deficit is defined as

MðZ; AÞ  MðZ; AÞ  Z ðMp þ me Þ  N Mn ð2:2Þ

and 
Mc2 is called the binding energy B. Binding energies may be calculated if
masses are measured accurately. One way of doing this is by using the techniques
of mass spectroscopy. The principle of the method is shown in Figure 2.1.

Figure 2.1 Schematic diagram of a mass spectrometer (adapted from Kr88 Copyright John
Wiley & Sons, Inc.)

A source of ions of charge q, containing various isotopes passes through a region

where there are uniform electric (E) and magnetic (B1 ) fields at right angles. The
electric field will exert a force qE in one direction and the magnetic field will exert
a force qvB1 in the opposite direction, where v is the speed of the ions. By
balancing these forces, ions of a specific speed v ¼ E=B1 can be selected and
allowed to pass through a collimating slit. Ions with other velocities (shown as
dashed lines) are deflected. The beam is then allowed to continue through a second
 MASS SPECTROSCOPY AND BINDING ENERGIES 35

uniform magnetic field B2 where it will be bent into a circular path of radius ,
given by

mv ¼ qB2  ð2:3Þ

and since q, B2 and v are fixed, particles with a fixed ratio q=m will bend in a path
with a unique radius. Hence isotopes may be separated and focused onto a detector
(e.g. a photographic plate). In the common case where B1 ¼ B2 ¼ B,

q E
¼ 2 : ð2:4Þ
m B

In practice, to achieve high accuracy, the device is used to measure mass

differences rather than absolute values of mass.1
Conventional mass spectroscopy cannot be used to find the masses of very short-
lived nuclei and in these cases the masses are determined from kinematic analysis
of nuclear reactions as follows. Consider the inelastic reaction Aða; aÞA , where A
is the short-lived nucleus whose mass is to be determined. The kinematics of this
are:

aðEi ; pi Þ þ AðmA c2 ; 0Þ ! aðEf ; pf Þ þ A ðE

~; p
~Þ; ð2:5Þ

where we use tilded quantities to denote the energy, mass, etc. of A . Equating the
total energy before the collision

Etot ðinitialÞ ¼ Ei þ ma c2 þ mA c2 ð2:6aÞ

to the total energy after the collision

~ þ ma c2 þ m
Etot ðfinalÞ ¼ Ef þ E ~ c2 ð2:6bÞ

gives the following expression for the change in energy of the nucleus:

p2i p2f ~p2

~¼
~  mA Þc2 ¼ Ei  Ef  E

E  ðm   ; ð2:7Þ
2ma 2ma 2m ~

where we have assumed non-relativistic kinematics. If the initial momentum of the

projectile is along the x-direction and the scattering angle is , then from
momentum conservation,

ð~pÞx ¼ pi  pf cos  ; ð~pÞy ¼ pf sin  ð2:8Þ

1
Practical details of mass spectroscopy may be found in, for example, Chapter 3 of Kr88.
36 CH2 NUCLEAR PHENOMENOLOGY

and using these in Equation (2.7) gives

ma 
ma  2ma  1=2

E ¼ Ei 1   Ef 1 þ þ Ei Ef cos : ð2:9Þ
~
m ~
m ~
m

This formula can be used iteratively to deduce
E and hence the mass of the excited
nucleus A , from measurements of the initial and final energy of the projectile by
~ ¼ mA on the right-hand side because
E is small in comparison
initially setting m
with mA . One final point is that the energies in Equation (2.9) are measured in the
laboratory system, whereas the final energies (masses) will be needed in the centre-of-
mass system.2 The necessary transformation is easily found to be

ECM ¼ Elab ð1 þ ma =mA Þ1 : ð2:10Þ

A similar formula to Equation (2.9) may be derived for the general reaction
A(a,b)B:
   
ma mb 2  1=2

E ¼ Ei 1   Ef 1 þ þ ma mb Ei Ef cos  þ Q; ð2:11Þ
mB mB mB

where Q is the kinetic energy released in the reaction.

A commonly used quantity of interest is the binding energy per nucleon B=A.
This is shown schematically in Figure 2.2 for nuclei that are stable or long-lived.

Figure 2.2 Binding energy per nucleon as a function of mass number A for stable and long-
lived nuclei

2
A discussion of these two systems is given in Appendix C.
 NUCLEAR SHAPES AND SIZES 37

This shows that B=A peaks at a value of 8.7 MeV for a mass number of about 56
(close to iron) and thereafter falls very slowly. Excluding very light nuclei, the
binding energy per nucleon is between 7 and 9 MeV over a wide range of the
periodic table. In Section 2.5 we will discuss a model that provides an explanation
for the shape of this curve.

2.2 Nuclear Shapes and Sizes

The shape and size of a nucleus may be found from scattering experiments; i.e. a
projectile is scattered from the nucleus and the angular distribution of the scattered
particles examined, as was done by Rutherford and his collaborators when they
deduced the existence of the nucleus. The interpretation is simplest in those cases
where the projectile itself has no internal structure, i.e. it is an elementary particle,
and electrons are often used. In this case the relevant force is electromagnetic and we
learn about the charge distribution in the nucleus. The first experiments of this type
were performed by Hofstader and his collaborators in the late 1950s.3 If instead of an
electron a hadron is used as the projectile, the force is dominantly the nuclear strong
interaction and we find information about the matter density. Neutrons are com-
monly used so that Coulomb effects are absent. We discuss these two cases in turn.

2.2.1 Charge distribution

To find the amplitude for electron–nucleus scattering, we should in principle solve

the Schrödinger (or Dirac) equation using a Hamiltonian that includes the full
electromagnetic interaction and use nuclear wavefunctions. This can only be done
numerically. However, in Appendix C we derive a simple formula that describes the
electromagnetic scattering of a charged particle in the so-called Born approxima-
tion, which assumes Z 1 and uses plane waves for the initial and final states.
This leads to the Rutherford cross-section, which in its relativistic form may be
written
 
d Z 2 2 ðhcÞ2
¼ 2 4 ; ð2:12Þ
dO Rutherford 4E sin ð=2Þ

where E is the total initial energy of the projectile and  is the angle through which
it is scattered. Note that Equation (2.12) is of order 2 because it corresponds to
the exchange of a single photon. Although Equation (2.12) has a limited range of
applicability, it is useful to discuss the general features of electron scattering.
Equation (2.12) actually describes the scattering of a spin-0 point-like projectile of
unit charge from a fixed point-like target with electric charge Ze, i.e. the charge

3
Robert Hofstader shared the 1961 Nobel Prize in Physics for his pioneering electron scattering experiments.
38 CH2 NUCLEAR PHENOMENOLOGY

distribution of the target is neglected. It therefore needs to be modified in a number of

ways before it can be used in practice. We will state the modifications without proof.
Firstly, taking account of the electron spin leads to the so-called Mott cross-
section
   
d d  
¼ 1  2 sin2 ð=2Þ ; ð2:13Þ
dO Mott dO Rutherford

where ¼ v=c and v is the velocity of the initial electron. At higher energies, the
recoil of the target needs to be taken into account and this introduces a factor E0 =E
on the right-hand side of Equation (2.13), where E0 is the final energy of the
electron. At higher energies we also need to take account of the interaction with
the magnetic moment of the target in addition to its charge. The final form for the
differential cross-section is
   

d d E0 2
¼ 1 þ 2
tan ; ð2:14Þ
dO spin 1 dO Mott E 2
2

where
q2

¼ ð2:15Þ
4M 2 c2

and M is the target mass. Because the energy loss of the electron to the recoiling
nucleus is no longer negligible, q, the previous momentum transfer, has been
replaced by the four-momentum transfer q, whose square is

 4EE0
q2 ¼ ðp  p0 Þ2 ¼ 2m2e c2  2ðEE0 c2  jpjjp0 j cos Þ   2 sin2 ð=2Þ; ð2:16Þ
c

where pð p0 Þ is the four-momentum of the initial (final) electron. (Because q2 0, it

is common practice to replace it with Q2 ¼ q2 , so as to work with positive
quantities.4) For the rest of this discussion it will be sufficient to ignore the magnetic
interaction, although we will use a variant of the full form (2.16) in Chapter 6.
The final modification is due to the spatial extension of the nucleus. If the spatial
charge distribution within the nucleus is written f (x) then we define the form factor
F(q2) by
ð ð
1
Fðq2 Þ  eiqx=h f ðxÞ d3 x with Ze ¼ f ðxÞ d3 x; ð2:17Þ
Ze

4
To remove any confusion, in the non-relativistic case, which we use in the rest of this chapter, q is
interpreted to be q ¼ jqj  0 where q  p  p0 , as was used in Section 1.6.1. We will need the four-
momentum definition of q in Chapter 6.
 NUCLEAR SHAPES AND SIZES 39

i.e. the Fourier transform of the charge distribution.5 In the case of a spherically
symmetric charge distribution, the angular integrations in Equations (2.17) may be
done using spherical polar coordinates to give

ð
1

qr 
2 4h
Fðq Þ ¼ rðrÞsin dr; ð2:18Þ
Zeq h
0

where q ¼ jqj and ðrÞ is the radial charge distribution. The final form of the
experimental cross-section in this approximation is given by6
   
d d
¼ Fðq2 Þ 2 : ð2:19Þ
dO expt dO Mott

Two examples of measured cross-sections are shown in Figure 2.3. Striking

features are the presence of a number of well-defined minima superimposed on a

Publisher's Note:
Permission to reproduce this image
online was not granted by the
copyright holder. Readers are kindly
requested to refer to the printed version
of this chapter.

Figure 2.3 Elastic differential cross-sections as a function of the scattering angle for 450 MeV
electrons from 58 Ni and 758 MeV electrons from 48 Ca; the solid lines are fits as described in the
text (adapted from Si75 (58Ni data) and Be67 (48Ca data), Copyright American Physical Society)

5
Strictly this formula assumes that the recoil of the target nucleus is negligible and the interaction is
relatively weak, so that perturbation theory may be used.
6
If the magnetic interaction were included, another form factor would be necessary, as is the case in high-
energy electron scattering discussed in Chapter 6.
40 CH2 NUCLEAR PHENOMENOLOGY

rapid decrease in the cross-section with angle. These features are common to all
elastic data, although not all nuclei show so many minima as those shown.
The minima are due to the form factor and we can make this plausible by taking
the simple case where the nuclear charge distribution is represented by a hard
sphere such that

ðrÞ ¼ constant; r a
ð2:20Þ
¼0 r>a

where a is a constant. In this case, evaluation of Equation (2.18) gives

Fðq2 Þ ¼ 3½sinðbÞ  bcosðbÞb3 ; ð2:21Þ

where b  qa= h. Thus Fðq2 Þ will be zero at values of b for which b ¼ tanðbÞ. In
practice, as we will see below, ðrÞ is not a hard sphere, and although it is
approximately constant for much of the nuclear volume, it falls smoothly to zero at
the surface. Smoothing the edges of the radial charge distribution (2.20) modifies
the positions of the zeros, but does not alter the argument that the minima in the
cross-sections are due to the spatial distribution of the nucleus. Their actual
positions and depths result from a combination of the form factor and the form of
the point-like amplitude. We shall see below that the minima can tell us about the
size of the nucleus.
If one measures the cross-section for a fixed energy at various angles (and hence
various q2 ), the form factor can in principle be extracted using Equation (2.19) and
one might attempt to find the charge distribution from the inverse Fourier
transform
ð
Ze
f ðxÞ ¼ Fðq2 Þ eiqx=h d3 q: ð2:22Þ
ð2Þ3

However, q2 only has a finite range for a fixed initial electron energy and even within
this range the rapid fall in the cross-section means that in practice measurements
cannot be made over a sufficiently wide range of angles for the integral in Equation
(2.22) to be evaluated accurately. Thus, even within the approximations used,
reliable charge distributions cannot be found from Equation (2.22). Therefore
different strategies must be used to deduce the charge distribution. In one approach,
plausible – but very general – parameterized forms (for example a sum of Gaussians)
are chosen for the charge distribution and are used to modify the point-like
electromagnetic interaction. The resulting Schrödinger (or Dirac) equation is solved
numerically to produce an amplitude, and hence a cross-section, for electron–
nucleus scattering. The parameters of the charge distribution are then varied to give
a good fit of the experimental data. The solid curves in Figure 2.3 are obtained in
this way.
 NUCLEAR SHAPES AND SIZES 41

Figure 2.4 Radial charge distributions ch of various nuclei, in units of e fm3 ; the thickness
of the curves near r ¼ 0 is a measure of the uncertaintity in ch (adapted from Fr83)

Some radial charge distributions for various nuclei obtained by these methods
are shown in Figure 2.4. They are well represented by the form

0ch
ch ðrÞ ¼ ; ð2:23Þ
1 þ eðraÞ=b

where a and b for medium and heavy nuclei are found to be

a  1:07A1=3 fm and b  0:54 fm: ð2:24Þ

From this we can deduce that the charge density is approximately constant in the
nuclear interior and falls fairly rapidly to zero at the nuclear surface, as anticipated
above. The value of 0ch is in the range 0.06–0.08 for medium to heavy nuclei and
decreases slowly with increasing mass number.
42 CH2 NUCLEAR PHENOMENOLOGY

A useful quantity is the mean square charge radius,

ð
1
 2
r  r 2 ch ðrÞ dr: ð2:25Þ
0

This can be found from the form factor as follows. Expanding Equation (2.17) for
Fðq2 Þ gives
ð X1  
2 1 1 ijqjrcos n 3
Fðq Þ ¼ f ðxÞ d x ð2:26Þ
Ze n¼0
n! h

and after carrying out the angular integrations this becomes

ð
1 ð
1
2 4 2 4q2
Fðq Þ ¼ f ðrÞ r dr  f ðrÞ r 4 dr þ    : ð2:27Þ
Ze 6Zeh2
0 0

From the normalization of f ðxÞ, we finally have

q2  2 
Fðq2 Þ ¼ 1  r þ  ð2:28Þ
6h2
and thus the mean square charge radius can be found from

2
 2 2 dFðq Þ
r ¼ 6h ; ð2:29Þ
dq2 q2 ¼0

provided the formfactor

 can be measured at very small values of q2 . For medium
2 1=2 7
and heavy nuclei r is given approximately by
 2 1=2
r ¼ 0:94A1=3 fm: ð2:30Þ

The nucleus is often approximated by a homogeneous charged sphere. The radius

R of this sphere is then quoted
 as the nuclear radius. The relation of this to the
mean square radius is R2 ¼ 53 r 2 , so that

Rcharge ¼ 1:21 A1=3 fm: ð2:31Þ

2.2.2 Matter distribution

Electrons cannot be used to obtain the distributions of neutrons in the nucleus. We

could, however, take the presence of neutrons into account by multiplying ch ðrÞ

7
The constant comes from a fit to a range of data, e.g. the compilation for 55 A 209 given in Ba77.
 NUCLEAR SHAPES AND SIZES 43

by A=Z. Then we find an almost identical nuclear density in the nuclear interior for
all nuclei, i.e. the decrease in 0ch with increasing A is compensated by the increase
in A=Z with increasing A. The interior nuclear density is given by

nucl  0:17 nucleons=fm3 : ð2:32Þ

Likewise, the effective nuclear matter radius for medium and heavy nuclei is

Rnuclear  1:2 A1=3 fm: ð2:33Þ

These are important results that will be used extensively later in this chapter and
elsewhere in this book.
To probe the nuclear (i.e. matter) density of nuclei experimentally, a strongly
interacting particle, i.e. a hadron, has to be used as the projectile. At high energies,
where elastic scattering is only a small part of the total interaction, the nucleus
behaves more like an absorbing sphere. In this case, the incident particle of
momentum p will have an associated quantum mechanical wave of wavelength
¼ h=p and will suffer diffraction-like effects, as in optics. To the extent that we are
dealing at high energies purely with the nuclear strong interaction (i.e. neglecting the
Coulomb interaction), the nucleus can be represented by a black disk of radius R and
the differential cross-section will have a Fraunhofer-like diffraction form, i.e.

d J1 ðqRÞ 2
/ ; ð2:34Þ
dO qR

where qR  pR for small  and J1 is a first-order Bessel function. For large qR,
 

2 
½J1 ðqRÞ2  sin2 qR  ; ð2:35Þ
qR 4

which has zeros at intervals
 ¼ =pR. The plausibility of this interpretation is

borne out by experiment, an example of which is shown in Figure 2.5. The data show
a succession of roughly equally spaced minima as suggested by Equation (2.35).
To go further requires solving the equations of motion, but this is far more
problematical than in the electron case because the hadrons are more likely to be
absorbed as they pass through the nucleus and the effective potential is far less
well known. However, the analogy with optics can be pursued further in the so-
called optical model. The essential idea in this model is that a hadron incident on a
nucleus may be elastically scattered, or it may cause a variety of different
reactions. As in the discussion above, if the incident particle is represented by a
wave, then in classical language it may be scattered or it may be absorbed. In
optics this is analogous to the refraction and absorption of a light wave by a
medium of complex refractive index, and just as the imaginary part of the
refractive index takes account of the absorption of the light wave, so in the
44 CH2 NUCLEAR PHENOMENOLOGY

Figure 2.5 Elastic differential cross-sections for 52 MeV deuterons on 54 Fe (adapted from Hi68,
copyright Elsevier, with permission)

nuclear case the imaginary part of a complex potential describing the interaction
takes account of all the inelastic reactions. It is an essential feature of the model
that the properties of nuclei are mainly determined by their size, as this implies
that the same potential can account for the interaction of particles of different
energies with different nuclei. Apart from the theoretical basis provided by
analogy with classical optics, the model is essentially phenomenological, in that
the values of the parameters of the optical potentials are found by optimizing the fit
to the experimental data. This type of semi-phenomenological approach is
common in both nuclear and particle physics.
In practice, the Schrödinger equation is solved using a parameterized complex
potential where the real part is a sum of the Coulomb potential (for charged
projectiles), an attractive nuclear potential and a spin-orbit potential, and the
imaginary part is assumed to cause the incoming wave of the projectile to be
attenuated within the nucleus, thereby allowing for inelastic effects. Originally,
mathematical forms like Equation (2.23) were used to parameterize the real and
imaginary parts of the potential, but subsequent work indicated substantial differ-
ences between the form factors of the real and imaginary parts of the potential and so
different forms are now used for the imaginary part. The free parameters of the total
potential are adjusted to fit the data.
The optical model has achieved its greatest success in the scattering of nucleons,
but analyses using data obtained from light nuclei targets are also possible. A wide
range of scattering data can be accounted for to a high degree of precision by
 NUCLEAR INSTABILITY 45

Figure 2.6 Differential cross-sections (normalized to the Rutherford cross-section) for the
elastic scattering of 30.3 MeV protons, for a range of nuclei compared with optical model
calculations; the solid and dashed lines represent the results using two different potentials
(adapted from Sa67, copyright Elsevier, with permission)

the model and examples of this are shown in Figure 2.6. The corresponding
wavefunctions are extensively used to extract information on nuclear structure.
The conclusions are in accord with those above deduced indirectly from electron
data.

2.3 Nuclear Instability

Stable nuclei only occur in a very narrow band in the Z N plane close to the line
Z ¼ N (see Figure 2.7). All other nuclei are unstable and decay spontaneously in
various ways. Isobars with a large surplus of neutrons gain energy by converting a
46 CH2 NUCLEAR PHENOMENOLOGY

Figure 2.7 The distribution of stable nuclei: the squares are the stable and long-lived nuclei
occurring in nature; other known nuclei lie within the jagged lines and are unstable. [adapted
from Ch97.)

neutron into a proton; conversely, a nucleus with a large surplus of protons converts
protons to neutrons. These are examples of -decays, already mentioned. A related
process is where an atomic electron is captured by the nucleus and a proton is
thereby converted to a neutron within the nucleus. This is electron capture and like
-decay is a weak interaction. The electron is usually captured from the innermost
shell and the process competes with -decay in heavy nuclei because the radius of
this shell (the K-shell) is close to the nuclei radius. The presence of a third particle in
the decay process, the neutrino (as first suggested by Fermi), means that the emitted
electrons (or positrons) have a continuous energy spectrum. The derivation and
analysis of the electron momentum spectrum will be considered in Chapter 7 when
we discuss the theory of -decay.
The maximum of the curve of binding energy per nucleon is at approximately
the position of iron (Fe) and nickel (Ni), which are therefore the most stable
nuclides. In heavier nuclei, the binding energy is smaller because of the larger
Coulomb repulsion. For still heavier nuclear masses, nuclei can decay sponta-
neously into two or more lighter nuclei, provided the mass of the parent nucleus is
larger than the sum of the masses of the daughter nuclei.
 RADIOACTIVE DECAY 47

Most such nuclei decay via two-body decays and the commonest case is when
one of the daughter nuclei is a 4 He nucleus (i.e. an -particle: 4 He  2p2n, with
A ¼ 4 ; Z ¼ N ¼ 2). The -particle is favoured in such decays because it is a very
stable, tightly bound structure. Because this is a two-body decay, the -particle has
a unique energy and the total energy released, the so-called Q-value, is given by:

Q ¼ ðMP  MD  M Þc2 ¼ ED þ E ; ð2:36Þ

where the subscripts refer to parent and daughter nuclei and the -particle, and E is
a kinetic energy.
The term fission is used to describe the rare cases where the two daughters have
similar masses. If the decay occurs without external action, it is called spontaneous
fission to distinguish it from induced fission, where some external stimulus is required
to initiate the decay. Spontaneous fission only occurs with a probability greater than that
for -emission for nuclei with Z  110. The reason for this is discussed in Section 2.7.
Finally, nuclei may decay by the emission of photons, with energies in the -ray
part of the electromagnetic spectrum (gamma emission). This occurs when an
excited nuclear state decays to a lower state and is a common way whereby excited
states lose energy. The lower energy state is often the ground state. A competing
process is internal conversion, where the nucleus de-excites by ejecting an electron
from a low-lying atomic orbit. Both are electromagnetic processes. Electromag-
netic decays will be discussed in more detail in Chapter 7.

2.4 Radioactive Decay

Before looking in more detail at different classes of instability, we will consider
the general formalism describing the rate of radioactive decay. The probability per
unit time that a given nucleus will decay is called its decay constant and is
related to the activity A by
A ¼ dN=dt ¼ N; ð2:37Þ

where N(t) is the number of radioactive nuclei in the sample at time t. The activity is
measured in becquerels (Bq), which is one decay per second.8 The probability here
refers to the total probability, because could be the sum of decay probabilities for a
number of distinct final states in the same way that the total decay width of an
unstable particle is the sum of its partial widths. Integrating Equation (2.37) gives
AðtÞ ¼ N0 expð tÞ; ð2:38Þ

where N0 is the initial number of nuclei, i.e. the number at t ¼ 0.

8
An older unit, the curie (1 Ci ¼ 3:7  1010 Bq) is also still in common use. A typical laboratory radioactive
source has an activity of a few tens of kBq, i.e. Ci.
48 CH2 NUCLEAR PHENOMENOLOGY

The mean lifetime

of an unstable state, such as a radioactive nucleus or a
hadron, follows from the general definition of a mean x of a distribution f(x):
ð
 ð

x  x f ðxÞdx f ðxÞdx : ð2:39Þ

Thus
Ð
1
Ð t exp½ t dt
t dNðtÞ 0 1

 Ð ¼ 1 ¼ : ð2:40Þ
dNðtÞ Ð
exp½ t dt
0

This is the quantity we simply called ‘the lifetime’ in Chapter 1. The mean
lifetime is always used in particle physics, but another measure more commonly
used in nuclear physics is the half-life t12 , defined as the time for the number of
nuclei to fall by one half. Thus t12 ¼ ln2= ¼
ln2. In this book, the term lifetime
will be used for the mean lifetime, both for radioactive nuclei and unstable
hadrons, unless explicitly stated otherwise.
A well-known use of the radioactive decay law is in dating ancient specimens
using the known properties of radioactive nuclei. For organic specimens, carbon is
usually used. Carbon-14 is a radioactive isotope of carbon that is produced by the
action of cosmic rays on nitrogen in the atmosphere.9 If the flux of cosmic rays
remains roughly constant over time, then the ratio of 14 C to the stable most abundant
isotope 12 C reaches an equilibrium value of about 1 : 1012 . Both isotopes will be
taken up by living organisms in this ratio, but when the organism dies there is no
further interaction with the environment and the ratio slowly changes with time as
the 14 C nuclei decay by -decay to 14 N with a lifetime of 8:27  103 years. Thus, if
the ratio of 14 C to 12 C is measured, the age of the specimen may be estimated.10 The
actual measurements can be made very accurately because modern mass spectro-
meters can directly measure very small differences in the concentrations of 14 C and
12
C using only milligrams of material. Nevertheless, in practice, corrections are
made to agree with independent calibrations if possible, using, for example, tree-
ring growth data, because cosmic ray activity is not strictly constant with time.
In many cases the products of radioactive decay are themselves radioactive and so
a decay chain results. Consider a decay chain A ! B ! C !   , with decay
constants A ; B ; C etc.. The variation of species A with time is given by Equation
(2.38), i.e.
NA ðtÞ ¼ NA ð0Þexpð A tÞ; ð2:41Þ

9
Cosmic rays are high-energy particles, mainly protons, that impinge on the Earth’s atmosphere from space.
The products of the secondary reactions they produce may be detected at the Earth’s surface. Victor Hess
shared the 1936 Nobel Prize in Physics for the discovery of cosmic radiation.
10
This method of using radioactive carbon to date ancient objects was devised by Willard Libby, for which he
received the 1960 Nobel Prize in Chemistry.
 RADIOACTIVE DECAY 49

but the differential equation for NB ðtÞ will have an extra term in it to take account
of the production of species B from the decay of species A:

dNB ðtÞ=dt ¼  B NB þ A NA : ð2:42Þ

The solution of this equation may be verified by substitution to be

A
NB ðtÞ ¼ NA ð0Þ½expð A tÞ  expð B tÞ: ð2:43Þ
B  A

Similar equations may be found for decay sequences with more than two stages.
Thus, for a three-stage sequence

expð A tÞ
NC ðtÞ ¼ A B NA ð0Þ
ð B  A Þð C  A Þ

expð B tÞ expð C tÞ
þ þ ð2:44Þ
ð A  B Þð C  B Þ ð A  C Þð B  C Þ

As an example, the variation of the components as a function of time is shown in

Figure 2.8 for the specific case:
79 þ
38 Sr ! 79
37 Rb þ e þ e ð2:25 minÞ
j þ
! 79
36 Kr þ e þ e ð22:9 minÞ ð2:45Þ
j 79 þ
! 35 Br þ e þ e ð35:04 hoursÞ

where the final nucleus is stable.

Figure 2.8 Time variation of the relative numbers of nuclei in the decay chain (2.45)
50 CH2 NUCLEAR PHENOMENOLOGY

This illustrates the general features that whereas NA ðtÞ for the initial species falls
monotonically with time and NC ðtÞ for the final stable species rises monotonically,
NB ðtÞ for an intermediate species rises to a maximum before falling. Note that at
any time the sum of the components is a constant, as expected.
In the following sections we consider the phenomenology of the various types of
radioactivity in more detail and in Chapter 7 we will return to discuss various
models and theories that provide an understanding of these phenomena.

2.5 Semi-Empirical Mass Formula: The Liquid Drop Model

Apart from the lightest elements and a few special isolated very stable nuclei, the
binding energy data of Figure 2.2 can be fitted by a simple formula containing just a
few free parameters. This is the semi-empirical mass formula (SEMF), first written
down in 1935 by Weizsäcker. It is a semi-empirical formula, because although it
contains a number of constants that have to be found by fitting experimental data, the
formula does have a theoretical basis. This arises from the two properties common to
all nuclei (except those with very small A values) that we have seen earlier: (1) the
interior mass densities are approximately equal, and (2) their total binding energies
are approximately proportional to their masses. There is an analogy here with a
classical model of a liquid drop, where for drops of various sizes: (1) interior densities
are the same, and (2) latent heats of vaporization are proportional to their masses.11
However, the analogy of a nucleus as an incompressible liquid droplet, with the
nucleons playing the role of individual molecules within the droplet, cannot be taken
too far because nucleons of course obey the laws of quantum, not classical, physics.
The semi-empirical mass formula will be taken to apply to atomic masses, as
these are the masses actually observed in experiment. The atomic mass M(Z, A)
may then be written as the sum of six terms fi ðZ; AÞ:
X
5
MðZ; AÞ ¼ fi ðZ; AÞ: ð2:46Þ
i¼0

The first of these is the mass of the constituent nucleons and electrons,

f0 ðZ; AÞ ¼ Z ðMp þ me Þ þ ðA  ZÞMn : ð2:47Þ

The remaining terms are various corrections, which we will write in the form ai
multiplied by a function of Z and A with ai > 0.
The most important correction is the volume term,

f1 ðZ; AÞ ¼ a1 A: ð2:48Þ

11
Latent heat is the average energy required to disperse the liquid drop into a gas and so is analogous to the
binding energy per nucleon.
 SEMI-EMPIRICAL MASS FORMULA: THE LIQUID DROP MODEL 51

This arises from the fact that the strong nuclear force is short-range and each
nucleon therefore feels the effect of only the nucleons immediately surrounding it
(the force is said to be saturated), independent of the size of the nucleus. Recalling
the important result deduced in Section 2.2 that the nuclear radius is proportional
1
to A3 , this leads immediately to the binding energy being proportional to the
volume, or nuclear mass. The coefficient is negative, i.e. it increases the binding
energy, as expected.
The volume term overestimates the effect of the nuclear force because nucleons
at the surface are not surrounded by other nucleons. Thus the volume term has to
be corrected. This is done by the surface term
2
f2 ðZ; AÞ ¼ þa2 A3 ; ð2:49Þ

which is proportional to the surface area and decreases the binding energy. In the
classical model of a real liquid drop, this term would correspond to the surface
tension energy.
The Coulomb term accounts for the Coulomb energy of the charged nucleus, i.e.
the fact that the protons repel each other. If we have a uniform charge distribution
1
of radius proportional to A3 , then this term is

ZðZ  1Þ Z2
f3 ðZ; AÞ ¼ þa3 1  þa3 1 ; ð2:50Þ
A3 A3

where the approximation is sufficiently accurate for the large values of Z we will
be considering. A similar effect would be present for a charged drop of a classical
liquid.
The next term is the asymmetry term.

ðZ  A=2Þ2
f4 ðZ; AÞ ¼ þa4 : ð2:51Þ
A

This accounts for the observed tendency for nuclei to have Z ¼ N. (There are no
stable nuclei with very large neutron or proton excesses – c.f. Figure 2.7.) This
term is purely quantum mechanical in origin and is due to the Pauli principle.
Part of the reason for the form (2.51) can be seen from the diagram of Figure 2.9,
which shows the energy levels of a nucleus near the highest filled level in the
approximation where all the energy levels are separated by the same energy
.
Keeping A fixed and removing a proton from level 3 and adding a neutron to level
4, gives ðN  ZÞ ¼ 2 and leads to an energy increase of
. Repeating this for more
protons, we find that the transfer of ðN  ZÞ=2 nucleons decreases the binding
energy by an amount 
ðN  ZÞ2 =4. Although we have assumed
is a constant,
in practice it decreases like A1 ; hence the final form of the asymmetry term.
If we start with an even number of nucleons and progressively fill states, then the
lowest energy will be when both Z and N are even. If, on the other hand, we have a
52 CH2 NUCLEAR PHENOMENOLOGY

Figure 2.9 Schematic diagram of nuclear energy levels near the highest filled levels

system where both Z and N are odd and the highest filled proton state is above the
highest filled neutron state, we can increase the binding energy by removing one
proton from the nucleus and adding one neutron. If the highest filled proton state is
below the highest filled neutron state, then we can produce the same effect by
removing a neutron and adding a proton. These observations are summarized in the
empirical pairing term, which maximizes the binding when both Z and N are even:

f5 ðZ; AÞ ¼ f ðAÞ; if Z even; A  Z ¼ N even

f5 ðZ; AÞ ¼ 0; if Z even; A  Z ¼ N odd; or; Z odd; A  Z ¼ N even ð2:52Þ
f5 ðZ; AÞ ¼ þf ðAÞ; if Z odd; A  Z ¼ N odd
1
The exact form of the function f ðAÞ is found by fitting the data; f ðAÞ ¼ a5 A2 is
often used.
To help remember these terms, the notation VSCAP is frequently used, with

a1 ¼ av ; a2 ¼ as ; a3 ¼ ac ; a4 ¼ aa ; a5 ¼ ap : ð2:53Þ

Precise values of the coefficients

 depend on the range of A fitted. One commonly
used set is, in units of MeV c2 :12

av ¼ 15:56; as ¼ 17:23; ac ¼ 0:697; aa ¼ 93:14; ap ¼ 12: ð2:54Þ

The fit to the binding energy data for A > 20 using these coefficients in the SEMF
is shown in Figure 2.10. Overall the fit to the data is remarkably good for such a
simple formula, but is not exact of course. For example, there are a small number of
regions where the binding energy curves show enhancements that are not repro-
duced. (These enhancements are due to the existence of a ‘shell structure’ of
nucleons within the nucleus and will be discussed in Chapter 7.) Nevertheless, the
SEMF gives accurate values for the binding energies for some 200 stable and many

12
Note that some authors write the asymmetry term proportional to ðZ  NÞ2 , which is equivalent to the form
used here, but their value for the coefficient aa will differ by a factor of four from the one in Equations (2.54).
 SEMI-EMPIRICAL MASS FORMULA: THE LIQUID DROP MODEL 53

Figure 2.10 Fit to binding energy data (shown as solid circles) for odd-A and even-A nuclei
using the SEMF with the coefficients given in the text; the predictions are shown as open circles
and do not lie on smooth curves because A is not a function of Z
54 CH2 NUCLEAR PHENOMENOLOGY

more unstable nuclei. We will use it to analyse the stability of nuclei with respect to
-decay and fission. The discussion of -decay is deferred until Chapter 7.
Using the numerical values of Equation (2.54), the relative sizes of each of
the terms in the SEMF may be calculated and for the case of odd-A are shown in
Figure 2.11. In this diagram, the volume term is shown as positive and the other terms
are subtracted from it to give the final SEMF curve.

Figure 2.11 Contributions to the binding energy per nucleon as a function of mass number
for odd-A from each term in the SEMF; the surface, asymmetry and Coulomb terms have been
plotted so that they subtract from the volume term to give the total SEMF result in the lowest
curve

Finally, from its definition, one might expect the binding energy per nucleon to
be equivalent to the energy needed to remove a nucleon from the nucleus.
However, to remove a neutron from a nucleus corresponds to the process
A
ZY ! A1Z Y þ n ð2:55aÞ

and requires an energy change

En ¼ ½MðZ; A  1Þ þ Mn  MðZ; AÞc2 ¼ BðZ; AÞ  BðZ; A  1Þ; ð2:55bÞ

whereas the removal of a proton corresponds to the process

A
ZY ! A1
Z1 X þ p; ð2:56aÞ
 -DECAY PHENOMENOLOGY 55

where X is a different chemical species to Y, and requires an energy change

 
Ep ¼ MðZ  1; A  1Þ þ Mp þ me  MðZ;AÞ c2 ¼ BðZ;AÞ  BðZ  1;A  1Þ þ me c2 :
ð2:56bÞ

Thus, Ep and En are only equal to the binding energy per nucleon in an average
sense. In practice, measurements show that Ep and En can differ substantially from
this average and from each other at certain values of (Z, A). We will see in
Chapter 7 that one reason for this is the existence of a shell structure for nucleons
within nuclei, similar to the shell structure of electrons in atoms, which is ignored
in the liquid drop model.

2.6 b-Decay Phenomenology

By rearranging terms, the SEMF (2.46) may be written


MðZ; AÞ ¼  A  Z þ  Z 2 þ 1 ; ð2:57Þ
A2

where
as
aa
 ¼ Mn  av þ 1 þ
A 34
¼ aa þ ðMn  Mp  me Þ
aa ac ð2:58Þ
¼ þ 1
A A3
 ¼ ap

MðZ; AÞ is thus a quadratic in Z at fixed A and has a minimum at Z ¼ =2. For a

fixed value of A, a stable nucleus will have an integer value of Z closest to the
solution of this equation. For odd A, the SEMF is a single parabola, but for even A
the even–even and odd–odd nuclei lie on two distinct vertically shifted parabolas,
because of the pairing term. The nucleus with the smallest mass in an isobaric
spectrum is stable with respect to -decay. We will consider the two cases of odd
and even A separately, using specific values of A to illustrate the main features.

2.6.1 Odd-mass nuclei

Odd-mass nuclei can arise from even-N, odd-Z, or even-Z, odd-N configurations
and in practice the number of nuclei that are stable against -decay are roughly
equally distributed between these two types. The example we take is the case of
56 CH2 NUCLEAR PHENOMENOLOGY

Figure 2.12 Mass parabola of the A ¼ 111 isobars: the circles are experimental data and the
curve is the prediction of the SEMF -- possible -decays are indicated by arrows

the A ¼ 111 isobars, which are shown in Figure 2.12. The circles show the
experimental data as mass excess values in atomic mass units, where

mass excess  MðZ; AÞ ðin atomic mass unitsÞ  A ð2:59Þ

and the atomic mass unit (u) is defined as one twelfth of the mass of the neutral
atom 126 C.
The curve is the theoretical prediction from the SEMF using the numerical
values of the coefficients (2.54). The exact form of the curve depends on the
precise values of these coefficients. The minimum of the parabola corresponds to
the isobar 111
48 Cd with Z ¼ 48.
Isobars with more neutrons, such as 111 111 111
45 Rh, 46 Pd and 47 Ag, decay by converting
a neutron to a proton, i.e.
n ! p þ e þ e ; ð2:60Þ

so that
111 
45 Rh ! 111
46 Pd þ e þ 
e ð11 sÞ; ð2:61aÞ

111 
46 Pd ! 111
47 Ag þ e þ 
e ð22:3 minÞ ð2:61bÞ

and
111 
47 Ag ! 111
48 Cd þ e þ 
e ð7:45 daysÞ ð2:61cÞ
 -DECAY PHENOMENOLOGY 57

This decay sequence is shown in Figure 2.12. Electron emission is energetically

possible whenever the mass of the daughter atom MðZ þ 1; AÞ is smaller than its
isobaric neighbour, i.e.
MðZ; AÞ > MðZ þ 1; AÞ: ð2:62Þ

Recall that we are referring here to atoms, so that the rest mass of the created
electron is automatically taken into account.
Isobars with proton excess decay via

p ! n þ eþ þ e ; ð2:63Þ

i.e. positron emission, which although not possible for a free proton, is possible in a
nucleus because of the binding energy. So for example, the nuclei
111 111 111
51 Sb; 50 Sn and 49 In could, in principle, decay by positron emission, which is
energetically possible if
MðZ; AÞ > MðZ  1; AÞ þ 2me ; ð2:64Þ

this takes account of the creation of a positron and the existence of an excess of
electrons in the parent atom.
It is also theoretically possible for this sequence of transitions to occur by
electron capture. This mainly occurs in heavy nuclei, where the electron orbits are
more compact. It is usually the electron in the innermost shell (i.e. the K-shell) that
is captured. Capture of such an electron gives rise to a ‘hole’ and causes electrons
from higher levels to cascade downwards and in so doing emit characteristic
X-rays. Electron capture is energetically allowed if
MðZ; AÞ > MðZ  1; AÞ þ "; ð2:65Þ

where " is the excitation energy of the atomic shell of the daughter nucleus. The
process competes with positron emission and in practice for the nuclei above this is
what happens. Thus, we have

e þ 111 111
51 Sb ! 50 Sn þ e ð75 sÞ; ð2:66aÞ

e þ 111 111
50 Sn ! 49 In þ e ð35:3 minÞ ð2:66bÞ

and

e þ 111 111
49 In ! 48 Cd þ e ð2:8 daysÞ; ð2:66cÞ

which are manifestations of the primary reaction

e þ p ! n þ e : ð2:67Þ

So once again we arrive at the stable isobar.

58 CH2 NUCLEAR PHENOMENOLOGY

2.6.2 Even-mass nuclei

Even-mass nuclei can arise from even-N, even-Z, or odd-Z, odd-N configurations,
but for reasons that are explained below, nearly all even-mass nuclei that are stable
against -decay are of the even–even type, with only a handful of odd–odd types
known. Consider as an example the case of A ¼ 102 shown in Figure 2.13. (Recall
that the plot is of mass excess, which is a very small fraction of the total mass.)

Figure 2.13 Mass parabolas of the A ¼ 102 isobars: the circles are experimental data (open
circles are even--even nuclei and closed circles are odd--odd nuclei); the curves are the
prediction of the SEMF (upper curve is for odd--odd nuclei and lower curve for even--even nuclei)
and possible -decays are indicated by arrows

The lowest isobar is 102 102

44 Ru and is -stable. The isobar 46 Pd is also stable since
its two odd–odd neighbours both lie above it. In principle, the two nuclei could be
connected by the reaction
102
46 Pd þ 2e ! 102
44 Ru þ 2e ; ð2:68Þ

but this would involve a ‘double electron capture’ and would be heavily suppressed.
The reaction has never been observed. Thus there are two -stable isobars. This is a
common situation for A-even, although no two neighbouring isobars are known to be
stable. Odd–odd nuclei always have at least one more strongly bound even–even
neighbour nucleus in the isobaric spectrum. They are therefore unstable. The only
exceptions to this rule are a few very light nuclei.
The lifetime of a free neutron is about 887 s. The free proton is believed to be
stable and can only ‘decay’ within a nucleus by utilizing the binding energy.
Lifetimes of emitters vary enormously from milliseconds to 1016 years. They
 FISSION 59

depend very sensitively on the Q-value for the decay and on the properties of the
nuclei involved, e.g. their spins.

2.7 Fission
Spontaneous fission has been defined as the process whereby a parent nucleus breaks
into two daughter nuclei of approximately equal masses without external action.
Precisely equal masses are very unlikely and in the most probable cases the daughter
nuclei have mass numbers that differ by about 45, with peaks around mass numbers
95 and 140. The reason for this is unknown. The binding energy curve shows that
spontaneous fission is energetically possible for nuclei with A > 100.13 An example is
238
92 U ! 145 90
57 La þ 35 Br þ 3n; ð2:69Þ

with a release of about 154 MeV of energy, which is carried off as kinetic energy of
the fission products. Heavy nuclei are neutron-rich and so necessarily produce
neutron-rich decay products, including free neutrons. The fission products are
themselves usually some way from the line of -stability and will decay by a series
of steps. For example, 145 145
57 La decays to the -stable 60 Nd by three stages, releasing a
further 8.5 MeV of energy, which in this case is carried off by the electrons and
neutrinos emitted in -decay. Although the probability of fission increases with
increasing A, it is still a very rare process. For example, in 238
92 U, the transition rate for
spontaneous fission is about 3  1024 s1 compared with about 5  1018 s1 for
-decay, a branching fraction of 6  107 . Spontaneous emission only becomes
dominant in very heavy elements with A  270, as we shall now show.
To understand spontaneous fission we can again use the liquid drop model. In the
SEMF we have assumed that the drop (i.e. the nucleus) is spherical, because this
minimizes the surface area. However, if the surface is perturbed for some reason from
spherical to prolate, the surface term in the SEMF will increase and the Coulomb term
will decrease (assuming the volume remains the same) and the relative sizes of these
two changes will determine whether the nucleus is stable against spontaneous fission.
For a fixed volume we can parametrize the deformation by the semi-major and
semi-minor axes of the ellipsoid a and b, respectively as shown in Figure 2.14. One
possible parametrization that preserves the volume is
1
a ¼ R ð1 þ "Þ; b ¼ R=ð1 þ "Þ2 ; ð2:70Þ

where " is a small parameter, so that

4 4
V ¼  R3 ¼  ab2 : ð2:71Þ
3 3

13
Fission in heavy nuclei was discovered by Otto Hahn, for which he received the 1944 Nobel Prize in
Chemistry.
60 CH2 NUCLEAR PHENOMENOLOGY

R
a

Figure 2.14 Deformation of a heavy nucleus

To find the new surface and Coulomb terms one has to find the expression for the
surface of the ellipsoid in terms of a and b and expand it in a power series in ". The
algebra is unimportant and the results are:
 
2 2
Es ¼ as A3 1 þ "2 þ . . . ð2:72aÞ
5

and
 
1 1
Ec ¼ ac Z 2 A3 1  "2 þ . . . : ð2:72bÞ
5

Hence the change in the total energy is

"2
2 1


E ¼ ðEs þ Ec Þ  ðEs þ Ec ÞSEMF ¼ 2as A3  ac Z 2 A3 : ð2:73Þ
5

If
E < 0, then the deformation is energetically favourable and fission can occur.
From Equation (2.73), this happens if

Z 2 2as
  49; ð2:74Þ
A ac
where we have used experimental values for the coefficients as and ac given in
Equations (2.54). The inequality is satisfied for nuclei with Z > 116 and A  270.
Spontaneous fission is a potential barrier problem and this is shown in Figure 2.15.
The solid line corresponds to the shape of the potential in the parent nucleus. The
activation energy shown in Figure 2.15 determines the probability of spontaneous
fission. To fission, the nucleus could in principle tunnel through the barrier, but the
fragments are large and the probability for this to happen is extremely small.14 For
heavy nuclei the activation energy is about 6 MeV, but disappears for very heavy
nuclei. For such nuclei, the shape of the potential corresponds closer to the dashed
line and the slightest deformation will induce fission.

14
The special case of -decay will be discussed in Chapter 7. There we will show that the lifetime for such
decays is expected to have an exponential dependence on the height of the fission barrier and this is observed
qualitatively in fission data.
 FISSION 61

Figure 2.15 Potential energy during different stages of a fission reaction

Another possibility for fission is to supply the energy needed to overcome the
barrier by a flow of neutrons. Because of the absence of a Coulomb force, a neutron
can get very close to the nucleus and be captured by the strong nuclear attraction.
The parent nucleus may then be excited to a state above the fission barrier and
therefore split up. This process is an example of induced fission. Neutron capture by
a nucleus with an odd neutron number releases not just some binding energy, but also
a pairing energy. This small extra contribution makes a crucial difference to nuclear
fission properties. For example, very low-energy (‘thermal’) neutrons can induce
fission in 235 U, whereas only higher energy (‘fast’) neutrons induce fission in 238 U.
This is because 235 U is an even–odd nucleus and 238 U is even–even. Therefore, the
ground state of 235 U will lie higher (less tightly bound) in the potential well of its
fragments than that of 238 U. Hence to induce fission, a smaller energy will be needed
for 235 U than for 238 U. In principle, fission may be induced in 235 U using even zero-
energy neutrons.15
We consider this quantitatively as follows. The capture of a neutron by 235 U
changes an even–odd nucleus to a more tightly bound even–even (compound)
nucleus of 236 U and releases the binding energy of the last neutron. In 235 U this is
6.5 MeV. As the activation energy (the energy needed to induce fission) is about
5 MeV for 236 U, neutron capture releases sufficient energy to fission the nucleus. The
kinetic energy of the incident neutron is irrelevant and even zero-energy neutrons
can induce fission in 235 U. In contrast, neutron capture in 238 U changes it from an
even–even nucleus to an even–odd nucleus, i.e. changes a tightly bound nucleus to a
less tightly bound one. The energy released (the binding energy of the last neutron) is
about 4.8 MeV in 239 U and is less than the 6.5 MeV required for fission. For this
reason, fast neutrons with energy of at least the difference between these two
energies are required to fission 238 U.

15
Enrico Fermi was a pioneer in the field of induced fission and received the 1938 Nobel Prize in Physics for
‘demonstrations of the existence of new radioactive elements produced by neutron irradiation, and for his
related discovery of nuclear reactions brought about by slow neutrons’. Fermi’s citation could equally have
been about his experimental discoveries and theoretical work in a wide range of areas from nuclear and
particle physics to solid-state physics and astrophysics. He was probably the last ‘universal physicist’.
62 CH2 NUCLEAR PHENOMENOLOGY

2.8 c-Decays
When a heavy nucleus disintegrates by either - or -decay, or by fission, the daughter
nucleus is often left in an excited state. If this state is below the excitation energy for
fission, it will de-excite, usually by emitting a high-energy photon. The energy of these
photons is determined by the average energy level spacings in nuclei and ranges from a
few to several MeV. They are in the gamma ray () part of the electromagnetic
spectrum. Because -decay is an electromagnetic process, we would expect the
typical lifetime of an excited state to be 1016 s. In practice, lifetimes are very
sensitive to the amount of energy released in the decay and in the nuclear case other
factors are also very important, particularly the quantity of angular momentum carried
off by the photon. Typical lifetimes of nuclear levels are about 1012 s.
The role of angular momentum in -decays is crucial. If the initial (excited)
state has a total spin Si and the final nucleus has a total spin Sf , then the total
angular momentum J of the emitted photon is given by

J ¼ Si  Sf ; ð2:75Þ

with

Si þ Sf  J  jSi  Sf j; ð2:76Þ

where S ¼ jSj; J ¼ jJj. In addition,

mi ¼ M þ mf ; ð2:77Þ

where m are the corresponding magnetic quantum numbers. Both total angular
momentum and its magnetic quantum number are conserved in -decays.
-decays are further complicated because parity is conserved in these electro-
magnetic processes. Both the initial and final nuclear level will have an intrinsic
parity, as does the photon, and in addition there is a parity associated with the angular
momentum carried off by the photon, which is of the form ð1ÞJ , reflecting the
symmetry of the angular part of the wavefunction (see Equation (1.14)). We will not
pursue this further here, but defer a more detailed discussion until Chapter 7.

2.9 Nuclear Reactions

In Chapter 1 and earlier sections of the present chapter we discussed various aspects
of reactions. In particle physics, because the projectiles and targets have relatively
simple structures, this is all that is required in classifying reactions. In nuclear
physics, however, because the target has a rich structure it is useful to classify
reactions in more detail. In this section we do this, drawing together our previous
work and also anticipating some reactions that will be encountered in later chapters.
 NUCLEAR REACTIONS 63

Elastic scattering reactions were defined in Chapter 1 as those interactions where

the initial and final particles are identical, i.e. a þ A ! a þ A. We also defined
inelastic scattering as the situation where the final particles are the same chemical
species, but one or more is in an excited state, e.g. a þ A ! a þ A and in Section 2.1
we showed how the kinematics of such reactions could be used to determine the mass
of the excited state. Elastic and inelastic scattering are examples of so-called direct
reactions. These are defined as ones where the incident particle interacts in a time
comparable to the time taken to transit the nucleus. They are more likely when the
incident particle has an energy corresponding to a de Broglie wavelength closer to
the size of a nucleon rather than that of the nucleus. The collisions are largely
peripheral, with only a relatively small fraction of the available energy transferred to
the target. Another direct reaction is 16 Oðp; dÞ15 O, i.e.

p þ 16 O ! d þ 15 O; ð2:78Þ

where we have used the notation Aða; bÞB for the general nuclear reaction
a þ A ! b þ B. This is an example of a pick-up reaction, because one or more
nucleons (in this case a neutron) is stripped off the target nucleus and carried away
by the projectile. The ‘inverse’ of this reaction is 16 Oðd; pÞ17 O. This is an example of
a stripping reaction, because one or more nucleons (in this case again a neutron) is
stripped off the projectile and transferred to the target nucleus.
The theoretical interpretation of direct reactions is based on the assumption that
the projectile experiences the average potential of the target nucleus. For example,
we have seen in the optical model of Section 2.2.2 how this approach can be used to
analyse differential cross sections for elastic scattering and be used to extract
information about nuclear shapes and sizes. It also leads to the prediction of
resonances of width typically of order 1 MeV separated by a few MeV, as observed
in cross-section as functions of centre-of-mass energy for nucleon scattering from
light nuclei. One way of viewing this is as a consequence of the reaction time for a
direct reaction, typically 1022 s , making use of the uncertainty relation between
energy and time,
E
t  h.
A second important class of interactions is where the projectile becomes loosely
bound in the nucleus and shares its energy with all the nuclear constituents. This is
called a compound nucleus reaction. The time for the system to reach statistical
equilibrium depends on the nuclear species, the type of projectile and its energy, but
will always be much longer than the transit time and is typically several orders of
magnitude longer. An important feature of these reactions is that the properties of the
compound nucleus determine its subsequent behaviour and not the mechanism by
which it was formed. The compound nucleus is in an excited state and is inherently
unstable. Eventually, by a statistical fluctuation, one or more nucleons will acquire
sufficient energy to escape and the nucleus either emits particles or de-excites by
radiating gamma rays.
If the compound nucleus is created in a region of excitation where its energy
levels are well separated, the cross-section will exhibit well-defined resonances
64 CH2 NUCLEAR PHENOMENOLOGY

described by the Breit–Wigner formula of Section 1.6.3. These processes are

depicted schematically in the energy-level diagram of Figure 2.16, which corre-
spond to a þ A ! C  ! b þ B, where C is the compound nucleus and
a þ A ! C  ! C þ , where C is the ground state corresponding to the excited
state C . In practice, there could be many final states to which C could decay.

Figure 2.16 Energy-level diagram showing the excitation of a compound nucleus C  and its
subsequent decay
Because the time for a compound nucleus to reach statistical equilibrium is much
longer than the transit time for a direct reaction, the cross-sections for a compound
nucleus process can show variations on much smaller energy scales than those for
direct reactions. The density of levels in the compound nucleus is high, and so a very
small change in the incident energy suffices to alter completely the intermediate
states, and hence the cross section. An example is shown in Figure 2.17, which gives
the total cross-section for neutron scattering from 12 C at neutron laboratory energies
of a few MeV. Peaks corresponding to resonance formation in 13 C are clearly
identified. Their widths vary from a few tens to a few hundreds of keV, consistent
with the characteristic times for compound nucleus formation and decay.

Figure 2.17 Total cross-section for n12 C interactions (adapted from Fo61. Copyright American
Physical Society.)
 NUCLEAR REACTIONS 65

The mean widths of compound nucleus excitations depend on the incident

energy and the target nucleus, decreasing both with energy and rapidly with
nuclear mass. Neutrons, because they are neutral, have a high probability of being
captured by nuclei and their cross-sections are rich in compound nucleus effects,
particularly at very low energies. This is discussed further below.
The division of reactions into direct and compound nucleus is not exhaustive
and situations can occur where particles are ejected from the nucleus before full
statistical equilibrium has been reached. Also, in the collisions of complex heavy
ions, there is an appreciable probability for an additional reaction mechanism
called deep inelastic scattering that is intermediate between direct and compound
nucleus reactions. In this case, the probability for complete fusion of the colliding
ions is small, but there can be substantial transfer of the incident kinetic energy to
internal excitations of the ions. We will not discuss this or other mechanisms further,
but we will encounter the concept of deep inelastic scattering again in Chapter 5 in
the context of exploring the internal structure of nucleons. In practice, the various
mechanisms feed the same final states as direct reactions. This is illustrated
schematically in Figure 2.18 for reactions initiated using protons as the projectile.

Figure 2.18 Direct and compound nucleus reactions in nuclear reactions initiated by protons

The general form of the yield NðEÞ of secondary particles at a fixed angle as a
function of the outgoing energy E, i.e. the number of particles with energy E
between E and E þ dE, is shown schematically in Figure 2.19 for the case of an
incident nucleon. At the upper end of the plot (which corresponds to low-incident
nucleon energies) there are a number of distinct peaks due to elastic, inelastic and
transfer reactions. Then as the excitation energy is reduced, the more closely-
spaced energy levels in the final nucleus are not fully resolved because of the
spread in energy of the incident beam and the uncertainty in the experimental
measurements of energy. At the lowest energies there is a broad continuum mainly
due to the decays of compound nuclei formed by the absorption of the projectile
nucleon by the target nucleus. The differential cross-sections for the two processes
will be very different. Direct reactions lead to a cross-section peaked in the
forward direction, falling rapidly with angle and with oscillations, as we have seen
66 CH2 NUCLEAR PHENOMENOLOGY

Figure 2.19 Typical spectrum of energies of the nucleons emitted at a fixed angle in inelastic
nucleon--nucleus reactions

in the case of elastic scattering in Section 2.2 (Figure 2.3). On the other hand, the
contribution from the compound nucleus at low energies where an isolated
compound nucleus is formed is fairly isotropic and symmetric about 90 .
Many medium- and large-A nuclei can capture low-energy ( ð10100Þ eV)
neutrons very readily. The neutron separation energy for the final nucleus is
 6 MeV and thus capture leads to a compound nucleus with an excitation energy
above the ground state by this separation energy. Such excitation often occurs in a
region of high density of narrow states that show up as a rich resonance structure in
the corresponding neutron total cross-section. An example is shown in Figure 2.20.
The value of the cross-section at the resonance peaks can be many orders of

Figure 2.20 Total cross-section for neutron interactions with 238 U, showing many very narrow
resonances (with intrinsic widths of order 102 eV) corresponding to excited states of 239 U (from
Ga76, courtesy of Brookhaven National Laboratory)
 PROBLEMS 67

magnitude greater than the geometrical cross-section based on the size of the
nucleus. This is because the cross-section is determined dominantly by the area
associated with the wavelength of the projectile, i.e.  2 , which is very large
because is large.
Once formed, the compound nucleus can decay to any final state consistent with
the relevant conservation laws. If this includes neutron emission, it will be the
preferred decay. However, for production by very slow (thermal) neutrons with
energies of the order of 0.02 eV, the available decay kinetic energy will reflect the
initial energy of the projectile, which is very small. Therefore, in these cases,
photon emission is often preferred. We shall see in Chapter 8 that the fact that
radiative decay is the dominant mode of decay of compound nuclei formed by
thermal neutrons is important in the use of nuclear fission to produce power in
nuclear reactors.

Problems
2.1 Electrons with momentum 330 MeV/c are elastically scattered through an angle of
10 by a nucleus of 56 Fe. If the charge distribution on the nucleus is assumed to be
that of a uniform hard sphere, and assuming the Born approximation is valid, by
what factor would you expect the Mott cross-section to be reduced?

2.2 Show explicitly that Equation (2.28) follows from Equation (2.26).

2.3 A beam of electrons with energies 250 MeV is scattered through an angle of 10 by a
heavy nucleus. It is found that the differential cross-section is 65 per cent of that
expected from scattering from a point nucleus. Estimate the root mean square radius
of the nucleus.

2.4 Find the form factor for a charge distribution ðrÞ ¼ 0 expðr=aÞ=r, where 0 and a
are constants.

2.5 A sample of 1 g of a radioactive isotope of atomic weight 208 decays via -emission
and 75 counts are recorded in a 24 h period. If the detector efficiency is 10 per cent,
estimate the mean life of the isotope.

2.6 A 1 g sample taken from an organic artefact is found to have a count rate of 2.1
counts per min, which are assumed to originate from the decay of 14 C with a mean
lifetime of 8270 years. If the abundance of 14 C in living matter is currently
1:2  1012 , what can you deduce about the approximate age of the artefact?

2.7 Nuclei of 212 208

86 Rn decay by -emission to 84 Po with a mean life of 23.9 min. The
208 204
84 Po nuclei in turn decay, also by -emission, to the stable isotope 82 Pb with a
212
mean life of 2.9 years. If initially the source is pure 86 Rn, how long will it take for
the rate of -emission in the final decay to reach a maximum?
68 CH2 NUCLEAR PHENOMENOLOGY

2.8 Natural lanthanum has an atomic weight of 138.91 and contains 0.09 per cent of the

isotope 138 138 138
57 La. This has two decay modes: 57 La ! 58 Ce þ e þ  e ð -decayÞ and
138  138 
57 La þ e ! 56 Ba þ e (electron capture), followed by the electromagnetic

decay of the excited state 138 138
56 Ba ! 56 Ba þ  (radiative decay). There are
2
7:8  10 -particles emitted per s per kg of natural lanthanum and there are
50 photons emitted per 100 -particles. Estimate the mean lifetime of 138
57 La.

2.9 Use the SEMF to estimate the energy released in the spontaneous fission reaction
235
92 U ! 87 145
35 Br þ 57 La þ 3n:

2.10 The most stable nucleus with A ¼ 111 is 111 48 Cd (see Figure 2.12). By what
percentage would the fine structure constant  have to change if the most stable
nucleus with A ¼ 111 were to be 11147 Ag? Assume that altering  does not change
particle masses.

2.11 The transuranic isotope 269

108 Hs decays 100 per cent via -emission with a lifetime of
27 s, i.e. 269
108 Hs ! 265
106 Sg þ , where the kinetic energy of the -particle is
E ¼ 9:23 MeV. Calculate the mass of the 269 108 Hs nucleus in atomic mass units.

2.12 The isotope 238 234

94 Pu decays via -emission to the essentially stable isotope 92 U with a
lifetime of 126.7 years and a release of 5.49 MeV of kinetic energy. This energy is
converted to electrical power in a space probe designed to reach planet X in a
journey planned to last 4 years. If the efficiency of power conversion is 5 per cent
and on reaching planet X the probe requires at least 200 W of power to perform its
landing tasks, how much 23894 Pu would be needed at launch?

2.13 On planet X it is found that the isotopes 205 Pbð

¼ 1:53  107 yÞ and 204 Pb (stable)
are present with abundances n205 and n204 , with n205 =n204 ¼ 2  107 . If at the time
of the formation of planet X both isotopes were present in equal amounts, how old is
the planet?

2.14 The reaction 45 46

21 Scðd; pÞ21 Sc has a Q-value of 6.54 MeV and a resonance when the
incident deuteron laboratory kinetic energy is 2.76 MeV. Would you expect the same
resonance to be excited in the reaction 43 46
20 Cað; nÞ22 Ti and if so at what value of the
laboratory kinetic energy of the alpha particle? You may use the fact that the -

decay 46 46
21 Sc ! 22 Ti þ e þ  e has a Q-value of 2.37 MeV and the mass difference
between the neutron and a hydrogen atom is 0.78 MeV=c2 .

2.15 A radioisotope with decay constant is produced at a constant rate P. Show that the
number of atoms at time t is NðtÞ ¼ P½1  expð tÞ= .

2.16 Radioactive 36 Cl (half-life 3  105 years) is produced by irradiating 1 g of natural

nickel chloride (NiCl2 , molecular weight 129.6) in a neutron beam of flux
F ¼ 1014 cm2 s1 . If the neutron absorption cross-section 35 Clðn; Þ36 Cl is
 ¼ 43:6 b and 75.8 per cent of natural chlorine is 35 Cl, use the result of Problem
2.15 to estimate the time it would take to produce a 3  105 Bq source of 36 Cl.
 PROBLEMS 69

2.17 Consider the total cross-section data for the n238 U interaction shown in Figure 2.20.
There is a resonance R at the centre-of-mass neutron kinetic energy En ¼ 10 eV with
width  ¼ 102 eV and the total cross-section there is max ¼ 9  103 b. Use this
information to find the partial widths n; for the decays R ! n þ 238 U and
R !  þ 238 U, if these are the only two significant decay modes. The spin of the
ground state of 238 U is zero.
3
Particle Phenomenology

In this chapter we shall look at some of the basic phenomena of particle physics –
the properties of leptons and quarks, and the bound states of the latter, the hadrons.
In later chapters we will discuss theories and models that attempt to explain these
and other particle data.

3.1 Leptons
We have seen that the spin-12 leptons are one of the three classes of elementary
particles in the standard model and we shall start with a discussion of their basic
properties. Then we shall look in more detail at the neutral leptons, the neutrinos
and, amongst other things, examine an interesting property they can exhibit, based
on simple quantum mechanics, if they have non-zero masses.

3.1.1 Leptons multiplets and lepton numbers

There are six known leptons and they occur in pairs, called generations, which we
write, for reasons that will become clear presently, as:
     

e


; ; : ð3:1Þ
e



Each generation comprises a charged lepton with electric charge
e, and a
neutral neutrino. The three charged leptons ðe
; 
; 
Þ are the familiar
electron, together with two heavier particles, the mu lepton (usually called the
muon, or just mu) and the tau lepton (usually called the tauon, or just tau). The
associated neutrinos are called the electron neutrino, mu neutrino, and tau

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
72 CH3 PARTICLE PHENOMENOLOGY

neutrino, respectively.1 In addition to the leptons, there are six corresponding

antileptons:




eþ þ þ
; ; : ð3:2Þ

e



Ignoring gravity, the charged leptons interact only via electromagnetic and weak
forces, whereas for the neutrinos, only weak interactions have been observed.2
Because of this, neutrinos, which are all believed to have extremely small masses,
can be detected only with considerable difficulty.
The masses and lifetimes of the leptons are listed in Table 3.1. The electron and
the neutrinos are stable, for reasons that will become clear shortly. The muons
decay by the weak interaction processes

þ ! eþ þ
e þ
 and 
! e
þ
e þ
 ; ð3:3aÞ

Table 3.1 Properties of leptons: all have spin 12 and masses are given units of MeV/c2; the
antiparticles (not shown) have the same masses as their associated particles, but the electric
charges (Q) and lepton numbers (L‘ ; ‘ ¼ e ;  ; ) are reversed in sign

Name and symbol Mass Q Le Ll Ls Lifetime (s) Major decays

Electron e 0.511
1 1 0 0 Stable None
Electron neutrino
e <2.2 eV/c2 0 1 0 0 Stable None
Muon (mu) 
105.7
1 0 1 0 2:197  10
6 e

e
 (100%)
Muon neutrino
 <0.19 0 0 1 0 Stable None
Tauon (tau) 
1777.0
1 0 0 1 2:906  10
13 


 (17.4%)
e

e
 (17.8%)

 +hadrons ( 64%)
Tauon neutrino
 <18.2 0 0 0 1 Stable None

with lifetimes ð2:19703  0:00004Þ  10
6 s. The tau also decays by the weak
interaction, but with a much shorter lifetime ð2:906  0:011Þ  10
13 s. (This
illustrates what we have already seen in nuclear physics, that lifetimes depend
sensitively on the energy released in the decay, i.e. the Q-value.) Because it is
heavier than the muon, the tau has sufficient energy to decay to many different
final states, which can include both hadrons and leptons. However, about 35 per cent

1
Leon Lederman, Melvin Schwartz and Jack Steinberger shared the 1988 Nobel Prize in Physics for their use
of neutrino beams and the discovery of the muon neutrino. Martin Perl shared the 1995 Nobel Prize in
Physics for his pioneering work in lepton physics and in particular for the discovery of the tau lepton.
2
Although neutrinos have zero electric charge they could, in principle, have a charge distribution that would
give rise to a magnetic moment (like neutrons) and hence electromagnetic interactions. This would of course
be forbidden in the standard model because the neutrinos are defined to be point-like.
 LEPTONS 73

of decays again lead to purely leptonic final states, via reactions which are very
similar to muon decay, for example:

 þ ! þ þ
 þ
 and 
! e
þ
e þ
 : ð3:3bÞ

Associated with each generation of leptons is a quantum number called a lepton

number. The first of these lepton numbers is the electron number, defined for any
state by

Le
Nðe
Þ
Nðeþ Þ þ Nð
e Þ
Nð


e Þ; ð3:4Þ

where Nðe
Þ is the number of electrons present, Nðeþ Þ is the number of positrons
present and so on. For single-particle states, Le ¼ 1 for e
and
e , Le ¼
1 for eþ
and
e , and Le ¼ 0 for all other particles. The muon and tauon numbers are defined
in a similar way and their values for all single particle states are summarized in
Table 3.1. They are zero for all particles other than leptons. For multiparticle
states, the lepton numbers of the individual particles are simply added. For
example, the final state in neutron -decay (i.e. n ! p þ e
þ
e ) has
Le ¼ Le ðpÞ þ Le ðe
Þ þ Le ð


e Þ ¼ ð0Þ þ ð1Þ þ ð
1Þ ¼ 0; ð3:5Þ

like the initial state, which has Le ðnÞ ¼ 0.

In the standard model, the value of each lepton number is postulated to be
conserved in any reaction. The decays (3.3) illustrate this principle of lepton
number conservation. Until recently this was considered an absolute conservation
law, but in Section 3.1.4 we will discuss growing evidence that neutrinos are not
strictly massless, which would imply that conservation of individual lepton
numbers is not an exact law. However, for the present we will assume lepton
numbers are conserved, as in the standard model. In electromagnetic interactions,
this reduces to the conservation of Nðe
Þ
Nðeþ Þ, Nð
Þ
Nðþ Þ and
Nð
Þ
Nð þ Þ, since neutrinos are not involved. This implies that the charged
leptons can only be created or annihilated in particle–antiparticle pairs. For
example, in the electromagnetic reaction

eþ þ e
! þ þ 
ð3:6Þ

an electron pair is annihilated and a muon pair is created by the mechanism of

Figure 3.1.

Figure 3.1 Single-photon exchange in the reaction eþ e
! þ 

74 CH3 PARTICLE PHENOMENOLOGY

Figure 3.2 Dominant Feynman diagram for the decay 
! e

e


In weak interactions more general possibilities are allowed, which still conserve
lepton numbers. For example, in the tau-decay process 
! e
þ
e þ
 , a tau
converts to a tau neutrino and an electron is created together with an antineutrino,
rather than a positron. The dominant Feynman graph corresponding to this process
is shown in Figure 3.2.
Lepton number conservation, like electric charge conservation, plays an impor-
tant role in understanding reactions involving leptons. Observed reactions conserve
lepton numbers, while reactions that do not conserve lepton numbers are ‘forbidden’
and are not observed. For example, the neutrino scattering reaction

 þ n ! 
þ p ð3:7Þ

is observed experimentally, while the apparently similar reaction

 þ n ! e
þ p; ð3:8Þ

which violates both Le and L conservation, is not. Another example which violates
both Le and L conservation is 
! e
þ . If this reaction were allowed, the
dominant decay of the muon would be electromagnetic and the muon lifetime would
be much shorter than its observed value. This is very strong evidence that even if
lepton numbers are not absolutely conserved, they are conserved to a high degree of
accuracy.
Finally, conservation laws explain the stability of the electron and the neutrinos.
The electron is stable because electric charge is conserved in all interactions and
the electron is the lightest charged particle. Hence decays to lighter particles that
satisfy all other conservation laws, like e
!
e þ , are necessarily forbidden by
electric charge conservation. In the same way, lepton number conservation implies
that the lightest particles with non-zero values of the three lepton numbers – the
three neutrinos – are stable, whether they have zero masses or not. Of course, if
lepton numbers are not conserved, then the latter argument is invalid.

3.1.2 Neutrinos

As we mentioned in Chapter 1, the existence of the electron neutrino
e was first

postulated by Pauli in 1930. He did this in order to understand the observed
 LEPTONS 75

nuclear -decays

ðZ ; NÞ ! ðZ þ 1; N
1Þ þ e
þ
e ð3:9Þ

and

ðZ; NÞ ! ðZ
1; N þ 1Þ þ eþ þ
e ð3:10Þ

that were discussed in Chapter 2. The neutrinos and antineutrinos emitted in these
decays are not observed experimentally, but are inferred from energy and angular
momentum conservation. In the case of energy, if the antineutrino were not present
in the first of the reactions, the energy Ee of the emitted electron would be a unique
value equal to the difference in rest energies of the two nuclei, i.e.

Ee ¼ Mc2 ¼ ½MðZ; NÞ
MðZ þ 1; N
1Þc2 ; ð3:11Þ

where for simplicity we have neglected the extremely small kinetic energy of the
recoiling nucleus. However, if the antineutrino is present, the electron energy
would not be unique, but would lie in the range

me c2 Ee ðM
m
e Þc2 ; ð3:12Þ

depending on how much of the kinetic energy released in the decay is carried away
by the neutrino. Experimentally, the observed energies span the whole of the above
range and in principle a measurement of the energy of the electron near its
maximum value of Ee ¼ ðM
m
e Þc2 determines the neutrino mass. The most
accurate results come from tritium decay and are compatible with zero mass for
electron antineutrinos. When experimental errors are taken into account, the
experimentally allowed range is

0 m
e < 2:2 eV=c2  4:3  10
6 me : ð3:13Þ

We will discuss this experiment in more detail in Chapter 7, after we have

considered the theory of -decay.
The masses of both
 and
 can similarly be directly inferred from the e
and


 energy spectra in the leptonic decays of muons and tauons, using energy
conservation. The results from these and other decays show that the neutrino
masses are very small compared with the masses of the associated charged leptons.
The present limits are given in Table 3.1.
Small neutrino masses, compatible with the above limits, can be ignored in most
circumstances, and there are theoretical attractions in assuming neutrino masses are
precisely zero, as is done in the standard model. However, we will show in the
following section that there is now strong evidence for physical phenomena that
76 CH3 PARTICLE PHENOMENOLOGY

could not occur if the neutrinos had exactly zero mass. The consequences of
neutrinos having small masses have therefore to be taken seriously.
Because neutrinos only have weak interactions, they can only be detected with
extreme difficulty. For example, electron neutrinos and antineutrinos of sufficient
energy can in principle be detected by observing the inverse -decay processes

e þ n ! e
þ p ð3:14Þ

and

e þ p ! eþ þ n : ð3:15Þ

However, the probability for these and other processes to occur is extremely small.
In particular, the neutrinos and antineutrinos emitted in -decays, with energies of
the order of 1 MeV, have mean free paths in matter of the order of 106 km.3
Nevertheless, if the neutrino flux is intense enough and the detector is large
enough, the reactions can be observed. In particular, uranium fission fragments are
neutron rich, and decay by electron emission to give an antineutrino flux that can
be of the order of 1017 m
2 s
1 or more in the vicinity of a nuclear reactor, which
derives its energy from the decay of nuclei. These antineutrinos will occasionally
interact with protons in a large detector, enabling examples of the inverse -decay
reaction to be observed. As mentioned in Chapter 1 (Footnote 12), electron
neutrinos were first detected in this way by Reines and Cowan in 1956, and
their interactions have been studied in considerable detail since.
The mu neutrino,
 , has been detected using the reaction
 þ n ! 
þ p and
other reactions. In this case, well-defined high-energy
 beams can be created in the
laboratory by exploiting the decay properties of pions, which are particles we have
mentioned briefly in Chapter 1 and which we will meet in more detail presently. The
probability of neutrinos interacting with matter increases rapidly with energy (this
will be demonstrated in Section 6.5.2) and, for large detectors, events initiated by
such beams are so copious that they have become an indispensable tool in studying
both the fundamental properties of weak interactions and the internal structure of the
proton. Finally, in 2000, a few examples of tau neutrinos were reported, so that
almost 70 years after Pauli first suggested the existence of a neutrino, all three types
had been directly detected.

3.1.3 Neutrino mixing and oscillations

Neutrinos are assumed to have zero mass in the standard model. However, as
mentioned above, data from the -decay of tritium are compatible with a non-zero

3
The mean free path is the distance a particle would have to travel in a medium for there to be a significant
probability of an interaction. A formal definition is given in Chapter 4.
 LEPTONS 77

mass. A phenomenon that can occur if neutrinos have non-zero masses is neutrino
mixing. This arises if we assume that the observed neutrino states
e ;
 and

which take part in weak interactions, i.e. the states that couple to electrons, muons
and tauons, respectively, are not eigenstates of mass, but instead are linear
combinations of three other states
1 ;
2 and
3 which do have definite masses
m1 ; m2 and m3 , i.e. are eigenstates of mass. For algebraic simplicity we will
consider the case of mixing between just two states, one of which we will assume
is
 and the other we will denote by
x. Then, in order to preserve the orthonormality
of the states, we can write

 ¼
1 cos  þ
2 sin  ð3:16Þ

and

x ¼

1 sin  þ
2 cos : ð3:17Þ

Here  is a mixing angle which must be determined from experiment. If  6¼ 0

then some interesting predictions follow.
Measurement of the mixing angle may be made in principle by studying the
phenomenon of neutrino oscillation. When, for example, a muon neutrino is
produced with momentum p at time t ¼ 0, the
1 and
2 components will have
slightly different energies E1 and E2 due to their slightly different masses. In
quantum mechanics, their associated waves will therefore have slightly different
frequencies, giving rise to a phenomenon somewhat akin to the ‘beats’ heard when
two sound waves of slightly different frequency are superimposed. As a result of
this, one finds that the original beam of muon neutrinos develops a
x component
whose intensity oscillates as it travels through space, while the intensity of the
muon neutrino beam itself is correspondingly reduced, i.e. muon neutrinos will
‘disappear’.
This effect follows from simple quantum mechanics. To illustrate this we will
consider a muon neutrino produced with momentum p at time t ¼ 0. The initial
state is therefore
 

 ; p ¼ j
1 ; pi cos  þ j
2 ; pi sin ; ð3:18Þ

where we use the notation jP; pi to denote a state of a particle P having

momentum p. After time t this will become

a1 ðtÞ j
1 ; pi cos  þ a2 ðtÞj
2 ; pi sin ; ð3:19Þ

where

ai ðtÞ ¼ e
iEi t=
h ði ¼ 1; 2Þ ð3:20Þ

78 CH3 PARTICLE PHENOMENOLOGY

are the usual oscillating energy factors associated with any quantum mechanical
stationary state.4 For t 6¼ 0, the linear combination (3.19) does not correspond to a
pure muon neutrino state, but can be written as a linear combination
 
AðtÞ
 ; p þ BðtÞj
x ; pi; ð3:21Þ

of
 and
x states, where the latter is

j
x ; pi ¼
j
1 ; pi sin  þ j
2 ; pi cos : ð3:22Þ

The functions AðtÞ and BðtÞ are found by solving Equations (3.18) and (3.22) for
j
1 ; pi and j
2 ; pi, then substituting the results into (3.19) and comparing at with
(3.21). This gives,

AðtÞ ¼ a1 ðtÞ cos2  þ a2 ðtÞ sin2  ð3:23Þ

and

BðtÞ ¼ sin  cos  ½a2 ðtÞ
a1 ðtÞ: ð3:24Þ

The probability of finding a
x state is therefore

Pð
 !
x Þ ¼ jBðtÞj2 ¼ sin2 ð2Þ sin2 ½ðE2
E1 Þ t=2
h ð3:25Þ

and thus oscillates with time, while the probability of finding a muon neutrino is
reduced by a corresponding oscillating factor. Similar effects are predicted if
instead we start from electron or tau neutrinos. In each case the oscillations vanish
if the mixing angle is zero, or if the neutrinos have equal masses, and hence equal
energies, as can be seen explicitly from Equation (3.25). In particular, such
oscillations are not possible if the neutrinos both have zero masses.
Returning to Equation (3.25), since neutrino masses are very small,
E1;2  mi c2 ði ¼ 1; 2Þ and we can write

 1=2  2 4 1=2 m2 c4
m21 c4

E2
E1 ¼ m22 c4 þ p2 c2
m1 c þ p2 c2  2 : ð3:26Þ
2pc

Also, E  pc and t  jxj=c

L=c, where L is the distance from the point of
production. Thus Equation (3.25) may be written
 
ðm2 c4 ÞL
Pð
 !
x Þ  sin2 ð2Þ sin2 ; ð3:27aÞ
4
hcE

4
See, for example, Chapter 1 of Ma92.
 LEPTONS 79

with

Pð
 !
 Þ ¼ 1
Pð
 !
x Þ; ð3:27bÞ

where ðm2 c4 Þ
m22 c4
m21 c4 . These formulae assume that the neutrinos are
propagating in a vacuum, whereas in real experiments they will be passing through
matter and the situation is more complicated than these simple results suggest.
This formalism can be extended to the general case of mixing between all three
neutrino species, but at the expense of additional free parameters.5
Attempts to establish neutrino oscillations rest on using the inverse -decay
reactions (3.14) and (3.15) to produce electrons and the analogous reactions for
muon neutrinos to produce muons, which are then detected. In addition, the time t
is determined by the distance of the neutrino detector from the source of the
neutrinos, since their energies are always much greater than their possible masses,
and they travel at approximately the speed of light. Hence, for example, if we start
with a source of muon neutrinos, the flux of muons observed in a detector should
vary with its distance from the source of the neutrinos, if appreciable oscillations
occur. In practice, oscillations at the few per cent level are very difficult to detect
for experimental reasons that we will not discuss here.

3.1.4 Neutrino masses

There are a number of different types of experiment that can explore neutrino
oscillations and hence neutrino masses. The first of these to produce definitive
evidence for oscillations was that of a Japanese group in 1998 using the giant
Super Kamiokande detector to study atmospheric neutrinos produced by the action
of cosmic rays.6 (Neutrinos of each generation are often referred to as having a
different flavour and so the observations are evidence for flavour oscillation.)
The Super Kamiokande detector is shown in Figure 3.3. (Detectors will be
discussed in detail in Chapter 4, so the description here will be brief.) It consists of
a stainless steel cylindrical tank of roughly 40 m diameter and 40 m height,
containing about 50 000 metric tons of very pure water. The detector is situated
deep underground in a mountain in Japan, at a depth equivalent to 2700 m of water.
This is to use the rocks above to shield the detector from cosmic ray muons. The
volume is separated into a large inner region, the walls of which are lined with
11 200 light-sensitive devices called photomultipliers (the physics of these will be
discussed in Chapter 4). These register the presence of electrons or muons

5
See, for example, the Review of Particle Properties published biannually by the Particle Data Group (2004
edition: Ei04). The PDG Review is also available at http://pdg.lbl.gov. This publication contains a wealth of
useful data about elementary particles and their interactions and we will refer to it in future simply as
PDG04.
6
Cosmic neutrinos were first detected (independently) by Raymond Davis Jr. and Masatoshi Koshiba, for
which they were jointly awarded the 2002 Nobel Prize in Physics.
80 CH3 PARTICLE PHENOMENOLOGY

Figure 3.3 A schematic diagram of the Super Kamiokande detector (adapted from an original
University of Hawaii, Manoa, illustration -- with permission)

indirectly by detecting the light (the so-called Čerenkov radiation – again, see
Chapter 4) emitted by relativistic charged particles (the electrons or muons) that
are created in, or pass through, the water. The outer region of water acts as a shield
against low-energy particles entering the detector from outside. An additional 1200
photomultipliers are located there to detect muons that enter or exit the detector.
When cosmic ray protons collide with atoms in the upper atmosphere they create
many pions, which in turn create neutrinos mainly by the decay sequences


! 
þ
 ; þ ! þ þ
 ð3:28Þ

and


! e
þ
e þ
 ; þ ! eþ þ
e þ
 : ð3:29Þ

From this, one would naively expect to detect two muon neutrinos for every
electron neutrino. However, the ratio was observed to be about 1.3 to 1 on average,
suggesting that the muon neutrinos produced might be oscillating into other species.
Clear confirmation for this was found by exploiting the fact that the detector
could measure the direction of the detected neutrinos to study the azimuthal
dependence of the effect. Since the flux of cosmic rays that lead to neutrinos with
energies above about 1 GeV is isotropic, the production rate for neutrinos should
 LEPTONS 81

be the same all around the Earth. In particular, one can compare the measured flux
from neutrinos produced in the atmosphere directly above the detector, which have
a short flight path before detection, with those incident from directly below, which
have travelled a long way through the Earth before detection, and so have had
plenty of time to oscillate (perhaps several cycles). Experimentally, it was found
that the yield of electron neutrinos from above and below were the same within
errors and consistent with the expectation for no oscillations. However, while
the yield of muon neutrinos from above accorded with the expectation for no
significant oscillations, the flux of muon neutrinos from below was a factor of
about two lower. This is clear evidence for muon neutrino oscillations.
In a later development of the experiment, the flux of muon neutrinos was
measured as a function of L=E by estimating L from the reconstructed neutrino
direction. Values of L range from 15 km to 13000 km. The results are shown in
Figure 3.4 in the form of the ratio of observed number of events to the theoretical
expectation if there were no oscillations. The data show clear evidence for a
deviation of this ratio from unity, particularly at large values of L=E.

Figure 3.4 Data from the Super Kamiokande detector showing evidence for neutrino
oscillations in atmospheric neutrinos (adapted from As04, copyright American Physical Society)

Other experiments also set limits on Pð
 !
e Þ and taking these into account
the most plausible hypothesis is that muon neutrinos are changing into tau neutrinos,
which for the neutrino energies concerned could not be detected by Super
Kamiokande. The data are consistent with this hypothesis and yield the values

1:9  10
3 ðm2 c4 Þ 3:0  10
3 ðeVÞ2 ; sin2 ð2Þ > 0:9 ð3:30Þ
82 CH3 PARTICLE PHENOMENOLOGY

at 90 per cent confidence level. This conclusion is supported by preliminary results

from laboratory-based experiments that start with a beam of
 and measure the
flux at a large distance (250 km) from the origin. Analysis of the data yields similar
parameters to those above.
A second piece of evidence for neutrino oscillations comes from our knowledge
of the Sun. We shall see in Chapter 8 that the energy of the Sun is due to various
nuclear reactions and these produce a huge flux of electron neutrinos that can be
detected at the surface of the Earth. Since the astrophysics of the Sun and nuclear
production processes are well understood, this flux can be calculated with some
confidence by what is known as the standard solar model.7 However, the measured
count rate is about a factor of two lower than the theoretical expectation. This is
the so-called solar neutrino problem. It was first investigated by Davis and co-
workers in the late 1960s who studied the reaction

e þ 37 Cl ! 37 Ar þ e
; ð3:31Þ

to detect the neutrinos. (This required sensitive radiochemical analysis to confirm

the production of 37 Ar.) This reaction has a threshold of 0:81 MeV and is therefore
only sensitive to relatively high-energy neutrinos from the Sun. Such neutrinos
come predominantly from the weak interaction decay

8
B ! 8 Be þ eþ þ
e ; ð3:32Þ

where the neutrinos have an average energy 7 MeV. More recent experiments
have studied the same process using the reactions

ðaÞ
x þ d ! e
þ p þ p; ðbÞ
x þ d !
x þ p þ n; ðcÞ
x þ e
!
x þ e
;

ð3:33Þ

to detect the neutrinos, where d is a deuteron. The first of these reactions clearly
can be initiated with electron neutrinos only, whereas the other two can be initiated
with neutrinos of any flavour. The measured flux of
e from reaction (a) agrees
well with the standard solar model, but the ratio of the flux for
e to that for
x,
where x could be a combination of  and , obtained by using data from all three
reactions, is less than unity. For example, the Sudbury Neutrino Observatory
(SNO) experiment finds a ratio of about 0.3. Thus there is a flux of neutrinos of a
type that did not come from the original decay process. The observations are
further clear evidence for flavour oscillation.
Although the neutrinos from (3.32) have been extensively studied, this decay
contributes only about 10
4 of the total solar neutrino flux. It is therefore important

7
See, for example, Chapter 4 of Ph94.
 LEPTONS 83

to detect neutrinos from other reactions and in particular from the reaction

p þ p ! d þ eþ þ
e ; ð3:34Þ

which is the primary reaction that produces the energy of the Sun and contributes
approximately 90 per cent of the solar neutrino flux. (It will be discussed in more
detail in Chapter 8.) The neutrinos in this reaction have average energies of
0:26 MeV and so cannot be detected by reaction (3.31). Instead, the reaction

e þ 71 Ga ! 71 Ge þ e
ð3:35Þ

has been used, which has a threshold of 233 keV. (The experiments can also detect
neutrinos from the solar reaction e
þ 7 Be ! 7 Li þ
e .) Just as for the original
experiments of Davis et al., there are formidable problems in identifying the
radioactive products from this reaction, which produces only about 1 atom of 71 Ge
per day in a target of 30 tons of gallium. Nevertheless, results from these
experiments confirm the deficit of electron neutrinos and find between 60 and
70 per cent of the flux expected from the standard solar model without flavour
changing.
These solar neutrino results require that interactions with matter play a
significant role in flavour changing and imply, for example, that a substantial
fraction of a beam of
e would change to antineutrinos of other flavours after
travelling a distance of the order of 100 km from its source. This prediction has
been tested by the KamLAND group in Japan. They have studied the
e flux from
more than 60 reactors in Japan and South Korea after the neutrinos have travelled
distances of between 150 and 200 km. They found that the
e flux was only about
60 per cent of that expected from the known characteristics of the reactors. A
simultaneous analysis of the data from this experiment and the solar neutrino data
yields the result:

7:6  10
5 ðm2 c4 Þ 8:8  10
5 ðeVÞ2 ; 0:32 tan2 ðÞ 0:48: ð3:36Þ

The existence of neutrino oscillations (flavour changing), and by implication non-

zero neutrino masses, is now generally accepted on the basis of the above set of
experiments.
What are the consequences of these results for the standard model? The
observation of oscillations does not lead to a measurement of the neutrino masses,
only (squared) mass differences, but combined with the tritium -decay experiment,
it would be natural to assume that neutrinos all had very small masses, with the mass
differences being of the same order-of-magnitude as the masses themselves. The
standard model can be modified to accommodate small masses, although methods
for doing this are not without their own problems.8 Unfortunately, the various

8
One possibility will be mentioned briefly in Chapter 9 as part of a discussion of the general question of how
masses arise in the standard model.
84 CH3 PARTICLE PHENOMENOLOGY

experiments – although producing compatible values for the mixing angle

40 –
yield wildly different values for the mass difference, as can be seen from Equations
(3.30) and (3.36). However, the analyses have been made in the framework of a two-
component mixing model, whereas there are of course three neutrinos. Thus it could
be, for example, that two of the neutrino states are separated by a small mass
difference given by Equation (3.36) and the third is separated from them by a
relatively large mass difference given by Equation (3.30). Progress will have to await
experiments currently being planned to detect oscillations directly using prepared
neutrino beams and which will make measurements at great distances from their
origin. These experiments are expected to produce data in the next few years and
should yield definitive values of the neutrino mass differences and the various
mixing angles involved.9
The consequences for lepton number conservation are unclear. In the simple
mixing model above, the total lepton number could still be conserved, but
individual lepton numbers would not. However, there are other theoretical
descriptions of neutrino oscillations and this is an open question. A definitive
answer would be to detect neutrinoless double -decay, such as

76
Ge ! 76 Se þ 2e ; ð3:37Þ

where the final state contains two electrons, but no antineutrinos. This could occur
if the neutrino emitted by the parent nucleus was internally absorbed by the
daughter nucleus (i.e. it never appears as a real particle) which is possible only if
e  e . A very recent experiment claims to have detected this decay, but the result
is not universally accepted and at present ‘the jury is out’. Experiments planned for
the next few years should settle important questions about lepton number
conservation and the nature of neutrinos.

3.1.5 Universal lepton interactions – the number of neutrinos

The three neutrinos have similar properties, but the three charged leptons are
strikingly different. For example: the mass of the muon is roughly 200 times
greater than that of the electron and consequently its magnetic moment is 200
times smaller; high-energy electrons are stopped by modest thicknesses of a
centimetre or so of lead, while muons are the most penetrating form of radiation
known, apart from neutrinos; and the tauon lifetime is many orders of magnitude
smaller than the muon lifetime, while the electron is stable. It is therefore a
remarkable fact that all experimental data are consistent with the assumption that
the interactions of the electron and its associated neutrino are identical to those of
the muon and its associated neutrino and of the tauon and its neutrino, provided the

9
For a review of these experiments see, for example, http://www.hep.anl.gov/ndk/hypertext/nuindustry.html.
 LEPTONS 85

mass differences are taken into account. This property, called lepton universality,
can be verified with great precision, because we have a precise theory of
electromagnetic and weak interactions (to be discussed in Chapter 6), which
enables predictions to be made of the mass dependence of all observables.
For example, when we discuss experimental methods in Chapter 4, we will show
that the radiation length, which is a measure of how far a charged particle travels
through matter before losing a certain fraction of its energy by radiation, is
proportional to the squared mass of the radiating particle. Hence it is about 4  104
times greater for muons than for electrons, explaining their much greater
penetrating power in matter. As another example, we have seen that the rates
for weak -decays are extremely sensitive to the kinetic energy released in the
decay (recall the enormous variation in the lifetimes of nuclei decaying via -
decay). From dimensional arguments and the fact that they are weak interactions,
the rates for muon and tau leptonic decays are predicted to be proportional to the
fifth power of the relevant Q-values multiplied by G2F , the square of the Fermi
coupling.10 Thus, from universality, the ratio of the decay rates  is given
approximately by

5
ð
! e
þ
e þ
 Þ Q
 ¼ 1:37  106 : ð3:38Þ
ð
! e
þ
e þ
 Þ Q

This is in excellent agreement with the experimental value of 1:35  106 (and is
even closer in a full calculation) and accounts very well for the huge difference
between the tau and muon lifetimes. The above are just some of the most striking
manifestations of the universality of lepton interactions.
A question that arises naturally is whether there are more generations of leptons,
with identical interactions, waiting to be discovered. This question has been
answered, under reasonable assumptions, by an experimental study of the decays
of the Z 0 boson. This particle, one of the two gauge bosons associated with the
weak interaction, has a mass of 91 GeV/c2. It decays, among other final states, to
neutrino pairs

Z 0 !
‘ þ
‘ ð‘ ¼ e; ; Þ: ð3:39Þ

If we assume universal lepton interactions and neutrino masses which are small
compared with the mass of the Z 0 ,11 the decay rates to a given neutrino pair will all
be equal and thus

neutrinos

e þ 
 þ 
 þ    ¼ N

; ð3:40Þ

10
The increase of the decay rate as the fifth power of Q is known as Sargent’s Rule.
11
More precisely, we assume m
MZ =2, so that the decays Z !

are not forbidden by energy
conservation.
86 CH3 PARTICLE PHENOMENOLOGY

where N
is the number of neutrino species and 
is the decay rate to any given
pair of neutrinos. The measured total decay rate may then be written
total ¼ hadrons þ leptons þ neutrinos ; ð3:41Þ
where the first two terms on the right are the measured decay rates to hadrons and
charged leptons, respectively. Although the rate to neutrinos 
is not directly
measured, it can be calculated in the standard model and combining this with
experimental data for the other decay modes, a value of N
may be found. The best
value using all available data is N
¼ 3:00  0:08, which is consistent with the
expectation for three neutrino species, but not four. The conclusion is that only
three generations (flavours) of leptons can exist, if we assume universal lepton
interactions and exclude very large neutrino masses.
Why there are just three generations of leptons remains a mystery, particularly
as the extra two generations seem to tell us nothing fundamental that cannot be
deduced from the interaction of the first generation.

3.2 Quarks
We turn now to the strongly interacting particles – the quarks and their bound states,
the hadrons. These also interact by the weak and electromagnetic interactions,
although such effects can often be neglected compared with the strong interactions.
To this extent we are entering the realm of ‘strong interaction physics’.

3.2.1 Evidence for quarks

Several hundred hadrons (not including nuclei) have been observed since pions
were first produced in the laboratory in the early 1950s and all have zero or integer
electric charges: 0; 1; or  2 in units of e. They are all bound states of the
fundamental spin-12 quarks, whose electric charges are either þ 23 or
13, and/or
antiquarks, with charges
23 or þ 13. The quarks themselves have never been
directly observed as single, free particles and, as remarked earlier, this fact initially
made it difficult for quarks to be accepted as anything other than convenient
mathematical quantities for performing calculations. Only later, when the funda-
mental reason for this was realized (it will be discussed in Chapter 6), were quarks
universally accepted as physical entities. Nevertheless, there is compelling
experimental evidence for their existence. The evidence comes from three main
areas: hadron spectroscopy, lepton scattering and jet production.

Hadron spectroscopy

This is the study of the static properties of hadrons: their masses, lifetimes and
decay modes, and especially the values of their quantum numbers, including spin,
 QUARKS 87

electric charge and several more that we define in Section 3.2.2. As mentioned in
Chapter 1, the existence and properties of quarks were first inferred from hadron
spectroscopy by Gell-Mann and independently by Zweig in 1964 and the close
correspondence between the experimentally observed hadrons and those predicted
by the quark model, which we will examine in more detail later, remains one of the
strongest reasons for our belief in the existence of quarks.

Lepton scattering

It was mentioned in earlier chapters that in the early 1960s experiments were first
performed where electrons were scattered from protons and neutrons. These
strongly suggested that nucleons were not elementary. By the late 1960s this work
had been extended to higher energies and with projectiles that included muons and
neutrinos. In much the same way as Rutherford deduced the existence of the nucleus
in atoms, high-energy lepton scattering, particularly at large momentum transfers,
revealed the existence of point-like entities within the nucleons, which we now
identify as quarks.

Jet production

High-energy collisions can cause the quarks within hadrons, or newly created
quark–antiquark pairs, to fly apart from each other with very high energies. Before
they can be observed, these quarks are converted into ‘jets’ of hadrons (a process
referred to as fragmentation) whose production rates and angular distributions
reflect those of the quarks from which they originated. They were first clearly
identified in experiments at the DESY laboratory in Hamburg in 1979, where
electrons and positrons were arranged to collide ‘head-on’ in a magnetic field. An
example of a ‘two-jet’ event is shown in Figure 3.5. The picture is a computer
reconstruction of an end view along the beam direction; the solid lines indicate the
reconstructed charged particle trajectories taking into account the known magnetic
field, which is also parallel to the beam direction; the dotted lines indicate the
reconstructed trajectories of neutral particles, which were detected outside this
device by other means.
The production rate and angular distribution of the observed jets closely matches
that of quarks produced in the reaction

eþ þ e
! q þ
q; ð3:42Þ

by the mechanism of Figure 3.6. Such jets have now been observed in many
reactions, and are strong evidence for the existence of quarks within hadrons.
88 CH3 PARTICLE PHENOMENOLOGY

Figure 3.5 Two-jet event in eþ e
collisions

Figure 3.6 Mechanism for two-jet production in eþ e
annihilation reaction

The failure to detect free quarks is not an experimental problem. Firstly, free
quarks would be easily distinguished from other particles by their fractional
charges and their resulting ionization properties.12 Secondly, electric charge
conservation implies that a fractionally charged particle cannot decay to a final
state composed entirely of particles with integer electric charges. Hence the
lightest fractionally charged particle, i.e. the lightest free quark, would be stable
and so presumably easy to observe. Finally, some of the quarks are not very
massive (see below) and because they interact by the strong interaction, one would
expect free quarks to be copiously produced in, for example, high-energy proton–
proton collisions. However, despite careful and exhaustive searches in ordinary
matter, in cosmic rays and in high-energy collision products, free quarks have

12
We will see in Chapter 4 that energy losses in matter due to ionization are proportional to the square of the
charge and thus would be ‘anomalously’ small for quarks.
 QUARKS 89

never been observed. The conclusion – that quarks exist solely within hadrons and
not as isolated free particles – is called confinement. It is for this reason that we are
forced to study the properties of hadrons, the bound states of quarks.
The modern theory of strong interactions, called quantum chromodynamics
(QCD), which is discussed in Chapter 5, offers at least a qualitative account of
confinement, although much of the detail eludes us due to the difficulty of
performing accurate calculations. In what follows, we shall assume confinement
and use the properties of quarks to interpret the properties of hadrons.

3.2.2 Quark generations and quark numbers

Six distinct types, or flavours, of spin-12 quarks are now known to exist. Like the
leptons, they occur in pairs, or generations, denoted




u c t
; ; : ð3:43Þ
d s b

Each generation consists of a quark with charge þ 23 (u, c, or t) together with a quark
of charge
13 (d, s, or b), in units of e. They are called the down (d), up (u), strange
(s), charmed (c), bottom (b) and top (t) quarks. The quantum numbers associated
with the s, c, b and t quarks are called strangeness, charm, beauty and truth,
respectively. The antiquarks are denoted




d

s
b
; ; ð3:44Þ

u
c
t

with charges þ 13 (d
,
s, or
b) and
23 (
u,
c,
t).

Approximate quark masses are given in Table 3.2. Except for the top quark,
these masses are inferred indirectly from the observed masses of their hadron

Table 3.2 Properties of quarks: all have spin 12 and masses are given units of GeV/c2; the
antiparticles (not shown) have the same masses as their associated particles, but the electric
charges (Q) are reversed in sign (in the major decay modes, X denotes other particles)
Name Symbol Mass Q Lifetime (s) Major decays
Down d md  0:3
1=3
Up u mu  md 2=3
Strange s ms  0:5
1=3 10
8 –10
10 s!uþX
Charmed c mc  1:5 2=3 10
12 –10
13 c!s þX
c!d þX
Bottom b mb  4:5
1=3 10
12 –10
13 b!c þX
Top t mt ¼ 180  12 2=3 10
25 t !b þX
90 CH3 PARTICLE PHENOMENOLOGY

bound states, together with models of quark binding.13 In this context they are also
referred to as constituent quark masses.
The stability of quarks in hadrons – like the stability of protons and neutrons in
atomic nuclei – is influenced by their interaction energies. However, for the s; c
and b quarks these effects are small enough for them to be assigned approximate
lifetimes of 10
8 –10
10 s for the s quark and 10
12 –10
13 s for both the c and b
quarks. The top quark is much heavier than the other quarks and its lifetime is of
the order of 10
25 s. This lifetime is so short that when top quarks are created they
decay too quickly to form observable hadrons. In contrast to the other quarks, our
knowledge of the top quark is based entirely on observations of its decay products.
When we talk about ‘the decay of quarks’ we always mean that the decay takes
place within a hadron, with the other bound quarks acting as ‘spectators’, i.e. not
taking part in the interaction. Thus, for example, in this picture neutron decay at
the quark level is given by the Feynman diagram of Figure 3.7 and no free quarks
are observed. Note that it is assumed that the exchanged particle interacts with
only one constituent quark in the nucleons. This is the essence of the spectator
model. (This is not dissimilar to the idea of a single nucleon decaying within a
radioactive nucleus.)

Figure 3.7 Spectator model quark Feynman diagram for the decay n ! pe

e

In strong and electromagnetic interactions, quarks can only be created or destroyed

as particle–antiparticle pairs, just like electrons as we discussed in Section 3.1.1. This
implies, for example, that in electromagnetic processes corresponding to the Feyn-
man diagram of Figure 3.8, the reaction eþ þ e
! c þ
c, which creates a c
c pair, is
allowed, but the reaction eþ þ e
! c þ
u producing a c
u pair, is forbidden.14
More generally, it implies conservation of each of the six quark numbers

Nf
Nð f Þ
Nð
f Þ ð f ¼ u; d; s; c; b; tÞ ð3:45Þ

where Nðf Þ is the number of quarks of flavour f present and Nð
f Þ is the number of
antiquarks of flavour
f present. For example, for single-particle states; Nc ¼ 1 for the

13
An analogy would be to deduce the mass of nucleons from the masses of nuclei via a model of the nucleus.
14
Again, these reactions and associated Feynman diagrams do not imply that free quarks are created.
Spectator quarks are implicitly present to form hadrons in the final state.
 QUARKS 91

Figure 3.8 Production mechanism for the reaction eþ e
! qq

c quark; Nc ¼
1 for the
c antiquark; and Nc ¼ 0 for all other particles. Similar
results apply for the other quark numbers Nf , and for multi-particle states the quark
numbers of the individual particles are simply added. Thus a state containing the
particles u, u, d, has Nu ¼ 2, Nd ¼ 1 and Nf ¼ 0 for the other quark numbers with
f ¼ s, c, b, t.
In weak interactions, more general possibilities are allowed, and only the total
quark number
Nq
NðqÞ
Nð
qÞ ð3:46Þ

is conserved, where NðqÞ and Nð
qÞ are the total number of quarks and antiquarks
present, irrespective of their flavour. This is illustrated by the decay modes of the
quarks themselves, some of which are listed in Table 3.2, which are all weak inter-
action processes, and we have seen it also in the decay of the neutron in Figure 3.7.
Another example is the main decay mode of the charmed quark, which is

c ! s þ u þ d
; ð3:47Þ

in which a c quark is replaced by an s quark and a u quark is created together with

a d
antiquark. This clearly violates conservation of the individual quark numbers
Nc , Ns , Nu and Nd , but the total quark number Nq is conserved.
In practice, it is convenient to replace the total quark number Nq in analyses by
the baryon number, defined by
B
Nq =3 ¼ ½NðqÞ
Nð
qÞ=3: ð3:48Þ

Like the electric charge and the lepton numbers introduced in the last section, the
baryon number is conserved in all known interactions, and unlike the lepton
number, there are no experiments that suggest otherwise.15

15
However, there are theories beyond the standard model that predict baryon number non-conservation,
although there is no experimental evidence to support this prediction. These will be discussed briefly in
Chapter 9.
92 CH3 PARTICLE PHENOMENOLOGY

3.3 Hadrons
In principle, the properties of atoms and nuclei can be explained in terms of their
proton, neutron and electron constituents, although in practice many details are too
complicated to be accurately calculated. However, the properties of these con-
stituents can be determined without reference to atoms and nuclei by studying
them directly as free particles in the laboratory. In this sense atomic and nuclear
physics are no longer fundamental, although they are still very interesting and
important if we want to understand the world we live in.
In the case of hadrons, the situation is more complicated. Their properties are
explained in terms of a few fundamental quark constituents; but the properties of
the quarks themselves can only be studied experimentally by appropriate measure-
ments on hadrons. Whether we like it or not, studying quarks without hadrons is
not an option.

3.3.1 Flavour independence and charge multiplets

One of the most fundamental properties of the strong interaction is flavour

independence. This is the statement that the strong force between two quarks at
a fixed distance apart is independent of which quark flavours u; d; s; c; b; t are
involved. Thus, for example, the strong forces between us and ds pairs are
identical. The same principle applies to quark–antiquark forces which are,
however, not identical to quark–quark forces, because in the former case
annihilations can occur. Flavour independence does not apply to the electromag-
netic interaction, since the quarks have different electric charges, but compared
with the strong force between quarks, the electromagnetic force is a small
correction. In addition, in applying flavour independence one must take proper
account of the quark mass differences, which can be non-trivial. However, there
are cases where these corrections are small or easily estimated, and the phenom-
enon of flavour independence is plain to see.
One consequence of flavour independence is the striking observation that
hadrons occur in families of particles with approximately the same masses, called
charge multiplets. Within a given family, all particles have the same spin-parity
and the same baryon number, strangeness, charm and beauty, but differ in their
electric charges. Examples are the triplet of pions, (þ ; 0 ; 
) and the nucleon
doublet ðp; nÞ. The latter behaviour reflects an approximate symmetry between u
and d quarks. This arises because, as we shall see in Section 3.3.2, these two
quarks have only a very small mass difference
md
mu ¼ ð3  1ÞMeV=c2 ; ð3:49Þ
so that in this case mass corrections can to a good approximation be neglected. For
example, consider the proton and neutron. We shall see in the next section that their
quark content is pð938Þ ¼ uud and nð940Þ ¼ udd. If we neglect the small mass
 HADRONS 93

difference between the u and d quarks and also the electromagnetic interactions,
which is equivalent to setting all electric charges to zero, so that the forces acting on
the u and d quarks are exactly equal, then replacing the u quark by a d quark in the
proton would produce a ‘neutron’ which would be essentially identical to the proton.
Of course the symmetry is not exact because of the small mass difference between
the u and d quarks and because of the electromagnetic forces, and it is these that give
rise to the small differences in mass within multiplets.
Flavour independence of the strong forces between u and d quarks also leads
directly to the charge independence of nuclear forces, e.g. the equality of the force
between any pair of nucleons, provided the two particles are in the same spin state.
Subsumed in the idea of charge independence is the idea of charge symmetry, i.e.
the equality of the proton–proton and neutron–neutron forces, again provided the
two particles are in the same spin state. Evidence for the latter is found in studies
of nuclei with the same value of A, but the values of N and Z interchanged (mirror
nuclei). An example is shown in Figure 3.9. The two nuclei 115 B and 116 C have the
same number of np pairs, but 115 B has 10 pp pairs and 15 nn pairs, whereas 116 C has
15 pp pairs and 10 nn pairs. Thus, allowing for the Coulomb interaction, the
approximate equality of the level structures of these two nuclei, as seen in
Figure 3.9, means charge symmetry is approximately verified. To test charge
independence in a nuclear context we would have to look at the level structure in
three related nuclei such as 114 Be, 115 B and 116 C.
Here the test is not so clear-cut because an np pair is not subject to the
restrictions of the Pauli principle like pp and nn pairs and there is evidence (to be
discussed briefly in Chapter 7) that the np force is stronger in the S ¼ 1 state than
in the S ¼ 0 state. Nevertheless, the measured energy levels in such triplets of
nuclei support the idea of approximate charge independence of nuclear forces.
The symmetry between u and d quarks is called isospin symmetry and greatly
simplifies the interpretation of hadron physics. It is described by the same mathe-
matics as ordinary spin, hence the name. For example, the proton and neutron are
viewed as the ‘up’ and ‘down’ components of a single particle, the nucleon N, that
has an isospin quantum number I ¼ 12, with I3 values 12 and
12, assigned to the proton
and neutron, where I3 is analogous to the magnetic quantum number in the case of
ordinary spin. Likewise, the three pions þ ; 
and 0 are part of a triplet  with
I ¼ 1 corresponding to I3 values 1, 0 and
1, respectively. In discussing the strong
interactions between pions and nucleons, it is then only necessary to consider the N
interaction with total isospin either 12 or 32.
As an example, we will consider some predictions for the hadronic resonance
ð1232Þ. The ð1232Þ has I ¼ 32 and four charge states þþ ; þ ; 0 and 
(see
Table 3.3) corresponding to I3 ¼ 32; 12; 
12;
32, respectively. If we use the notation
jN; I; I3 i for a N state, then N; 32; 32 is the unique state þ p and may be written

  
  
N; 3; 3 ¼ ; 1; 1 N; 1; 1 : ð3:50Þ
 2 2   2 2
94 CH3 PARTICLE PHENOMENOLOGY

Figure 3.9 Low-lying energy levels with spin-parity JP of the mirror nuclei 11
5B and 116 C. (data
from Aj90)

The other N states may then be obtained by applying quantum mechanical shift
(ladder) operators to Equation (3.50), as is done when constructing ordinary spin
states. This gives16  rffiffiffi rffiffiffi
  
N; 3; 1 ¼
1þ n þ 20 p ð3:51Þ
 2 2 3  3
and hence isospin invariance predicts
ðþ ! þ nÞ 1
¼ ; ð3:52Þ
ðþ ! 0 pÞ 2
which is in good agreement with experiment.
16
The reason for the minus sign and other details are given in, for example, Appendix D of Ma97.
 HADRONS 95

Secondly, by constructing all the N isospin states by analogy with Equations

(3.50) and (3.51) we can show that

  rffiffiffi

 
 p ¼ p1ffiffiffiN; 3;
1
2N; 1;
1 ð3:53aÞ
 3 2 2 3 2 2

and
 rffiffiffi 
 0  
 n ¼ 2N; 3;
1 þ p1ffiffiffiN; 1;
1 : ð3:53bÞ
 3 2 2 3 2 2

Then, if MI is the amplitude for scattering in a pure isospin state I,

1 2
Mð
p ! 
pÞ ¼ M3 þ M1 ð3:54aÞ
3 3

and
pffiffiffi pffiffiffi

0 2 2
Mð p !  nÞ ¼ M3
M1 : ð3:54bÞ
3 3

At the ð1232Þ, the available energy is such that the total cross-section is
dominated by the elastic (
p ! 
p) and charge-exchange (
p ! 0 n) reac-
tions. In addition, because the ð1232Þ has I ¼ 32, M3  M1 , so

1
total ð
pÞ ¼ ð
p ! 
pÞ þ ð
p ! 0 nÞ / jM3 j2 ð3:55aÞ
3

and

total ðþ pÞ / jM3 j2 : ð3:55bÞ

Thus, neglecting small kinematic corrections due to mass differences (phase space
corrections), isospin symmetry predicts

total ðþ pÞ
¼ 3: ð3:56Þ
total ð
pÞ

Figure 3.10 shows the two total cross-sections at low energies. There are clear
peaks with Breit–Wigner forms at a mass of 1232 MeV corresponding to
the production of the ð1232Þ and the ratio of the peaks is in good agreement
with the prediction of Equation (3.56).
96 CH3 PARTICLE PHENOMENOLOGY

Figure 3.10 Total cross-sections for 
p and þ p scattering

3.3.2 Quark model spectroscopy

The observed hadrons are of three types. There are baryons and their antiparticles
antibaryons, which have half-integral spin, and mesons, which have integral spin. In
the quark model of hadrons the baryons are assumed to be bound states of three
quarks (3q), antibaryons are assumed to be bound states of three antiquarks (3
q) and
mesons are assumed to be bound states of a quark and an antiquark (q
q).17 The
17
In addition to these so-called ‘valence’ quarks there could also, in principle, be other constituent quarks present
in the form of a cloud of virtual quarks and antiquarks – the so-called ‘sea’ quarks – the origin of which we will
discuss in Chapter 5. In this chapter we consider only the valence quarks which determine the static properties of
hadrons. The masses of the constituent quarks could be quite different from those that appear in the fundamental
strong interaction Hamiltonian for quark–quark interactions via gluon exchange (i.e. QCD), because those
quarks are free of the dynamical effects experienced in hadrons. The latter are referred to as ‘current’ quarks.
 HADRONS 97

baryons and antibaryons have baryon numbers 1 and
1 respectively, while the
mesons have baryon number 0. Hence the baryons and antibaryons can annihilate
each other in reactions which conserve baryon number to give mesons or, more
rarely, photons or lepton–antilepton pairs, in the final state.
The lightest known baryons are the proton and neutron, with the quark
compositions given in Section 3.3.1:

p ¼ uud and n ¼ udd: ð3:57Þ

These particles have been familiar as constituents of atomic nuclei since the
1930s. The birth of particle physics as a new subject, distinct from atomic and
nuclear physics, dates from 1947, when hadrons other than the neutron and proton
were first detected. These were the pions, already mentioned, and the kaons,
discovered in cosmic rays by groups in Bristol and Manchester Universities, UK,
respectively.
The discovery of the pions was not totally unexpected, since Yukawa had
famously predicted their existence and their approximate masses in 1935, in order
to explain the observed range of nuclear forces (recall the discussion in Section
1.5.2). This consisted of finding what mass was needed in the Yukawa potential to
give the observed range of the strong nuclear force (which was poorly known at the
time). After some false signals, a particle with the right mass and suitable properties
was discovered – this was the pion. Here and in what follows we will give the hadron
masses in brackets in units of MeV/c2 and use a superscript to indicate the electric
charge in units of e. Thus the pions are  ð140Þ; 0 ð135Þ. Pions are the lightest
known mesons and have the quark compositions

þ ¼ ud
; 0 ¼ u
u; d d
; 
¼ d
u: ð3:58Þ

While the charged pions have a unique composition, the neutral pion is composed of
both u
u and d d
pairs in equal amounts. Pions are copiously produced in high-energy
collisions by strong interaction processes such as p þ p ! p þ n þ þ .
In contrast to the discovery of the pions, the discovery of the kaons was totally
unexpected, and they were almost immediately recognized as a completely new
form of matter, because they had supposedly ‘strange’ properties. Eventually, after
several years, it was realized that these properties were precisely what would be
expected if kaons had non-zero values of a hitherto unknown quantum number,
given the name strangeness, which was conserved in strong and electromagnetic
interactions, but not necessarily conserved in weak interaction. Particles with non-
zero strangeness were named strange particles, and with the advent of the quark
model in 1964, it was realized that strangeness S was, apart from a sign, the
strangeness quark number introduced earlier, i.e.

S ¼
Ns : ð3:59Þ
98 CH3 PARTICLE PHENOMENOLOGY

Kaons are the lightest strange mesons, with the quark compositions:

K þ ð494Þ ¼ u
s and K 0 ð498Þ ¼ d
s; ð3:60Þ

where K þ and K 0 have S ¼ þ1 and their antiparticles K
and K
0 have S ¼
1,

while the lightest strange baryon is the lambda, with the quark composition
 ¼ uds. Subsequently, hadrons containing c and b quarks have also been dis-
covered, with non-zero values of the charm and beauty quantum numbers defined by

~

Nb

NðbÞ
Nð
bÞ:
C
Nc
NðcÞ
Nð
cÞ and B ð3:61Þ

The above examples illustrate just some of the many different combinations of
quarks that form baryons or mesons. These and some further examples are shown
in Table 3.3 and a complete listing is given in the PDG Tables.

Table 3.3 Some examples of baryons and mesons, with their major decay
modes; masses are in MeV/c2
Particle Mass Lifetime (s) Major decays
 ðud
Þ
þ
140 2:6  10
8
þ
 ( 100%)
0 ðu
u; d d
Þ 135 8:4  10
17 ( 100%)
K þ ðu
sÞ 494 1:2  10
8 þ
 (64%)
þ 0 (21%)
K þ ðu
sÞ 892 1:3  10
23 K þ 0 ; K 0 þ ( 100%)
D
ðd
cÞ 1869 1:1  10
12 Several seen
B
ðb
uÞ 5278 1:6  10
12 Several seen
pðuudÞ 938 Stable None
nðuddÞ 940 887 pe

e (100%)
ðudsÞ 1116 2:6  10
10 p
(64%)
n0 (36%)
þþ ðuuuÞ 1232 0:6  10
23 pþ (100%)

ðsssÞ 1672 0:8  10
10 K
(68%)
0 
(24%)
þ
c ðudcÞ 2285 2:1  10
13 Several seen

To proceed more systematically one could, for example, construct all the mesons
states of the form q
q, where q can be any of the six quark flavours. Each of these is
labelled by its spin and its intrinsic parity P. The simplest such states would have
the spins of the two quarks antiparallel with no orbital angular momentum between
them and so have spin-parity J P ¼ 0
. (Recall from Chapter 1 that quarks and
antiquarks have opposite parities.) If, for simplicity, we consider those states
composed of just u, d and s quarks, there will be nine such mesons and they have
quantum numbers which may be identified with the observed mesons ðK 0 ; K þ Þ,

0 ; K
Þ, ð ; 0 Þ and two neutral particles, which are called  and 0 . This
ðK
supermultiplet is shown Figure 3.11(a) as a plot of Y, the hypercharge, defined as
 HADRONS 99

þ
Figure 3.11 The lowest-lying states with (a) JP ¼ 0
and (b) J ¼ 12 that are composed of u, d
and s quarks

Y
BþSþCþB ~ þ T, against I3 , the third component of isospin. This can be
extended to the lowest-lying qqq states and the lowest-lying supermultiplet
þ
consists of the eight J P ¼ 12 baryons shown in Figure 3.11(b).18
It is a remarkable fact that the states observed experimentally agree with those
predicted by the simple combinations qqq;
q
q
q and q
q and until very recently there
was no evidence for states corresponding to any other combinations. However, some
recent experiments have claimed evidence for the existence of a few states outside this
scheme, possibly ones involving five quarks, although other experiments have failed to
confirm this. Nevertheless, it is still a fact that hadron states are overwhelmingly
composed of the simplest quark combinations of the basic quark model. This was one
of the original pieces of evidence for the existence of quarks and remains one of the
strongest today.
The scheme may also be extended to more quark flavours, although the diagrams
become increasingly complex. For example, Figure 3.12 shows the predicted
þ
J P ¼ 32 baryon states formed from u, d, s and c quarks when all three quarks
have their spins aligned, but still with zero orbital angular momentum between them.
All the states in the bottom plane have been detected as well as many in the higher
planes and with the possible exception of the five-quark states mentioned previously,
no states have been found that are outside this scheme. The latest situation may be
found in the PDG Tables.
For many quark combinations there exist not one, but several states. For example,
the lowest-lying state of the ud
system has spin-parity 0
and is the þ meson. It can
be regarded as the ‘ground state’ of the ud
system. Here the spins of the quark

18
If you try to try to verify Figure 3.11, you will find that it is necessary to assume that the overall hadronic
wavefunctions ¼ space spin are symmetric under the exchange of identical quarks, i.e. opposite to the
symmetry required by the Pauli principle. This apparent contradiction will be resolved in Chapter 5.
100 CH3 PARTICLE PHENOMENOLOGY

þ
Figure 3.12 The J ¼ 32 baryon states composed of u, d, s and c quarks

constituents are anti-aligned to give a total spin S ¼ 0 and there is no orbital angular
momentum L between the two quarks, so that the total angular momentum, which
we identify as the spin of the hadron, is J ¼ L þ S ¼ 0. Other ‘excited’ states can
have different spin-parities depending on the different states of motion of the quarks
within the hadron.
An example is the K þ ð890Þ meson shown in Table 3.3 with J P ¼ 1
. In this state
the u and
s quarks have their spins aligned so that S ¼ 1 and there is no orbital
angular momentum between them, i.e. L ¼ 0, so that the spin of the K þ is
J ¼ L þ S ¼ 1. This is a resonance and such states usually decay by the strong
interaction, with very short lifetimes, of order 10
23 s. The mass distribution of their
decay products is described by the Breit–Wigner formula we met in Section 1.6.3.
The spin of a resonance may be found from an analysis of the angular distributions of
its decay products. This is because the distribution will be determined by the
wavefunction of the decaying particle and this will contain an angular part
proportional to a spherical harmonic labelled by the orbital angular momentum
between the decay products. Thus from a measurement of the angular distribution of
the decay products, the angular momentum may be found, and hence the spin of the
resonance. It is part of the triumph of the quark model that it successfully accounts
for the excited states of the various quark systems, as well as their ground states,
when the internal motion of the quarks is properly taken into account.
From experiments such as electron scattering we know that hadrons have typical
radii r of the order of 1 fm and hence associated time scales r/c of the order of
10
23 s. The vast majority are highly unstable resonances, corresponding to excited
 HADRONS 101

states of the various quark systems, and decay to lighter hadrons by the strong
interaction, with lifetimes of this order. The K þ ð890Þ ¼ u
s resonance, mentioned
above, is an example. It decays to K þ 0 and K 0 þ final states with a lifetime of
1:3  10
23 s. The quark description of the process K þ ! K 0 þ þ , for example, is

u
s ! d
s þ ud
: ð3:62Þ

From this we see that the final state contains the same quarks as the initial state,
plus an additional d d
pair, so that the quark numbers Nu and Nd are separately
conserved. This is characteristic of strong and electromagnetic processes, which
are only allowed if each of the quark numbers Nu ; Nd ; Ns ; Nc and Nb is separately
conserved.
Since leptons and photons do not have strong interactions, hadrons can only decay
by the strong interaction if lighter states composed solely of other hadrons exist with
the same quantum numbers. While this is possible for the majority of hadrons, it is
not in general possible for the lightest state corresponding to any given quark
combination. These hadrons, which cannot decay by strong interactions, are long-
lived on a timescale of the order of 10
23 s and are often called stable particles. It
is more accurate to call them long-lived particles, because except for the proton
they are not absolutely stable, but decay by either the electromagnetic or weak
interaction.
The proton is stable because it is the lightest particle with non-zero baryon
number and baryon number is conserved in all known interactions. A few of the
other long-lived hadrons decay by electromagnetic interactions to final states that
include photons. These decays, like the strong interaction, conserve all the
individual quark numbers. An example of this is the neutral pion, which has
Nu ¼ Nd ¼ Ns ¼ Nc ¼ Nb ¼ 0 and decays by the reaction

0 ðu
u; d d
Þ ! þ ; ð3:63Þ

with a lifetime of 0:8  10
16 s. However, most of the long-lived hadrons have non-
zero values for at least one of the quark numbers, and can only decay by the weak
interaction, in which quark numbers do not have to be conserved. For example, the
positive pion decays with a lifetime of 2:6  10
8 s by the reaction

þ ! þ þ
 ; ð3:64Þ

while the ð1116Þ ¼ uds baryon decays mainly by the reactions

 ! p þ 
and n þ 0 ; ð3:65Þ

with a lifetime of 2:6  10
10 s. The quark interpretations of these reactions are

ðud
Þ ! þ þ
 ; ð3:66Þ
102 CH3 PARTICLE PHENOMENOLOGY

in which a u quark annihilates with a d
antiquark, violating both Nu and Nd

conservation; and for lambda decay to charged pions,

sud ! uud þ d
u; ð3:67Þ

in which an s quark turns into a u quark and a ud
pair is created, violating Nd and
Ns conservation.
We see from the above that the strong, electromagnetic or weak nature of a given
hadron decay can be determined by inspecting quark numbers. The resulting
lifetimes can then be summarized as follows. Strong decays lead to lifetimes that
are typically of the order of 10
23 s. Electromagnetic decay rates are suppressed by
powers of the fine structure constant  relative to strong decays, leading to observed
lifetimes in the range 10
16
10
21 s. Finally, weak decays give longer lifetimes,
which depend sensitively on the characteristic energy of the decay.
A useful measure of the decay energy is the Q-value, the kinetic energy released in
the decay of the particle at rest, which we metioned before in Section 2.3. In the weak
interactions of hadrons, Q-values of the order of 102
103 MeV are typical, leading
to lifetimes in the range 10
7 –10
13 s, but there are some exceptions, notably
neutron decay, n ! p þ e
þ
e , for which

Q ¼ mn
mp
me
m
e ¼ 0:79 MeV ð3:68Þ

is unusually small, leading to a lifetime of about 103 s. Thus hadron decay lifetimes
are reasonably well understood and span some 27 orders of magnitude, from about
10
24 s to about 103 s. The typical ranges corresponding to each interaction are
summarized in Table 3.4.

Table 3.4 Typical lifetimes of hadrons decaying by

the three interactions
Interaction Lifetimes (s)
Strong 10
22 –10
24
Electromagnetic 10
16 –10
21
Weak 10
7 –10
13

3.3.3 Hadron masses and magnetic moments

The quark model can make predictions for hadronic magnetic moments and masses
in a way that is analogous to the semi-empirical mass formula for nuclear masses, i.e.
the formulae have a theoretical basis, but contain parameters that have to be
determined from experiment. We start by examining the case of baryon magnetic
moments.
 HADRONS 103
þ
These have been measured only for the 12 octet of states composed of u, d and s
quarks and so we will consider only these. In this supermultiplet, the quarks have
zero orbital angular momentum and so the hadron magnetic moments are just the
sums of contributions from the constituent quark magnetic moments, which we will
assume are of the Dirac form, i.e.
  
1  1  
q
q; Sz ¼ ^z q; Sz ¼ ¼ eq e
h=2mq ¼ eq M p =mq N ; ð3:69Þ
2 2

where eq is the quark charge in units of e and N

e
h=2Mp is the nuclear
magneton. Thus

2Mp Mp Mp
u ¼ N ; d ¼
N and s ¼
N : ð3:70Þ
3mu 3md 3ms

Consider, for example, the case of the ð1116Þ ¼ uds. It is straightforward to

show that the configuration that ensures that the predicted quantum numbers of the
supermultiplet agree with experiment is to have the ud pair in a spin-0 state. Hence
it makes no contribution to the  spin or magnetic moment. Thus we have the
immediate prediction

Mp
 ¼ s ¼
N : ð3:71Þ
3ms

þ
For 12 baryons B with quark configuration aab, the aa pair is in the symmetric
spin-1 state with parallel spins (again this is to ensure that the predicted quantum
numbers of the supermultiplet agree with experiment) and magnetic moment 2a .
The ‘spin-up’ baryon state is given by

 rffiffiffi 
 1 1 2 1 1 
B; S ¼ ; Sz ¼ ¼ b; S ¼ ; S ¼
aa; S ¼ 1; S ¼ 1
 2 2 3 2
z
2 
z
rffiffiffi 
1 1 1 

b; S ¼ ; S ¼ aa; S ¼ 1; S ¼ 0 ð3:72Þ
3 2 
z z
2

The first term corresponds to a state with magnetic moment 2a
b , since the b
quark has Sz ¼
12; the second term corresponds to a state with magnetic moment
b , since the aa pair has Sz ¼ 0 and does not contribute. Hence the magnetic
moment of B is given by

2 1 4 1
B ¼ ð2a
b Þ þ b ¼ a
b : ð3:73Þ
3 3 3 3
104 CH3 PARTICLE PHENOMENOLOGY

For example, the magnetic moment of the proton is

4 1 Mp
p ¼ u
d ¼ N ; ð3:74Þ
3 3 m

where we have neglected the mass difference between the u and d quarks, as
suggested by isospin symmetry, and set mu  md
m. The predictions for the
þ
magnetic moments of all the other members of the 12 octet may be found in a
similar way in terms of just two parameters, the masses m and ms . A best fit to the
measured magnetic moments (but not taking account of the errors on the data19)
yields the values m ¼ 0:344 GeV=c2 and ms ¼ 0:539 GeV=c2 . The predicted
moments are shown in Table 3.5. The agreement is good, but by no means perfect
and suggests that the assumption that baryons are pure three-quark states with zero
orbital angular momentum between them is not exact. For example, there could be
small admixtures of states with non-zero orbital angular momentum.

þ
Table 3.5 Magnetic moments of the 12 baryon octet as predicted by the
constituent quark model, compared with experiment in units of N , the
nuclear magneton; these have been obtained using m ¼ 0:344 GeV/c2 and
ms ¼ 0:539 GeV/c2 -- errors on the nucleon moments are of the order of 10
7
Particle Moment Prediction Experiment
4 1
p(938) 3u
3d 2.73 2.793
4 1
n(940) 3d
3u
1.82
1.913
(1116) s
0.58
0:613  0:004
þ (1189) 4 1
3u
3s 2.62 2:458  0:010

4 1
 (1197) 3d
3s
1.02
1:160  0:025
4 1
0 ð1315Þ 3s
3u
1.38
1:250  0:014

4 1
 ð1321Þ 3s
3d
0.47
0:651  0:003

We now turn to the prediction of hadron masses. The mass differences between
members of a given supermulitplet are conveniently separated into the small mass
differences between members of the same isospin multiplet and the much larger
mass differences between members of different isospin multiplets. The size of the
former suggests that they have their origin in electromagnetic effects, and if we
neglect them then a first approximation would be to assume that the mass
differences are due solely to differences in the constituent quark masses. If we
concentrate on hadrons with quark structures composed of u, d and s quarks, since

19
If we had fitted taking account of the errors, the fit would be dominated by the proton and neutron moments
because they have very small errors.
 HADRONS 105

their masses are the best known from experiment, this assumption leads directly to
the relations

M
M ¼ M
M ¼ M
MN ¼ ms
mu;d ð3:75Þ

1þ
for the 2 baryon octet and

M
M ¼ M
M ¼ M
M ¼ ms
mu;d ð3:76Þ

þ
for the 32 decuplet. These give numerical estimates for ms
mu;d in the range 120
to 200 MeV=c2 , which are consistent with the estimate from magnetic moments
above.
These results support the suggestion that baryon mass differences (and by analogy
meson mass differences) are dominantly due to the mass differences of their
constituent quarks. However, this cannot be the complete explanation, because if
þ þ
it were then the 12 nucleon would have the same mass as the 32 ð1232Þ, as they
þ
have the same quark constituents, and similarly for other related particles in the 12
þ
octet and 32 decuplet. The absence of orbital angular momentum in these states
means that there is nothing equivalent to the ‘fine structure’ of atomic physics. The
difference lies in the spin structures of these states.
If we take the case of two spin-12 particles with magnetic moments mi and mj
separated by a distance rij then the interaction energy is proportional to
mi  mj =rij 3 . If, in addition, the particles are point-like and have charges
ei and ej , the moments will be of the Dirac form mi ¼ ðei =mi ÞSi . Then for two
particles in a relative S-state it can be shown that the interaction energy is given by

8 ei ej
E ¼ j ð0Þj2 Si  Sj ; ð3:77Þ
3 mi mj

where ð0Þ is the wavefunction at the origin, rij ¼ 0. (When averaged over all
space, the interaction is zero except at the origin.) In atomic physics this is known
as the hyperfine interaction and causes very small splittings in atomic energy
levels. In the hadron case, the electric charges must be replaced by their strong
interaction equivalents with appropriate changes to the overall numerical factor.
The resulting interaction is called (for reasons that will be clear in Chapter 5) the
chromomagnetic interaction. As we cannot calculate the equivalent quark–quark
wavefunction, for the purposes of a phenomenological analysis we will write the
contribution to the hadron mass as

S1  S2
M / : ð3:78Þ
m1 m2

This of course assumes that j ð0Þj2 is the same for all states, which will not be
exactly true.
106 CH3 PARTICLE PHENOMENOLOGY

Consider first the case of mesons. By writing the total spin squared as

S2
ðS1 þ S2 Þ2 ¼ S21 þ S22 þ 2S1  S2 ; ð3:79Þ

we easily find the expect values of S1  S2 are
34
h2 for the S ¼ 0 (pseudoscalar)
mesons and 14
h2 for the S ¼ 1 (vector) mesons. Then the masses may be written
MðmesonÞ ¼ m1 þ m2 þ M; ð3:80Þ
where m1;2 are the masses of the constituent quarks and
3a 1 a 1
MðJ P ¼ 0
mesonÞ ¼
; MðJ P ¼ 1
mesonÞ ¼ ð3:81Þ
4 m1 m2 4 m1 m2
and a is a constant to be found from experiment. The masses of the members of the
0
and 1
meson supermultiplets then follow from a knowledge of their quark
compositions. For example, the K-mesons have one u or d quark and one s quark and so


3a 1 1
MK ¼ m þ ms
þ : ð3:82Þ
8 m2 m2s

Predictions for the masses of all the mesons are shown in Table 3.6, which also
gives the best fit to the measured masses (again ignoring the relative errors on the
latter) using these formulae. The predictions correspond to the values

m ¼ 0:308 GeV=c2 ; ms ¼ 0:482GeV=c2 ; a ¼ 0:0588 ðGeV=c2 Þ3 : ð3:83Þ

Note that the quark mass values are smaller than those obtained from fitting the
baryon magnetic moments. There is no contradiction in this, because there is no
reason that quarks should have the same effective masses in mesons as in baryons.

Table 3.6 Meson masses (in Gev/c2) in the constituent quark model compared
with experimental values
Particle Mass Prediction Experiment
3a
 2m
0.15 0.137
4m2

3a 1 1
K m þ ms
þ 0.46 0.496
8 m2 m2s


2 4 a 1 2
 m þ ms
þ 0.57 0.549
3 3 4 m2 m2s
a
 2m þ 2 0.77 0.770
4m
a
! 2m þ 2 0.77 0.782
4m

a 1 1
K m þ ms þ þ 0.87 0.892
8 m2 m2s
a
 2ms þ 2 1.03 1.020
4ms
 HADRONS 107

The comparison with the measured values is very reasonable, but omitted from the fit
is the 0 state where the fit is very poor indeed. Unlike the atomic case, the spin–spin
interaction in the strong interaction case leads to substantial corrections to the meson
masses.
The baryons are somewhat more complicated, because in this case we have three
pairs of spin–spin couplings to consider. In general the spin–spin contribution to
the mass is
X Si  Sj
M / ; i; j ¼ 1; 3: ð3:84Þ
i<j
mi mj

þ
In the case of the 32 decuplet, all three quarks have their spins aligned and every
pair therefore combines to make spin-1. Thus for example,

ðS1 þ S2 Þ2 ¼ S21 þ S22 þ 2S1  S2 ¼ 2
h2 ; ð3:85Þ

giving S1  S2 ¼
h2 =4 and in general

S1  S2 ¼ S1  S3 ¼ S2  S3 ¼
h2 =4: ð3:86Þ

Using this result, the mass of the  ð1385Þ, for example, may be written


b 1 2
M ¼ 2m þ ms þ þ ; ð3:87Þ
4 m2 mms

where b is a constant to be determined from experiment. (There is no reason for b

to be equal to the constant a used in the meson case because the quark
wavefunctions and numerical factors in the baryonic equivalent of Equation
(3.77) will be different in the two cases.)
þ
In the case of the 12 octet, we have

ðS21 þ S22 þ S23 Þ ¼ S21 þ S22 þ S23 þ 2ðS1  S2 þ S1  S3 þ S2  S3 Þ ¼ 3
h2 =4 ð3:88Þ

and hence

S1  S2 þ S1  S3 þ S2  S3 ¼
3
h2 =4: ð3:89Þ

In addition, we have to consider the symmetry of the spin wavefunctions of

individual hadrons. For example, without proof (this will be given in Chapter 5),
the spins of the u and d pair in the  must combine to give S ¼ 0. Thus,
ðSu þ Sd Þ2 ¼ 0, so that Su  Sd ¼
3
h2 =4. Then,
 
Su  Sd Su  Ss Sd  Ss
M ¼ mu þ md þ ms þ b þ þ : ð3:90Þ
mu md mu ms md ms
108 CH3 PARTICLE PHENOMENOLOGY

Table 3.7 Baryon masses (in Gev/c2) in the constituent quark model compared
with experimental values
Particle Mass Prediction Experiment
3b
N 3m
2 0.89 0.939
4m

3b 1
 2m þ ms
1.08 1.116
4
m2 
b 1 4
 2m þ ms þ
1.15 1.193
4
m2 mms 
b 1 4
 m þ 2ms þ
1.32 1.318
4 m2s mms
3b
 3m þ 2 1.07 1.232
4m

b 1 2
 2m þ ms þ þ 1.34 1.385
4 m mms


b 2 1
 m þ 2ms þ þ 2 1.50 1.533
4 mms ms
3b
 3ms þ 2 1.68 1.673
ms

Finally, setting mu ¼ md ¼ m and absorbing factors of
h2 into the constant b, gives
 
Su  Sd ðS1  S2 þ S1  S3 þ S2  S3
Su  Sd Þ 3b
M ¼ 2m þ ms þ b 2
þ ¼ 2m þ ms
2 ;
m m ms 4m
ð3:91Þ
þ
where we have used Equation (3.89). The resulting formulae for all the 12 octet and
3þ
2 decuplet masses are shown in Table 3.7. Also shown are the predicted masses,
where for consistency we have used the same quark mass values obtained earlier in
fitting baryon magnetic moments, i.e. m ¼ 0:308 GeV=c2 and ms ¼ 0:482 GeV=c2 ,
and varied only the parameter b, giving a value 0.0225 ðGeV=c2 Þ3 . The fit is quite
reasonable and although better fits can be obtained by allowing the masses to vary
there is little justification for this, given the approximations of the analysis.
Overall, what we learn from the above is that the constituent quark model is
capable of giving a reasonably consistent account of hadron masses and magnetic
moments, at least for the low-lying states (the 0 is an exception), provided a few
parameters are allowed to be found from experiment.

Problems
3.1 Which of the following reactions are allowed and which are forbidden by the
conservation laws appropriate to weak interactions?
(a)
 þ p ! þ þ n;
(b)
e þ p ! n þ e
þ þ ;
(c)  ! þ þ e
þ
e ;
(d) K þ ! 0 þ þ þ
 .
 PROBLEMS 109

3.2 Draw the lowest-order Feynman diagram at the quark level for the following
decays:

(a) D
! K 0 þ 
;

(b)  ! p þ e
þ
e .

~ ) where
3.3 Consider the following combinations of quantum numbers (Q; B; S; C; B
Q ¼ electric charge, B ¼ baryon number, S ¼ strangeness, C ¼ charm and B ~¼
beauty:

(a) ð
1; 1;
2; 0;
1Þ;

(b) (0, 0, 1, 0, 1).

Which of these possible states are compatible with the postulates of the quark
model?

3.4 Consider a scenario where overall hadronic wavefunctions consist of just spin and
space parts, i.e. ¼ space spin . What would be the resulting multiplet structure of
the lowest-lying baryon states composed of u; d and s quarks?

3.5 Draw Feynman diagrams at the quark level for the reactions:

0 þ B0 , where B is a meson containing a b-quark;

(a) eþ þ e
! B

(b) 
þ p ! K 0 þ 0 .

3.6 Find the parity P and charge conjugation C values for the ground-state ðJ ¼ 0Þ
meson  and its first excited ðJ ¼ 1Þ state . Why does the charged pion have a
longer lifetime than the ? Explain also why the decay 0 ! þ 
has been
observed, but not the decay 0 ! 0 0.

3.7 The particle Y
can be produced in the strong interaction process

K
þ p ! K þ þ Y
. Deduce its baryon number, strangeness, charm and beauty,
and using these, its quark content. The Y
ð1311Þ decays by the reaction
Y
!  þ 
. Give a rough estimate of its lifetime.
þ
3.8 Verify the expression in Table 3.7 for the mass of the 12  baryon, given that the
spins of the two non-strange quarks combine to give S ¼ 1.

3.9 Consider the reaction K
þ p ! 
þ K þ þ K 0 followed by the sequence of decays


! 0 þ 

K þ ! þ þ 0
j
! 0 þ ; j and K 0 ! þ þ 
þ 0
j ! þ þ

! þ

Classify each process as strong, weak or electromagnetic and give your reasons.
110 CH3 PARTICLE PHENOMENOLOGY

3.10 Draw the lowest-order Feynman diagram for the decay K þ ! þ þ
 þ and
hence deduce the form of the overall effective coupling.

3.11 A KamLAND-type experiment detects
e neutrinos at a distance of 200 m from a

nuclear reactor and finds that the flux is ð90  10Þ per cent of that expected if there
were no oscillations. Assuming maximal mixing and a mean neutrino energy of
3 MeV, use this result to estimate upper and lower bounds on the squared mass of
the
e .

3.12 Comment on the feasibility of the following reactions:

(a) p þ
p ! þ þ 
;

(b) p ! eþ þ ;

(c) 0 !  þ ;

(d) p þ p ! þ þ n þ K 0 þ þ ;

(e) 
!  þ 
;

(f) þ ! p þ 0 .

3.13 Use the results of Section 3.3.1 to deduce a relation between the total cross-sections
for the reactions 
p ! K 0 0 ; 
p ! K þ 
and þ p ! K þ þ at a fixed energy.

3.14 At a certain energy ðþ nÞ  ð
pÞ, whereas ðK þ nÞ 6¼ ðK
pÞ. Explain this.
4
Experimental Methods

In earlier chapters we have discussed the results of a number of experiments, but

said almost nothing about how such experiments are done. In this chapter we will
take a brief look at experimental methods. This is a very extensive subject and the
aim will not be to give a comprehensive review, but rather to emphasize the
physical principles behind the methods. More details may be found in specialized
texts.1

4.1 Overview
To explore the structure of nuclei (nuclear physics) or hadrons (particle physics)
requires projectiles whose wavelengths are at least as small as the effective radii of
the nuclei or hadrons. This determines the minimum value of the momentum
p ¼ h= and hence the energy required. The majority of experiments are
conducted using beams of particles produced by machines called accelerators.
This has the great advantage that the projectiles are of a single type, and have
energies that may be controlled by the experimenter.2 For example, beams that are
essentially mono-energetic may be prepared, and can be used to study the energy
dependence of interactions. The beam, once established, is directed onto a target so
that interactions may be produced. In a fixed-target experiment the target is
stationary in the laboratory. Nuclear physics experiments are almost invariably of
this type, as are many experiments in particle physics.
In particle physics, high energies are also required to produce new and unstable
particles and this reveals a disadvantage of fixed-target experiments when large

1
See, for example, Fe86 and Kl86.
2
Nevertheless, important experiments are still performed without using accelerators, for example some of
those described in Chapter 3 on neutrino oscillations used cosmic rays and nuclear reactors. In fact cosmic
rays are still the source of the very highest-energy particles.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
112 CH4 EXPERIMENTAL METHODS

centre-of-mass energies are required. The centre-of-mass energy is important

because it is a measure of the energy available to create new particles. In the
laboratory frame at least some of the final-state particles must be in motion to
conserve momentum. Consequently, at least some of the initial beam energy must
reappear as kinetic energy of the final-state particles, and is therefore unavailable
for particle production. In contrast, in the centre-of-mass frame the total momen-
tum is zero and, in principle, all the energy is available for particle production.
To find the centre-of-mass energy we use the expression
2
ECM ¼ ð Pt þ P b Þ 2 c 2 ; ð4:1Þ
where P is the particle’s four-momentum and the subscripts t and b refer to target
and beam, respectively.3 For a fixed-target experiment in the laboratory we have
Pt ¼ ðmt c ; 0Þ; Pb ¼ ðEL =c; pb Þ: ð4:2Þ
Expanding Equation (4.1) gives
2


ECM ¼ P2t þ P2b þ 2Pt Pb c2 ð4:3Þ

and using P2t ¼ m2t c2 etc., together with the general result
Pi Pj ¼ Ei Ej =c2  pi  pj ; ð4:4Þ

we have
 1=2
ECM ¼ m2b c4 þ m2t c4 þ 2mt c2 EL : ð4:5Þ
1
At high energies this increases only as ðEL Þ and so an increasingly smaller
2

fraction of the beam energy is available for particle production, most going to
impart kinetic energy to the target.
In a colliding-beam accelerator, two beams of particles travelling in almost
opposite directions are made to collide at a small or zero crossing angle. If for
simplicity we assume the particles in the two beams have the same mass and
laboratory energy EL and collide at zero crossing angle, then the total centre-
of-mass energy is
ECM ¼ 2EL : ð4:6Þ
This increases linearly with the energy of the accelerated particles, and hence is a
significant improvement on the fixed-target result. Colliding-beam experiments are
not, however, without their own disadvantages. The colliding particles have to be
stable, which limits the interactions that can be studied, and the collision rate in the
intersection region is smaller than that achieved in fixed-target experiments,
because the particle densities in the beams are low compared with a solid or
liquid target.

3
A brief summary of relativistic kinematics is given in Appendix B.
 ACCELERATORS AND BEAMS 113

Finally, details of the particles produced in the collision (e.g. their momenta) are
deduced by observing their interactions with the material of detectors, which are
placed in the vicinity of the interaction region. A wide range of detectors is
available. Some have a very specific characteristic, others serve more than one
purpose. Modern experiments, particularly in particle physics, typically use several
types in a single experiment.
In this chapter we start by describing some of the different types of accelerator
that have been built, the beams that they can produce and also how beams of
neutral and unstable particles can be prepared. Then we discuss the ways in which
particles interact with matter, and finally review how these mechanisms are
exploited in the construction of a range of particle detectors.

4.2 Accelerators and Beams

All accelerators use electromagnetic forces to boost the energy of stable charged
particles. These are injected into the machine from a device that provides a high-
intensity source of low-energy particles, for example an electron gun (a hot
filament), or a proton ion source. The accelerators used for nuclear structure
studies may be classified into those that develop a steady accelerating field
(DC machines) and those in which radio frequency electric fields are used
(AC machines). All accelerators for particle physics are of the latter type. We
start with a brief description of DC machines.

4.2.1 DC accelerators

The earliest type of DC accelerator was the Cockcroft–Walton machine, in which

ions pass through sets of aligned electrodes that are operated at successively higher
potentials. These machines are limited to energies of about 1 MeV, but are still
sometimes used as injectors as part of the multistage process of accelerating
particles to higher energies.4
The most important DC machine in current use is the van de Graaff accelerator
and an ingenious version of this, known as the tandem van de Graaff, that doubles
the energy of the simple machine, is shown schematically in Figure 4.1. The key
to this type of device is to establish a very high voltage. The van de Graaff
accelerator achieves this by using the fact that the charge on a conductor resides
on its outermost surface and hence if a conductor carrying charge touches another
conductor it will transfer its charge to the outer surface of the second conductor.

4
Sir John Cockcroft and Ernest Walton received the 1951 Nobel Prize in Physics for the development of their
accelerator and the subsequent nuclear physics experiments they did using it.
Figure 4.1 Principle of the tandem van de Graaff accelerator
 ACCELERATORS AND BEAMS 115

In Figure 4.1, a high voltage source at I passes positive ions to a belt via a comb
arrangement at C. The belt is motor driven via the pulleys at P and the ions are
carried on the belt to a second pulley where they are collected by another comb
located within a metal vessel T. The charges are then transferred to the outer
surface of the vessel, which acts as an extended terminal. In this way a high
voltage is established on T. Singly-charged negative ions are injected from a
source and accelerated along a vacuum tube towards T. Within T there is a stripper
S (for example a thin carbon foil) that removes two or more electrons from the
projectiles to produce positive ions. The latter then continue to accelerate through
the second half of the accelerator increasing their energy still further and finally
may be bent and collimated to produce a beam of positive ions. This brief account
ignores many technical details. For example, an inert gas at high pressure is used to
minimize electrical breakdown by the high voltage. The highest energy van de
Graaff accelerator can achieve a potential of about 30–40 MeV for singly-charged
ions and greater if more than one electron is removed by the stripper. It has been a
mainstay of nuclear research.

4.2.2 AC accelerators

Accelerators using radio frequency (r.f.) electric fields may conveniently be

divided into linear and cyclic varieties.

Linear accelerators

In a linear accelerator (or linac) for acclerating ions, particles pass through a series
of metal pipes called drift tubes, that are located in a vacuum vessel and connected
successively to alternate terminals of an r.f. oscillator, as shown in Figure 4.2.
Positive ions accelerated by the field move towards the first drift tube. If the

Figure 4.2 Acceleration in a linear ion accelerator

116 CH4 EXPERIMENTAL METHODS

alternator can change its direction before the ions passes through that tube, then
they will be accelerated again on their way between the exit of the first and entry
to the second tube, and so on. Thus the particles will form bunches. Because
the particles are accelerating, their speed is increasing and hence the lengths
of the drift tubes have to increase to ensure continuous acceleration. To produce
a useful beam the particles must keep in phase with the r.f. field and remain
focused.
In the case of electrons, their velocity very rapidly approaches the speed of light.
In this case a variation of the linac method that is more efficient is used to
accelerate them. Bunches of particles pass through a straight evacuated waveguide
with a periodic array of gaps, similar to the ion accelerator. Radio frequency
oscillations in the gaps are used to establish a moving electromagnetic wave in the
structure, with a longitudinal component of the electric field moving in phase with
the particles. As long as this phase relationship can be maintained, the particles
will be continuously accelerated. Radio frequency power is pumped into the
waveguide at intervals to compensate for resistive losses and gives energy to the
electrons. The largest electron linac is at the SLAC laboratory in Stanford, USA,
and has a maximum energy of 50 GeV. It is over 3 km long.
An ingenious way of reducing the enormous lengths of high-energy linacs has
been developed at the Continuous Electron Beam Accelerator Facility (CEBAF) at
the Jefferson Laboratory in the USA. This utilizes the fact that above about
50 MeV, electron velocities are very close to the speed of light and thus electrons
of very different energies can be accelerated in the same drift tube. Instead of a
single long linac, the CEBAF machine consists of two much shorter linacs and the
beam from one is bent and passed through the other. This can be repeated for up to
four cycles. Even with the radiation losses inherent in bending the beams, very
intense beams can be produced with energies between 0.5 and 6.0 GeV. CEBAF is
proving to be an important machine in the energy region where nuclear physics
and particle physics descriptions overlap.

Cyclic accelerators

Cyclic accelerators used for low-energy nuclear physics experiments are of a type
called cyclotrons. They are also used to produce beams of particles for medical
applications, including proton beams for radiation therapy. Cyclotrons operate in a
somewhat different way to cyclic accelerators used in particle physics, which are
called synchrotrons. In a cyclotron,5 charged particles are constrained to move
in near-circular orbits by a magnetic field during the acceleration process. There
are several types of cyclotron; we will describe just one. This is illustrated

5
The cyclotron was invented by Ernest Lawrence, who received the 1939 Nobel Prize in Physics for this and
the experimental work he did using it.
 ACCELERATORS AND BEAMS 117

Figure 4.3 Schematic diagram of a cyclotron (adapted from Kr88, copyright John Wiley &
Sons)

schematically in Figure 4.3. The accelerator consists of two ‘dee’-shaped sections

across which an r.f. electric field is established. Charged ions are injected into the
machine near its centre and are constrained to traverse outward in spiral
trajectories by a magnetic field. The ions are accelerated each time they pass
across the gap between the dees. At the maximum radius, which corresponds to the
maximum energy, the beam is extracted. The shape of the magnetic field, which is
also shown in Figure 4.3, ensures that forces act on particles not orbiting in the
medium plane to move them closer to this plane. This brief description ignores the
considerable problems that have to be overcome to ensure that the beam remains
focused during the acceleration.
The principle of a synchrotron is analogous to that of a linear accelerator, but
where the acceleration takes place in a near circular orbit rather than in a straight
line. The beam of particles travels in an evacuated tube called the beam pipe and is
constrained in a circular or near circular path by an array of dipole magnets called
bending magnets (Figure 4.4). Acceleration is achieved as the beam repeatedly
traverses one or more cavities placed in the ring where energy is given to the
particles. Since the particles travel in a circular orbit they continuously emit
radiation, called in this context synchrotron radiation. The amount of energy
radiated per turn by a relativistic particle of mass m is proportional to 1=m4 .
For electrons the losses are thus very severe, and the need to compensate for these
by the input of large amounts of r.f. power limits the energies of electron
synchrotrons.
The momentum in GeV/c of an orbiting particle assumed to have unit charge is
given by p ¼ 0:3B, where B is the magnetic field in Tesla and , the radius of
curvature, is measured in metres. Because p is increased during acceleration, B
must also be steadily increased if  is to remain constant, and the final momentum
is limited both by the maximum field available and by the size of the ring. With
118 CH4 EXPERIMENTAL METHODS

Figure 4.4 Cross-section of (a) a typical bending (dipole) magnet, and (b) a focusing
(quadrupole) magnet; the thin arrows indicate field directions; the thick arrows indicate the
force on a negative particle travelling into the paper

conventional electromagnets, the largest field attainable over an adequate region is

about 1.5 T, and even with superconducting coils it is only of the order of 10 T.
Hence the radius of the ring must be very large to achieve very high energies.
For example, the Tevatron accelerator, located at the Fermi National Laboratory,
Chicago, USA, which accelerates protons to an energy of 1 TeV, has a radius of
l km. A large radius is also important to limit synchrotron radiation losses in
electron machines.
In the course of its acceleration, a beam may make typically 105 traversals of its
orbit before reaching its maximum energy. Consequently stability of the orbit is
vital, both to ensure that the particles continue to be accelerated, and that they do
not strike the sides of the vacuum tube. In practice, the particles are accelerated in
bunches each being synchronized with the r.f. field. In equilibrium, a particle
increases its momentum just enough to keep the radius of curvature constant as
the field B is increased during one rotation, and the circulation frequency of the
particle is in step with the r.f. of the field. This is illustrated in Figure 4.5. With
obvious changes, a similar principle is used in linear accelerators.
In practice, the particles remain in the bunch, but their trajectories oscillate
about the stable orbits. These oscillations are controlled by a series of focusing
magnets, usually of the quadrupole type, which are placed at intervals around the
beam and act like optical lenses. A schematic diagram of one of these is shown in
Figure 4.4. Each focuses the beam in one direction and so alternate magnets have
their field directions reversed.
In addition to the energy of the beam, one is also concerned to produce a beam
of high intensity, so that interactions will be plentiful. The intensity is ultimately
limited by defocussing effects, e.g. the mutual repulsion of the particles in the
 ACCELERATORS AND BEAMS 119

Figure 4.5 Magnitude of the electric field as a function of time at a fixed point in the rf cavity:
particle B is synchronous with the field and arrives at time tB ; particle A (C) is behind (ahead of)
B and receives an increase (decrease) in its rotational frequency -- thus particles oscillate about
the equilibrium orbit

beam, and a number of technical problems have to be overcome which are outside
the scope of this brief account.6

Fixed-target machines and colliders

Both linear and cyclic accelerators can be divided into fixed-target and colliding-
beam machines. The latter are also known as colliders, or sometimes in the case of
cyclic machines, storage rings.7 In fixed-target machines, particles are accelerated
to the highest operating energy and then the beam is extracted from the machine
and directed onto a stationary target, which is usually a solid or liquid. Much
higher energies have been achieved for protons than electrons, because of the large
radiation losses inherent in electron machines mentioned earlier. The intensity of
the beam is such that large numbers of interactions can be produced, which can
either be studied in their own right or used to produce secondary beams.
The main disadvantage of fixed-target machines for particle physics has been
mentioned earlier: the need to achieve large centre-of-mass energies to produce

6
Very recently (2005) significant progress has been made on an ‘induction synchrotron’ in which a changing
magnetic field produces the electric field that accelerates the particles. This device has the potential to
overcome certain effects that limit the intensity achievable in conventional synchrotrons.
7
The use of the terms storage rings and colliders as synonymous is not strictly correct, because we will see
that the former can also describe a machine that stores a single beam for use on both internal and external
fixed targets.
120 CH4 EXPERIMENTAL METHODS

new particles. Almost all new machines for particle physics are therefore colliders,
although some fixed-target machines for specialized purposes are still constructed.
The largest collider currently under construction is the Large Hadron Collider
(LHC), which is being built at CERN, Geneva, Switzerland. This is a massive pp
accelerator of circumference 27 km, with each beam having an energy of 7 TeV.
A schematic diagram of the CERN site showing the LHC and some of its other
accelerators is shown in Figure 4.6. The acceleration process starts with a linac

Figure 4.6 A schematic diagram of the CERN site showing the LHC and some of its other
accelerators (CERN photo, reproduced with permission)
 ACCELERATORS AND BEAMS 121

whose beam is boosted in energy in the Proton Synchrotron Booster (PSB) and
passed to the Proton Synchrotron (PS), a machine that is still the source of beams
for lower-energy experiments. The beam energy is increased still further in the
Super Proton Synchrotron (SPS) that also provides beams for a range of experi-
ments as well as the injection beams for the LHC itself. Four beam intersection
points are shown in the LHC and experiments (ALICE, CMS, LHC-b and ATLAS)
will be located at each of these. The extracted neutrino beam shown at the bottom
of the diagram is sent to the Gran Sasso laboratory 730 km away and is used,
amongst other things, for experiments on neutrino oscillations of the type
discussed in Chapter 3.
Another very large collider we should mention is the Relativistic Heavy Ion
Collider (RHIC), located at Brookhaven National Laboratory, USA. This unique
machine, which began operation in 2000 following 10 years of development and
construction, is the first collider in the world capable of accelerating heavy ions.
Like the LHC above, there are several stages, involving a linac, a tandem van de
Graaff and a synchrotron, before the ions are injected into the main machine.
There they form two counter-circulating beams controlled by two 4-km rings of
superconducting magnets and are accelerated to an energy of 100 GeV/nucleon.
Thus the total centre-of-mass energy is 200 GeV/nucleon. Collisions occur at six
intersection points, where major experiments can be sited. RHIC primarily
accelerates ions of gold and is used to study matter at extreme energy-densities,
where a new state of matter called a ‘quark–gluon plasma’ is predicted to occur.
We will return to this briefly in Chapter 9.
The performance of a collider is characterized by its luminosity, which was
defined in Chapter 1. The general formula for luminosity given there is shown in
Problem 1.10 to reduce in the case of a collider to the useful form

N1 N2
L¼n f; ð4:7Þ
A

where Ni ði ¼ 1,2) are the numbers of particles in the n colliding bunches, A is the
cross-sectional area of the beam and f is the frequency, i.e. f ¼ 1=T, where T is the
time taken for the particles to make one traversal of the ring.
An interesting proton synchrotron for nuclear physics studies is the COSY
facility located at the Research Centre Jülich, Germany. Low-energy protons are
pre-accelerated in a cyclotron, then cooled to reduce their transverse momentum
and injected into a synchrotron, where they are further accelerated to momenta in
the range 600–3700 MeV/c (corresponding to energies of 175–2880 MeV). The
protons can be stored in the ring for appreciable times and are available for
experiments not only in the usual way by extracting the beam, but also by using the
circulating beam to interact with a very thin internal target. Thus we have a
mixture of storage rings and fixed targets. The fact that the circulating beam may
make as many as 1010 traversals through the target compensates to some extent for
its low particle density.
122 CH4 EXPERIMENTAL METHODS

4.2.3 Neutral and unstable particle beams

The particles used in accelerators must be stable and charged, but one is also
interested in the interaction of neutral particles, e.g. photons and neutrons, as well
as those of unstable particles, such as charged pions. Beams appropriate for
performing such experiments are produced in a number of ways.
We have seen that neutrons are the natural product of radioactive decays and we
will see in Chapter 8 that a large flux of neutrons is present in a nuclear reactor.
Typically these will have a spectrum concentrated at low energies of 1–2 MeV, but
extending as high as 5–6 MeV. Purpose-built reactors exist for research purposes,
such as the ILL reactor at the Institut Laue-Langevin, France. Another source of
neutrons is via the spallation process. The most important neutron spallation
source at present is ISIS located at the Rutherford Appleton Laboratory, UK.
Protons which have been accelerated in a linac to 70 MeV are injected into a
synchrotron that further accelerates them to 800 MeV, where they collide with a
heavy metal target of tantalum. The interaction drives out neutrons from the target
and provides an intense pulsed source. In each case, if beams of lower-energy
neutrons are required these are produced by slowing down faster neutrons in
moderators, which are materials with a large cross-section for elastic scattering,
but a small cross-section for absorption. In Chapter 8 we will see that moderators
are vital for the successful extraction of power from fission nuclear reactors.
Beams of unstable particles can be formed provided their constituents live long
enough to travel appreciable distances in the laboratory. One way of doing this is to
direct an extracted primary beam onto a heavy target. In the resulting interactions
with the target nuclei, many new particles are produced which may then be
analysed into secondary beams of well-defined momentum. Such beams will
ideally consist predominantly of particles of one type, but if this cannot
be achieved, then the wanted species may have to be identified by other means.
In addition, if these secondary beams are composed of unstable particles, they can
themselves be used to produce further beams formed from their decay products.
Two examples will illustrate how, in principle, such secondary particle beams can
be formed.
Consider firstly the construction of a K  beam from a primary beam of protons.
By allowing the protons to interact with a heavy target, secondary particles will be
produced. Most of these will be pions, but a few will be kaons (that have to be
produced with a hyperon to conserve strangeness – this an example of so-called
associated production). A collimator can be used to select particles in a particular
direction, and the K  component can subsequently be removed and focused into a
mono-energetic beam by selective use of electrostatic fields and bending and
focusing magnets.
The pion beam may also be used to produce a beam of neutrinos. For example,
the  is unstable and as we have seen, one of its weak interaction decays modes is
 !  þ  . So if the pions are passed down a long vacuum pipe, many will
 PARTICLE INTERACTIONS WITH MATTER 123

decay in flight to give muons and anti-neutrinos, which will mostly travel in
essentially the same direction as the initial beam. The muons and any remaining
pions can then be removed by passing the beam through a very long absorber,
leaving the neutrinos. In this case the final neutrino beam will have a momentum
spectrum reflecting the initial momentum spectrum of the pions and, since
neutrinos are neutral, no further momentum selection using magnets is possible.

4.3 Particle Interactions with Matter

In order to be detected, a particle must undergo an interaction with the material of
a detector. In this section we discuss these interactions, but only in sufficient detail
to be able to understand the detectors themselves.
The first possibility is that the particle interacts with an atomic nucleus. For
example, this could be via the strong nuclear interaction if it is a hadron, or by the
weak interaction if it is a neutrino. We know from the work of Chapter 1 that both
are short-range interactions. If the energy is sufficiently high, new particles may
be produced, and such reactions are often the first step in the detection process. In
addition to these short-range interactions, a charged particle will also excite and
ionize atoms along its path, giving rise to ionization energy losses, and emit
radiation, leading to radiation energy losses. Both of these processes are due to the
long-range electromagnetic interaction. They are important because they form the
basis of most detectors for charged particles. Photons are also directly detected by
electromagnetic interactions, and at high energies their interactions with matter
lead predominantly to the production of eþ e pairs via the pair production process
! eþ þ e , which has to occur in the vicinity of a nucleus to conserve energy
and momentum. (Recall the discussion in Chapter 1 on the range of forces.) All
these types of interactions are described in the following sections.

4.3.1 Short-range interactions with nuclei

For hadrons, the most important short-range interactions with nuclei are due to the
strong nuclear force which, unlike the electromagnetic interaction, is as important
for neutral particles as for charged ones, because of the charge-independence of the
strong interaction. Both elastic scattering and inelastic reactions may result. At
high energies, many inelastic reactions are possible, most of them involving the
production of several particles in the final state.
Many hadronic cross-sections show considerable structure at low energies due to
the production of hadronic resonances, but at energies above about 3 GeV, total
cross-sections are usually slowly varying in the range 10–50 mb and are much
larger than the elastic cross-section. (The example of  p scattering is shown
in Figure 4.7.) This is of the same order-of-magnitude as the ‘geometrical’
124 CH4 EXPERIMENTAL METHODS

Figure 4.7 Total and elastic cross-sections for  p scattering as functions of the pion
laboratory momentum

cross-section r 2  30 mb, where r  1 fm is the approximate range of the strong

interaction between hadrons. Total cross-sections on nuclei are much larger see
for example Figure (2.17), increasing roughly as the square of the nuclear radius,
i.e. as A2=3 .
A special case is the detection of thermal neutrons (defined as those with kinetic
energies energies below about 0.02 eV). We have seen in Chapter 2 that neutrons in
this region have very large cross-sections for being absorbed, leading to the
production of a compound nucleus which decays by delayed emission of a -ray.
Examples of these so-called neutron activation reactions are 63 Cuðn; Þ64 Cu and
55
Mnðn; Þ56 Mn.
The probability of a hadron-nucleus interaction occurring as the hadron
traverses a small thickness dx of material is given by n
tot dx, where n is the
number of nuclei per unit volume in the material. Consequently, the mean distance
travelled before an interaction occurs is given by
‘c ¼ 1=n
tot : ð4:8Þ

This is called the collision length. An analogous quantity is the absorption length,
defined by

‘a ¼ 1=n
inel ; ð4:9Þ

that governs the probability of an inelastic collision. In practice, ‘c  ‘a at high

energies. As examples, the interaction lengths are between 10 and 40 cm
for nucleons of energy in the range 100–300 GeV interacting with metals such
as iron.
 PARTICLE INTERACTIONS WITH MATTER 125

Neutrinos and antineutrinos can also be absorbed by nuclei, leading to reactions

of the type
‘ þ p ! ‘þ þ X; ð4:10Þ
where ‘ is a lepton and X denotes any hadron or set of hadrons allowed by the
conservation laws. Such processes are weak interactions (because they involve
neutrinos) and the associated cross-sections are extremely small compared with the
cross-sections for strong interaction processes. The corresponding interaction
lengths are therefore enormous. Nonetheless, in the absence of other possibilities
such reactions are the basis for detecting neutrinos. Finally, photons can be
absorbed by nuclei, giving photoproduction reactions such as þ p ! X. How-
ever, these electromagnetic interactions are only used to detect photons at low
energies, because at higher energies there is a far larger probability for eþ e pair
production in the Coulomb field of the nucleus. We will return to this in
Section 4.3.4.

4.3.2 Ionization energy losses

Ionization energy losses are important for all charged particles, and for particles
other than electrons and positrons they dominate over radiation energy losses at all
but the highest attainable energies. The theory of such losses, which are due
dominantly to Coulomb scattering from the atomic electrons, was worked out by
Bethe, Bloch and others in the 1930s. The result is called the Bethe–Bloch
formula, and for spin-0 bosons with charge q (in units of e), mass M and velocity
v, it takes the approximate form (neglecting small corrections for highly relativistic
particles)
   
dE D q2 ne 2me c2  2 2 2
 ¼ ln   ; ð4:11Þ
dx 2 I

where x is the distance travelled through the medium;

h2
4 2 
D¼ ¼ 5:1
1025 MeV cm2 ; ð4:12Þ
me

me is the electron mass,  ¼ v=c and ¼ ð1  2 Þ1=2 . The other constants refer
to the properties of the medium: ne is the electron density; I is the mean ionization
potential of the atoms averaged over all electrons, which is given approximately by
I ¼ 10 Z eV for Z greater than 20. The corresponding formula for spin-12 particles
differs from this, but in practice the differences are small and may be neglected in
discussing the main features of ionization energy loses.
Examples of the behaviour of dE=dx for muons, pions and protons traversing a
range of materials is shown in Figure 4.8. It is common practice to absorb the
density  of the medium by dividing by  and expressing dE=dx in terms of an
126 CH4 EXPERIMENTAL METHODS

Figure 4.8 Ionization energy loss for muons, pions and protons on a variety of mate-
rials (reprinted from Ei04, copyright Elsevier, with permission)

equivalent thickness of gm cm2 – hence the units in Figure 4.8. As can be seen,
dE=dx falls rapidly as the velocity increases from zero because of the 1= 2 factor
in the Bethe–Bloch equation. All particles have a region of ‘minimum ionization’
for  in the range 3–4. Beyond this,  tends to unity, and the logarithmic factor in
the Bethe–Bloch formula gives a ‘relativistic rise’ in dE=dx.
The magnitude of the energy loss depends on the medium. The electron density
is given by ne ¼ NA Z=A, where NA is Avogadro’s number, and  and A are the
mass density and atomic weight of the medium, so the mean energy loss is
proportional to the density of the medium. The remaining dependence on the
medium is relatively weak because Z=A  0:5 for all atoms except the very light
 PARTICLE INTERACTIONS WITH MATTER 127

and the very heavy elements, and because the ionization energy I only enters the
Bethe–Bloch formula logarithmically. In the ‘minimum ionization’ region where
  3–4, the minimum value of dE=dx can be calculated from Equation (4.11)
and for a particle with unit charge is given approximately by
 
dE Z
  3:5 MeVg1 cm2 : ð4:13Þ
dx min A
Ionization losses are proportional to the squared charge of the particle, so that a
fractionally charged particle with   3 would have a much lower rate of energy
loss than the minimum energy loss of any integrally charged particle. This has
been used as a means of identifying possible free quarks, but without success.
From the knowledge of the rate of energy loss, we can calculate the attenuation
as a function of distance travelled in the medium. This is called the Bragg curve.
Most of the ionization loss occurs near the end of the path where the speed is
smallest and the curve has a pronounced peak (the Bragg peak) close to the end
point before falling rapidly to zero at the end of the particle’s path length.
The range R, i.e. the mean distance a particle travels before it comes to rest is
defined as
ð
xmax

R dxðÞ; ð4:14Þ
0

which, using Equation (4.11), may be written

ð 
initial 1
dE dE M
R¼  d ¼ 2 Fðinitial Þ; ð4:15Þ
dx d q ne
0

where F is a function of the initial velocity and we have used the relation
E ¼ Mc2 to show the dependence on the projectile mass M.
The range as given by Equation (4.15) is actually an average value because
scattering is a statistical process and there will therefore be a spread of values for
individual particles. The spread will be greater for light particles and smaller for
heavier particles such as -particles. These properties have implications for the use
of radiation in therapeutic situations, where it may be necessary to deposit energy
within a small region at a specific depth of tissue, for example to precisely target a
cancer. The biological effects of radiation will be discussed in Chapter 8.
Because neutrons are uncharged, direct detection is not possible by ionization
methods. However, they can be detected via the action of the charged products of
induced direct nuclear reactions. Commonly used reactions are 6 Liðn; Þ3 H,
10
Bðn; Þ7 Li and 3 Heðn; pÞ3 H. All these reactions are exothermic and so are
very suitable for detecting neutrons with energies below about 20 MeV. Moreover,
as nuclear cross-sections tend to increase as v1 at low energies, detection
becomes more efficient the slower the neutron.
128 CH4 EXPERIMENTAL METHODS

4.3.3 Radiation energy losses

When a charged particle traverses matter it can also lose energy by radiative
collisions, especially with nuclei. The electric field of a nucleus will accelerate and
decelerate the particles as they pass, causing them to radiate photons, and hence
lose energy. This process is called bremsstrahlung (literally ‘braking radiation’ in
German) and is a particularly important contribution to the energy loss for
electrons and positrons.
The dominant Feynman diagrams for electron bremsstrahlung in the field of a
nucleus, i.e.
e þ ðZ; AÞ ! e þ þ ðZ; AÞ; ð4:16Þ
are shown in Figure 4.9 and are of the order of Z 2 3 . The function of the nucleus
is to absorb the recoil energy and so ensure that energy and momentum are
simultaneously conserved (recall the discussion of Feynman diagrams in Chapter 1).

Figure 4.9 Dominant Feynman diagrams for the bremsstrahlung process e þ ðZ; AÞ ! e þ þ ðZ; AÞ

There are also contributions from bremsstrahlung in the fields of the atomic
electrons, each of the order of 3 . Since there are Z atomic electrons for each
nucleus, these give a total contribution of the order of Z 3 , which is small
compared with the contribution from the nucleus for all but the lightest elements.
A detailed calculation shows that for relativistic electrons with E mc2 = Z 1=3 ,
the average rate of energy loss is given by
dE=dx ¼ E=LR : ð4:17Þ
The constant LR is called the radiation length and is a function of Z and na , the
number density of atoms/cm3 in the medium. Integrating Equation (4.17) gives
E ¼ E0 expðx=LR Þ; ð4:18Þ
where E0 is the initial energy. It follows that the radiation length is the average
thickness of material that reduces the mean energy of an electron or positron by a
factor e. For example, the radiation length in lead is 0.566 cm.
From these results, we see that at high energies the radiation losses are
proportional to E=m2p for an arbitrary charged particle of mass mp . On the other
 PARTICLE INTERACTIONS WITH MATTER 129

hand, the ionization energy losses are only weakly dependent on the projectile
mass and energy at very high energies. Consequently, radiation losses completely
dominate the energy losses for electrons and positrons at high enough energies, but
are much smaller than ionization losses for all particles other than electrons and
positrons at all but the highest energies.
Taking into account the above and the results of Section 4.3.2, we see that at low
energies, particles with the same kinetic energy but different masses can have
substantially different ranges. Thus, for example, an electron of 5 MeV has a range
that is several hundred times that of an -particle of the same kinetic energy.

4.3.4 Interactions of photons in matter

In contrast to heavy charged particles, photons have a high probability of being

absorbed or scattered through large angles by the atoms in matter. Consequently, a
collimated monoenergetic beam of I photons per second traversing a thickness dx
of matter will lose
dx
dI ¼ I ð4:19Þ

photons per second, where
 ¼ ðna
Þ1 ð4:20Þ
is the mean free path before absorption or scattering out of the beam, and
is
the total photon interaction cross-section with an atom. The mean free path  is
analogous to the collision length for hadronic reactions. Integrating Equation (4.19)
gives
IðxÞ ¼ I0 ex= ð4:21Þ
for the intensity of the beam as a function of distance, where I0 is the initial
intensity.
The main processes contributing to
are: Rayleigh scattering, in which the
photon scatters coherently from the atom, the photoelectric effect, in which the
photon is absorbed by the atom as a whole with the emission of an electron;
Compton scattering,8 where the photon scatters from an atomic electron; and
electron–positron pair production in the field of a nucleus or of an atomic electron.
The corresponding cross-sections on carbon and lead are shown in Figure 4.10,
where it can be seen that above a few MeV the cross-section is dominated by pair
production from the nucleus. The pair production process is closely related to

8
Arthur Compton shared the 1927 Nobel Prize in Physics for the discovery of the increase in wavelength that
occurs when photons with energies of around 0.5–3.5 MeV interact with electrons in a material – the original
Compton effect.
130 CH4 EXPERIMENTAL METHODS

Figure 4.10 Total experimental photon cross-section

on (a) a carbon atom, and (b) a lead
atom, together with the contributions from (a) the photoelectric effect, (b) Rayleigh (coherent
atomic) scattering, (c) Compton scattering, (d) pair production in the field of the nucleus, and
(e) pair production in the field of the atomic electrons (adapted from Ei04, copyright Elsevier,
with permission)

Figure 4.11 The pair production process þ ðZ; AÞ ! e þ eþ þ ðZ; AÞ

electron bremsstrahlung, as can be seen by comparing the Feynman diagrams

shown in Figures 4.9 and 4.11.
The cross-section for pair production rises rapidly from threshold, and is given
to a good approximation by
7 1

pair ¼ ; ð4:22Þ
9 na LR
for E mc2 = Z 1=3, where LR is the radiation length. Substituting these results
into Equation (4.21), gives
IðxÞ ¼ I0 expð7x=9LR Þ; ð4:23Þ
so that at high energies, photon absorption, like electron radiation loss, is
characterized by the radiation length LR .
 PARTICLE DETECTORS 131

4.4 Particle Detectors

The detection of a particle means more than simply its localization. To be useful
this must be done with a resolution sufficient to enable particles to be separated
in both space and time in order to determine which are associated with a parti-
cular event. We also need to be able to identify each particle and measure its
energy and momentum. No single detector is optimal with respect to all these
requirements, although some are multifunctional. For example, calorimeters,
primarily used for making energy measurements, can also have very good space
and time resolution. Many of the devices discussed below are commonly used both
in nuclear and particle physics, but in the former a small number of types of
detector is often sufficient, whereas in particle physics, both at fixed-target
machines and colliders, modern experiments commonly use very large multi-
component detectors which integrate many different sub-detectors in a single
device. Such systems rely heavily on fast electronics and computers to monitor
and control the sub-detectors, and to coordinate, classify and record the vast
amount of information flowing in from different parts of the apparatus. In this
section we will briefly introduce some of the most important detectors currently
available, but detector development is a rapidly-moving major area of research
and new devices are frequently developed, so the list below is by no means
exhaustive.9

4.4.1 Gas detectors

Most gas detectors detect the ionization produced by the passage of a charged
particle through a gas, typically an inert one such as argon, either by collecting the
ionization products or induced charges onto electrodes, or (historically) by making
the ionization track visible in some form. The average energy needed to produce
an electron–ion pair is 30 10 eV, with a weak dependence on the gas used and
the energy of the incident particle. In practice, the output is a pulse at the anode
(which is amplified by electronic means), with the bulk of the signal being due to
the positive ions because of their longer drift distance. For a certain range of
applied voltages – the so-called ‘proportional region’ (see below) – these devices
are primarily used to provide accurate measurements of a particle’s position. As
position detectors, gas detectors largely replaced earlier detectors which used
visual techniques, such as cloud chambers, bubble chambers and stacks of
photographic emulsions, although the latter are still an ingredient in some neutrino
experiments. Although historically important, none of these visual devices are now

9
For more detailed discussions of particle detectors see, for example, Gr96 and the references in Footnote 1.
There are also useful reviews in Chapter 5 of Ho97 and Ei04.
132 CH4 EXPERIMENTAL METHODS

Figure 4.12 Gas amplification factor as a function of voltage V applied in a single-wire gas
detector, with a wire radius typically 20 mm, for a strongly ionizing particle ( ) and a weakly
ionizing particle (electron)

in general use and they have been superceded by electronic detectors.10 In particle
physics experiments being planned at the new accelerators currently being built,
gas detectors themselves are being replaced by a new generation of solid-state
detectors based on silicon.
To understand the principles of gas detectors we refer to Figure 4.12, which
shows the number of ion pairs produced per incident charged particle (the gas

10
These early detector techniques produced many notable discoveries and their importance has been
recognized by the award of no less than five Nobel Prizes in Physics: a share of the 1927 Prize to Charles
Wilson for the invention and use of the cloud chamber; the 1948 Prize to Patrick Blackett for further
developments of the cloud chamber and discoveries made with it; the 1950 Prize to Cecil Powell for
development of the photographic emulsion technique and its use to discover pions; the 1960 Prize to Donald
Glaser for the invention of the bubble chamber; and the 1968 Prize to Luis Alvarez for developing the bubble
chamber and associated data analysis techniques resulting in the discovery of a large number of hadronic
resonances.
 PARTICLE DETECTORS 133

amplification factor) as a function of the applied voltage V for two cases: a heavily
ionizing particle (e.g. an alpha particle – upper curve) and a lightly ionizing
particle (e.g. an electron – lower curve).

Ionization chamber

At low applied voltages, the output signal is very small because electron–ion pairs
recombine before reaching the electrodes, but as the voltage increases the number
of pairs increases to a saturation level representing complete collection. This is the
region of the ionization chamber. The simplest type of chamber is a parallel plate
condenser filled with an inert gas and having an electric field E ¼ V=d, where d is
the distance between the plates. In practice the gas mixture must contain at lease
one ‘quenching’ component that absorbs ultraviolet light and stops a plasma
forming and spreading throughout the gas.
Another arrangement is cylindrical with an inner anode of radius ra and an outer
cathode of radius rc , giving an electric field
V
EðrÞ ¼ ð4:24Þ
rlnðrc =ra Þ
at a radial distance r from the centre of the anode wire. The output signal is
proportional to the number of ions formed and hence the energy deposited by the
radiation, but is independent of the applied voltage. However, the signal is very
small compared with the noise of all but the slowest electronic circuits and
requires considerable amplification to be useful. Overall, the energy resolution and
the time resolution of the chamber are relatively poor and ionization chambers are
of very limited use in recording individual pulses. They are used, for example, as
beam monitors, where the particle flux is very large, and in medical environments
to calibrate radioactive sources.
As mentioned previously, neutrons cannot be directly detected by ionization
methods, but neutron flux measurements can be made with ionization chambers
(or proportional chambers – see below) filled with BF3 by utilizing the neutron
activation reactions of Section 4.3.1.

Proportional counters

If the voltage is increase beyond the region of operation of the ionization chamber,
we move into the proportional region. In this region, a cylindrical arrangement
as used in the ionization chamber will produce electric field strengths of the order
of 104 –105 V/cm near the wire and this is strong enough for electron–ion pairs
released in the primary ionization to gain sufficient energy to cause secondary
ionization. The rapid increase in amplification due to secondary ionization is called
a Townsend avalanche. The output signal at the anode is still proportional to the
134 CH4 EXPERIMENTAL METHODS

energy lost by the original particle. There are a number of different types of device
working in the proportional region and they are sometimes generically referred to
as track chambers or wire chambers.
The earliest detector using this idea was the proportional counter, which
consists of a cylindrical tube filled with gas (again a quenching component in
the gas is required) and maintained at a negative potential, and a fine central anode
wire at a positive potential. Again, neutrons can be detected indirectly by using the
direct nuclear reaction 3 Heðn; pÞ3 H mentioned in Section 4.3.2 in a proportional
chamber filled with a mixture of 3 He and krypton. Subsequently, the resolution of
proportional counters was greatly improved as a result of the discovery that if
many anode wires were arranged in a plane between a common pair of cathode
plates, each wire acts as an independent detector. This device is called a multiwire
proportional chamber (MWPC), and was introduced in 1968.11 An MWPC can
achieve spatial resolutions of 200 mm or less, and has a typical time resolution of
about 3 ns.
A schematic diagram of an MWPC is shown in Figure 4.13. The planes (a) have
anode wires into the page and those in plane (b) are at right angles. The wire
spacings are typically 2 mm. The cathodes are the faces of the chambers. A
positive voltage applied to the anode wires generates a field as shown in the upper
corner. A particle crossing the chamber ionizes the gas and the electrons drift along
the field lines to the anode wires. In this particular example, there would be signals
from one wire in the upper (a) chamber and two in the lower (a) chamber.

Figure 4.13 A group of three planes of an MWPC (from Po00 with kind permission of Springer
Science and Business Media)

11
The MWPC was invented by Georges Charpak and for this and other developments in particle detectors he
was awarded the 1992 Nobel Prize in Physics.
 PARTICLE DETECTORS 135

Even better spatial resolutions are obtained in a related device called a drift
chamber, which has now largely replaced the MWPC as a general detector.12 This
uses the fact that the liberated electrons take time to drift from their point of
production to the anode. Thus the time delay between the passage of a charged
particle through the chamber and the creation of a pulse at the anode is related to
the distance between the particle trajectory and the anode wire. In practice,
additional wires are incorporated to provide a relatively constant electric field in
each cell in a direction transverse to normal incidence. A reference time has to be
defined, which, for example, could be done by allowing the particle to pass through
a scintillator positioned elsewhere in the experiment (scintillation counters are
discussed in Section 4.4.2). The electrons drift for a time and are then collected at
the anode, thus providing a signal that the particle has passed. If the drift time
can be measured accurately (to within a few ns) and if the drift velocity is known,
then spatial resolutions of 100–200 mm can easily be achieved, and specialized
detectors can reduce this still further.
Drift chambers are constructed in a variety of geometries to suit the nature of the
experiment, and arrangements where the wires are in planar, radial or cylindrical
configurations have all been used. The latter type is also called a ‘jet chamber’ and
a two-jet event in a jet chamber was shown in Figure 3.5 as evidence for the
existence of quarks.
One of the most advanced applications of proportional and drift chamber
principles is embodied in the time projection chamber (TPC) illustrated sche-
matically in Figure 4.14. This device consists of a cylindrical barrel, typically

Figure 4.14 Schematic diagram of a time projection chamber (TPC) (adapted from Kl86,
copyright Cambridge University Press)

12
In the new generation of colliders, drift chambers are largely being replaced by detectors based on silicon.
136 CH4 EXPERIMENTAL METHODS

2 m long and 1 m in diameter, surrounding the beam pipe of a collider. At each

end of the chamber is a segmented layer of proportional counters. The electric
drift field E, due to a negative high-voltage electrode plane at the centre of
the chamber, and a strong magnetic field B are aligned parallel and anti-parallel
to the axis of the cylinder. Because of this, the Lorentz forces on the drifting
electrons vanish and electrons formed along the track of an ionizing particle
emerging from the interaction point at the centre of the barrel, drift towards
one of the endcaps along helical trajectories whose direction is parallel to
the axis of the barrel. Their locations are measured by a set of anode wires
located between rectangular cathodes in the endcaps. The remaining third
coordinate necessary to reconstruct the position of a point on the track
is found from the time it takes for the electrons to drift from the point of
production to the endcaps where they are detected. The TPC has excellent spatial
resolution.
Recently a more robust form of chamber has evolved, in which the wires are
replaced by conductive metal strips on a printed circuit board. This is called a
microstrip gas chamber (MSGC) and is being incorporated in experiments being
designed for the new generation of accelerators currently planned or under
construction.

Beyond the region of proportionality

Referring again to Figure 4.12, by increasing the external voltage still further one
moves into a region where the output signal ceases to be proportional to the
number of ion pairs produced and hence the incident energy. This is the region of
limited proportionality. In this region a type of gas detector called a streamer tube
operates, but this will not be discussed here. Eventually the process runs out of
control and we enter the Geiger–Müller region where the output signal is
independent of the energy lost by the incident particle. In this region a quenching
agent is not used. Detectors working in this region are called Geiger–Müller
counters. Physically they are similar to the simple cylindrical proportional counter
and are widely used as portable radiation monitors in the context of safety
regulations.
For completeness, we can mention that if the gas amplification factor is taken
beyond the Geiger–Müller region, the avalanche develops moving plasmas or
streamers. Recombination of ions then leads to visible light which can be made to
generate an electrical output. Eventually complete breakdown occurs and a spark
is emitted as the incident particle traverses the gas. Detectors in this region, called
streamer and spark chambers (these were of parallel plate construction, rather than
cylindrical), were widely used in the 1970s and 1980s and played an important role
in hadron physics, but are no longer in general use.
 PARTICLE DETECTORS 137

4.4.2 Scintillation counters

For charged particles we have seen that energy losses occur due to excitation and
ionization of atomic electrons in the medium of the detector. In suitable materials,
called scintillators, a small fraction of the excitation energy re-emerges as visible
light (or sometimes in the UV region) during de-excitation. In a scintillation
counter this light passes down the scintillator and onto the face of a photodetector –
a device that converts a weak photon signal to a detectable electric impulse. An
important example of a photodetector is the photomultiplier tube, a schematic
diagram of which is shown in Figure 4.15.

Figure 4.15 Schematic diagram of the main elements of a photomultiplier tube (adapted from
Kr88, copyright John Wiley & Sons)

Electrons are emitted from the cathode of the photomultiplier by the photo-
electric effect and strike a series of focusing dynodes. These amplify the electrons
by secondary emission at each dynode and accelerate the particles to the next
stage. The final signal is extracted from the anode at the end of the tube. The
electronic pulse can be shorter than 10 ns if the scintillator has a short decay time.
The scintillation counter is thus an ideal timing device and it is widely used for
‘triggering’ other detectors, i.e. its signal is used to decide whether or not to
activate other parts of the detector, and whether to record information from the
event. Commonly used scintillators are inorganic single crystals (e.g. caesium
iodide) or organic liquids and plastics, and a modern complex detector in particle
physics may use several tons of detector in combination with thousands of
photomultiplier tubes.13 The robust and simple nature of the scintillation counter

13
For example, the Super Kamiokande experiment mentioned in Chapter 3, which detected neutrino
oscillations, although not using scintillation counters, has 13 000 photomultiplier tubes.
138 CH4 EXPERIMENTAL METHODS

has made it a mainstay of experimental nuclear and particle physics since the
earliest days of the subject.
Just as direct detection of neutrons is not possible by ionization methods, so the
same is true using scintillators. However, the -particle and the 3 H nucleus from
the direct nuclear reaction 6 Liðn; Þ3 H mentioned in Section 4.3.2 can produce
light in a LiI crystal scintillator and forms the basis for detecting neutrons with
energies up to about 20 MeV.

4.4.3 Semiconductor detectors

Solid-state detectors operate through the promotion of electrons from the valence
band of a solid to the conduction band as a result of the entry of the incident
particle into the solid. The resulting absence of an electron in the valence band
(a ‘hole’) behaves like a positron. Semiconductor detectors are essentially solid-
state ionization chambers with the electron–hole pairs playing the role of electron–
ion pairs in gas detectors. In the presence of an electric field, the electrons and
holes separate and collect at the electrodes, giving a signal proportional to the
energy loss of the incident charged particle. Most semiconductor detectors use
the principle of the junction diode. Since the band gap in some solids is as small as
1 eV and the energy loss required to produce a pair is only 3–4 eV on average
(cf. the 30 eV required in a gas detector), a very large number of electron–hole
pairs with only a small statistical fluctuation will be produced by a low-energy
particle. Solid-state detectors are therefore very useful in detecting low-energy
particles. Semiconductors (principally silicon or germanium) are used as a
compromise between materials that have residual conductivity sufficient to enable
conduction pulses due to single particles to be distinguished above background and
those in which the charges carriers are not rapidly trapped in impurities in the
material.
Such detectors have long been used in nuclear physics, where, for example, their
excellent energy resolution and linearity, plus their small size and consequent fast
response time, make them ideal detectors for -ray spectroscopy. Only recently
have thin planar detectors become important in particle physics, because of the
expense of covering large areas. Nevertheless, several square metres of semicon-
ductor detector are being planned for experiments at the LHC.
One example of a solid-state detector is a silicon microstrip detector, where
narrow strips of active detector are etched onto a thin slice of silicon, with gaps of
the order of l0 mm, to give a tiny analogue of an MWPC. Arrays of such strips
can then be used to form detectors with resolutions of the order of 5 mm. These
are often placed close to the interaction vertex in a colliding beam experiment,
with a view to studying events involving the decay of very short-lived particles.
Another example is the pixel detector. A single-plane strip detector only gives
position information in one dimension (orthogonal to the strip). A pixel detector
improves on this by giving information in two dimensions from a single plane.
 PARTICLE DETECTORS 139

Solid-state ‘vertex detectors’ are becoming increasingly important in particle

physics and have been incorporated in several of the multi-component detectors
designed for use in the new generation of colliders. Their main advantage is their
superb spatial resolution; a disadvantage is their limited ability to withstand
radiation damage.

4.4.4 Particle identification

Methods of identifying particles are usually based on determining the mass of the
particle by a simultaneous measurement of its momentum together with some
other quantity. At low values of ¼ E=mc2 , measurements of the rate of energy
loss dE=dx can be used, while muons may be characterized by their unique
penetrating power in matter, as we have already seen. Here we concentrate on
methods based on measuring the velocity or energy, assuming always that the
momentum is known. We thus need to start with explaining how momenta are
measured.

Measurement of momentum

The momentum of a charged particle is usually determined from the curvature of

its track in an applied magnetic field. It is common practice to enclose track
chambers in a magnetic field to perform momentum analysis. An apparatus that is
dedicated to measuring momentum is called a spectrometer. It consists of a magnet
and a series of detectors to track the passage of the particles. The precise design
depends on the nature of the experiment being undertaken. For example, in a
fixed-target experiment at high energies, the reaction products are usually
concentrated in a narrow cone around the initial beam direction, whereas in
colliding-beam experiments spectrometers must completely surround the interac-
tion region to obtain full angular coverage.
Magnet designs vary. Dipole magnets typically have their field perpendicular to
the beam direction. They have their best momentum resolution for particles
emitted forward and backward with respect to the beam direction, and are often
used in fixed-target experiments at high energies. However, the beam will be
deflected, and so at colliders this must be compensated for elsewhere to keep the
particles in orbit. Compensating magnets are present in the ‘layered detectors’
shown in Section 4.6 below. At colliders, the most usual magnet shape is the
solenoid, where the field lines are essentially parallel to the beam direction. This
device is used in conjunction with cylindrical tracking detectors, like jet chambers,
and has its best momentum resolution for particles perpendicular to the beam
direction.
We now turn to methods of measuring velocity.
140 CH4 EXPERIMENTAL METHODS

Time-of-flight

The simplest method, in principle, is to measure the time of flight between, for
example, two scintillation counters. If the distance between them is L, the time
difference for two particles of masses m1 and m2 travelling with velocities v1 and
v2 , is
 
L 1 1
t ¼ t2  t1 ¼  ; ð4:25Þ
c 1 2

where   v=c. For a common momentum p, Equation (4.25) may be written,

using14 E ¼ pc=,
L h
1=2
2 4 1=2 i
t ¼ 2 m22 c4 þ p2 c2  m1 c þ p2 c2 : ð4:26Þ
pc

We are interested in the situation where m2  m1  m and v2  v1  v. In this

case, setting m  m2  m1 , the non-relativistic limit of Equation (4.26) is
t ¼ tm=m and using v ¼ L=t, we have
m c
¼ t : ð4:27Þ
m L
Thus, for example, taking typical values of  ¼ 0:2, L ¼ 100 cm and t ¼
2
1010 s (assuming the timing is done using a scintillation counter), m=m
can be determined to about 1 per cent for low-energy particles. This method is
used, for example, in nuclear physics experiments using very low-energy neutron
beams.
However, since all high-energy particles have velocities close to the speed of
light, the method ceases to be useful for even quite moderate momenta. This can
been seen by taking the relativistic limit of Equation (4.26), when we have
Lc
2 
t  2
m2  m21 ; ð4:28Þ
2p

which for m2  m1  m, v2  v1  c, and using p ¼ mc becomes

m 2 ct
¼ : ð4:29Þ
m L
For example, using our previous values for L and t, Equation (4.29) shows that
the method is not useful for values of above about three, which corresponds to a
momentum of only about 3 GeV/c for nucleons. Of course this could be extended
by taking longer flight paths, but only at greater expense in instrumentation.

14
See Appendix B.
 PARTICLE DETECTORS 141

C̆erenkov counters

The most important identification method for high-energy particles is based on the
 erenkov effect. When a charged particle with velocity v traverses a dispersive
C
medium of refractive index n, excited atoms in the vicinity of the particle become
polarized, and if v is greater than the speed of light in the medium c=n, a part of the
excitation energy reappears as coherent radiation emitted at a characteristic angle 
to the direction of motion. The necessary condition v > c=n implies n > 1 and by
considering how the waveform is produced15 it can be shown that cos  ¼ 1=n for
the angle , where  ¼ v=c as usual. A determination of  is thus a direct
measurement of the velocity.16
 erenkov radiation appears as a continuous spectrum and may be collected onto
C
a photosensitive detector. Its main limitation from the point of view of particle
detection is that very few photons are produced. The number of photons NðÞd
radiated per unit path length in a wavelength interval d can be shown to be
   
1 d 1 d
NðÞd ¼ 2 1  2 2 < 2 1  2 ð4:30Þ
 n 2 n 2

and so vanishes rapidly as the refractive index approaches unity. The maximum
value occurs for  ¼ 1, which for a particle with unit charge, corresponds to about
200 photons/cm in the visible region in water and glass. These numbers should be
compared with the 104 photons/cm emitted by a typical scintillator. Because the
yield is so small, appreciable lengths are needed to give enough photons, and gas
 erenkov counters in particular can be several metres long.
C
 erenkov counters are used in two different modes. The first is as a threshold
C
counter to detect the presence of particles whose velocities exceed some minimum
value. Suppose that two particles with  values 1 and 2 at some given
momentum p are to be distinguished. If a medium can be found such that
1 n > 1  2 n, then particle 1 will produce C  erenkov radiation but particle 2
will not. Clearly, to distinguish between highly relativistic particles with 1
also requires n  1, so that from Equation (4.30) very few photons are produced.
Nevertheless, common charged particles can be distinguished in this way up to at
least 30 GeV/c.
Another device is the so-called ring-image C  erenkov detector which is a very
important device in both fixed-target machines and colliders. If we assume that the
particles are not all travelling parallel to a fixed axis, then the radiating medium
can be contained within two concentric spherical surfaces of radii R and 2R
centred on the target or interaction region where the particles are produced, as
illustrated in Figure 4.16. The outer surface is lined with a mirror, which focuses

15
This is Huygens’ construction in optics.
16  erenkov, Ilya Frank and Igor Tamm were
For the discovery and interpretation of this effect, Pavel C
awarded the 1958 Nobel Prize in Physics.
142 CH4 EXPERIMENTAL METHODS

Figure 4.16 Two particles P1 and P2 , produced from the target T, emit Cerenkov radiation on
traversing a medium contained between two spheres of radius R and 2R. The mirror M on the
outer sphere focuses the radiation into ring images at aa0 and bb0 on the inner detector sphere D.
The radii of the ring images depend on the angle of emission of the Cerenkov radiation and hence
on the velocities of the particles

 erenkov radiation into a ring at the inner detector surface. The radius of this
the C
ring depends on the angle  at which the C  erenkov radiation is emitted, and hence
on the particle velocity v. It is determined by constructing an image of the ring
electronically. This was the technique used in the Super Kamiokande detector
discussed in Chapter 3 to detect relativistic electrons and muons produced by
neutrino interactions. In that experiment the radiating medium was pure water.

4.4.5 Calorimeters

Calorimeters are an important class of detector used for measuring the energy and
position of a particle by its total absorption and are widely used. They differ from
most other detectors in that (1) the nature of the particle is changed by the detector,
and (2) they can detect neutral as well as charged particles. A calorimeter may be a
homogeneous absorber/detector to detect photons and electrons. In early devices this
was often a block of lead glass, but is now more likely to be scintillator such as CsI.
Alternatively, it can be a sandwich construction with separate layers of absorber (e.g.
a metal such as lead) and detector (scintillator, MWPC etc.). The latter are also
 PARTICLE DETECTORS 143

known as ‘sampling calorimeters’. During the absorption process, the particle will
interact with the material of the absorber, generating secondary particles which will
themselves generate further particles and so on, so that a cascade or shower,
develops. For this reason calorimeters are also called ‘shower counters’.
The shower is predominantly in the longitudinal direction due to momentum
conservation, but will be subject to some transverse spreading due both to multiple
Coulomb scattering and the transverse momentum of the produced particles.
Eventually all, or almost all, of the primary energy is deposited in the calorimeter,
and gives a signal in the detector part of the device.
There are several reasons why calorimeters are important, especially at high
energies:

 they can detect neutral particles, by detecting the charged secondaries;

 the absorption process is statistical (and governed by the Poisson distribution),

1
so that the relative precision of energy measurements E=E varies as E2 for
large E, which is a great improvement on high-energy spectrometers where
E=E varies as E2 ;

 the signal produced can be very fast, of the order of (10–100) ns, and is ideal for
making triggering decisions.

Although it is possible to build calorimeters that preferentially detect just one class
of particle (electrons and photons, or hadrons) it is also possible to design detectors
that serve both purposes. Since the characteristics of electromagnetic and hadronic
showers are somewhat different it is convenient to describe each separately. In
practice, it is common to have both types in one experiment with the hadron
calorimeter stacked behind the electromagnetic one.

Electromagnetic showers

When a high-energy electron or positron interacts with matter we have seen that
the dominant energy loss is due to bremsstrahlung, and for the photons produced
the dominant absorption process is pair production. Thus the initial electron will,
via these two processes, lead to a cascade of e pairs and photons, and this will
continue until the energies of the secondary electrons fall below the critical energy
EC where ionization losses equal those from bremsstrahlung. This energy is
roughly given by EC  600 MeV=Z.
Most of the correct qualitative features of shower development may be obtained
from the following very simple model. We assume:

 each electron with E > EC travels one radiation length and then gives up half of
its energy to a bremsstrahlung photon;
144 CH4 EXPERIMENTAL METHODS

 each photon with E > EC travels one radiation length and then creates an
electron–positron pair with each particle having half the energy of the photon;

 electrons with E < EC cease to radiate and lose the rest of their energy by
collisions;

 ionization losses are negligible for E > EC.

A schematic diagram of the approximate development of a shower in an

electromagnetic calorimeter assuming this simple model is shown in Figure 4.17.
If the initial electron has energy E0 EC, then after t radiation lengths the
shower will contain 2t particles, which consist of approximately equal numbers of
electrons, positrons and photons each with an average energy
EðtÞ ¼ E0 =2t : ð4:31Þ
The multiplication process will cease abruptly when EðtÞ ¼ EC , i.e. at t ¼ tmax where
lnðE0 =EC Þ
tmax ¼ tðEC Þ  ð4:32Þ
ln2

Figure 4.17 Approximate development of an electromagnetic shower in a sampling calorimeter

assuming the simple model described in the text; the calorimeter consists of alternate layers of lead
(Pb) and a scintillator (Sc), the latter attached to photomultipliers (one only shown)
 LAYERED DETECTORS 145

and the number of particles at this point will be

Nmax ¼ exp½tmax ln2 ¼ E0 =EC : ð4:33Þ

The main features of this simple model are observed experimentally, and in
particular the maximum shower depth increases only logarithmically with primary
energy. Because of this, the physical sizes of calorimeters need increase only
slowly with the maximum energies of the particles to be detected. The energy
resolution of a calorimeter, however, depends on statistical fluctuations, which are
neglected in this simple model, but for an electromagnetic calorimeter it is
1
typically E=E  0:05=E2, where E is measured in GeV.

Hadronic showers

Although hadronic showers are qualitatively similar to electromagnetic ones,

shower development is far more complex because many different processes
contribute to the inelastic production of secondary hadrons. The scale of the
shower is determined by the nuclear absorption length defined earlier. Since this
absorption length is larger than the radiation length, which controls the scale of
electromagnetic showers, hadron calorimeters are thicker devices than electro-
magnetic ones. Another difference is that some of the contributions to the total
absorption may not give rise to an observable signal in the detector. Examples are
nuclear excitation and leakage of secondary muons and neutrinos from the
calorimeter. The loss of ‘visible’ or measured energy for hadrons is typically
20–30 per cent greater than for electrons.
The energy resolution of calorimeters is in general much worse for hadrons
than for electrons and photons because of the greater fluctuations in the develop-
ment of the hadron shower. Depending on the proportion of 0 s produced in the
early stages of the cascade, the shower may develop predominantly as an
electromagnetic one because of the decay 0 ! . These various features lead
to an energy resolution typically a factor of 5–10 poorer than in electromagnetic
calorimeters.

4.5 Layered Detectors

As stated earlier, in particle physics it is necessary to combine several detectors in
a single experiment to extract the maximum amount of information from it.
Typically, working out from the interaction region, there will be a series of wire
chambers, followed further out by calorimeters and at the outermost limits,
detectors for muons, the most penetrating particles to be detected. The whole
device is usually in a strong magnetic field so that momentum measurements may
be made. We will illustrate the general features by three examples.
146 CH4 EXPERIMENTAL METHODS

Figure 4.18 The CDF detector at the pp collider at Fermilab, USA (Fermilab Graphic, reproduced
with permission)

The first is the pp Collider Detector at Fermilab (CDF), which is shown
schematically in Figure 4.18. The detection of the top quark and the measurement
of its mass were first made using this device. The dashed lines indicate some
particles produced in the collision. CDF is a large device, being approximately 8 m
wide and 26 m in overall length. The beams of protons and antiprotons enter from
each end through focusing quadrupole magnets and interact in the central
intersection region where there is a silicon vertex detector (1) to detect very
short-lived particles. The intersection point is surrounded by a 2000 tonne detector
system which, in addition to the vertex detector, consists of inner drift chambers
(2), electromagnetic calorimeters (4), hadron calorimeters (5) time-of-flight
detectors (not indicated) and further drift chambers (2) on the outside to detect
muons. The whole system is in a magnetic field with the solenoid coil shown at
(7) and steel shielding at (6). The rest of the detector consists of two symmetrical
sets of drift chambers (2) sandwiched between scintillation counters (3) and
magnetic toroids (8) to provide momentum measurements, primarily for muons.
The second example, shown schematically in Figure 4.19, is the ATLAS
detector currently under construction for use at the Large Hadron Collider
(LHC). It is hoped that this and other detectors at the LHC will be able to detect
the important Higgs boson, if it exists, and so help solve one the outstanding
current problems in particle physics – the origin of mass. The ATLAS detector is
even larger than the CDF detector and measures about 22 m high and 44 m long,
with an overall weight of approximately 7000 tonnes.
Figure 4.19 The ATLAS detector under construction for use at the pp collider LHC (also under construction) at CERN, Geneva, Switzerland (CERN photo,
reproduced with permission)
148 CH4 EXPERIMENTAL METHODS

Figure 4.20 The STAR detector at the RHIC accelerator at Brookhaven National Laboratory,
USA. (Courtesy of Brookhaven National Laboratory)

Finally, Figure 4.20 shows the STAR detector at the RHIC accelerator at
Brookhaven National Laboratory. This detects events resulting from the collisions
of heavy ions, typically those of fully-stripped gold nuclei, where the final state
may contain many thousands of particles. An example of an event is shown in
Figure 9.12.

Problems
4.1 At a collider, a 20 GeV electron beam collides with a 300 GeV proton beam at a
crossing angle of 10 . Evaluate the total centre-of-mass energy and calculate what
beam energy would be required in a fixed-target electron machine to achieve the
same total centre-of-mass energy.

4.2 What is the length L of the longest drift tube in a linac which operating at a
frequency of f ¼ 20 MHz is capable of accelerating 12 C ions to a maximum energy
of E ¼ 100 MeV?

4.3 Alpha particles are accelerated in a cyclotron operating with a magnetic field of
magnitude B ¼ 0:8 T. If the extracted beam has an energy of 12 MeV, calculate the
extraction radius and the orbital frequency of the beam (the so-called cyclotron
frequency).
 PROBLEMS 149

4.4 Protons with momentum 50 GeV/c are deflected through a collimator slit 2 mm wide
by a bending magnet 1.5 m long which produces a field of 1.2 T. How far from the
magnet should the slit be placed so that it accepts particles with momenta in the
range 49–51 GeV/c?

4.5 Estimate the minimum length of a gas C  erenkov counter that could be used in
threshold mode to distinguish between charged pions and charged kaons with
momentum 20 GeV/c. Assume that a minimum of 200 photons need to be radiated
to ensure a high probability of detection. Assume also that the radiation covers the
whole visible spectrum between 400 nm and 700 nm and neglect the variation with
wavelength of the refractive index of the gas.

4.6 An eþ e collider has a diameter of 8 km and produces beams of energy 45 GeV.

Each beam consists of 12 bunches each containing 3
1011 particles. The bunches
have a cross-sectional area of 0:02 mm2 . What is the luminosity of the machine in
units of cm2 s1 ?

4.7 What are the experimental signatures and with what detectors would one measure:
(a) the decay Z ! b
b, and (b) W ! e and W ! .

4.8 The reaction eþ e !  þ   is studied using a collider with equal beam energies of
5 Gev. The differential cross-section is given by
d
2  h 2 c2

¼ 2
1 þ cos2 
d 4Ecm
where Ecm is the total centre-of-mass energy and  is the angle between the incoming
e and the outgoing   . If the detector can only record an event if the  þ   pair
makes an angle of at least 30 relative to the beam line, what fraction of events will
be recorded? What is the total cross-section for this reaction in nanobarns? If the
reaction is recorded for 107 s at a luminosity of L ¼ 1031 cm2 s1 , how many events
are expected?
Suppose the detector is of cylindrical construction and at increasing radii from the
beam line there is a drift chamber, an electromagnetic calorimeter, a hadronic
calorimeter and finally muon chambers. If in a particular event the   decays via
  !  þ  þ  and the  þ decays to  þ ! eþ þ  þ e , what signals would
be observed in the various parts of the detector?

4.9 A charged particle with speed v moves in a medium of refractive index n. By considering
the wavefronts emitted at two different times, derive a relation for the angle  of the
emitted C erenkov radiation relative to the particle’s direction in terms of  ¼ v=c and n.
What is the maximum angle of emission and to what limit does it correspond?
If the momentum p of the particle is known from other detectors, show that the
mass squared x of the particle is given by x ¼ ðmc2 Þ2 ¼ p2 c2 ðn2 cos2   1Þ. If the
error on the momentum is negligible, show, by taking derivatives of this expression,
that for very relativistic particles, the standard error
x on x is approximately
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

x  2p2 c2 ðn2  1Þ
 ;
where
 is the standard error on .
150 CH4 EXPERIMENTAL METHODS

4.10 Estimate the thickness of iron through which a beam of neutrinos with energy
300 GeV must travel if 1 in 109 of them is to interact. Assume that at high energies
the neutrino-nucleon total cross-section is given approximately by
  1038 E cm2
where E is given in GeV. The density of iron is  ¼ 7:9 g cm3 .

4.11 An electron with an initial energy of 2 GeV traverses 10 cm of water with a radiation
length of 36.1 cm. Calculate its final energy. How would the energy loss change if
the particle were a muon rather than an electron?

4.12 A beam of neutrons with kinetic energy 0.1 eV and intensity 106 s1 is incident
normally on a thin foil of 23592 U of effective thickness 10
1
kg m2 . The beam can
undergo (1) isotropic elastic scattering, with a cross section
el ¼ 3
102 b, (2)
radiative capture, with a cross-section
cap ¼ 102 b, or (3) it can fission a 235 92 U
nucleus, with a cross-section
f ¼ 3
102 b. Calculate the attenuation of the beam
and the flux of elastically-scattered particles 5 m from the foil.

4.13 A positron with laboratory energy 50 GeV interacts with the atomic electrons in a
lead target to produce þ  pairs. If the cross-section for this process is given by

¼ 4 2 h2 c2 =3ðECM Þ2 , calculate the positron’s interaction length. The density of
lead is  ¼ 1:14
107 kg m3 .

4.14 A liquid hydrogen target of volume 125 cm3 and density 0:071 g cm3 is bombarded
with a mono-energetic beam of negative pions with a flux 2
107 m2 s1 and the
reaction  þ p ! 0 þ n observed by detecting the photons from the decay of the
0 . Calculate the number of photons emitted from the target per second if the cross-
section is 40 mb.

4.15 Assuming the Bethe–Bloch formula is valid for low energies, show that the rate of
ionization has a maximum (the Bragg peak) and find the kinetic energy of protons in
iron for which this maximum would occur.

4.16 A cylindrical proportional chamber has a central anode wire of radius 0.02 mm and
an outer cathode of radius 10 mm with a voltage of 500 V applied between them.
What is the electric field at the surface of the anode? If the threshold for ionization
by collision is 750 kV m1 and the mean free path of the particles being detected is
4
106 m, estimate the number of ion pairs produced per primary particle.
5
Quark Dynamics:
the Strong Interaction

In Chapter 3 we described the basic properties of quarks and in particular their

static properties and how these are used to construct the quark model of hadrons.
We now look in more detail at how quarks interact and the role of gluons in the
strong interactions. Thus we will be considering dynamical properties and the
theoretical framework that describes these interactions.

5.1 Colour
We saw in Chapter 3 that the quark model account of the hadron spectrum is very
successful. However, it begs several questions. One is: why are the observed states
overwhelmingly of the form 3q, 3
q and q
q? Another arises from a particular
assumption that was implicitly made in Chapter 3. This is: if two quarks of the
same flavour uu, dd, ss . . . are in the same spatial state, they must also be in the
same spin state, with their spins parallel. This can be seen very easily by
considering the baryon state omega-minus 
that is shown in Table 3.3 and
Figure 3.12.1 From its decay products, it may be deduced that this state has
strangeness S ¼
3 and spin J ¼ 32 and thus in the quark model it has the
composition 
¼ sss, where all three quarks have their spins parallel and there
is no orbital angular momentum between them. This means that all three like-
quarks have the same space and spin states, i.e. the overall wavefunction must be
symmetric, which violates the fundamental requirement of the Pauli principle. The
latter states that a system of identical fermions has a wavefunction that is overall
antisymmetric under the interchange of any two particles, because identical

1
The discovery of the 
was a crucial step in gaining acceptance of the quark model of hadron
spectroscopy. The experiment is described in Chapter 15 of Tr75.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
152 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

fermions cannot simultaneously be in the same quantum state. The three s quarks
in the 
therefore cannot be in the same state. So how do they differ?
The 
is an obvious example of the contradiction, but it turns out that in order
for the predictions of the quark model to agree with the observed spectrum of
hadron multiplets, it is necessary to assume that overall baryon wavefunctions are
symmetric under the interchange of like quarks.2 In order to resolve this contra-
diction, it is necessary to assume that a new degree of freedom exists for quarks,
but not leptons, which is somewhat whimsically called colour. The basic properties
of colour are as follows.

1. Any quark u, d, s, . . . can exist in three different colour states.3 We shall see
later that there is direct experimental evidence that just three such states exist,
which we denote r, g, b for ‘red’, ‘green’ and ‘blue’ respectively.

2. Each of these states is characterized by the values of two conserved colour

charges, denoted I3C and Y C , which are strong interaction analogues of the
electric charge in electromagnetic interactions.4 These charges depend only on
the colour states r, g, b and not on the flavours u, d, s, . . . The particular values
for quarks and antiquarks are given in Table 5.1, and are a consequence of a
fundamental symmetry of the strong interaction (called SU(3) colour symme-
try), which we will not pursue here. For multiparticle states, the colour charges
of the individual states are simply added.

3. Only states with zero values for the colour charges are observable as free
particles; these are called colour singlets. This is the hypothesis of colour
confinement. It can be derived, at least approximately, from the theory of strong
interactions we shall describe.

Table 5.1 Values of the colour charges I3C and Y C for the
colour states of quarks and antiquarks
(a) Quarks (b) Antiquarks

I3C YC I3C YC
1 1
r 2 3
r
12
13
g
12 1
3

g 1
2
13
b 0
23

b 0 2
3

2
In Problem 3.4 it was shown explicitly that otherwise the wrong hadron spectrum is predicted.
3
Needless to say, nothing to do with ‘real’ colour!
4
This is one reason we were careful to use the qualifier ‘electric’ when talking about charge in the context of
electromagnetic interactions in earlier chapters.
 QUANTUM CHROMODYNAMICS (QCD) 153

Returning to the quark model, it can be seen from Table 5.1 that a 3q state
can only have both I3C ¼ 0 and Y C ¼ 0 if we have one quark in an r state, one in a
g state and one in a b state. Hence in the 
, for example, all three s quarks
are necessarily in different colour states, and thus the Pauli principle can be
satisfied. Formally, we are assuming that the total wavefunction is the product of a
spatial part spatial ðxÞ and a spin part spin , as usual, but also a colour wavefunction
colour , i.e.

¼ spatial ðxÞ spin colour  ð5:1Þ

The Pauli principle is now interpreted as applying to the total wavefunction

including the colour part colour . The combined space and spin wavefunctions can
then be symmetric under the interchange of quarks of the same flavour (to agree
with experiment) provided the colour wavefunction is antisymmetric. The struc-
ture of colour is

1
colour ¼ pffiffiffi ½R1 G2 B3 þ G1 B2 R3 þ B1 R2 G3
R1 B2 G3
B1 G2 R3
G1 R2 B3 ;
6
ð5:2Þ

where R, G and B represent quarks with colour red, green and blue, respectively.
One can also see from Table 5.1 part of the answer to the first question of this
section. Free quarks and fractionally charged combinations like qq and qq
q are
forbidden by colour confinement, in accordance with experimental observation. On
the other hand, the combinations q
q and 3q used in the simple quark model are
allowed. More unusual combinations like qq
q
q and qqqq
q, which could give rise to
so-called ‘exotic’ mesons and baryons, respectively, are not in principle forbidden
by colour confinement and, as mentioned in Chapter 3, recent experiments may
possibly have provided some evidence for a small number of these, but this has yet
to be confirmed.

5.2 Quantum Chromodynamics (QCD)

The theory that describes strong interactions in the standard model is called
quantum chromodynamics, or QCD for short (chromos means colour in Greek).
Although QCD is not tested to the same extent or precision as quantum
electrodynamics (QED), the quantum theory of electromagnetic interactions, it
is nevertheless in impressive agreement with a large body of experimental data.
QCD is similar to QED in that both describe interactions that are mediated by
massless spin-1 bosons; gluons in the former case and photons in the latter. Both
theories are of the type called gauge theories which, as mentioned in Chapter 1,
refer to a particular symmetry of the theory. However, there is a very important
difference in the two interactions that we now discuss.
154 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

Gluons, the force carriers of the strong interaction, have zero electric charge,
like photons, but unlike photons, which couple to electric charge, gluons couple to
colour charges. This leads immediately to the flavour independence of strong
interactions discussed in Chapter 3; that is, the different quark flavours a ¼ u,
d, s, c, b and t have identical strong interactions. We now see that this is because
they are postulated to exist in the same three colour states r, g, b, with the
same possible values of the colour charges. Flavour independence has its most
striking consequences for u and d quarks, which have almost equal masses, where
it leads to the phenomenon of isospin symmetry. This results, among other
things, in the near equality of the masses of the proton and neutron, and charge
states within other multiplets such as pions and kaons, all of which we have
seen in Chapter 3 are confirmed by experiment. We will examine the con-
sequence of flavour independence for the bound states of the heavy quarks c
and b in Section 5.3.
Although QED and QCD both describe interactions, albeit of very different
strengths, that are mediated by massless spin-1 bosons which couple to conserved
charges, there is a crucial difference between them that profoundly effects
the characters of the resulting forces. While the photons which couple to the
electric charge are themselves electrically neutral, gluons have non-zero values of
the colour charges to which they couple. This is illustrated in Figure 5.1, which
shows a particular example of a quark–quark interaction by gluon exchange.

Figure 5.1 Example of quark--quark scattering by gluon exchange; in this diagram, the quark
flavours u and s are unchanged, but their colour states can change, as shown

In this diagram, the colour states of the two quarks are interchanged, and the
gluon has non-zero values of the colour quantum numbers, whose values follow
from colour charge conservation at the vertices, i.e.

1
I3C ðgÞ ¼ I3C ðrÞ
I3C ðbÞ ¼ ð5:3Þ
2

and

Y C ðgÞ ¼ Y C ðrÞ
Y C ðbÞ ¼ 1: ð5:4Þ

 QUANTUM CHROMODYNAMICS (QCD) 155

Figure 5.2 The two lowest-order contributions to gluon--gluon scattering in QCD: (a) one-
gluon exchange, (b) contact interaction

Just as quarks can exist in three colour states, gluons can exist in eight colour
states, although we will not need the details of these. The first thing implied by the
non-zero values of the colour charges is that gluons, like quarks, are confined and
cannot be observed as free particles. The second is that since gluons couple to
particles with non-zero colour charges, and since gluons themselves have non-zero
colour charges, it follows that gluons couple to other gluons. The two types of
gluon self-coupling that occur in QCD are given in Figure 5.2, which shows the
two lowest-order contributions to gluon–gluon scattering.
The first is a gluon exchange process in analogy to gluon exchange in quark–
quark scattering, which we have encountered previously (see Figure 1.3), while the
second involves a so-called ‘zero range’ or ‘contact’ interaction. If the forces
resulting from these interactions were attractive and sufficiently strong, they could
in principle lead to bound states of two or more gluons. These would be a new type
of exotic state called glueballs. Although some experiments claimed to have
detected these, at present there is little compelling evidence that they exist.5
The gluon–gluon interactions have no analogue in QED (photons couple to
electrically charged particles and hence do not couple directly to other photons)
and it can be shown that they lead to properties of the strong interaction that differ
markedly from those of the electromagnetic interaction. These properties are
colour confinement, which we have discussed above, and a new property called
asymptotic freedom. The latter is the statement that the strong interaction gets
weaker at short distances; conversely, as the distance between the quarks increases,
the interaction gets stronger.6 In this strong interaction regime the situation is very
complicated, and it has not yet been possible to evaluate the theory precisely. We
therefore have to rely on results obtained by numerical simulations of the theory;
the approach is called lattice gauge theory. In these simulations, the theory is

5
A critical review is given in Ei04.
6
Asymptotic freedom was postulated in 1973 by David Gross, David Politzer and Frank Wilczek, who were
subsequently awarded the 2004 Nobel Prize in Physics.
156 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

evaluated at a grid of discrete points on a three-dimensional lattice and by making

the lattice spacing small enough it is hoped that the results of the true continuum
theory will be approximated. The calculations require very large ultra-fast
computers and precise results are difficult to obtain because of the approximations
that have to be made. Nevertheless, at present, the demonstration of confinement in
QCD rests largely on such simulations.7

5.3 Heavy Quark Bound States

Some of the features of QCD discussed above are illustrated by considering the
static potential between a heavy quark and an antiquark. Such systems give rise
to bound states and because the quarks are so heavy they move slowly enough
within the resulting hadrons to be treated non-relativistically to a first approxima-
tion. (This is one of the few places in particle physics where a non-relativistic
calculation is adequate.) This means that the rest energies of the bound states, and
hence their masses, can be calculated from the static potential between the quarks
in exactly the same way that the energy levels in the hydrogen atom are calculated,
although of course the potential is not Coulombic. In the present case, however, the
procedure is reversed, with the aim of determining the form of the static potential
from the rather precisely measured energies of the bound states.
The first such state to be discovered, the J= ð3097Þ8, is a bound state of the c
c
system and is part of a family of such states given the name charmonium, by
analogy with positronium, the bound state of an electron and a positron. It is
identified with the n ¼ 1, 3 S1 state of the c
c system, where n is the principal
quantum number and we use the notation 2Sþ1 LJ , with ðL, SÞ the angular
momentum between the quarks and their total spin, respectively. The discovery
of the J= ð3097Þ caused considerable excitement because it confirmed the
existence of the charm quantum number that had been predicted many years
earlier, even though the J= ð3097Þ itself has zero overall charm. It was hence a
very important piece of evidence in favour of the standard model.
The interpretation of the J= ð3097Þ as a c
c bound state follows from its
unusually narrow width. For a state decaying predominantly (86 per cent) to
hadrons (mostly pions) by the strong interaction one would expect a width
measured in MeV, whereas the width of the J= ð3097Þ was only about 90 keV.
This meant that there was no possibility of an explanation in terms of just u, d and
s quarks. The preferred decay of the J= ð3097Þ would be via the mechanism
shown in Figure 5.3(a). However, this is forbidden by energy conservation because

7
Lattice calculations also support the view that gluon–gluon forces are strong enough to give rise to
glueballs.
8
The rather clumsy notation is because it was discovered independently by two groups, led by Burton Richer
and Samuel Ting. Richer’s group was studying the reactions eþ e
! hadrons and named it the particle.
Ting’s group discovered it in pBe reactions and called it the J. It is now known as the J= . Richer and Ting
shared the 1976 Nobel Prize in Physics for the discovery.
 HEAVY QUARK BOUND STATES 157

Figure 5.3 Quark diagrams for (a) the decay of a charmonium state to a pair of charmed
mesons, and (b) an example of a decay to non-charmed mesons

MJ= < 2MD , where MD is the mass of the lightest meson having non-zero charm,
the Dð1870Þ. (These latter states had already been seen in neutrino experiments,
but not clearly identified.) The mass 2MD is referred to as the charm threshold.
Since the direct decay to charmed mesons is forbidden, the only hadronic decays
allowed must proceed via mechanisms such as that of Figure 5.3(b) and diagrams
like this where initial and final quark lines are disconnected are known to be
heavily suppressed.9
The explanation for this in QCD is that since both the decaying particle and
the three pions in the final state are colour singlets, they can only be connected by
the exchange of a combination of gluons that is also a colour singlet, i.e. not the
exchange of a single gluon. Moreover, the J= ð3097Þ is known to be produced in
eþ e
annihilations via photon exchange, so it must have a charge conjugation
C ¼
1. Thus the minimum number of gluons exchanged is three. This is
illustrated in Figure 5.4. In contrast, if M > 2MD then the decay may proceed
via the exchange of low-momentum gluons as usual.
Subsequently, higher-mass charmonium states also with J PC ¼ 1

, where
P ¼ ð
1ÞLþ1 and C ¼ ð
1ÞLþS , were discovered in eþ e
reactions and states
with other J PC values were identified in their radiative decays. Thus the n ¼ 1; 1 S0

Figure 5.4 OZI-suppressed decay of a charmonium state below the DD
threshold

9
This is known as the OZI Rule after Okubo, Zweig and Iizuka who first formulated it. Another example

.
where it acts is the suppression of the decay  ! þ 
0 compared with  ! K K
158 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

ground state c ð2980Þ has been found from the decays

ð3686Þ ! c ð2980Þ þ
and J= ð3097Þ ! c ð2980Þ þ
ð5:5Þ

and a series of states ci ði ¼ 1; 3Þ have been found in the decays

ð3686Þ ! ci þ
 ð5:6Þ

The latter themselves decay and from an analysis of their decay products they are
identified with the n ¼ 1 states 3 P0 , 3 P1 and 3 P2 . Some of these states lie below the
charm threshold and like the J= ð3097Þ are forbidden by energy conservation to
decay to final states with ‘open’ charm and thus have widths measured in keV.
Others lie above the charm threshold and therefore have ‘normal’ widths measure
in MeV. The present experimental situation for charmonium states with L  2 is
shown in Table 5.2.

Table 5.2 Predicted c c
and bb
states with L  2 and masses

up to and just above the charm and bottom thresholds
(3.74 GeV/c2 and 10.56 GeV/c2, respectively), compared with
experimentally observed states (masses are given in MeV/c2)

n2Sþ1 LJ J PC c
c state b

b state
11 S0 0
þ c ð2980Þ b ð9300Þ?
13 S1 1

J= ð3097Þ ð9460Þ
11 P1 1þ
hc ð3526Þ?
13 P0 0þþ c0 ð3415Þ b0 ð9860Þ
13 P1 1þþ c1 ð3511Þ b1 ð9893Þ
13 P2 2þþ c2 ð3556Þ b2 ð9913Þ
21 S0 0
þ c ð3654Þ?
23 S1 1

ð3686Þ ð10 023Þ
23 P0 0þþ b0 ð10 232Þ
23 P1 1þþ b1 ð10 255Þ
23 P2 2þþ b3 ð10 269Þ
33 S1 1

ð4040Þ ð10 355Þ
3


4 S1 1 ð4160Þ ð10 580Þ

Later experiments established a spectrum of bottomium states, i.e. bound states

of the b

b system. These are also shown in Table 5.2. By analogy with charmonium,
those bottomium states below the bottom threshold 2MB ¼ 10:56 GeV=c2 , where
MB is the mass of the lightest meson with non-zero beauty quantum number, have
widths measured in keV, whereas those above this threshold have ‘normal’ widths
expected of resonances decaying via the strong interaction
The charmonium and bottomium states with L  2 are shown in Figure 5.5 as
conventional energy level diagrams, where the energies are plotted relative to those
Figure 5.5 Energy levels of the charmonium (c c
) and bottomium (bb
) states for L  2; the masses are given in units of GeV=c2
160 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

of the 3 S1 ground states. There is a striking similarity in the levels of the two
systems, which suggests that the forces in the c
c and b
b are flavour independent, as
discussed in Chapter 3 and now seen to follow from the postulates of QCD. The
level structure is also very similar to that seen in positronium which suggests that,
as in positronium, there is a major contribution from a single-particle exchange
with the Coulomb-like form. In fact at short interquark distances r <  0:1 fm, the
interaction is dominated by one-gluon exchange that we can write as
a
VðrÞ ¼
; ð5:7Þ
r
where a is proportional to the strong interaction analogue of the fine structure
constant  in QED. Because of asymptotic freedom, the strength of the interaction
decreases with decreasing r, but for r < 0:1 fm this variation is slight and can in
many applications be neglected.10
In strong interactions we also have to take account of the fact that the quarks are
confined. The latter part of the potential cannot at present be calculated from QCD
and several forms are used in phenomenological applications. All reasonable forms
are found to give very similar results for the region of interest. If we choose a
linear form, then

VðrÞ b r: ð5:8Þ

This is an example of a confining potential, in that it does not die away with
increasing separation and the force between the quark and antiquark cannot be
neglected, even when they are very far apart. The full potential is thus
a
VðrÞ ¼
þ br: ð5:9Þ
r

If the form (5.9) is used in the Schrödinger equation for the c
c and b
b systems,
taking account of their different masses, it is found that a good fit to both sets of
energy levels can be obtained for the same values a 0:48 and b 0:18 GeV2 ,
which confirms the flavour independence of the strong interaction and is evidence
for QCD and the standard model.

5.4 The Strong Coupling Constant and Asymptotic Freedom

The strong interaction derives its name from the force that, among other things,
binds quarks into hadrons. However, some remarkable phenomena depend on the
fact that the interaction gets weaker at short distances; that is, on asymptotic

10
The equivalent coupling in QED also varies with distance, but the variation is very small and can usually be
neglected.
 THE STRONG COUPLING CONSTANT AND ASYMPTOTIC FREEDOM 161

freedom. Such short-distance interactions are associated with large momentum

transfers jqj between the particles, with jqj ¼ Oð
h=rÞ, where r is the distance at
which the interaction occurs. Hence in discussing scattering from a static potential,
like the one above, we can regard the strong coupling s as decreasing with
increasing momentum transfer, rather than with decreasing r.
In general, the strength of the interaction can be shown to depend on the squared
four-momentum transfer
Q2  Eq2 =c2
q2 ; ð5:10Þ
which was introduced in Chapter 2. Specifically, it can be shown that the QCD
coupling constant s is given to a good approximation by
12
s ¼ ; ð5:11Þ
ð33
2Nf Þ lnðQ2 =2 Þ

where Nf is the number of quark flavours u, d, s, . . . . , with 4m2q c4 < Q2 , and

Q2  2 . The constant  is a scale parameter that must be determined from
experiment. Thus QCD does not predict the absolute value of s , but rather its
dependence on Q2 . The value of  may be found by measuring the coupling
constant in a variety of processes (two of which will be discussed later in this
chapter) giving values consistent with
 ¼ 0:2  0:1 GeV=c: ð5:12Þ

Because s varies with Q2 , it is often referred to as the running coupling constant.

The values of s ðQ2 Þ corresponding to Equation (5.12) are plotted in Figure 5.6. The

Figure 5.6 The running coupling constant s corresponding to four flavours and a scale
parameter  ¼ 0:2  0:1 GeV/c; the dashed, solid and dot-dashed curves correspond to
 ¼ 0:1 ; 0:2 and 0:3, respectively
162 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

variation with Q2 is small at large Q2 and over limited Q2 regions it can often be
neglected. In this large Q2 region, the coupling is sufficiently weak that calcula-
tions can be performed with reasonable accuracy by retaining only diagrams of
lowest and next-to-lowest order; and sometimes the short-range strong interaction
can be neglected to a first approximation, as we shall see.
Although there are other forces that increase with increasing separation (for
example, the force between two particles connected by a spring or elastic string),
the difference between those and the present case is that in the former cases
eventually something happens (for example, the string breaks) so that the particles
(or the ends of the string) become free. This does not happen with the strong force.
Instead, the energy stored in the colour field increases until it becomes sufficiently
large to create q
q pairs and eventually combinations of these will appear as physical
hadrons. This latter process is called fragmentation and is rather poorly understood.
The behaviour of the strong interaction as a function of distance (or equivalently
momentum transfer) is so unlike the behaviour of other forces we are familiar with
(e.g. gravity and electromagnetism) that it is worth looking at why this is.
In QED, single electrons are considered to emit and reabsorb photons con-
tinually, as shown in Figure 5.7(a). Such a process is an example of a so-called
quantum fluctuation, i.e. one particle converting to two or more particles for a
finite time. This is allowed provided the time and the implied violation of energy
conservation are compatible with the uncertainty principle. Of course if another
electron is nearby, then it may absorb the photon and we have the usual one-photon
exchange scattering process of Figure 5.7(b).

Figure 5.7 (a) The simplest quantum fluctuation of an electron, and (b) the associated
exchange process

The emitted photon may itself be subject to quantum fluctuations, leading to more
complicated diagrams like those shown in Figure 5.8(a). Thus the initial electron
emits not only photons, but also indirectly electron–positron pairs. These are
referred to as a ‘sea’ of virtual electrons (cf. comments in Chapter 3 in the context
of the quark model). The equivalent contribution to elastic electron–electron
scattering is shown in Figure 5.8(b).
 THE STRONG COUPLING CONSTANT AND ASYMPTOTIC FREEDOM 163

Figure 5.8 (a) A more complicated quantum fluctuation of the electron, and (b) the
associated exchange process

These virtual processes are collectively referred to as vacuum polarization

effects.11 The production of virtual eþ e
pairs produces a shielding effect, so that
the charge and the strength of the interaction , as seen from a distance, will appear
altered. Detailed calculations show that if we write the Coulomb potential as
eff ðrÞ
hc
eff ðrÞ ¼ ; ð5:13Þ
r
then
eff ¼  1=137 ð5:14Þ

for

r  rC 
h=mc ¼ 3:9  10
13 m; ð5:15Þ

but for r  rC , the value of  is somewhat larger and increases as r becomes

smaller. In other words, the strength of the interaction increases at very short
distances. Formally, without proof, the QED coupling em ðQ2 Þ is given by
  2 
1
1 Q
em ðQ2 Þ ¼ ð 2 Þ 1
ð 2 Þln 2 ; ð5:16Þ
3

where 2 is a low-energy value of Q2 at which the value of  is known. If, for

example, we take ¼ 1 MeV=c and  ¼ 1=137, i.e. the value of the fine structure
constant as found from low-energy interactions, then at the mass of the Z 0 boson,
 1=135. Thus the electromagnetic coupling increases with energy-transfer, but
only very slowly.

11
The name arises from the analogy of placing a charge in a dielectric medium. This aligns the particles of
the medium and produces a net polarization.
164 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

Figure 5.9 The two lowest-order vacuum polarization corrections to one-gluon exchange in
quark--quark scattering

Vacuum polarization effects have measurable consequences. For example, the

2S state in hydrogen is predicted to be more tightly bound than it would be in a
pure Coulomb potential. The increased binding is only 2:2  10
7 eV, but never-
theless it is confirmed by extremely accurate measurements on the hydrogen
spectrum. There are also very small corrections to the magnetic moment of the
electron that have been verified experimentally to extraordinary precision.
Quantum fluctuations also exist in QCD and also give rise to a variation of the
interaction strength with distance. If, by analogy with QED, we consider quark–
quark scattering, then the two lowest-order vacuum polarization corrections are
shown in Figure 5.9. The first of these (Figure 5.9(a)) is analogous to virtual eþ e

production in QED and also leads to a screening effect. However, the second
diagram (Figure 5.9(b)) has no counterpart in QED, because there are no direct
photon self-couplings. Calculations show that this diagram leads to an antiscreen-
ing effect that is larger than the screening effect from Figure 5.9(a) and so the net
effect is that the interaction grows weaker at short distances, i.e. asymptotic
freedom. Formally, the strong interaction coupling s is given by a formula
analogous to that for em above, except the coefficient of the logarithmic term is
different and, crucially, its sign is positive:
 
1
s ð 2 Þ
s ðQ2 Þ ¼ s ð 2 Þ 1 þ ð33
2Nf ÞlnðQ2 = 2 Þ ; ð5:17Þ
12

where again 2 is a low-energy value of Q2 at which the value of s is known and

Nf is the number of quark flavours that take part in the interaction.

5.5 Jets and Gluons

A striking feature of many high-energy particle collisions is the occurrence of jets
of hadrons in the final state. We have already mentioned these in Section 3.2.1
when we discussed the experimental evidence for quarks and again when we
 JETS AND GLUONS 165

discussed basic properties of quarks and gluons interactions earlier in this chapter.
They have been extensively studied in the reaction
eþ þ e
! hadrons ð5:18Þ

at high energies using colliding beam experiments, which were discussed in

Chapter 4. High-energy electrons and positrons collide head-on, with equal and
opposite momenta, so that the total momentum of the hadrons produced cancels
out to zero in order to conserve momentum. This is a particularly ‘clean’ reaction,
because the initial particles are elementary, without internal structure.
In the centre-of-mass energy range 15–40 GeV, electron–positron annihilation
into hadrons is dominated by the production of jets. These can be regarded as
occurring in two stages: (1) a primary electromagnetic process eþ þ e
! q þ
q
(due to photon exchange) leading to the production of a quark–antiquark pair,
followed by (2) fragmentation (the concept we met in discussing asymptotic
freedom) which converts the high-energy q
q pair into two jets of hadrons. This is
illustrated in Figure 5.10.

Figure 5.10 Basic mechanism of two-jet production in electron--positron annihilation

The fragmentation process that converts the quarks into hadrons is very
complicated, and the composition of the jets – i.e. the numbers and types of
particles in the jet and their momenta – varies from event to event. However, the
direction of a jet, defined by the total momentum vector
X
P¼ pi ; ð5:19Þ
i

where the sum extends over all the particles within the jet, closely reflects the
parent quark or antiquark direction. This is because the QCD interaction is
relatively weak at very short distances (asymptotic freedom), and the quark and
antiquark do not interact strongly until they are separated by a distance r of order
1 fm. At these relatively large distances, only comparatively small momenta can be
transferred, and hence the jets that subsequently develop point almost exactly in
the initial quark and antiquark directions. That is, the jet angular distribution
relative to the electron beam direction reflects the angular distributions of the
quark and antiquark in the basic reaction eþ þ e
! q þ
q. The latter can easily
166 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

be calculated in QED as it is a purely electromagnetic process, and is in excellent

agreement with the observed angular distribution of the jets. This is one of the
pieces of evidence for the existence of quarks that was cited in Chapter 3 and again
at the start of the present chapter.
Although the dominant process in electron–positron annihilation into hadrons is
the formation of two ‘back-to-back’ jets, occasionally we would expect a high-
momentum gluon to be emitted by the quark or anti-quark before fragmentation
occurs, in much the same way as a high-energy electron sometimes emits a photon
(i.e. bremssrahlung). The quark, antiquark and gluon then all fragment into
hadrons, leading to a three-jet event. A computer reconstruction of such an
event in a jet chamber is shown in Figure 5.11.

Figure 5.11 Computer reconstruction of a three-jet event in electron--positron annihilation

Events like these provided the first unambiguous evidence for gluons, because
the angular distributions of the jets are found to be in good agreement with the
theoretical expectation for spin-1 gluons, but are inconsistent with what would
be expected if, for example, the third jet originated from a particle of spin-0. The
ratio of three-jet to two-jet events can also be calculated, assuming that the third jet is
a gluon, because the probability that a quark or antiquark will emit a gluon is
determined by the strong coupling s , in the same way that the probability that an
electron or positron will emit a photon is determined by the fine structure constant .
This leads to a value of s and hence , the QCD scale parameter. The value obtained
is consistent with Equation (5.12) found from other determinations and lends further
support for the whole picture of quarks interacting via the exchange of gluons.

5.6 Colour Counting

What evidence is there that quarks exist in just three colour states? This question can
be settled by using data from electron–positron annihilation. The cross-sections for
 COLOUR COUNTING 167

electron–positron annihilation to hadrons and for electron–positron annihilation to

muons12 both decrease rapidly with energy, but their ratio

ðeþ e
! hadronsÞ
R ð5:20Þ
ðeþ e
! þ
Þ

is almost energy independent. The near constancy of this ratio follows from the
dominance of the two-step mechanism of Figure 5.10, with the total annihilation
rate being determined by that of the initial reaction eþ e
! q þ
q. The value of
the ratio R then directly confirms the existence of three colour states, each with the
same electric charge, for each quark flavour.
To understand this, let us suppose that each quark flavour f ¼ u, d, s . . . exists
in NC colour states, so that NC ¼ 3 according to QCD, while NC ¼ 1 if the colour
degree of freedom does not exist. Since the different colour states all have the same
electric charge, they will all be produced equally readily by the mechanism of
Figure 5.10, and the rate for producing quark pairs of any given flavour f ¼ u, d,
s, . . . will be proportional to the number of colours NC . The cross-section is also
proportional to the squared charge of the produced pair (because this is a first-order
electromagnetic process), and since muon pairs are produced by an identical
mechanism, we obtain

ðeþ e
! q
qÞ ¼ NC e2f ðeþ e
! þ
Þ; ð5:21Þ

where ef is the electric charge, in units of e, on a quark of flavour f.

The cross-section for eþ þ e
! hadrons will receive an additional contribution
of the form of Equation (5.21) when the energy passes a threshold for a new quark
flavour to be produced. Thus R at low energies will have a series of ‘steps’
corresponding to the production of pairs of new quarks and this is what is observed
experimentally. At high energies above the threshold for the production of b
b pairs
and assuming that hadron production is completely dominated by the two-step
process of Figure 5.10, we would have13

R ¼ R0  NC ðe2u þ e2d þ e2s þ e2c þ e2b Þ ¼ 11NC =9: ð5:22Þ

When the small contribution from the three-jet events and other corrections of
order s are taken into account, this expression for R is modified to

R ¼ R0 ð1 þ s =Þ; ð5:23Þ

12
The cross-section for the production of muon pairs is essentially a purely electromagnetic one, except at
very high energies where the effect of the weak interaction may be seen. This will be discussed in Chapter 6.
13
There is no contribution from the top quark because it is too heavy to be produced, even at the high
energies we are considering.
168 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

Figure 5.12 Measured values of the cross-section ratio R and the theoretical prediction from
QCD for NC ¼ 3 colours; the dashed line shows the prediction without QCD corrections

giving rise to a weak energy dependence of R from the energy dependence of s

discussed earlier (Equation (5.17)). Although these corrections of order s are
small compared to the dominant contribution, they must be included if the most
precise experimental data on R are to be accounted for. The data are in excellent
agreement with the theoretical prediction for the value NC ¼ 3 (see Figure 5.12)
and hence prove that quarks exist in just three colour states.

5.7 Deep Inelastic Scattering and Nucleon Structure

In Chapter 2 we discussed the scattering of electrons from nuclei to determine their
radial charge distributions. This was done by assuming a form for the charge
distribution, calculating the resulting form factor (i.e. the Fourier transform of the
charge distribution) and using it to fit experimental cross-sections. In a somewhat
similar way we can use high-energy inelastic scattering to investigate the charge
distribution within nucleons. This is referred to as deep inelastic scattering,
because the projectiles probe deep into the internal structure of the nucleon.
This type of interaction was mentioned in Section 2.9 in the context of classifying
nuclear reaction mechanisms. The original experiments of this type in particle
physics were done in the 1960s and provided the first definitive evidence for the
existence of quarks. We will deduce that nucleons have a sub-structure of point-
like charged constituents.14

14
The pioneering work on deep inelastic scattering done by Jerome Friedman, Henry Kendall and Richard
Taylor resulted in their receiving the 1990 Nobel Prize in Physics.
 DEEP INELASTIC SCATTERING AND NUCLEON STRUCTURE 169

Figure 5.13 Dominant one-photon exchange mechanism for inelastic lepton--proton

scattering where ‘ ¼ e or

The dominant one-photon contribution to the inelastic scattering of a charged

lepton from a proton in the spectator quark model is shown in Figure 5.13. Unlike
elastic scattering, where at a given lepton energy E there is only one free variable
(e.g. the scattering angle), in inelastic scattering the excitation energy of the
nucleon adds a further degree of freedom, so we can define two independent
variables. These are usually taken to be , defined by

2M  W 2 c2 þ Q2
M 2 c2 ð5:24Þ

and a dimensionless quantity (called the scaling variable) given by

x  Q2 =2M: ð5:25Þ

Here, M is the proton mass, W is the invariant mass of the final-state hadrons and
Q2 is the squared energy–momentum transfer

Q2 ¼ ðE
E0 Þ2 =c2
ðp
p0 Þ2 : ð5:26Þ

The physical interpretation of x will be discussed below. In the rest frame of the
initial proton,  reduces to

 ¼ E
E0 ð5:27Þ

and so is the Lorentz-invariant generalization for the energy transferred from the
lepton to the proton.
In Chapter 2 we discussed several modifications to the formalism for describing
the structure of nuclei obtained from scattering experiments. Here we are dealing
with high-energy projectiles and so we will need to take all those corrections into
account. In particular, the magnetic interaction introduces a second form factor.
(cf Equation (2.14)). The two form factors, denoted W1 and W2 , are called structure
170 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

functions in this context. In terms of these, the differential cross-section may be

written
 
d2  d
¼ ½W2 ðQ2 ; Þ þ 2W1 ðQ2 ; Þ tan2 ð=2Þ ; ð5:28Þ
ddE0 d Mott

where  is the lepton scattering angle. For values of W  2:5 GeV=c2 , the cross-
sections show considerable structure due to the excitation of nucleon resonances, but
above this mass they are smoothly varying. In the latter region, the values of the
structure functions can be extracted from the data by choosing suitable parameter-
izations and fitting the available data in an analogous way to the way charge
distributions of nuclei were deduced in Chapter 2.
Rather than W1 and W2 , it is usual to work with two related dimensionless
structure functions defined by

F1 ðx; Q2 Þ  Mc2 W1 ðQ2 ; Þ and F2 ðx; Q2 Þ  W2 ðQ2 ; Þ: ð5:29Þ

It is a remarkable fact that at fixed values of x the structure functions have only a
very weak dependence on Q2 . This behaviour is referred to as scaling and is
illustrated in Figure 5.14. As the Fourier transform of a spherically symmetric
point-like distribution is a constant, we conclude that the proton has a sub-structure
of point-like charge constituents.
The interpretation of scaling is simplest in a reference frame where the target
nucleon is moving with a very high velocity, so that the transverse momenta and

Figure 5.14 The structure function F2 of the proton as a function of x, for Q2 between 2 and
18 ðGeV=cÞ2 (reproduced from At82 with kind permission of Springer Science and Business
Media)
 DEEP INELASTIC SCATTERING AND NUCLEON STRUCTURE 171

Figure 5.15 The parton model of deep inelastic scattering

rest masses of its constituents may be neglected. The structure of the nucleon is
then given by the longitudinal momentum of its constituents. This approach was
first adopted by Feynman and Bjorken, who called the constituents partons. (We
now identify charged partons with quarks and neutral partons with gluons.) In
the parton model, deep inelastic scattering is visualized as shown in Figure 5.15.
The target nucleon is a stream of partons each with four-momentum xP, where
P ¼ ðp; pÞ is the four-momentum of the nucleon and p ¼ jpj, is its (very large)
three-momentum, so that the nucleon mass may be neglected.
Suppose now that one parton of mass m is scattered elastically by the exchanged
photon of four-momentum Q. Then

ðxP þ QÞ2 ¼ ðx2 P2 þ 2xP  Q þ Q2 Þ ¼ m2 c4 0: ð5:30Þ

 
If x2 P2  ¼ x2 M 2 c4  Q2 , then

Q2 Q2
x¼
¼ ; ð5:31Þ
2P  Q 2M

where the invariant scalar product has been evaluated in the laboratory frame
in which the energy transfer is  and the nucleon is at rest. This is our previous
definition Equation (5.25). Thus, the physical interpretation of x is the fractional
three-momentum of the parton in the reference frame where the nucleon has
a very high velocity. This is equivalent to having a parton of mass m stationary
in the laboratory system, with the elastic relation Q2 ¼ 2m. So provided
Q2  M 2 ,

Q2 m
x¼ ¼ ; ð5:32Þ
2M M

i.e. x may also be interpreted as the fraction of the nucleon mass carried by the
struck parton.
172 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

To identify the constituent partons with quarks we need to know their spins and
charges. For the spin, it can be shown that

F1 ðx; Q2 Þ ¼ 0 ðspin 0Þ ð5:33Þ

and

2xF1 ðx; Q2 Þ ¼ F2 ðx; Q2 Þ ðspin 1=2Þ: ð5:34Þ

The latter relation, known as the Callan–Gross relation, follows by comparing the
coefficients in the equation for the double differential cross-section Equation (5.28)
with that in Chapter 2 (Equation (2.14)). This gives

2W1 =W2 ¼ 2; ð5:35Þ

where  ¼ Q2 =4m2 c2 and m is the mass of the target, in this case the mass of the
struck parton. Replacing W1 by F1 =Mc2 and W2 by F2 =, gives

 F1 Q2
2
¼ 2 2 ð5:36Þ
Mc F2 4m c

and since now Q2 ¼ 2m, we have m ¼ Q2 =2 ¼ xM. Finally, using this mass in
Equation(5.36) yields the Callan–Gross relation. Figure 5.16 shows some results
for the ratio 2xF1 =F2 . It is clear that spin-12 is strongly favoured.
To deduce the parton charges is more complicated. We will assume that the
constituent partons are quarks and show that this is consistent with experimental
data. We start by defining qf ðxÞ to be the momentum distribution of a quark of
flavour f, i.e. qf ðxÞdx is the probability of finding in a nucleon a quark of flavour f,
with momentum fraction in the interval x to x þ dx. A given nucleon will consist of
a combination of valence quarks (i.e. those that give rise to the observed quantum
numbers in the quark model) and additional quark–antiquark pairs that are
continually produced and annihilated by the radiation of virtual gluons by the
quarks.15 (Recall the discussion of quantum fluctuations in electrodynamics in
Section 5.4.) Thus, in general, a structure function can be written as the sum of
contributions from quarks and antiquarks of all flavours. Also, from the cross-
section formula Equation (5.28), we would expect the structure functions to
involve the quark distributions weighted by the squares of the quark charges zf
(in units of e) for a given quark flavour f.

15
These are the ‘sea’ quarks referred to in the discussion of the static quark model in Chapter 3.
 DEEP INELASTIC SCATTERING AND NUCLEON STRUCTURE 173

Figure 5.16 The ratio 2xF1 =F2 at fixed x

Thus, for example, F2 is

X 
F2 ðxÞ ¼ x z2f qf ðxÞ þ
qf ðxÞ : ð5:37Þ
f

If we concentrate on the scattering of charged leptons, i.e. electrons or muons, and

consider just the possibility of light quarks u, d and s within nucleons, then we
have (for ‘ ¼ e; )
 
1 p
p 4 p 1 p
F2‘p ðxÞ p p
¼ x ðd þ d Þ þ ðu þ
u Þ þ ðs þ
s Þ ð5:38aÞ
9 9 9

and
 
1 n
n 4 n 1 n
F2‘n ðxÞ n n
¼ x ðd þ d Þ þ ðu þ
u Þ þ ðs þ
s Þ ; ð5:38bÞ
9 9 9

where, for example, un;p is the distribution of u quarks in the neutron and proton.
Using isospin symmetry, interchanging u and d quarks changes neutron to proton,
174 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

i.e. u $ d implies n $ p. Thus,

up ðxÞ ¼ dn ðxÞ  uðxÞ; ð5:39aÞ

p n
d ðxÞ ¼ u ðxÞ  dðxÞ; ð5:39bÞ

and

s p ðxÞ ¼ s n ðxÞ  sðxÞ; ð5:39cÞ

with similar relations for the antiquarks. Then if we work with a target nucleus
with equal numbers of protons and neutrons (an isoscalar target), its structure
function will have the approximate form (neglecting purely nuclear effects)

1 5 X 1
F2‘N ðxÞ ¼ ½F2‘p ðxÞ þ F2‘n ðxÞ ¼ x ½qðxÞ þ
qðxÞ þ x½sðxÞ þ
sðxÞ : ð5:40Þ
2 18 q¼d;u 9

The second term is small because s quarks are only present in the sea component at
the level of a few percent. Thus the mean squared value of the charges of the u and
5
d quarks is approximately 18 .
The final step is to extract information from deep inelastic scattering using
neutrinos and antineutrinos as projectiles. This is more complicated because, as we
shall see in Chapter 6, neutrinos and antineutrinos couple differently to the different
quarks and antiquarks and there is also a third form factor involved. Without proof,
we shall just quote the result:
X
F2N ðxÞ ¼ x ½qðxÞ þ
qðxÞ : ð5:41Þ
q¼d;u

There is no electric charge factor outside the summation because, just as quarks form
strong interaction isospin multiplets with different electric charges, the leptons also
form weak isospin multiplets, but in this case the resulting weak charge is the same
for all quarks.16
From Equation (5.40) and (5.41), we expect

18 ‘N
F2N ðxÞ  F ðxÞ: ð5:42Þ
5 2

The experimental data illustrated in Figure 5.17 show that F2‘N ðxÞ and F2N ðxÞ are
equal within errors except possibly at small values of x where antiquarks are more
important. Thus one can conclude that the partons do have charges 23 and
13, which
completes the evidence for identifying partons with quarks.

16
Weak isospin will be discussed briefly in Chapter 6.
 DEEP INELASTIC SCATTERING AND NUCLEON STRUCTURE 175

Figure 5.17 Comparison of F2 ðxÞ from deep inelastic muon (data from Ar97) and neutrino
(data from Se97) scattering experiments; the data points are the average over a range of
Q2 > 2 ðGeV=cÞ2 and the error bars express the range of data values within the Q2 ranges

Combining data from different experiments, with electrons, muons, neutrinos and
antineutrinos as projectiles, enables individual quark/parton momentum distribu-
tions to be extracted from combinations of cross-sections. Some typical results at
Q2 ¼ 10 ðGeV=cÞ2 are shown in Figure 5.18 for the combinations

QðxÞ ¼ dðxÞ þ uðxÞ ð5:43aÞ

and

ðxÞ ¼ d
ðxÞ þ
uðxÞ:

Q ð5:43bÞ

The difference

ðxÞ
Qv ðxÞ  QðxÞ
Q ð5:44Þ

can be identified as the distribution of the valence quarks of the quark model. It can
be seen that Qv is concentrated around x 0:2 and dominates except at small
values of x where the antiquarks
q in the sea distribution are important.
The results of Figure 5.18 reveal an interesting and unexpected result concerning
gluons within the nucleon. If we integrate the momentum distributions for quarks
and antiquarks over all x we might expect to recover the total momentum of the
nucleon, whereas the curves of Figure 5.18 yield a value of approximately 0.5.
Thus it follows that about 50 per cent of the momentum is carried by gluons.
Although scaling is approximately correct, it is certainly not exact. In Fig-
ure 5.19 we show some deep inelastic scattering data plotted in more detail. The
176 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

Figure 5.18 Quark and antiquark momentum distributions in the nucleon

deviations from scaling are due to QCD corrections to the simple quark model, i.e.
the quark in the proton that is struck by the exchanged photon can itself radiate
gluons. Again, without further details, the scaling violations are explained by QCD
using a value of the strong interaction parameter  that is consistent with that
obtained from other sources (e.g. the three-jet events that we have discussed
above).17
Finally, it is worth noting that the nucleon structure functions and hence the
quark densities are found from lepton scattering experiments using a range of
different nuclear targets. We have seen that the average binding energy of nucleons
in heavy nuclei is of the order of 7–8 MeV per nucleon. As this energy is much
smaller than those used in deep inelastic scattering experiments, it might be
thought safe to ignore nuclear effects (except those due to the internal motion of
the nucleons – the Fermi momentum – that are typically about 200 MeV/c).
However, experiments have shown that the structure functions do in fact depend

17
Scaling violations are discussed in detail, but at a more advanced level than here in, for example Ha84.
 PROBLEMS 177

Publisher's Note:
Permission to reproduce this image
online was not granted by the
copyright holder. Readers are kindly
requested to refer to the printed version
of this chapter.

Figure 5.19 A compilation of values of F2 measured in deep inelastic electron and muon
scattering from a deuterium target -- different symbols denote different experiments; for clarity,
the data at different values of x have been multiplied by the factors shown in brackets and the
solid line is a QCD fit with  ¼ 0:2 GeV (adapted from Mo94, copyright the American Physical
Society)

slightly on the nuclear medium. Although the effects are very small and not
enough to alter the conclusions of this chapter, it is a reminder that there are still
things to be learnt about the role of nuclear matter and that this may hold
information on the nuclear force in terms of the fundamental quark–gluon
interaction.

Problems
5.1 The general combination of m quarks and n antiquarks qm q
n , with baryon number




B > 0 has a colour wavefunction that may be written r  g b

r 

g

b , where r 
means that there are  quarks in the r colour state, etc.. By imposing the condition of
colour confinement, show that m
n ¼ 3p, where p is a non-negative integer and
hence show that states with the structure qq are not allowed.
178 CH5 QUARK DYNAMICS: THE STRONG INTERACTION

5.2 Draw the lowest-order Feynman diagrams for the following processes:

(a) the interaction of a quark and a gluon to produce a quark and a photon;

(b) the production of a single Z 0 -boson in a collision of protons and antiprotons;

(c) the annihilation of an electron and a positron to produce a pair of W-bosons.

5.3 A p
p collider with equal beam energies is used to produce a pair of top quarks. Draw
a Feynman diagram for this process that involves a single gluon. If the three quarks
of the proton (or antiproton) carry between them 50 per cent of the hadron total
energy–momentum, calculate the minimum beam momentum required to produce
the t
t pair.

5.4 The lowest Feynman diagram for inelastic electron–proton scattering at high
energies
e
ðE; pcÞ þ pðEP ; Pp cÞ ! e
ðE0 ; p0 cÞ þ XðhadronsÞ

is shown in Figure 5.20.

Figure 5.20 Kinematics of inelastic electron--proton scattering

Use energy–momentum conservation to show that the variable  defined in

Equation (5.24) becomes  ¼ E
E0 in the rest frame of the proton. Hence show
that the variable x defined in Equation (5.25) lies in the range 0  x  1 if the mass
of the electron is neglected.

5.5 If hadron–hadron total cross-sections are assumed to be the sum of the cross-sections
between their constituent quarks, show that the quark model predicts the
relationship:

ðpÞ ¼ ðppÞ þ ðK
nÞ
ðþ pÞ:

5.6 The 3
decay of positronium (the bound state of eþ e
) has a width that in QED is
predicted to be ð3
Þ ¼ 2ð2
9Þ6 me c2 =9, where  is the fine structure constant.
If the hadronic decay of the c
c bound state J=ð3100Þ proceeds via an analogous
 PROBLEMS 179

mechanism, but involving three gluons, use the experimental hadronic width
ð3gÞ ¼ 80 keV to estimate the strong interaction coupling constant s . Use an
analogous assumption to estimate s from the radiative width ðgg
Þ ¼ 0:16 keV of
the b
b bound state ð9460Þ.

5.7 Use Equations (5.38) and (5.39) to derive the Gottfried sum rule,

ð1 ð1
dx 1 2
½F2ep ðxÞ
F2en ðxÞ ¼ þ ½
uðxÞ
d
ðxÞ dx;
x 3 3
0 0

where the quark distributions refer to the proton.

5.8 Estimate the cross-section ratio R defined in Equation (5.20) at centre-of-mass

energies ECM ¼ 2:8 GeV and 15 GeV. How would R change if the energy were
increased so that top quark pairs could be produced?

5.9 Common forms assumed for the momentum distributions of valence quarks in the
proton are:

Fu ðxÞ ¼ xuðxÞ ¼ að1
xÞ3 ; Fd ðxÞ ¼ xdðxÞ ¼ bð1
xÞ3 :

If the valence quarks account for half the proton’s momentum, find the values of a
and b.

5.10 The cross-section ðud
! W þ Þ near the mass of the W þ is given by the Breit–
Wigner form
hcÞ2 
ð

2 ud

¼ ;
3½4ðE
MW c2 Þ2 þ 2

where ðMW ; Þ are the mass and total width of the W þ , ud
is the partial width for
W þ ! ud
, E is the total centre-of-mass energy of the ud
pair and 
¼ 2=E. Find the
maximum value of , i.e. max , given that the branching ratio for W þ ! ud
is 1=3.
Use this result and the quark distributions of Question 5.9 to find an expression for
pffiffi cross-section ðp

the p ! W þ þ   Þ in terms
pffiffi of the p

p total centre-of-mass energy
s and max and evaluate your result for s ¼ 1 TeV. (Use the narrow width, i.e.
delta function, approximation
!
W E2
ud
ðEÞ ¼   max  1

MW c2 ðMW c2 Þ2

in integrals.)
6
Electroweak Interactions

We have already discussed some aspects of weak and electromagnetic interactions

when we discussed nuclear stability in Chapter 2 and again when we introduced
the basic properties of leptons in Chapter 3. In this chapter we will consider wider
aspects of the weak interaction and also its unification with electromagnetism to
produce the spectacularly successful electroweak theory.

6.1 Charged and Neutral Currents

Like the strong and electromagnetic interactions, the weak interaction is also
associated with elementary spin-1 bosons, which act as ‘force carriers’ between
quarks and/or leptons. Until 1973 all observed weak interactions were consistent
with the hypothesis that they were mediated by the exchange of the charged bosons
W
only. However, in the 1960s, a theory was developed which unified electro-
magnetic and weak interactions in a way that is often compared with the
unification of electric and magnetic interactions by Maxwell a century earlier.
This new theory made several remarkable predictions, including the existence of
the heavy neutral vector boson Z 0 and of weak reactions arising from its exchange.
The latter processes are called neutral current reactions (the word neutral referring
to the charge of the exchanged particle) to distinguish them from the so-called
charged current reactions arising from charged W
boson exchange. In particular,
neutral current reactions of the type  þ N !  þ X were predicted to occur via
the mechanism of Figure 6.1, where N is a nucleon and X is any set of hadrons
allowed by the conservation laws. Although difficult to detect, such reactions were
first observed in a bubble chamber experiment in 1973.
The prediction of the existence and properties of neutral currents, prior to their
discovery, is only one of many remarkable successes of the unified theory of
electromagnetic and weak interactions. Others include the prediction of the
existence of the charmed quark, prior to its discovery in 1974 and the prediction

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
182 CH6 ELECTROWEAK INTERACTIONS

Figure 6.1 Feynman diagram for the weak neutral current reaction
 þ N !
 þ X

of the masses of the W
and Z 0 bosons prior to the long-awaited detection of these
particles in 1983. In general, the theory is in agreement with all data on both weak
and electromagnetic interactions, which are now referred to collectively as the
electroweak interaction, in the same way that electric and magnetic interactions
are referred to collectively as electromagnetic interactions. Furthermore, the theory
predicts the existence of a new spin-0 boson, the so-called Higgs boson, which is
associated with the origin of particle masses within the model. This was mentioned
in passing in earlier chapters. Although a detailed discussion of the Higgs boson is
beyond the scope of this book, there is a brief discussion of the role of this very
important particle in Chapter 9.
The new unification only becomes manifest at high energies, and at low energies
weak and electromagnetic interactions can still be clearly separated. This follows
from the general form of the amplitude Equation (1.41):

g2
h2
Fðq2 Þ ¼ ; ð6:1Þ
jqj2 þ MX2 c2

where MX2 is the mass of the exchanged particle and g is the appropriate coupling.
For weak interactions, MX ¼ MW;Z 80 GeV=c2 and for the electromagnetic
interaction MX ¼ M ¼ 0. Thus, even with gweak
gem , the amplitudes for the
two interactions will only become of comparable size for jqj2
MX2 c2, i.e. at high
energies. We therefore start by considering the weak interaction at low energies
and deduce some of its general properties that are valid at all energies. Later we
will consider how unification arises and some of its consequences.

6.2 Symmetries of the Weak Interaction

In this section we will discuss the parity (P) and charge conjugation (C ) operators,
which were introduced in Chapter 1. These are conserved in the strong and
 SYMMETRIES OF THE WEAK INTERACTION 183

electromagnetic interactions. The first indication that parity might be violated in

weak interactions came from observations on the pionic decays of K-mesons, i.e.
K !  and K ! ,1 and these led Lee and Yang in 1956 to make a thorough
study of all previous experiments in which parity conservation had been assumed
or apparently proved. They came to the startling conclusion that there was in fact
no firm evidence for parity conservation in weak interactions; and they suggested
experiments where the assumption could be tested.2 This led directly to the classic
demonstration of parity violation from a study of the -decay of polarized 60 Co
nuclei. We shall just describe the principles of this experiment.3
The experiment was done in 1957 by Wu and co-workers, who placed a sample
of 60 Co inside a magnetic solenoid and cooled it to a temperature of 0.01 K. At
such temperatures, the interaction of the magnetic moments of the nuclei with the
magnetic field overcomes the tendency to thermal disorder, and the nuclear spins
tend to align parallel to the field direction. The polarized 60 Co nuclei produced in
this way decay to an excited state of 60 Ni by the -decay

60
Co ! 60
Ni þ e þ
e : ð6:2Þ

Parity violation was established by the observation of a ‘forward–backward

decay asymmetry’, i.e. the fact that fewer electrons were emitted in the forward
hemisphere than in the backward hemisphere with respect to the spins of the
decaying nuclei.
We can show that this implies parity violation as follows. The parity transforma-
tion reverses all particle momenta p while leaving their orbital angular momenta
r  p, and by analogy their spin angular momenta, unchanged. Hence in the rest
frame of the decaying nuclei its effect is to reverse the electron velocity while
leaving the nuclear spins unchanged, as shown in Figure 6.2. Parity invariance would
then require that the rates for the two processes Figure 6.2(a) and Figure 6.2(b) were
equal, so that equal numbers of electrons would be emitted in the forward and
backward hemispheres with respect to the nuclear spins, in contradiction to what
was observed. The discovery of parity violation was a watershed in the history of
weak interactions because the effect is large, and an understanding of weak
interactions is impossible if it is neglected.
The charge conjugation operator C changes all particles to antiparticles and as
we will see presently is also not conserved in weak interactions. In examining
these operators, two interconnected themes will emerge. The first is that these
effects have their origin in the spin dependence of weak interactions; the second is

1
Two particles, called at that time and
, were observed to decay via the weak interaction to  and 
final states, respectively, which necessarily had different final-state parities. However, the and
had
properties, including the near equality of their masses, which strongly suggested that they were in fact the
same particle. Analysis of the ‘ –
puzzle’ suggested that parity was not conserved in the decays.
2
For their work on parity non-conservation, Chen Yang and Tsung-Dao Lee were awarded the 1957 Nobel
Prize in Physics.
3
This classic experiment is described in readable detail in Chapter 10 of Tr75.
184 CH6 ELECTROWEAK INTERACTIONS

Figure 6.2 Effect of a parity transformation on 60 Co decay: the thick arrows indicate the
direction of the spin of the 60 Co nucleus, while the thin arrows show the direction of the
electron’s momentum

that while P-violation and C-violation are large effects, there is a weaker combined
symmetry, called CP-invariance, which is almost exactly conserved. This has its
most striking consequences for the decays of neutral mesons, which are also
discussed below. We start by considering the P and C operators in more detail.
C-violation and P-violation are both conveniently illustrated by considering the
angular distributions of the electrons and positrons emitted in the decays
 ! e þ
e þ
 ð6:3aÞ

and
þ ! eþ þ
e þ
 ð6:3bÞ

of polarized muons. In the rest frame of the decaying particle these were found to
be of the form


1 


ðcos
Þ ¼ 
1  cos
; ð6:4Þ
2 3

where
is the angle between the muon spin direction and the direction of the
outgoing electron or positron, as shown in Figure 6.3(a). The quantities 
are
called the asymmetry parameters, and 
are the total decay rates, or equivalently
the inverse lifetimes, i.e.
ð
þ1

1 d cos

ðcos
Þ ¼ 
; ð6:5Þ
1

as may easily be checked by direct substitution.

 SYMMETRIES OF THE WEAK INTERACTION 185

e+−

µ+
− P π−θ µ+
−

e+
−
(a) (b)

Figure 6.3 Effect of a parity transformation on muon decays: the thick arrows indicate the
direction of the muon spin, while the thin arrows indicate the direction of the electron’s
momentum

We consider now the consequences of assuming parity and charge conjugation

for these decays, starting with the latter as it is the simpler. Under charge
conjugation,  decay converts to þ decay. C-invariance then implies that the
rates and angular distributions for these decays should be the same, i.e.

þ ¼  ðC-invarianceÞ ð6:6Þ

and
þ ¼  ðC-invarianceÞ: ð6:7Þ

The parity transformation preserves the identity of the particles, but reverses
their momenta while leaving their spins unchanged. Its effect on muon decay is
shown in Figure 6.3, where we see that it changes the angle
to  
, so that cos

changes sign. Hence P-invariance implies


ðcos
Þ ¼ 
ð cos
Þ ðP-invarianceÞ: ð6:8Þ

Substituting Equation (6.4), leads to the prediction that the asymmetry parameters
vanish,


¼ 0 ðP-invarianceÞ: ð6:9Þ

Experimentally, the 
lifetimes are equal to a very high level of precision, so

that the prediction for the lifetimes is satisfied; but the measured values of the
186 CH6 ELECTROWEAK INTERACTIONS

asymmetry parameters are

 ¼ þ ¼ 1:00
0:04; ð6:10Þ

which shows that both C-invariance and P-invariance are violated. The violation is
said to be ‘maximal’, because the asymmetry parameters are defined to lie in the
range 1  
 1.
In view of these results, a question that arises is: why do the þ and  have
the same lifetime if C-invariance is violated? The answer lies in the principle of
CP-conservation, which states that the weak interaction is invariant under the
combined operation CP even though both C and P are separately violated. The CP
operator transforms particles at rest to their corresponding antiparticles at rest, and
CP-invariance requires that these states should have identical properties. Thus, in
particular, the masses of particles and antiparticles are predicted to be the same.
Specifically, if we apply the CP operator to muon decays, the parity operator
changes
to  
as before, while the C operator changes particles to anti-
particles. Hence CP-invariance alone implies that the condition obtained from
P-invariance is replaced by the weaker condition

þ ðcos
Þ ¼  ð cos
Þ: ð6:11Þ

Again, substituting Equation (6.4) into Equation (6.11), gives

þ ¼  ðCP-invarianceÞ; ð6:12Þ

implying equal lifetimes and also

þ ¼  ðCP-invarianceÞ; ð6:13Þ

in agreement with the experimental results. Thus CP-invariance retains the

symmetry between particles and antiparticles as observed by experiment, at
least for -decays. In fact CP-invariance has been verified in a wide variety of
experiments involving weak interactions, and it is believed to be exact for purely
leptonic processes (i.e. ones involving only leptons) and a very good approxima-
tion for those involving hadrons. (The only known violations will be discussed in
Section 6.6.1.) Particles and antiparticles have the same masses and lifetimes even
if CP is not conserved.

6.3 Spin Structure of the Weak Interactions

We turn now to the spin structure of the weak interactions, which is closely related
to the symmetry properties discussed above. As this spin structure takes its
simplest form for zero-mass particles, we will discuss the case of neutrinos and
 SPIN STRUCTURE OF THE WEAK INTERACTIONS 187

antineutrinos first, assuming that they have zero mass for the purpose of this
discussion.

6.3.1 Neutrinos

In discussing neutrinos, it is convenient to use the so-called helicity states, in

which the spin is quantized along the direction of motion of the particle, rather
than along some arbitrarily chosen ‘z-direction’. For a spin-12 particle, the spin
component along the direction of its motion can be either þ 12 or  12 (in units of
h),
as illustrated in Figure 6.4, corresponding to positive or negative helicity
respectively. These states are called right-handed or left-handed, respectively,
since the spin direction corresponds to rotational motion in a right-handed or left-
handed sense when viewed along the momentum direction.

Figure 6.4 Helicity states of a spin-12 particle: the long thin arrows represent the momenta of
the particles and the shorter thick arrows represent their spins

We will denote these states by a subscript R or L, so that, for example,
L means

a left-handed neutrino. The remarkable fact about neutrinos and antineutrinos,
which only interact via the weak interaction, is that only left-handed neutrinos
L
and right-handed antineutrinos
R are observed in nature. This obviously violates
C-invariance, which requires neutrinos and antineutrinos to have identical weak
interactions. It also violates P-invariance, which requires the states
L and
R to
also have identical weak interactions, since the parity operator reverses the
momentum while leaving the spin unchanged and so converts a left-handed
neutrino into a right-handed neutrino. It is, however, compatible with CP-
invariance, since the CP operator converts a left-handed neutrino to a right-handed
antineutrino, as illustrated in Figure 6.5.
The helicity of the neutrino was first measured in an ingenious experiment by
Goldhaber and co-workers in 1958. Again, we will only discuss the principles of
the experiment. They studied electron capture in 152 Eu, i.e.

e þ 152 EuðJ ¼ 0Þ ! 152 Sm ðJ ¼ 1Þ þ
e ; ð6:14Þ

188 CH6 ELECTROWEAK INTERACTIONS

Figure 6.5 Effect of C, P and CP transformations; only the states shown in boxes are observed
in nature

where the spins of the nuclei are shown in brackets. The excited state of samarium
that is formed decays to the ground state by -emission

152
Sm ðJ ¼ 1Þ ! 152 SmðJ ¼ 0Þ þ  ð6:15Þ

and it is these -rays which were detected in the experiment. In the first reaction
(Equation (6.14)), the electrons are captured from the K-shell and the initial
state has zero momentum, so that the neutrino and the 152 Sm nucleus recoil in
opposite directions. The experiment selected events in which the photon was
emitted in the direction of motion of the decaying 152 Sm nucleus, so that overall
the observed reaction was

e þ 152 EuðJ ¼ 0Þ ! 152 SmðJ ¼ 0Þ þ
e þ ; ð6:16Þ

where the three final-state particles were co-linear, and the neutrino and photon
emerged in opposite directions, as shown in Figure 6.6.
The helicity of the neutrino can then be deduced from the measured helicity of
the photon by applying angular momentum conservation about the event axis to
the overall reaction. In doing this, no orbital angular momentum is involved,
because the initial electron is captured from the atomic K-shell and the final-state
particles all move along the event axis. Hence the spin components of the neutrino
and photon, which can be
12 and
1 respectively, must add to give the spin
component of the initial electron, which can be
12. This gives two possible spin
configurations, as shown in Figures 6.6(a) and 6.6(b). In each case the photon and
neutrino have the same helicities. In the actual experiment, the polarization of the
photons was determined by studying their absorption in magnetized iron (which
depends on the polarization of the photon) and the results obtained were consistent
with the occurrence of left-handed neutrinos only, corresponding to Figure 6.6(a).
 SPIN STRUCTURE OF THE WEAK INTERACTIONS 189

Figure 6.6 Possible helicities of the photon and neutrinos emitted in the reaction
e þ 152 EuðJ ¼ 0Þ ! 152 SmðJ ¼ 0Þ þ
e þ  for those events in which they are emitted in
opposite directions. Experiment selects configuration (a)

Later experiments have shown that only right-handed antineutrinos take part in
weak interactions.

6.3.2 Particles with mass: chirality

To see the effect of the spin dependence in weak interactions involving particles
with mass, we will look at the decays of the pion and muon which are, of course,
examples of charged current reactions. The spin dependence is of a special form,
called a V–A interaction. This name is derived from the behaviour under a parity
transformation of the weak interaction analogue of the electromagnetic current.
The letter V denotes a proper vector, which is one whose direction is reversed by a
parity transformation (an example is momentum p). The familiar electric current,
to which photons couple, transforms as a proper vector under parity. Because
parity is not conserved in weak interactions, the corresponding weak current, to
which W
-bosons couple, has in addition to a vector (V) component another
component whose direction is unchanged by a parity transformation. Such a
quantity is called an axial-vector (A) (an example of an axial-vector is orbital
angular momentum L ¼ r  p). Since observables are related to the modulus
squared of amplitudes, either term would lead by itself to parity conservation.
Parity non-conservation is an interference effect between the two components.
Here we shall consider only the most important characteristic of this spin
dependence, which is that the results discussed above for neutrinos, hold for all
fermions in the ultra-relativistic limit. That is, in the limit that their velocities
approach that of light, only left-handed fermions
L , e L etc. and right-handed
antifermions
R , eþR etc. are emitted in charged current interactions. These
190 CH6 ELECTROWEAK INTERACTIONS

right-handed and left-handed particles are called chiral states and these are the
eigenstates that take part in weak interactions. In general, chiral states are linear
combinations of helicity states,4 with the contributions of the ‘forbidden’ heli-
þ
city states e
R , eL etc. suppressed by factors which are typically of the order of
2
ðmc2 =EÞ , where m is the appropriate fermion mass and E its energy. For massless
neutrinos this is always a good approximation and chiral states and helicity states
are identical. However, for particles with mass, it is only a good approximation for
large energies E. These spin properties can be verified most easily for the electrons
and muons emitted in weak decays, by directly measuring their spins. Here we
shall assume them to hold and use them to understand some interesting features of
pion and muon decays.
We start by considering the pion decay mode

þ ! ‘þ þ
‘ : ð‘ ¼ e; Þ ð6:17Þ

In the rest frame of the decaying pion, the charged lepton and the neutrino recoil in
opposite directions, and because the pion has zero spin, their spins must be
opposed to satisfy angular momentum conservation about the decay axis. Since the
neutrino (assumed to be zero mass) is left-handed, it follows that the charged
lepton must also be left-handed, as shown in Figure 6.7, in contradiction to the
expectations for a relativistic antilepton.

+ π+ ν

Figure 6.7 Helicities of the charged leptons in pion decays: the short arrows denote spin
vectors and the longer arrows denote momentum vectors

For the case of a positive muon this is unimportant, since it is easy to check
that it recoils non-relativistically and so both chirality states are allowed. However,
if a positron is emitted it recoils relativistically, implying that this mode is
suppressed by a factor that we can estimate from the above to be of the order
of ðme =m Þ2 105 . Thus the positron decay mode is predicted to be much rarer
than the muonic mode. This is indeed the case, and the measured ratio

ðþ ! eþ þ
e Þ
¼ ð1:218
0:014Þ  104 ð6:18Þ
ðþ ! þ þ
 Þ

is in excellent agreement with a full calculation that takes into account both the
above suppression and the difference in the density-of-final states (i.e. the
difference in the Q-values) for the two reactions.

4
This is another example where linear combinations of states are the ones of physical interest; compare
neutrino mixing (Section 3.1.3).
 SPIN STRUCTURE OF THE WEAK INTERACTIONS 191

A second consequence of the chirality argument is that the muons emitted in

pion decays are 100 per cent polarized (see Figure 6.7).5 We have mentioned this
earlier in connection with measuring the muon decay asymmetries. These have
their origins in the spin structure of the interaction, as we shall illustrate for the
highest-energy electrons emitted in the decay of the muon. These have energy
!
m c2 m2
E¼ 1 þ 2e  me c2 ð6:19Þ
2 m

and correspond to decays in which the neutrino and antineutrino are both emitted
in the direction opposite to the electron. This is illustrated in Figure 6.8 for the two
simplest cases in which the electron is emitted in the muon spin direction
(Figure 6.8(a)) and opposite to it (Figure 6.8(b)).

Figure 6.8 Muon decays in which electrons of the highest possible energy are emitted: (a) in
the muon spin direction, and (b) opposite to the muon spin direction

Since the neutrino and antineutrino have opposite helicities, the muon and
electron must have the same spin component along the event axis in order to
conserve angular momentum, implying the electron helicities shown in Figure 6.8.
When combined with the fact that the relativistic electrons emitted must be left-
handed, this implies that electrons cannot be emitted in the muon spin direction.
We thus see that the spin structure of the interaction automatically gives rise to a
forward–backward asymmetry in polarized muon decays. Of course not all the
electrons have the maximum energy and the actual asymmetry, averaged over all
electron energies, can only be calculated by using the full form of the V–A

5
This is in the rest frame of the decaying pion and assumes that the neutrino has zero mass. The degree of
polarization in the laboratory frame is a function of the muon momentum.
192 CH6 ELECTROWEAK INTERACTIONS

interaction.6 The resulting prediction is in excellent agreement with the measured

values.
Finally, we have seen in earlier chapters that there is increasing evidence
that neutrinos are not strictly massless. How then can we ensure that the
weak interactions only couple to
L and
R ? To understand this we return to
the Dirac equation, which was mentioned in Chapter 1. As was stated there, the
solution of this equation for a massive spin-12 particle is in the form of a four-
component spinor, whose components are interpreted as the two possible spin
projections for the particle and its antiparticle of a given energy (see Section 1.2
and Equation (1.4)). However, in the case of a massless fermion the Hamiltonian
of Equation (1.2) consists only of a spin projection term and there is a simpler
solution of the Dirac equation consisting of two independent two-component
wavefunctions. If we assume for definiteness the case of neutrinos (assumed to
be massless), then these would correspond to the pairs ð
L ,
R ) and ð
R ,
L ).
This observation was first made by Weyl in 1929, but was rejected as unphysical
because under a parity transformation
L !
R (see Figure 6.5) and hence the
interaction would not be invariant under parity. However, we now know that
parity is not conserved in the weak interactions, so this objection is no longer valid.
A possible solution is therefore to make the neutrino its own antiparticle. In this
case ð
L ,
R Þ are identified as two helicity components of a four-component spinor
and the other two components ð
R ,
L Þ, if they exist, can then be a fermion of a
different mass. This scheme is due to Majorana and is very different to the
structure of a spinor describing a massive spin-12 fermion such as an electron. A test
of this idea would be the observation of neutrinoless double -decay, such as that
given in Equation (3.37), which is only possible if
e
e .

6.4 W
and Z 0 Bosons

The three intermediate vector bosons mediating weak interactions, the two charged
bosons W þ and W  and the neutral Z 0 , were all discovered at CERN in 1983 in the
reactions

p þ p ! W þ þ X ;


p þ p ! W  þ X þ ; and
p þ p ! Z 0 þ X 0 ; ð6:20Þ

where X
and X 0 are arbitrary hadronic states allowed by the conservation laws.
The beams of protons and antiprotons were supplied by a proton–antiproton
collider, which was specifically built for this purpose. At the time it had proton and
antiproton beams with maximum energies of 270 GeV each, giving a total centre-
of-mass energy of 540 GeV. Two independent experiments were mounted (called

6
See, for example, Chapter 12 of Ha84.
 W
AND Z 0 BOSONS 193

UA1 and UA2), both of which were examples of the ‘layered’ detector systems
that were discussed in Chapter 4.7 One of the main problems facing the
experimenters was that for each event in which a W
or Z 0 is produced and
decays to leptons, there were more than 107 events in which hadrons alone are
produced and so the extraction of the signal required considerable care.
In contrast to the zero mass photons and gluons, the W
and Z 0 bosons are both
very massive particles, with measured masses

MW ¼ 80:6 GeV=c2 ; MZ ¼ 91:2 GeV=c2 ; ð6:21Þ

while their lifetimes are about 3  1025 s. Their dominant decays lead to jets of
hadrons, but the leptonic decays

W þ ! ‘þ þ
‘ ; W  ! ‘ þ
‘ ð6:22Þ

and

Z 0 ! ‘þ þ ‘ ; Z 0 !
‘ þ
l ; ð6:23Þ

where ‘ ¼ e,  or as usual, are also important. The particles are detected as

resonance-like enhancements in plots of the invariant mass of suitable final states
seen in reactions such as Equation (6.20).8
We have seen that an important feature of an exchange interaction is its strength.
As in the case of electromagnetism, Feynman diagrams for weak interactions are
constructed from fundamental three-line vertices. Those for lepton–W
interac-
tions are shown in Figure 6.9.

Figure 6.9 The two basic vertices for W
-lepton interactions

7
Simon van der Meer lead the team that built the accelerator and Carlo Rubbia lead the UA1 experimental
team that subsequently discovered the bosons. They shared the 1984 Nobel Prize in Physics for their work.
8
A more detailed description of the UA1 experiment is given in, for example, Section 8.1 of Ma97.
194 CH6 ELECTROWEAK INTERACTIONS

At each vertex a boson is emitted or absorbed; while both fermion lines belong
to the same generation ‘ ¼ e,  or , with one arrow pointing inwards and one
outwards to guarantee conservation of each lepton number Ne, N and N .
Finally, associated with each vertex is a dimensionless parameter with the same
value

W ¼ g2W =4
hc 1=400 ð6:24Þ

at high energies for all three generations (because of lepton universality). This
constant is the weak analogue of the fine structure constant 1=137 in
electromagnetic interactions, with gW the weak analogue of the electronic charge
e in appropriate units.
We see from the above that, despite its name, the weak interaction has a similar
intrinsic strength to the electromagnetic interaction. Its apparent weakness in many
low-energy reactions, is solely a consequence of its short range, which arises
because the exchange bosons are heavy. From Equation (6.1) we see that the
scattering amplitude has a denominator that contains the squared mass of the
exchanged particle and so at energies where the de Broglie wavelengths  ¼ h=p
of the particles are large compared with the range of the weak interaction, which is
an excellent approximation for all lepton and hadron decays, the range can be
neglected altogether. In this approximation the weak interaction becomes a point
or zero range interaction, whose effective interaction strength can be shown to be
 
eff ¼ W E
=MW c2 2 ;
 MW c2 ;
E ð6:25Þ

where E
is a typical energy scale for the process in question. (For example in muon
decay it would be the mass of the muon.) Thus we see that the interaction is both weak
and very energy dependent at ‘low energies’, but becomes comparable in strength
with the electromagnetic interaction at energies on the scale of the W-boson mass.

6.5 Weak Interactions of Hadrons

The weak decays of hadrons are understood in terms of basic processes in which
W
bosons are emitted or absorbed by their constituent quarks. In this section we
will consider both decays and scattering processes, starting with the former.

6.5.1 Semileptonic decays

A typical semileptonic decay (i.e. one that involves both hadrons and leptons) is
that of the neutron, which at the quark level is

d ! u þ e þ
e ; ð6:26Þ
 WEAK INTERACTIONS OF HADRONS 195

Figure 6.10 Quark diagram for the decay n ! pe
e

as illustrated in Figure 6.10, where the other two quarks play the role of spectators.
Similarly, in the pion decay process

 ðd
uÞ !  þ
 ð6:27Þ

the initial quarks annihilate to produce a W boson as shown in Figure 6.11.

However, the weak interactions of quarks are more complicated than those of
leptons, and are best understood in terms of two ideas: lepton–quark symmetry, and
quark mixing.
For simplicity, we will look firstly at the case of just two generations of quarks
and leptons. In this case, lepton–quark symmetry asserts that the first two
generations of quarks
 
u c
and ð6:28Þ
d s

Figure 6.11 Quark diagram for the process  !  þ


196 CH6 ELECTROWEAK INTERACTIONS

and the first two generations of leptons

 

e

and ð6:29Þ
e 

have identical weak interactions. That is, one can obtain the basic W
quark
vertices by making the replacements
e ! u ; e ! d ;
 ! c ;  ! s in
the basic W
lepton vertices, leaving the coupling constant gW unchanged. The
resulting W
quark vertices are shown in Figure 6.12.

Figure 6.12 The W
quark vertices obtained from quark--lepton symmetry, without quark
mixing

Quark symmetry in the simple form stated above then implies that the
fundamental processes d þ
u ! W  and s þ
c ! W  occur with the same
couplings gW as the corresponding leptonic processes, i.e. in Figure (6.12) we
have gcs ¼ gud ¼ gW , while the processes s þ
u ! W  and d þ
c ! W  are
forbidden. This works quite well for many reactions, like the pion decay
 !  þ
 , but many decays that are forbidden in this simple scheme are
observed to occur, albeit at a rate which is suppressed relative to the ‘allowed’
decays. An example of this is the kaon decay K  !  þ
 , which requires a
sþ
u ! W  vertex, which is not present in the above scheme.
All these suppressed decays can be successfully incorporated into the theory by
introducing quark mixing. According to this idea, the d and s quarks participate in
the weak interactions via the linear combinations

d 0 ¼ d cos
C þ s sin
C ð6:30aÞ

and

s0 ¼ d sin
C þ s cos
C ; ð6:30bÞ
 WEAK INTERACTIONS OF HADRONS 197

where the parameter

C is called the Cabibbo angle.9 That is, lepton–quark
symmetry is assumed to apply to the doublets
 
u c
and : ð6:31Þ
d0 s0

This then generates new vertices previously forbidden. For example, the usW vertex
required for the decay K  !  þ
 arises from the interpretation of the ud 0 W
vertex shown in Figure 6.13. In a similar way a new cdW vertex is also generated.

Figure 6.13 The ud0 W vertex and its interpretation in terms of udW and usW vertices

Quark mixing enables theory and experiment to be brought into good agreement
by choosing a value
C 13 for the Cabibbo angle. One then finds that the rates
for the previously ‘allowed’ decays occur at rates which are suppressed by a factor
cos2
C 0:95, while the previously ‘forbidden’ decays are now allowed, but with
rates which are suppressed by a factor sin2
C 0:05.
Historically, the most remarkable thing about these ideas is that they were
formulated before the discovery of the charmed quark. In 1971 seven fundamental
fermions were known: the four leptons
e , e ,
 and  , and the three quarks u, d
and s. This led Glashow, Iliopolous and Maiani to propose the existence of a fourth
quark c to complete the lepton–quark symmetry and to solve problems associated
with neutral currents that we will discuss in Section 6.7. The existence of the
charmed quark was subsequently confirmed in 1974 with the discovery of the first
charmonium states (this is why their discovery was so important – see the
discussion in Section 5.3) and its measured weak couplings are consistent with
the predictions of lepton–quark symmetry and quark mixing.
We now know that there are six leptons
  

e


ð6:32Þ
e  

9
This is yet another example of physical states being mixtures of other states.
198 CH6 ELECTROWEAK INTERACTIONS

and six known quarks

  
u c t
: ð6:33Þ
d s b

When the third generation is taken into account, the mixing scheme becomes more
complicated, as we must allow for the possibility of mixing between all three
‘lower’ quarks d, s and b instead of just the first two and more parameters are
involved. In general the mixing can be written in the form
0 1 0 10 1
d0 Vud Vus Vub d
@ s0 A ¼ @ Vcd Vcs Vcb A@ s A; ð6:34Þ
b0 Vtd Vts Vtb b

where Vij ði ¼ u; c; t ; j ¼ d; s; bÞ the so-called CKM matrix,10 is unitary to ensure

the orthonormality of the new states generated by the transformation. The matrix
elements Vij are all obtainable from various decay processes and values exist for
them, although the smaller off-diagonal terms are not very well measured.11 For
the first two generations, the changes introduced by this more complex mixing are
very small. However, a new feature that emerges is the possibility of CP violation.
We shall see in the Section 6.6.1 that CP violation does actually occur in the
decays of neutral K-mesons and neutral B-mesons and it is of considerable interest
to see if the size of the violation is consistent with the CKM mixing formalism and
the standard model.

6.5.2 Neutrino scattering

Consider the elastic scattering process
e þ e !
e þ e at high energies,

proceeding via the exchange of a W-meson, i.e. a charged current weak interaction.
We know the W-meson couples only to left-handed fermions and from the
discussion of Section 6.3.1 that neutrinos have negative helicity, i.e. they are
polarized along the direction of their motion (which we will take to be the z-axis).
We also know from the work of Section 6.3.2 that in the relativistic limit, the same
is true of electrons. We are therefore led to the centre-of-mass spin/momentum
configurations before the collision shown in Figure 6.14(a). If the interaction
scatters the particles through an angle of 180 , then the centre-of-mass spin/
momentum configurations after the collision are those shown in Figure 6.14(b). In

10
The initials stand for Cabibbo, Kobayashi and Maskawa, the last two of whom extended the original
Cabibbo scheme to three generations of quarks.
11
A review is given Ei04.
 WEAK INTERACTIONS OF HADRONS 199

Figure 6.14 Spin (thick arrows) and momentum (thin arrows) configurations for
e e and
e e
interactions: (a)
e e before collision; (b)
e e after scattering through 180 ; (c)
e e before
collision; (d)
e e after scattering through 180

both cases the total spin component along the z-axis is zero. This result is true for
all angles and the scattering is isotropic.
From this we can calculate the differential cross-section using the formulae of
Chapter 1. We will assume that the squared momentum transfer Q2 is such that
Q2max  MW 2 2
c , so that the matrix element may be written [cf. Equation (6.1)]

f ð
e þ e !
e þ e Þ ¼ GF ; ð6:35Þ

where GF is the Fermi coupling constant of Equation (1.42), i.e.

4ð
hcÞ3 W
GF ¼ ð6:36Þ
ðMW c2 Þ2

and W ¼ g2 =4
hc is the equivalent of the fine structure constant for charged
current weak interactions. Hence, using Equation (1.57) and recalling that the
velocities of both the neutrino and electron are equal to c,

d 1 G2 2
ð
e e Þ ¼ 2 F 4 ECM : ð6:37Þ
d 4 ð

hcÞ

2
At high energies ECM is given by
2
ECM 2me c2 E
; ð6:38Þ

where E
is the energy of the neutrino. So finally the total cross-section is

2me c2 G2F
tot ð
e e Þ ¼ E
ð6:39Þ
ð
hcÞ4
200 CH6 ELECTROWEAK INTERACTIONS

and increases linearly with E
.12

If we apply the same argument to the scattering of antineutrinos, we are lead to
the configurations shown in Figures 6.14(c) and 6.14(d). The initial state has
Jz ¼ 1, but the final state has Jz ¼ 1. Thus scattering through 180 is forbidden
by angular momentum conservation and the amplitude must contain a factor
ð1 þ cos
Þ. This is borne out by a full calculation using the V–A formalism which
gives, in the same approximation,
d 1 G2F 2

e e Þ ¼
ð
E ð1 þ cos
Þ2 : ð6:40Þ
d 162 ð
hcÞ4 CM

Integrating over angles gives

1

e e Þ ¼ tot ð
e e Þ:
tot ð
ð6:41Þ
3
Neutrino–electron scattering is not, of course, a very practical reaction to study
experimentally, but these ideas may be taken over to deep inelastic neutrino
scattering from nucleons, where the latter are assumed to be composed of
constituent quarks whose masses may be neglected at high energies. This will
enable us to extend the discussion of Section 5.7 for charged leptons. In this case
the neutrino is assumed to interact with a single quark within the nucleon (this is
again the spectator model) and we must take account of all relevant quarks and
antiquarks. In practice we can neglect interactions with s and
s quarks as these will
be suppressed by the Cabibbo factor. So, taking into account only the u and d
quarks and their antiparticles, we can generalize Equations (6.39) and (6.41) to
give

MN c2 G2F Ev 1
tot ð
e NÞ ¼ Hþ H ð6:42aÞ
ð
hcÞ4 3

and

MN c2 G2F Ev 1
tot ð


e NÞ ¼ H þ H ; ð6:42bÞ
ð
hcÞ4 3

for scattering from an isoscalar nucleus, i.e. one with an equal number of neutrons
and protons, where MN is the mass of the nucleon. The quantities H and H are
given by
ð1 ð1
H x½uðxÞ þ dðxÞdx and H x½
uðxÞ þ d
ðxÞdx; ð6:43Þ
0 0

12
This behaviour has arisen because of the approximation Equation (6.35). It cannot of course continue
indefinitely. At very high values of Q2 the full form of the propagator would have to be taken into account
and this would introduce an energy dependence in the denominator of Equation (6.39).
 NEUTRAL MESON DECAYS 201

where uðxÞ etc. are the quark densities defined in Section 5.7 and the integral is
over the scaling variable x.
Setting y ¼ H=H, we have from Equations (6.42)

e NÞ 1 þ 3y
ð

R ¼ : ð6:44Þ
ð
e NÞ 3þy

Some data for R are shown in Figure 6.15 from an experiment using muon–
neutrinos. These show that R is approximately constant, as predicted by
Equation (6.44), and has a value of about 0.51, which implies y 0:2, i.e.
antiquarks exist in the nucleon at the level of about 20 per cent. Other experiments
yield similar results in the range 15–20 per cent.

Figure 6.15 Neutrino and antineutrino total cross-sections (data from Se97)

6.6 Neutral Meson Decays

Neutral mesons are of particular interest not only because they enable very
sensitive tests of CP-conservation to be made, but also because the application
of basic quantum mechanics leads to surprising effects that, for example, allow the
symmetry between particles and antiparticles to be tested with extraordinary
precision. In both cases the crucial ingredient is the phenomenon of particle
mixing that we have met before in connection with the mixing of neutrino flavour
202 CH6 ELECTROWEAK INTERACTIONS

states. Because most work has been done on the neutral kaons, we will mainly
discuss this system as an example. The equivalent formalisms for B- and D-decays
are similar. We start with a discussion of CP violation.

6.6.1 CP violation

We have seen that there are two neutral kaon states

0
K 0 ð498Þ ¼ d
s and K ð498Þ ¼ sd
; ð6:45Þ
which have strangeness S ¼ þ1 and S ¼ 1 respectively. However, because
strangeness is not conserved in weak interactions, these states can be converted
into each other by higher-order weak processes like those shown in Figure 6.16.

Figure 6.16 Example of a process that can convert a K 0 state to a K
0 state

This is in marked contrast to most other particle–antiparticle systems, for which

such transitions are forbidden, because the particle and its antiparticle differ by
quantum numbers that are conserved in all known interactions. For example, the
þ and  have opposite electric charges, and the neutron and antineutron have
opposite baryon numbers. For neutral kaons, however, there is no conserved
0
quantum number to distinguish the K 0 and K states when weak interactions are
taken into account and the observed physical particles correspond not to the K 0
and K 0 states themselves, but to linear combinations of them. Similar mixing can
0
occur between B0  B0 and D0  D states. We have met the idea that observed
states can be linear combinations of other states in the CKM mixing scheme for
quarks above and earlier when we discussed neutrino oscillations in the absence of
lepton number conservation in Chapter 3. In the present case it leads to the
0
phenomena of K 0  K mixing, and strangeness oscillations.
We start by assuming that CP-invariance is exact and that neutral kaons are
eigenstates of the combined CP operator. In this case, using the standard phase
convention, we can define
0 0
CjK 0 ; pi ¼ jK ; pi; CjK ; pi ¼ jK 0 ; pi; ð6:46Þ

where jK 0 ; pi denotes a K 0 state with momentum p, etc.. Since kaons have

negative intrinsic parity, we also have for p ¼ 0
0 0
PjK 0 ; 0i ¼ jK 0 ; 0i ; PjK ; 0i ¼ jK ; 0i; ð6:47Þ
 NEUTRAL MESON DECAYS 203

so that
0 0
CPjK 0 ; 0i ¼ jK ; 0i; CPjK ; 0i ¼ jK 0 ; 0i: ð6:48Þ

0
Thus CP eigenstates K1;2 are

0 1 n 0
o
jK1;2 ; 0i ¼ pffiffiffi jK 0 ; 0i
jK ; j0i ðCP ¼
1Þ: ð6:49Þ
2

If CP is conserved, then K10 should decay entirely to states with CP ¼ 1 and K20
should decay entirely into states with CP ¼ 1. We examine the consequences of
this for decays leading to pions in the final state.
Consider the state 0 0 . Since the kaon has spin-0, by angular momentum
conservation the pion pair must have zero orbital angular momentum in the rest
frame of the decaying particle. Its parity is therefore given by [cf. Equation (1.14)]

P ¼ P2 ð1ÞL ¼ 1; ð6:50Þ

where P ¼ 1 is the intrinsic parity of the pion. The C-parity is given by

C ¼ ðC0 Þ2 ¼ 1; ð6:51Þ

where C0 ¼ 1 is the C-parity of the neutral pion. Combining these results gives
CP ¼ 1. The same result holds for the þ  final state.
The argument for three-pion final states þ  0 and 0 0 0 is more compli-
cated, because there are two orbital angular momenta to consider, If we denote by
L12 the orbital angular momentum of one pair (either þ  or 0 0 ) in their
mutual centre-of-mass frame, and L3 is the orbital angular momentum of the third
pion about the centre-of-mass of the pair in the overall centre-of-mass frame, then
the total orbital angular momentum L L12 þ L3 ¼ 0, since the decaying particle
has spin-0. This can only be satisfied if L12 ¼ L3 . This implies that the parity of the
final state is

P ¼ P3 ð1ÞL12 ð1ÞL3 ¼ 1: ð6:52Þ

For the 0 0 0 final state, the C-parity is

C ¼ ðC0 Þ3 ¼ 1 ð6:53Þ

and combining these results gives CP ¼ 1 overall. The same result can be shown
to hold for the þ  0 final state.
The experimental position is that two neutral kaons are observed, called K 0 -short
and K 0 -long, denoted KS0 and KL0 , respectively. They have almost equal masses of
about 499 MeV/c2, but very different lifetimes and decay modes. The KS0 has a
204 CH6 ELECTROWEAK INTERACTIONS

lifetime of 0:89  1010 s and decays overwhelmingly to two pions; the longer-
lived KL0 has a lifetime of 0:52  107 s with a significant branching ratio to three
pions, but not to two. In view of the CP analysis above, this immediately suggests
the identification

KS0 ¼ K10 ; and KL0 ¼ K20 : ð6:54Þ

However, in 1964 it was discovered that the KL0 also decayed to two pions13

KL0 ! þ þ  ; ð6:55Þ

but with a very small branching ratio of the order of 103 . This result is clear
evidence of CP violation. This was confirmed in later experiments on the decay
K 0 ! 0 0 .
Because CP is not conserved, the physical states KS0 and KL0 need not correspond
to the CP-eigenstates K10 and K20 , but can contain small components of states with
the opposite CP, i.e. we may write

1  
jKS0 ; 0i ¼ jK10 ; 0i  "jK20 ; 0i ð6:56aÞ
ð1 þ j"j2 Þ1=2

and

1  
jKL0 ; 0i ¼ 2 1=2
"jK10 ; 0 i þ jK20 ; 0 i ; ð6:56bÞ
ð1 þ j"j Þ

where e is a small complex parameter. (The factor in front of the brackets is to

normalize the states.) The CP-violating decays can then occur in two different
ways: either (a) the CP-forbidden K10 component in the KL0 decays via a CP-
allowed processe, giving a contribution proportional to the probability
j" j2 ½1 þ j" j2 1 j" j2 of finding a K10 component in the KL0 ; or (b) the CP-
allowed K20 component in the KL0 decays via a CP-violating reaction. A detailed
analysis of the data for the  decay modes14 shows that it is the former
mechanism that dominates, with j"j 2:2  103 .
This is confirmed in the semileptonic decays

K 0 !  þ eþ þ
e ð6:57aÞ

13
The experiment was led by James Cronin and Val Fitch. They received the 1980 Nobel Prize in Physics for
their discovery.
14
See, for example, Ei04.
 NEUTRAL MESON DECAYS 205

and

0 ! þ þ e þ
e :
K ð6:57bÞ

For example, if we start with a beam of K 0 particles, with initially equal amounts
of KS0 and KL0 , then after a time that is large compared with the KS0 lifetime, the KS0
component will have decayed leaving just the KL0 component, which itself will be
an equal admixture of K 0 and K
0 components. We would therefore expect to
observe identical numbers of electrons (N  ) and positrons (N þ ) from the decays of
Equations (6.57). However, if KL0 is not an eigenstate of CP, then there will be an
asymmetry in these numbers, which will depend on the relative strengths of the K 0
and K
0 components in KL0 . The asymmetry is given by 2Re ", where " is the CP-
violating parameter defined in Equation (6.56).
Figure 6.17 shows data on the asymmetry ðN þ  N  Þ=ðN þ þ N  Þ as a function
of proper time. After the initial oscillations there is seen to be an asymmetry whose
value is 2Re " 3:3  103 , which is consistent with the value of " obtained
from the  modes. Thus CP-violation in K-decay occurs mainly, though not
entirely, by the mixing of the CP-eigenstates in the physical states rather than by
direct CP-violating decays, both of which are allowed in the CKM mixing scheme.
What do these results mean for the CKM mixing scheme? The CKM matrix is a
3  3 matrix and in general contains nine complex elements. However, the unitary
nature of the matrix implies that there are relations between the elements, such as
 
Vud Vub þ Vcd Vcb þ Vtd Vtb ¼ 0: ð6:58Þ

Figure 6.17 The charge asymmetry observed for K 0 !  eþ
e and K
0 ! þ e
e as a function
of proper time, for a beam that is initially predominantly K 0 (adapted from Gj74, copyright
Elsevier, with permission)
206 CH6 ELECTROWEAK INTERACTIONS

Using these and exploiting the freedom to define the phases of the basic quark
states, the matrix may be parameterized by just four quantities. A number of
different parameterizations are used, but an approximate form that is commonly
used to discuss CP violation is
0 1
1  12 2  A3 ð  iÞ
V ¼@  1  12 2 A2 A þ Oð4 Þ; ð6:59Þ
3 2
A ð1    iÞ A 1

with parameters A; ;  and . The quantity  ¼ jVus j 0:22 plays the role of an
expansion parameter in this approximation and a non-zero value of  would be
indicative of CP violation.
The parameter " in Equations (6.56) is just one CP-violating parameter that may
be measured in various K-decay modes. We will not pursue this further, but note
that by combining the values of the parameters with information on other elements
of the CKM matrix, a value of the CP-violating parameter  may be deduced and
used to predict the size of CP-violating effects in other decays. There are very few
other places where such mixing effects can occur, but in principle they should be

0 and B0 B
possible in the D0 D
0 systems, which are analogues of the K-mesons, but
with a strange quark replaced by a charmed and bottom quark, respectively.
Mixing in the B0 B
0 states due to B0 B
0 oscillations has in fact been observed and
also very recently direct CP violation. The latter was established by comparing the
decay B0 ! K þ  with the decay B
0 ! K  þ. Moreover, the size of the effect is
much stronger than in neutral kaon decays and this is in agreement with the
predictions of the CKM mixing scheme.
There is still much to be done in studying CP violation. For example,
the cleanest measurement of the CP-violating parameter would be from the
decays B0 =B
! ðJ= ÞKS0 , where J= is the 3 S1 ground state of charmonium,
but the present limits on these decays are orders of magnitude from those required
to test the predictions. On the theoretical side, although the CKM mixing model
accounts for all CP-violating data to date, it fails by several orders of magnitude to
account for the observed matter–antimatter asymmetry observed in the universe
(which will be discussed in Chapter 9) and so there is probably a CP-violating
mechanism beyond the standard model awaiting to be discovered.

6.6.2 Flavour oscillations

One interesting consequence of flavour mixing for the K 0 – K
0 system is the

phenomenon of strangeness oscillation, which occurs whenever a neutral kaon is
produced in a strong interaction process. For example, the neutral kaon produced
in the strong interaction
 þ p ! K 0 þ 0
ð6:60Þ
S¼0 0 1 1
 NEUTRAL MESON DECAYS 207

must necessarily be a K 0 state with S ¼ 1, in order to conserve strangeness.

However, if the produced particle is allowed to travel through free space and its
strangeness is measured, one finds that it no longer has a definite strangeness
S ¼ 1, but has components with both S ¼ 1 and S ¼ 1 whose intensities oscillate
with time. These are called strangeness oscillations. The phenomenon is very
similar mathematically to that describing the flavour oscillations of neutrinos we
met in Chapter 3 and enables the mass difference between KS0 and KL0 particles to
be measured with extraordinary precision, as we will now show.
In what follows, we shall measure time in the rest frame of the produced
particle, and define t ¼ 0 as the moment when it is produced. If we ignore the very
small CP violations, the initial state produced in the  p reaction above is
1  
jK 0 ; pi ¼ pffiffiffi jKS0 ; pi þ jKL0 ; pi : ð6:61Þ
2
At later times, however, this will become
1 
pffiffiffi aS ðtÞjKS0 ; pi þ aL ðtÞjKL0 ; pig; ð6:62Þ
2
where

a ðtÞ ¼ eim t e t=2 ða ¼ S; LÞ ð6:63Þ

and m and  are the mass and decay rate of the particle concerned. Here the first
exponential factor is the usual oscillating time factor eiEt associated with any
quantum mechanical stationary state, evaluated in the rest frame of the particle.
The second exponential factor reflects the fact that the particles decay, and it
ensures that the probability
 2
 1 
 pffiffiffi a ðtÞ ¼ 1 e t ða ¼ S; LÞ ð6:64Þ
 2  2

of finding a KS0 or KL0 decreases exponentially with a mean lifetime ¼ 1

(a ¼ S, L). Because S  L , for times t such that S  t <
L only the KL0

0 components.
component survives, implying equal intensities for the K 0 and K
0
components at shorter
Here we are interested in the intensities of the K and K 0

times, and to deduce these we rewrite the expression

1  
pffiffiffi aS ðtÞjKS0 ; pi þ aL ðtÞjKL0 ; pi ð6:65Þ
2

in the form
 

0 ðtÞjK 0 ; pi ;
A0 ðtÞjK 0 ; pi þ A ð6:66Þ
208 CH6 ELECTROWEAK INTERACTIONS

where
1
0 ðtÞ ¼ 1 ½aS ðtÞ  aL ðtÞ:
A0 ðtÞ ¼ ½aS ðtÞ þ aL ðtÞ and A ð6:67Þ
2 2
The intensities of the two components are then given by
1h i
IðK 0 Þ jA0 ðtÞj2 ¼ eS t þ eL t þ 2eðS þL Þt=2 cosðmtÞ ð6:68aÞ
4
and
0
0 ðtÞj2 ¼ 1 h S t i
IðK Þ jA e þ eL t  2eðS þL Þt=2 cosðmtÞ ð6:68bÞ
4
where m jmS  mL j and we have used Equation (6.63) explicitly to evaluate
the amplitudes.
0 0
The variation of the K intensity IðK Þ with time can be determined experi-
mentally by measuring the rate of production of hyperons (baryons with non-zero
strangeness) in strangeness-conserving strong interactions such as
0
K þ p !  þ þ 0
ð6:69Þ
! 0 þ  þ

as a function of the distance from the K 0 source. The data are then fitted by
Equations (6.68) with m as a free parameter and the predictions are in good
agreement with the experiments for a mass difference

m ¼ ð3:522
0:016Þ  1012 MeV=c2 : ð6:70Þ

0
The states KS0 and KL0 are not antiparticles but the K 0 and K are, of course, and the
mass difference m can be shown to arise solely from the possibility of transitions
0
K $ K , whose magnitude can be calculated from diagrams like that shown in
Figure 6.16. We shall not discuss this further, but merely note that the resulting
agreement between the predicted and measured values confirms the equality
mK 0 ¼ mK
0 to better than one part in 1017 . (This should be compared with the
next most precisely tested particle–antiparticle mass relation meþ ¼ me which
only verified to within an experimental error of the order of one part in 107 .) This
equality is a prediction of the so-called CPT theorem, which states that under very
general conditions any relativistic quantum theory will be invariant under the
combined operations of C, P and T.

6.7 Neutral Currents and the Unified Theory

Neutral current reactions are those that involve the emission, absorption or
exchange of Z 0 bosons. The unified electroweak theory predicted the existence
 NEUTRAL CURRENTS AND THE UNIFIED THEORY 209

Figure 6.18 Higher order contribution to the reaction eþ  ! eþ  from the exchange of
two W-bosons

of such reactions before their discovery in 1973. This theory15 was proposed
mainly in order to solve problems associated with Feynman diagrams in which
more than one W boson was exchanged, like that shown in Figure 6.18, which
contributes to the reaction eþ  ! eþ  .
Such contributions are expected to be small because they are higher order in the
weak interaction and this appears to be confirmed by experimental data, which are
in good agreement with theoretical predictions that neglect them entirely. (For
example, in the experimentally accessible reaction eþ þ e ! þ þ  .) How-
ever, when these higher-order contributions are explicitly calculated, they are
found to be proportional to divergent integrals, i.e. they are infinite. In the unified
theory, this problem is solved when diagrams involving the exchange of Z 0 bosons
and photons are taken into account. These also give infinite contributions, but
when all the diagrams of a given order are added together the divergences
cancel (!), giving a well-defined and finite contribution overall.16 This cancellation
is not accidental, but is a consequence of a fundamental symmetry relating the
weak and electromagnetic interactions. Here we will simply comment on some
phenomenological consequences of the theory.

15
The formulation of the theory is in terms of four massless vector bosons arranged as multiplets of ‘weak
isospin’ and ‘weak hypercharge’. Specifically, three states are a weak isospin triplet and the fourth is a weak
isospin singlet. The fact that they all have zero masses ensures that gauge invariance is satisfied. These fields
then interact with additional scalar fields associated with new postulated particles called Higgs bosons, which
we have mentioned elsewhere. This process, known as ‘spontaneous symmetry breaking’ generates the
observed masses of the W, Z and  bosons, while still preserving gauge invariance. (For further details see,
for example, Section 8.4 of Pe00.) The originators of this theory, Sheldon Glasow, Abdus Salam and Steven
Weinberg, shared the 1979 Nobel Prize in Physics for their contributions to formulating the electroweak
theory and the prediction of weak neutral currents.
16
The first people to demonstrate that this occurred were Gerardus ‘t Hooft and Martinus Veltman. They
shared the 1999 Nobel Prize in Physics for this discovery.
210 CH6 ELECTROWEAK INTERACTIONS

The first is that to ensure the cancellation, the theory requires a relation between
the weak and electromagnetic couplings, called the unification condition. This is
e
pffiffiffi 1=2 ¼ gw sin
W ¼ gz cos
W ; ð6:71Þ
2 2"0

where the weak mixing angle

W (also called the Weinberg angle after one of the
authors of the theory) is given by

cos
W MW =MZ ð0 <
< =2Þ ð6:72Þ

and gz is a coupling constant which characterizes the strength of the neutral current
vertices. The unification condition relates the strengths of the various interactions
to the W and Z masses, and historically was used to predict the latter from the
former before the W
and Z 0 bosons were discovered.
Secondly, just as all the charged current interactions of leptons can be under-
stood in terms of the basic W
-lepton vertices, in the same way all known neutral
current interactions can be accounted for in terms of basic Z 0 -lepton vertices
shown in Figures 6.19(a) and 16.19(b). The corresponding quark vertices can be
obtained from the lepton vertices by using lepton–quark symmetry and quark
mixing, in the same way that W
-quark vertices are obtained from the W
-lepton
vertices. Thus, making the replacements

e ! u;
 ! c; e ! d 0 ;  ! s0 ð6:73Þ

in the lepton vertices

e
e Z 0 ;

 Z 0 ; e e Z 0 ;   Z 0 ; ð6:74Þ

Figure 6.19 Z 0 and  couplings to leptons and quarks in the unified electroweak theory, where
‘ ¼ e;  and and a denotes a quark
 NEUTRAL CURRENTS AND THE UNIFIED THEORY 211

leads to the quark vertices

uuZ 0 ; ccZ 0 ; d0 d0 Z 0 ; s0 s0 Z 0 : ð6:75Þ

Finally, we interpret the latter two of these using Equations (6.30). Thus, for
example,

d 0 d 0 Z 0 ¼ ðd cos
C þ s sin
C Þ ðd cos
C þ s sin
C ÞZ 0
¼ ddZ 0 cos2
C þ ssZ 0 sin2
C þ ðdsZ 0 þ sdZ 0 Þ sin
C cos
C ð6:76Þ

When all the terms in Expression (6.75) are evaluated, ones obtains a set of
vertices equivalent to

uuZ 0 ; ccZ 0 ; ddZ 0 ; ssZ 0 ; ð6:77Þ

which are shown in Figure 6.19(c).

One important difference from charged current reactions that follows from
Figure 6.19 is that neutral current interactions conserve individual quark numbers.
Thus, for example, strangeness-changing weak neutral current reactions are
forbidden. An example is the decay K 0 ! þ  and indeed this is not seen
experimentally, although nothing else forbids it.
It follows from the above that in any process in which a photon is exchanged, a
Z 0 boson can be exchanged as well. At energies that are small compared with the
Z 0 mass, the Z 0 -exchange contributions can be neglected compared to the
corresponding photon exchange contributions, and these reactions can be regarded
as purely electromagnetic to a high degree of accuracy. However, at very high
energy and momentum transfers, Z 0 -exchange contributions become comparable
with those of photon exchange and we are therefore dealing with genuinely
electroweak processes which involve both weak and electromagnetic interactions
to a comparable degree.
These points are clearly illustrated by the cross-section for the muon pair
production reaction eþ þ e ! þ þ  . If we assume that the energy is large
enough for the lepton masses to be neglected, then the centre-of-mass energy E is
the only quantity in the system that has dimensions. Because a cross-section has
the dimensions of area, on dimensional grounds the electromagnetic cross-section
for one-photon exchange is of the form  2 ð
hcÞ2 =E2 . For Z 0 -exchange with
E  MZ c2 , a similar argument gives for the weak interaction cross-section
hcÞ2 =ðMZ c2 Þ4 . (The factor in the denominator comes from the propa-
Z 2Z E2 ð

gator of the Z 0 -boson.) Thus the one-photon exchange diagram dominates at low
energies, and the cross-section falls as E2 . This is in agreement with the observed
behaviour shown in Figure 6.20 and justifies our neglect of the Z 0 -exchange
contribution at low energies. However, the relative importance of the Z 0 -exchange
contribution increases rapidly with energy and at beam energies of about 35 GeV it
212 CH6 ELECTROWEAK INTERACTIONS

Figure 6.20 Total cross-section for the reaction eþ e ! þ 

begins to make a significant contribution to the total cross-section. At still higher

energies, the cross-section is dominated by a very large peak at an energy
corresponding to the Z 0 mass, as illustrated in Figure 6.20. At this energy the
low-energy approximation is irrelevant and Figure 6.20 corresponds to
the formation of physical Z 0 bosons in the process eþ þ e ! Z 0 followed
by the subsequent decay Z 0 ! þ þ  to give the final-state muons. Finally,
beyond the peak we once again regain the electroweak regime where both
contributions are comparable.
If the exchange of a Z 0 boson always accompanies the exchange of a photon,
then there will also in principle be parity-violating effects in reactions that at first
sight we would expect to be purely electromagnetic. Their observation would be
further unambiguous evidence for electroweak unification. This was first tested in
1978 by scattering polarized electrons from a deuterium target and measuring the
parity-violating asymmetry

R  L
APV ; ð6:78Þ
R þ L

where R ðL Þ is the cross-section for incident right (left)-handed electrons. To

produce polarized electrons is a complicated multistage process that starts with
linearly polarized photons from a laser source that are then converted to states with
circular polarization. Finally these are used to pump a GaAs crystal (photocathode)
to produce the require electrons. Polarizations of about 80 per cent are obtained by
this means. The asymmetry is very small and in this experiment APV is predicted to
be only a few parts per million. Nevertheless, a non-zero value was definitely
established. Moreover, APV was also measured as a function of the fractional
energy loss of the initial electron. This is a function of the weak mixing angle and a
 PROBLEMS 213

value was found in agreement with other determinations, e.g. from deep inelastic
neutrino scattering. A later experiment confirmed the effect in polarized electron–
proton scattering.
A very recent experiment (2004) has measured APV for e e scattering. This
was done using electrons of about 50 GeV primary energy from the SLAC linear
accelerator in Stanford, USA, and scattering them from a liquid hydrogen target.
The experiment was able to distinguish final-state electrons scattered from the
atomic electrons from those scattered from protons. Taking account of all sources
of error, the measured value was APV ¼ ð175
40Þ  109 (note the exponent –
parts per billion) and the experiment also yielded a value of sin2
W consistent with
other determinations. These remarkable experiments provide unambiguous evi-
dence for parity violation in ‘electromagnetic’ processes at the level predicted by
theory and hence for the electroweak unification as specified in the standard
model.17
It should also in principle be possible to detect parity violating effects in atomic
physics, where the electromagnetic interactions of the electrons dominate. For
example, measurements have been made of the slight change in the polarization
angle of light passing through a vapour of metallic atoms. In this case the rotation
angle is extremely small (
107 rad), but very sensitive experiments can measure
the effect to an accuracy of
1 per cent. However, to predict the size of the effects
requires a detailed knowledge of the atomic theory of the atom and in all cases to
date the uncertainties on the predictions are such that a null effect cannot be ruled
out. Thus at present, atomic physics does not compete with particle physics
experiments in detecting parity-violating effects and measuring sin2
W , although
this could change in the future.

Problems
6.1 Define charged and neutral current reactions in weak interactions and give an
example of each in symbol form. How do they differ in respect of conservation of
the strangeness quantum number? Why does observation of the process

 þ e !
 þ e constitute unambiguous evidence for weak neutral currents,
whereas the observation of
e þ e !
e þ e does not?

6.2 The reaction eþ e ! þ  is studied using colliding beams each of energy 7 GeV
and at these energies the reaction is predominantly electromagnetic. Draw its lowest
order Feynman diagram. The differential cross-section is given by

d h 2 c2 
2

¼ 2
1 þ cos2
;
d 4ECM

17
Incidentally, all these experiments are of the fixed-target type, showing that this type of experiment still has
a lot to offer.
214 CH6 ELECTROWEAK INTERACTIONS

where ECM is the total centre-of-mass energy and

is the scattering angle with
respect to the beam direction. Calculate the total cross-section in nanobarns at this
energy.
The weak interaction also contributes to this process. Draw the corresponding
lowest-order Feynman diagram. The weak interaction adds an additional term to the
differential cross-section of the form

d h 2 c2
2

¼ 2
Cwk cos
:
d 4ECM

The constant Cwk may be determined experimentally by measuring the ‘forward–

backward’ asymmetry, defined by

F  B
AFB ¼ ;
F þ B

where F ðB Þ is the total cross-section for scattering in the forward (backward)
hemisphere, i.e. 0  cos
 1 ð1  cos
 0Þ. Derive a relation between Cwk
and AFB .

6.3 Draw a Feynman diagram at the quark level for the decay  ! p þ . If nature
were to double the weak coupling constant and decrease the mass of the W boson by
a factor of four, what would be the effect on the decay rate ð ! p þ  Þ?

6.4 Neglecting the electron mass, the energy spectrum for the electrons emitted in muon
decay is give by

d! 2G2F ðm c2 Þ2 Ee2 4Ee
¼ 1 :
dEe ð2Þ3 ð
hcÞ6 3m c2

What is the most probable energy for the electron? Draw a diagram showing the
orientation of the momenta of the three outgoing particles and their helicities for the
case when Ee m c2 =2. Show also the helicity of the muon. Integrate the energy
spectrum to obtain an expression for the total decay width of the muon. Hence
calculate the muon lifetime in seconds ðGF =ð
hcÞ3 ¼ 1:166  105 GeV2 Þ:

6.5 Use lepton universality and lepton–quark symmetry to estimate the branching ratios
for (a) the decays b ! c þ e þ
e (where the b and c quarks are bound in hadrons)
and (b)  ! e þ
e þ
. Ignore final states that are Cabibbo-suppressed relative
to the lepton modes.

6.6 The couplings of the Z 0 to right-handed (R) and left-handed (L) fermions are given by

gR ðf Þ ¼ qf sin2
W ; gL ðf Þ ¼
1=2  qf sin2
W ;

where qf is the electric charge of the fermion f in units of e and

W is the weak
mixing angle. The positive sign in gL is used for neutrinos and the q ¼ u; c; t
 PROBLEMS 215

quarks; the negative sign is used for charged leptons and the q ¼ d; s; b quarks. If
the partial width for Z 0 ! f
f is given by
GF MZ3 c6  2 
f ¼ pffiffiffi 3
gR ðf Þ þ g2L ðf Þ ;
3 2ð
hcÞ
calculate the partial widths to neutrinos 
and to q
q pairs q and explain the
relation of q to the partial width to hadrons hadron .
The widths to hadrons and to charged leptons are measured to be
had ¼ ð1738
12Þ MeV and lep ¼ ð250
2Þ MeV, and the total width to all
final states is measured to be tot ¼ ð2490
7Þ MeV. Use these experimental results
and your calculated value for the decay width to neutrinos to show that there are
only three generations of neutrinos with masses M
< MZ =2.

6.7 Explain, with the aid of Feynman diagrams, why the decay D0 ! K  þ þ can
occur as a charged-current weak interaction at lowest order, but the decay
Dþ ! K 0 þ þ cannot.

6.8 Why is the decay rate of the charged pion much smaller than that of the neutral pion?
Draw Feynman diagrams to illustrate your answer.

6.9 Draw the lowest-order Feynman diagrams for the decays  !  þ
and
K  !  þ
 . Use lepton–quark symmetry and the Cabibbo hypothesis with the
Cabibbo angle
C ¼ 12 to estimate the ratio

RateðK  !  þ
 Þ
R ;
Rateð !  þ
 Þ

ignoring all kinematic and spin effects. Comment on your result.

6.10 Estimate the ratio of decay rates

ð ! n þ e þ
e Þ
R
ð !  þ e þ
e Þ

and explain why the decay ðþ ! n þ eþ þ
e Þ has never been seen.

6.11 One way of looking for the Higgs boson H is in the reaction eþ e ! Z 0 H. If this
reaction is studied at a centre-of-mass energy of 500 GeV in a collider operating for
107 s per year and the cross-section at this energy is 60 fb, what instantaneous
luminosity (in units of cm2 s1) would be needed to collect 2000 events in one year
if the detection efficiency is 10 per cent. For a Higgs boson with mass
MH < 120 GeV, the branching ratio for H ! b
b is predicted to be 85 per cent.
Why will looking for b quarks help distinguish eþ e ! Z 0 H from the background
reaction eþ e ! Z 0 Z 0 ?

6.12 Hadronic strangeness-changing weak decays approximately obey the so-called

‘I ¼ 12 rule’, i.e. the total isospin changes by 12 in the decay. By assuming a
216 CH6 ELECTROWEAK INTERACTIONS

fictitious strangeness zero I ¼ 12 particle S0 in the initial state, find the prediction of
this rule for the ratio

ð !  þ  Þ
R :
ð0 !  þ 0 Þ

Assume that the state j0 ; S0 i is an equal mixture of states with I ¼ 0 and I ¼ 1.

6.13 The charged-current differential cross-sections for
and
scattering from a spin-12
target are given by generalizations of Equations (6.37) and (6.40) and may be written

dCC ð
Þ 1 G2 Hs dCC ð

Þ dCC ð
Þ
¼ ; ¼ ð1  yÞ2 ;
dy hcÞ4
 ð
dy dy

2
where s ¼ ECM , y ¼ 12 ð1  cos
Þ and H is the integral of the quark density for the
target (cf. Equation (6.43)). The corresponding cross-sections for neutral current
scattering are

dNC ð
Þ dCC ð
Þ 2

¼ ½gL þ g2R ð1  yÞ2 ;
dy dy
dNC ð


Þ dCC ð
Þ 2
¼ ½gL ð1  yÞ2 þ g2R ;
dy dy

where the right- and left-hand couplings to u and d quarks are given by

1 2 2
gL ðuÞ ¼  sin2
W ; gR ðuÞ ¼  sin2
W ;
2 3 3
1 1 2 1
gL ðdÞ ¼  þ sin
W ; gR ðdÞ ¼ sin2
W :
2 3 3

Derive expressions for the ratios NC ð
Þ=CC ð
Þ and NC ð


Þ=CC ð


Þ in the case of
an isoscalar target consisting of valence u and d quarks only.
7
Models and Theories
of Nuclear Physics

Nuclei are held together by the strong nuclear force between nucleons, so we start
this chapter by looking at the form of this, which is more complicated than that
generated by simple one-particle exchange. Much of the phenomenological evi-
dence comes from low-energy nucleon–nucleon scattering experiments which we
will simply quote, but we will interpret the results in terms of the fundamental strong
interaction between quarks. The rest of the chapter is devoted to various models and
theories that are constructed to explain nuclear data in particular domains.

7.1 The Nucleon -- Nucleon Potential

The existence of stable nuclei implies that overall the net nucleon–nucleon force
must be attractive and much stronger than the Coulomb force, although it cannot
be attractive for all separations, or otherwise nuclei would collapse in on
themselves. So at very short ranges there must be a repulsive core. However,
the repulsive core can be ignored in low-energy nuclear structure problems
because low-energy particles cannot probe the short-distance behaviour of the
potential. In lowest order, the potential may be represented dominantly by a central
term (i.e. one that is a function only of the radial separation of the particles),
although there is also a smaller non-central part. We know from proton–proton
scattering experiments1 that the nucleon–nucleon force is short-range, of the same
order as the size of the nucleus, and thus does not correspond to the exchange of
gluons, as in the fundamental strong interaction. A schematic diagram of the
resulting potential is shown in Figure 7.1. In practice of course this strong

1
For reviews see, for example, Chapter 7 of Je90 and Chapter 14 of Ho97.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
218 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

Figure 7.1 Idealized square well representation of the strong interaction nucleon--nucleon
potential. The distance R is the range of the nuclear force and
 R is the distance at which the
short-range repulsion becomes important. The depth V0 is approximately 40 MeV

interaction potential must be combined with the Coulomb potential in the case in
the case of protons.
A comparison of nn and pp scattering data (after allowing for the Coulomb
interaction) shows that the nuclear force is charge-symmetric (pp ¼ nn) and
almost charge-independent ðpp ¼ nn ¼ pnÞ.2 We have commented in Chapter 3
that there is also evidence for this from nuclear physics. Charge-symmetry is seen in
comparisons of the energy levels of mirror nuclei (see, for example, Figure 3.9) and
evidence for charge-independence comes from the energy levels of triplets of
related nuclei with the same A values. Nucleon–nucleon forces are, however,
spin-dependent. The force between a proton and neutron in an overall spin-1 state
(i.e. with spins parallel) is strong enough to support a weakly bound state (the
deuteron), whereas the potential corresponding to the spin-0 state (i.e. spins
antiparallel) has no bound states. Finally, nuclear forces saturate. This describes
that fact that a nucleon in a typical nucleus experiences attractive interactions only
with a limited number of the many other nucleons and is a consequence of the short-
range nature of the force. The evidence for this is the form of the nuclear binding
energy curve and was discussed in Chapter 2.
Ideally one would like to be able to interpret the nucleon–nucleon potential in
terms of the fundamental strong quark–quark interactions. It is not yet possible to
give a complete explanation along these lines, but it is possible to go some way in
this direction. If we draw an analogy with atomic and molecular structure, with

2
For a discussion of these data see, for example, the references in Footnote 1.
 THE NUCLEON--NUCLEON POTENTIAL 219

quarks playing the role of electrons, then possibilities are: an ionic-type bond, a van
der Waals type of force, or a covalent bond.3 The first can be ruled out because the
confining forces are too strong to permit a quark to be ‘lent’ from one nucleon to
another and the second can also be ruled out because the resulting two-gluon
exchange is too weak. This leaves a covalent bond due to the sharing of single quarks
between the nucleons analogous to the covalent bond that binds the hydrogen
molecule. However, nucleons have to remain ‘colourless’ during this process and so
the shared quark from one nucleon has to have the same colour as the shared quark
from the other nucleon. The effect of this is to reduce the effective force (because
there are three possible colour states) and by itself it is unable to explain the depth
of the observed potential. In addition to the three (valence) quarks within the nucleon
there are also present quark–antiquark pairs due to vacuum fluctuations.4 Such pairs
can be colourless and so can also be shared between the nucleons. These quarks
actually play a greater role in generating the nuclear strong interaction than single
quarks. The lightest such diquarks will be pions and this exchange gives the largest
contribution to the attractive part of the nucleon–nucleon force (see, for example, the
Feynman diagram Figure 1.4).
In principle, the short-range repulsion could be due to the exchange of heavier
diquarks (i.e. mesons), possibly also in different overall spin states. Experiment
provides many suitable meson candidates, in agreement with the predictions of the
quark model, and each exchange would give rise to a specific contribution to the
overall nucleon–nucleon potential, by analogy with the Yukawa potential resulting
from the exchange of a spin-0 meson, as discussed in Chapter 1. It is indeed possible
to obtain excellent fits to nucleon–nucleon scattering data in a model with several
such exchanges.5 Thus this approach can yield a satisfactory potential model, but is
semi-phenomenological only, as it requires the couplings of each of the exchanged
particles to be found by fitting nucleon–nucleon scattering data. (The couplings that
result broadly agree with values found from other sources.) Boson-exchange models
therefore cannot give a fundamental explanation of the repulsion. The reason for
the repulsion at small separations in the quark model lies in the spin dependence
of the quark–quark strong interaction, which like the phenomenological nucleon–
nucleon interaction, is strongly spin-dependent. We have discussed this in the
context of calculating hadron masses in Section 3.3.3. When the two nucleons are
very close, the wavefunction is effectively that for a 6-quark system with zero
angular momentum between the quarks, i.e. a symmetric spatial wave function.
Since the colour wave function is antisymmetric, (recall the discussion of
Chapter 5), it follows that the spin wavefunction is symmetric. However, the

3
Recall from chemistry that in ionic bonding, electrons are permanently transferred between constituents to
form positive and negative ions that then bind by electrostatic attraction; in covalent bonding the constituents
share electrons; and the van der Waals force is generated by the attraction between temporary charges
induced on the constituents by virtue of slight movements of the electrons.
4
These are the ‘sea’ quarks mentioned in connection with the quark model in Chapter 3 and which are probed
in deep inelastic lepton scattering that was discussed in Chapter 6.
5
This approach is discussed in, for example, Chapter 3 of Co01 and also in the references given in Footnote1.
220 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

potential energy increases if all the quarks remain in the L ¼ 0 state with
spins aligned.6 The two-nucleon system will try to minimize its ‘chromomagnetic’
energy, but this will compete with the need to have a symmetric spin wavefunction.
The optimum configuration at small separations is when one pair of quarks is in
an L ¼ 1 state, although the excitation energy is comparable to the decrease in
chromomagnetic energy, so there will still be a net increase in energy at small
separations.
Some tantalizing clues exist about the role of the quark–gluon interaction in
nuclear interactions, such as the small nuclear effects in deep inelastic lepton
scattering mentioned in Chapter 5. There is also a considerable experimental
programme in existence (for example at CEBAF, the superconducting accelerator
facility at the Jefferson Laboratory, Virginia, USA, mentioned in Chapter 4) to learn
more about the nature of the strong nucleon–nucleon force in terms of the
fundamental quark–gluon strong interaction and further progress in this area may
well result in the next few years. Meanwhile, in the absence of a fundamental theory
to describe the nuclear force, specific models and theories are used to interpret the
phenomena in different areas of nuclear physics. In the remainder of this chapter we
will discuss a number of such approaches.

7.2 Fermi Gas Model

In this model, the protons and neutrons that make up the nucleus are assumed to
comprise two independent systems of nucleons, each freely moving inside the
nuclear volume subject to the constraints of the Pauli principle. The potential felt by
every nucleon is the superposition of the potentials due to all the other nucleons. In
the case of neutrons this is assumed to be a finite-depth square well; for protons, the
Coulomb potential modifies this. A sketch of the potential wells in both cases is
shown in Figure 7.2.
For a given ground state nucleus, the energy levels will fill up from the bottom of
the well. The energy of the highest level that is completely filled is called the
Fermi level of energy EF and has a momentum pF ¼ ð2MEF Þ1=2 , where M is the
mass of the nucleon. Within the volume V, the number of states with a momentum
between p and p þ dp is given by the density of states factor

4V
nðpÞdp ¼ dn ¼ p2 dp; ð7:1Þ
ð2
hÞ3

6
Compare the mass of the (1232) resonance, where all three quarks spins are aligned, to that of the lighter
nucleon, where one pair of quarks spins is anti-aligned to give a total spin of zero. This is discussed in detail
in Section 3.3.3.
 FERMI GAS MODEL 221

Figure 7.2 Proton and neutron potentials and states in the Fermi gas model

which is derived in Appendix A. Since every

Ð p state can contain two fermions of the
same species, we can have (using n ¼ 2 0 F dnÞ

VðpnF Þ3 VðppF Þ3
N¼ and Z ¼ ð7:2Þ
32
h3 32
h3

neutrons and protons, respectively, with a nuclear volume

4 4
V ¼ R3 ¼ R30 A; ð7:3Þ
3 3

where experimentally R0 ¼ 1:21 fm, as we have seen from electron and hadron
scattering experiments discussed in Chapter 2. Assuming for the moment that the
depths of the neutron and proton wells are the same, we find for a nucleus with
Z ¼ N ¼ A=2, the Fermi momentum
 
h
9 1=3
pF ¼ pnF ¼ ppF ¼  250 MeV=c: ð7:4Þ
R0 8

Thus the nucleons move freely within the nucleus with quite large momenta.
The Fermi energy is

p2F
EF ¼  33 MeV: ð7:5Þ
2M

The difference between the top of the well and the Fermi level is constant
for most heavy nuclei and is just the average binding energy per nucleon
222 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

~  B=A ¼ 7–8 MeV. The depth of the potential and the Fermi energy are to a
B
good approximation independent of the mass number A:
~  40 MeV:
V0 ¼ EF þ B ð7:6Þ

Heavy nuclei generally have a surplus of neutrons. Since the Fermi levels of the
protons and neutrons in a stable nucleus have to be equal (otherwise the nucleus
can become more stable by -decay) this implies that the depth of the potential
well for the neutron gas has to be deeper than for the proton gas, as shown in
Figure 7.2. Protons are therefore on average less tightly bound in nuclei than are
neutrons.
We can use the Fermi gas model to give a theoretical expression for some of the
dependence of the binding energy on the surplus of neutrons, as follows. First, we
define the average kinetic energy per nucleon as
ð p F  ð p F 
1
2 2
hEkin i  Ekin p dp p dp : ð7:7Þ
0 0

Evaluating the integrals gives

3 p2F
hEkin i ¼  20 MeV: ð7:8Þ
5 2M
The total kinetic energy of the nucleus is then

3
Ekin ðN; ZÞ ¼ NhEn i þ ZhEp i ¼ ½NðpnF Þ2 þ ZðppF Þ2 ; ð7:9Þ
10 M

which may be re-expressed as

   
3
h2 9 2=3 N 5=3 þ Z 5=3
Ekin ðN; ZÞ ¼ ; ð7:10Þ
10 M R20 4 A2=3

where again we have taken the radii of the proton and neutron wells to be equal.
This expression is for fixed A but varying N and has a minimum at N ¼ Z. Hence
the binding energy gets smaller for N 6¼ Z. If we set N ¼ ðA þ Þ=2,
Z ¼ ðA
Þ=2, where   N
Z, and expand Equation (7.10) as a power series
in =A, we obtain
  " #
3
h2 9 2=3 5 ðN
ZÞ2
Ekin ðN; ZÞ ¼ Aþ þ ...: ; ð7:11Þ
10 M R20 8 9 A

which gives the dependence on the neutron excess. The first term contributes to the
volume term in the semi-empirical mass formula (SEMF), while the second
 SHELL MODEL 223

describes the correction that results from having N 6¼ Z. This is a contribution to the
asymmetry term we have met before in the SEMF and grows as the square of the
neutron excess. Evaluating this term from Equation (7.11) shows that its contribu-
tion to the asymmetry coefficient defined in Equation (2.51) is about 44 MeV=c2 ,
compared with the empirical value of about 93 MeV=c2 given in Equation (2.54). In
practice, to reproduce the actual term in the SEMF accurately we would have to take
into account the change in the potential energy for N 6¼ Z.

7.3 Shell Model

The nuclear shell model is based on the analogous model for the orbital structure
of atomic electrons in atoms. In some areas it gives more detailed predictions than
the Fermi gas model and it can also address questions that the latter model cannot.
Firstly, we recap the main features of the atomic case.

7.3.1 Shell structure of atoms

The binding energy of electrons in atoms is due primarily to the central Coulomb
potential. This is a complicated problem to solve in general because in a multi-
electron atom we have to take account of not only the Coulomb field of the nucleus,
but also the fields of all the other electrons. Analytic solutions are not usually
possible. However, many of the general features of the simplest case of hydrogen
carry over to more complicated cases, so it is worth recalling the former.
Atomic energy levels are characterized by a quantum number n = 1, 2, 3, 4, . . . :
called the principal quantum number. This is defined so that it determines the energy
of the system.7 For any n there are energy-degenerate levels with orbital angular
momentum quantum numbers given by

‘ ¼ 0; 1; 2; 3; . . . ; ðn
1Þ ð7:12Þ

(this restriction follows from the form of the Coulomb potential) and for any value
of ‘ there are ð2‘ þ 1Þ sub-states with different values of the projection of orbital
angular momentum along any chosen axis (the magnetic quantum number):

m‘ ¼
‘;
‘ þ 1; . . . ; 0; 1; . . . ; ‘
1; ‘: ð7:13Þ

Due to the rotational symmetry of the Coulomb potential, all such sub-states are
degenerate in energy. Furthermore, since electrons have spin-12, each of the above

7
In nuclear physics we are not dealing with the same simple Coulomb potential, so it would be better to call n
the radial node quantum number, as it still determines the form of the radial part of the wavefunction.
224 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

states can be occupied by an electron with spin ‘up’ or ‘down’, corresponding to

the spin-projection quantum number
ms ¼ 1=2: ð7:14Þ

Again, both these states will have the same energy. So finally, any energy eigenstate
in the hydrogen atom is labelled by the quantum numbers ðn; ‘; m‘ ; ms Þ and for any n,
there will be nd degenerate energy states, where

X
n
1
nd ¼ 2 ð2‘ þ 1Þ ¼ 2n2 : ð7:15Þ
‘¼0

The high degree of degeneracy can be broken if there is a preferred direction in

space, such as that supplied by a magnetic field, in which case the energy levels
could depend on m‘ and ms . One such interaction is the spin–orbit coupling, which is
the interaction between the magnetic moment of the electron (due to its spin) and the
magnetic field due to the motion of the nucleus (in the electron rest frame). This
leads to corrections to the energy levels called fine structure, the size of which are
determined by the electromagnetic fine structure constant .
In atomic physics the fine-structure corrections are small and so, if we ignore them
for the moment, in hydrogen we would have a system with electron orbits
corresponding to shells of a given n, with each shell having degenerate sub-shells
specified by the orbital angular momentum ‘. Going beyond hydrogen, we can
introduce the electron–electron Coulomb interaction. This leads to a splitting in any
energy level n according to the ‘ value. The degeneracies in ml and ms are unchanged.
It is straightforward to see that if a shell or sub-shell is filled, then we have
X X
ms ¼ 0 and m‘ ¼ 0; ð7:16Þ

i.e. there is a strong pairing effect for closed shells. In these cases it can be shown
that the Pauli principle implies

L ¼ S ¼ 0 and J ¼ L þ S ¼ 0: ð7:17Þ

For any atom with a closed shell or a closed sub-shell structure, the electrons are
paired off and thus no valence electrons are available. Such atoms are therefore
chemically inert. It is straightforward to work out the atomic numbers at which this
occurs. These are

Z ¼ 2; 10; 18; 36; 54: ð7:18Þ

For example, the inert gas argon ArðZ ¼ 18Þ has closed shells corresponding to
n ¼ 1, 2 and closed sub-shells corresponding to n ¼ 3; ‘ ¼ 0; 1. These values of Z
are called the atomic magic numbers.
 SHELL MODEL 225

Figure 7.3 Binding energy per nucleon for even values of A: the solid curve is the SEMF (from
Bo69)

7.3.2 Nuclear magic numbers

In nuclear physics, there is also evidence for magic numbers, i.e. values of Z and N at
which the nuclear binding is particularly strong. This can been seen from the B=A
curves of Figure 2.10 where at certain values of N and Z the data lie above the SEMF
curve. This is also shown in Figure 7.3, where the inset shows the low-A region
magnified. (The figure only shows results for even values of the mass number A.)
The nuclear magic numbers are found from experiment to be

N ¼ 2; 8; 20; 28; 50; 82; 126

ð7:19Þ
Z ¼ 2; 8; 20; 28; 50; 82

and correspond to one or more closed shells, plus eight nucleons filling the s and p
sub-shells of a nucleus with a particular value of n. Nuclei with both N and Z having
one of these values are called doubly magic, and have even greater stability. An
example is the helium nucleus, the -particle.
Shell structure is also suggested by a number of other phenomena. For example:
‘magic’ nuclei have many more stable isotopes than other nuclei; they have very
226 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

small electric dipole moments, which means they are almost spherical, the most
tightly bound shape; and neutron capture cross-sections show sharp drops compared
with neighbouring nuclei. However, to proceed further we need to know something
about the effective potential.
A simple Coulomb potential is clearly not appropriate and we need some form that
describes the effective potential of all the other nucleons. Since the strong nuclear
force is short-ranged we would expect the potential to follow the form of the density
distribution of nucleons in the nucleus. For medium and heavy nuclei, we have seen
in Chapter 2 that the Fermi distribution fits the data and the corresponding potential
is called the Woods–Saxon form

V0
Vcentral ðrÞ ¼ : ð7:20Þ
1 þ eðr
RÞ=a

However, although these potentials can be shown to offer an explanation for the
lowest magic numbers, they do not work for the higher ones. This is true of all purely
central potentials.
The crucial step in understanding the origin of the magic numbers was taken in
1949 by Mayer and Jensen who suggested that by analogy with atomic physics there
should also be a spin–orbit part, so that the total potential is

Vtotal ¼ Vcentral ðrÞ þ V‘s ðrÞL  S; ð7:21Þ

where L and S are the orbital and spin angular momentum operators for a single
nucleon and V‘s ðrÞ is an arbitrary function of the radial coordinate.8 This form for the
total potential is the same as that used in atomic physics except for the presence of
the function V‘s ðrÞ. Once we have coupling between L and S then m‘ and ms are no
longer ‘good’ quantum numbers and we have to work with eigenstates of the total
angular momentum vector J, defined by J ¼ L þ S. Squaring this, we have

J2 ¼ L2 þ S2 þ 2L  S; ð7:22Þ

i.e.
1
L  S ¼ ðJ2
L2
S2 Þ ð7:23Þ
2
and hence the expectation value of L  S, which we write as h‘si, is

h2

‘=2 for j ¼ ‘ þ 12
h‘si ¼ ½ jð j þ 1Þ
‘ð‘ þ 1Þ
sðs þ 1Þ ¼ :
2
ð‘ þ 1Þ=2 for j ¼ ‘
12
ð7:24Þ

8
For their work on the shell structure of nuclei. Maria Goeppert-Mayer and J. Hans Jensen were awarded a
half share of the 1963 Nobel Prize in Physics. (They shared the prize with Wigner, mentioned in Chapter 1
for his development of the concept of parity.)
 SHELL MODEL 227

(We are always dealing with a single nucleon, so that s ¼ 12.) The splitting between
the two levels is thus

2‘ þ 1 2
Els ¼ h
hV‘s i: ð7:25Þ
2

Experimentally, it is found that V‘s ðrÞ is negative, which means that the state with
j ¼ ‘ þ 12 has a lower energy than the state with j ¼ ‘
12. This is the opposite of the
situation in atoms. Also, the splittings are substantial and increase linearly with ‘.
Hence for higher ‘, crossings between levels can occur. Namely, for large ‘, the
splitting of any two neighbouring degenerate levels can shift the j ¼ ‘
12 state of the
initial lower level to lie above the j ¼ ‘ þ 12 level of the previously higher level.
An example of the resulting splittings up to the 1G state is shown in Figure 7.4,
where the usual atomic spectroscopic notation has been used, i.e. levels are written
n‘j with S, P, D, F, G, . . . : used for ‘ ¼ 0, 1, 2, 3, 4, . . .. Magic numbers occur when
there are particularly large gaps between groups of levels. Note that there is no
restriction on the values of ‘ for a given n because, unlike in the atomic case, the
strong nuclear potential is not Coulomb-like.
The configuration of a real nuclide (which of course has both neutrons and
protons) describes the filling of its energy levels (sub-shells), for protons and for
neutrons, in order, with the notation ðn‘j Þk for each sub-shell, where k is the
occupancy of the given sub-shell. Sometimes, for brevity, the completely filled

Figure 7.4 Low-lying energy levels in a single-particle shell model using a Woods--Saxon
potential plus spin--orbit term; circled integers correspond to nuclear magic numbers
228 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

sub-shells are not listed, and if the highest sub-shell is nearly filled, k can be given as
a negative number, indicating how far from being filled that sub-shell is. Using the
ordering diagram above, and remembering that the maximum occupancy of each
sub-shell is 2j þ 1, we predict, for example, the configuration for 178 O to be:

ð1s12 Þ2 ð1p32 Þ4 ð1p12 Þ2 for the protons ð7:26aÞ

and

ð1s12 Þ2 ð1p32 Þ4 ð1p12 Þ2 ð1d52 Þ1 for the neutrons: ð7:26bÞ

Notice that all the proton sub-shells are filled, and that all the neutrons are in filled
sub-shells except for the last one, which is in a sub-shell on its own. Most of the
ground state properties of 178 O can therefore be found from just stating the neutron
configuration as ð1d52 Þ1 .

7.3.3 Spins, parities and magnetic dipole moments

The nuclear shell model can be used to make predictions about the spins of ground
states. A filled sub-shell must have zero total angular momentum, because j is always
an integer-plus-a-half, so the occupancy of the sub-shell, 2j þ 1, is always an even
number. This means that in a filled sub-shell, for each nucleon of a given mj ð¼ jz Þ
there is another having the opposite mj . Thus the pair have a combined mj of zero and
so the complete sub-shell will also have zero mj . Since this is true whatever axis we
choose for z, the total angular momentum must also be zero. Since magic number
nuclides have closed sub-shells, such nuclides are predicted to have zero contribu-
tion to the nuclear spin from the neutrons or protons or both, whichever are magic
numbers. Hence magic-Z/magic-N nuclei are predicted to have zero nuclear spin.
This is indeed found to be the case experimentally.
In fact it is found that all even-Z/even-N nuclei have zero nuclear spin. We can
therefore make the hypothesis that for ground state nuclei, pairs of neutrons and
pairs of protons in a given sub-shell always couple to give a combined angular
momentum of zero, even when the sub-shell is not filled. This is called the pairing
hypothesis. We can now see why it is the last proton and/or last neutron that
determines the net nuclear spin, because these are the only ones that may not be
paired up. In odd-A nuclides there is only one unpaired nucleon, so we can predict
precisely what the nuclear spin will be by referring to the filling diagram. For even-
A/odd-Z/odd-N nuclides, however, we will have both an unpaired proton and an
unpaired neutron. We cannot then make a precise prediction about the net spin
because of the vectorial way that angular momenta combine; all we can say is that
the nuclear spin will lie in the range jjp
jn j to ðjp þ jn Þ.
Predictions can also be made about nuclear parities. First, recall the following
properties of parity: (1) parity is the transformation r !
r; (2) the wavefunction
 SHELL MODEL 229

of a single-particle quantum state will contain an angular part proportional to the

spherical harmonic Yml ð ,
Þ, and under the parity transformation

PYml ð ;
Þ ¼ ð
Þ‘ Yml ð ;
Þ; ð7:27Þ

(3) a single-particle state will also have an intrinsic parity, which for nucleons is
defined to be positive. Thus the parity of a single-particle nucleon state depends
exclusively on the orbital angular momentum quantum number with P ¼ ð
1Þ‘ .
The total parity of a multiparticle state is the product of the parities of the individual
particles. A pair of nucleons with the same ‘ will therefore always have a combined
parity of þ1. The pairing hypothesis then tells us that the total parity of a nucleus is
found from the product of the parities of the last proton and the last neutron. So we
can predict the parity of any nuclide, including the odd/odd ones, and these
predictions are in agreement with experiment.
Unless the nuclear spin is zero, we expect nuclei to have magnetic (dipole)
moments, since both the proton and the neutron have intrinsic magnetic moments,
and the proton is electrically charged, so it can produce a magnetic moment when it
has orbital motion. The shell model can make predictions about these moments.
Using a notation similar to that used in atomic physics, we can write the nuclear
magnetic moment as
 ¼ gj jN ; ð7:28Þ

where N is the nuclear magneton that was used in the discussion of hadron
magnetic moments in Section 3.3.3, gj is the Landé g-factor and j is the nuclear spin
quantum number. For brevity we can write simply  ¼ gj j nuclear magnetons.
We will find that the shell model does not give very accurate predictions for
magnetic moments, even for the even–odd nuclei where there is only a single
unpaired nucleon in the ground state. We will therefore not consider at all the much
more problematic case of the odd–odd nuclei having an unpaired proton and an
unpaired neutron.
For the even–odd nuclei, we would expect all the paired nucleons to contribute
zero net magnetic moment, for the same reason that they do not contribute to the
nuclear spin. Predicting the nuclear magnetic moment is then a matter of finding
the correct way to combine the orbital and intrinsic components of magnetic moment
of the single unpaired nucleon. We need to combine the spin component of the
moment, gs s, with the orbital component, g‘ ‘ (where gs and g‘ are the g-factors for
spin and orbital angular momentum.) to give the total moment gj j. The general
formula for doing this is9
jðj þ 1Þ þ ‘ð‘ þ 1Þ
sðs þ 1Þ jðj þ 1Þ
‘ð‘ þ 1Þ þ sðs þ 1Þ
gj ¼ g‘ þ gs ; ð7:29Þ
2jðj þ 1Þ 2jðj þ 1Þ

9
See, for example, Section 6.6 of En66.
230 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

which simplifies considerably because we always have j ¼ ‘  12. Thus

jgj ¼ g‘ ‘ þ gs =2 for j ¼ ‘ þ 12 ð7:30aÞ

and
   
1 1
jgj ¼ g‘ j 1 þ
gs j for j ¼ ‘
12: ð7:30bÞ
2‘ þ 1 2‘ þ 1

Since g‘ ¼ 1 for a proton and 0 for a neutron, and gs is approximately þ5.6 for the
proton and
3.8 for the neutron, Equations (7.30) yield the results (where
gprotonðneutronÞ is the g-factor for nuclei with an odd proton(neutron))

1
jgproton ¼ ‘ þ  5:6 ¼ j þ 2:3 for j ¼ ‘ þ 12
2
   
1 1 2:3j
jgproton ¼ j 1 þ
5:6  j ¼j
for j ¼ ‘
12
2‘ þ 1 2‘ þ 1 jþ1
ð7:31Þ
1
jgneutron ¼
 3:8 ¼
1:9 for j ¼ j ¼ ‘ þ 12
2  
1 1:9j
jgneutron ¼ 3:8  j ¼ for j ¼ ‘
12:
2‘ þ 1 jþ1

Accurate values of magnetic dipole moments are available for a wide range of
nuclei and plots of a sample of measured values for a range of odd-Z and odd-N
nuclei across the whole periodic table are shown in Figure 7.5. It is seen that for a
given j, the measured moments usually lie somewhere between the j ¼ ‘
12 and the
j ¼ ‘ þ 12 values (the so-called Schmidt lines), but beyond that the model does not
predict the moments accurately. The only exceptions are a few low-A nuclei where
the numbers of nucleons are close to magic values.
Why should the shell model work so well when predicting nuclear spins and
parities, but be poor for magnetic moments? There are several likely problem areas,
including the possibility that protons and neutrons inside nuclei may have effective
intrinsic magnetic moments that are different to their free-particle values, because of
their very close proximity to one another.

7.3.4 Excited states

In principle, the shell model’s energy level structure can be used to predict nuclear
excited states. This works quite well for the first one or two excited states when there
is only one possible configuration of the nucleus. However, for higher states the
 SHELL MODEL 231

Figure 7.5 Magnetic moments for odd-N, even-Z nuclei (upper diagram) and odd-Z, even-N
(lower diagram) as functions of nuclear spin compared with the predictions of the single-particle
shell model (the Schmidt lines)

spectrum becomes very complicated because several nucleons can be excited

simultaneously into a superposition of many different configurations to produce a
given nuclear spin and parity. When trying to predict the first one or two excited
states using a filling diagram like Figure 7.4, we are looking for the configuration
that is nearest to the ground state configuration. This will normally involve either
232 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

moving an unpaired nucleon to the next highest level, or moving a nucleon from the
sub-shell below the unpaired nucleon up one level to pair with it. Thus it is necessary
to consider levels just above and below the last nucleons (protons and neutrons).
As an example, consider the case of 178 O. Its ground-state configuration is given in
Equations (7.26). All the proton sub-shells are filled, and all the neutrons are in filled
sub-shells except for the last one, which is in a sub-shell on its own. There are three
possibilities to consider for the first excited state:

1. promote one of the 1p12 protons to 1d52 , giving a configuration of ð1p12 Þ

1 ð1d52 Þ1 ,
where the superscript
1 means that the shell is one particle short of being
filled;

2. promote one of the 1p12 neutrons to 1d52 , giving a configuration of ð1p12 Þ

1 ð1d52 Þ2 ;

3. promote the 1d52 neutron to the next level, which is probably 2s12 (or the nearby
1d32 ), giving a configuration of ð1s12 Þ1 or ð1d32 Þ1.

Following the diagram of Figure 7.4, the third of these possibilities would
correspond to the smallest energy shift, so it should be favoured over the others.
The next excited state might involve moving the last neutron up a further level to 1d32 ,
or putting it back where it was and adopting configurations (1) or (2). Option (2) is
favoured over (1) because it keeps the excited neutron paired with another, which
should have a slightly lower energy than creating two unpaired protons. When
comparing these predictions with the observed excited levels it is found that the
expected excited states do exist, but not necessarily in precisely the order predicted.
The shell model has many limitations, most of which can be traced to its
fundamental assumption that the nucleons move independently of one another in
a spherically symmetric potential. The latter, for example, is only true for nuclei that
are close to having doubly-filled magnetic shells and predicts zero electric quadruple
moments, whereas in practice many nuclei are deformed and quadruple moments are
often substantial. We discuss this important observation in the next section.

7.4 Non-Spherical Nuclei

So far we have discussed only spherical nuclei, but with non-sphericity new
phenomena are allowed, including additional modes of excitation and the possibility
of an electric quadrupole moment.

7.4.1 Electric quadrupole moments

The charge distribution in a nucleus is described in terms of electric multipole

moments and follows from the ideas of classical electrostatics. If we have a localized
 NON-SPHERICAL NUCLEI 233

classical charge distribution with charge density ðxÞ within a volume , then the
first moment that can be non-zero is the electric quadrupole Q, defined by
ð
eQ  ðxÞð3z2
r 2 Þd3 x; ð7:32Þ

where we have taken the axis of symmetry to be the z-axis. The analogous definition
in quantum theory is
ð
1X 
Q¼ qi ð3z2i
r2 Þ d3 x; ð7:33Þ
e i

where is the nuclear wavefunction and the sum is over all relevant nucleons,
each with charge qi .10 The quadrupole moment is zero if j j2 is spherically
symmetric and so a non-zero value of Q would be indicative of a non-spherical
nuclear charge distribution.
If we consider a spheroidal distribution with semi-axes defined as in Figure 2.14,
then evaluation of Equation (7.32) leads to the result

2
Qintrinsic ¼ Zeða2
b2 Þ; ð7:34Þ
5

where Qintrinsic is the value of the quadrupole moment for a spheroid at rest and Ze
is its total charge. For small deformations,

6
Qintrinsic  ZeR2 "; ð7:35Þ
5

where " is defined in Equation (2.70) and R is the nuclear radius. Thus, for a
prolate distribution ða > bÞ, Q > 0 and for an oblate distribution ða < bÞ, Q < 0,
as illustrated in Figure 7.6. The same results hold in the quantum case.
If the nucleus has a spin J and magnetic quantum number M, then Q will depend
on M because it depends on the shape and hence the orientation of the charge
distribution. The quadrupole moment is then defined as the value of Q for which M
has its maximum value projected along the z-axis. This may be evaluated from
Equation (7.33) in the single-particle shell model and without proof we state the
resulting prediction that for odd-A, odd-Z nuclei with a single proton having a total
angular moment j outside closed sub-shells, the value of Q is given by

ð2j
1Þ
Q 
R2 : ð7:36Þ
2ðj þ 1Þ

10 PÐ 
The electric dipole moment dz ¼ 1e qi zi d vanishes because it will contain a sum of terms of the
i
form h i jzi j i i, all of which are zero by parity conservation.
234 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

Figure 7.6 Shapes of nuclei leading to (a) Q > 0 (prolate), and (b) Q < 0 (oblate)

Thus, Q ¼ 0 for j ¼ 12. For odd-A, odd-N nuclei with a single neutron outside
closed sub-shells Q is predicted to be zero because the neutron has zero electric
charge, as will all even-Z, odd-N nuclei because of the pairing effect.
Unlike magnetic dipole moments, electric quadrupole moments are not always well
measured and the quoted experimental errors are often far larger than the differences
between the values obtained in different experiments. Significant (and difficult to
apply) corrections also need to be made to the data to extract the quadrupole moment
and this is not always done. The compilation of electric dipole moment data shown in
Figure 7.7 is therefore representative. The solid lines are simply to guide the eye and

Figure 7.7 Some measured electric quadrupole moments for odd-A nuclei, normalized by dividing
by R2, the squared nuclear radius: grey circles denote odd-N nuclei and black circles odd-Z nuclei;
the solid lines have no theoretical significance and the arrows denote the position of closed shells
 NON-SPHERICAL NUCLEI 235

have no theoretical significance. The arrows indicate the positions of major closed
shells. A change of sign of Q at these points is expected because a nucleus with one
proton less than a closed shell behaves like a closed-shell nucleus with a negatively
charged proton (a proton hole) and there is some evidence for this in the data.
Two features emerge from this diagram. Firstly, while odd-A, odd-Z nuclei with
only a few nucleons outside a closed shell do have moments of order
R2, in general
the measured moments are larger by factors of two to three and for some nuclei the
discrepancy can be as large as a factor of 10. Secondly, odd-A, odd-N nuclei also
have non-zero moments, contrary to expectations and, moreover, there is little
difference between these and the moments for odd-A, odd-Z nuclei, except that
the former tend to be somewhat smaller. These results strongly suggest that for some
nuclei it is not a good approximation to assume spherical symmetry and that these
nuclei must be considered to have non-spherical mass distributions.
The first attempt to explain the measured electric quadrupole moments in terms
of non-spherical nuclei was due to Rainwater. His approach can be understood
using the model we discussed in Chapter 2 when considering fission and used
above to derive the results of Equations (7.34) and (7.35). There the sphere was
deformed into an ellipsoid (see Figure 2.14) with axes parameterized in terms of a
small parameter " via Equation (2.70). The resulting change in the binding energy
EB was found to be

EB ¼
"2 ; ð7:37Þ

where

1 2 1
 ¼ ð2as A3
ac Z 2 A
3 Þ ð7:38Þ
5

and the coefficients as and ac are those of the SEMF with numerical values given
in Equation (2.54). Rainwater assumed that this expression only held for closed-
shell nuclei, but not for nuclei with nucleons in unfilled shells. In the latter cases he
showed that distortion gives rise to an additional term in EB that is linear in ", so
that the total change in binding energy is

EB ¼
"2
"; ð7:39Þ

where  is a parameter that could be calculated from the Fermi energy of the
nucleus. This form has a minimum value 2 =4 where " ¼
=2. The ground
state would therefore be deformed and not spherical.
Finally, once the spin of the nucleus is taken into account in quantum theory, the
measured electric quadrupole moment for ground states is predicted to be

jð2 j
1Þ
Q¼ Qintrinsic : ð7:40Þ
ð j þ 1Þð2 j þ 1Þ
236 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

This model gives values for Q that are of the correct sign, but overestimates them by
typically a factor of two. Refined variants of the model are capable of bringing the
predictions into agreement with the data by making better estimates of the parameter .

7.4.2 Collective model

The Rainwater model is equivalent to assuming an aspherical liquid drop and Aage
Bohr (the son of Neils Bohr) and Mottelson showed that many properties of heavy
nuclei could be ascribed to the surface motion of such a drop. However, the single-
particle shell model cannot be abandoned because it explains many general features
of nuclear structure. The problem was therefore to reconcile the shell model with the
liquid-drop model. The outcome is the collective model.11
This model views the nucleus as having a hard core of nucleons in filled shells, as
in the shell model, with outer valence nucleons that behave like the surface
molecules of a liquid drop. The motions of the latter introduce non-sphericity in
the core that in turn causes the quantum states of the valence nucleons to change
from the unperturbed states of the shell model. Such a nucleus can both rotate and
vibrate and these new degrees of freedom give rise to rotational and vibrational
energy levels. For example, the rotational levels are given by EJ ¼ JðJ þ 1Þ
h2 =2I,
where I is the moment of inertia and J is the spin of the nucleus. The predictions of
this simple model are quite good for small J, but overestimate the energies for larger
J. Vibrational modes are due predominantly to shape oscillations, where the nucleus
oscillates between prolate and oblate ellipsoids. Radial oscillations are much rarer
because nuclear matter is relatively incompressible. The energy levels are well
approximated by a simple harmonic oscillator potential with spacing E ¼
h!,
where ! is the oscillator frequency.
In practice, the energy levels of deformed nuclei are very complicated, because
there is often coupling between the various modes of excitation, but nevertheless
many predictions of the collective model are confirmed experimentally.12

7.5 Summary of Nuclear Structure Models

The shell model is based upon the idea that the constituent parts of a nucleus move
independently. The liquid-drop model implies just the opposite, since in a drop of
incompressible liquid, the motion of any constituent part is correlated with the
motion of all the neighbouring pairs. This emphasizes that models in physics have a
limited domain of applicability and may be unsuitable if applied to a different set of
phenomena. As knowledge evolves, it is natural to try and incorporate more

11
For their development of the collective model, Aage Bohr, Ben Mottelson and Leo Rainwater shared the
1975 Nobel Prize in Physics.
12
The details are discussed, for example, in Section 2.3 of Je90 and Chapter 17 or Ho97.
 SUMMARY OF NUCLEAR STRUCTURE MODELS 237

phenomena by modifying the model to become more general, until (hopefully) we

have a model with firm theoretical underpinning which describes a very wide range
of phenomena, i.e. a theory. The collective model, which uses the ideas of both the
shell and liquid drop models, is a step in this direction. We will conclude this section
with a brief summary of the assumptions of each of the nuclear models we have
discussed and what each can tell us about nuclear structure.

Liquid-drop model

This model assumes that all nuclei have similar mass densities, with binding
energies approximately proportional to their masses, just as in a classical
charged liquid drop. The model leads to the SEMF, which gives a good description
of the average masses and binding energies. It is largely classical, with some
quantum mechanical terms (the asymmetry and pairing terms) inserted in an
ad hoc way. Input from experiment is needed to determine the coefficients of the
SEMF.

Fermi gas model

The assumption here is that nucleons move independently in a net nuclear

potential. The model uses quantum statistics of a Fermi gas to predict the depth
of the potential and the asymmetry term of the SEMF.

Shell model

This is a fully quantum mechanical model that solves the Schrödinger equation with
a specific spherical nuclear potential. It makes the same assumptions as the Fermi
gas model about the potential, but with the addition of a strong spin–orbit term. It is
able to successfully predict nuclear magic numbers, spins and parities of ground-
state nuclei and the pairing term of the SEMF. It is less successful in predicting
magnetic moments.

Collective model

This is also a fully quantum mechanical model, but in this case the potential is
allowed to undergo deformations from the strictly spherical form used in the shell
model. The result is that the model can predict magnetic dipole and electric
quadrupole magnetic moments with some success. Additional modes of excitation,
both vibrational and rotational, are possible and are generally confirmed by
experiment.
238 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

It is clear from the above that there is at present no universal nuclear model. What
we currently have is a number of models and theories that have limited domains of
applicability and even within which they are not always able to explain all the
observations. For example, the shell model, while able to give a convincing account
of the spins and parities of the ground states of nuclei, is unable to predict the spins of
excited states with any real confidence. And of course the shell model has absolutely
nothing to say about whole areas of nuclear physics phenomena. Some attempt has
been made to combine features of different models, such as is done in the collec-
tive model, with some success. A more fundamental theory will require the full
apparatus of many-body theory applied to interacting nucleons and some progress
has been made in this direction for light nuclei, as we will mention in Chapter 9.
A theory based on interacting quarks is a more distant goal.

7.6 a-Decay
To discuss -decays, we could return to the semiempirical mass formula of Chapter
2 and by taking partial derivatives with respect to A and Z find the limits of -
stability, but the result is not very illuminating. To get a very rough idea of the
stability criteria, we can write the SEMF in terms of the binding energy B. Then
-decay is energetically allowed if

Bð2; 4Þ > BðZ; AÞ

BðZ
2; A
4Þ: ð7:41Þ

If we now make the approximation that the line of stability is Z ¼ N (the actual
line of stability deviates from this, see Figure 2.7), then there is only one independent
variable. If we take this to be A, then

dB
Bð2; 4Þ > BðZ; AÞ
BðZ
2; A
4Þ  4 ; ð7:42Þ
dA

and we can write

 
dB dðB=AÞ B
4 ¼4 A þ : ð7:43Þ
dA dA A

From the plot of B=A (Figure 2.2), we have dðB=AÞ=dA 

7:7  10
3 MeV for
A  120 and we also know that Bð2; 4Þ ¼ 28:3 MeV, so we have

28:3  4½B=A
7:7  10
3 A; ð7:44Þ

which is a straight line on the B=A versus A plot which cuts the plot at A  151.
Above this value of A, Equation (7.41) is satisfied by most nuclei and -decay
becomes energetically possible.
 a-DECAY 239

Figure 7.8 Schematic diagram of the potential energy of an -particle as a function of its
distance r from the centre of the nucleus

Lifetimes of -emitters span an enormous range, and examples are known from
10 ns to 1017 years. The origin of this large spread lies in the quantum mechanical
phenomenon of tunelling. Individual protons and neutrons have binding energies in
nuclei of about 8 MeV, even in heavy nuclei (see Figure 2.2), and so cannot in
general escape. However, a bound group of nucleons can sometimes escape because
its binding energy increases the total energy available for the process. In practice,
the most significant decay process of this type is the emission of an -particle,
because unlike systems of two and three nucleons it is very strongly bound by
7 MeV/ nucleon. Figure 7.8 shows the potential energy of an -particle as a function
of r, its distance from the centre of the nucleus.
Beyond the range of the nuclear force, r > R, the -particle feels only the
Coulomb potential

2Z
hc
VC ðrÞ ¼ ; ð7:45Þ
r

where we now use Z to be the atomic number of the daughter nucleus. Within the
range of the nuclear force, r < R, the strong nuclear potential prevails, with its
strength characterized by the depth of the well. Since the -particle can escape from
the nuclear potential, E > 0. It is this energy that is released in the decay. Unless E
is larger than the Coulomb barrier (in which case the decay would be so fast as to be
unobservable) the only way the -particle can escape is by quantum mechanical
tunelling through the barrier.
The probability T for transmission through a barrier of height V and thickness
r by a particle of mass m with energy E is given approximately by

T  e
2r ; ð7:46Þ
240 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

where
h ¼ ½2mjVC
E j1=2 . Using this result, we can model the Coulomb
barrier as a succession of thin barriers of varying height. The overall transmission
probability is then

T ¼ e
G ; ð7:47Þ

where the Gamow factor G is

rðC
2
G¼ ½2mjVC ðrÞ
E j1=2 dr; ð7:48Þ

h
R

with  ¼ v=c and v is the velocity of the emitted particle.13 This assumes that
the orbital angular momentum of the -particle is zero, i.e. we ignore
possible centrifugal barrier corrections.14 Since rC is the value of r where
E ¼ VC ðrC Þ,
rC ¼ 2Ze2 =4"0 E ð7:49Þ

and hence

VC ðrÞ ¼ 2Ze2 =4"0 r ¼ rC E =r: ð7:50Þ

So, substituting into Equation (7.48) gives

rðC
2ð2mE Þ1=2 hrC i1=2
G¼
1 dr; ð7:51Þ
h
r
R

where m is the reduced mass of the -particle and the daughter nucleus, i.e.
m ¼ m mD =ðm þ mD Þ  m . Evaluating the integral in Equation (7.51) gives
 1=2 " rffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 #
2mc2
1 R R R
G ¼ 4Z cos
1
: ð7:52Þ
E rC rC rC

Finally, since E is typically 5 MeV and the height of the barrier is typically
40 MeV, rC  R and from (7.52), G  4Z=, where  ¼ v =c and v is the
velocity of the alpha particle within the nucleus.
The probability per unit time  of the -particle escaping from the nucleus
is proportional to the product of: (a) the probability wðÞ of finding the -particle in
the nucleus; (b) the frequency of collisions of the -particle with the barrier (this

13
These formulae are derived in Appendix A.
14
The existence of an angular momentum barrier will suppress the decay rate (i.e. increase the lifetime)
compared with a similar nucleus without such a barrier. Numerical estimates of the suppression factors,
which increase rapidly with angular momentum, have been calculated by Blatt and Weisskopf and are given
in their book B152.
 a-DECAY 241

is v =2R); and (c) the transition probability. Thus, combining these factors,  is
given by
v
 ¼ wðÞ e
G ð7:53Þ
2R
and since
Z Z
G/ / pffiffiffiffiffiffi ; ð7:54Þ
 E
small differences in E have strong effects on the lifetime.
To examine this further we can take logarithms of Equation (7.53) to give

1
log10 t12 ¼ a þ bZE 2 ; ð7:55Þ

where t12 is the half-life. The quantity a depends on the probability wðÞ and so is a
function of the nucleus, whereas b is a constant that may be estimated from the above
equations to be about 1.7. Equation (7.55) is a form of a relation that was found
empirically by Geiger and Nuttall in 1911 long before its theoretical derivation in
1928. It is therefore called the Geiger-Nuttall relation. It predicts that for fixed Z,

12
the log of the half-life of -emitters varies linearly with E .
Figure 7.9 shows lifetime data for the isotopes of four nuclei. The very strong
variation with -particle energy is evident; changing E by a factor of about 2.5
changes the lifetime by 20 orders of magnitude. In all cases the agreement with the
Geiger–Nuttall relation is very reasonable and the slopes are compatible with the

Figure 7.9 Comparison of the Geiger--Nuttall relation with experimental data for some
-emitters
242 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

estimate for b above. Thus the simple barrier penetration model is capable of
explaining the very wide range of lifetimes of nuclei decaying by -emission.

7.7 b-Decay
In Chapter 2 we discussed in some detail the phenomenology of -decay using the
SEMF. In this section we return to these decays and examine their theoretical
interpretation.

7.7.1 Fermi theory

The first successful theory of nuclear -decay was proposed in the 1930s by Fermi,
long before the W and Z bosons were known and the quark model formulated. He
therefore had to construct a theory based on very general principles, working by
analogy with the quantum theory of electromagnetic processes (QED), the only
successful theory known at the time for quantum particles.
The general equation for electron -decay is

A
ZX ! Zþ1A Y þ e
þ 
e : ð7:56Þ

In Chapter 2, we interpreted this reaction as the decay of a bound neutron, i.e.

n ! p þ e
þ 
e , and in Chapter 3 we gave the quark interpretation of this decay. In
general, it is possible for the internal state of the nucleus to change in other ways
during the transition, but we will simplify matters by considering just the basic
neutron decay process.
We have also met the Second Golden Rule, which enables transition rates to be
calculated provided the interaction is relatively weak. We will write the Golden
Rule as
2
!¼ jMfi j2 nðEÞ; ð7:57Þ

h

where ! is the transition rate (probability per unit time), Mfi is the transition
amplitude (also called the matrix element because it is one element of a matrix
whose elements are all the possible transitions from the initial state i to different final
states f of the system) and nðEÞ is the density of states, i.e. the number of quantum
states available to the final system per unit interval of total energy. The density-
of-states factor can be calculated from purely kinematical factors, such as energies,
momenta, masses and spins where appropriate.15 The dynamics of the process is
contained in the matrix element.

15
This is done explicitly in Appendix A.
 b-DECAY 243

The matrix element can in general be written in terms of five basic Lorentz
^:
invariant interaction operators, O
ð
Mfi ¼ f ðgO
^ Þi d3 x; ð7:58Þ

where f and i are total wavefunctions for the final and initial states, respectively,
g is a dimensionless coupling constant, and the integral is over three-dimensional
space The five categories are called scalar (S), pseudo-scalar (P), vector (V), axial-
vector (A), and tensor (T ); the names having their origin in the mathematical
transformation properties of the operators. (We have met the V and A forms
previously in Chapter 6 on the electroweak interaction.) The main difference
between them is the effect on the spin states of the parent and daughter nuclei.
When there are no spins involved, and at low energies, ðgO ^ Þ is simply the interaction
potential, i.e. that part of the potential that is responsible for the change of state of the
system.
Fermi guessed that O ^ would be of the vector type, since electromagnetism is a
vector interaction, i.e. it is transmitted by a spin-1 particle – the photon. (Decays of
the vector type are called Fermi transitions.) We have seen from the work of
Chapter 6 that we now know that the weak interaction violates parity conservation
and is correctly written as a mixture of both vector and axial-vector interactions (the
latter are called Gamow–Teller transitions in nuclear physics), but as long as we are
not concerned with the spins of the nuclei, this does not make much difference, and
we can think of the matrix element in terms of a classical weak interaction potential,
like the Yukawa potential. Applying a bit of modern insight, we can assume the
potential is of extremely short range (because of the large mass of the W boson), in
which case we have seen that we can approximate the interaction by a point-like
form and the matrix element then becomes simply a constant, which we write as

GF
Mfi ¼ ; ð7:59Þ
V

where GF is the Fermi coupling constant we met in Chapter 6. It has dimensions

[energy][length]3 and is related to the charged current weak interaction coupling
W by

4ð
hcÞ3 W
GF ¼ : ð7:60Þ
ðMW c2 Þ2

In Equation (7.59) it is convenient to factor out an arbitrary volume V, which is used

to normalize the wavefunctions. (It will eventually cancel out with a factor in the
density-of-states term.)
In nuclear theory, the Fermi coupling constant GF is taken to be a universal
constant and with appropriate corrections for changes of the nuclear state this
244 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

assumption is also used in -decay. Experimental results are consistent with the
theory under this assumption. However, the theory goes beyond nuclear -decay,
and can be applied to any process mediated by the W boson, provided the energy is
not too great. In Chapter 6, for example, we used the same ideas to discuss
neutrino scattering. The best process to determine the value of GF is one not
complicated by hadronic (nuclear) effects and muon decay is usually used. The
lifetime of the muon  is given to a very good approximation by (ignoring the
Cabbibo correction)
1 ðm c2 Þ5
¼ G2 ;
6 F
ð7:61Þ
 1923
hð

hcÞ

from which we can deduce that the value of GF is about 90 eV fm3. It is usually
quoted in the form GF =ð
hcÞ3 ¼ 1:166  10
5 GeV
2 .

7.7.2 Electron momentum distribution

We see from Equation (7.58) that the transition rate (i.e. -decay lifetime) depends
essentially on kinematical factors arising through the density-of-states factor, nðEÞ.
To simplify the evaluation of this factor, we consider the neutron and proton to be
‘heavy’, so that they have negligible kinetic energy, and all the energy released in the
decay process goes into creating the electron and neutrino and in giving them kinetic
energy. Thus we write

E ¼ Ee þ E ; ð7:62Þ

where Ee is the total (relativistic) energy of the electron, E is the total energy
of the neutrino, and E is the total energy released. (This equals ðmÞc2 , if m is
the neutron–proton mass difference, or the change in mass of the decaying
nucleus.)
The transition rate ! can be measured as a function of the electron momentum, so
we need to obtain an expression for the spectrum of -decay electrons. Thus we will
fix Ee and find the differential transition rate for decays where the electron has
energy in the range Ee to Ee þ dEe . From the Golden Rule, this is

2
d! ¼ jMj2 n ðE
Ee Þne ðEe ÞdEe ; ð7:63Þ

h

where ne and n are the density of states factors for the electron and neutrino,
respectively. These may be obtained from our previous result:

V
nðpe Þdpe ¼ 4p2e dpe ; ð7:64Þ
ð2
hÞ3
 b-DECAY 245

with a similar expression for n , by changing variables using

dp E
¼ 2; ð7:65Þ
dE pc

so that
4V
nðEe ÞdEe ¼ pe Ee dEe ; ð7:66Þ
ð2
hÞ3 c2

with a similar expression for nðE Þ. Using these in Equation (7.57) and setting
M ¼ GF =V, gives

d! G2
¼ 3 F7 4 pe Ee p E ð7:67Þ
dEe 2
h c

where in general
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
p c ¼ E2
m2 c4 ¼ ðE
Ee Þ2
m2 c4 : ð7:68Þ

Finally, it is useful to change the variable to pe by writing

d! dEe d! G2
¼ ¼ 3 F7 2 p2e p E : ð7:69Þ
dpe dpe dEe 2

hc

If we take the antineutrino to be precisely massless, then p ¼ E =c and Equation

(7.69) reduces to

d! G2F p2e p2 G2F p2e E2 G2F p2e ðE

Ee Þ2
¼ ¼ ¼ : ð7:70Þ
dpe 23
h7 c 23
h7 c3 23
h7 c3

This expression gives rise to a bell-shaped electron momentum distribution,

which rises from zero at zero momentum, reaches a peak and falls to zero again at an
electron energy equal to E, as illustrated in the curve labelled Z ¼ 0 in Figure 7.10.
Studying the precise shape of the distribution near its upper end-point is one way in
principle of finding a value for the antineutrino mass. If the neutrino has zero mass,
then the gradient of the curve approaches zero at the end-point, whereas any non-
zero value results in an end-point that falls to zero with an asymptotically infinite
gradient. We will return to this later.
There are several factors that we have ignored or over-simplified in deriving this
momentum distribution. The principal ones are to do with the possible changes
of nuclear spin of the decaying nucleus, and the electric force acting between
the -particle (electron or positron) and the nucleus. In the first case, when the
electron–antineutrino carry away a combined angular momentum of 0 or 1, the
246 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

Figure 7.10 Predicted electron spectra: Z ¼ 0, without Fermi screening factor;   , with Fermi
screening factor
above treatment is essentially correct: these are the so-called ‘allowed transitions’.
However, the nucleus may change its spin by more than 1 unit, and then the
simplified short-range potential approach to the matrix element is inadequate. The
decay rate in these cases is generally suppressed, although not completely
forbidden, despite these being traditionally known as the ‘forbidden transitions’.16
In the second case, the electric potential between the positive nucleus and a
positive -particle will cause a shift of the low end of its momentum spectrum to
the right, since it is propelled away by electrostatic repulsion. Conversely, the low
end of the negative -spectrum is shifted to the left (see Figure 7.10). The precise
form of these effects is difficult to calculate, and requires quantum mechanics, but
the results are published in tables of a factor FðZ; Ee Þ, called the Fermi screening
factor, to be applied to the basic -spectrum.

7.7.3 Kurie plots and the neutrino mass

The usual way of experimentally testing the form of the electron momentum
spectrum given by the Fermi theory is by means of a Kurie plot. From
Equation (7.70), with the Fermi screening factor included, we have

d! FðZ; Ee ÞG2F p2e ðE

Ee Þ2
¼ ; ð7:71Þ
dpe 23
h7 c3

which can be written as

  12
d! 1
HðEe Þ  ¼ E
Ee ; ð7:72Þ
dpe p2e KðZ; pe Þ

16
For a discussion of forbidden transitions see, for example, Co01.
 b-DECAY 247

Figure 7.11 Kurie plot for the -decay of 36 Cl (the y-axis is proportional to the function HðEe Þ
above)

where KðZ; pe Þ includes FðZ; Ee Þ and all the fixed constants in Equation (7.71). A
plot of the left-hand side of this equation – using the measured d!=dpe and pe ,
together with the calculated value of KðZ; pe Þ – against the electron energy Ee should
then give a straight line with an intercept of E. An example is shown in Figure 7.11.
If the neutrino mass is not exactly zero then it is straightforward to repeat the above
derivation and to show that the left-hand side of the Kurie plot is proportional to
1 1
fðE
Ee Þ½ðE
Ee Þ2
m2 c4 2 g2 : ð7:73Þ

This will produce a very small deviation from linearity extremely close to the end-
point of the spectrum and the straight line will curve near the end point and cut the
0
axis vertically at E0 ¼ E0
m c2 . In order to have the best conditions for measuring
the neutrino mass, it is necessary to use a nucleus where a non-zero mass would have
a maximum effect, i.e. the maximum energy release E ¼ E0 should only be a few
keV. Also at such low energies atomic effects have to be taken into account, so the
initial and final atomic states must be very well understood. The most suitable case is
the decay of tritium,
3
H ! 3 He þ e
þ 
e ; ð7:74Þ

where E0 ¼ 18:6 keV. The predicted Kurie plot very close to the end-point is
shown in Figure 7.12.
Since the counting rate near E0 is vanishingly small, the experiment is extremely
difficult. In practice, the above formula is fitted to data close to the end-point of the
spectrum and extrapolated to E0 . The best experiments are consistent with a zero
248 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

Figure 7.12 Expected Kurie plot for tritium decay very close to the end-point of the electron
energy spectrum for two cases: m ¼ 0 and m ¼ 5 eV=c2

neutrino mass, but when experimental and theoretical uncertainties are taken into
account, an upper limit of about 2–3 eV/c2 results.

7.8 c-Emission and Internal Conversion

In Chapter 2 we mentioned that excited states of nuclei frequently decay to lower
states (often the ground state) by the emission of photons in the energy range
appropriate to -rays and that in addition it is possible for the nucleus to de-excite by
ejecting an electron from a low-lying atomic orbit. We shall discuss this only briefly
because a proper treatment requires using a quantized electromagnetic radiation
field and is beyond the scope of this book. Instead, we will outline the results,
without proof.

7.8.1 Selection rules

Gamma emission is a form of electromagnetic radiation and like all such radiation is
caused by a changing electric field inducing a magnetic field. There are two
possibilities, called electric (E) radiation and magnetic (M) radiation. These
names derive from the semiclassical theory of radiation, in which the radiation
field arises because of the time variation of charge and current distributions. The
classification of the resulting radiation is based on the fact that total angular
momentum and parity are conserved in the overall reaction, the latter because it is
an electromagnetic process.
 g-EMISSION AND INTERNAL CONVERSION 249

The photon carries away a total angular momentum, given by a quantum number
L17, which must include the fact that the photon is a spin-1 vector meson. The
minimum value is L ¼ 1. This is because a real photon has two possible polarization
states corresponding, for example, to Lz ¼ 1. Thus in the transition, there must be a
change of Lz for the emitting nucleus of 1 and this cannot happen if L ¼ 0. Hence,
if the spins of the initial and final nuclei states are denoted by Ji and Jf respectively,
the transition Ji ¼ 0 ! Jf ¼ 0 is strictly forbidden. In general, the photons are said
to have a multipolarity L and we talk about multipole radiation; transitions are called
dipole ðL ¼ 1Þ, quadrupole ðL ¼ 2Þ, octupole ðL ¼ 3Þ etc.. Thus, for example, M2
stands for magnetic quadrupole radiation. The allowed values of L are restricted by
the conservation equation relating the photon total angular momentum L and the
spins of the initial and final nuclei states, i.e.

Ji ¼ Jf þ L: ð7:75Þ

Thus, L may lie in the range

Ji þ Jf  L  jJi
Jf j: ð7:76Þ

It is also necessary to take account of parity. In classical physics, an electric dipole

qr is formed by having two equal and opposite charges q separated by a distance r. It
therefore has negative parity under r !
r. Similarly, a magnetic dipole is
equivalent to a charge circulating with velocity v to form a current loop of radius
r. The magnetic dipole is then of the form qr  v, which does not change sign under a
parity inversion and thus has positive parity. The general result, which we state
without proof, is that electric multipole radiation has parity ð
1ÞL, whereas magnetic
multipole radiation has parity ð
1ÞLþ1. We thus are led to the selection rules for 
emission shown in Table 7.1. Using this table we can determine which radiation
types are allowed for any specific transition. Some examples are shown in Table 7.2.

Table 7.1 Selection rules for  emission

Multipolarity Dipole Quadrupole Octupole
Type of radiation E1 M1 E2 M2 E3 M3
L 1 1 2 2 3 3
P Yes No No Yes Yes No

Although transitions Ji ¼ 0 ! Jf ¼ 0 are forbidden because the photon is a real

particle, such transitions could occur if a virtual photon is involved, provided parity
does not change. The reason for this is that a virtual photon does not have the

17
As this is the total angular momentum, logically it would be better to employ the symbol J. However, as L
is invariably used in the literature, it will be used in what follows.
250 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

Table 7.2 Examples of nuclear electromagnetic transitions

P
jPi i Jf f P L Allowed transitions
0þ 0þ No – None
1þ 1

2 2 Yes 1 E1
þ þ
1 0 No 1 M1
2þ 0þ No 2 E2
3
1þ
2 2 Yes 1, 2 E1, M2
2þ 1þ No 1, 2, 3 M1, E2, M3
3
5þ
2 2 Yes 1, 2, 3, 4 E1, M2, E3, M4

restriction on its states of polarization that a real photon does. In practice, the energy
of the virtual photon can be transferred to an orbital atomic electron that can thereby
be ejected. This is the process of internal conversion. There is another possibility
whereby the virtual photon can create an internal eþ e
pair. This is referred to as
internal pair production.

7.8.2 Transition rates

In semi-classical radiation theory, the transition probability per unit time, i.e. the
emission rate, is given by18
 
E;M 1 8ðL þ 1Þ 1 E 2Lþ1 E;M
Tfi ðLÞ ¼ Bfi ðLÞ; ð7:77Þ
4"0 L½ð2L þ 1Þ!!2
h
hc

where E is the photon energy, E and M refer to electric and magnetic radiation,
and for odd-n, n!!  nðn
2Þðn
4Þ . . . 3:1. The function BE;Mfi ðLÞ is the so-called
reduced transition probability and contains all the nuclear information. It is
essentially the square of the matrix element of the appropriate operator causing
the transition producing photons with multipolarity L, taken between the initial and
final nuclear state wave functions. For electric transitions, B is measured in units of
e2 fm2L and for magnetic transitions in units of ðN =cÞ2 fm2L
2 where N is the
nuclear magneton.
To go further requires knowledge of the nuclear wave functions. An approxima-
tion due to Weisskopf is based on the single-particle shell model. This approach
assumes that the radiation results from the transition of a single proton from an initial
orbital state of the shell model to a final state of zero angular momentum. In this
model the general formulas reduce to
 2
e2 3RL
BE ðLÞ ¼ ð7:78aÞ
4 L þ 3

18
See, for example, Chapter 16 of Ja75.
 g-EMISSION AND INTERNAL CONVERSION 251

for electric radiation and

 2
M
h
B ðLÞ ¼ 10 BE ðLÞ ð7:78bÞ
mp cR

for magnetic radiation, where R is the nuclear radius and mp is the mass of the
proton. Finally, from the work in Chapter 2 on nuclear sizes, we can substitute
R ¼ R0 A1=3 , with R0 ¼ 1:21 fm, to give the final results:
 2
e2 3
E
B ðLÞ ¼ ðR0 Þ2L A2L=3 ð7:79aÞ
4 L þ 3

and
  2
10 e
h 2 3
BM ðLÞ ¼ ðR0 Þ2L
2 Að2L
2Þ=3 : ð7:79bÞ
 2mp c Lþ3

Figure 7.13 shows an example of the transition rates T E;M calculated from
Equation (7.77) using the approximations of Equations (7.79). Although these are
only approximate predictions, they do confirm what is observed experimentally: for
a given transition there is a very substantial decrease in decay rates with increasing
L, and electric transitions have decay rates about two orders of magnitude higher
than the corresponding magnetic transitions.
Finally, it is often useful to have simple formulas for radiative widths  . These
follow from Equations (7.77), (7.78) and (7.79) and for the lowest multipole
transitions may be written

 ðE1Þ ¼ 0:068E3 A2=3 ;  ðM1Þ ¼ 0:021E3 ;  ðE2Þ ¼ ð4:9  10

8 ÞE5 A4=3 ;
ð7:80Þ

Figure 7.13 Transition rates using single-particle shell model formulas of Weisskopf as a
function of photon energy for a nucleus of mass number A ¼ 60
252 CH7 MODELS AND THEORIES OF NUCLEAR PHYSICS

where  is measured in eV, the transition energy E is measured in MeVand A is the

mass number of the nucleus. These formulae are based on the single-particle
approximation and in practice collective effects often give values that are much
greater than those predicted by Equations (7.80).

Problems
7.1 Assume that in the shell model the nucleon energy levels are ordered as shown in
Figure 7.4. Write down the shell-model configuration of the nucleus 73 Li and hence
find its spin, parity and magnetic moment (in nuclear magnetons). Give the two most
likely configurations for the first excited state, assuming that only protons are
excited.

7.2 A certain odd-parity shell-model state can hold up to a maximum of 16 nucleons;

what are its values of j and ‘?
þ
7.3 The ground state of the radioisotope 179 F has spin-parity jP ¼ 52 and the first excited

state has jP ¼ 12 . By reference to Figure 7.4, suggest two possible configurations for
the latter state.

7.4 What are the configurations of the ground states of the nuclei 93 33
41 Nb and 16 S and what
values are predicted in the single-particle shell model for their spins, parities and
magnetic dipole moments?

7.5 Show explicitly that a uniformly charged ellipsoid at rest with total charge Ze and
semi-axes defined in Figure 2.14, has a quadrupole moment Q ¼ 25Zeða2
b2 Þ.

7.6 The ground state of the nucleus 165

67 Ho has an electric quadrupole moment Q  3:5 b.
If this is due the fact that the nucleus is a deformed ellipsoid, use the result of
Question 7.5 to estimate the sizes of its semi-major and semi-minor axes.

þ þ
7.7 The decay 244 240
98 Cfð0 Þ ! 96 Cmð0 Þ þ  has a Q-value of 7.329 MeVand a half-life of
19.4 mins. If the frequency and probability of forming -particles (see Equation
þ þ
(7.53)) for this decay are the same as those for the decay 228 224
90 Thð0 Þ ! 88 Rð0 Þ þ ,
228
estimate the half-life for the -decay of 90 Th, given that its Q-value is 5.520 MeV.

7.8 The hadrons

0 and 0 can both decay via photon emission:

0 ð1193Þ ! ð1116Þ þ  (branching ratio  100 per cent); 0 ð1232Þ ! n þ 
(branching ratio 0.56 per cent). If the lifetime of the 0 is 0:6  10
23 s, estimate
the lifetime of the
0 .

7.9 The reaction 34 Sðp, nÞ34 Cl has a threshold proton laboratory energy of 6.45 MeV.
Calculate non-relativistically the upper limit of the positron energy in the -decay of
34
Cl, given that the mass difference between the neutron and the hydrogen atom is
0.78 MeV.
 PROBLEMS 253

7.10 To determine the mass of the electron neutrino from the -decay of tritium requires
measurements of the electron energy spectrum very close to the end-point where
there is a paucity of events (see Figure 7.12.). To see the nature of the problem,
estimate the fraction of electrons with energies within 10 eV of the end-point.

7.11 The electron energy spectra of -decays with very low-energy end-points E0 may be
approximated by d!=dE ¼ E1=2 ðE0
EÞ2. Show that in this case the mean energy is
1
3E0 .

7.12 The ground state of 35 P

73 Br has J ¼ 2
1
and the first two excited states have
P 5
P 3

J ¼ 2 ð26:92 keVÞ and J ¼ 2 ð178:1 keVÞ. List the possible -transitions

between these levels and estimate the lifetime of the 32 state.

7.13 Use the Weisskopf formulas of Equations (7.79) to calculate the radiative width
 ðE3Þ expressed in a form analogous to Equations (7.80).
8
Applications of Nuclear Physics

Nuclear physics impinges on our everyday lives1 in a way that particle physics
does not, at least not yet. A minor example of this is radioactive dating of historical
artefacts which we discussed in Chapter 2. It is appropriate, therefore, to discuss
some of these applications. For reasons of space, we will consider just three
important areas: fission, fusion and biomedical applications, concentrating in the
latter on medical imaging and the therapeutic use of radiation.

8.1 Fission
Fission was discussed in Chapter 2 in the context of the semi-empirical mass
formula and among other things we showed that spontaneous fission only occurs
for very heavy nuclei. In this section we discuss fission in more detail in the
context of its use in the production of energy.

8.1.1 Induced fission -- fissile materials

In Chapter 2 we saw that for a nucleus with A  240, the Coulomb barrier, which
inhibits spontaneous fission, is between 5 and 6 MeV. If a neutron with zero kinetic
energy enters a nucleus to form a compound nucleus, the latter will have an
excitation energy above its ground state equal to the neutron’s binding energy in
that state. For example, a zero-energy neutron entering a nucleus of 235 U forms a
state of 236 U with excitation energy of 6.5 MeV. This energy is well above the
fission barrier and the compound nucleus quickly undergoes fission, with decay
products similar to those found in the spontaneous decay of 236 U. To induce fission

1
This is literally true, because we shall see that the energy of the Sun has its origins in nuclear reactions.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
256 CH8 APPLICATIONS OF NUCLEAR PHYSICS

in 238 U on the other hand, requires a neutron with kinetic energy of at least
1.2 MeV. The binding energy of the last neutron in 239 U is only 4.8 MeV and an
excitation energy of this size is below the fission threshold of 239 U.
The differences in the binding energies of the last neutron in even-A and odd-A
nuclei are given by the pairing term in the semi-empirical mass formula.
Examination of the value of this term (see Equation (2.52)) leads to the explana-
tion of why the odd-A nuclei

233 235 239 241

92 U; 92 U; 94 Pu; 94 Pu ð8:1Þ

are ‘fissile’, i.e. fission may be induced by even zero-energy neutrons, whereas the
even-A (even-Z/even-N) nuclei

232 238 240 242

90 Th; 92 U; 94 Pu; 94 Pu ð8:2Þ

require an energetic neutron to induce fission.

A nuclear reactor is a device designed to produce useful energy. The most
commonly used fuel in reactors is uranium, so we will focus on this element.
Natural uranium consists of 99.3 per cent 238 U and only 0.7 per cent 235 U. The
total and fission cross-sections, tot and fission , respectively, for neutrons incident
on 235 U and 238 U are shown in Figure 8.1.
The most important features of these figures are (cf. the discussion of nuclear
reactions in Section 2.9) as follows.

1. At energies below 0.1 eV, tot for 235 U is much larger than that for 238 U and the
fission fraction is large ( 84 per cent). (The other 16 per cent is mainly
radiative capture with the formation of an excited state of 236 U, plus one or
more photons.)

2. In the region between 0.1 eV and 1 keV, the cross-sections for both isotopes
show prominent peaks corresponding to neutron capture into resonances. The
widths
of these states are  0:1 eV and thus their lifetimes are of the order of
h f
1014 s. In the case of 235 U these compound nuclei lead to fission,
f


whereas in the case of 238 U, neutron capture leads predominantly to radiative
decay of the excited state.

3. Above 1 keV, the ratio fission =tot for 235 U is still significant, although smaller
than at very low energies. In both isotopes, tot is mainly due to contributions
from elastic scattering and inelastic excitation of the nucleus.

The fission fragments (which are not unique – several final states are possible)
carry away about 90 per cent of the energy of the primary fission reaction. The
accompanying neutrons in the primary fission process (referred to as prompt
 FISSION 257

Figure 8.1 Total cross-section tot and fission cross-section fission as functions of energy for
neutrons incident on (a)235 U and (b)238 U (adapted from Ga76, Courtesy of Brookhaven National
Laboratory)

neutrons) carry away only about 2 per cent of the energy. For 235 U, the average
number of prompt neutrons per fisson is n
2:5, with the value depending a little
on the incident neutron energy and they have an average energy of about 2 MeV.
In addition to the neutrons produced in the primary fission, the decay products
will themselves decay by chains of -decays and some of the resulting nuclei will
258 CH8 APPLICATIONS OF NUCLEAR PHYSICS

themselves give off further neutrons. This delayed component constitutes about
13 per cent of the energy release in the fission of 235 U. Although the mean delay is
about 13 s, some components have very long lifetimes and may not decay until
many years later. One consequence of this is that a reactor will still produce heat
even after it has ceased to be used for power production and another is that the
delayed component may be emitted after the fuel has been used and removed from
the reactor, leading to the biological hazard of radioactive waste.2

8.1.2 Fission chain reactions

We have seen in Chapter 2 that in each fission reaction a large amount of energy is
produced, which of course is what is needed for power production. However, just
as important is the fact that the fission decay products contain other neutrons. For
example, we have said that in the case of fission of 235 U, on average n ¼ 2:5
neutrons are produced. Since neutrons can induce fission, the potential exists for a
sustained chain reaction, although a number of conditions have to be fulfilled for
this to happen in practice. If we define

number of neutrons produced in the ðn þ 1Þ th stage of fission

k ; ð8:3Þ
number of neutrons produced in the nth stage of fission

then for k ¼ 1 the process is said to be critical and a sustained reaction can occur.
This is the ideal situation for the operation of a power plant based on nuclear
fission. If k < 1, the process is said to be subcritical and the reaction will die out; if
k > 1, the process is supercritical and the energy will grow very rapidly, leading to
an uncontrollable explosion, i.e. a nuclear fission bomb.
Again we will focus on uranium as the fissile material and consider the length
and timescales for a chain reaction to occur. If we assume that the uranium is a
mixture of the two isotopes 235 U and 238 U with an average neutron total cross-
section 
tot , then the mean free path, i.e. the mean distance the neutron travels
between interactions (see Chapter 4), is given by

‘ ¼ 1=ð
nucl 
tot Þ; ð8:4Þ

where
nucl ¼ 4:8 1028 nuclei/m3 is the nuclei density of uranium metal. For
example, the average energy of a prompt neutron from fission is 2 MeV and at this
energy we can see from Figure 8.1 that 
tot
7 barns, so that ‘
3 cm. A 2 MeV
neutron will travel this distance in about 1:5 109 s.
Consider first the case of the explosive release of energy in a nuclear bomb,
using the highly enriched isotope 235 U (for simplicity we will assume a sample of

2
We will return to the effect of radiation on living tissue later in this chapter, in Section 8.3.1.
 FISSION 259

100 per cent 235 U). From Figure 8.1, we see that a neutron with energy of 2 MeV
has a probability of about 18 per cent to induce fission in an interaction with a 235 U
nucleus. Otherwise it will scatter and lose energy, so that the probability for a
further interaction will be somewhat increased (because the cross-section increases
with decreasing energy). If the probability of inducing fission in a collision is p, the
probability that a neutron has induced fission after n collisions is pð1  pÞn1 and
the mean number of collisions to induce fission will be
X
1

n ¼ npð1  pÞn1 ; ð8:5Þ
n¼1

provided the neutron does not escape outside the target. The value of n
can be
estimated using the measured cross-sections
pffiffiffi and is about six. Thus the neutron will
move a linear (net) distance of 3 6 cm
7 cm in a time tp
108 s before
inducing a further fission and being replaced on average by 2.5 new neutrons with
average energy of 2 MeV.3
The above argument suggests that the critical mass of uranium 235 U that would
be necessary to produce a nuclear explosion is a sphere of radius about 7 cm.
However, not all neutrons will be available to induce fission. Some will escape
from the surface and some will undergo radiative capture. If the probability that a
newly created neutron induces fission is q, then each neutron will on average lead
to the creation of ðnq  1Þ additional neutrons in the time tp . If there are NðtÞ
neutrons present at time t, then at time t þ t there will be


Nðt þ tÞ ¼ NðtÞ 1 þ ðnq  1Þ ðt tp Þ ; ð8:6Þ

neutrons and hence

Nðt þ tÞ NðtÞ ðnq  1Þ
¼ : ð8:7Þ
t tp

In the limit as t ! 0, this gives

dN ðnq  1Þ
¼ NðtÞ; ð8:8Þ
dt tp

3
The square root appears because we are assuming that at each collision the direction changes randomly, i.e.
the neutron executes a random walk. Thus if the distance travelled in the ith collision is li , the displacement
P
n
vector d after n collisions will be d ¼ li and the net distance travelled d will be given by
i¼1

n X
X n
d2 ¼ ðli lj Þ ¼ l21 þ l22 þ l23 þ þ l2n þ 2ðl1 l2 þ l1 l3 þ Þ:
i¼1 j¼1

When the average is taken over several collisions, the scalar products will cancel because the direction of
pffiffiffi
each step is random. Finally, setting li ¼ l, the mean distance travelled per collision, gives d ¼ l n.
260 CH8 APPLICATIONS OF NUCLEAR PHYSICS

and hence by integrating Equation (8.8)



NðtÞ ¼ Nð0Þ exp ðnq  1Þt tp : ð8:9Þ

Thus the number increases or decreases exponentially, depending on whether

nq > 1 or nq < 1. For 235 U, the number increases if q > 1=n
0:4 (recall that
n
2:5). Clearly if the dimensions of the metal are substantially less than 7 cm, q
will be small and the chain reaction will die out exponentially. However, a
sufficiently large mass brought together at t ¼ 0 will have q > 0:4. There will
always be some neutrons present at t ¼ 0 arising from spontaneous fission and
since tp
108 s, an explosion will occur very rapidly. For a simple sphere of 235 U
the critical radius at which nq ¼ 1 is actually close to 9 cm and the critical mass is
about 50 kg.
Despite the above simple analysis, it is not easy to make a nuclear bomb! This is
because the thermal energy released as the assembly becomes critical will produce
an outward pressure that is sufficient to blow apart the fissile material unless
special steps are taken to prevent this. In early ‘atom bombs’, a sub-critical mass
was assembled and a small plug fired into a prepared hollow in the material so that
the whole mass became supercritical. In later devices, the fissile material was a
sub-critical sphere of 239 Pu surrounded by chemical explosives. These were
specially designed (‘shaped’) so that when they exploded, the resulting shock
wave imploded the plutonium, which as a result became supercritical.

8.1.3 Nuclear power reactors

The production of power in a controlled way for peaceful use is carried out in a
nuclear reactor and is just as complex as producing a bomb. There are several
distinct types of reactor available. We will discuss just one of these, the thermal
reactor, which uses uranium as the fuel and low-energy neutrons to establish a
chain reaction. The discussion will concentrate on the principles operating such a
reactor and not on practical details.
A schematic diagram of the main elements of a generic example of a thermal
reactor is shown in Figure 8.2. The most important part is the core, shown
schematically in Figure 8.3. This consists of fissile material (fuel elements),
control rods and the moderator. The roles of the control rods and the moderator
will be explained later. The most commonly used fuel is uranium and many
thermal reactors use natural uranium, even though it has only 0.7 per cent of 235 U.
Because of this, a neutron is much more likely to interact with a nucleus of 238 U.
However, a 2 MeV neutron from the primary fission has very little chance of
inducing fission in a nucleus of 238 U. Instead it is much more likely to scatter
inelastically, leaving the nucleus in an excited state and after a couple of such
collisions the energy of the neutron will be below the threshold of 1.2 MeV for
inducing fission in 238 U.
 FISSION 261

Figure 8.2 Sketch of the main elements of a thermal reactor -- the components are not to scale
(after Li01, Copyright, John Wiley & Sons)

Figure 8.3 Sketch of the elements of the core of a reactor

262 CH8 APPLICATIONS OF NUCLEAR PHYSICS

A neutron with its energy so reduced will have to find a nucleus of 235 U if it is
to induce fission, but its chances of doing this are very small unless its energy
has been reduced to very low energies below 0.1 eV, where the cross-section is
large (see Figure 8.1). Before that happens it is likely to have been captured into
one of the 238 U resonances with the emission of photons. Thus, to sustain a chain
reaction, either the fuel must be enriched with a greater fraction of 235 U (2–
3 per cent is common in some types of commercial reactor), or if natural uranium
is to be used, some method must be devised to overcome this problem.
This is where the moderator comes in. This surrounds the fuel elements and
its volume is much greater than that of the fuel elements themselves. Its main
purpose is to slow down fast neutrons produced in the fission process. Fast
neutrons will escape from the fuel rods into the moderator and are reduced to very
low energies by elastic collisions. In this way the absorption into resonances of
238
U is avoided. The moderator must therefore be a material with a negligible
cross-section for absorption and ideally should also be inexpensive. In practice,
heavy water (a form of water where the hydrogen atoms are replaced by atoms of
deuterium), or carbon (in the form of graphite), are the moderators of choice in
many thermal reactors using natural uranium. For enriched reactors, ordinary
water may be used.
Consider now the stability of the chain reaction. This is where the control rods
play their part. They are usually made of cadmium, which has a very high
absorption cross-section for neutrons. By mechanically manipulating the control
rods, i.e. by retracting or inserting them, the number of neutrons available to
induce fission can be regulated. This mechanism is the key to maintaining a
constant k value of unity and therefore a constant power output. However, safe
working of the reactor is not possible with prompt neutrons alone. To see this, we
return to Equation (8.9) and set nq  1 ¼ k  1, so that



NðtÞ ¼ Nð0Þ exp ðk  1Þt tp : ð8:10Þ

The value of tp is determined by the mean free path for neutron absorption and
unlike the case of pure 235 U we considered in Section 8.1.2, is given approximately
by tp  103 s. Thus, for example, if we take k ¼ 1:001, i.e. an increase of only
0.1 per cent, the reactor flux would increase by e60
1026 in only one minute.
Clearly a much smaller rate of increase has to be achieved for safe manipulation of
the control rods if a disaster is to be averted. This is where the delayed neutrons
play a crucial role.
In a nuclear weapon, the delayed neutrons are of no consequence, because the
explosion will have taken place long before they would have been emitted, but in a
power reactor they are vital for reactor safety. Taking account of delayed neutrons,
each fission leads to ½ðn þ  nÞq  1 additional neutrons, where we have defined
 n as the number of delayed neutrons per fission. In practice n  0:02. In the
steady-state operation, with constant energy output, the neutron density must
 FISSION 263

remain constant (i.e. k ¼ 1 in Equation (8.3)). Thus q must satisfy the critical
condition

ðn þ nÞq  1 ¼ 0: ð8:11Þ

Equation (8.10) is now modified to have an additional term that depends on the
mean time td of the delayed neutrons, which is about 13 s. Provided
nðk  1Þ  n, it is the latter term that dominates and (without proof) the modified
form of Equation (8.10) is given approximately by
 
nðk  1Þt
NðtÞ
Nð0Þ exp : ð8:12Þ
½n  nðk  1Þtd

Thus the timescale to manipulate the control rods is determined by that of the
delayed neutrons. For example, using n ¼ 2:5; n ¼ 0:02; k ¼ 1:001 and td ¼ 13 s
in Equation (8.12) gives an increase in reactor flux of less than a factor two in one
minute. Clearly, the precise increase is sensitive to the parameters chosen, but
factors of this size are manageable. The reactor design therefore ensures that
nq  1 < 0 always, so that the reactor can only become critical in the presence of
delayed neutrons.
This simple discussion has ignored many practical details that will modify the
real formulas used in reactor dynamics, such as the fact that the fuel and moderator
are not uniformly distributed throughout the core and that some of the fission
products themselves have appreciable cross-sections for neutron absorption and
will therefore reduce the flux of neutrons available to sustain the chain reaction.4
Returning to Figure 8.2, the core is surrounded by a coolant (often water), which
removes the heat generated in the core from the energy deposited by the fission
fragments. A thick concrete shield to prevent radiation leaks surrounds the entire
set-up. At start-up, the value of k is set slightly higher than unity and is kept at that
value until the desired power output is achieved and the operating temperature is
reached, after which the value of k is lowered by adjusting the control rods. It is
very important for safety reasons that dq=dT < 0, so that an increase in tempera-
ture T leads to a fall in reaction rate. The rest of the plant is conventional
engineering. Thus, the heated coolant gives up its heat in a heat exchanger and is
used to boil water and drive a steam turbine, which in turn produces electricity.
It is worth calculating the efficiency with which one can expect to produce
energy in a nuclear reactor. We can use the SEMF to calculate the energy released
during fission, by finding the binding energies of the two fission products and
comparing their sum to the binding energy of the decaying nucleus. For the fission
of a single 235 U nucleus this is 200 MeV or 3:2 1011 J. (As we have mentioned
above, about 90 per cent of this is in the form of ‘prompt’ energy.) We also know

4
More details of reaction dynamics are discussed in, for example, Section 10.3 of Li01. In Section 10.6 of
this reference there is also a discussion of several other types of commercial reactor.
264 CH8 APPLICATIONS OF NUCLEAR PHYSICS

that 1 g of any element contains NA =A atoms, where NA is Avogadro’s number.

Thus 1g of 235 U has about 6 1023 =235
3 1021 atoms and if fission were
complete would yield a total energy of about 1011 J, or 1 MW-day. This is about
3 106 times greater than the yield obtained by burning (chemical combustion)
1 g of coal. In practice only about 1 per cent of the energy content of natural
uranium can be extracted (the overall efficiency is greatly reduced by the
conventional engineering required to produce electricity via steam turbines), but
this can be increased significantly in another type of reactor, called a fast breeder
discussed briefly below.
We can also calculate the power output from an ideal thermal reactor for a given
mass of uranium. From Equation (1.44) of Chapter 1, the reaction rate for fission
Wf is given by

Wf ¼ J N fission ; ð8:13Þ

where J is the flux, N is the number of nuclei undergoing fission and fission is the
fission cross-section. Consider, for example, a reactor containing 100 tonnes of
natural uranium, generating a neutron flux of 1013 cm2 s1 and with a fission
cross-section for 235 U of 580 b at the appropriate energy (see Figure 8.1). Since the
fraction of 235 U in natural uranium is 0.072 per cent, the number of 235 U nuclei
undergoing fission is given by

100 103 0:0072 NA

N¼ ¼ 1:82 1027 ; ð8:14Þ
A

where A ¼ 238:03 is the mass number of natural uranium. The power generated is
thus

P ¼ Wf E; ð8:15Þ

where E ¼ 200 MeV is the total energy released per fission (see above). Evaluating
Equation (8.15) gives P
340 MW. In addition to causing fission, neutrons will be
absorbed by 235 U without causing fission. If the total absorption cross-section a is
680 b, then the number of 235 U nuclei that will be consumed per second will be
N Ja , i.e. 1:24 1019 s1 . Since we started with 1:82 1027 nuclei, the fuel will
be used at the rate of about 1.8 per cent per month.
We turn now to the fast breeder reactor mentioned above. In this reactor there
is no large volume of moderator and no large density of thermal neutrons is
established. In such a reactor, the proportion of fissile material is increased to
about 20 per cent and fast neutrons are used to induce fission. The fuel used is
239
Pu rather than 235 U, the plutonium being obtained by chemical separation from
the spent fuel rods of a thermal reactor. This is possible because some 238 U nuclei
in the latter will have captured neutrons to produce 239 U, which subsequently
 FISSION 265

decays via a chain of -decays to plutonium. The whole sequence is as follows:

n þ 238 U ! 239 U ð23 minsÞ ! 239 Np ð2:4 daysÞ ! 239 Pu ð2:4 104 yearsÞ: ð8:16Þ

The mean number of neutrons produced in the fission of 239 Pu is 2.96, so this
nucleus is very suitable for use in a fast reactor. In practice, the core is a mixture of
20 per cent 239 Pu and 80 per cent 238 U surrounded by a blanket of more 238 U,
where more plutonium is made. The 238 U obtained from spent fuel rods in thermal
reactors is called depleted uranium. Such a reactor can produce more fissile 239 Pu
than it consumes, hence the name ‘breeder’. In principle such a reactor can
consume all the energy content of natural uranium, rather than the 1 per cent used
in thermal reactors, although in practice there are limits to its efficiency.
Whatever type of reactor is used, a major problem is the generation of
radioactive waste, including transuranic elements and long-lived fission fragments,
which in some cases may have to be stored safely for hundreds of years.5 Much
effort has been expended on this problem, but a totally satisfactory solution is still
not available. Short-lived waste with low activity (for example, consumables such
as protective clothing) is simply buried in the ground. One idea for long-lived
waste with high activity is to ‘glassify’ it into stable forms that can be stored
underground without risk of spillage, leakage into the water table, etc..
A particularly ingenious idea is to ‘defuse’ long-lived fission fragments by using
the resonance capture of neutrons to convert them to short-lived, or even stable,
nuclei. For example, 99 Tc (technetium), which concentrates in several organs of
the body and also in the blood, has a very long half-life. However, it has a large
resonant cross-section for neutron capture to a completely stable isotope 100 Ru
(ruthenium) and in principle this reaction could be used to ‘neutralize’ 99 Tc.
Needless to say, the problems to be overcome are far from trivial. First, the amount
of radioactive waste is very large, so one problem is to find a source of neutrons
capable of handling it. (Reactors themselves are one possible source!) Secondly,
the neutron energy has to be matched to the particular waste material, which
therefore would have to be separated and prepared before being bombarded by
the neutrons. All this would take energy and would increase the overall cost of
energy production by nuclear power, which is already more expensive than
conventional burning of fossil fuels. Nevertheless, there is considerable interest
in the principle of this method and proposals have been made to exploit it without
the attendant drawbacks above. We will return to this in Chapter 9, where we
discuss some of the outstanding problems in nuclear physics and their possible
future solutions. However, until such time as this, or some other, method is
realized in practice, the safe long-term disposal of radioactive waste remains a
serious unsolved problem.

5
In principle, there would be no such problem with fast breeder reactors, but in practice the ideal is not
realized.
266 CH8 APPLICATIONS OF NUCLEAR PHYSICS

8.2 Fusion
We have seen that the plot of binding energy per nucleon (Figure 2.2) has a
maximum at A
56 and slowly decreases for heavier nuclei. For lighter nuclei,
the decrease is much quicker, so that with the exception of magic nuclei, lighter
nuclei are less tightly bound than medium size nuclei. Thus, in principle, energy
could be produced by two light nuclei fusing to produce a heavier and more
tightly bound nucleus – the inverse process to fission. Just as for fission, the energy
released comes from the difference in the binding energies of the initial and final
states. This process is called nuclear fusion. Since light nuclei contain fewer
nucleons than heavier nuclei, the energy released per fusion is smaller than in
fission. However, as a potential source of power, this is more than balanced by the
far greater abundance of stable light nuclei in nature than very heavy nuclei. Thus
fusion offers enormous potential for power generation, if the huge practical
problems could be overcome.

8.2.1 Coulomb barrier

The practical problem to obtaining fusion, whether in power production or more

generally, has its origin in the Coulomb repulsion, which inhibits two nuclei
getting close enough together to fuse. This is given by the Coulomb potential

1 ZZ 0 e2
VC ¼ ; ð8:17Þ
4"0 R þ R0

where Z and Z 0 are the atomic numbers of the two nuclei and R and R0 are their
effective radii. The quantity ðR þ R0 Þ is therefore classically the distance of closest
approach. Recalling, from the work on nuclear structure in Chapter 2, that for
1
medium and heavy nuclei R ¼ 1:2A3 fm, we have

e2
hc Z Z 0 Z Z0
VC ¼  1=3  ¼ 1:198 MeV: ð8:18Þ
4"0
hc 1:2 A1=3 þ ðA0 Þ fm A1=3 þ ðA0 Þ1=3

If, for illustration, we set A
A0
2Z
2Z 0 , then

5
VC
0:15A3 MeV: ð8:19Þ

Thus, for example, with A
8, VC
4:8 MeV and this energy has to be supplied
to overcome the Coulomb barrier.
This is a relatively small amount of energy to supply and it might be thought
that it could be achieved by simply colliding two accelerated beams of light nuclei,
but in practice nearly all the particles would be elastically scattered. The only
 FUSION 267

practical way is to heat a confined mixture of the nuclei to supply enough thermal
energy to overcome the Coulomb barrier. The temperature necessary may
be estimated from the relation E ¼ kT, where kB is Boltzmann’s constant, given
by kB ¼ 8:6 105 eV K1. For an energy of 4.8 Mev, this implies a temperature
of 5:6 1010 K. This is well above the typical temperature of 108 K found in
stellar interiors.6 It is also the major hurdle to be overcome in achieving a
controlled fusion reaction in a reactor, as we shall see later.
Fusion actually occurs at a lower temperature than this estimate due to a
combination of two reasons. The first and most important is the phenomenon of
quantum tunnelling, which means that the full height of the Coulomb barrier does
not have to be overcome. In Chapter 7 we discussed a similar problem in the
context of -decay, and we can draw on that analysis here. The probability of
barrier penetration depends on a number of factors, but the most important is the
Gamow factor, which is a function of the relative velocities and the charges of the
reaction products. In particular, the probability is proportional to exp½GðEÞ,
where GðEÞ is apgeneralization
ffiffiffiffiffiffiffiffiffiffiffiffi of the Gamow factor of Chapter 7. This may be
written as G ¼ EG =E, where again, generalizing the equations in Chapter 7,
EG ¼ 2mc2 ðZ1 Z2 Þ2 : ð8:20Þ
Here, m is the reduced mass of the two fusing nuclei and they have electric charges
Z1 e and Z2 e. Thus the probability of barrier penetration increases as E increases.
Nevertheless, the probability of fusion is still extremely small. For example, if we
consider the fusion of two protons (which we will see below is an important
ingredient of the reactions that power the Sun), at a typical stellar temperature of
107 K, we find EG
490 keV and E
1 keV. Hence the probability of fusion is
proportional to exp½ðEG =EÞ1=2 
expð22Þ
109:6 which is a very large
suppression factor and so the actual fusion rate is still extremely slow.
The other reason that fusion occurs at a lower temperature than expected is that
a collection of nuclei at a given mean temperature, whether in stars or elsewhere,
will have a Maxwellian distribution of energies about the mean and so there will be
some with energies substantially higher than the mean energy. Nevertheless, even a
stellar temperature of 108 K corresponds to an energy of only about 10 keV, so the
fraction of nuclei with energies of order 1 MeV in such a star would only be of the
order of expðE=kTÞ  expð100Þ  1043 , a minute amount. We will return to
these questions in more detail in Section. 8.2.3.

8.2.2 Stellar fusion

The energy of the Sun comes from nuclear fusion reactions, foremost of which is
the so-called proton–proton cycle. This has more than one branch, but one of these,

6
Because of this, many scientists refused to accept that fusion occurred in stars when the suggestion was first
made. When challenged on this, Eddington’s reposte was simple: ‘. . . . go and find a hotter place’.
268 CH8 APPLICATIONS OF NUCLEAR PHYSICS

the PPI cycle, is dominant. This starts with the fusion of hydrogen nuclei to
produce deuterium via the weak interaction:
1
H þ 1 H ! 2 H þ eþ þ e þ 0:42 MeV: ð8:21Þ

The deuterium then fuses with more hydrogen to produce 3 He via the electro-
magnetic interaction:
1
H þ 2 H ! 3 He þ  þ 5:49 MeV ð8:22Þ

and finally, two 3 He nuclei fuse to form 4 He via the nuclear strong interaction:
3
He þ 3 He ! 4 He þ 2ð1 HÞ þ 12:86 MeV: ð8:23Þ

The relatively large energy release in the last reaction is because 4 He is a doubly
magic nucleus and so is very tightly bound. The first of these reactions, being a
weak interaction, proceeds at an extremely slow rate and sets the scale for the long
lifetime of the Sun. Combining these equations, we have overall

4ð1 HÞ ! 4 He þ 2eþ þ 2e þ 2 þ 24:68 MeV: ð8:24Þ

Because the temperature of the Sun is  107 K, all its material is fully ionized.
Matter in this state is referred to as a plasma. The positrons produced above will
annihilate with electrons in the plasma to release a further 1.02 MeV of energy per
positron and so the total energy released is 26.72 MeV. However, of this each
neutrino will carry off 0.26 MeV on average, which is lost into space.7 Thus on
average, 6.55 MeV of electromagnetic energy is radiated from the Sun for every
proton consumed in the PPI chain.
The PPI chain is not the only fusion cycle contributing to the energy output of
the Sun, but it is the most important. Another interesting cycle is the carbon, or
CNO chain. Although this contributes only about 3 per cent of the energy output
of the Sun, it plays an important role in the evolution of other stellar objects. In the
presence of any of the nuclei 126 C, 136 C, 147 N or 157 N, hydrogen will catalyse burning
via the reactions
12
C þ 1 H ! 13 N þ  þ 1:95 MeV
ð8:25Þ
13
N ! 13 C þ eþ þ e þ 1:20 MeV
13
C þ 1 H ! 14 N þ  þ 7:55 MeV ð8:26Þ
14
N þ 1 H ! 15 O þ  þ 7:34 MeV
ð8:27Þ
15
O ! 15 N þ eþ þ e þ 1:68 MeV

7
These are the main contributors to the neutrino flux observed at the surface of the Earth that was discussed
in Chapter 3.
 FUSION 269

and

15
N þ 1 H ! 12 C þ 4 He þ 4:96 MeV ð8:28Þ

Thus, overall in the CNO cycle we have

4 ð1 HÞ ! 4 He þ 2eþ þ 2e þ 3 þ 24:68 MeV: ð8:29Þ

These and other fusion chains all produce electron neutrinos as final-state products
and using detailed models of the Sun, the flux of such neutrinos at the surface of
the Earth can be predicted.8 The actual count rate is far lower than the theoretical
expectation. This is the solar neutrino problem that we met in Section 3.1.4. The
solution to this problem is almost certainly neutrino oscillations, where some e
are converted to neutrinos of other flavours in their passage from the Sun to the
Earth. We saw in Chapter 3 that this is only possible if neutrinos have mass, so a
definitive measurement of neutrino masses would be an important piece of
evidence to finally resolve the solar neutrino problem. Such measurements should
be available in a few years.
The process whereby heavier elements (including the 12 C required in the
CNO cycle) are produced by fusion of lighter ones can continue beyond the
reactions above. For example, when the hydrogen content is depleted, at high
temperatures helium nuclei can fuse to form an equilibrium mixture with 8 Be via
the reaction

4
He þ 4 He Ð 8 Be ð8:30Þ

and the presence of 8 Be allows the rare reaction

4
He þ 8 Be ! 12 C ð8:31Þ

to occur, where C is an excited state of carbon. A very small fraction of the latter
will decay to the ground state, so that overall we have9

3ð4 HeÞ ! 12 C þ 7:27 MeV: ð8:32Þ

8
The expectations are based on a detailed model of the Sun known as the standard solar model that we met in
Chapter 3.
9
The occurrence of this crucial reaction depends critically on the existence of a particular excited state of 12 C.
For a discussion of this and the details of the other reactions mentioned below see, for example, Section 4.3
of Ph94. Very recent experiments (2005) have found evidence for other nearby excited states that change the
accepted energy dependence (or equivalently the temperature dependence) of this reaction which will have
implications for theories of stellar evolution.
270 CH8 APPLICATIONS OF NUCLEAR PHYSICS

The presence of 12 C enables another series of fusion reactions to occur, in addition

to the CNO cycle. Thus 16 O can be produced via the reaction
4
He þ 12 C ! 16 O þ  ð8:33Þ
and the production of neon, sodium and magnesium is possible via the reactions
12
C þ 12 C ! 20 Ne þ 4 He ; 23
Na þ p ; 23
Mg þ n: ð8:34Þ
Fusion processes continue to synthesize heavier elements until the core of the
stellar object is composed mainly of nuclei with A
56, i.e. the peak of the
binding energy per nucleon curve. Heavier nuclei are produced in supernova
explosions, but this is properly the subject of astrophysics and we will not pursue it
further here, although we will return to it briefly in Chapter 9.

8.2.3 Fusion reaction rates

We have mentioned in Section 8.2.1 that quantum tunnelling and the Maxwellian
distribution of energies combine to enable fusion to occur at a lower temperature
than might at first be expected. The product of the increasing barrier penetration
factor with energy and the Maxwellian decreasing exponential actually means that
in practice fusion takes place over a rather narrow range of energies. To see this we
will consider the fusion between two types of nuclei, a and b, having number
densities na and nb (i.e. the number of particles per unit volume) and at a
temperature T. We assume that the temperature is high enough so that the nuclei
form a plasma, with uniform values of number densities and temperature. We also
assume that the velocities of the two nuclei are given by the Maxwell–Boltzmann
distribution, so that the probability of having two nuclei with a relative speed v in
the range v to v þ dv is

1=2   
2 m 3=2 mv2 2
PðvÞ dv ¼ exp v dv; ð8:35Þ
 kT 2kT
where m is the reduced mass of the pair. The fusion reaction rate per unit volume is
then
Rab ¼ na nb hab vi; ð8:36Þ
where ab is the fusion cross-section10 and the brackets denote an average, i.e.
ð
1

hab vi  ab vPðvÞ dv: ð8:37Þ

0

10
The product nA nB is the number of pairs of nuclei that can fuse. If the two nuclei are of the same type, with
nA ¼ nB ¼ n, then the product must be replaced by 12 nðn  1Þ
12 n2, because in quantum theory such nuclei
are indistinguishable.
 FUSION 271

The fusion cross-section may be written

"
#
SðEÞ EG 1=2
ab ðEÞ ¼ exp  ; ð8:38Þ
E E

where the exponential follows from the previous discussion of quantum tunnelling
and SðEÞ contains the details of the nuclear physics. The term 1=E is conveniently
factored out because many nuclear cross-sections have this behaviour at low
energies. Using (8.35) and (8.38) in (8.37) gives, from (8.36):

1=2
3=2 1
ð "
1=2 #
8 1 E EG
Rab ¼ na nb SðEÞexp   dE: ð8:39Þ
m kT kT E
0

Because the factor 1=E has been taken out of the expression for ðEÞ, the quantity
SðEÞ is slowly varying and the behaviour of the integrand is dominated by the
behaviour of the exponential term. The falling exponential of the Maxwellian
energy distribution combines with the rising exponential of the quantum tunnelling
effect to produce a maximum in the integrand at E ¼ E0 where

 1=3
1
E0 ¼ EG ðkTÞ2 ð8:40Þ
4

and fusion takes place over a relatively narrow range of energies E0  E0 where

4 1=6
E0 ¼ EG ðkTÞ5=6 : ð8:41Þ
31=2 21=3

The importance of the temperature and the Gamow energy EG ¼ 2mc2 ðZa Zb Þ2
is clear. A schematic illustration of the interplay between these two effects is
shown in Figure 8.4.
As a real example, consider the pp reaction (Equation (8.21)), at a temperature
of 2 107 K. We have EG ¼ 493 keV and kT ¼ 1:7 keV, so that fusion is most
likely at E0 ¼ 7:2 keV and the half-width of the distribution is E0 =2 ¼ 4:1 keV.
The resulting function exp½E=kT  ðEG =EÞ1=2  is shown in Figure 8.5.
In the approximation where SðEÞ is taken as a constant SðE0 Þ, the integral in
Equation (8.39) may be done and gives

1=2
8 2
hab vi
SðE0 Þ  2 exp½ ; ð8:42Þ
9 3mEG
 2=3
where  ¼ 3 12 ðEG =kTÞ1=3 :
272 CH8 APPLICATIONS OF NUCLEAR PHYSICS

Figure 8.4 The right-hand dashed curve is proportional to the barrier penetration factor and
the left-hand dashed curve is proportional to the Maxwell distribution. The solid curve is the
combined effect and is proportional to the overall probability of fusion with a peak at E0 and a
width of E0

Figure 8.5 The exponential part of the integrand in Equation (8.39) for the case of pp fusion at
a temperature of 2 107 K
 FUSION 273

If we take the masses to be Aa;b in atomic mass units we can evaluate

Equation (8.36) using the expression (8.20) for EG to give

7:21 1022 ðAa þ Ab Þ SðE0 Þ

Rab ¼ na nb  2 exp½  m3 s1 ; ð8:43Þ
Za Zb Aa Ab 1 MeV b

with

1=3

Aa Ab 1 keV 1=3
 ¼ 18:8ðZa Zb Þ2=3 : ð8:44Þ
Aa þ Ab kT

The rate depends very strongly on both the temperature and the nuclear species
because of the factor  2 exp½ . This is illustrated in Figure 8.6 for the pp and
p12 C reactions, the initial reactions in the pp and CNO cycles.

Figure 8.6 The function  2 expðÞ of Equation (8.43) for the pp and p12 C reactions

8.2.4 Fusion reactors

There is currently an international large-scale effort to achieve controlled fusion in

the laboratory, with the eventual aim of producing power. For this, the pp reactions
are far too slow to be useful. However, the Coulomb barrier for the deuteron 21 H is
the same as for the proton and the exothermic reactions
2
1H þ 21 H ! 32 H þ n þ 3:27 MeV ð8:45aÞ

and
2
1H þ 21 H ! 31 H þ p þ 4:03 MeV ð8:45bÞ
274 CH8 APPLICATIONS OF NUCLEAR PHYSICS

Figure 8.7 Values of the quantity h i for the d--t reaction of Equation (8.46) and the
combined d--d reactions of Equations (8.45) (adapted from Ke82 and reproduced by permission
of Annual Reviews)

suggest that deuterium might be a suitable fuel for a fusion reactor. Deuterium is
also present in huge quantities in sea water and is easy to separate at low cost.
An even better reaction in terms of energy output is deuterium–tritium fusion:
2
1H þ 31 H ! 42 He þ n þ 17:62 MeV: ð8:46Þ

The values of hvi for the d–t reaction of Equation (8.46) and the combined d–d
reactions of Equations (8.45) are shown in Figure 8.7. It can be seen that the
deuterium–tritium (d–t) reaction has the advantage over the deuterium–deuterium
(d–d) reaction of a much higher cross-section. The heat of the reaction is also
greater. The principal disadvantage is that tritium does not occur naturally (it has a
mean life of only 17.7 years) and has to be manufactured, which increases the
overall cost. From Figure 8.7 it can be seen that the rate for the d–t reaction peaks
at about E ¼ kT ¼ 3040 keV and a working energy where the cross-section is
still considered reasonable is about 20 keV, i.e. 3 108 K.
The effective energy produced by the fusion process will be reduced by the heat
radiated by the hot plasma. The mechanism for this is predominantly electron
bremmstrahlung. The power loss per unit volume due to this process is propor-
tional to T 1=2 Z 2 , where Z is the atomic number of the ionized atoms. Thus for a
plasma with given constituents and at a fixed ion density, there will be a minimum
 FUSION 275

temperature below which the radiation losses will exceed the power produced
by fusion. For example, for the d–t reaction with an ion density 1021 m3,
kTmin
4 keV. It would be 10 times larger for the d–d reaction of Equation (8.45a)
because of the form of hvi (see Figure 8.7), which is another reason for using the
d–t reaction. In practice, the situation is worse than this because most of the
neutrons in Equation (8.46) will escape, so even at the theoretical ‘break-even’
temperature, external energy would have to be supplied to sustain the fusion
process. Only when the energy deposited in the plasma by the -particles exceeds
the radiation losses would the reaction be self-sustaining. This is referred to as the
‘ignition point’.
A numerical expression that embodies these ideas is the so-called Lawson
criterion, which provides a measure of how close to practicality is a particular
reactor design. We will assume a d–t reaction. To achieve a temperature T in a
deuterium–tritium plasma, there has to be an input of energy 4nd ð3kT=2Þ per unit
volume. Here nd is the number density of deuterium ions and the factor of 4 comes
about because nd is equal to the number density of tritium ions and the electron
density is twice this, giving 4nd particles per unit volume. The reaction rate in the
plasma is n2d hdt vi. If the plasma is confined for time tc , then per unit volume of
plasma,

energy output n2d hdt vi tc Q nd hdt vi tc Q

L ¼ ¼ ; ð8:47Þ
energy input 6nd k T 6kT

where Q is the energy released in the fusion reaction. For a useful device, L > 1.
For example, If we assume k T ¼ 20 keV and use the experimental value
hdt vi
1022 m3 s1 , then the Lawson criterion may be written

nd tc > 7 1019 m3 s: ð8:48Þ

Thus either a very high particle density or a long confinement time, or both, is
required.
At the temperatures required for fusion, any material container will vaporize and
so the central problem is how to contain the plasma for sufficiently long times for
the reaction to take place. The two main methods are magnetic confinement and
inertial confinement. Both techniques present enormous technical challenges. In
practice, most work has been done on magnetic confinement and so this method
will be discussed in more detail than the inertial confinement method.
In magnetic confinement, the plasma is confined by magnetic fields and heated
by electromagnetic fields. Firstly we recall the behaviour of a particle of charge q
in a uniform magnetic field B, taking the two extreme cases where the velocity v of
the particle is (a) at right angles to B and (b) parallel to B. In case (a) the particle
traverses a circular orbit of fixed radius (compare the principle of the cyclotron
discussed in Chapter 4) and in case (b) the path is a helix of fixed pitch along the
direction of the field (compare the motion of electrons in a time projection
276 CH8 APPLICATIONS OF NUCLEAR PHYSICS

chamber, also discussed in Chapter 4). Two techniques have been proposed to stop
particle losses: magnetic ‘mirrors’ and a geometry that would ensure a stable
indefinite circulation. In the former, it is arranged that the field in a region is
greater at the boundaries of the region than in the interior. Then as the particle
approaches the boundary, the force it experiences will develop a component that
points into the interior where the field is weaker. Thus the particle is trapped and
will oscillate between the interior and the boundaries.11 However, most practical
work has been done on case (b) and for that reason we will restrict our discussion
to this technique.
The simplest configuration is a toroidal field produced by passing a current
through a doughnut-shaped solenoid. In principle, charged particles in such a field
would circulate endlessly, following helical paths along the direction of the
magnetic field. In practice, the field would be weaker at the outer radius of the
torus and the non-uniformity of the field would produce instabilities in the orbits of
some particles and hence lead to particle loss. To prevent this a second field is
added called a poloidal field. This produces a current around the axis of the torus
and under the combined effect of both fields, charged particles in the plasma
execute helical orbits about the mean axis of the torus. Most practical realizations
of these ideas are devices called tokamaks, in which the poloidal field is generated
along the axis of the torus through the plasma itself.
One of the largest tokamaks in existence is the Joint European Torus (JET), which
is a European collaboration and sited at the Culham Laboratory in Berkshire, UK. A
schematic view of the arrangement of the fields in JET is shown in Figure 8.8(a).
This shows the external coils that generate the main toroidal field. The poloidal field
is generated by transformer action on the plasma. The primary windings of the
transformer are shown with the plasma itself forming the single-turn secondary.
The current induced in the plasma not only generates the poloidal field, but also
supplies several megawatts of resistive heating to the plasma. However, even this is
insufficient to ensure a sufficient temperature for fusion and additional energy is
input via other means, including rf sources.
In the inertial confinement method, small pellets of the deuterium–tritium ‘fuel’
mixture are bombarded with intense energy from several directions simultaneously
which might, for example, be supplied by pulsed lasers. As material is ejected from the
surface, other material interior to the surface is imploded, compressing the core of the
pellet to densities and temperatures where fusion can take place. The laser pulses are
extremely short, typically 107 109 s, which is many orders of magnitude shorter
than the times associated with the pulsed poloidal current in a tokamak (which could
be as long as 1s), but this is compensated for by much higher plasma densities.
Considerable progress has been made towards the goal of reaching the ignition
point. However, although appropriate values of nd , tc , and T have been obtained

11
The Van Allen radiation belts that occur at high altitudes consist of charged particles from space that have
become trapped by a magnetic mirror mechanism because the Earth’s magnetic field is stronger at the poles
than at the equator.
Figure 8.8 Schematic diagrams showing: (a) the main magnetic field components of the JET tokamak; (b) how these elements are incorporated into the JET
device (courtesy of EFDA--JET)
278 CH8 APPLICATIONS OF NUCLEAR PHYSICS

separately, to date no device has yet succeeded in achieving the Lawson criterion.
Tokamaks have reached the break-even point, but the best value of the Lawson
ratio that has been achieved is still about a factor of two too small. Much work
remains to be done on this important problem and in recognition of this at least one
major new tokamak machine is planned as a global collaboration. Even when the
ignition point is achieved, experience with fission power reactors suggests that it
will probably take decades of further technical development before fusion power
becomes a practical reality.

8.3 Biomedical Applications

The application of nuclear physics to biomedicine is a very large subject and for
reasons of space we will therefore concentrate on just two topics: the therapeutic
uses of radiation and medical imaging.

8.3.1 Biological effects of radiation: radiation therapy

Radiation therapy is a long-standing treatment for cancer, often combined with

chemotherapy and/or surgery. By damaging DNA, the ability of the cell to
reproduce is inhibited and so tumour tissue can, in principle, be destroyed.
However, the same of course applies to healthy tissue so, when using radiation
in a medical environment, a balance has to be struck between the potential
diagnostic and/or therapeutic benefits and the potential deleterious effects of
damage done by the radiation. This is a particularly delicate balance for cancer
treatment because, as we shall see below, highly oxygenated tissue has a greater
sensitivity to radiation and unfortunately many tumours are less oxygenated than
healthy tissue and therefore more resistant to radiation. We start by reviewing the
biological effects of radiation and then describe the use of various types of
radiation for cancer treatment.
Exposure of living tissue to radiation is a complex process. Immediate physical
damage may be caused by the initial deposition of energy, but in addition there can
also be secondary damage due to the production of highly active chemicals. The
latter may not be evident in full for several hours after exposure. For low levels of
radiation this effect is the only one. High levels of damage may lead to the rapid
death of living cells, but cells that survive in a damaged form may still have serious
consequences. However caused, damage to the DNA of the nucleus of cells can
result in long-term biological effects, such as cancer or genetic abnormalities,
which may not reveal themselves for years, even decades, after the original
exposure.12

12
This has been known for a long time. For example, Hermann Muller was awarded the 1946 Nobel Prize in
Physiology and Medicine for his discovery that mutations can be induced by X-rays.
 BIOMEDICAL APPLICATIONS 279

To make descriptions like ‘low-level’ and ‘high-level’ used above meaningful

needs a more detailed discussion, including the question of how dosages are
defined. We will do this only very briefly. Roughly speaking, the average absorbed
dose is the total energy deposited per unit mass of tissue. This is measured in
‘grays’, defined by 1 Gy ¼ 1 J kg1, which has largely replaced the older unit of
the ‘rad’ (1 Gy ¼ 100 rads). However, in practice, biological effects depend not
only on the total dose, but also on other factors, including the type of radiation, the
rate of deposition and whether the whole organ is uniformly radiated. These
considerations lead to the definitions used in medical applications of equivalent
and effective doses, where multiplicative weighting factors are included to take
account of different types of radiation and different organs being radiated. To
distinguish these latter doses from the simple absorbed dose, the sievert (Sv) unit is
used, also defined as 1 J kg1 because the weighting factors are dimensionless. For
example, the dose rate absorbed in tissue at a distance r from an external source of
activity A emitting -rays of energy E is given approximately by

dD AðMBqÞ E ðMeVÞ
ðSv h1 Þ
ð8:49aÞ
dt 6r2 ðm2 Þ

and for an internal source emitting radiation of energy ER, the effective dose rate
for an organ of mass M is

dD AER f
¼ ; ð8:49bÞ
dt M

where f is the fraction of the energy deposited in the organ.

To get some idea of scale, the total annual effective dose to the UK population is
approximately 2600 mSv, of which 85 per cent is due to naturally occurring
background radiation, although much higher doses can occur in specific cases –
for example, workers whose occupational activities expose them to radiation on a
daily basis, or people who live in areas rich in granite rocks (which emit radon, the
source of about half of the background radiation). The recommended limit for
additional whole-body exposure of the general population is 1 mSv y1 .13
The primary deposition of energy is due, as in non-living matter, to ionization
and excitation of atoms and molecules in the path of the radiation. This occurs on a
timescale of 1016 s or less and was described in Chapter 4. We can draw on that
discussion here, bearing in mind that living tissue consists mainly of light elements
and in particular has a high proportion (about 80 per cent) of water.
For heavy particles, such as protons and -particles, the most important process is
ionization via interactions with electrons and the energy losses are given by the
Bethe–Bloch formula Equation (4.11). The rate of energy loss by a heavy particle is

13
For a discussion of Equations (8.49) and quantitative issues of acceptable doses for various sections of the
population and to different organs, see for example, Chapter 7 of Li01 and Chapter 11 of De99.
280 CH8 APPLICATIONS OF NUCLEAR PHYSICS

high, peaking near the end of its range and so the penetrating power is low. For
example, a 1 MeV -particle travels only a few tens of microns and is easily stopped
by skin. However, considerable damage can be caused to sensitive internal organs if
an -emitting isotope is ingested. An exception to the above is neutron radiation,
which being electrically neutral does not produce primary ionization. Its primary
interaction is via the nuclear strong force and it will mainly scatter from protons
contained in the high percentage of water present. The scattered protons will,
however, produce ionization as discussed above. The overall effect is that neutrons
are more penetrating than other heavy particles and at MeV energies can deposit
their energy to a depth of several centimetres. Electrons also lose energy by
interaction with electrons, but the rate of energy loss is smaller than for heavy
particles. Also, as they have small mass, they are subject to greater scatter and so their
paths are not straight lines. In addition, electrons can in principle lose energy by
bremsstrahlung, but this is not significant in the low Z materials that make up the
patient. The overall result is that electrons are more penetrating than heavy particles
and deposit their energy over a greater volume. Finally, photons lose energy via a
variety of processes (see Section 4.4.4), the relative importance of which depends on
the photon energy. Photons are very penetrating and deposition of their energy is not
localized.
In addition to the physical damage that may be caused by the primary ionization
process, there is also the potential for chemical damage, as mentioned above. This
comes about because most of the primary interactions result in the ionization of
simple molecules and the creation of neutral atoms and molecules with an
unpaired electron. The latter are called free radicals (much discussed
in advertising material for health supplements). These reactions occur on much
longer timescales of about 106 s. For example, ionization of a water molecule
produces a free electron and a positively charged molecule:
H2 O ! H2 Oþ þ e ð8:50aÞ
radiation

and the released electron is very likely to be captured by another water molecule
producing a negative ion:
e þ H2 O ! H2 O : ð8:50bÞ

Both ions are unstable and dissociate to create free radicals (denoted by black
circles):
H2 Oþ ! Hþ þ OH ð8:51aÞ

and
H2 O ! H þ OH : ð8:51bÞ

Free radicals are chemically very active, because there is a strong tendency for
their electrons to pair with one in another free radical. Thus the free radicals in
 BIOMEDICAL APPLICATIONS 281

Equations (8.51) will interact with organic molecules (denoted generically by RH)
to produce organic free radicals:

RH þ OH ! R þ H2 O ð8:52aÞ

and

RH þ H ! R þ H2 : ð8:52bÞ

The latter may then induce chemical changes in critical biological structures (e.g.
chromosomes) some way from the site of the original radiation interaction that
produced them. Alternatively, the radiation may interact directly with the molecule
RH again releasing a free radical R :

RH ! RHþ þ e ; RHþ ! R þ H : ð8:53Þ

radiation

Finally, if the irradiated material is rich in oxygen, yet another set of reactions is
possible:

R þ O2 ! RO2 ; ð8:54aÞ

followed by

RO2 þ RH ! RO2 H þ R ; ð8:54bÞ

with the release of another free radical. This is the oxygen effect mentioned above
that complicates the treatment of tumours.
Fortunately, for low-level radiation, living matter itself has the ability to repair
much of the damage caused by radiation and so low-level radiation does not lead to
permanent consequences. Indeed, if this were not so, then life may not have
evolved in the way it has, because we are all exposed to low levels of naturally-
occurring radiation throughout our lives (which may well have been far greater in
the distant past) and the modern use of radiation for a wide range of industrial and
medical purposes has undoubtedly increased that exposure. However, the repair
mechanism is not effective for high levels of exposure.
In the context of radiation therapy an important quantity is the linear energy
transfer (LET) which measures the energy deposited per unit distance over the
path of the radiation. Except for bremsstrahlung, LET is the same as dE=dx
discussed in Chapter 4. High-LET particles are heavy ions and -particles, which
lose their energy rapidly and have short ranges. LET values of the order of
100 keV=mm and ranges 0.1–1.0 mm are typical. Low-LET particles are electrons
and photons with LET values of the order of 1 keV=mm and ranges of the order of
1 cm. Much cancer therapy work uses low-LET particles. Treatment consists of
directing a beam at a cancer site from several directions to reduce the exposure
282 CH8 APPLICATIONS OF NUCLEAR PHYSICS

of healthy tissue, while maintaining the total dose to the tumour. Other techniques
include giving the dose in several stages so that the outer regions of the tumour,
which are relatively oxygen-rich, are successively destroyed as they become re-
oxygenated. Other treatments, particularly for localized cancers, involve the
introduction of a radionuclide either physically via a needle or by ingestion/
injection of a compound containing the radionuclide. Chemicals that preferentially
target specific organs or bones are commonly used.
Neutron therapy, as an example of a high-LET particle, is not widely used
because of the problem of producing a strongly collimated beam plus the difficulty
of ensuring that the energy is deposited primarily at the tumour site. Neutrons also
share with low-LET radiation the drawback that their attenuation in matter is
exponential. On the other hand, the rate of energy loss of protons and other
charged particles increases with penetration depth, culminating at a maximum, the
Bragg peak, close to the end of their range. In principle, this means that a greater
fraction of the energy would be deposited at the tumour site and less damage
would be caused along the path length to the site. There is also an increasing
interest in using heavy charged particles. For example, carbon ions at the
beginning of their path in tissue have a low rate of energy loss more like an
LET particle, but near the end of their range the local ionization increases
dramatically as it approaches the Bragg peak. Thus considerable energy can be
deposited at a precise depth without the danger of massive destruction of healthy
tissue en route to the target. Another potential advantage is that nuclear interac-
tions along the path length will convert a small fraction of the nuclei to radioactive
positron-emitting isotopes which could then be used to image the irradiated region
(using the PET technique described below) and thus monitor the effectiveness of
the treatment programme. Unfortunately, the use of protons and heavy particles
requires access to an accelerator and for this reason proton and heavy ion therapy
is not commonly used.

8.3.2 Medical imaging using radiation

There are several techniques for producing images useful for diagnostic purposes
and in this section we will describe the principles of the main ones, but without
technical details.14

Imaging using projected images

The use of an external source of radiation for medical imaging is of long-standing

and well known. Basically, the system consists of a source placed some distance in

14
A readable account of medical imaging at the appropriate level is given in Chapter 7 of Li01 and a short
useful review of the whole field is He97.
 BIOMEDICAL APPLICATIONS 283

Figure 8.9 Basic layout for imaging using an external source

front the patient and a detector (usually a special type of sensitive film) placed
immediately behind the patient. Because the radiation is absorbed according to an
exponential law, a measurement of the intensities just before and after the patient
yields information on the integrated mean free path (or equivalently the attenuation
coefficient   1=) of the photons in the body.
Thus, referring to Figure 8.9, we have for the ray shown, using Equation (4.17),
xð2

lnðI1 =I2 Þ ¼ ðxÞ dx: ð8:55Þ

x1

The full image reveals variations of this integral only in two dimensions and thus
contains no depth information. A three-dimensional effect comes from overlapping
shadows in the two-dimensional images and part of the skill of a radiologist is to
interpret these effects.
The most commonly used radiation is X-rays. The attenuation coefficient is
dependent on the material and is greater for elements with high Z than for elements
with low Z. Thus X-rays are good for imaging bone (which contains calcium with
Z ¼ 20), but far less useful for imaging soft tissue (which contains a high
proportion of water).
Images can also be obtained using an internal source of radiation. This is done
by the patient ingesting, or being injected with, a substance containing a radio-
active -emitting isotope. As photon detectors are very sensitive, the concentration
of the radioisotope can be very low and any risk to the patient is further minimized
by choosing an isotope with a short lifetime. If necessary, the radioisotope can be
combined in a compound that is known to be concentrated preferentially in a
specific organ if that is to be investigated, for example iodine in the thyroid. In
practice, more than 90 per cent of routine investigations use the first excited state
of 99
43 Tc as the radioisotope. This has a lifetime of about 6 hours and is easily
produced from the -decay of 99 42 Mo which has a lifetime of 67 hours. The
284 CH8 APPLICATIONS OF NUCLEAR PHYSICS

Figure 8.10 Schematic diagram of a -camera

usefulness of this metastable state (written 99 Tcm ) is that it emits a single 140 keV
photon with negligible -decay modes, decaying to the very long-lived
(2 105 years) ground state.
Because the radiation is emitted in all directions, a different technique is used
to detect it. The patient is stationary and is scanned by a large-area detector consisting
of a collimated single-crystal scintillator, usually NaI, the output from which is
viewed by an array of photomultipliers (PMTs) via a light guide (see Section 4.4.2).
A schematic diagram of such a -camera is shown in Figure 8.10. The output from
the scintillator is received in several PMTs and the relative intensities of these signals
depend on the point of origin. The signals can be analysed to locate the point to
within a few millmetres. The collimator restricts the direction of photons that can
be detected and, combined with the information from the PMTs, the overall
spatial resolution is typically of the order of 10 mm, provided the region being
examined has an attenuation coefficient that differs by at least 10 per cent from its
surroundings.
Radioisotope investigations principally demonstrate function rather than anat-
omy, in contrast to X-ray investigations that show mainly anatomical features.
Thus better images of soft tissue, such as tumours, can be obtained than those
obtained using external X-rays, because the ability of the tumour to metabolize has
been exploited, but the exact location of the tumour with respect to the anatomy is
often lost or poorly defined.
Figure 8.11 shows part of a whole body skeletal image of a patient who had been
injected with a compound MDP which moves preferentially to sites of bone
cancer, labelled with the isotope 99 Tcm . The image clearly shows selective take-up
of the isotope in many tumours distributed throughout the body. (The concentra-
tion in the bladder is probably not significant.)
 BIOMEDICAL APPLICATIONS 285

Figure 8.11 Part of a whole-body skeletal image obtained using 99 Tcm MDP (image courtesy of
Prof. R. J. Ott, Royal Marsden Hospital, London, UK)
286 CH8 APPLICATIONS OF NUCLEAR PHYSICS

Computed tomography

A radiographic image is a two-dimensional display of a three-dimensional

structure and although the overlapping images give a useful three-dimensional
effect, details are always partially obscured by the superposition of information
from underlying and overlying planes. The result is loss of contrast. Thus while
images from the projection methods have good spatial resolution they have poor
resolution in depth. A major advance which addresses this problem was made in
1971 with the introduction of a new scanning technique called computed
tomography (CT).15 This enables a series of two-dimensional sections to be
imaged as small as a millimetre across, even when the attenuation coefficient
differs by less than 1 per cent from its surroundings.
The principle behind the CT technique is the observation that all the information
needed to construct an image of a two-dimensional section of tissue is contained in
the one-dimensional projections that cover all possible directions within the plane
of the slice. Thus for example, if the slice is in the x–y plane, a projected image of
the slice contains information on ðx,yÞ in the form of a set of line integrals of 
taken through the region in a particular direction. As the angle in the plane of the
slice is varied, a different representation of ðx, yÞ is obtained in the form of a
different set of line integrals. Once a complete set of line integrals has been
obtained there are mathematical methods (including some that have been devel-
oped by particle physicists to reconstruct events from high-energy collisions) that
allow the required two-dimensional function to be reconstructed. Modern high-
speed computers are able to perform this construction very rapidly, so that images
can now be obtained in real time and motion as fast as heartbeats can be captured.
Computed tomography may be used in conjunction with both external and
internal radiation. As an example, the arrangement for a CT X-ray scan is shown
schematically in Figure 8.12. In this example (known as a fourth-generation
machine), the patient remains stationary within a ring of several hundred detectors
(solid-state scintillators are frequently used). Within this ring there is an X-ray
source that moves on another ring and provides a fan of X-rays. Each alignment of
the source and a detector in the ring defines a line through the patient and the
recorded count rate enables a line integral to be computed from Equation (8.55).
By moving the source through its full angular range, a complete set of such line
integrals is generated, enabling a two-dimensional section to be computed through
the patient. This type of scanner is relatively expensive in both capital and
maintenance costs and another type (known as a third-generation machine) is
more common. This differs from Figure 8.12 in having a single bank of detectors
opposite the source and both source and detectors are rotated to cover the full
angular range. Although the CT method can produce scans of soft tissue better
than conventional X-ray projections, the images are achieved at the expense of the

15
The CT system was devised independently by Godfrey Hounsfield and Allan Cormack who were jointly
awarded the 1979 Nobel Prize in Physiology and Medicine for their work.
 BIOMEDICAL APPLICATIONS 287

Figure 8.12 Schematic diagram of the arrangement for a CT X-ray scanner

patient receiving a higher dose of potentially harmful radiation. An example of a

CT X-ray scan is shown in Figure 8.13(a).
CT can also be used to construct images obtained from projections from internal
radiation using radioisotopes that emit a single -ray. This technique is called
single-photon emission computed tomography (SPECT). The arrangement is in

Figure 8.13 (a) X-ray CT scan of the brain, and (b) SPECT brain scan using a 99 Tcm labelled
blood flow tracer, showing high perfusion in the tumour (indicated by arrows) (image courtesy of
Prof. R. J. Ott, Royal Marsden Hospital, London, UK)
288 CH8 APPLICATIONS OF NUCLEAR PHYSICS

some sense the ‘inverse’ of that in Figure 8.12. Thus the source is now within the
patient and the fixed ring of detectors is replaced by one or more -cameras
designed so that they can rotate in a circle about the patient. An example of an
image obtained using SPECT is shown in Figure 8.13(b).
For a number of technical reasons, including the fact that the emitted radiation is
isotropic, there are more stringent requirements on the -cameras and SPECT
images have a resolution of only about 10 mm. However, although not suitable for
accurate quantitative measurement of anatomy, they are of great use for clinical
diagnostic work involving function. For example, the technique is used to make
quantitative measurements of the functioning of an organ, i.e. clearance rates in
kidneys, lung volumes, etc..
Since radionuclide imaging provides functional and physiological information,
it would be highly desirable to be able to image the concentrations of elements
such as carbon, oxygen and nitrogen that are present in high abundances in
the body. The only radioisotopes of these elements that are suitable for imaging
are short-lived positron emitters: 11 C (half-life 20 min), 13 N (10 min) and 15 O
( 2 min). For these emitters, the radiation detected is the two -rays emitted when
the positron annihilates with an electron. This occurs within a few millimetres
from the point of production of the positron, whose initial energy is typically less
than 0.5 MeV. The photons each have energies equal to the rest mass of an
electron, i.e. 0.511 MeV and emerge ‘back-to-back’ to conserve momentum. This
technique is called positron emission tomography (PET) and was mentioned
earlier in connection with radiation treatment using heavy ions.
The arrangement of a PET scanner is shown in Figure 8.14. If the detectors D1
and D2 detect photons of the correct energy in coincidence, then the count rate is a
measure of the integral of the source activity within the patient along the line AB
passing through P. The ring of detectors defines a plane through the patient and the

Figure 8.14 Schematic diagram of the arrangement of a PET scanner

 BIOMEDICAL APPLICATIONS 289

Figure 8.15 Part of a whole-body PET scan showing uptake of the chemical FDG (labelled by
99 m
Tc ) in lung cancer (image courtesy of Prof. R. J. Ott, Royal Marsden Hospital, London, UK)

complete set of data from all combinations of detector pairs contains all
the information needed to generate the set of line integrals which can be
converted into a two-dimensional image of the source using standard CT image
reconstruction techniques. An example of an image using the PET technique is
shown in Figure 8.15.
This account of medical imaging has ignored many technical points. For
example, there are a number of corrections that have to be made to the raw
data, particularly in the SPECT technique, and the most useful radioisotopes used
in PET are produced in a cyclotron, so the scanner has to be near such a facility,
which considerably limits is use. The interested reader is referred to specialized
texts for further details.16

16
For a more detailed discussion see, for example, De99.
290 CH8 APPLICATIONS OF NUCLEAR PHYSICS

8.3.3 Magnetic resonance imaging

We conclude this brief description of imaging with an account of a remarkable

technique which in a relatively short time has become one of the most sophisti-
cated tools for clinical diagnostic work and medical research. It is not only capable
of producing images of unprecedented clarity, but it does so in a way that is
intrinsically safe and without using potentially harmful ionising radiation.
Magnetic resonance imaging (MRI) is based on the phenomenon of nuclear
magnetic resonance that was discovered independently by Bloch and Purcell and
used by them to study the structure and diffusion properties of molecules.17 It is
based on the fact that the quantum spin states of nuclei (strictly their associated
magnetic moments) can be manipulated by magnetic fields. A brief overview of
the method is as follows. First, nuclear spins in tissue are aligned by a powerful
static magnetic field, typically in the range 0.2–3 T, usually supplied by a super-
conducting magnet. As living tissue is predominantly water, the spins in question
are mainly those of protons. (Oxygen is an even–even nucleus and so plays no
role.) Secondly, oscillating magnetic field pulses at radio frequency are applied in
a plane perpendicular to the magnetic field lines of the static field, which causes
some of the protons to change from their aligned positions. After each pulse, the
nuclei relax back to their original configuration and in so doing they generate
signals that can be detected by coils wrapped around the patient. Differences in the
relaxation rates and associated signals are the basis of contrast in MRI images. For
example, water molecules in blood have different relaxation rates from water
molecules in other tissues.
There are many different types of MRI scan, each with their own specialized
procedures and the full mathematical analysis of these is complex. We will
therefore give only a rather general account concentrating on the basic physics.
The interested reader is referred to more detailed texts at an appropriate level.18
The proton has spin-12 and magnetic moment lP . In the absence of an external
magnetic field, the two states corresponding to the two values of the magnetic
quantum number ms ¼  12 are equally populated and the net magnetization M (i.e.
the average magnetic moment per unit volume) is zero. In the presence of a static
magnetic field B, taken to be in the z-direction, there is an interaction energy
ðlP BÞ and the two states have different energies with different probabilities
given by the Boltzmann distribution. The energy difference between the states is
E ¼ 2P B ¼ h f , where f is the Larmor (or nuclear resonance) frequency,19

17
Felix Bloch and Edward Purcell were awarded the 1952 Nobel Prize in Physics for their discovery of
nuclear magnetic resonance (NMR) and subsequent researches. Although the term NMR is still used in
research environments, the term magnetic resonance imaging (MRI) is preferred in clinical environments to
prevent patients associating the technique with ‘harmful nuclear radiation’.
18
See, for example, De99, McR03 and Ho97a.
19
In general, the nuclear resonance frequency is defined by f ¼ jljB=j h, where j is the spin of the particle
involved and  is its magnetic dipole moment.
 BIOMEDICAL APPLICATIONS 291

Figure 8.16 (a) Precession of the magnetization M in the xy-plane under the action of a torque
M B resulting from an external field B; (b) motion viewed in a frame of reference ðx 0 ; y0 ; z0 Þ
rotating at the Larmor frequency about the z-axis  the r.f. pulse Brf applied in the x 0 -direction
has rotated M so that it points in the y0 -direction

which is the frequency of a photon that would correspond to a transition between

the two nuclear spin states. The energy difference is small. For example, for a field
of 1 T, E
1:8 107 eV and f is about 43 MHz, i.e. in the radio region of the
electromagnetic spectrum. Although there is a net magnetization in the z-direction,
the resultant magnetization M0 is too small to be measured.
The situation changes, however, if M no longer points along the z-axis and a
signal is generated if the magnetization has a component in the plane orthogonal to
B. This is illustrated in Figure 8.16(a). In this figure, M has been rotated to lie in
the xy-plane and since there is an angular momentum associated with the
magnetization, M will precess about B under the action of the torque M B.
The rotation can be achieved by applying an alternating r.f. magnetic field Brf
to the sample at right angles to B and at the Larmor frequency. As M precesses
about B, one component of Brf rotates in phase with it. The resulting motion is
complicated and is best viewed in a frame of reference rotating at the Larmor
frequency about the z-axis, which we label by ðx0 ; y0 ; z0 Þ with z0 parallel to z. This is
shown in Figure 8.16(b). The full mathematical analysis is given, for example, in
the book by Hobbie (see Footnote 18) and we will just quote the result. This is that
the magnetization vector can be rotated through an arbitrary angle depending on the
strength and duration of the r.f. pulse. In particular, it is possible to rotate it
through 90 so that the magnetization vector precesses about the the x0 -axis, i.e.
rotating with a frequency that depends on the magnitude of the r.f. field. As the r.f.
pulse forces all the protons to precess exactly in phase, there will be a component
of magnetization along the y-axis in the rotating frame. When the r.f. pulse is
turned off, the system returns to equilbrium with M aligned along the z-axis by re-
emitting the energy absorbed from the r.f. pulse. As it does so, the external field
due to M will vary with time with the same frequency and can be detected as an
induced emf in a coil surrounding the patient. This is the basic MRI signal.
292 CH8 APPLICATIONS OF NUCLEAR PHYSICS

Crucially, the frequency of the external r.f. field must match the Larmor frequency
of the protons to be excited.
The induced signal will decay as equilibrium is restored. If B were uniform
throughout the selected region, all the protons would precess at the same frequency
and remain in phase. In that case the interaction of the proton spins with the
surrounding lattice, the so-called spin-lattice interactions, would cause M to relax
to its equilibrium state M0 parallel to B. Under reasonable assumptions the
radiated signal is proportional to the difference ðM0  MÞ and decreases expo-
nentially with a characteristic spin-lattice, or longitudinal, relaxation time T1 .
Typical spin-lattice relaxation times are of the order of a few hundred milliseconds
and are significantly different for different materials, such as muscle, fat and water.
However, because there are always small irregularities in the field due to local
atomic and nuclear effects, individual protons actually precess at slightly different
rates and the signal decays because the component of M orthogonal to B (i.e. in the
xy-plane) decreases as the individual moments loose phase coherence. This
decrease is characterized by a second time T2 , called the spin–spin, or transverse,
relaxation time. This is normally much shorter than T1 , but again varies with
material. Both relaxation times can be measured.
The above assumes that the external field B is perfectly uniform, but of course
the ideal is not realized in practice. The effects of macroscopic inhomogeneity in
the magnetic field can be eliminated by generating so-called spin echoes, which
may crudely be described as making two ‘orthogonal’ measurements such that the
unwanted effects cancel out exactly in the sum. Many MRI imaging sequences use
this technical device and again we refer the interested reader to the literature cited
in Footnote 18 for further details.
All the above assumes we are scanning the whole body. The original develop-
ment of the method as a medical diagnostic technique is due to the realization that
gradients in the static magnetic field could be used to encode the signal with
precise spatial information and be processed to generate two-dimensional images
corresponding to slices through the tissue of the organ being examined.20 The
patient is placed in the fixed field B pointing along the z-direction. A second static
field Bg parallel to z, but with a gradient in the z-direction is then applied so that
the total static field is a function of z. This means that the Larmor frequency (which
is proportional to the magnetic field) will vary as a function of z. Thus when the r.f.
field Brf is applied with a narrow band of frequencies about fr:f: , the only protons to
be resonantly excited will be those within a narrow slice of thickness dz at the
particular value of z corresponding to the narrow band of frequencies. The field
Brf is applied until the magnetization in the slice has been rotated through either
90 or 180 depending on what measurements are to be taken. Both Brf and Bg are
then turned off.

20
This discovery was first made by Paul Lauterbur and an analysis of the effect was first made by Sir Peter
Mansfield. They shared the 2003 Nobel Prize in Physiology and Medicine for their work is establishing MRI
as a medical diagnostic technique.
 BIOMEDICAL APPLICATIONS 293

The final step is to obtain a spatial image of the magnetization as a function of x

and y. This entails encoding the MRI signal with information linking it to a point
of origin in real space. There are many ways this can be done (one utilizes the CT
method encountered earlier) and again we refer the interested reader to the
specialized texts quoted earlier for the details. The outcome is that M and the
two relaxation times can both be measured. All three quantities vary spatially
within the body and can give valuable biomedical information. For example,
relaxation times are usually different for tumour tissue compared with normal
tissue.
In some areas MRI scans have considerable advantages over other forms of
imaging. For example, the contrast of soft tissue is much better than CT scans,
leading to very high quality images, especially of the brain. Examples of such
images are shown in Figure 8.17.
As ionizing radiation is not used, MRI is intrinsically safe at the field intensities
used. The only exception to this is that because of the presence of high magnetic
fields, care must be taken to keep all ferromagnetic objects away from the scanner.
This means that patients with heart pacemakers, or metal implants cannot in
general be scanned and care has to be taken to screen out people who have had an
occupational exposure to microscopic fragments of steel (such as welders) as these
may well have lodged in critical organs such as the eyes and the latter could be
seriously damaged if the fragments moved rapidly under the action of the very
strong magnetic field.

Figure 8.17 Two MRI scans of a brain  (a) T1-weighted, and (b) T2-weighted  showing a
frontal lobe tumour (images courtesy of the MRI Unit of the Royal Marsden NHS Foundation
Trust, London, UK)
294 CH8 APPLICATIONS OF NUCLEAR PHYSICS

Problems
8.1 The fission of 235 U is induced by a neutron and the fission fragments are 92 37 Rb and
140
55 Cs. Use the SEMF to calculate the energy released (in MeV) per fission. Ignore
the (negligible) contributions from the pairing term. The reaction is used to power a
100 MW nuclear reactor whose core is a sphere of radius 100 cm. If an average of
one neutron per fission escapes the core, what is the neutron flux at the outer surface
of the reactor in m2 s1 ? The core is surrounded by 1:3 m3 of ideal gas maintained
at a pressure of 1 105 Pa and a temperature of 298 K. All neutrons escaping the
reactor core pass through the gas. If the interaction cross-section between the
neutrons and the gas is 1 mb, calculate the rate of neutron interactions in the gas.

8.2 A neutron with non-relativistic laboratory speed v collides elastically with a nucleus
of mass M. If the scattering is isotropic, show that the average kinetic energy of the
neutron after the collision is

M 2 þ m2
Efinal ¼ Einitial ;
ðM þ mÞ2

where m  mn . Use this result to estimate the number of collisions necessary to

thermalize neutrons from the fission of 235 U using a graphite moderator (assume this
is pure 12 C).

8.3 A thermal fission reactor uses natural uranium. The energy released from fission is
200 MeV per atom of 235 U and the total power output is 500 MW. If all neutrons
captured by 238 U lead to the production of 239 Pu, calculate the rate of production of
plutonium in kg/year. The cross-sections at the relevant neutron energy are
capture ¼ 3 b and fission ¼ 600 b; and the relative abundance of 238 U to 235 U in
natural uranium is 138:1.

8.4 In a particular thermal reactor, each fission releases 200 MeV of energy with an
instantaneous power output 3 t1:2 , where t is measured in seconds. After burning
with a steady power output P0 ¼ 2 GW for a time T, the reactor is shut-down. Show
that the mean thermal power P from a fuel rod of the reactor after time t ð> 1 sÞ is
approximately
h i
PðtÞ ¼ 0:075P0 t0:2  ðT þ tÞ0:2

and, taking the mean age of the fuel rods to be 1 year, calculate the power output
after 6 months.

8.5 If the Sun were formed 4.6 billion years ago and initially consisted of 9 1056
hydrogen atoms and since then has been radiating energy via the PPI chain at a
detectable rate of 3:86 1026 W, how much longer will it be before the Sun’s supply
of hydrogen is exhausted (assuming that the nature of the Sun does not change)?
 PROBLEMS 295

8.6 In the PPI cycle, helium nuclei are produced by the fusion of hydrogen nuclei and
6.55 MeV of electromagnetic energy is produced for every proton consumed. If the
electromagnetic radiation energy at the surface of the Earth is 8:4 J cm2 s1 and is
due predominantly to the PPI cycle, what is the expected flux of solar neutrinos at
the Earth in cm2 s1 ?

8.7 In a plasma of equal numbers of deuterium and tritium atoms (in practice, deuteron
and triton nuclei) at an energy kT ¼ 10 keV, the Lawson criterion is just satisfied for
a total of 5 s. Estimate the number density of deuterons.

8.8 A thermal power station operates using inertial confinement fusion. If the ‘fuel’
consists of 1 mg pellets of frozen deuterium–tritium mixture, how many would have
to be supplied per second to provide an output of 750 MW if the efficiency for
converting the material is 25 per cent?

8.9 In some extensions of the standard model (to be discussed in Chapter 9) the proton is
unstable and can decay, e.g. via p ! 0 þ eþ . If all the energy in such decays is
deposited in the body and assuming that an absorbed dose of 5 Gy per year is lethal
for humans, what limit does the existence of life place on the proton lifetime?

8.10 The main decay mode of 60 27 Co is the emission of two photons, one with energy
1.173 MeV and the other with 1.333 MeV. In an experiment, an operator stands 1 m
away from an open source of 40 KBq of 6027 Co for a total period of 18 h. Estimate the
approximate whole-body radiation dose received.

8.11 A bone of thickness b cm is surrounded by tissue with a uniform thickness of t cm. It

is irradiated with 140 keV -rays. The intensities through the bone (Ib ) and through
the tissue only (It ) are measured and their ratio R  Ib =It is found to be 0.7. If the
attenuation coefficients of bone and tissue at this energy are b ¼ 0:29 cm1 and
t ¼ 0:15 cm1 , calculate the thickness of the bone.

8.12 The flux of relativistic cosmic ray muons at the surface of the Earth is approximately
250 m2 s1 . Use Figure 4.8 to make a rough estimate of their rate of ionization
energy loss as they traverse living matter. Hence estimate in grays the annual human
body dose of radiation due to cosmic ray muons.

55
8.13 Calculate the nuclear magnetic resonance frequency for the nucleus 25 Mn in a field
of 2 T if its magnetic dipole moment is 3.46 N .
9
Outstanding Questions and Future
Prospects

In this chapter we shall describe a few of the outstanding questions in both nuclear
and particle physics and future prospects for their solution. The list is by no means
exhaustive (particularly for nuclear physics, which has a very wide range of
applications) and concentrates mainly on those areas touched on in earlier
chapters. The examples should be sufficient to show that nuclear and particle
physics remain exciting and vibrant subjects with many interesting phenomena
being discovered and questions awaiting explanations.

9.1 Particle Physics

Unlike nuclear physics, particle physics does have a comprehensive theory, but
although the standard model is very successful at explaining a wide range of
phenomena, there are still questions that remain to be answered and some hints
from experiments of phenomena that lie outside the model, for example neutrino
oscillations and the possibility of lepton number violation. In addition, the success
of the standard model has spurred physicists to construct theories that incorporate
the strong interaction, and even in some cases gravity, in wider unification
schemes. A full discussion of these topics is beyond the scope of this book, but
in this chapter we will briefly review some of these questions and also look at the
rapidly growing field of particle astrophysics.

9.1.1 The Higgs boson

The Higgs boson is an electrically neutral spin-0 boson whose existence is

predicted by the unified electroweak theory, but which has not yet been observed.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
298 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

It is required because of a fundamental symmetry associated with theories in

which the force carriers are spin-1 bosons. This symmetry is called gauge
invariance and has been mentioned in previous chapters. Gauge invariance can
be shown to require that the spin-1 ‘gauge bosons’ have zero masses if they are the
only bosons in the theory. This is acceptable for QED and QCD, since the gauge
bosons are photons and gluons and they do indeed have zero masses. Gauge
invariance also plays a crucial role in the unified electroweak theory, where it is
needed to ensure the cancellation of the divergences that occur in individual
higher-order Feynman diagrams. In this case the result is even stronger and it can
be shown that gauge invariance requires that the all the fundamental particles –
quarks, leptons and gauge bosons – have zero masses if gauge bosons are the only
bosons in the theory. This prediction is clearly in contradiction with experiment,
because the W and Z bosons have masses about 80–90 times that of the nucleon.
This problem, known as the origin of mass, is overcome by assuming that the
various particles interact with a new field, called the Higgs field, whose existence
can be shown to allow the gauge bosons to acquire masses without violating the
gauge invariance of the interaction.1 The ‘price’ of this is that there must exist
electrically neutral quanta associated with the Higgs field, called Higgs bosons, in
the same way that there are quanta associated with the electromagnetic field, i.e.
photons.
We saw in Chapter 3, that there is evidence that neutrinos, originally assumed to
have zero masses in the standard model, are in fact not massless. The Higgs
mechanism can also, in principle, be invoked to generate masses for neutrinos.
However, it would be natural to expect that such masses would then be roughly the
same size as the masses generated for the gauge bosons and we have seen that this is
clearly not the case. This problem can only be avoided if the coupling of the
neutrinos to the Higgs field is at least 12 orders of magnitude smaller than that of the
coupling of the top quark. Many physicists reject such an explanation as implausible
and alternative mechanisms have been suggested for generating very small neutrino
masses. All have problems of their own and at present none is universally accepted.
Experiments currently being planned should help resolve the matter.
The existence of the Higgs boson is the most important prediction of the
standard model that has not been verified by experiment, and searches for it are of
the highest priority. A problem in designing suitable experiments is that its mass is
not predicted by the theory. However, its couplings to other particles are predicted
and are essentially proportional to the masses of the particles to which it couples.
The Higgs boson therefore couples very weakly to light particles like neutrinos,
electrons, muons and u, d, s quarks; and much more strongly to heavy particles like
W
and Z 0 bosons, and presumably b and t quarks. Hence attempts to produce
Higgs bosons are made more difficult by the need to first produce the very heavy
particles to which they couple.

1
This process is called ‘spontaneous symmetry breaking’ and was mentioned in Chapter 6.
 PARTICLE PHYSICS 299

The failure to observe Higgs bosons in present experiments leads to limits on

their mass. The best results come from the Large Electron–Positron (LEP)
accelerator at CERN. This machine (which is no longer operational) had a
maximum energy of 208 GeV, which is enough to produce Higgs bosons with
masses up to almost 120 GeV/c2 in the reaction

eþ þ e ! H 0 þ Z 0 ; ð9:1Þ

which is expected to occur by the dominant mechanism of Figure 9.1.

Figure 9.1 Dominant mechanism for Higgs boson production in eþ e annihilation

Attempts were made to detect Higgs bosons by their decays to bb
pairs, where
the quarks would be observed as jets containing short-lived hadrons with non-zero
beauty. The results were tantalizing. By the time LEP closed down in November
2000 to make way for another project, it had shown that no Higgs bosons existed
with a mass less than 113.5 GeV/c2; and some evidence had been obtained for the
existence of a Higgs boson with a mass of 115 GeV/c2. This is very close to the
upper limit of masses that were accessible by LEP, but because the Higgs boson
would have a width, its mass distribution would extend down to lower energies and
would give a signal. Unfortunately, while this signal was statistically likely to be a
genuine result rather than a statistical fluctuation, the latter cannot be completely
ruled out.
Future investigations will involve the use of new accelerators currently under
construction, particularly the LHC proton–proton collider mentioned in Chapter 4.
(One of the detectors at the LHC, ATLAS was shown in Figure 4.19.) This will
enable searches to be made for Higgs bosons with masses up to 1 TeV/c2 via
reactions of the type

p þ p ! H 0 þ X; ð9:2Þ

where X is any state allowed by the conservation laws. The mechanism for this
reaction is the weak interaction between the constituent quarks of the protons, an
example of which is shown in Figure 9.2, where the other quarks in the protons are
spectators, as usual.
300 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

Figure 9.2 An example of a process that can produce Higgs bosons in pp collisions

The reaction in Equation (9.2) will take place against a large background of
strong interaction processes and the method of detecting it will depend on the
actual Higgs boson mass. If MH > 2MW , then the Higgs boson can decay to a pair
of W-mesons or Z 0 -mesons, which themselves decay. For example, from the
leptonic decay of the Z 0 s, we could have overall the reaction
H 0 ! ‘þ þ ‘ þ ‘þ þ ‘ ; ð‘ ¼ e; Þ: ð9:3Þ
This would enable the mass range 200 GeV=c2  MH  500 GeV=c2 to be explored.
However, the branching ratios are such that only a few per cent of decays will have
such a distinctive signal and other decays modes will also have to be explored. For
lower masses such that MH < 2MW where these decays are energetically forbidden,
one might think of looking for decays to fermion–antifermion pairs. Because the
Higgs boson preferentially couples to heavy particles, the dominant decay of this
type will be H 0 ! b þ
b with accompanying jets. This was the method used in the
LEP experiments referred to above. Unfortunately, it is very difficult to distinguish
these jets from those produced by other means. Rarer decay modes, but with more
distinctive signals, will have to be sought, such as H 0 !  þ , which in the
standard model has a branching ratio of about 103 .
All the above is based on the standard model with a single neutral Higgs boson,
but we will see in Section 9.1.3 that realistic extensions of the standard model
require several Higgs bosons, not all of which are electrically neutral. Experi-
mental investigations of the Higgs sector will undoubtedly play a central role in the
future of particle physics for many years to come.

9.1.2 Grand unification

Whether or not the Higgs boson exists is the most pressing unanswered question of
the standard model but, even if it is found with its predicted properties, this is not
 PARTICLE PHYSICS 301

the end of the story, because one of the goals of particle theory is to have a single
universal theory that explains all the phenomena of the subject. Since we already
have a unified theory for the weak and electromagnetic interactions, the next
logical step is to try to include the strong interaction. Attempts to do this are called
grand unified theories (GUTs).
We have seen that unification of the weak and electromagnetic interactions does
not manifest itself until energies of the order of the W and Z masses. To get some
idea of the energy scale of a grand unified theory, we show in Figure 9.3 the
couplings2 pffiffiffi pffiffiffi
g 2 2gW ; g0 2 2gZ ð9:4Þ

Figure 9.3 Idealized behaviour of the strong and electroweak coupling as functions of the
squared energy--momentum transfer Q2 in a simple grand unified theory; gU is the unification
coupling

and the strong coupling gs (this is related to s by s ¼ g2s =4

Þ as functions of Q2 ,
the squared energy–momentum transfer in a typical GUT. A naı̈ve extrapolation in
Q2 (using, for example, Equation (5.11)) from the region where these couplings
are presently known suggests that they become approximately equal to a single
value gU at the enormous energy Q2 ¼ MX2 c4, where MX , the so-called unification
mass, is of the order of 1015 GeV/c2. In practice, which couplings to extrapolate
depends on which version of GUT one considers, but if the extrapolation is done
accurately the three curves actually fail to meet at a point by an amount that cannot
be explained by uncertainties in the models.

2
Recall that the electromagnetic coupling e is related to these couplings by the unification condition
Equation (6.71).
302 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

There are many potential grand unified theories, but the simplest incorporates
the known quarks and leptons into common families. For example, one way is to
put the three coloured d-quarks and the doublet ðeþ ; e Þ (strictly their right-handed
components) into a common family, i.e.

ðdr ; db ; dg ; eþ ; e Þ: ð9:5Þ

The fundamental vertex interactions allowed in this model are shown in Figure 9.4.

Figure 9.4 Fundamental vertices that can occur for the multiplet of particles in Equation (9.5)

In addition to the known QCD interaction in (a) and the electroweak interaction
in (b), there are two new interactions represented by (c) and (d) involving the
emission or absorption of two new gauge bosons X and Y with electric charges  43
and  13, respectively, and masses of the order of MX . In this theory all the
processes of Figure 9.4 are characterized by a single GUT coupling given by
g2U 1
U   ; ð9:6Þ
4 42
which is found by extrapolating the known coupling of the standard model to the
energy MX c2.
This simple model has a number of attractive features. For example, it can be
shown that the sum of the electric charges of all the particles in a given multiplet is
zero. So, using the multiplet ðdr ; db ; dg ; eþ ; e Þ, it follows that

3qd þ e ¼ 0; ð9:7Þ
 PARTICLE PHYSICS 303

where qd is the charge of the down quark. Thus qd ¼ e=3 and the fractional
charges of the quarks is seen to originate in the fact that they exist in three colour
states. By a similar argument, the up quark has charge qu ¼ 2e=3 and so with the
usual quark assignment p ¼ uud, the proton charge is given by

qp ¼ 2qu þ qd ¼ e: ð9:8Þ

Thus, we also have an explanation of the long-standing puzzle of why the proton
and positron have precisely the same electric charge.
GUTs make a number of predictions that can be tested at presently accessible
energies. For example, if the three curves of Figure 9.3 really did meet at a point, then
the three low-energy couplings of the standard model would be expressible in terms
of the two parameters U and MX . This could be used to predict one of the former, or
equivalently the weak mixing angle W . The result is sin2 W ¼ 0:214
0:004,
which is close to the measured value of 0:2313
0:0003, although not strictly
compatible with it. (This is true even if the effect of the Higgs boson is taken into
account when evaluating the evolution of the coupling constants.)

Figure 9.5 The three fundamental vertices predicted by the simplest GUT involving the gauge
bosons X and Y (these are in addition to those shown in Figure 9.4)

In addition to the interactions of the X and Y bosons shown in Figure 9.4, there
are a number of other possible vertices, which are shown in Figure 9.5. (There is
also another set where particles are changed to antiparticles.) A consequence of
these interactions and those of Figure 9.4(c) and (d) is the possibility of reactions
that conserve neither baryon nor lepton numbers. The most striking prediction of
this type is that the proton would be unstable, with decay modes such as
p !
0 þ eþ and p !
þ þ 
e . Examples of Feynman diagrams for these decays
are shown in Figure 9.6 and are constructed by combining the vertices of Figure 9.4
and 9.5. In all such processes, although lepton number L and baryon number B are
not conserved, the combination
X
R B L‘ ð‘ ¼ e; ; Þ ð9:9Þ
‘

is conserved.
304 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

Figure 9.6 Examples of processes that contribute to the proton decay mode p !
0 þ eþ

Since the masses of the X and Y bosons are far larger than the quarks and
leptons, we can use the zero-range approximation to estimate the lifetime of proton
decay. In this approximation, and by analogy with the lifetime for the muon
Equation (7.62), we have for the proton lifetime

ðMX c2 Þ4
p  : ð9:10Þ
g4U ðMp c2 Þ5

Taking account of reasonable uncertainties on gU and MX , this gives

p  1030
1 years: ð9:11Þ

Proton decay via these modes has been looked for experimentally. The most
extensive search has been made using the Kamiokande detector described in
Chapter 4. To date no events have been observed and this enables a lower limit to
be put on the proton lifetime of about 1032 years, which rules out the simplest
version of a grand unified theory. However, there are other, more complicated,
versions that still cannot be completely ruled out by present experiments. Some of
these incorporate the idea of supersymmetry which is described below.
Finally, GUTs may offer an explanation for the very small neutrino masses
observed in the oscillation experiments discussed in Chapter 3. In Section 6.3 we
discussed the possibility that the neutrino was its own antiparticle (a so-called
Majorana neutrino). In GUTs the right-handed neutrino states are predicted to be
very massive (of order 1017 GeV/c2) and mix with the massless left-handed
neutrinos of the standard model to give physical neutrinos with masses
m m2L =MX ; ð9:12Þ
where mL is the typical mass of a charged lepton or quark.

9.1.3 Supersymmetry

One of the problems with GUTs is that if there are new particles associated with
the unification energy scale, then they would have to be included as additional
 PARTICLE PHYSICS 305

contributions in the higher-order calculations in the electroweak theory, for

example for the mass of the W-boson. These contributions would upset the delicate
cancellations that ensure finite results from higher-order diagrams in the standard
model, unless there were some way of cancelling these new contributions.
Supersymmetry (SUSY) does exactly this.
Supersymmetry is the proposal that every known elementary particle has a
partner, called a superpartner, which is identical to it all respects except its spin.
Spin-12 particles have spin-0 superpartners and spin-1 particles have spin-12 super-
partners. To distinguish between a spin-12 particle and its superpartner, an ‘s’ is
attached to the front of its name in the latter case. Thus, for example, a spin-12
electron has a spin-0 selectron as its superpartner. The full set of elementary
particles and their superpartners in the simplest SUSY model (the so-called
Minimal Supersymmetric Standard Model – MSSM) is shown in Table 9.1.
(This is actually a simplification because even the simplest SUSY requires a
number of different Higgs bosons, not all electrically neutral.)

Table 9.1 The particles of the MSSM and their superpartners

Particle Symbol Spin Superparticle Symbol Spin
1
Quark q 2 Squark ~
q 0
1
Electron e 2 Selectron ~e 0
1
Muon  2 Smuon ~ 0
1
Tauon 2 Stauon ~ 0
W-boson W 1 Wino We 1
2
Z-boson Z 1 Zino Ze 1
2
1
Photon  1 Photino ~ 2
1
Gluon g 1 Gluino ~
g 2
Higgs boson H 0 Higgsino He 1
2

If the symmetry were exact then a particle and its superparticle would have equal
masses. This is clearly not the case or such states would have already been found. So
supersymmetry is at best an approximate symmetry of nature. Nevertheless, even in
an approximate symmetry, the couplings of the two states are equal and opposite,
thereby ensuring the required cancellation, providing their masses are not too large.
In practice, it is usually assumed in GUTs that incorporate supersymmetry that the
masses of the superparticles are of the same order as the masses of the W and Z
bosons. With the inclusion of superparticles, the evolution of the coupling constants
of the standard model as functions of Q2 changes slightly and when extrapolated
they meet very close to a single point. The unification mass is increased somewhat to
about 1016 GeV/c2, while the value of gU remains roughly constant. Thus the
predicted lifetime of the proton is increased to about 1032 – 1033 years, conveniently
beyond the ‘reach’ of current experiments. At the same time, the value of the weak
mixing angle is brought into almost exact agreement with the measured value.
Whether this is simply a coincidence or not is unclear.
306 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

To verify supersymmetry it will of course be essential to detect the superparticles

and that will not be easy. For example, the virtual exchange of superparticles
could contribute to the deviation of the muon magnetic dipole moment from its
Dirac value, although it would be difficult to separate these contributions from
other corrections. To date, activity has concentrated on the direct detection of
superparticles in reactions. In the simplest version of a SUSY theory, super-
particles are produced in pairs (like leptons or strange particles in strong
interactions, i.e. associated production) so that the decay of a superparticle must
have at least one superparticle in the final state and the lightest such particle will
necessarily be stable. Most versions of SUSY theories assume that the lightest
particle will be a neutralino  e0 , which is the name given to a mixture of the
photino, the higgsino and the zino, the three spin-12 superparticles that interact
purely by the electroweak interaction. If this is the case, a possible reaction that
could be studied is

eþ þ e ! ~eþ þ ~e ; ð9:13Þ

followed by the decays

~e
! e
þ 
e0 ; ð9:14Þ

giving overall

eþ þ e ! eþ þ e þ 
e0 þ 
e0 : ð9:15Þ

The cross-section for Equation (9.13) is predicted to be comparable to that for

producing pairs of ordinary charged particles. As the neutralinos only have weak
interactions they will be undetectable in practice and so the reaction would be
characterized by eþ e pairs in the final state with only a fraction (typically
50 per cent) of the initial energy and not emerging ‘back-to-back’ (because it is
not a two-body reaction). This and many other reactions have been studied, mainly
in experiments at LEP, but to date no evidence for the existence of superparticles
has been found. The null results have enabled lower limits to be set on the masses
of neutralinos and sleptons of various flavours in the range, 40 – 100 GeV/c2. This
is not very useful in practice, as the masses are believed to be of the order of the W
and Z masses. Much larger lower limits for the masses of gluinos and squarks have
been obtained in experiments using the CDF detector that was described in
Chapter 4 (see Figure 4.18). The search for supersymmetric particles will be a
major activity of detectors at the LHC accelerator currently under construction at
CERN.3
Undeterred by the lack of immediate success of supersymmetry, some bold
physicists have attempted to incorporate gravity into even larger unified schemes.

3
For a review of the current state of experimental searches for superparticles see, for example, Ei04.
 PARTICLE PHYSICS 307

The problems here are formidable, not least of which is that the divergences
encountered in trying to quantize gravity are far more severe than those in either
QCD or the electroweak theory and there is at present no successful ‘stand-alone’
quantum theory of gravity analogous to the former two. The theories that have
been proposed that include gravity invariably replace the idea of point-
like elementary particles with tiny quantized strings as a device to reduce these
technical problems and are formulated in many more dimensions (usually 10 or
11) than we observe in nature. More recently, even strings have been superceded
by theories based on mathematical objects called membranes, or simply branes.
The problem with these theories, leaving aside their formidable mathematical
complexity, is that they apply at an energy scale where gravitational effects are
comparable to those of the gauge interactions, i.e. at energies defined by the so-
called Planck mass MP , which is given by

 1=2

hc
MP ¼ ¼ 1:2  1019 GeV=c2 ; ð9:16Þ
G

where G is the gravitational constant.4 This energy is so large that it is difficult to

think of a way that the theories could be tested at currently accessible energies, or
even indeed at energies accessible in the conceivable future. Their appeal at
present is, therefore, the mathematical beauty and ‘naturalness’ that their sponsors
claim for them. Needless to say, experimentalists will remain sceptical until
definite experimental tests can be suggested and carried out.

9.1.4 Particle astrophysics

Particle physics and astrophysics interact in an increasing number of areas and the
resulting field of particle astrophysics is a rapidly expanding one. The interactions
are particularly important in the field of cosmology where, for example, the
detection of neutrinos can provide unique cosmological information. Another
reason is because the conditions in the early Universe implied by standard
cosmological theories (the big bang model) can only be approached, however
remotely, in high-energy particle collisions. At the same time, these conditions
occurred at energies that are relevant to the grand unified and SUSY theories of
particle physics and so offer a possibility of testing the predictions of such
theories. This is important because, as mentioned above, it is difficult to see
other ways of testing such predictions. For reasons of space, we will discuss just
three examples of particle astrophysics. We will return to the question of
conditions in the early universe in Section 9.2.2.

4
This implies that strings have dimensions of order ‘P
h=MP c ¼ 1:6  1035 m.
308 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

Neutrino astrophysics

We have seen in Chapter 3 that cosmic rays and emissions from the Sun are
important sources of information about neutrinos and have led us to revise the view
that neutrinos are strictly massless, as is assumed in the standard model. At the
same time, there is considerable interest in studying ultra high-energy neutrinos as
a potential source of information about galactic and extra-galactic objects and
hence cosmology in general.
One of the first neutrino astrophysics experiments was the observation of
neutrinos from a supernova. Supernovas are very rare events where a star literally
explodes with a massive output of energy over a very short timescale measured in
seconds. The mechanism for this (briefly) is as follows. If a star has a mass greater
than about 11 solar masses, it can evolve through all stages of fusion, ending in a
core of iron surrounded by shells of lighter elements. Because energy cannot be
released by the thermonuclear fusion of iron, the core will start to contract under
gravity. Initially this is resisted by the pressure of the dense gas of degenerate
electrons in the core (electron degeneracy pressure), but as more of the outer core is
burned and more iron deposited in the core, the resulting rise in temperature makes
the electrons become increasingly relativistic. When the core mass reaches about
1.4 solar masses (the so-called Chandrasekhar limit), the electrons become ultra
relativistic and they can no longer support the core. At this point the star is on the
brink of a catastrophic collapse.
The physical reactions that lead to this are as follows. Firstly, photodisintegra-
tion of iron (and other nuclei) takes place,

 þ 56 Fe ! 134 He þ 4n; ð9:17Þ

which further heats the core and enables the photodisintegration of the helium
produced, i.e.

 þ 4 He ! 2p þ 2n: ð9:18Þ

As the core continues to collapse, the energy of the electrons present increases to a
point where the weak interaction

e þ p ! n þ e ð9:19Þ

becomes possible and eventually the hadronic matter of the star is predominantly
neutrons. This stage is therefore called a neutron star. The collapse ceases when
the gravitational pressure is balanced by the neutron degeneracy pressure. At this
point the radius of the star is typically just a few kilometres. The termination of the
collapse is very sudden and as a result the core material produces a shock wave
that travels outwards through the collapsing outer material, leading to a supernova
(actually a so-called Type II supernova). Initially there is an intense burst of e
 PARTICLE PHYSICS 309

with energies of a few MeV from the reaction of Equation (9.19). This lasts for a
few milliseconds because the core rapidly becomes opaque even to neutrinos and
after this the core material enters a phase where all its constituents (nucleons,
electrons, positrons and neutrinos) are in thermal equilibrium. In particular, all
flavours of neutrino are present via the reactions

 Ð eþ e Ð ‘ 
‘ ; ð‘ ¼ e; ; Þ ð9:20Þ

and these will eventually diffuse out of the collapsed core and escape. Neutrinos
of all flavours, with average energies of about 15 MeV, will be emitted in all
directions over a period of 0.1–10 s. Taken together, the neutrinos account for
about 99 per cent of the total energy released in a supernova. Despite this, the
output in the optical region is sufficient to produce a spectacular visual effect.
The first experiments to detect neutrinos from a supernova were an earlier
version of the Kamiokande experiment described in Chapter 3 and the IMB
collaboration, which also used a water C̆erenkov detector. Both had been
constructed to search for proton decay as predicted by GUTs, but by good fortune
both detectors were ‘live’ in 1987 at the time of a spectacular supernova explosion
(now named SN1987A) and both detected a small number of antineutrino events.
The data are shown in Figure 9.7. The Kamiokande experiment detected 12 
e
events and the IMB experiment eight events, both over a time interval of
approximately 10 s and with energies in the range 0–40 MeV. These values are
consistent with the estimates for the neutrinos that would have been produce by the
reaction in Equation (9.20) and then diffused from the supernova after the initial
pulse.

Figure 9.7 Data for neutrinos from SN1987A detected in the Kamiokande and IMB experiments:
the threshold for detecting neutrinos in the experiments are 6 MeV (Kamiokande) and 20 MeV
(IMB) -- in each case the first neutrino detected is assigned the time zero
310 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

The data can be used to make an estimate of the neutrino mass as follows. The
time of arrival on Earth of a neutrino i is given by
  
L m2 c4
ti ¼ t0 þ 1þ ; ð9:21Þ
c 2Ei2

where t0 is the time of emission from the supernova and ðm; Ei Þ are the mass and
total energy of the neutrino. Thus
" #
L m2 c4 1 1
ðtÞij ti  tj ¼  : ð9:22Þ
2c Ei2 Ej2

Using data for pairs of neutrinos, Equation (9.22) leads to the result

m
e  20 eV; ð9:23Þ

which, although larger than the value from tritium decay, is still a remarkable
measurement.
The neutrinos from SN1987A were of low energy, but there is also a great
interest in detecting ultra high-energy neutrinos. For example, it is known that
there exist point sources of -rays with energies in the TeV range, many of which
have their origin within so-called active galactic nuclei. It is an open question
whether this implies the existence of point sources of neutrinos with similar
energies. The neutrinos to be detected would be those travelling upwards through
the Earth, as the signal from downward travelling particles would be swamped
by neutrinos produced via pion decay in the atmosphere above the detector. Like
all weak interactions the intrinsic rate would be very low, especially so for such
high-energy events, but this is partially compensated by the fact that the –nucleon
cross-section increases with energy, as we showed in Chapter 6.
To detect neutrinos in the TeV energy range using the C̆erenkov effect in water
requires huge volumes, orders-of-magnitude larger than used in the Super-
Kamiokande detector. An ingenious solution to this problem is to use the vast
quantities of water available in liquid form in the oceans, or frozen in the form of
ice at the South Pole, and several experiments have been built, or are being built,
using these sources. The largest so far is the Antartic Muon and Neutrino Detector
Array (AMANDA) which is sited at the geographical South Pole. A schematic
diagram of this detector is shown in Figure 9.8.
The detector consists of strings of optical modules containing photomultiplier
tubes that convert the C̆erenkov radiation to electrical signals. The enlarged inset
in Figure 9.8 shows the details of an optical module. They are located in the ice at
great depths by using a novel hot-water boring device. The ice then refreezes
around them. In the first phase of the experiment in 1993/94 (AMANDA-A) four
detector strings were located at depths of between 800 and 1000 m. The ice at
 PARTICLE PHYSICS 311

Figure 9.8 A schematic diagram of the AMANDA neutrino detector

these depths is filled with air bubbles and so the detectors are not capable of
precision measurements, but they proved the validity of the technique. In the next
phase a few years later (AMANDA-B10), 10 more strings containing 320 optical
modules were located at depths between 1.5 and 2.0 km, where the properties of
ice are suitable for muon detection. Finally, the current version of the detector
(AMANDA-II) has an additional nine strings extending to a depth of 2.35 km. In
total there are 680 optical modules covering a cylindrical volume with a cross-
sectional diameter of 120 m.
The AMANDA detector has successfully detected atmospheric neutrinos and
has produced the most detailed map of the high-energy neutrino sky to date.
However, no source of continuous emission has yet been observed that would be a
candidate for a point source.
312 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

AMANDA can detect neutrinos with energies up to about 1015 eV, but an even
bigger detector, called IceCube, is under construction at the South Pole. This uses 80
strings each containing 60 optical modules regularly spaced over an area of 1 km2 at
depths of between 1.4 and 2.4 km (the volume covered by AMANDA is only
1.5 per cent of the volume to be covered by IceCube) and will be capable of detecting
neutrinos with energies as high as 1018 eV. IceCube is due for completion in 2010.

Dark matter

The modern description of the universe is based on the observation that it is

expanding and assumes that the origin of this is a sudden explosion at some time in
the past. For this reason the description is called the big bang model. However, this
does not mean an explosion from a singular space–time point. Because the universe
appears isotropic at large distance scales, there can be no preferred points in space
and so the big bang must have occurred everywhere at once, thus ensuring that the
expansion appears the same to all observers irrespective of their locations. Two pieces
of evidence for this model are the existence of a cosmic background radiation
consistent with a black-body spectrum at an effective temperature of 2.7 K, and the
cosmic abundance of light elements.5 Whether the expansion will continue indefi-
nitely depends on the average density of the universe . The critical density c at
present times, below which the expansion will continue indefinitely, and above which
it will eventually halt and the universe start to contract, can be written

3H02
c ¼ 1026 kg m3  5:1 ðGeV=c2 Þm3 ; ð9:24Þ
8
G

where G is the gravitational constant and we have used the best current value for
Hubble’s constant H0 to evaluate Equation (9.24). In the most popular version of
the model, called the inflationary big bang model, the relative density

 =c ¼ 1: ð9:25Þ

The relative density is conveniently written as the sum of three components,

 ¼ r þ m þ  ; ð9:26Þ

where r is the contribution due to radiation, m is that due to matter and  is

related to a term in the equation governing the evolution of the universe that

5
For an accessible discussion of the big bang model and other matters discussed in this section see, for
example, Pe03.
 PARTICLE PHYSICS 313

contains a so-called cosmological constant . The latter contribution can be

estimated from various cosmological observations, including recently measured
temperature fluctuations in the microwave background radiation. Its value is about
0.7 and is the largest contribution to . The value of the radiation term is of the
order of 105 and so makes a negligible contribution to . Finally, the total matter
density contribution can be deduced from the gravitational energy needed for
consistency with observations on the rotation of galaxies and the kinematics of
large-scale structures in the universe. Its value is about 0.3. Thus we see that the
value of  is consistent with Equation (9.25), although the uncertainties are
considerable. An unsatisfactory feature is that the origin of the largest term, also
referred to as dark energy, is totally unknown.
The contribution of baryons to the mass term may be inferred from knowledge
of how nuclei are formed in the universe (nucleosynthesis) and its value is about
0.05, of which only about 20 per cent is accounted for in the form of stars, gas and
dust, i.e. in the form of visible luminous baryonic matter. There could be other
sources of non-luminous baryonic matter, for example in the form of brown dwarfs
and small black holes the size of planets, and there is experimental evidence that
such ‘massive, compact halo objects’ (MACHOs) do indeed exist, but in unknown
quantities. However, it is not thought that they alone can account for the ‘missing’
matter. Thus we are forced to conclude that the bulk of matter, as much as
85 per cent, is non-baryonic. It is referred to collectively as dark matter.
There are several dark matter candidates. Massive neutrinos might be one
possibility. Such particles would have to be heavy enough to have been non-
relativistic in the early stages of the universe (so-called cold dark matter), because
if they were relativistic (hot dark matter) they would have rapidly dispersed,
giving rise to a uniform energy distribution in space. Calculations suggest that in
this case there would have been insufficient time for the observed galaxies to have
formed. Although neutrinos may still play a minor role in contributing to the
matter deficit, it is now believed that the bulk of the contribution comes from cold
dark matter in the form of ‘weakly interacting massive particles’ (WIMPs).
Although there are no known particles that have the required properties, for
various reasons the most likely candidates are SUSY particles and in particular the
lightest such state, usually taken to be the neutralino.
Experiments such as AMANDA can search for WIMPs, but they were not
designed to do so as a priority. However, several dedicated experiments have been
mounted to detect WIMPs by detecting the recoil energy of interacting nuclei,
which is about 50 keV. Such recoils can, in principle, be detected in a number of
ways. For example, in semiconductors such as GaAs, free charge will be produced
that can be detected electronically; in a scintillator such as NaI the emission of
photons can be detected using photomultipliers; and in crystals at low temperatures
the energy can be converted to phonons that can be detected by a very small rise in
temperature. In practice, the problems are formidable because of the very low
expected event rate. This can be calculated from the expected WIMP velocities
and assumed masses. For example, if WIMPs are identified with neutralinos, then
314 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

expectations range from 1–10 events per kg of detector per week. This is very
small compared with the event rate from naturally occurring radioactivity,
including that in the materials of the detectors themselves. The former is
minimized by working deep underground to shield the detector from cosmic
rays and in areas with geological structures where radioactive rocks are absent; and
the latter is minimized by building detectors of extreme purity. Finally, WIMP
recoils should exhibit a small seasonal time variation due to the motion of the
Earth around the Sun and the motion of the Sun within the galaxy. One experiment
claims to have seen this variation. Present experiments are at an early stage, but
some versions of SUSY theories with low-mass neutralinos can probably already
be ruled out.6

Matter–antimatter asymmetry

One of the most striking facts about the universe is the paucity of antimatter
compared with matter. There is ample evidence for this. For example, cosmic
rays are overwhelmingly composed of matter and what little antimatter is present
is compatible with its production in intergalactic collisions of matter with photons.
Neither do we see intense outbursts of electromagnetic radiation that would
accompany the annihilation of clouds of matter with similar clouds of antimatter.
The absence of antimatter is completely unexpected because, in the original
big bang, it would be natural to assume a total baryon number B ¼ 0.7 Then
during the period when kT was large compared with hadron energies, baryons
and antibaryons would be in equilibrium with photons via reversible reactions
such as

p þ
p Ð  þ  ð9:27Þ

and this situation would continue until the temperature fell to a point where the
photons no longer had sufficient energy to produce p
p pairs and the expansion had
proceeded to a point where the density of protons and antiprotons was such that
their mutual annihilation became increasingly unlikely. The critical temperature is
kT  20 MeV and at this point the ratios of baryons and antibaryons to photons
‘freezes’ to values that can be calculated to be

NB =N ¼ NB =N 1018 ; ð9:28Þ

6
An up-to-date review of the status of dark matter searches is given in Pe03 and Ei04.
7
One could of course simply bypass the problem by arbitrarily assigning an initial non-zero baryon number
to the universe, but it would have to be exceedingly large to accommodate the observed asymmetry, as well
as being an unaesthetic solution.
 NUCLEAR PHYSICS 315

with of course NB =NB ¼ 1. These ratios would then be maintained in time,

whereas the actual observed ratios are

NB =N  109 ; NB =N 1013 ; ð9:29Þ

with NB =NB 104 . The simple big bang model fails spectacularly.
The conditions whereby a baryon–antibaryon asymmetry could arise were first
stated by Sakharov. It is necessary to have: (a) an interaction that violates baryon
number conservation, (b) an interaction that violates charge conjugation, and (c) a
non-equilibrium situation must exist at some point to ‘seed’ the process. We have
seen in Chapter 6 that there is evidence that CP is violated in the decays of
some neutral mesons, but its source and size are not compatible with that required
for the observed baryon–antibaryon asymmetry and we must conclude that there
is another, as yet unknown, source of CP violation. Likewise a method for
generating a non-equilibrium situation is also unknown, although it may be that
the baryon-violating interactions of GUTs, which are necessary for condition (a),
may provide one. Clearly, matter–antimatter asymmetry remains a serious
unsolved problem.

9.2 Nuclear Physics

Despite more than a century of research, nuclear physics is by no means a ‘closed’
subject. Even the basic strong nucleon–nucleon force is not fully understood at a
phenomenological level, let alone in terms of the fundamental quark–gluon strong
interaction. Indeed one of the outstanding problems of nuclear physics is to
understand how models of interacting nucleons and mesons arise as approxima-
tions to the quark–gluon picture of QCD and where these two descriptions merge.
A related question is whether the nuclear environment modifies the quark–gluon
structure of nucleons and mesons. It follows from our lack of knowledge in these
areas that the properties of nuclei cannot at present be calculated from first
principles, although some progress has been made in this direction. Meanwhile, in
the absence of a fundamental theory to describe the nuclear force, we have seen
in earlier chapters that specific models and theories are used to interpret the
phenomena in different areas of nuclear physics. Current nuclear physics models
must break down at very high energy-densities and at sufficiently high tempera-
tures the distinction between individual nucleons in a nucleus should disappear.
This is the regime that is believed to have existed in the very early times of the
universe and is of great interest to astrophysicists.
Nuclear physics is a mature subject and has implications in many other areas of
physics and wide applications in industry, biology and medicine that are at the core
of the subject. Examples include: the nuclear physics input required to understand
many processes that occur in cosmology and astrophysics, such as supernovae and
316 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

the production of chemical elements; the many applications of NMR, such as

studies of protein structure and its use in medical diagnostics; and industrial
applications such as the production of power. In Chapter 8 we touched on just three
applications and the ‘applied’ problems to be solved in those – safe disposal of
nuclear waste, better medical imaging diagnostics and therapeutics, controlled
nuclear fusion, etc. – are as demanding as the ‘fundamental’ ones, simply different.
They are also vitally important for the future well-being of everyone. In the
sections that follow we will take a brief look at a few of these pure and applied
problems.8

9.2.1 The structure of hadrons and nuclei

In the standard model, the structure of nucleons is specified in terms of quarks and
gluons, but questions remain. One concerns the spin of the proton. This must be
formed from combining the spins and the relative orbital angular momenta of its
constituent quarks and gluons. Measuring these various contributions can be done
in deep inelastic scattering experiments of the type described in Chapter 5, but
using spin-polarized targets, sometimes with spin-polarized beams. Experiments to
date have shown the surprising result that the spins of all the quarks and antiquarks
together contribute only about 20–30 per cent to the total spin of the proton (the
so-called ‘proton spin crisis’). There is some information that the angular
momentum contributions of the quarks play an important role, but very little is
known about the contribution of the total angular momentum of the gluon. This is
an area where the type of experiment that can be pursued at the CEBAF accelerator
described in Section 4.2.2 will be vital in unravelling the details of each
contribution and thus further testing QCD.
Nucleons and mesons are the building blocks of nuclear matter, but there is no
guarantee that the properties of these particles in nuclei are identical to those
exhibited as free particles. According to QCD the properties of hadrons are
strongly influenced by the sea of quark–antiquark pairs and gluons that we have
seen in Chapter 5 are always present around confined quarks due to quantum
fluctuations. However, these influences could well be different in the case of
closely spaced nucleons in nuclear matter from those for a free nucleon. Indeed
there are theoretical predictions that the probability of finding a qq pair decreases
as the density of the surrounding nuclear matter increases. If such effects could be
established they would have a profound influence on our understanding of quark
confinement.

8
A comprehensive overview of the field as at 1999 is a report of the Board on Physics and Astronomy of the
National Research Council, USA: ‘‘Nuclear Physics: The Core of Matter, The Fuel of Stars’’, National
Academy Press, Washington, D.C. (1999) – NRC99. Other useful sources are the publications of the Nuclear
Physics European Collaboration Committee (NuPecc) and in particular its ‘‘Report on Impact, Applications,
Interactions of Nuclear Science’’ (2002) and the NuPecc Long-Range Plan 2004.
 NUCLEAR PHYSICS 317

Figure 9.9 The ratios of the F2 structure function found from nuclear targets to that found
from deuterium, as a function of the scaling variable x (Carbon data from Ar95, calcium data from
Am95)

Another consequence of these predictions is that the effective masses of hadrons

will in general change in nuclear matter, as will their sizes and interactions. There
is already some evidence in favour of this suggestion from deep inelastic scattering
from nucleons (see Section 5.7) where the structure functions obtained using
targets of light and heavier nuclei differ slightly, even after allowing for calculable
effects such as nuclear binding energies and the internal Fermi motion of the
nucleons. (This is the so-called ‘EMC effect’, named after the group that first
discovered it.) It is illustrated in Figure 9.9, which shows the ratios F2Ca =F2D and
F2C =F2D , i.e. the F2 nucleon structure function deduced from calcium and carbon
targets divided by the structure function deduced using a deuterium target.
A number of other experiments have been performed to detect the effect of the
nuclear environment on effective masses (for example, by determining the mass of
mesons produced in nuclear matter) but nothing significant has been found
elsewhere. This will be a continuing field of study.
It is also important to study how the interactions of lower-energy hadrons
change when they are embedded in nuclear matter. For example, there is
considerable interest in the interactions of hadrons containing a strange valence
quark. (One reason is that they may play an important role in the high-density
matter present in neutron stars.) The lightest mesons that contain a strange valence
quark or antiquark are the kaons and these can be implanted in nuclei by nuclear
reactions that substitute a strange quark for an up or down quark. (This is an
example of a so-called ‘hypernucleus’.) Experiments at CEBAF and other
318 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

laboratories will provide information on the interaction of implanted, negatively

charged kaons with the surrounding nucleons in a nucleus.
The facilities at CEBAF and RHIC (the relativistic heavy ion accelerator
described in Section 4.2.2) will enable a range of new experimental possibilities
to be explored, in addition to those above. One is the intriguing question of the
existence of glueballs (mesons made of gluons alone) and hybrid quark–gluon
mesons, mentioned in Section 5.2 and vital for the theory of confinement via QCD.
The results may well help to find a solution to one of the central questions posed
at the start of this section: how are the properties of the strong nuclear force related
to the standard model formulation in terms of quarks and gluons?
There are also questions to be answered in the realm of nuclear structure,
many with implications elsewhere. For example, can the properties of nuclei
be related to those of an underlying nucleon–nucleon interaction and can they be
derived from many-body theory? At present we have a good knowledge from
scattering experiments of the long-range part of the nucleon–nucleon force in
terms of meson exchanges (see Section 7.1), but models that fit data differ
about the short-range part. This is not surprising because at separations of less
than 1 fm a description in terms of quarks and gluons is necessary and the
interface with QCD is critical. Experiments on meson production in nucleon–
nucleon collisions are sensitive to the short-range part of the forces and should
provide information on this region. On the theoretical side, advances in computer
power and calculational techniques have enabled the binding energies of all light
nuclei to be successfully calculated using the best available parameterization of
the nucleon–nucleon force. However, this is only possible by including an
explicit weaker three-nucleon force, which has to be adjusted to obtain the
correct binding energies. A satisfactory theory of the three-body force between
nucleons is lacking. This work also needs to be extended to heavier nuclei, but
present computer power is inadequate to the task using current computational
techniques.
One approach to the latter problem is to work within the framework of the shell
model, where each nucleon moves in the average potential (the mean field) generated
by its interactions with all the other nucleons in the nucleus. We have seen the
successes of this approach in simple applications in Section 7.3. When combined
with further computational improvements, it has enabled nuclear structure calcula-
tions to be extended to A ¼ 56. This is an important point for astrophysics, because
the details of the nuclear reactions of iron control the critical process occurring in the
collapse of a supernova, as we have seen above in Section 9.1.4.
Very often in science new insights are achieved by pushing experiments to their
limits. Nuclear physics is no exception. One such limit is the quest for super-heavy
elements. Discovery of elements beyond those currently known could explore
questions about possible limits on nuclear charges and masses. According to
nuclear models there should exist a new group of super-heavy elements with
charges Z in the range approximately 114 to 126 that are stabilized by shell effects.
 NUCLEAR PHYSICS 319

Figure 9.10 An energetic particle (typically several tens of MeV/u to GeV/u) is fragmented in
a nuclear reaction in a thin target, and radioactive reaction products are separated in-flight and
transported as a secondary beam to the experiment

The heaviest element made to date has Z ¼ 116 and was produced by fusion in the
reaction 48 248 292
20 Ca þ 96 Cm ! 116 Uuh þ 4n (the symbol Uuh is used as the element has
yet to be named). Strenuous efforts are being made to reach the predicted new
island of relative stability. This will require facilities to produce exotic short-lived
nuclear beams and there is much development work going on in this area. One
example of how such a beam can be formed is shown in Figure 9.10. The other
main method employs two independent accelerators: a high-power driver accel-
erator for production of the short-lived nuclei in a thick target that is directly
connected to an ion-source, and a second post-accelerator. Radioactive atoms
diffuse out of a hot target into an ion source where they are ionized for acceleration
in the post-accelerator.
Fewer than 300 stable nuclei occur naturally (see Figure 2.7) and outside the
stability region nuclei decay by the mechanisms discussed in Chapters 2 and 7. In
the uncharted regions there are many fundamental questions to be answered,
such as what are the limiting conditions under which nuclei can remain bound
and do new structures emerge near these limits? The answers to these questions
are important because theoretical descriptions of nuclei far from the line of
stability suggest that their structures are different from what has been seen in
stable nuclei. Nuclei far from stability also play an important role in astro-
physics, for example in understanding the processes in supernovae and how
elements are synthesized in stars. Another limiting region that is expected to
yield interesting information is that of angular momentum. Super-deformed
nuclei have been discovered with highly elongated shapes and very rapid
rotational motion. The states associated with these shapes are extremely stable.
Further investigation of these is expected to yield important information about
nuclear structure.
320 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

9.2.2 Quark--gluon plasma, astrophysics and cosmology

We have touched on the implications of nuclear physics for astronomy at various

places above. Here we look at other areas where improvements in our nuclear
physics knowledge would help astrophysics and cosmology.
In QCD, quarks and gluons are confined within hadrons, although the nature of
this confinement is still not fully understood. At extremely high energy-densities
the quarks and gluons are expected to become deconfined across a volume that
is large compared with that of a hadron. They would then exist in a new state
of matter, called a quark–gluon plasma, which is the state of nuclear matter
believed to have existed in the first few microseconds after the big bang (see
Figure 9.11).
It is possible to probe this state of matter using the RHIC facility (and also in a
few years at the LHC when its construction is complete). RHIC typically collides
two counter-circulating beams of fully-stripped gold ions at a maximum energy of
200 GeV per nucleon. If the ions collide centrally (i.e. head-on) several thousand
final-state particles are produced. An example of an event seen in the STAR
detector (which was shown in Figure 4.20) is illustrated in Figure 9.12. A key

Figure 9.11 Stages in the formation of a quark--gluon plasma and subsequent hadron
emission: two heavy nuclei collide at high energies (a) and interact via the colour field (b); the
very high energy-density produced causes the quarks and gluons to deconfine and form a plasma
that can radiate photons and lepton pairs (c); finally, as the plasma cools, hadrons condense and
are emitted (d) (after NRC99, with permission of the National Academics Press)
 NUCLEAR PHYSICS 321

Figure 9.12 View of a 200 GeV gold--gold interaction in the STAR detector at the RHIC
accelerator (Courtesy of Brookhaven National Laboratory)

question is whether the energy-density in the collisions is sufficient to have created

a quark–gluon plasma and its subsequent cooling phases. There are many
signatures for this, including the relative abundances of different final-state particle
types (for example, production of the c
c meson J= would be suppressed) and
measurements are all consistent with the expected temperature at which hadrons
would be formed (about 176 MeV, corresponding to about 1012 K, close to that
predicted by QCD) and that the temperature of the initial fireball is considerably
higher.
Future experiments at RHIC will play a crucial role in understanding the basic
nature of deconfinement. Questions to be addressed include: what is the nature of
matter at the highest densities (very recent experiments at RHIC suggest that the
plasma behaves more like a liquid than a gas); under what conditions can a quark–
gluon plasma be made; and what are the rules governing the evolution and the
transition to and from this kind of matter?
322 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

Information gathered from high-energy heavy-ion collisions is potentially

important in astrophysics. It will help constrain the equation of state that relates
the density of matter in neutron stars and supernovae (as well as in the first
microseconds of the early universe) to pressure and temperature. This information
will place stronger theoretical constraints on the maximum mass of a neutron star,
improving the ability to distinguish neutron stars and black holes.
The synthesis of nuclei in the very early universe is one of the cornerstones
of modern astrophysics, but even here there are still surprises. For example, in
the discussion of stellar fusion in Chapter 8, we saw that the production of
heavy elements involves the rare reaction 3 ð4 HeÞ ! 12 C (Equation (8.31)), the
occurrence of which depends critically on the existence of a particular excited
state of 12 C. We also noted that very recently another excited state has been
discovered at a somewhat higher energy which has the effect of significantly
altering the energy dependence (or equivalently the temperature dependence) of
this reaction from the values usually assumed. This could have major conse-
quences for models and theories of stellar evolution. Another recent experiment
has measured for the first time the lifetime of the doubly-magic nucleus 78 Ni and
finds it to be shorter than expected, implying that supernova explosions may
produce gold and other heavier elements much faster than had previously been
thought. This is important because 78 Ni is believed to produce more than half the
elements heavier than iron in the universe. A reaction of great current interest is
the synthesis of 16 O from the reaction of 4 He with 12 C (Equation (8.32)), which
determines the relative sizes of the carbon and oxygen shells of massive stars that
later explode in supernovae. The sizes of these shells are a crucial factor in
predicting the nucleosynthesis that occurs during the explosion. Nuclear physi-
cists are currently trying to measure the rate of this reaction with sufficient
accuracy to constrain astrophysical models.
One of the outstanding theoretical challenges in nuclear astrophysics is to
understand the process by which a massive, fully-evolved star ejects its mantle
while its core collapses to a neutron star or black hole. In Section 9.1.4 above we
gave a simple description of this process involving the collapse of the iron core to
several times the density of nuclear-matter, thereby producing a powerful shock
wave that travels outward through the mantle of the star. This shock wave, aided
by the heating of the matter by neutrinos emitted by the newly formed neutron
star, is responsible for the ejection of the mantle. This is the broad-brush picture,
but there is still no satisfactory theory that can account for the observed
frequency of supernovae. Efforts to understand dense nuclear matter and to predict
the properties of neutron stars depend on knowledge of nuclear interactions gained
in the laboratory. Heavy-ion collisions will help us better understand the interac-
tions of mesons in hot, dense nuclear matter, which is crucial to the issue of meson
condensation in neutron stars. Future studies of neutron-rich nuclei, near the
limit of stability, in radioactive ion beam facilities, as mentioned in Section 9.2.1
above, will allow more accurate modelling of nuclear forces in neutron star crusts.
 NUCLEAR PHYSICS 323

9.2.3 Symmetries and the standard model

An important symmetry that can be tested in nuclear physics is time reversal. We

have seen in Section 6.6 that CP invariance is violated in the weak decays of K and
B mesons and by inference so is T invariance, provided CPT invariance holds.
However, we have also seen in Section 9.1.4 above that the mechanism of violation
that can explain meson decays is unable to explain the observed matter/antimatter
asymmetry in the universe. Thus it is likely there exists another CP-violating
mechanism and hence another source of T violation.
There are several ways in principle of exploring CP violation in the context of
atomic and nuclear physics. One way involves antihydrogen. This was first
produced in a controlled experiment in 2002 by mixing cold antiprotons with a
dense positron plasma confined by electromagnetic fields in a so-called ‘Penning
trap’. If atoms of antihydrogen could be trapped for extended periods their
properties could be compared with those of hydrogen and this might shed light
on matter–antimatter asymmetry. CP violation can also be probed by searching for
electric dipole moments (EDMs) of the neutron, atoms or the electron. In the case
of atoms, an EDM could arise if the electron had an electric dipole moment or if
there were a T-violating interaction within the nucleus. Static EDMs are forbidden
if T invariance is exact and so a non-zero value would imply CP violation,
assuming CPT invariance holds. The present 90 per cent confidence limit on the
EDM dn of the neutron is dn < 6:3  1026 e cm and that for the electron
is de < 1:6  1027 e cm. Improving these presents formidable experimental
challenges. Nevertheless, several experiments are planned or are underway to
measure EDMs, with the aim of reducing the bounds to regions where they could
test the predictions of current theories. For the standard model these are
dn 1031 e cm and de 1038 e cm, although some extensions of the standard
model discussed in Section 9.1.3 above predict considerably larger values. Limits
on the existence of atomic and neutron EDMs already provide constraints on
some of the most plausible extensions to the standard model. It is also possible that
T-violation might show up in the decay of an unstable system. Modern experiments
are searching for T-violating correlations in the -decay of neutrons, mesons and
particular nuclei.
Atomic/nuclear physics can also provide information on the standard model
in other areas of the weak interactions. For example, a recent (2005) study of the
-decay of a metastable state of 38 K in an atomic trap has enabled severe limits to
be placed on a possible spin-0 particle to augment the spin-1 W-meson exchange.
The mixing between the weak and electromagnetic interactions can also be
studied. This is characterized by the Weinberg angle, which can be measured in
the parity-violating interaction between electrons and the nuclei of particular
atoms. This was mentioned at the end of Section 6.7. Parity mixing has been seen
in several atomic systems. The best measurement at present has been made using
133
Cs atoms, although the limits on the Weinberg angle do not yet compete with
324 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

those obtained from particle physics experiments. Other experiments plan to study
this effect in atomic francium, where the parity-mixing effect should be about 18
times larger. (The effect of an electric dipole moment of the electron is also
expected to be greatly enhanced in francium.) Unfortunately, francium is an
extremely rare element with no stable isotopes and so experiments will be carried
out with a small number of radioactive atoms collected in a magneto-optic trap.

9.2.4 Nuclear medicine

In Section 8.3.1, we reviewed the use of radiation techniques for cancer therapy.
We also briefly mentioned that in principle heavier particles had advantages over
photons. For example, because of the form of the Bragg curve, protons deposit
more of their energy where they stop, not where they enter the body. Also their
depth of penetration can be precisely controlled so that they stop within the
tumour, thus allowing radiologists to increase the radiation dose to the tumour
while reducing the dose to healthy tissues.
This is illustrated in Figure 9.13, which compares the treatment plans (i.e.
simulations of the pattern of radiation that the patient would receive) for treating a
case of advanced pancreatic cancer. Figure. 9.13(a) shows an X-ray plan using
a ‘state-of-the-art’ nine-beam X-ray system. The amount of radiation received by
nearby organs and other critical areas (kidneys, liver and spinal chord) is seen to be
a substantial fraction of the dose received by the region of the cancer. This is
contrasted with the results of Figure 9.13(b), which is for treatment using a
single proton beam. Although there is some unwanted exposure at the input site

Figure 9.13 Treatment plans for a large pancreatic tumour: (a) using a nine-beam X-ray
system; (b) using a single proton beam. The diffuse grey areas in (a) indicate the spread of
energy deposition outside the region of the tumour (adapted from Zu00, copyright Elsevier, with
permission)
 NUCLEAR PHYSICS 325

(which could be lessened by a system of multiple beams or a rotating beam), the

radiation energy is concentrated much more within the area of the tumour.
Although they have great potential, the problem with using particle beams is the
practical one of access to suitable accelerators. There is considerable effort being
made to design proton accelerators for cancer therapy and more than 20 centres
now exist worldwide specifically for proton therapy. Research is also continuing
with other forms of radiation therapy using neutrons and heavy ions. Neutrons
produce a high linear energy transfer (LET) and this overcomes a cancer cell’s
resistance to radiation damage more effectively than low-LET photon, electron or
proton radiation. Thus neutrons appear to be more biologically effective in killing
cancers than many other forms of radiation, especially in oxygen-poor cells.
Beams of heavy ions, such as carbon or neon, with energies of 400–800 MeV per
nucleon, are nearly ideal dose delivery vehicles for radiation therapy. Limited
studies with carbon and neon beams have been conducted and doubtless these
studies will increase in the future.
Progress in the NMR technique in medicine continues. For example, recent
advances have enabled a variation known as functional MRI (fMRI) to be
developed that exploits the paramagnetic behavior of deoxyhaemoglobin in red
blood cells. When in a magnetic field, a blood vessel containing deoxyhaemoglo-
bin distorts the field in its immediate environs, with the degree of distortion
increasing with the concentration of deoxyhaemoglobin. This distortion affects the
behaviour of water protons in the environs and, consequently, the magnetic-
resonance signal arising from these protons. Neural activation of a region of the
brain stimulates increased arterial flow of oxygenated blood, thereby decreasing
the concentration of deoxyhemoglobin in the region. Changes in the magnetic-
resonance signal can be detected and displayed as functional-MRI images. These
so-called BOLD (blood-oxygen-level dependent) images enable studies to be made
of the way the brain works by taking MRI images in real time while the patient is
performing specific tasks. In this way areas of the brain can be studied that are
associated with particular activities or sensations.
As another example, the gases 3 He and 129 Xe have the magnetic properties
needed for MRI and the atomic structure needed to retain their polarization for
hours at a time. They can be introduced into lungs, allowing MRI studies of lung
function. Because of the strong signal provided by the polarized nuclei in the
gas atoms, the MRI scans are short and can be synchronized with breathing.
Developments are also being made towards general high-speed imaging, which
would be useful for claustrophobic patients and children who are unable to be in
the confined environment of a conventional MRI magnet for sometimes up to an
hour.
Perhaps the greatest potential of all lies in the imaging of nuclei other than
hydrogen, particularly the phosphorus nucleus. Phosphorus is a major constituent
of the molecules adenosine triphosphate (ATP) and phosphocreatine, which
mediate the transfer of energy in living cells. From knowledge of such concentra-
tions it is possible to infer the metabolic status of internal organs, and it may
326 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

eventually be possible to add this capability to an imaging instrument. The future

will undoubtedly see both an improvement in the quality of NMR images and a
growing diversity of applications for nuclear magnetic resonance in clinical
practice.
An area that was not mentioned in Chapter 8 is the use of radioactive nuclear
isotopes produced by accelerators or nuclear reactors in many areas of biological
and biomedical research. For example, by inserting such radioisotopes as 14 C and
tritium, it is possible to obtain information on how molecules move through
the body, what types of cells contain receptors, and what kinds of compounds
bind to these receptors. Radioisotopes are also used directly to treat disease and
radioactive tracers are indispensable tools for the new forensic technique of DNA
fingerprinting, as well as for the Human Genome Project.

9.2.5 Power production and nuclear waste

Nuclear fusion still holds the promise of unlimited power without the problem of
radioactive waste, but the road to realization of this goal is long and we are far
from the end. In Section 8.2 we introduced the Lawson criterion as a measure of
how close a design was to the ignition point, i.e. the point at which a fusion
reaction becomes self-sustaining. To date no device has yet succeeded in achieving
the Lawson criterion and much work remains to be done on this important
problem. In recognition of this, at least one major new tokamak machine (to be
built in France) is planned as a global collaboration, but even when the ignition
point is attained, based on experience with fission reactors, it could be many
decades before that achievement is translated into a practical power plant.
In the shorter term and assuming that renewable sources of energy are insufficient
to fulfil the world’s increasing energy needs, it does seem as if power plants based on
fission reactions are the only hope of replacing fossil fuels in the future. The
problems of reactor safety and the safe disposal of radioactive waste are therefore
paramount.
The waste from light-water reactors, the most common type of power reactor,
has two major components: the actinides, i.e. any of the series of radioactive
elements with atomic numbers between 89 and 103 (mainly uranium but also
smaller amounts of heavier elements, the transuranic elements like plutonium and
the minor actinides such as neptunium, americium and curium) and fission
products, which are medium-weight elements from fission processes in the nuclear
fuel. While it is generally agreed that radioactive nuclei with relatively short
lifetimes can be safely stored in deep geological disposal facilities, the same is not
true of waste with very long lifetimes, some of which are water-soluble and so
have the potential to contaminate ground water. An additional problem is the
disposal of material that could be used for nuclear weapons, i.e. 239 Pu and 235 U.
One option for handling waste with very long lifetimes, which was mentioned as a
theoretical possibility in Section 8.1.3, is to transmute it by neutron reactions into
 NUCLEAR PHYSICS 327

shorter-lived, or even stable, isotopes that can be dealt with by conventional

storage.
The idea of using an accelerator to produce materials that can only be made
artificially has been around for more than 40 years, but more recently there has
been considerable interest and research in this idea to ‘incinerate’ nuclear waste
with the aim of reducing the waste lifetimes to less than 100 years. This is referred
to as ADS – Accelerator Driven System. In one proposed scheme, uranium
and most of the plutonium would be separated prior to proton irradiation and
used again as reactor fuel. The most important long-lived components of
the remaining waste would be isotopes of neptunium, americium, curium and
iodine, some with half-lives of 10 000 years or more. The approach would be to
irradiate this material with a new source of fast neutrons produced by spallation
reactions (cf. the discussion of producing neutron beams by this process in
Section 4.2.3) initiated using protons from a high-current accelerator. In this
way the capacity to ‘burn’ long-lived fission products and actnides is greatly
increased, leaving waste with much shorter lifetimes which can be disposed of by
conventional means. The accelerator would deliver a high-power (10–20 mA)
proton beam of about 1 GeV energy to a heavy metal (spallation) target surrounded
by the nuclear waste to be incinerated. The accelerator–waste combination would
be operated at a subcritical level – by itself it could not sustain a chain reaction –
so that no reactor-core meltdown accident could occur.
It has been suggested that this concept might be carried one step further, and a
particle beam might be used to produce additional neutrons directly in a nuclear-
reactor-like core. Versions of this concept have been studied in America and by a
European group. The latter is based on a proposal by Rubbia9 and is called the
Energy Amplifier. In this scheme, the core of the reactor would again be sub-
critical, and the accelerator beams would provide sufficient additional neutrons via
the spallation reaction to run the reactor. An idealized possible set-up is shown in
Figure 9.14.
Because the spallation neutrons would have high energy, a less enriched
element, such as natural thorium, could serve as the fuel. Thorium has the great
advantage over uranium in being an abundant element that does not require costly
isotope separation. Although the thorium fuel would not require enrichment, it
would need to be recharged every 5 years or so. The proposal has a number of
other advantages over a conventional power reactor, including: it is sub-critical
without the spallation neutrons and so is inherently safe – a meltdown or explosion
is not possible; radioactive waste is consumed in the reactor and no long-lived
waste is produced; there is no overlap with the nuclear weapons fuel cycle and so
the energy amplifier cannot be used as the basis for producing materials for nuclear
weapons, making installations politically acceptable worldwide.

9
The same man who shared the 1984 Nobel Prize in Physics for the discovery of the W and Z bosons.
328 CH9 OUTSTANDING QUESTIONS AND FUTURE PROSPECTS

Figure 9.14 Schematic diagram of a possible configuration of an energy amplifier; in this

design the coolant and spallation metal is molten lead (from Sc01, copyright Cavendish Press
Ann Arbor 2001, with permission)

The possible energy flow in a commercial system is shown in Figure 9.15. This
assumes a 1 GeV, 20 ma proton beam requiring about 20 MW of input power. The
latter is taken from the output of the reactor leaving a net electrical output of
580 MW, i.e. a gain factor of about 30.
The current situation on the energy amplifier is that a European collaboration
has shown that initial partitioning at the level of 95–99 per cent is possible

Figure 9.15 Possible energy flows in an energy amplifier system; the conversion efficiencies
are denoted by h
 NUCLEAR PHYSICS 329

depending on the actinide species. They have also carried out a number of
successful reactor transmutation and spallation studies and the first full ADS
experiment (TRADE) is currently under construction. This consists of coupling a
cyclotron delivering a 140 MeV, 0.5–1.0 ma proton beam to an existing 1MW
water-cooled reactor sited in Italy and uses a spallation target of tantalum. The
operation date is planned for 2007/08. Additional work is being carried out in
Belgium on coupling a 350 MeV, 5 ma proton beam to a 100 MW subcritical
reactor (the Myrrha experiment) and has already shown that some long-lived
isotopes can be successfully incinerated. Although ADS has enormous potential,
there are still a great many problems to be overcome and questions to be answered.
The estimated time for completion of research and development work and
commencement of an industrial plant based on ADS could be as long as 50 years.
Appendix A
Some Results in Quantum Mechanics

A.1 Barrier Penetration

Consider the one-dimensional potential shown in Figure A.1(a). Free particles of
mass m and energy E represented by plane waves are incident from the left and
encounter the constant rectangular barrier of height V, where V > E.
In region I ðx < 0Þ, there is an incoming wave eikx , where the wave number k is
given by
h2 k2 ¼ 2mE; ðA:1Þ

and also a wave reflected at the barrier travelling from right to left of the form
eikx . Thus the total wavefunction in region I is

1 ðxÞ ¼ Aeikx þ Beikx ; ðA:2Þ

where A and B are complex constants. Within the barrier, region II ð0 < x < aÞ,
the solution of the Schrödinger equation is a decaying exponential plus an
exponential wave reflected from the boundary at x ¼ a, i.e. the total wavefunction
is
x
2 ðxÞ ¼ Ce þ Dex ; ðA:3Þ

where C and D are complex constants and  is given by

h2 2 ¼ 2mðV  EÞ: ðA:4Þ

Finally, in region III ðx > aÞ to the right of the barrier, there is only an outgoing
wave of the form
ikx
3 ðxÞ ¼ Fe ; ðA:5Þ

where again F is a complex constant.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
332 APPENDIX A: SOME RESULTS IN QUANTUM MECHANICS

Figure A.1 Rectangular barrier with (a) wavefunction solutions, and (b) form of the incoming
and outgoing waves; (c) modelling an arbitrary smooth barrier as a series of rectangular barriers

We are interested in the transmission coefficient T, defined by

T  jF=Aj2 : ðA:6Þ

The values of F and A are found by imposing continuity of the wavefunction and
its first derivative, i.e. matching the values of these quantities at the two boundaries
x ¼ 0 and x ¼ a. The algebra may be found in any introductory book on quantum
mechanics.1 The result is


2

2keika



T ¼

: ðA:7Þ
2k coshðaÞ  iðk   Þ sinhðaÞ

2 2

The corresponding incident and transmitted waves are shown in Figure A.1(b)
(the reflected waves are not shown).
For large a, which corresponds to small penetrations, we can make the
replacement
1
sinhðaÞ  coshðaÞ  ea ðA:8Þ
2
and hence
 2
4k
T e2a : ðA:9Þ
k þ 2
2

1
See, for example, Chapter 6 of Me61.
 DENSITY OF STATES 333

The first factor is due to the reflection losses at the two boundaries x ¼ 0 and x ¼ a
and the decreasing exponential describes the amplitude decay within the barrier.
The first factor is slowly varying with energy and is usually neglected.
The result of Equation (A.9), ignoring the first factor, may be used to find the
transmission coefficient for an arbitrary smoothly-varying barrier by modelling it
as a series of thinPrectangular barriers. This is illustrated in Figure A.1(c). Thus by
replace 2a by 2 ðxÞ
x and taking the limit of small
x, the summation goes
over to an integral, i.e.
ð  12
2m
2a ! 2 dx 2 ½VðxÞ  E ðA:10Þ
h

and
" ð 12 #
2m
T  exp 2 dx 2 ½VðxÞ  E : ðA:11Þ
h

This is the essence of what is known as the WKB approximation in quantum

mechanics. Equation (A.11) was used in Section 7.6 to discuss -decay and in
Section 8.2.1 to discuss nuclear fusion.

A.2 Density of States

Consider a spinless particle of mass m confined within a cube of sides L and
volume V ¼ L3 , oriented so that one corner is at the origin (0,0,0) and the edges
are parallel to the x, y and z axes. If the potential is zero within the box, then the
walls represent infinite potential barriers and the solutions of the Schrödinger
equation must therefore vanish on all faces of the cube. It is straightforward to
show that the solutions of the Schrödinger equation satisfying these boundary
conditions are standing waves of the form
ðx; y; zÞ ¼ C sinðkx xÞ sinðky yÞ sinðkz zÞ; ðA:12Þ

where C is a constant and the components of the wave number k ¼ ðkx ; ky ; kz Þ take
the values
nx ny nz
kx ¼ ; ky ¼ ; kz ¼ ; ðnx ; ny ; nz Þ ¼ 1; 2; 3:::: ðA:13Þ
2 2 2
The energy of the particle is given by

h2 2 h2 k2 ðh Þ2 2

E¼ ðkx þ ky2 þ kz2 Þ ¼ ¼ ðnx þ n2y þ n2z Þ; ðA:14Þ
2m 2m 8m
334 APPENDIX A: SOME RESULTS IN QUANTUM MECHANICS

where k  jkj ¼ p=h and p is the particle’s momentum. Negative values of the
integers do not lead to new states since they merely change the sign of the wave
function Equation (A.12) and phase factors have no physical significance.
The allowed values of k form a cubic lattice in the quadrant of ‘k-space’ where
all the values of ðnx ; ny ; nz Þ are positive. Since each state corresponds to
one combination of ðnx ; ny ; nz Þ, the number of allowed states is equal to the
number of lattice points. The spacing between the lattice points is ðL= Þ, so the
density of points per unit volume in k-space is ðL= Þ3 . The number of lattice
points nðk0 Þ with k less than some fixed value k0 is the number contained within a
volume that for large values of k0 may be well approximated by the quadrant of a
sphere of radius k0 , i.e.
 
14 3 L 3 V 4 k03
nðk0 Þ ¼ k0 ¼ : ðA:15Þ
83 ð2 Þ3 3

Hence the number of points with k in the range k0 < k < ðk0 þ dk0 Þ is

V
dnðk0 Þ ¼ 3
4 k02 dk0 : ðA:16Þ
ð2 Þ

The density of states is defined as

ðk0 Þ  dnðk0 Þ=dk0 and so is given by
V

ðk0 Þ ¼ 3
4 k02 : ðA:17Þ
ð2 Þ

Thus
ðk0 Þdk0 is the number of states with k between k0 and k0 þ dk0 , or
equivalently
4 V 2

ðpÞdp ¼ p dp ðA:18Þ
ð2 hÞ3

is the number of states with momentum between p and p þ dp. This is the form
used in Equation (7.1) when discussing the Fermi energy in the Fermi gas model.
Equation (A.18) can also be written in terms of energy using E ¼ p2 =2m, when it
becomes
4 V

ðEÞdE ¼ m p dE ðA:19Þ
hÞ3
ð2 

and this was the form used in discussing -decay in Section 7.7.2.
Although the above derivation is for a particle confined in a box, the same
technique can be used for scattering problems. In this case we can consider a large
volume V ¼ L3 and impose ‘periodic’ boundary conditions

ðx þ L; y; zÞ ¼ ðx; y þ L; zÞ ¼ ðx; y; z þ LÞ ¼ ðx; y; zÞ: ðA:20Þ

 PERTURBATION THEORY AND THE SECOND GOLDEN RULE 335

Instead of standing waves, the solutions of the Schrödinger equation consistent

with Equation (A.20) are the travelling waves

eik:r ¼ eikx x eiky y eikz z ðA:21Þ

where

2nx 2ny 2nz

kx ¼ ; ky ¼ ; kz ¼ ; nx ; ny ; nz ¼ 0; 1; 2 :::: ðA:22Þ
L L L

The density of lattice points in k-space now becomes ðL=2 Þ3 , but unlike the
standing wave case, permutations of signs in Equation (A.22) do produce new
states and the whole quadrant of lattice points has to be considered. Thus these two
effects ‘cancel out’ and we arrive at the same result for the density of states in
Equations (A.18) and (A.19). This approach was used in discussing the formal
definitions of cross sections in Chapter 1.
All the above is for spinless particles. If the particle has spin then the density of
states must be multiplied by the appropriate spin multiplicity factor, taking account
of the Pauli principle as necessary. Thus, for example, for spin-12 particles, with two
spin states, Equation (A.19) becomes

8 V

ðEÞdE ¼ mp dE: ðA:23Þ
ð2 hÞ3

A.3 Perturbation Theory and the Second Golden Rule

Without detailed proof, we will outline the derivation from perturbation theory of
the important relationship between the transition probability per unit time for a
process and its matrix element.2
In perturbation theory, the Hamiltonian at time t may be written in general as

HðtÞ ¼ H0 þ VðtÞ; ðA:24Þ

where H0 is the unperturbed Hamiltonian and VðtÞ is the perturbation, which we

will assume is small. The solution for the eigenfunctions of H starts by expanding
in terms of the complete set of energy eigenfunctions jun i of H0 , i.e.
X
j ðtÞi ¼ cn ðtÞjun ieiEn t=h ; ðA:25Þ
n

2
We follow the derivation given in Chapter 9 of Ma92.
336 APPENDIX A: SOME RESULTS IN QUANTUM MECHANICS

where En are the corresponding energies. If j ðtÞi is normalized to unity, then

the squared coefficient jcn ðtÞj2 is the probability that at time t the system is in a
state jun i. Substituting Equation (A.25) into the Schrödinger equation leads to a
differential equation for the transition coefficients:

dcf ðtÞ X
ih ¼ Vfn ðtÞei!fn t cn ðtÞ; ðA:26Þ
dt n


where the matrix element Vfn  uf jVðtÞjun i and the angular frequency
!fn  ðEf  En Þ=h. If we assume initially ðt ¼ 0Þ that the system is in a state
jui i, then cn ð0Þ ¼ ni and the solutions for cf ðtÞ are found by substituting this result
into the right-hand side of Equation (A.22) giving, to first-order in V,

ðt
1
ci ðtÞ ¼ 1 þ Vii ðt0 Þdt0 ðA:27aÞ
ih
0

and

ðt
1 0
cf ðtÞ ¼ Vfi ðt0 Þei!fi ðt Þ dt0 ðf 6¼ iÞ: ðA:27bÞ
ih
0


2
For f 6¼ i, the quantity
cf ðtÞ
is the probability, in first-order perturbation theory,
that the system has made a transition from state i to state f.
The above is for a general time-dependent perturbation VðtÞ, but the results can
also be used to describe other situations, for example where the perturbation is
zero up to some time t0 and a constant thereafter. In this case, the integrals in
Equations (A.27) can be evaluated and, in particular, Equation (A.27b) gives, again
to first-order in V,

Vfi 
cf ðtÞ ¼ 1  ei!fi t ðA:28Þ
h!fi

and hence the probability of the transition i ! f is

" #


2 4
Vfi
2 sin2 ð12 !fi tÞ
Pfi ðtÞ ¼
cf ðtÞ
¼ : ðA:29Þ
h2 !2fi

The function in the square brackets in Equation (A.29) is shown in Figure A.2.
 PERTURBATION THEORY AND THE SECOND GOLDEN RULE 337

" #
sin2 ð12 !fi tÞ
Figure A.2 The function
!2fi

For sufficiently large values of t, it has the form of a large central peak with
much smaller side oscillations. In this case Pfi is only appreciable if

h
!fi
¼
Ef  Ei
 2 h=t ðA:30Þ

and then the square bracket can be replaced by a Dirac delta function3, i.e.

sin2 ð12 !fi tÞ 1

lim ¼ ht ðEf  Ei Þ; ðA:31Þ
t!1 !2fi 2

where the external factors are to preserve the normalization. Then

2



2
Pfi ðtÞ ¼ t Vfi ðEf  Ei Þ ðA:32Þ
h

3
The Dirac delta function was theÐ first so-called ‘generalized function’. It is defined by the two conditions:
þ1
(i) ðx0  xÞ ¼ 0 if x 6¼ x0 and 0
Ð x2 (ii) 0 1 0 ðx  xÞdx
0
¼ 1: It follows that if f ðxÞ is a function continuous in the
interval x1 < x < x2 , then x1 f ðx Þ ðx  xÞdx0 ¼ f ðxÞ if x1 < x < x2 or ¼ 0 if x < x1 or x > x2.
338 APPENDIX A: SOME RESULTS IN QUANTUM MECHANICS

and the transition probability per unit time is

dPfi ðtÞ 2



2

¼ Vfi ðEf  Ei Þ: ðA:33Þ
dt h

The above assumes that the final state is discrete, but it is more common for the
final states to form a continuum defined by the density of states
ðEÞ derived in
Section A.2 above. In this case, since
ðEÞdE is the number of states with energy
between E and E þ dE, we can write the transition rate per unit time dTfi =dt to a
group of states f with energies in this range as
ð
dTfi dPfi ðtÞ 2 h



2 i
¼
ðEf ÞdEf ¼ Vfi
ðEf Þ ; ðA:34Þ
dt dt h Ef ¼Ei

where the integral has been evaluated using the properties of the delta function.
Equation (A.34) is called the Second Golden Rule (sometimes Fermi’s Second
Golden Rule, although strictly the result is not due to Fermi) and has been used
in several places in this book, for example in Chapter 7 when discussing nuclear
-decay.
Appendix B
Relativistic Kinematics

In particle physics, most scattering interactions take place between particles whose
speeds are comparable with the speed of light c. This is often true even in decays,
particularly if light particles are emitted. The requirements of special relativity
therefore cannot be ignored. In nuclear physics accurate predictions can also often
only be obtained if relativistic effects are taken into account. In this appendix we
review (usually without proof) some relativistic kinematical results and the use of
invariants to simplify calculations.

B.1 Lorentz Transformations and Four-Vectors

Consider a particle of mass m in an inertial frame of reference S. Its co-ordinates
are ðt; rÞ  ðt; x; y; zÞ and its speed is u ¼ juj, where u is its velocity. In a second
inertial frame S0 its co-ordinates are ðt0 ; r0 Þ  ðt0 ; x0 ; y0 ; z0 Þ and its speed is u0 ¼ ju0 j
where u0 is its velocity. If S and S0 coincide at t ¼ 0 and S0 is moving with uniform
speed v in the positive z-direction with respect to S, then the two sets of
coordinates are related by the Lorentz transformation

x0 ¼ x
y0 ¼ y
ðB:1Þ
z0 ¼ ðvÞðz  vtÞ
t0 ¼ ðvÞðt  vz=c2 Þ
1
where ðvÞ ¼ ð1  2 Þ2 is the Lorentz factor and   v=c. From the definition of
velocity and using these transformations, the particle’s speed in S0 is related to its
speed in S by
uv
u0 ¼ ðB:2Þ
1  uv=c2

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
340 APPENDIX B: RELATIVISTIC KINEMATICS

and hence
1
ðu0 Þ  ½1  ðu0 =cÞ2 2 ¼ ðuÞðvÞð1  uv=c2 Þ: ðB:3Þ

As v ! 0, the transformations in Equations (B.1) approach the Galilean transfor-

mations.
The most general Lorentz transformation has its simplest form in terms of four-
vectors, whose general form is a ¼ ða0 ; a1 ; a2 ; a3 Þ ¼ ða0 ; aÞ. Then Equations (B.1)
become

a00 ¼ ða0  va3 =cÞ; a01 ¼ a1 ; a02 ¼ a2 ; a03 ¼ ða3  va0 =cÞ: ðB:4Þ

For example, the space-time four-vector is x ¼ ðct; xÞ and when used in Equations
(B.4) reproduces Equations (B.1). The scalar product of two four-vectors a and b is
defined as
ab  a0 b0  a  b ðB:5Þ

and is an invariant, i.e. is the same in all inertial frames of references.

The basic four-vector in particle kinematics is the four-momentum, defined by

P  mu; ðB:6Þ

where m is the rest mass and u is the four-velocity, whose components are
u ¼ ðvÞðc, vÞ, where v is the three-velocity and v  jvj. In terms of the total
energy E (i.e. including the rest mass) and the three-momentum p,

P ¼ ðE=c; pÞ: ðB:7Þ

Thus for two four-momenta P1 and P2 the invariant scalar product is

P1 P2 ¼ E1 E2 =c2  p1 :p2 ðB:8Þ

and for P1 ¼ P2 ¼ P,

P2 ¼ E2 =c2  p2 : ðB:9Þ

However, from Equations (B.5) and (B.6) we have u2 ¼ c2 and hence P2 ¼ m2 c2 ,

so combining this with Equation (B.9) gives

E2 ¼ p2 c2 þ m2 c4 : ðB:10Þ

It follows that

E ¼ ðvÞmc2 ; p ¼ ðvÞmv; v ¼ c2 p=E: ðB:11Þ

 FRAMES OF REFERENCE 341

The Lorentz transformations for energy and momentum follow from these
definitions and Equations (B.4). Thus, in S0 we have

E0 ¼ m c2 ðu0 Þ ¼ ðvÞðE  vpÞ ðB:12aÞ

and

p0 ¼ mu0 ðu0 Þ ¼ ðvÞðp  vE=c2 Þ; ðB:12bÞ

where p ¼ jpj and p0 ¼ jp0 j. For a set of N non-interacting particles,

p0z ¼ ðvÞðpz  vE=c2 Þ; p0x ¼ px ; p0y ¼ py ; ðB:13aÞ

and

E0 ¼ ðvÞðE  vpz Þ; ðB:13bÞ

where

X
N X
N
E¼ Ei and p¼ pi : ðB:13cÞ
i¼1 i¼1

In the general case where the relative velocity v of the two frames is in an arbitrary
direction, the transformations in Equations (B.12) become
 
0 vp 1
p ¼ p þ v E 2; E0 ¼ ðE  v  pÞ: ðB:14Þ
þ1 c

B.2 Frames of Reference

The two most commonly used frames of reference for particle kinematics are the
laboratory system (LS) and the centre-of-mass system (CMS). We will start by
discussing these in the context of two-particle scattering. In the LS, a moving
projectile a in a beam strikes a target particle b at rest, i.e.

Pa ¼ ðEa =c; pa Þ; Pb ¼ ðmb c; 0Þ: ðB:15Þ

In the CMS, the three-momenta of the two particles a and b are equal and opposite,
so that the total momentum is zero,1 i.e.

Pa ¼ ðEa =c; pa Þ; Pb ¼ ðEb =c; pb Þ; ðB:16aÞ

1
Although ‘centre-of-mass’ system is the most frequently used name, some authors refer to this as the
‘centre-of-momentum’ system. Logically, a better name would be ‘zero-momentum’ frame.
342 APPENDIX B: RELATIVISTIC KINEMATICS

with
pa þ pb ¼ 0: ðB:16bÞ

In a colliding beam accelerator, these two views become mixed. The colliding
particles are both moving, but only if they have equal momenta and collide at zero
crossing angle is the system identical to the centre-of-mass system.
The four-vectors of the initial-state particles in the two systems may be written
(L ¼ laboratory, T ¼ target)

Pa ¼ ðEL =c; 0; 0; pL Þ; PT ¼ ðmT c; 0; 0; 0Þ LS ðB:17aÞ

with EL2 ¼ m2B c4 þ p2L c2 (B ¼ beam), and

Pa ¼ ðEa =c; 0; 0; pÞ; Pb ¼ ðEb =c; 0; 0; pÞ CMS ðB:17bÞ

with Ea2 ¼ m2B c4 þ p2 c2 and Eb2 ¼ m2T c4 þ p2 c2 .

The Lorentz transformations between them are

p ¼ ðpL  vEL =c2 Þ; Ea ¼ ðEL  vpL Þ; ðB:18Þ

where

c2 pL EL þ mT c2 pL
v¼ ; ¼ pffiffi ; v ¼ pffiffi ðB:19Þ
EL þ mT c2 c2 s s

and s is the invariant mass squared of the system defined by

s  ðpa þ pb Þ2 =c2 ¼ ½ðEa þ Eb Þ2  ðpa c þ pb cÞ2 =c4 : ðB:20Þ

In particular, in the LS,

s ¼ m2T þ m2B þ 2mT EL =c2 : ðB:21Þ

This result was used in Chapter 4 when discussing the relative merits of fixed-
target and colliding beam accelerators.
Substituting Equations (B.19) into Equations (B.18) gives

pL mT m2B c2 þ mT EL
p ¼ pffiffi ; Ea ¼ pffiffi ðB:22aÞ
s s

and similarly for particle b:

pL mT m2T c2 þ mT EL
p ¼ pffiffi ; Eb ¼ pffiffi : ðB:22bÞ
s s
 FRAMES OF REFERENCE 343

Finally we state, without proof, the transformation of scattering angles for the
specific case of laboratory and centre-of-mass systems. Consider the general
scattering reaction

BðEL ; pL Þ þ Tðm2T ; 0Þ ! PðE; qÞ þ       ; ðB:23Þ

where B is a beam particle incident on a target particle T at rest in the laboratory

system and P is one of a number of possible particles in the final state. If pL is
taken along the z-direction, then
pL ¼ ð0; 0; pL Þ and q ¼ ð0; q sin L ; q cos L Þ; ðB:24Þ

where L is the scattering angle in the laboratory system, i.e. the angle between the
beam direction and q. In the CMS,

p0B þ p0T ¼ 0; ðB:25Þ

where p0B and p0T are the CMS momenta of the beam and target, respectively. The
relation between the scattering angle C in this system and L is
1 q0 sin C
tan L ¼ ; ðB:26Þ
ðvÞ q0 cos C þ vE0 =c2

where

E0 ¼ mP c2 ðuÞ and q0 ¼ mP uðuÞ ðB:27Þ

and u is the magnitude of the velocity of P in the centre-of-mass frame.

It is instructive to consider the form of Equation (B.26) at high energies. From
Equation (B.19) the velocity of the transformation is

v ¼ pL c2 ½EL þ mT c2 1 ; ðB:28Þ

so at high energies where EL2  pL c  mB c2 , mT c2 , v  cð1  mT c=pL Þ  c and

 
pL 1=2
ðvÞ  : ðB:29Þ
2mT c

Substituting Equations (B.27), (B.28) and (B.29) into Equation (B.26) gives
 
2mT c 1=2 u sin C
tan L   : ðB:30Þ
pL u cos C þ c

Thus, unless u  c and cos C  1, the final-state particles will lie in a narrow
cone about the beam direction in the laboratory system. Similarly, when a
344 APPENDIX B: RELATIVISTIC KINEMATICS

high-energy particle decays, its decay products will emerge predominantly at small
angles to the initial beam direction.

B.3 Invariants
The transformations between laboratory and centre-of-mass systems for energy
and momentum have been worked out explicitly above, but a more efficient way is
to work with quantities that are invariants, i.e. have the same values in all inertial
frames. We have already met one of these: s the invariant mass squared, defined in
Equation (B.20). We will now find expressions for the energy and momentum in
terms of invariants for both the LS and the CMS.
First, in the LS, from Equations (B.15), we have
pB ¼ 0; EB ¼ mB c: ðB:31Þ

However, from Equation (B.23),

s ¼ m2B þ m2T þ 2mT EL =c2 ðB:32Þ

i.e.

ðs  m2T  m2B Þc2

EL ¼ ðB:33Þ
2mT

and so

EL2 ðs  m2B  m2T Þ2 c2  4m2B m2T c2

p2L ¼  m2 2
B c ¼ : ðB:34Þ
c2 4m2T

This can be written in the useful compact form

c 1
pL ¼ 2 ðs; m2B ; m2T Þ; ðB:35aÞ
2mT

where the triangle function  is defined by

ðx; y; zÞ  ðx  y  zÞ2  4yz: ðB:35bÞ

This function is invariant under all permutations of its arguments and in particular
Equation (B.35a) can be written in the form
h ih i12
c
pL ¼ s  ðmT þ mB Þ2 s  ðmT  mB Þ2 : ðB:36Þ
2mT
 PROBLEMS 345

In a similar way it is straightforward to show that, in the CMS,

h ih i12
c 2 2
p ¼ pffiffi s  ðmT þ mB Þ s  ðmT  mB Þ ðB:37Þ
2 s

from which it follows that

ðs þ m2B  m2T Þc2 ðs  m2B þ m2T Þc2

Ea ¼ pffiffi ; Eb ¼ pffiffi : ðB:38Þ
2 s 2 s

The above formulae have many applications. For example, if we wish to produce
particles with a certain mass M, the minimum laboratory energy of the beam
particles is, from Equation (B.33),

M 2 c2  m2B c2  m2T c2
EL ðminÞ ¼ : ðB:39Þ
2mT

In the case of the decay of a particle A to a set of final-state particles i ¼

1; 2; 3; . . . ; N, i.e.
A ! 1 þ 2 þ 3 þ    þ N; ðB:40Þ

the invariant mass W of the final-state particles is given by

!2 !2
X X
W 2 c4 ¼ Ei  pi c ¼ EA2  ðpA cÞ2 ¼ MA2 c4 : ðB:41Þ
i i

Hence the mass of the decaying particle is equal to the invariant mass of its decay
products. The latter can be measured if the particle is too short-lived for its mass to
be measured directly.

Problems
B.1 The Mandelstam variables s, t and u are defined for the reaction A þ B ! C þ D by

s ¼ ðpA þ pB Þ2 =c2 ; t ¼ ðpA  pC Þ2 =c2 ; u ¼ ðpA  pD Þ2 =c2 ;

where pA etc. are the relevant energy-momentum four-vectors.

(a) Show that

X
sþtþu¼ m2j :
j¼A;B;C;D
346 APPENDIX B: RELATIVISTIC KINEMATICS

(b) In the case of elastic scattering show that t ¼ 2p2 ð1  cos Þ=c2 , where
p  jpj, p is the centre-of-mass momentum of particle A and  is its scattering
angle in the CMS.

B.2 A pion travelling with speed v  jvj in the laboratory decays via  ! þ
. If the
neutrino emerges at right angles to v, find an expression for the angle  at which the
muon emerges.

B.3 A pion at rest decays via  ! þ

. Find the speed of the muon in terms of the
masses involved.

B.4 A neutral particle X 0 decays via X 0 ! Aþ þ B . The momentum components of the

final-state particles are measured to be (in GeV/c):

px py pz
þ
A 0:488 0:018 2.109

B 0.255 0.050 0.486

Test the hypotheses that the decay is (a) D0 ! þ þ K  and (b)  ! p þ  .

B.5 In a fixed-target e p scattering experiment, show that the squared four-momentum

transfer is given by Q2  2E2 ð1  cos Þ=c2, where E is the total laboratory energy
of the initial electron and  is the laboratory scattering angle.

B.6 Calculate the minimum laboratory energy Emin of the initial proton for the
production of antiprotons in a fixed-target experiment using the reaction
pp ! ppp p. If the protons are bound in nuclei, show that taking the internal motion
of the nucleons into account leads to a smaller minimum energy given by

0
Emin  ð1  p=mP cÞEmin ;

where p is the modulus of the average internal longitudinal momentum of a

0
nucleon. Use a typical value of p to calculate Emin .

B.7 A particle A decays at rest via A ! B þ C. Find the total energy of B in terms of the
three masses.

B.8 A meson M decays via M ! . Find an expression for the angle in the laboratory
between the two momentum vectors of the photons in terms of the photon energies
and the mass of M.

B.9 Pions and protons, both with momentum 2 GV/c, travel between two scintillation
counters distance L m apart. What is the minimum value of L necessary to
 PROBLEMS 347

differentiate between the particles if the time-of-flight can be measured with an

accuracy of 200 ps?

B.10 A photon is Compton scattered off a stationary electron through a scattering angle of
60 and its final energy is half its initial energy. Calculate the value of the initial
energy in MeV.
Appendix C
Rutherford Scattering

C.1 Classical Physics

In Chapter 1 we commented on the experiments of Geiger and Marsden that
provided evidence for the existence of the nucleus. They scattered low-energy

-particles from thin gold foils and observed that sometimes the projectiles were
scattered through large angles, in extreme cases close to 180
. If we start for the
moment by ignoring the fact that there is a Coulomb interaction present, then it is
easy to show that this behaviour is incompatible with scattering from light particles
such as electrons.
Consider the non-relativistic elastic scattering of an
-particle of mass m
and
initial velocity vi from a target of mass mt stationary in the laboratory. If the final
velocities are vf and vt , respectively, then we have the situation as shown in Figure C.1.
Conservation of linear momentum and kinetic energy are:

m
vi ¼ m
vf þ mt vt ðC:1Þ

and

m
v2i ¼ m
v2f þ mt v2t ; ðC:2Þ

where vi ¼ jvi j etc.. Squaring Equation (C.1) we obtain

m2t 2
m
v2i ¼ m
v2f þ v þ 2mt ðvf  vt Þ ðC:3Þ
m
t

and hence from Equation (C.2),

 
mt
v2t 1 ¼ 2vf  vt : ðC:4Þ
m

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
350 APPENDIX C: RUTHERFORD SCATTERING

Figure C.1 Kinematics of the Geiger and Marsden experiment

Thus, if the target is an electron, with mt ¼ me m
, the directions of motion of

the outgoing
-particle and the recoiling target are essentially along the direction
of the initial
-particle and no large-angle scatterings are possible. Such events
could, in principle, be due to multiple small-angle scattering, but the thinness of
the gold foil target rules this out.1 If, however, mt ¼ mAu
m
, then the left-hand
side of Equation (C.4) will be negative and large scattering angles are possible.
The above only makes plausible the existence of a heavy nucleus, because it has
ignored the existence of the Coulomb force, so we now have to take this into
account. We will do this first using non-relativistic classical mechanics.
Consider the non-relativistic Coulomb scattering of a particle (the projectile) of
mass m and electric charge ze from a target particle of mass M and electric charge
Ze. The kinematics of this are shown in Figure C.2. The target mass is assumed to
be sufficiently large that its recoil may be neglected. The initial velocity of the
projectile is v and it is assumed that in the absence of any interaction it would
travel in a straight line and pass the target at a distance b (called the impact

Figure C.2 Kinematics of Rutherford scattering

1
For completeness one should also show that the observations cannot be due to scattering from the diffuse
positive charge present. This was done by the authors of the original experiment.
 CLASSICAL PHYSICS 351

parameter). The derivation follows from considering the implications of linear and
angular momentum conservation.
Angular momentum conservation implies that
d
mvb ¼ mr 2 ; ðC:5Þ
dt
where v ¼ jvj. Since the scattering is symmetric about the y-axis, the component of
linear momentum in the y-direction is initially p ¼ mv sinð=2Þ and changes to
þmv sinð=2Þ after the interaction, i.e. the total change in momentum in the y-
direction is

p ¼ 2mv sinð=2Þ: ðC:6Þ

The change in momentum may also be calculated by integrating the impulse in the
y-direction due to the Coulomb force on the projectile. This gives
ð
þ1
zZe2

p ¼ cos  dt; ðC:7Þ
4"0 r2
1

where we have taken t ¼ 0 to coincide with the origin of the x-axis. Using
Equation (C.5) to change variables, Equation (C.7) may be written

 þ
ð
zZe2 1
2mv sinð=2Þ ¼ cos  d; ðC:8Þ
4"0 bv


which, using  ¼ ð  Þ=2, yields

zZe2 1
b¼  cotð=2Þ; ðC:9Þ
8"0 Ekin

where Ekin ¼ 12 mv2 is the kinetic energy of the projectile.

Finally, we need to calculate the differential cross-section. If the initial flux of
projectile particles crossing a plane perpendicular to the beam direction is J, then
the intensity of particles having impact parameters between b and b þ db is
2b Jdb and this is equal to the rate dW at which particles are scattered into a solid
angle d
¼ 2 sin d between  and  þ d. Thus

dW ¼ 2bJ db: ðC:10Þ

However, from Equation (1.47) and considering a single target particle,

d d
dW ¼ J d
¼ 2J sin  d ; ðC:11Þ
d
d

352 APPENDIX C: RUTHERFORD SCATTERING

i.e.

d b db
¼  : ðC:12Þ
d
sin  d

The right-hand side of Equation (C.12) may be evaluated from Equation (C.9) and
gives

2
d zZe2
¼ cosec4 ð=2Þ: ðC:13Þ
d
16 "0 Ekin

This is the final form of the Rutherford differential cross-section for non-
relativistic scattering.

C.2 Quantum Mechanics

While Equation (C.13) is adequate to describe the Geiger and Marsden experi-
ments, in the case of electron scattering we need to take account of both relativity
and quantum mechanics. This may be done using the general formalism for the
differential cross-section in terms of the scattering potential that was derived in
Chapter 1.
The starting point is Equation (1.55), which in the present notation is

d 1 p0 2  2
¼ 2 4 0 Mðq2 Þ ; ðC:14Þ
d
4 h vv

where v and p are the velocity and momentum respectively of the projectile (which
for convenience we take to have a unit negative charge) with v ¼ jvj, p ¼ jpj and
the primes refer to the final-state values. The matrix element is given by
ð
MðqÞ ¼ VðxÞeiqx=h dx; ðC:15Þ

where q ¼ p  p0 is the momentum transfer. VðxÞ is the Coulomb potential

ZðhcÞ
VðxÞ ¼ VC ðxÞ ¼  ; ðC:16Þ
r

where r ¼ jxj and Ze is the charge of the target nucleus. Inspection of the integral
in Equation (C.15) shows that it diverges at large r. However, in practice, charges
 QUANTUM MECHANICS 353

are always screened at large distances by intervening matter and so we will

interpret the integral as
ð

Lt Z
ðhcÞe
r iqx=h 3
MC ðqÞ ¼  e d x: ðC:17Þ

!0 r

To evaluate this, take q along the x-axis, so that in spherical polar coordinates
q  x ¼ qr cos . The angular integration may then be done and yields
ð
1
4ðhcÞZ
h Lt
MC ðqÞ ¼  e
r sinðqr=hÞdr: ðC:18Þ
q
!0
0

The remaining integral may be done by parts (twice) and taking the limit
! 0
gives

4ðhcÞZ
h2
MC ðqÞ ¼  : ðC:19Þ
q2

Finally, substituting Equation (C.19) into Equation (C.14) gives

d p0 2
¼ 4Z 2
2 ðhcÞ2 0 4 ; ðC:20Þ
d
vv q

which is the general form of the Rutherford differential cross-section. To see that
this is the same as Equation (C.13) in the non-relativistic limit, we may substitute
the non-relativistic approximations
2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
p2 ¼ p0 ¼ 2mEkin ; and v ¼ v0 ¼ 2Ekin =m; ðC:21Þ

together with the kinematic relation for the scattering angle

q ¼ 2p sinð=2Þ; ðC:22Þ

into Equation (C.20). The result in Equation (C.13) follows immediately.

Because we are assuming that the target mass is heavy so that its recoil may be
neglected, to a good approximation p ¼ p0 and E ¼ E0 , where E is the total energy
of the electron. Also for relativistic electrons v ¼ v0  c and E  pc. Using these
conditions together with Equation (C.22) in Equation (C.20), gives the relativistic
result for the Rutherford differential cross-section in the convenient form:

d Z 2
2 ðhcÞ2
¼ 2 4 ; ðC:23Þ
d
4E sin ð=2Þ

which is the form used in Chapter 2 and elsewhere.

354 APPENDIX C: RUTHERFORD SCATTERING

Problems
C.1 Calculate the differential cross-section in mb/sr for the scattering of a 20 MeV

-particle through an angle 20
by a nucleus 209
83 Bi, stating any assumptions made.
Ignore spin and form factor effects.

C.2 Show that in Rutherford scattering at a fixed impact parameter b, the distance of
closest approach d to the nucleus is given by d ¼ b½1 þ cosec ð=2Þ=cosec ð=2Þ,
where  is the scattering angle.

C.3 Find an expression for the impact parameter b in the case of small-angle Rutherford
scattering. A beam of protons with speed v ¼ 4  107 ms1 is incident normally on a
thin foil of 194 78 Pt, thickness 10
5
m (density ¼ 2:145  104 kg m3 ). Estimate the
proportion of protons that experience double scattering, where each scattering angle
is at least 5
.
Appendix D
Solutions to Problems

Chapter 1
1.1 Substituting the operators p ¼ i
h@=@x and E ¼ i
h@=@t into the mass–energy
relation E2 ¼ p2 c2 þ M 2 c4 and allowing the operators to act on the function
ðx, tÞ, leads immediately to the Klein–Gordon equation. To verify that the Yukawa
potential VðrÞ is a static solution of the equation, set VðrÞ ¼ ðxÞ, where r ¼ jxj,
and use
@2 2 @
r2 ¼ þ
@r2 r @r

together with the expression for the range, R ¼

h=Mc.

1.2 Using Equation (1.11), gives

rffiffiffi rffiffiffi
^ Y11 ¼ 3 iðþÞ 3
P sinð  Þe ¼ sinðÞei ¼ Y11 ;
8 8

and hence Y11 is an eigenfunction of parity with eigenvalue 1.

1.3 Because the initial state is at rest, it has L ¼ 0 and thus its parity is
Pi ¼ Pp P
p ð1ÞL ¼ 1, where we have used the fact that the fermion–antifermion
pair has overall negative intrinsic parity. In the final state, the neutral pions are
identical bosons and so their wavefunction must be totally symmetric under their
interchange. This implies even orbital angular momentum L0 between them and
2 L0
hence Pf ¼ P ð1Þ ¼ 1 6¼ Pi . The reaction violates parity conservation and is thus
forbidden as a strong interaction.
 
1.4 Since ^ 2 ¼ 1, we must have C
C ^ 

^ 2 jb; b i ¼ Cb C b;
b ¼ jb; b i, implying that
 
^ 
b;
b ¼ C
b jb; b i with Cb C
b ¼ 1 independent of Cb . The result follows because
C
an eigenstate of C ^ must contain only particle–antiparticle pairs b
b, leading to the
intrinsic parity factor Cb C
b ¼ 1, independent of Cb .

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
356 APPENDIX D: SOLUTIONS TO PROBLEMS

1.5 The parity of the deuteron is Pd ¼ Pp Pn ð1ÞLpn . Since the deuteron is an

S-wave bound state, Lpn ¼ 0 and so, using Pp ¼ Pn ¼ 1, gives Pd ¼ 1. The
parity of the initial state is therefore Pi ¼ P Pd ð1ÞLd ¼ P , because the
pion is at rest and so Ld ¼ 0. The parity of the final state is
Pf ¼ Pn Pn ð1ÞLnn ¼ ð1ÞLnn and therefore P ¼ ð1ÞLnn . To find Lnn impose the
condition that nn ¼ space spin must be antisymmetric. Examining the spin,
Equation (1.17) shows that there are two possibilities for spin : either the symmetric
S ¼ 1 state or the S ¼ 0 antisymmetric state. If S ¼ 0, then space would have to be
symmetric, implying Lnn would be even, but the total angular momentum would not
then be conserved. Thus S ¼ 1 is implied and space is antisymmetric, i.e.
Lnn ¼ 1; 3;    . The only way to combine Lnn and S to give J ¼ 1 is with Lnn ¼ 1
and hence P ¼ 1.

1.6 (a) e þ eþ ! e þ eþ ;

(b) p þ p ! p þ p þ 0 þ 0 ;

(c)
p þ n !  þ 0 þ 0 ;  þ þ þ  .

1.7 (a) e þ  ! e þ  .

(b) n ! p þ e þ 
e .
 CHAPTER 1 357

(c) eþ þ e ! eþ þ e .

(d)
þ
! eþ þ e .

1.8 If an exchanged particle approaches to within a distance d fm, this is equivalent to a

momentum transfer q ¼
h=d ¼ ð0:2=dÞ GeV/c. Thus, q ¼ 0:2 GeV/c for d ¼ 1 fm
and q ¼ 200 GeV/c for  d ¼ 103 fm. The scattering amplitude is given by
1
h2 q2 þ m2x c2 , where mx is the mass of the exchanged particle. Thus,
f ðq2 Þ ¼ g2

fEM ðq2 Þ q2 c2 þ m2W c4

Rðq2 Þ ¼ 2 2 ;
fWeak ðq Þ q c þ m2
c4
2

since gEM  gWeak . Using m

¼ 0 and mW ¼ 80 GeV/c2, gives
Rð0:2 GeV=cÞ  1:6  105 fm but Rð200 GeV=cÞ  1:2 fm:
1.9 Using spherical polar coordinates, we have q  x ¼ qr cos  and d3 x ¼ r2 dr d cos  d,
where q ¼ jqj. Thus, from Equation (1.38),

ð
2 ð
1 ð
þ1
2 g2 er=R
2
f ðq Þ ¼ d dr r hÞ
d cos  expðiqr cos =

4 r
0 0 1
ð
1 ð
1
h i
g2

h g2

h
¼ dre r=R
hÞþ1
½expðiqr cos =
1 ¼ drer=R eiqr=
h  eiqr=
h
2iq 2iq
0 0
2 2
g

h
¼
q2 þ m2 c2
358 APPENDIX D: SOLUTIONS TO PROBLEMS

1.10 Let one of the beams (labelled by 1) refer to the ‘beam’ and let the other beam
(labelled by 2) refer to the ‘target’. Then in Equation (1.43), nb ¼ nN1 =2RA and
vi ¼ 2R=T, where R is the radius of the circular path. Thus the flux is
J ¼ nb vi ¼ nN1 f =A, where f is the frequency. Also N ¼ N2 , so finally the luminosity
is L ¼ JN ¼ nN1 N2 f =A.

1.11 From Equation (1.44c),  ¼ WMA =Ið tÞNA . Since the scattering is isotropic, the
total number of protons emitted from the target is W ¼ 20  ð4=2  103 Þ
¼ 1:25  105 s1 . I can be calculated from the current, noting that the -particles
carry two units of charge, and is I ¼ 3:13  1010 s1 . The density of the target is
t ¼ 1 mg cm2 ¼ 1032 kg fm2 . Putting everything together gives  ¼ 161 mb.

Chapter 2
2.1 From Equation (2.21),

ðr
ðr 1
4

h
Fðq2 Þ ¼ r sin bðrÞdr 4 r2 dr ¼ 3½sin bðaÞ  bðaÞ cos bðaÞb3 ;
q 0 0

where bðrÞ ¼ qr=

h. To evaluate this we need to find a and q. For the latter, we have

1
from which q ¼ 2p sinð#=2Þ ¼ 57:5 MeV=c. Also, we know that a ¼ 1:21A3 fm and
so for A ¼ 56, a ¼ 4:63 fm and qa=
h ¼ 1:35 radians. Finally, using this in the
integral, gives F ¼ 0:829 and hence the reduction is F 2 ¼ 0:69.

2.2 Setting q ¼ jqj in Equation (2.26), we have

ð X1 
2 1 1 iqr cos  n 3
Fðq Þ ¼ f ðxÞ d x:
Ze n¼0
n!

h

Using d3 x ¼ r2 d cos  d and doing the  integral, gives

ðð

2 2 2 iqr cos  q2 r2 cos2 
Fðq Þ ¼ f ðrÞr 1 þ  þ . . . dr d cos 
Ze h
h2


ð
1 ð
1
4 2 4q2
¼ f ðrÞr dr  f ðrÞr4 dr þ . . .
Ze h2
6Ze

0 0
 CHAPTER 2 359
Ð
1
However, from Equation (2.17), Z e ¼ 4 f ðrÞ r 2 dr and from Equation (2.25),
  Ð
1
q2
0 2 
Ze r 2 ¼ 4 f ðrÞ r 4 dr so Fðq2 Þ ¼ 1  6

h2
r þ 
0

 
2.3 From Equation (2.28), r2 ¼ 6
h2 ½1  Fðq2 Þ=q2 , where q ¼ 2E sinð=2Þ: Thus,
pffiffiffiffiffiffiffiffi
q ¼ 43:6 MeV=c. Also, F 2 ¼ 0:65 and so hr 2 i ¼ 6:56 fm.

2.4 The charge distribution is spherical, so the angular integrations in the general result
of Equation (2.17) may be done, giving

21 32 1 31
ð ð
Fðq2 Þ ¼ 4 ðrÞ½sinðqr=
hÞ4r2 dr54 ðrÞ4r2 dr5 :
hÞ=ðqr=

0 0

Ð
1
Substituting for ðrÞ, setting x ¼ r=a and using x expðxÞ dx ¼ 1, gives, after
integrating by parts (twice), 0

1
ð qax


h 1
2
Fðq Þ ¼ ex sin dx ¼  :
qa

h h2
1 þ q2 a2 =

0

 
2.5 In 1 g of the isotope there are initially N0 ¼ 1 g=208  1:66  1024 g . Thus
N0 ¼ 2:9  1021 atoms. At time t there are NðtÞ ¼ N0 et= atoms, where  is the
mean life of the isotope. Thus, provided t  , the average decay rate is

N0  NðtÞ N0 75
 ¼ h1 :
t  0:1  24

Thus,  ¼ 2:4N0 =75 h  1016 years:

2.6 The count rate is proportional to the number of 14 C atoms present in the sample.
If we assume that the abundance of 14 C has not changed with time, the artefact
was made from living material and is predominantly carbon, then at the time it
was made ðt ¼ 0Þ, 1 g would have contained 5  1022 carbon atoms of which
N0 ¼ 6  1010 would have been 14 C. Thus the average count rate would have been
N0 = ¼ 13:8 m1 . At time t, the number of 14 C atoms would be NðtÞ ¼ N0 expðt=Þ
and NðtÞ=N0 ¼ et= ¼ 2:1=13:8, from which t ¼  ln 6:57 ¼ 1:56  104 years. The
artefact is approximately 16 000 years old.

2.7 If the transition rate for 212 208

86 Rn decay is !1 and that for 84 Po is !2 and if the numbers of
each of these atoms at time t is N1 ðtÞ and N2 ðtÞ , respectively, then the decays are
governed by Equation (2.43), i.e. N2 ðtÞ ¼ !1 N1 ð0Þ ½expð!1 tÞexpð!2 tÞ½!2 !1 1 .
The latter is a maximum when dN2 ðtÞ=dt ¼ 0, i.e. when !2 expð!2 tÞ ¼ !1 expð!1 tÞ,
with tmax ¼ lnð!1 =!2 Þð!1  !2 Þ1 . Using !1 ¼ 4:12  102 min1 and
!2 ¼ 6:58  107 min1 , gives tmax ¼ 265 min.
360 APPENDIX D: SOLUTIONS TO PROBLEMS

138
2.8 The total decay rate of both modes of 57 La is
1 1
2
ð1 þ 0:5Þ  ð7:8  10 Þ kg s ¼ 1:17  103 kg1 s1 :
Also, since this isotope is only 0.09 per cent of natural lanthanum, the number
4
of 138 23
57 La atoms per kg is N ¼ ð9  10 Þ  ð1000=138:91Þ  ð6:022  10 Þ, i.e.
N ¼ 3:90  1021 kg1 . The rate of decays is dN=dt ¼ ! N, where ! is the
transition rate, and in terms of this the mean lifetime  ¼ 1=!. Thus,

N 3:90  1021
¼ ¼ s ¼ 3:33  1018 s ¼ 1:06  1011 years:
dN=dt 1:17  103

2.9 The energy released is the increase in binding energy. Now from the SEMF,
Equations (2.46)–(2.52),

ð35Þ2 ð87  70Þ2

BEð35; 87Þ ¼ av ð87Þ  as ð87Þ2=3  ac 1=3
 aa ;
ð87Þ 348
ð57Þ2 ð145  114Þ2
BEð57; 145Þ ¼ av ð145Þ  as ð145Þ2=3  ac 1=3
 aa ;
ð145Þ 580
ð92Þ2 ð235  184Þ2
BEð92; 235Þ ¼ av ð235Þ  as ð235Þ2=3  ac  aa :
ð235Þ1=3 940

The energy released is thus

E ¼ BEð35; 87Þ þ BEð57; 145Þ  BEð92; 235Þ

¼ 3 av  9:153 as þ 476:7ac þ 0:280 aa
which using the values given in Equation (2.54) gives E ¼ 154 MeV.

2.10 The most stable nucleus for fixed A has a Z-value given by Z ¼ =2
, where
from Equation (2.58),  ¼ aa þ ðMn  Mp  me Þ and
¼ aa =A þ ac =ðAÞ1=3 .
Changing  would not change aa , but would effect the Coulomb coefficient
because ac is proportional to . For A ¼ 111, using the value of aa from Equation
(2.54) gives  ¼ 93:93 MeV=c2 and
¼ 0:839 þ 0:208 ac MeV=c2 . For Z ¼ 47,
ac ¼ 0:770 MeV=c2 . This is a change of about 10 per cent from the value given in
Equation (2.54) and so  would have to change by the same percentage.

2.11 In the rest frame of the 269

108 Hs nucleus, m v ¼ mSg vSg . The
 ratio of thekinetic energies
is ESg =E ¼ m =mSg and the total kinetic energy is E 1 þ m =mSg ¼ 9:370 MeV.
Thus, mHs c2 ¼ ðmSg þ m Þc2 þ 9:370 MeV ¼ 269:154 u.

2.12 If there are N0 atoms of 238

94 Pu at launch, then after t years the activity of the source
will be AðtÞ ¼ N0 expðt=Þ=, where  is the lifetime. The instantaneous power is
then PðtÞ ¼ AðtÞ  0:05  5:49  1:602  1013 W > 200 W. Substituting the value
given for , gives N0 ¼ 1:88  1025 and hence the weight of 238 94 Pu at launch would
 
1:88  1025 238
have to be at least kg ¼ 7:43 kg.
6:02  1023 1000
 CHAPTER 2 361

2.13 If there were N0 atoms of each isotope at the formation of the planet ðt ¼ 0Þ,
then after time t the numbers of atoms are N205 ðtÞ ¼ N0 expðt=205 Þ and
N204 ðtÞ ¼ N0 expðt=204 Þ, with


N205 ðtÞ 1 1 n205
¼ exp t  ¼ ¼ 2  107 :
N204 ðtÞ 205 204 n204

Now 204  205, so t ¼ 205 lnð2  107 Þ ¼ 2:6  108 years.

2.14 We first calculate the mass difference between ½p þ 46 46

21 Sc and ½n þ 22 Ti. Using the
information given, we have

Mð21;46Þ  ½Mð22;46Þ þ me  ¼ 2:37MeV=c2 and Mn  ðMp þ me Þ ¼ 0:78MeV=c2

and hence ½Mp þ Mð21; 46Þ  ½Mn þ Mð22; 46Þ ¼ 1:59 MeV=c2 . We also need the
mass differences ½M þ Mð20; 43Þ  ½Mn þ Mð22; 46Þ ¼ 0:07 MeV=c2 . We can
now draw the energy level diagram where the centre-of-mass energy of the
resonance is at (see Equation (2.10)) 2:76  ð45=47Þ ¼ 2:64 MeV.

Thus the resonance could be excited in the 43 46

20 Cað; nÞ 22 Ti reaction at an -
particle laboratory energy of 10:7  ð47=43Þ ¼ 11:7 MeV.

2.15 We have dNðtÞ=dt ¼ P  N, from which


dNðtÞ d  t 
Pet ¼ et N þ ¼ Ne :
dt dt

Integrating and using the fact that N ¼ 0 at t ¼ 0 to determine the constant of

integration, gives the required result.
362 APPENDIX D: SOLUTIONS TO PROBLEMS

35
2.16 The number of Cl atoms in 1 g of the natural chloride is
N ¼ 2  0:758  NA =molecular weight ¼ 7:04  1021 :
 
The activity AðtÞ ¼ N ¼ P 1  et  Pt, since t  1. So
AðtÞ AðtÞt1=2
t¼ ¼ :
P ln2    F  N
Substituting AðtÞ ¼ 3  105 Bq and using the other constants given, yields t ¼ 1:55
days.

2.17 At very low energies we may assume the scattering has ‘ ¼ 0 and so in Equation
(1.63) we have j ¼ 12; sn ¼ 12 and su ¼ 0. Thus,
h2 ðn n þ n 
Þ 4


h2 n
max ¼ ¼ ;
q2n 2 =4 q2n 
Therefore, n ¼ q2n max =4
h2 ¼ 0:35  103 eV and 
¼  n ¼ 9:65  103 eV.

Chapter 3
3.1 (a) Forbidden: violates L conservation, because L ð Þ ¼ 1, but L ð þ Þ ¼ 1.

(b) Forbidden: violates electric charge conservation, because Q (left-hand side) ¼ 1,

but Q (right-hand side) ¼ 0.

(c) Forbidden: violates baryon number conservation because B (left-hand side) ¼ 1,

but B (right-hand side) ¼ 0.

(d) Allowed: conserves L ; B ; Q etc. (violates S, but this is allowed because it is a

weak interaction).

3.2 (a) The quark compositions are: D ¼ d
c ; K 0 ¼ d
s ;  ¼ d

u and since the
dominant decay of a c-quark is c ! s, we have
 CHAPTER 3 363

(b) The quark compositions are:  ¼ sud; p ¼ uud and since the dominant decay
of an s-quark is s ! u, we have

3.3 (a) This would be a baryon because B ¼ 1 and the quark composition would be ssb
which is allowed in the quark model.

(b) This would be a meson because B ¼ 0, but would have to have both an
s- and a

b-quark. However, Qð
s þ

bÞ ¼ 2=3, which is incompatible with the quark model
and anyway combinations of two antiquarks are not allowed. Thus this
combination is forbidden.

3.4 ‘Low-lying’ implies that the internal orbital angular momentum between the
quarks is zero. Hence the parity is P ¼ þ and space is symmetric. Since the
Pauli principle requires the overall wavefunction to be antisymmetric under
the interchange of any pair of like quarks, it follows that spin is antisymmetric.
Thus, any pair of like quarks must have antiparallel spins, i.e. be in a spin-0
state.
Consider all possible baryon states qqq, where q ¼ u; d; s. There are six
combinations with a single like pair: uud; uus; ddu; dds; ssu; ssd, with the spin
of (uu) etc. equal to zero. Adding the spin of the third quark leads to six states
þ
with J P ¼ 12 . In principle, there could be six combinations with all three quarks
the same – uuu; ddd; sss – but in practice these do not occur because it is
impossible to arrange all three spins in an antisymmetric way. Finally, there is one
combination where all three quarks are different: uds. Here there are no restric-
tions from the Pauli principle, so for example, the ud pair could have spin-0 or
þ
spin-1. Adding the spin of the s-quark leads to two states with J P ¼ 12 and 1 with
þ
J P ¼ 32 .
þ þ
Collecting the results, gives an octet of J P ¼ 12 states and a singlet J P ¼ 32 state.
This is not what is observed in nature. In Chapter 5 we will see what additional
assumptions have to be made to reproduce the observed spectrum.
364 APPENDIX D: SOLUTIONS TO PROBLEMS

3.5 (a)

(b)

3.6 The ground state mesons all have L ¼ 0 and S ¼ 0. Therefore they all have P ¼ 1.
Only in the case of the neutral pion is their constituent quark and antiquark also
particle and antiparticle. Thus C is only defined for the 0 and is C ¼ 1. For the
excited states, L ¼ 0 still and thus P ¼ 1 as for the ground states. However, the
total spin of the constituent quarks is S ¼ 1 and so for the 0 , the only state for which
C is defined, C ¼ 1.
For the excited states, by definition there is a lower mass configuration with the
same quark flavours. As the mass differences between the excited states and their
ground states is greater than the mass of a pion, they can all decay by the strong
interaction. In the case of the charged pions and kaons and the neutral kaon ground
states, there are no lower mass configurations with the same flavour structure and so
the only possibility is to decay via the weak interaction, with much longer lifetimes.
In the case of 0 decay, the initial state has a total angular momentum of 1 and
since the pions have zero spin, the  final state must have L ¼ 1. While this is
possible for þ , for the case of 0 0 it violates the Pauli Principle and so is
forbidden.

3.7 In the initial state, S ¼ 1 and B ¼ 1. To balance strangeness (conserved in strong

interactions), in the final state SðY  Þ ¼ 2 and to balance baryon number,
 CHAPTER 3 365

BðY  Þ ¼ 1. As charm and beauty for the initial state are both zero, these quantum
numbers are zero for the Y. The quark content is therefore dss. In the decay, the
strangeness of the  is 1 and so strangeness is not conserved. This is therefore a
weak interaction and its lifetime will be in the range 107 –1013 s:

3.8 The quark composition is  ¼ uds, then ðSu þ Sd Þ2 ¼S2u þ S2d þ 2Su  Sd ¼ 2

h2 and
2
hence Su  Sd ¼

h =4. Then, from the general formula given in Equation (3.84),
setting mu ¼ md ¼ m, we have


Su  Sd Sd  Ss þ Su  Ss
M ¼ 2m þ ms þ b þ
m2 mms


Su  Sd S1  S2 þ S1  S3 þ S2  S3  Su  Sd
¼ 2m þ ms þ b þ
m2 mms

h2 =4 from Equation (3.89), gives

which, using S1  S2 þ S1  S3 þ S2  S3 ¼ 3



b 1 4
M ¼ 2m þ ms þ  :
4 m2 mms

3.9 The initial reacton is strong because it conserves all individual quark numbers. The
 decay is weak because strangeness changes by one unit and the same is true for
the decays of the
0 , K þ and K 0 . The decay of the þ is also weak because it
involves neutrinos and finally the decay of the 0 is electromagnetic because only
photons are involved.

3.10 The Feynman diagram is:

The two vertices where the W-boson couples are weak interactions and have
pffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffi
strengths W . The remaining vertex is electromagnetic and has strength EM .
pffiffiffiffiffiffiffiffiffi
So the overall strength of the diagram is W EM .

e ! x Þ ¼ sin2 ð2Þsin2 ½ðm2 c4 ÞL=ð4

3.11 From Equation (3.27a), we have Pð
hcEÞ, which
for maximal mixing ( ¼ =4) gives Pð
e ! x Þ ¼ sin2 ½1:27ðm2 c4 ÞL=E where L
is measured in m, E in MeV and ðm2 c4 Þ in ðeVÞ2 . If Pð
e ! 
e Þ ¼ 0:90  0:10,
then at 95 per cent confidence level, 1:0  Pð
e ! x Þ  0:70 and hence
0:012  ðm2 c4 Þ  0:019ðeVÞ2 .
366 APPENDIX D: SOLUTIONS TO PROBLEMS

3.12 Reactions (a), (d) and (f) conserve all quark numbers individually and hence are
strong interactions. Reaction (e) violates strangeness and is a weak interaction.
Reaction (c) conserves strangeness and involves photons and hence is an electro-
magnetic interaction. Reaction (b) violates both baryon number and electron lepton
number and is therefore forbidden.

3.13 The doublet of S ¼ þ1 mesons ðK þ ; K 0 Þ has isospin I ¼ 12, with I3 ðK þ Þ ¼ 12 and

I3 ðK 0 Þ ¼ 12. The triplet of S ¼ 1 baryons ðþ ; 0 ;  Þ has I ¼ 1, with
I3 ¼ 1; 0; 1 for þ ; 0 and , respectively. Thus ðK þ ; K 0 Þ is analogous to the
ðp; nÞ isospin doublet and ðþ ; 0 ;  Þ is analogous to the ðþ ; 0 ;  Þ isospin
triplet. Hence, by analogy with Equations (3.54a) and (3.54b),
pffiffiffi pffiffiffi
 1 þ 2  0 20 2
Mð p !  K Þ ¼ M3 þ M1 ; Mð p !  K Þ ¼ M3  M1
3 3 3 3

and

Mðþ p ! þ K þ Þ ¼ M3 ;

where M1;3 are the amplitudes for scattering in a pure isospin state I ¼ 12; 32,
respectively. Thus,

ðþ p ! þ K þ Þ : ð p !  K þ Þ : ð p ! 0 K o Þ

1 2
¼ jM3 j2 : jM3 þ 2M1 j2 : jM3  M1 j2 :
9 9

3.14 Under charge symmetry, nðuddÞ Ð pðduuÞ and þ ðud
Þ Ð  ðd
uÞ and since the
strong interaction is approximately charge symmetry, we would expect
ðþ nÞ  ð pÞ at the same energy, with small violations due to electromagnetic
effects and quark mass differences. However, K þ ðu
sÞ and K  ðs

uÞ are not charge
symmetric and so there is no reason why ðK þ nÞ and ðK  pÞ should be equal.

Chapter 4
4.1 In an obvious notation,

2
ECM ¼ ðEe þ Ep Þ2  ðpe c þ pp cÞ2 ¼ ðEe2  p2e c2 Þ  ðEp2  p2p c2 Þ þ 2Ee Ep  2pe  pp c2
¼ m2e c4 þ m2p c4 þ 2Ee Ep  2pe  pp c2

At the energies of the beams, masses may be neglected and so with p ¼ jpj,

2
ECM ¼ 2Ee Ep  2pe pp c2 cosð  Þ ¼ 2Ee Ep ½1  cosð  Þ;

where  is the crossing angle. Using the values given, gives ECM ¼ 154 GeV. In a
2
fixed-target experiment, and again neglecting masses, ECM ¼ 2Ee Ep  2pe  pp c2 ,
 CHAPTER 4 367
 1=2
where Ee ¼ EL ; Ep ¼ mp c2 ; pp ¼ 0. Thus, ECM ¼ 2mp c2 EL and for
ECM ¼ 154 GeV, this gives EL ¼ 1:26  104 GeV.

4.2 For constant acceleration, the ions must travel the length of the drift tube in half a
cycle of the rf field. Thus, L ¼ v=2 f , where v is the velocity of the ion. Since the
energy is far less than the rest mass of the ion, we can use non-relativistic kinematics
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
to find v, i.e. v ¼ c 200=ð12  931:5Þ ¼ 4:01 107 m s1 and finally L ¼ 1 m.

4.3 A particle with mass m, charge q and speed v moving in a plane perpendicular to a
constant magnetic field of magnitude B will traverse a circular path with radius of
curvature r ¼ mv=qB and hence the cyclotron frequency is f ¼ v=2r ¼ qB=2m.
At each traversal the particle will receive energy from the rf field, so if f is kept fixed,
r will increase (i.e. the trajectory will be a spiral). Thus if the final energy is E, the
pffiffiffiffiffiffiffiffiffi
extraction radius will be R ¼ 2mE=qB. To evaluate these expressions we use
q ¼ 2e ¼ 3:2  1019 C, together with B ¼ 0:8 T ¼ 0:45  1030 ðMeV=c2 Þs1 C1
and thus f ¼ 6:15 MHz and R ¼ 62:3 cm.

4.4 A particle with unit charge e and momentum p in the uniform magnetic field B of the
bending magnet will traverse a circular trajectory of radius R, given by p ¼ BR. If B
is in T, R in m and p in GeV/c, then p ¼ 0:3BR. Referring to the figure below, we
have   L=R ¼ 0:3 LB=p and  ¼ s=d ¼ 0:3BLp=p2 . Solving for d using the
data given, gives d ¼ 9:3 m.

erenkov condition is n  1. So, for the pion to give a signal, but not the
4.5 The C
kaon, we have  n  1  K n. The momentum is given by p ¼ mv
where
1=2

¼ ð1  v2 =c2 Þ , so eliminating
gives  ¼ v=c ¼ ð1 þ m2 c2 =p2 Þ1=2 . For
p ¼ 20 GeV=c, m ¼ 0:14 GeV=c2 and mK ¼ 0:49 GeV=c2 ,  ¼ 0:99997 and
K ¼ 0:99970, so the condition on the refractive index is 3  104  ðn  1Þ=n
 3  105 . Using the largest value of n ¼ 1:0003, we have
 
1 1 1
N ¼ 2 1  2 2 
 n 1 2
as the number of photons radiated per metre, where 1 ¼ 400 nm and 2 ¼ 700 nm.
Numerically, N ¼ 26:5 photons/m and hence to obtain 200 photons requires a
detector of length 7.5 m. (You could also use
 
1 2  1
N ¼ 2 1  2 2
 n 2
368 APPENDIX D: SOLUTIONS TO PROBLEMS

where  is the mean of 1 and 2 , which would give 24.5 photons/m and a length of
8.2 m.)

4.6 Luminosity may be calculated from the formula for colliders, L ¼ n N1 N2 f =A,
where n is the number of bunches, N1 and N2 are the numbers of
particles in each bunch, A is the cross-sectional area of the beam and f is its
frequency. We have, n ¼ 12; N1 ¼ N2 ¼ 3  1011 ; A ¼ ð0:02  102 Þ cm2 and
f ¼ ð3  1010=8  105 Þ s1 , so finally L ¼ 6:44  1031 cm2 s1.

4.7 (a) The b quarks are not seen directly but, instead, they fragment (hadronize) to B-
hadrons, i.e. hadrons containing b quarks. So one characteristic is the presence of
hadrons with non-zero beauty quantum numbers. As these hadrons are unstable
and the dominant decay of b-quarks is to c-quarks, a second characteristic is the
presence of hadrons with non-zero values of the charm quantum number.
We need to observe the point where the eþ e collision occurred and the point
of origin of the decay products of the B-hadrons. The difference between these
two is due to the lifetime of the B-hadrons. As the difference will be very small,
precise position measurements are required. The daughter particles may be
detected using a silicon micro-vertex detector and an MWPC. In addition, any
electrons from the decays could be detected by an MWPC or an electromagnetic
calorimeter. The same is true for muons in the decay products, except they are
not readily detected in the calorimeter as they are very penetrating. However, if
one places an MWPC behind a hadron calorimeter then one can be fairly
confident that any particle detected is a muon, as everything else (except
neutrinos) will have been stopped in the calorimeter.

(b) In the electronic decay mode, the electron can be measured in both a MWPC
and an EM calorimeter. For high energies the better measurement is made in the
calorimeter. The neutrino does not interact unless there is a very large mass of
material (thousands of tons) and so its presence must be inferred by imposing
conservation of energy and momentum. In a colliding beam machine, the original
colliding particles have zero transverse momentum and a fixed energy. If one
adds up all the energy and momentum of all the final-state particles, then
any imbalance compared to the initial system can be attributed to the neutrino.

For the muonic mode, the muon can be measured in the MWPC but cannot be
measured well in the calorimeter because it only ionizes to a very small extent.
Since the muons only interact to a small extent they (along with neutrinos) are
generally the only particles that emerge from a hadronic calorimeter. So if one
registers a signal in a small MWPC placed behind a calorimeter then one can be
confident that the particle is a muon.

4.8 To be detected, the event must have 150 <  < 30 , i.e. jcos j < 0:866. Setting
x ¼ cos , the fraction of events in this range is

ð
þ0:866  þ1:0
ð 
d d  þ0:866  þ1:0
f ¼ dx dx ¼ x þ x3 =3 0:866 x þ x3 =3 1:0 ¼ 0:812:
dx dx
0:866 1:0
 CHAPTER 4 369

The total cross-section is given by

ð ð
2 ð
þ1 ð
þ1
d d h2 c2
2
 
¼ d ¼ d d cos  ¼ 2 2
1 þ cos2  d cos :
d d 4Ecm
0 1 1

Using Ecm ¼ 10 GeV, gives  ¼ 42
h2 c2 =3Ecm

2
¼ 0:866 nb. The rate of production
of events is given by L and since L is a constant, the total number of events
produced will be L t ¼ 86 600:
The   decay too quickly to leave a visible track in the drift chamber. The eþ and
the  will leave tracks in the drift chamber and the eþ will produce a shower in
the electromagnetic calorimeter. If it has enough energy, the  will pass through the
calorimeters and leave a signal in the muon chamber. There will be no signal in the
hadronic calorimeter.

4.9 Referring to the figure below, the distance between two positions of the particle t
apart in time is v t. The wave fronts from these two positions have a difference in
their distance travelled of c t=n.

These constructively interfere at an angle , where

c t=n 1
cos  ¼ ¼ :
v t n

The maximum value of  corresponds to the minimum of cos  and hence the
maximum of . This occurs as  ! 1, when max ¼ cos1 ð1=nÞ. This value occurs
in the ultra-relativistic or massless limit.
1=2
The quantity  may be expressed as  ¼ pc=E ¼ pc½p2 c2 þ m2 c4  . Hence,
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 p 2 c 2 þ m2 c 4
cos  ¼ ;
n pc

which rearranging, gives x ðmc2 Þ2 ¼ p2 c2 ðn2 cos2   1Þ. Differentiating this

formula gives dx=d ¼ 2p2 c2 n2 cos  sin  and the error on x is then given by
370 APPENDIX D: SOLUTIONS TO PROBLEMS

x ¼ jdx=dj  . For very relativistic particles, the derivative

pffiffiffiffiffiffiffiffiffiffiffiffiffi can be approximated
by using max , for which cos max ¼ 1=n, sin max ¼ n2  1=n. Hence
pffiffiffiffiffiffiffiffiffiffiffiffiffi
1 n2  1 pffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2 2
x  2p c n  ¼ 2p2 c2 n2  1  :
n n

4.10 The average distance between collisions of a neutrino and an iron nucleus is the
mean free path  ¼ 1=n , where n  =mp c2 is the number of nucleons per cm3 .
Using the data given, n  4:7  1024 cm3 and   3  1036 cm2 , so that
  7:1  1010 cm. Thus if 1 in 109 neutrinos is to interact, the thickness of iron
required is 71 cm.

4.11 Radiation energy losses are given by dE=dx ¼ E=LR, where LR is the radiation
length. This implies that E ¼ E0 expðx=LR Þ, where E0 is the initial energy. Using
E0 ¼ 2 GeV, LR ¼ 36:1 cm, x ¼ 10 cm, gives E ¼ 1:51 GeV. Radiation losses at
fixed E are proportional to m2 , where m is the mass of the projectile. Thus for
muons, they are negligible at this energy.

4.12 The total cross section is tot ¼ el þ cap þ f ¼ 4  102 b and the attenuation is
expðnxtot Þ where nx ¼ 101 NA =A ¼ 2:56  1023 m2 . Thus expðnxtot Þ ¼ 0:9898,
i.e 1.02 per cent of the incident particles interact and of these the fraction
that elastically scatter is given by the ratio of the cross-sections, i.e.
3  102 =4  102 ¼ 0:75  104 . Thus the intensity of elastically-scattered neu-
trons is 0:75  104  0:0102  106 ¼ 0:765 s1 and finally the flux at 5 m is
0:765=ð4    52 Þ ¼ 2:44  103 m2 s1 .
1
4.13 The total centre-of-mass energy is given by ECM  ð2mc2 EL Þ2 ¼ 0:23 GeV and so
the cross-section is  ¼ 1:64  1034 m2 . The interaction length is ‘ ¼ 1=n, where
n is the number density of electrons in the target. This is given by n ¼ NA Z=A,
where NA is Avogadro’s number and for lead, ¼ 1:14  107 kg m3 is the density,
Z ¼ 82 and A ¼ 208. Thus n ¼ 2:7  1033 m3 and ‘ ¼ 2:3 m.

4.14 The target contains n ¼ 1:07  1025 protons and so the total number of interactions
per second is N ¼ n  flux  tot ¼ ð1:07  1025 Þ  ð2  107 Þ  ð40  1031 Þ ¼
856 s1 . There are thus 856 photons/s produced from the target.

4.15 For small v, the Bethe–Bloch formula may be written



dE 1 2me v2 dS 2 2me v2
S  / 2 ln with / 3 1  ln :
dx v I dv v I

The latter has a maximum for v2 ¼ eI=2me. Thus for a proton in iron we can use
I ¼ 10Z eV ¼ 260 eV, so that Ep ¼ 12mp v2 ¼ mp Ie=4me ¼ 324 keV.

4.16 From Equation (4.24), EðrÞ ¼ V=r lnðrc =ra Þ and at the surface of the anode this is
0:5=ð20  106 Þ lnð500Þ ¼ 4023 kV m1 . Also, if Ethreshold ðrÞ ¼ 750 kV m1 , then
from Equation (4.24) r ¼ 0:107 mm and so the distance to the anode is 0:087 mm.
 CHAPTER 5 371

This contains 22 mean free paths and so assuming each collision produces an ion
pair, the multiplication factor is 222 ¼ 4:2  106 ¼ 106:6 .

Chapter 5
5.1 We have m ¼  þ  þ
> n ¼ 
þ 
þ

, where the inequality is because baryon
number B > 0. Using the values of the colour charges I3C and Y C from Table 5.1, the
colour charges for the state are:

Þ=2  ð  
Þ=2 and

I3C ¼ ð  
Þ=3 þ ð  
Þ=3  2ð


Þ=3:
Y C ¼ ð  

By colour confinement, both these colour charges must be zero for observable
hadrons, which implies   
¼   
¼


p and hence m  n ¼ 3p, where
p is a non-negative integer. Thus the only combinations allowed by colour
confinement are of the form

ð3qÞp ðq

qÞn ðp; n  0Þ:

It follows that a state with the structure qq is not allowed, as no suitable values of p
and n can be found.

5.2 (a)

(b)
372 APPENDIX D: SOLUTIONS TO PROBLEMS

(c)

5.3 The Feynman diagram is:

The four-momenta are:

PðpÞ ¼ ðE=c; pÞ and pÞ ¼ ðE=c; pÞ;

Pð

with

P2 ¼ m2 c2 ¼ E2 =c2  p2 and m ¼ mp ¼ m
p :

Now PðqÞ ¼ ðxE=c; xpÞ qÞ ¼ ðxE=c; xpÞ with x ¼ 16, so

and Pð

 
2
ECM pÞ2 ¼ x2 2m2 c4 þ 2E2 þ 2p2 c2 :
¼ x2 c2 ½PðpÞ þ Pð

Neglecting the masses of the proton and the antiproton at these energies, gives

E ¼ 3ECM and p ¼ 3  350 ¼ 1050 GeV=c:

5.4 Energy–momentum conservation gives,

2 2
W 2 c4 ¼ ½ðE  E0 Þ þ EP  ½ðp  p0 Þ þ P c2 ¼ invariant mass of X:
 CHAPTER 5 373

Using, Q2 ¼ ðp  p0 Þ2  ðE  E0 Þ2 =c2 and M 2 c4 ¼ EP2  P2 c2 , where M is the mass

of the proton, gives

W 2 c4 ¼ Q2 c2 þ M 2 c4 þ 2EP ðE  E0 Þ  2P  ðp  p0 Þc2 :

Also, 2M W 2 c2 þ Q2  M 2 c2 and so, in the rest frame of the proton
ðP ¼ 0; EP ¼ Mc2 Þ,  ¼ E  E0 .
Since some energy must be transferred to the outgoing electron, it follows that
E  E0 , i.e.   0. Also, since the lightest state X is the proton, W 2  M 2 . Thus,

2M ¼ Q2 þ ðW 2  M 2 Þc2  Q2 :
From the definition of x, it follows that x  1. Finally, x > 0 because both Q2 and
2M are positive.

5.5 In the quark model,  ¼ uds; p ¼ uud; K  ¼ s
u; n ¼ udd and þ ¼ ud
. From the
flavour independence of the strong interaction, we can set ðqqÞ ¼ ðudÞ ¼
ðsdÞ etc. and ðq
qÞ ¼ ðud
Þ ¼ ðs

uÞ etc.. Then ðpÞ ¼ ðppÞ ¼ 9ðqqÞ and
ðK  nÞ ¼ ðþ pÞ ¼ 3ðqqÞ  3ðq
qÞ. The result follows directly.

5.6 By analogy with the QED formula, we have ð3gÞ ¼ 2ð2  9Þ6s mc c2 =9, where
mc  1:5 GeV=c2 is the constituent mass of the c-quark. Evaluating this gives
s ¼ 0:31. In the case of the radiative decay, ðgg
Þ ¼ 2ð2  9Þ4s 2 mb c2 =9,
where mb  4:5 GeV=c2 is the constituent mass of the b-quark. Evaluating this gives
s ¼ 0:32. (These values are a little too large because in practice  is replaced by
4
3s .)

5.7 From Equation (5.38a)


1 4 1
F2‘p ðxÞ ¼ x ðd þ d
Þ þ ðu þ

uÞ þ ðs þ
sÞ
9 9 9

and from Equations (5.38b) and (5.39)


4 1 1
F2‘n ðxÞ ¼ x ðd þ d
Þ þ ðu þ

uÞ þ ðs þ
sÞ ;
9 9 9
so that

ð1 ð1 ð1
  dx 1 1
F2ep ðxÞ  F2en ðxÞ ¼ ½uðxÞ þ

uðxÞ dx  ½dðxÞ þ d
ðxÞdx:
x 3 3
0 0 0

However, summing over all contributions we must recover the quantum numbers of
the proton, i.e.

ð1 ð1
½uðxÞ 

uðxÞ dx ¼ 2; ½dðxÞ  d
ðxÞ dx ¼ 1:
0 0
374 APPENDIX D: SOLUTIONS TO PROBLEMS

Eliminating the integrals over u and d gives the Gottfried sum rule.

5.8 Substituting Equation (5.22) into Equation (5.23) and setting NC ¼ 3, gives
X
R ¼ 3ð1 þ s =Þ e2q ;

where s is given by Equation (5.11) evaluated at Q2 ¼ ECM 2

and the sum is over
those quarks that can be produced in pairs at the energy considered. At 2.8 GeV the
u, d and s quarks can contribute and at 15 GeV the u, d, s, c and b quarks can
contribute. Evaluating R then gives R  2:17 at ECM ¼ 2:8 GeV and R  3:89 at
ECM ¼ 15 GeV. When ECM is above the threshold for t
t production, R rises to
R ¼ 5ð1 þ s =Þ.

5.9 A proton has the valence quark content p ¼ uud. Thus from isospin invariance the u
quarks in the proton carry twice as much momentum as the d quarks, which implies
a ¼ 2b. In addition, we are told that

ð1 ð1
1
xFu ðxÞdx þ xFd ðxÞdx ¼ :
2
0 0

Using the form of the quark distributions with a ¼ 2b gives a ¼ 43 and b ¼ 23.

5.10 The peak value of the cross-section is where E ¼ MW c2 , i.e.

hcÞ2 ð2=MW c2 Þ2 ud
4 ð

ð
hcÞ2
max ¼ ¼ brðW þ ! ud
Þ ¼ 84 nb:
3 3 ðMW c2 Þ2

The required integral is

ð1 ð1
p
p ðsÞ ¼ ud
ðEÞ uðxu Þ dðxd Þ dxu dxd
0 0

where we have used C-invariance to relate the distribution functions for protons and
antiprotons. In the narrow width approximation and using the quark distributions
from Question 5.9,
ð1 ð1 !
ð1  xu Þ3 ð1  xd Þ3 xu s
p
p ðsÞ ¼ C  1 xd dxu dxd
xu xd ðMW c2 Þ2
0 0

where C ð8W max Þ=ð9MW c2 Þ and we have used E2 ¼ xu xd s. Thus,

ð1 3
ð1  xu Þ3 k
p
p ðsÞ ¼ C 1 dxu ;
xu xu
k
 CHAPTER 6 375

where k ðMW c2 Þ=s and the lower limit is because k < xu < 1. The integral yields
 
8 W 2 3 11 2 11 3
p
p ðsÞ ¼  max ð1 þ 9k þ 9k þ k ÞlnðkÞ   9k þ 9k þ k :
9 MW c2 3 3
pffiffi
Evaluating this for s ¼ 1 TeV gives k ¼ 0:0064 and p
p ¼ 9:3 nb, which is about a
factor of two larger than experiment.

Chapter 6
6.1 A charged current weak interaction is one mediated by the exchange of charged W 
boson. A possible example is n ! p þ e þ 
e . A neutral current weak interaction is
one mediated by a neutral Z 0 boson. An example is  þ p !  þ p. Charged
current weak interactions do not conserve the strangeness quantum number, whereas
neutral current weak interactions do. For  þ e !  þ e, the only Feynman
diagram that conserves both Le and L is:

which is a weak neutral current. However, for e þ e ! e þ e, there are two
diagrams:

Thus the reaction has both neutral and charged current components and is not
unambiguous evidence for weak neutral currents.
376 APPENDIX D: SOLUTIONS TO PROBLEMS

6.2 The lowest-order electromagnetic Feynman diagram is

The total cross-section is given by

2ð ð1
1
d 22
h2 c2 1 3
¼ d d cos  ¼ 2
cos  þ cos 
d 4ECM 3 1
0 1
2 2 2
4

h c
¼ 2
¼ 0:44 nb:
3ECM

The lowest-order weak interaction diagram is

With the addition of the weak interaction term,

  
d d d 2
h2 c2  
¼ þ ¼ 2
1 þ Cwk cos  þ cos2  :
d d em d wk 4ECM

Then, using
ð1
 
F ¼ C 1 þ Cwk cos  þ cos2  d cos 
0
and
ð0
 
B ¼ C 1 þ Cwk cos  þ cos2  d cos :
1
2 2 2
2
where C 2

h c =4ECM , gives



4 Cwk 4 Cwk
F ¼ C þ and B ¼ C 
3 2 3 2
 CHAPTER 6 377

and so

Cwk
AFB ¼ ; i:e: 8AFB ¼ 3Cwk :
2ð4=3Þ

6.3 The Feynman diagram is

The amplitude has two factors of the weak coupling gW and one W propagator
carrying a momentum q, i.e.

g2W g2
amplitude / 2
/ W2 ;
q2 c2  MW c 4 MW

because qc  M c2  MW c2 . Now, ð ! p Þ / ðamplitudeÞ2 / g4W =MW 4

and so
doubling gW and reducing MW by a factor of four will increase the rate by a factor
½24 =½ð1=4Þ4  ¼ 4096:

6.4 The most probable energy is given by

 
d d! 2G2F m2 4Ee2
¼ 0; which gives 2Ee  ¼ 0; i:e Ee ¼ m c2 =2:
dEe dEe ð2Þ3 ð

hcÞ6 m c 2

When Ee  m c2 =2, the electron has its maximum energy and the two neutrinos
must be recoiling in the opposite direction. Only left-handed particles (and right-
handed antiparticles) are produced in weak interactions. Since the masses of all
particles are neglected, states of definite handiness are also states of definite helicity,
so the orientations of the momenta and spins are therefore as shown:
378 APPENDIX D: SOLUTIONS TO PROBLEMS

Integrating the spectrum gives

c2 =2

m ð 
2G2F ðm c2 Þ2 4Ee3 G2F ðm c2 Þ5
¼ Ee2  dE e ¼ :
ð2Þ3 ð

hcÞ6 3m c2 1923 ð
hcÞ6
0

Numerically,   3:0  1019 GeV, which gives a lifetime  ¼

h=  2:2  106 s.

6.5 (a) In addition to the decay b ! c þ e þ 
e, there are two other leptonic decays
ð‘ ¼  ;   Þ and by lepton universality they will all have equal decay rates.
There are also hadronic decays of the form b ! c þ X where QðXÞ ¼ 1.
Examining the allowed Wq
q vertices using lepton–quark symmetry shows that
the only forms that X can have, if we ignore Cabibbo-suppressed modes, are
d
u and s
c. Each of these hadronic decays has a probability three times that
of a leptonic decay because the quarks exist in three colour states. Thus,
there are effectively six hadronic channels and three leptonic ones. So finally,
BRðb ! c þ e þ 
e Þ ¼ 19.

(b) The argument is similar to that of (a) above. Thus, in addition to the decay
  ! e þ 
e þ  , there is also the leptonic decay   !  þ 
þ  with
equal probability and the hadronic decays   !  þ X. In principle, X ¼ d
u
and s
c, but the latter is not allowed because ms þ mc > m . So the only allowed
hadronic decay is   ! d þ
u þ  with a relative probability of three because
of colour. So finally, BRð  ! e þ 
e þ  Þ ¼ 15. (The measured rate is 0.18,
but we have neglected kinematic corrections.)

6.6 For neutrinos, gR ðÞ ¼ 0; gL ðÞ ¼ 12. So, e ¼  ¼  ¼ 0 =4, where

GF MZ3 c6
0 ¼ pffiffiffi ¼ 668 MeV:
3 2ð
hcÞ3

Thus the partial width for decay to neutrino pairs is  ¼ 501 MeV. For quarks,
gR ðu; c; tÞ ¼ 16 and gL ðu; c; tÞ ¼ 13. Thus, u ¼ c ¼ 10
720 . Also, gR ðd;
P s; bÞ ¼ 12
1
5
and gL ðb; s; dÞ ¼ 12 . Thus, d ¼ s ¼ b ¼ 13 
72 0 . Finally,  q ¼ i , where
i ¼ u; c; d; s; b – no top quark because 2Mt > MZ . So, i


3  13 2  10 59
q ¼ þ 0 ¼ 0 ¼ 547 MeV:
72 72 72

Hadron production is assumed to be equivalent to the production of q
q pairs

followed by fragmentation with probability unity. Thus hadron ¼ 3q , where the
factor of three is because each quark exists in one of three colour states. Thus
hadron ¼ 1641 MeV.
 CHAPTER 6 379

If there are N generations of neutrinos with M < MZ =2, so that Z 0 !  
is

allowed, then tot ¼ had þ lep þ N 
where 
is the width to a specific  

pair. Thus
tot  had  lep ð2490  7Þ  ð1738  12Þ  ð250  2Þ
N ¼ ¼

167
¼ 3:01  0:05;

which rules out values of N greater than 3.

u; þ ¼ ud
. Since preferentially

u; K  ¼ s

6.7 The quark compositions are: D0 ¼ c

c ! s, we have

i.e. a lowest-order charge current weak interaction. However, for Dþ ! K 0 þ þ,

we have Dþ ¼ cd
; K 0 ¼ d
s ; þ ¼ ud
. Thus we could arrange c ! d via W
emission and the W þ could then decay to ud
, i.e. þ . However, this would leave the
d
quark in the Dþ to decay to an
s quark in the K 0 which is not possible as they both
have the same charge.

6.8 The relevant Feynman diagrams are:

pffiffiffiffiffiffiffi
In the case of the charged pion, there are two vertices of strength W , and there
will be a propagator
1 1
2 c2
 2 2;
Q2 þ MW MW c
380 APPENDIX D: SOLUTIONS TO PROBLEMS

because the momentum transfer (squared) Q2 carried by the W is very small. Thus
the decay rate will be proportional to
pffiffiffiffiffiffiffipffiffiffiffiffiffiffi2
W W 2
2
¼ W
4
:
MW MW
pffiffiffiffiffiffiffiffi
In the case of the neutral pion, there are two vertices of strength em , but no
2
propagator. Thus the decay rate will be proportional to em and since em  W , the
decay rate for the charged pion will be much smaller than that for the neutral decay,
i.e. the lifetime of the 0 will be much shorter.

6.9 The two Feynman diagrams are:

Using lepton–quark symmetry and the Cabibbo hypothesis, the two hadron vertices
are given by gudW ¼ gW cos C and gusW ¼ gW sin C . So, if we ignore kinematic
differences and spin effects, we would expect the ratio of decay rates is given by

Rate ðK  !  þ 
Þ g2usW
R¼ / ¼ tan2 C  0:05
Rate ð !  þ 
Þ g2udW

The measured ratio is actually about 1.3, which shows the importance of the
neglected effects. For example, the Q-value for the kaon decay is almost 20 times
that for pion decay.

6.10 To a first approximation the difference in the two decay rates is due to two effects.
First,  ! n þ e þ 
e has jSj ¼ 1 and hence is proportional to sin2 C , where
C is the Cabbibo angle, whereas  !  þ e þ 
e has jSj ¼ 0 and is propor-
tional to cos2 C . Secondly, the Q-values are different for the two reactions. Thus,
using Sargent’s Rule,
 5
sin2 C Qn 5 257
R  0:053 ¼ 17:0:
cos2 C Q 81

(The experimental value is 17.8.) Whereas,  ! n þ e þ 
e is a first-order weak

interaction, no Feynman diagram with a single W-boson exchanged can be drawn for
þ ! n þ eþ þ e (try it), i.e. it is higher-order and hence very heavily suppressed –
in practice not seen.
 CHAPTER 6 381

6.11 The required number of events produced must be 20 000, taking account of the
detection efficiency. If the cross-section is 60 fb ¼ 6  1038 cm2 , then the inte-
grated luminosity required is 2  104 =6  1038 ¼ ð1=3Þ  1042 cm2 and hence
the instantaneous luminosity must be 3:3  1034 cm2 s1 .
The branching ratio for Z 0 ! b
b is found from the partial widths to be 15 per cent.
Thus, if b quarks are detected, the much greater branching ratio for H ! b
b will
help distinguish this decay from the background of Z 0 ! b
b.

6.12 By ‘adding’ an I ¼ 12 particle to the initial state we can assume isospin invariance
holds. Consider
 þ S0 !  þ . The final state is jI ¼ 1; I3 ¼ 1i and so is the
initial state because I3 ðS0 Þ ¼ 12. Thus the transition is pure I ¼ 1 and the rate is
jM1 j2 . For
0 þ S0 !  þ 0, the final state is again pure I ¼ 1 but with I3 ¼ 0.
However, the initial state is an equal mixture of I ¼ 0 and I ¼ 1, i.e.
  0

S ¼ p1ffiffiffijI ¼ 1; I3 ¼ 0i  p1ffiffiffijI ¼ 0; I3 ¼ 0i
2 2
and so the rate is 12jM1 j2 . Thus R ¼ 2: (The measured value is about 1.8.)

6.13 Integrating the differential cross-sections over y (from 0 to 1) gives for a spin-12 target
with a specific quark distribution
21 32 1 31
ð ð
NC ðÞ 4 2 2 2 54 dy5 ¼ g2 þ 1g2
¼ ½gL þ gR ð1  yÞ  dy L
CC ðÞ 3 R
0 0

and
21 32 1 31
ð ð
NC ð

Þ 4 2
¼ ½gL ð1  yÞ2 þ g2R  dy54 ð1  yÞ2 dy5 ¼ g2L þ 3g2R :
CC ð

Þ
0 0

For an isoscalar target, we must add the contributions for u and d quarks in equal
amounts, i.e.

NC ðÞ 1 1
ðisoscalarÞ ¼ g2L ðuÞ þ g2R ðuÞ þ g2L ðdÞ þ g2R ðdÞ
CC ðÞ 3 3

and

NC ð

Þ
ðisoscalarÞ ¼ g2L ðuÞ þ 3g2R ðuÞ þ g2L ðdÞ þ 3g2R ðdÞ:
CC ð

Þ

Substituting for the couplings finally gives for an isoscalar target

NC ðÞ 1 2 20 4 NC ð

Þ 1 20
¼  sin W þ sin W ; ¼  sin2 W þ sin4 W :
CC ðÞ 2 27 CC ð

Þ 2 9
382 APPENDIX D: SOLUTIONS TO PROBLEMS

Chapter 7
7.1 For the 73 Li nucleus, Z ¼ 3 and N ¼ 4. Hence the configuration is
 2  1  2  2
protons: 1s1=2 1p3=2 ; neutrons: 1s1=2 1p3=2 :

By the pairing hypothesis, the two neutrons in the 1p3=2 sub-shell will have a total
orbital angular momentum and spin L ¼ S ¼ 0 and hence J ¼ 0. Therefore they will
not contribute to the overall nuclear spin, parity or magnetic moment. These will be
determined by the quantum numbers of the unpaired proton in the 1p3=2 sub-shell.

This has J ¼ 32 and ‘ ¼ 1, hence for the spin-parity we have J P ¼ 32 . The magnetic
moment is given by
1
¼ j gproton ¼ j þ 2:3 ðsince j ¼ ‘ þ Þ ¼ 1:5 þ 2:3
2
¼ 3:8 nuclear magnetons:

If only protons are excited, the two most likely excited states are:
 2  1  2  2
protons: 1s1=2 1p1=2 ; neutrons: 1s1=2 1p3=2 ;

which corresponds to exciting a proton from the p3=2 sub-shell to the p1=2 sub-shell,
and

 1  2  2  2
protons: 1s1=2 1p3=2 ; neutrons: 1s1=2 1p3=2 ;

which corresponds to exciting a proton from the s1=2 sub-shell to the p3=2 sub-shell.

7.2 A state with quantum number jð¼ ‘  12Þ can contain a maximum number
Nj ¼ 2ð2j þ 1Þ nucleons. Therefore, if Nj ¼ 16 it follows that j ¼ 72 and ‘ ¼ 3 or
4. However, we know that the parity is odd and since P ¼ ð1Þ‘ , it follows that
‘ ¼ 3.

7.3 The configuration of the ground state is

protons: ð1s1=2 Þ2 ð1p3=2 Þ4 ð1p1=2 Þ2 ð1d5=2 Þ;

neutrons: ð1s1=2 Þ2 ð1p3=2 Þ4 ð1p1=2 Þ2 :

To get jP ¼ 12 , one could promote a p1=2 proton to the d5=2 shell, giving

protons: ð1s1=2 Þ2 ð1p3=2 Þ4 ð1p1=2 Þ1 ð1d5=2 Þ2 :

Then by the pairing hypothesis, the two d5=2 protons could combine to give jP ¼ 0þ ,
so that the total spin-parity would be determined by the unpaired p1=2 neutron, i.e.

jP ¼ 12 . Alternatively, one of the p3=2 protons could be promoted to the d5=2 shell,
giving

protons: ð1s1=2 Þ2 ð1p3=2 Þ1 ð1p1=2 Þ2 ð1d5=2 Þ2

 CHAPTER 7 383

and the two d5=2 protons could combine to give jP ¼ 2þ , so that when this combines
 
with the single unpaired jP ¼ 32 proton the overall spin-parity is jP ¼ 12 . There are
many other possibilities.

7.4 For 93
41 Nb, Z ¼ 41 and N ¼ 52. From the filling diagram Figure 7.4, the configuration
is predicted to be:

proton: . . . ð2p3=2 Þ4 ð1f5=2 Þ6 ð2p1=2 Þ2 ð1g9=2 Þ1 ; neutron : . . . ð2d5=2 Þ2 :

þ
So ‘ ¼ 4, j ¼ 92 ) jP ¼ 92 (which agrees with experiment). The magnetic dipole
moment follows from the expression for jproton in Equations (7.31) with j ¼ ‘ þ 12, i.e.
¼ ð j þ 2:3Þ N ¼ 6:8 N . (The measured value is 6:17 N .)
For 33
16 S, Z ¼ 16 and N ¼ 17. From the filling diagram Figure 7.4, the configura-
tion is predicted to be:

proton    ð1d5=2 Þ6 ð2s1=2 Þ2 ; neutron:    ð1d5=2 Þ7 ð2s1=2 Þ2 ð1d3=2 Þ1 :

þ
So ‘ ¼ 2; j ¼ 32 ) jP ¼ 32 (which agrees with experiment). The magnetic dipole
moment follows from the expression for jneutron in Equations (7.31) with j ¼ ‘  12,
i.e. ¼ ð1:9jÞ=ð j þ 1Þ N ¼ 1:14 N . (The measured value is 0:64 N .)

7.5 From Equation (7.32),

ð
eQ ¼ ð2z2  x2  y2 Þd

with ¼ Ze=ð43b2 aÞ and the integral is through the volume of the spheroid
ðx2 þ y2 Þ=b2 þ z2 =a2  1. The integral can be transformed to one over the volume
of a sphere by the transformations x ¼ bx0 ; y ¼ by0 and z ¼ az0 . Then
ððð
3Z
Q¼ dx0 dy0 dz0 ð2a2 z02  b2 x02  b2 y02 Þ:
4

But
ððð ð1
02 0 0 1 02 0 4
x dx dy dz ði:e: z Þ ¼ r 4r02 dr0 ¼ ;
3 15
0

and similarly for the other integrals. Thus, by direct substitution, Q ¼ 25Zða2  b2 Þ.

7.6 From Question 7.5 we have Q ¼ 25Zeða2  b2 Þ and using Z ¼ 67 this gives
a2  b2 ¼ 13:1 fm2 . Also, from Equation (2.32) we have A ¼ 43ab2 , where
¼ 0:17 fm3 is the nuclear density. Thus, ab2 ¼ 231:7 fm3 . The solution of
these two equations gives a  6:85 fm and b  5:82 fm.

7.7 From Equation (7.53), t1=2 ¼ ln2= ¼ CR ln2 expðGÞ, where C is a constant
formed from the frequency and the probability of forming -particles in the nucleus.
384 APPENDIX D: SOLUTIONS TO PROBLEMS

Thus t1=2 ðThÞ ¼ t1=2 ðCfÞ exp½GðThÞ  GðCfÞ. The Gamow factors may be calcu-
lated from the data given. Some intermediate quantities are: rC ¼ 45:96 fm (Th);
37.72 fm (Cf); R ¼ 9:268 fm (Th); 9.439 fm (Cf) (using R ¼ 1:21 ðA1=3 þ 41=3 Þ and
recalling that ðZ; AÞ refer to the daughter nucleus). These give G ¼ 66:5 (Th); 54.9 (Cf)
and t1=2 ðThÞ ¼ e11:6 t1=2 ðCfÞ ¼ 4:0 years. (The measured value is 1.9 years).
þ
7.8 The J P values of the 0 and the  are both 12 (see Chapter 3), so the photon has
L ¼ 1 and as there is no change of parity the decay proceeds via an M1 transition.
þ
The 0 has J P ¼ 32 and again there is no parity change. Therefore both M1 and E2
multipoles could be involved, with M1 dominant (see Section 7.8.2). If we assume
that the reduced transition probabilities are equal in the two cases, then from
Equations (7.80), in an obvious notation,

3
E
ð0 Þ
ð0 Þ ¼ ð0 Þ;
E
ð0 Þ

i.e. ð0 Þ ¼ ð292=77Þ3  ð0:6  1023 Þ=0:0056 ¼ 5:8  1020 s (the measured
value is ð7:4  0:7Þ  1020 s).

7.9 In the centre-of-mass system, the threshold for 34 S þ p ! n þ 34 Cl is

6:45  ð34=35Þ ¼ 6:27 MeV. Correcting for the neutron–proton mass difference
gives the Cl–S mass difference as 5.49 MeV and since in the positron decay
34
Cl ! 34 S þ eþ þe , Q ¼ MðA; ZÞ  MðA; Z  1Þ  2me , the maximum positron
energy is 4.47 MeV.

7.10 From Equation (7.71) the electron energy spectrum may be written IðEÞ ¼ AE1=2 
ðE0  EÞ2 , where E is the electron energy, E0 is the end-point, A is a constant and we
have neglected the Fermi screening correction and set the neutrino mass to be zero.
We need to calculate the fraction
2 32 31
ð0
E Eð0
6 74
F 4 IðEÞ dE5 IðEÞ dE5
E0  0

Ð 
where  is a small quantity. Using x1=2 ða  xÞ2 dx ¼ 12a2 x2  23ax3 þ 14x4 1=2 , gives,
using E0 ¼ 18:6  103 eV and  ¼ 10 eV, F ¼ 3:1  1010 .

is defined by
7.11 The mean energy E
2E 32E 31
ð0 ð0

4 E d!ðEÞ54 d!ðEÞ5 :
E
0 0

The integrals are:

ð
2 5=2  
E3=2 ðE0  EÞ2 dE ¼ E 63E02  90E0 E þ 35E2
315
 CHAPTER 8 385

and
ð
2 3=2  
E1=2 ðE0  EÞ2 dE ¼ E 35E02  42E0 E þ 15E2 :
105

¼ 1E0 , as required.
Substituting the limits gives E 3

7.12 The possible transitions are as follows:

Initial Final L P Multipoles

3 5
2 2 1, 2, 3, 4 No M1, E2, M3, . . .
3 1
2 2 1, 2 No M1, E2
5 1
2 2 2, 3 No E2, M3

From Figure 7.13, the dominant multipole for a fixed transition energy will be M1
     
for the 32 ! 52 and 32 ! 12 transitions and E2 for the 52 ! 12 transition. Thus
we need to calculate the rate for an M1 transition with E
¼ 178 keV. This can be
done using Equations (7.80) and gives 1=2  3:9  1012 s. The measured value is
3:5  1010 s, which confirms that the Weisskopf approximation is not very
accurate.

7.13 Set L ¼ 3 in Equation (7.78a), substitute the result into Equation (7.77) and use

¼
hT to give 
ðE3Þ ¼ ð2:3  1014 ÞE
7 A2 eV, where E
is expressed in MeV.

Chapter 8
8.1 To balance the number of protons and neutrons, the fission reaction must be

n þ 235 92 140
92 U ! 37 Rb þ 55 Cs þ 4n;

i.e. four neutrons are produced. The energy released is the differences in binding
energies of the various nuclei, because the mass terms in the SEMF cancel out. We
have, in an obvious notation,
" #

2=3 ðZ  NÞ2 Z2
ðAÞ ¼ 3; ðA Þ ¼ 9:26;  ¼ 0:28;  ¼ 485:0:
4A A1=3

The contribution from the pairing term is negligible (about 1 MeV). Using the
numerical values for the coefficients in the SEMF, the energy released per fission
EF ¼ 157:9 MeV.
The power of the nuclear reactor is P ¼ nEF ¼ 100 MW ¼ 6:25 1020 MeV s1 ,
where n is the number of fissions per second. Since one neutron escapes per fission and
386 APPENDIX D: SOLUTIONS TO PROBLEMS

contributes to the flux, the flux F is equal to the number of fissions per unit area per
second, i.e.
n P 6:25  1020 MeVs1
F¼ ¼ ¼ ¼ 3:15  1017 s1 m2 :
4r2 4r2 EF ð157:9 MeVÞ  ð12:57 m2 Þ

The interaction rate R is given by R ¼   F  ðnumber of target particlesÞ. The

latter is given by nT ¼ n  NA, where NA is Avogadro’s number and n is found from
the ideal gas law to be n ¼ PV=RT, where R is the ideal gas constant. Using
T ¼ 298 K, P ¼ 1  105 Pa and R ¼ 8:31 Pa m3 mol1 K1 , gives n ¼ 52:5 mol
and hence nT ¼ 3:2  1025 . Using the cross-section  ¼ 1031 m2 , the rate is
1:0  1012 s1 .

8.2 The neutron speed in the CM system is v  mv=ðM þ mÞ ¼ Mv=ðM þ mÞ and if the
scattering angle in the CM system is , then after the collision the neutron will have
a speed vðm þ M cos Þ=ðM þ mÞ in the original direction and Mv sin =ðM þ mÞ
perpendicular to this direction. Thus the kinetic energy is

mv2 ðM 2 þ 2mM cos  þ m2 Þ

Eðcos Þ ¼
2ðM þ mÞ2

and the average value is

2 32 31
ð1 ð1

4
Efinal ¼ E Eðcos Þ d cos 5 4 d cos 5 ¼ REinitial ;
1 1

where the reduction factor is R ¼ ðM 2 þ m2 Þ=ðM þ mÞ2 . For neutron scattering from
graphite, R  0:86 and after N collisions the energy will be reduced to
Efinal ¼ RN Einitial . The average initial energy of fission neutrons from 235 U is
2 MeV and to thermalize them their energy would have to be reduced to about
0.025 eV. Thus N  lnðEfinal =Einitial Þ=lnð0:86Þ ¼ 116.

8.3 From Equation (1.44a), for the fission of 235 U, Wf ¼ JNð235Þf and the total power
output is P ¼ Wf Ef , where Ef is the energy released per fission. For the capture by
238
U, Wc ¼ JNð238Þc . Eliminating the flux J, gives

Nð238ÞC P
Wc ¼ :
Nð235Þf Ef

Using the data supplied, gives Wc ¼ 1:08  1019 atoms s1  135 kg year1 .

8.4 Consider fissions occurring sequentially separated by a small time interval t. The
instantaneous power is the sum of the power released from all the fissions up to that
time. If E is the energy released in each fission, then over the lifetime of the reactor,
i.e. up to time T, the power is given by P0 ¼ nE=T, where n is the total number of
fissions and t ¼ E=P0 .
 CHAPTER 8 387

The power after some time t after the reactor has been shut down is

PðtÞ ¼ 3ðT þ tÞ1:2 þ 3ðT þ t  tÞ1:2 þ 3ðT þ t  2tÞ1:2    þ 3t1:2 :

In this formula, the first term is the power released from the first fission and the last
term is the power released from the last fission before the reactor was shut down. To
sum this series, we convert it to an integral:

TP0ð=EF
X
n¼P 0 T=E
1:2
PðtÞ ¼ 3 ðT þ t  nE=P0 Þ 3 ðT þ t  nE=P0 Þ1:2 dn:
n¼0
0

Setting u ¼ ðT þ t  nE=P0 Þ, gives

ðt h i
P0
PðtÞ ¼ 3 u1:2 du ¼ 0:075P0 t0:2  ðT þ tÞ0:2 :
E
Tþt

Using T ¼ 1 year and t ¼ 0:5 year, gives a power output of approximately 1.1 MW
after 6 months.

8.5 The PPI chain overall is: 4ð1 HÞ ! 4 He þ 2eþ þ 2e þ 2

þ 24:68 MeV. Two
corrections have to be made to this. Firstly, the positrons will annihilate with
electrons in the plasma releasing a further 2me ¼ 1:02 MeV per positron.
Secondly, each neutrino carries off 0.26 MeV of energy into space that will not
be detected. So, making these corrections, the total output per hydrogen atom is
1
4ð24:68 þ 2:04  0:52Þ ¼ 6:55 MeV. The total energy produced to date is
5:60  1043 J ¼ 3:50  1056 MeV. Thus, the total number of hydrogen atoms
consumed is 5:34  1055 and so the fraction of the Sun’s hydrogen used is
5:34  1055 =9  1056 ¼ 5:9 per cent and as this corresponds to 4.6 billion years,
the Sun has another 73 billion years to burn before its supply of hydrogen is
exhausted.

8.6 A solar constant of 8:4 J cm2 s1 is equivalent to 5:25  1013 MeV cm2 s1 of
energy deposited. If this is due to the PPI reaction 4ð1 HÞ ! 4 He þ 2eþ þ 2e þ 2
,
then this rate of energy deposition corresponds to a flux of ð5:25  1013=2 
6:55Þ  4  1012 neutrinos cm2 s1 .

8.7 For the Lawson criterion to be just satisfied, from Equation (8.46),

nd hdt  itc Q
L¼ ¼ 1:
6kT

We have kT ¼ 10 keV and from Figure 8.7 we can estimate hdt  i  1022 m3 s1 .
Also, from Equation (8.45), Q ¼ 17:6 MeV. So, finally, nd ¼ 6:8  1018 m3 .
388 APPENDIX D: SOLUTIONS TO PROBLEMS

8.8 The mass of a d–t pair is 5:03 u ¼ 4:69  109 eV=c2 ¼ 8:36  1024 g. The number
of d–t pairs in a 1 mg pellet is therefore 1:2  1020 . From Equation (8.45), each d–t
pair releases 17.6 MeV of energy. Thus, allowing for the efficiency of conversion,
each pellet releases 5:3  1026 eV. The output power is 750 MW ¼ 4:7  1027 eV=s.
Thus the number of pellets required is 8:9  9 s1 .

8.9 Assume a typical body mass of 70 kg, half of which is protons. This corresponds to
2:1  1028 protons and after 1 year the number that will have decayed is
2:1  1028 ½1  expð1=Þ, where  is the lifetime of the proton in years. Each
proton will eventually deposit almost all of its rest energy, i.e. approximately 0.938
GeV, in the body. Thus in 1 year the total energy in Joules deposited per kg of body
mass would be 4:5  1016 ½1  expð1=Þ and this amount will be lethal if greater
than 5 Gy. Expanding the exponential gives the result that the existence of humans
implies  > 0:9  1016 years.

8.10 The approximate rate of whole-body radiation absorbed is given by Equation (8.48a).
Substituting the data given, we have

dD AðMBqÞ  E
ðMeVÞ ð40  103 Þ  ð1:173 þ 1:333Þ
ð Sv h1 Þ ¼ ¼
dt 6r2 ðm2 Þ 6
¼ 1:67  102 Sv h1

and so in 18 h, the total absorbed dose is 0:30 Sv.

8.11 If the initial intensity is I0 , then from Equation (4.18), the intensities after passing
through bone, Ib , and tissue, It , are

Ib  I0 exp½ð b b þ 2 t tÞ and It  I0 exp½ t ðb þ 2tÞ:

Thus R ¼ exp½bð b  t Þ ¼ 0:7 and hence b ¼ lnð0:7Þ=ð b  t Þ ¼ 2:5 cm.

8.12 From Figure 4.8, the rate of ionization energy losses is only slowly varying for
momenta above about 1 GeV/c and given that living matter is mainly water and
hydrocarbons a reasonable estimate is 3 MeV g1 cm2 . Thus the energy deposited in
1 year is 2:37  109 MeV kg1 , which is 3:8  104 Gy.

8.13 In general, the nuclear magnetic resonance frequency is f ¼ jljB=jh. The numerical
input we use is:

j ¼ 7=2; B ¼ 1 T; ¼ 3:46 N ; N ¼ 3:15  1014 MeV T1

and h ¼ 4:13  1021 MeV s;

giving f ¼ 7:5 MHz.

 APPENDIX B 389

Appendix B
B.1 (a) From the definitions of s, t and u, we have

ðs þ t þ uÞc2 ¼ ðp2A þ 2pA pB þ p2B Þ þ ðp2A  2pA pC þ p2C Þ þ ðp2A  2pA pD þ p2D Þ

which, using p2A ¼ m2A c2 etc., becomes

ðs þ t þ uÞc2 ¼ 3m2A c2 þ m2B c2 þ m2C c2 þ m2D c2 þ 2pA ðpB  pC  pD Þ:

However, from four-momentum conservation, pA þ pB ¼ pC þ pD , so that

ðs þ t þ uÞc2 ¼ 3m2A c2 þ m2B c2 þ m2C c2 þ m2D c2  2p2A

and hence
X
ðs þ t þ uÞ ¼ mj 2 :
j¼A;B;C;D

(b) From the definition of t,


EA EC
c2 t ¼ p2A þ p2C  2pA pC ¼ m2A c2 þ m2C c2  2  pA : pC :
c2

For elastic scattering, A C. Thus EA ¼ EC and jpA j ¼ jpC j ¼ p, so that

pA  pC ¼ p2 cos . Then c2 t ¼ 2m2A c2  2 EA2 =c2  p2 cos  and using
EA2 ¼ p2 c2 þ m2A c4 , gives t ¼ 2p2 ð1  cos Þ=c2 .

B.2 Energy conservation gives E ¼ E þ E , where

E ¼
m c2 ; E ¼ cðm2 c2 þ p2 Þ1=2 ; E ¼ p c

and hence
 2  

m c2  p c ¼ c2 m2 c2 þ p2 : ð1Þ

However, three-momentum conservation gives

p cos  ¼ p ¼
m v; p sin  ¼ pv ¼ E =c: ð2Þ

Eliminating p and p between (1) and (2) and simplifying, gives

ðm2  m2 Þ
tan  ¼ :
2
2 m2
390 APPENDIX D: SOLUTIONS TO PROBLEMS

B.3 Conservation of four-momentum is p ¼ p  p , from which p2 ¼ p2 þ p2  2p p .

Now p2j ¼ m2j c2 for j ¼ ; and , and

E E
p pv ¼  p  p ¼ m E ¼ m jp jc;
c2
 
because p ¼ 0 and E ¼ m c2 in the rest frame of the pion. However, jp j ¼ p  p
because the muon and neutrino emerge
h back-to-back.
i 1 Thus, p ¼ ðm2  m2 Þ c=2m ;
2
but p ¼
m v, from which v ¼ pc p2 þ m2 c2 . Finally, substituting for p gives
!
m2  m2
v¼ c:
m2 þ m2

B.4 By momentum conservation, the momentum components of X 0 are:

px ¼ 0:743 (GeV/c), py ¼ 0:068 (GeV/c), pz ¼ 2:595 (GeV/c) and hence
p2X ¼ 7:291. Also, p2A ¼ 4:686 (GeV/c)2 and p2B ¼ 0:304 (GeV/c)2 .

Under hypothesis (a):

EA ¼ ðm2 c4 þ p2A c2 Þ1=2 ¼ 2:169 GeV and EB ¼ ðm2K c4 þ p2B c2 Þ1=2 ¼ 0:740 GeV.
Thus EX ¼ 2:909 GeV and MX ¼ ðEX2  p2X c2 Þ1=2 c2 ¼ 1:082 GeV/c2 .
Under hypothesis (b):
EA ¼ ðm2p c4 þ p2A c2 Þ1=2 ¼ 2:359 GeV and EB ¼ ðm2 c4 þ p2B c2 Þ1=2 ¼ 0:569 GeV.
Thus EX ¼ 2:928 GeV and MX ¼ ðEX2  p2X c2 Þ1=2 c2 ¼ 1:132 GeV/c2 .
Since MD ¼ 1:86 GeV/c2 and M ¼ 1:12 GeV/c2 , the decay is  ! p þ  .

B.5 If the four-momenta of the initial and final electrons are p ¼ ðE=c; qÞ and
p0 ¼ ðE0 =c; q0 Þ, respectively, the squared four-momentum transfer is defined by

Q2 ðp0  pÞ2 ¼ 2m2 c2 þ 2EE0 =c2  2q  q0 :

However, E ¼ E0 and jqj ¼ jq0 j q, so neglecting the electron mass,

Q2  2q2  ð1  cos Þ. The laboratory momentum may be found from Equation
(B.36):
c2 h 2
ih
2
i c2 ðs  m2 Þ2
p
q2 ¼ s  ðm p  m e Þ s  ðmp þ m e Þ  ;
4m2p 4m2p

where the invariant mass squared s is defined by s ðp þ PÞ2 =c2 and P is the four-
momentum of the initial proton, i.e. P ¼ ðmp c; 0Þ. Thus,

s ¼ m2e þ m2p þ 2mp E=c2  m2p þ 2mp E=c2 :

Substituting into the expression for Q2 gives Q2  2E2 ð1  cos Þ=c2 .

  
B.6 The total four-momentum of the initial state is ptot ¼ E þ mp c2 =c; pL . Hence the
invariant mass W is given by ðWc2 Þ2 ¼ ðEL þ mp c2 Þ2  p2L c2, where pL jpL j. The
 APPENDIX B 391

invariant mass squared in the final state evaluated in the centre-of-mass frame has a
minimum value ð4mp cÞ2 when all four particles are stationary. Thus, Emin is given by

ðEmin þ mp c2 Þ2  p2L c2 ¼ ð4mp c2 Þ2

2
which expanding and using Emin  p2L c2 ¼ m2p c4 , gives Emin ¼ 7mp c2 ¼ 6:6 GeV.
For a bound proton, the initial four-momentum of the projectile is ðEL0 =c; p0 L Þ and
that of the target is ðE=c; pÞ, where p is the internal momentum of the nucleons,
which we have taken to be in the opposite direction to the beam because this gives
the maximum invariant mass for a given EL0 . The invariant mass W 0 is now given by

ðW 0 c2 Þ2 ¼ ðEL0 þ EÞ2  ðp0L  pÞ2 c2 ¼ 2m2p c4 þ 2EEL0 þ 2pp0L c2 :

0
Since the thresholds Emin and Emin correspond to the same invariant mass 4mp , we
have 2mp c Emin ¼ 2EEmin þ 2pp0min c2 . Finally, since the internal momentum of the
2 0

nucleons is  250 MeV=c (see Chapter 7), E  mp c2 , while for the relativistic
incident protons pmin  Emin =c, so using these gives
0
 
Emin  1  p=mp c Emin ¼ 4:8 GeV:

B.7 The initial total energy is Ei ¼1 EA ¼ mA c2 and the final total1energy is Ef ¼ EB þ EC ,

where EB ¼ ðm2B c4 þ p2B c2 Þ2 , and EC ¼ ðm2C c4 þ p2C c2 Þ2 , with pB ¼ jpB j and
pC ¼ jpC j. However, by momentum conservation, pB ¼ pC p and so
h i 
1 2 
mA c2  ðm2B c4 þ p2 c2 Þ2 ¼ m2C c4 þ p2 c2 ;

 
which on expanding gives EB ¼ m2A þ m2B  m2C c2 =2mA .

B.8 If the four-momenta of the photons are pi ¼ ðEi =c; pi Þði ¼ 1; 2Þ, then the invariant
mass of M is given by M 2 c4 ¼ ðE1 þ E2 Þ2  ðp1 þ p2 Þc2 ¼ 2E1 E2 ð1  cos Þ, since
p1  p2 ¼ E1 E2 ð1  cos Þ=c2 for zero-mass photons. Thus, cos  ¼ 1  M 2 c4 =2E1 E2 .

B.9 A particle with velocity v will take time t ¼ L=v to pass 1 between the two
counters. Relativistically,
1
p ¼ mv
with
¼ ð1  v2 =c2 Þ2 . Solving, gives

v ¼ cð1 þ m2 c2 =p2 Þ 2 and hence the difference in times-of-flight (assuming
m1 > m2 ) is
" 1 1 #
L m21 c2 2 m22 c2 2
t ¼ 1þ 2  1þ 2 :
c p p

Using m1 c2 ¼ mp c2 ¼ 0:983 GeV, m2 c2 ¼ m c2 ¼ 0:140 GeV and pc ¼ 2 GeV

gives t ¼ ½1:114  1:002ðL=cÞ and Lmin ¼ 0:54 m.

B.10 In an obvious notation, the kinematics in the lab frame are:

ðE
; p
Þ þ e ðmc2 ; 0Þ !
ðE
0 ; p0
Þ þ e ðE; pÞ:
392 APPENDIX D: SOLUTIONS TO PROBLEMS

Energy conservation gives E

þ mc2 ¼ E
0 þ E and momentum conservation
gives p
¼ p0
þ p. From the latter we have E2  m2 c4 ¼ c2 ðp2
þ p0
2  2p
 p0
Þ.
But p
c ¼ E
; p0
c ¼ E
0 and the scattering angle is , so we have
E2  m2 c4 ¼ E
2 þ E
2  2E
E
0 cos . Eliminating E between this equation and the
equation for energy conservation gives E
0 ¼ E
½1 þ E
ð1  cos Þ=mc2 1 . Finally,
using E
¼ E
0 =2 and  ¼ 600 , gives E
¼ 2mc2 ¼ 1:02 MeV.

Appendix C
C.1 The assumptions are: ignore the recoil of the target nucleus because its mass is
much greater than the total energy of the projectile -particle; use non-relativistic
kinematics because the kinetic energy of the -particle is very much less that its rest
mass; assume the Rutherford formula (i.e. the Born approximation) is valid for
small-angle scattering. The relevant formula is then Equation (C.13) and it may
be evaluated using z ¼ 2; Z ¼ 83, Ekin ¼ 20 MeV and  ¼ 20 . The result is
d=d ¼ 98:3 b=sr.

C.2 From Figure C.2, the distance of closest approach d is when x ¼ 0. For x < 0, the
sum of the kinetic and potential energies is Etot ¼ 12mv2 and the angular momentum
is mvb. At x ¼ 0, the total mechanical energy is 12mu2 þ Zze2 =4"0 d and the
angular momentum is mud, where u is the instantaneous velocity. From angular
momentum conservation, u ¼ vb=d and using this in the conservation of total
mechanical energy gives d 2  Kd  b2 ¼ 0 where, using Equation (C.9),
K 2b=cotð=2Þ. The solution for d  0 is d ¼ b½1 þ cosecð=2Þ=cotð=2Þ.

C.3 The result for small-angle scattering follows directly from Equation (C.9) in the
limit  ! 0. Evaluating b, we have, using the data given,
2 
zZe2 e

hc 1
b¼ ¼ 2zZ ¼ 1:55  1013 m:
2
2"0 mv  4"0
hc mc ðv=cÞ2 
2

The cross-section for scattering through an angle greater than 5 is thus

 ¼ b2 ¼ 7:55  1026 m2 and the probability that the proton scatters through
an angle greater than 5is P ¼ 1  exp½nt,
 where n is the number density of
the target. Using n ¼ 6:022  1026 =194  21450 ¼ 6:658  1028 m3 , gives
P ¼ 4:91 102 . Since P is very small but the number of scattering centres is
very large, the scattering is governed by the Poisson distribution and the probability
for a single scatter is P1 ðmÞ ¼ mem ¼ 4:91  102 , giving m  0:052. Finally, the
probability for two scattering is P2 ¼ m2 expðmÞ=2!  1:3  103 .
References

Aj90 F. Ajzenberg-Selove (1990) Energy levels of light nuclei., Nucl. Phys., A506
1–158.
Am95 P. Amaudruz et al. (1995) A re-evaluation of the nuclear structure function ratios
for D, He, 6Li, C and Ca, Nucl. Phys., B441, 3–11.
Ar95 M. Arneodo et al. (1995) The structure function ratios F2C /F2D and F2Ca /F2D at small
x. Nucl. Phys., B441, 12–30.
Ar97 M. Arneodo et al. (1997) Measurements of the proton and deuteron structure
functions F2p and F2d and of the ratio L /T . Nucl. Phys., B483, 3–43.
As04 Y. Ashie. et al. (2004) Evidence for an oscillatory signature in atmospheric
neutrino oscillations. Phys. Rev. Letters, 93, 101801/1–5.
At82 W. B. Atwood (1982) Lectures on Lepton Nucleon Scattering and Quantum
Chromodynamics, Progress in Physics Vol. 4.
Ba77 R. C. Barrett and D. F. Jackson (1977) Nuclear Sizes and Structures, Clarendon
Press, Oxford, UK.
Be67 J. B. Bellicard et al. (1967) Scattering of 750-MeV electrons by calcium isotopes.
Phys. Rev. Letters, 19, 527–529.
Bl52 J. M. Blatt and V. F. Weisskopf (1952), Theoretical Nuclear Physics, John Wiley
and Sons, New York, USA.
Bo69 A. Bohr and B. R. Mottelson (1969) Nuclear Structure Vol. 1, W. A. Benjamin Inc.,
New York, USA.
Ch97 Chart of the Nuclides (1997) General Electric Company, Schenectady, New York,
USA.
Co01 W. N. Cottingham and D. A. Greenwood (2001) An Introduction to Nuclear
Physics, 2nd edn. Cambridge University Press, Cambridge, UK.
De99 P. P. Dendy and B. Heaton (1999) Physics for Diagnostic Radiology, 2nd edn.,
Institute of Physics Publishing, UK.
Ei04 S. Eidelman et al. (2004) Review of Particle Physics. Physics Letters B592, 1–
1109.
En66 H. A. Enge (1966) Introduction to Nuclear Physics, Addison-Wesley Publishing
Company.
Fe86 R. Fernlow (1986) Introduction to Experimental Particle Physics, Cambridge
University Press, Cambridge, UK.
Fo61 D. B. Fossan et al. (1961) Neutron total cross sections of Be, B10 , B, C, and O.
Phys. Rev., 123, 209–218.
Fr83 B. Frois (1983) Proc. Int. Conf. Nucl. Phys., Florence eds. P. Blasi and R. A. Ricci,
(Tipografia Compositori, Bolgona) Vol. 2, p. 221.

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
394 BIBLIOGRAPHY

Ga76 D. I. Garber and R. R. Kinsey (1976) Neutron Cross-sections Vol 2, Brookhaven

National Laboratory Report BNL-325, Upton, NY, USA.
Gj74 S. Gjesdal et al. (1974) A measurement of the KL KS mass difference from the
charge asymmetry in semi-leptonic kaon decays. Phys. Letters, B52, 113–119.
Go86 K. Gottfried and V. F. Weisskopf (1986) Concepts of Particle Physics, vol. 2,
Oxford University Press, Oxford, UK.
Gr96 C. Grupen (1996) Particle Detectors. Cambridge University Press, Cambridge,
UK.
Gr87 D. Griffiths (1987) Introduction to Elementary Particles. John Wiley and Sons,
New York, USA.
Ha84 F. Halzen and A. D. Martin (1984) Quarks and Leptons. John Wiley and Sons,
New York, USA.
He97 W. R. Hendee (1997) Physics and applications of medical imaging. Rev. Mod.
Phys., 71, S444–S450.
Hi68 F. Hintenberger et al. (1968) Inelastic scattering of 52 MeV deuterons. Nucl. Phys.,
A115, 570–592.
Ho97 P. E. Hodgson, E. Gadioli and E. Gadioli Erba (1997) Introductory Nuclear
Physics. Oxford University Press, Oxford, UK.
Ho97a R. K. Hobbie (1997) Intermediate Physics for Medicine and Biology, 3rd edn..
Springer and AIP Press, Germany.
Ja75 J. D. Jackson (1975) Classical Electrodynamics, 2nd edn.. John Wiley & Sons,
New York, USA.
Je90 N. A. Jelley (1990) Fundamentals of Nuclear Physics. Cambridge University
Press, Cambridge, UK.
Ke82 D. Keefe (1982) Inertial confinement fusion. Ann. Rev. Nucl. Part. Sci., 32, 391–
441.
Kl86 K. Kleinknecht (1986) Detectors for Particle Radiation. Cambridge University
Press, Cambridge, UK.
Kr88 K. S. Krane (1988) Introductory Nuclear Physics. John Wiley and Sons, New
York, USA.
Li01 J. Lilley (2001) Nuclear Physics. Principles and Applications, John Wiley and
Sons. Ltd., Chichester, UK.
Ma92 F. Mandel (1992) Quantum Mechanics, 2nd edn.. John Wiley and Sons Ltd.,
Chichester, UK.
Ma97 B. R. Martin and G. Shaw (1997) Particle Physics, 2nd edn.. John Wiley and Sons
Ltd., Chichester, UK.
McR03 D. W. McRobbie, et al. (2003) MRI From Picture to Proton. Cambridge University
Press, Cambridge, UK.
Mo94 L. Montanet et al. (1994) Review of particle properties. Phys. Rev., D50, 1173–
1814.
Me61 E. Merzbacher (1961) Quantum Mechanics. John Wiley and Sons, New York,
USA.
NRC99 Report of the Board on Physics and Astronomy of the National Research Council,
USA (1999) Nuclear Physics: The Core of Matter, The Fuel of Stars. National
Academies Press, Washington, D.C., USA.
Pe00 D. H. Perkins (2000) Introduction to High Energy Physics, 4th edn. Cambridge
University Press, Cambridge, UK.
 395

Pe03 D. H. Perkins (2003) Particle Astrophysics. Oxford University Press, Oxford, UK.
Ph94 A. C. Phillips (1994) The Physics of Stars. John Wiley and Sons. Ltd., Chichester,
UK.
Po99 B. Povh et al. (1999) Particles and Nuclei. An Introduction to the Physical
Concepts. Springer, Germany.
Sa67 G. R. Satchler (1967) Optical model for 30 MeV proton scattering. Nucl. Phys.,
A92, 273–305.
Sc68 L. I. Schiff (1968) Quantum Mechanics, 3rd edn.. McGraw-Hill, New York, USA.
Sc01 W. Scheider (2001) News Update to A Serious But Not Ponderous Book About
Nuclear Energy, Cavendish Press, Ann Arbour, USA.
Se80 E. Segrè (1980) From X-Rays to Quarks. W. H. Freeman and Company, USA.
Se97 W. G. Seligman et al. (1997) Improved determination of s from neutrino-nucleon
scattering. Phys. Rev. Letters, 79, 1213–1216.
Si75 I. Sick et al. (1975) Shell structure of the 58 Ni charge density. Phys. Rev. Letters,
35, 910–913.
Tr75 G. L. Trigg (1975) Landmark Experiments in 20th Century Physics. Crane, Russak
and Co. Ltd, New York, USA.
Zu00 A. Zurlo et al. (2000) The role of proton therapy in the treatment of large
irradiation volumes: a comparative planning study of pancreatic and biliary
tumors. Int. J. Radiat. Oncol. Biol. Phys, 48, 277–288.
Bibliography

Below are brief notes on a few books on nuclear and particle physics at the
appropriate level which I have found particularly useful. Other, more specialized,
texts are listed in the References section which follows.

1 Nuclear Physics
Two very readable concise texts at about the level of the present book although
covering more topics are: W. N. Cottingham and D. A. Greenwood, An Introduc-
tion to Nuclear Physics 2nd edn., Cambridge University Press, 2001, and N. A.
Jelley, Fundamentals of Nuclear Physics, Cambridge University Press, 1990. Both
deal with theoretical aspects only – there is nothing about experimental methods.
Both provide some problems for each chapter with either answers or brief hints on
solutions. Another good book at this level is: J. Lilley, Nuclear Physics –
Principles and Applications, John Wiley and Sons, 2001. This is in two parts.
The first covers the principles of nuclear physics, including experimental techni-
ques, and the second discusses an unusually wide range of applications, including
industrial and biomedical uses. An extensive range of problems is provided, with
detailed notes on their solutions.
Two good examples of comprehensive texts covering both theory and experi-
ment are: K. S. Krane, Introductory Nuclear Physics, John Wiley and Sons, 1988,
and P. E. Hodgson, E. Gadioli and E. Gadioli Erba, Introductory Nuclear Physics,
Oxford University Press, 1997. Both provide problems, but without solutions.
Finally there is the unique set of (hand written!) lecture notes by Fermi: E.
Fermi, Nuclear Physics, University of Chicago Press, 1950. Although old, these
are still well worth reading.

2 Particle Physics
There are several books covering particle physics at the appropriate level, For
obvious reasons, the one closest to the present book is: B. R. Martin and G. Shaw,

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
398 BIBLIOGRAPHY

Particle Physics, 2nd edn., John Wiley and Sons, 1997, and some of the material
on particle physics in the present book has been developed from this previous
book. It covers both theory and experimental methods. Problems with full
solutions are provided for each chapter.
Some of the other texts available are now rather dated, but one that is not is: D.
H. Perkins, Introduction to High Energy Physics, 4th edn., Cambridge University
Press, 2000. This book is well-established and has changed substantially over the
years. It goes further than the present book in its use of relativistic calculations.
The latest edition (the fourth) has far less discussion of experimental methods
than earlier editions, but an expanded chapter on astroparticle physics. It is
therefore worth looking at the third edition also. Problems are provided, some
with answers.
Another older book, but still relevant, is: D. Griffiths, Introduction to Elemen-
tary Particle Physics, John Wiley and Sons, 1987. Griffiths’ book is written in an
unconventional conversational style, with interesting footnotes (and extensive
notes at the end of most chapters), giving further details and background. It is
exclusively theoretical – there is nothing on experimental techniques. It goes well
beyond the present text, as at least half of the book involves the detailed evaluation
of Feynman diagrams. A wealth of interesting problems is provided at the end of
each chapter, but no solutions are given.

3 Nuclear and Particle Physics

There are not many books that treat nuclear and particle physics together and some
of those are out-of-date. Five that are appropriate are:

R. A. Dunlap, The Physics of Nuclei and Particles, Thomson Learning – Brooks/

Cole, 2004;

A. Das and T. Ferbel, Introduction to Nuclear and Particle Physics, John Wiley
and Sons, 1994;

W. S. C. Williams, Nuclear and Particle Physics, Oxford University Press, 1991;

W. E. Burcham and M. Jobes, Nuclear and Particle Physics, Longman Scientific

and Technical, 1995;

B. Povh, K. Rih, C. Scholz and F. Zetsche, Particles and Nucle, 2nd edn., Springer,
1995.

The first two books are concise readable introductions, although in Dunlap’s case
the particle physics part is very short – just 50 pages. This book is exclusively
about theory, whereas the book of Das and Ferbel also discusses experimental
 NUCLEAR AND PARTICLE PHYSICS 399

methods. Both books provide problems, but neither supplies solutions. The book
by Williams is fairly comprehensive, although now a little old. The style is rather
discursive. There is a wealth of illustrations and many problems are given, with
answers to some of them supplied. (A full solutions manual is available as a
separate volume.) The book by Burcham and Jobes is also comprehensive and goes
further than the present text. There are many problems, all with solutions. Both of
the latter two books treat nuclear and particle physics as almost independent
subjects. The book by Povh et al. is closest in its coverage to the present book and
at a similar level, although experimental methods are only discussed in a brief
appendix. Some problems with solutions are provided for all chapters.
Index

Absorption length 124, 145 Antiparticles

Accelerator driven systems (ADS) discovery 9
(see Nuclear power) predicted from Dirac equation 8
Accelerators 113–121 quark structure 96
alternating current 115–121 relative parity of 11
CEBAF, Jefferson Laboratory, USA Antiscreening 164
116 Associated production 122
CERN complex 120 Asymmetry term in SEMF 51
Cockcroft-Walton 113 Asymptotic freedom 155, 160–161, 164
colliders 112, 119 ATLAS detector at LHC 146–147
COSY, Jülich, Germany 121 Atomic mass (unit) 56
cyclic 116–119 Atomic number 33
cyclotron 116–117
direct current 113–115 Barn (unit) 29
fixed-target 119 Barrier penetration
LHC, CERN, Switzerland 120–121 alpha-decay 239–242
linear 115–116 fusion 267, 271
RHIC, BNL, USA 121 quantum theory 331–333
SLAC, Stanford, USA 116 Baryon-antibaryon asymmetry in universe
storage rings 119 314–315
synchrotron 116–119 Baryon number 91
Van de Graaff machine 113–115 conservation 91
Activation energy in fission 60–61 violation in grand unified theories
Activity 47–48, 265, 279 303–304
ADS (see Nuclear power) Baryons
Allowed transitions in beta decay 246 colour wavefunction 153
-decay 238–242 magnetic moments in quark model
Gamow factor 240 102–104
Geiger-Nuttall relation 241 masses in quark model 105–108
potential well 239 mass splittings within multiplets
tunnelling theory 239–242 104–105
-rays 1 multiplets in quark model 96–102
Amplification factor in gas detectors 132–133 Beams (see Particle beams)
Amplitudes 21–23 Beauty quantum number 89
coupling constant 22 Becquerel (unit) 47
propagator 22 Bending magnet 117–118

Nuclear and Particle Physics B. R. Martin

# 2006 John Wiley & Sons, Ltd. ISBN: 0-470-01999-9
402 INDEX

-decay Centre-of mass system 341–342

allowed transitions 246  erenkov counter 141–142
C
condition for stability from SEMF 57  erenkov radiation 141
C
double 58 Chandrasekar limit 308
electron momentum distribution 244–246 Charge distribution of nuclei 37–42
even-mass nuclei 58–59 Charge conjugation
Fermi screening factor 246 C-parity 12
Fermi theory 242–244 fermion-antifermion pair 12–13
Fermi transitions 243 violation in weak interactions 183–185
forbidden transitions 246 Charged current weak interactions 181
Gamow-Teller transitions 243 violation of quark numbers 194–196
Kurie plot 246–248 W-lepton vertices 193–194
Lorentz invariant amplitudes 243 W-quark vertices 196–198
neutrinoless double -decay 84 Charge distribution
neutrino mass from -decay of tritium mean square charge radius 42
247–248 nuclear 38–42
odd-mass nuclei 55–57 Charm quantum number 90–91
-rays 1 Charmed particles 98
Bethe-Bloch formula 125–126 Charmonium 156–160
Big bang model 312 Chirality 189–190
Binding energy 34 Chromomagnetic interaction 105, 219
experimental curve 36 CKM matrix 198
semi-empirical mass formula 50–55 Cloud chamber 132
theoretical predictions 318 Cockcroft-Walton machine 113
Biological effects of radiation Collective model 236–237
cell damage due to 278 Colliders 112, 119
oxygen effect 281 advantages and disadvantages 112
production of free radicals 280–281 luminosity 121
units of radiation 279 Collision length 124
Born approximation 25 Colour 152
Mott Scattering 38 confinement 152
Rutherford scattering 37 evidence from eþ e annihilation
Bottomium 158–160 166–168
Bragg curve 127 hadron spectroscopy 151, 153
Bragg peak 127, 282 hypercharge and isospin 152
Branching ratio 27 quantum number for gluons 154–155
isospin predictions for 93–95 role in QCD 153–156
Branes 307 singlet 152
Breit-Wigner formula 27–29 wavefunction 153
Bremsstrahlung 128 Colour confinement 152
Bubble chamber 132 confining potential 160
Complex potential 44–45
Cabibbo allowed/suppressed decays 197 Compound nucleus 63
Cabibbo angle 197 Compton scattering 129–130
Cabibbo hypothesis 196 Computed tomography 286–289
Callan-Gross relation 172 Confinement (see Colour)
Calorimeters 142–145 Conservation laws
Carbon dating 48 angular momentum 9
CDF detector at Fermilab 146 baryon number 91
 INDEX 403

B-L 303 MACHOs 313

charge conjugation 12–13 WIMPs 313–314
colour 151–153 Data
CP 202–206 nuclear and particle xiii
isospin 93–96 physical constants and conversion
lepton number 73–74 factors xv
linear momentum 9 Decay constant 47
parity 10–12 Decay width 27
quark numbers 90–91 Deep inelastic scattering
strangeness, charm, beauty and truth 89 Callan-Gross relation 172
Constituent quarks 90 charged lepton-nucleon 168–177
Control rods 260–262 EMC effect 317
Coulomb barrier Gottfried sum rule 179
alpha decay 239–242 neutrino-nucleon 198–201
fusion 266–267 nuclear 65
Coulomb term in SEMF 51 parton model 171
Cosmic baryon asymmetry 314–315 scaling 169–171
Cosmic rays 48 scaling violations 175–177
C-parity structure functions 169
fermion-antifermion pairs 12–13 Deformation parameter 59–60, 233–234
violation in weak interactions 183–185 Degeneracy pressure 308
CP symmetry Delayed neutron emission 258, 262–263
consequences for muon decay 185–186 Delta resonance 95–96
violation in B0 B 0 system 206 Density-of-states 333–335
violation in neutral K decays 202–205 Detectors (see Particle detectors)
CPT theorem 208 Deuteron 218
Cross-sections Differential cross-section 24
Born approximation 25 Dipole transitions 249
deep inelastic scattering 169–170, 200 Diquarks 219
differential 24 Dirac equation
elastic scattering from nuclei 39 magnetic moments 6
flux 23 parity for fermion-antifermion pairs 11–12
luminosity 24 solutions as spinors 8
Mott scattering 38 Direct nuclear reaction 63
neutron-uranium 256–257 Drift chamber 135
partial 24 Drift tube 115
Rutherford scattering 37
spin factors 26 Effective dose 279
total 24 Elastic scattering 14
CT (see Computed tomography) Electric dipole moments 323
Curie (unit) 47 Electric quadruple moments 232–236
Current quarks 96 Electromagnetic interactions 16
Cyclotron 116–117 pair production and annihilation 129–130
Cyclotron frequency 148 typical lifetimes 102
Electromagnetic showers 143–145
Dark matter 313 Electron 2
cosmological constant 313 magnetic moment 6
dark energy 313 Electron capture 46
inflationary big bang model 312–313 condition for occurrence 57
404 INDEX

Electron neutrinos 71–72 importance of pairing energy 61

Electron number 73 induced 47
Electron-positron annihilation prompt neutrons 256
evidence for colour 166–168 spontaneous 47
hadron production 164–166 Fixed-target experiment 111–112
three-jet events 166 Flavour 79, 89
two-jet events 87–88 Flux 23
Electron-positron pair production 129–130 Focusing magnet 118
Electron volt (unit) 29 Forbidden transitions in beta decay 246
Electroweak unification Forces of nature 6
confirmation in experiments 211–213 Four-momentum 340
manifestation at high energies 182 Four-vector 340
Elementarity 5 Fragmentation 87, 162
EMC effect 317 Fusion
Emulsions 132 Coulomb barrier 266–267
Energy losses by particles in matter Maxwell-Boltzmann distribution 270
ionization losses 125–127 reaction rates 270–273
photons 129–130 reactions 267–270, 273–274
photoproduction reactions 125 reactors 273–278
radiation losses 128–129 stellar fusion 267–270
short-range interactions with nuclei tunnelling 267, 271
123–125 Fusion reactors
Equivalent dose 279 deuterium-deuterium reaction 273–274
Exotic hadrons 153 deuterium-tritium reaction 274
Lawson criterion 275
Fast breeder reactor 264–265 inertial confinement 276
Fermi (unit) 29 magnetic confinement 275–278
Fermi coupling constant 22 tokamak 276–277
Fermi gas model 220–221 -emission
contributions to binding energy 221–223 mutipole radiation 248–250
Fermi’s Golden Rule (see Golden Rule) radiative width 251–252
Fermi transitions 243 role of angular momentum 62, 249
Feynman diagrams 15–18 selection rules 248–250
Feynman rules 20 transition rates 250–252
gluon vertices 154–166 Weisskopf approximation 250
multiparticle-exchange 22 -rays 4
order of 22 attenuation in matter 129–130
vertices in electroweak interactions use in medical imaging 283–284
193–194, 196–198, 210–211
Feynman rules 20 Gamow factor 240, 267
Fine structure constant 20 Gamow-Teller transitions 243
Fissile nuclei 256–257 Gauge bosons 6–7
Fission Gauge invariance
activation energy 60–61 lattice gauge theory 155
chain reactions 258–260 role in QCD 154
condition for stability 60 standard model 6
delayed neutrons in 257 Geiger-Müller counter 136
fissile materials 256–257 Geiger-Müller region 136
fragments 59 Geiger-Nuttall relation 241
 INDEX 405

g-factor 229 Hypercharge 98–99

Glueballs 155 Hyperfine interaction 105
Gluino 305 Hypernucleus 318
Gluons
determination of spin in electron-positron Imaging (see Medical imaging)
annihilation 166 Impact parameter 350–351
gluon-gluon scattering 155 Inelastic scattering 14
Golden Rule 333–338 Internal conversion 250
Grand unified theories Invariant amplitude (see Amplitude)
B-L quantum number 303 Invariant mass 342
fundamental vertices 302 Ionization chamber 133
non-conservation of baryon and lepton Isobar 33
numbers 302–303 Isospin
proton decay 304 branching ratio predictions
unification mass 301 93–96
X and Y bosons 302–304
I ¼ 12 rule 215
weak mixing angle 303 hadron multiplets 97–100
Gray (unit) 279 Isotone 33
GUTs (see Grand unified theories) Isotope 33
production for medical use 326
Hadrons
charge independence of nuclear forces 93 Jet chambers 135
charge symmetry 93 Jets 87
decays in quark model 101–102 eþ e annihilation 164–166
excited states in quark model 100 evidence for quarks 87–88
exotic states 153 fragmentation 87
flavour independence 92–93 gluon spin determination 166
glueballs 155 3-jet events 166
heavy quark bound states 156–160 2-jet events 88
magnetic moments in quark model QCD 165
102–104
masses in quark model 104–108 Kaons 98
multiplets 97–100 K-capture 46, 57
quark model structure 7 Klein-Gordon equation 8
semileptonic weak decays 194–198 Kurie plot 246–248
Half-life 48
Helicity 187 Large Hadron Collider 120–121
measurement for neutrinos 187–18 Larmor frequency 290
right-handed and left-handed states 187 Lattice gauge theory 155
role in muon decay 191 Lawson criterion 275
role in pion decay 190 Leptons 5, 71–74
states for massless particles 187 decays 72–73
states for particles with mass 189–192 lepton numbers 73
Higgs boson 6 lepton-quark symmetry 195
detection at LHC 299–300 scattering as evidence for quarks 87
mass limits from experiment 298–300 universal weak interactions 84–85
origin of mass problem 298 Lepton-quark symmetry 195
Higgsino 305 Lepton universality 84–85
Hubble constant 312 LET 281–282
406 INDEX

LHC (see Large Hadron Collider) Mesons

Lifetime 48 bound states of heavy quarks 156–160
Linac (see Linear accelerator) glueballs 155
Linear accelerator 115–116 kaons 98–99
Linear energy transfer (LET) masses in quark model 106
281–282 pions 7, 97–99
Liquid drop model 50–54 quark structure 96–99
Luminosity 24 Microstrip detector 136
Minimum ionization 126
MACHOs 313 Mirror nuclei 93–94
Magic number 224–225 Moderator 261–262
Magnetic moments Mott cross-section 38
Dirac equation predictions 6 MRI (See Magnetic resonance imaging)
nuclei 228–231 MSSM (see Supersymmetry)
nucleon 7 Multiplication factor in gas detectors 132
quark model predictions 102–104 Multipole radiation 249–250
Magnetic resonance imaging Multiwire proportional chamber MWPC (see
BOLD images 325 Particle detectors)
imaging elements other than hydrogen
325–326 Natural units 30
Larmor frequency 290 Neutral current weak interactions 181
nuclear resonance frequency 290 conservation of quark numbers 211
relaxation times 291–293 electroweak unification 211–213
spatial encoding 293 prediction from electroweak theory 209
spin echos 292 Z 0 – vertices 210–211
Magnetic spectrometer 34, 139 Neutralino 306
Mass deficit 34 Neutral K decays 202–205
Masses in quark model 105–108 CPT theorem 208
Mass number 33 CP violation 203–205
Mass spectrometer 34 flavour oscillation 206–208
Mass spectroscopy 34–35 strangeness oscillation 202
Matter-antimatter asymmetry 314–315 Neutrino astrophysics
Matter distribution of nuclei 42 active galactic nuclei 310
Mean life 48 Amanada experiment 310–312
Medical imaging Chandrasekhar limit 308
attenuation coefficient 283 detection of supernova neutrinos 309–310
computed tomography (CT) 286–289 IceCube experiment 312
CT scanner 287 neutron star 308
functional MRI 325 supernovas 308
-camera 284 Neutrinos 4
magnetic resonance imaging (MRI) emitted in supernova explosions 308–309
290–293 Limits on number of neutrinos 84–86
oxygen effect 281 Majorana neutrino 192
positron emission tomography (PET) masses from oscillation measurements
288–289 79–83
projected images 282–285 masses of  and  75–76
proton beams 324–325 mass of e from beta decay of tritium 75
single-photon emission CT (SPECT) mixing 77–79
287–288 multiplets 71–72
 INDEX 407

oscillations 77–79 Nuclear shape and sizes 37–45

scattering from electrons 198–200 Nucleon 3
scattering from nucleons 200–201 Nucleon-nucleon force 217–219
solar neutrinos 82–83 three-body contribution 318
two-component theory 192 Nucleon number 33
Neutron 3 Nucleosynthesis 322
capture 64, 66 Nuclide 33
decay 194–195 N-Z nuclide distribution 45–46
magnetic moment 7, 102–104
scattering 64, 66, 256–257 Oblate nuclei 233–234
Neutron number 33 Observables
Neutron stars 308 amplitudes 21–23
Non-spherical nuclei cross-sections 23–26
electric quadruple moments 232–236 decay rates 27–29
collective model 236 Omega-minus particle 152–153
Nuclear chain reaction Optical model 43–45
criticality condition 258 Order of Feynman diagrams 22
critical size for explosive release of Oxygen effect 281
energy 258–259 OZI Rule 157
use in power production 260–265
Nuclear density 43 Pairing term in SEMF 52
Nuclear force (see Strong nuclear force) Pair production 129–130
Nuclear form factor 38–42 Parity 10
Nuclear fusion (see Fusion) associated with angular momentum 11
Nuclear instability fermion-antifermion pair 11
-decay 238–242 gamma emission 249
-decay 242–248 leptons 11
fission 47, 59–61 quarks 12
-emission and internal conversion violation in weak interactions 183
47, 248–252 Partial width 27
range of stability 46 Particle accelerators (see Accelerators)
Nuclear magneton (unit) 103 Particle astrophysics
Nuclear models dark matter 312–314
collective model 236 matter-antimatter asymmetry 314–315
Fermi gas 220–223 neutrino astrophysics 308–312
liquid drop 50–54 Particle beams
shell 223–232 neutral and unstable particles 122–123
Nuclear physics primary 112
data xiii stability 118–119
origins and history 1–3 Particle detectors
Nuclear power bubble chamber 132
accelerator driven system (ADS) calorimeters 142–145
327–329  erenkov counter 141–142
C
fast breeder reactor 264–265 cloud chamber 132
thermal nuclear reactors 260–265 drift chamber 135
Nuclear radius 42–42 emulsion 132
Nuclear reactions 62–67 gas detectors 131–136
Nuclear reactors (see Thermal nuclear Geiger-Müller counter 136
reactor and Fast breeder reactor) ionization chamber 133
408 INDEX

Particle detectors (Continued ) Proton

jet chamber 135 decay 303–304
layered detectors 145–148 magnetic moment 7
microstrip gas chamber 136 quark distributions 173–176
multiwire proportional chamber 134 ‘spin crisis’ 316
pixel detector 138 Proton-proton cycle 267–268
proportional counter 133–134
scintillation counter 137–138 QCD
semiconductor detectors 138–139 asymptotic freedom 155, 160–161, 164
silicon microstrip detector 138 colour confinement 155–153, 155
spark chamber 136 energy dependence of strong coupling
spectrometer 34, 139 160–164
streamer chamber 136 exchange of gluons 17, 154–155
Super Kamiokande 79 jets in QCD 165
time-projection chamber 135–136 quantum fluctuations 164
track chambers 134 role of gauge invariance 154
vertex detector 139 scaling violations 177
wire chambers 134 vacuum polarization 164
Particle identification 139–142 QED 162–163
measurement of energy 142–145 Quadrupole moment 232–236
measurement of momentum 139 Quadrupole transition 249–250
measurement of velocity 140–142 Quantum chromodynamics (see QCD)
Particle interactions with matter 123–130 Quantum electrodynamics (see QED)
Particle physics Quark-gluon plasma
data xiii formation 320
emergence from nuclear physics 4 importance for astrophysics 320–323
Parton model 171 signature for 321
Perturbation theory 335–338 Quark model
PET (see Medical imaging) excited states 99–100
Photino 305 mass predictions 105–108
Photodetector 137 magnetic moment predictions 102–104
Photomultiplier tube 137 spectroscopy 96–102
Photon 4 Quarks
interactions 128–130 constituent quarks 90
Physical constants and conversion current quarks 96
factors xv determination of quark charges 172–174
Pickup reaction 63 determination of quark spin 172
Pions 7 distributions in nuclei 176
decays 101, 195 distributions in nucleons 173–176
role in nuclear forces 16–17 evidence from hadron spectroscopy
Planetary model of atoms 2 86–87
Plum pudding model of atoms 2 evidence from jet production 87–88
Positron 9 evidence from lepton scattering 87
Positron emission tomography flavours 89
(see Medical imaging) generations 89
Prolate nuclei 233–234 mixing of quark states 195–197
Prompt neutron emission 256–257 quark numbers 90–91
Propagator 22 sea quarks 201
Proportional counter 133–134 spectator model 90
 INDEX 409

static properties 5, 89 eþ e annihilation 166–168

valence quarks 96, 172 diffraction 43–44
inelastic 14
Rad (unit) 279 neutrino-electron 198–200
Radiation (see Biological effects of radiation) neutrino-nucleon 200–201
Radiation length 128 Mott 38
Radiation therapy 278–280 optical model 43–45
Radioactive dating 48 Rutherford 37
Radioactive decay Schmidt lines 230
activity 47 Sea quarks 96, 172–176
chains 48–50 Selectron 305
constant 47 SEMF (see Semi-empirical mass formula)
discovery 1 Semiconductor detectors 138–139
law 47 Semi-empirical mass formula (SEMF)
Raleigh scattering 129–130 condition for stability against -decay 55
Range-energy relation 127 Coulomb term 51
Range of forces fit to binding energy data 53
from particle exchange 17–19 mass term 50
fundamental interactions 19 numerical values of coefficients 52
strong nuclear force 19 pairing term 52
Yukawa potential 19–20 physical basis of terms 50–52
Range of particle in matter 127 size of terms 54
Relativistic kinematics surface term 51
centre-of-mass system 341 use in analysing -decay 55–59
four-vectors 339–341 volume term 50
frames of reference 341–343 Scintillation counters 137–138
invariant mass 342 Screening 163
invariants 344–345 Shell model
laboratory system 341 applied to atoms 223–224
Lorentz transformations 339–341 configuration 227–228
Relativistic wave equations evidence for 225–226
Dirac equation 8 excited states 230–232
Klein-Gordon equation 8 magic numbers 225–228
Resonance magnetic moments 229–230
Breit-Wigner formula 27 parities 228–229
quark model predictions 99–100 spin-orbit potential 226–227
Rotational states 236 spins 228
Running coupling in QCD 161 Showers 143–145
Rutherford scattering Sievert (unit) 279
classical derivation 349–352 Silicon strip detector 138
quantum mechanical derivation 352–353 Single-photon emission computed
tomography (SPECT) 287–288
Sargent’s Rule 85 Sleptons 305
Saturation of nuclear force 51 Solar neutrinos 82–83
Scaling 170–171 Solar neutrino problem 269
Scaling violations 176–177 Spark chamber 136
Scattering SPECT 287–288
charged lepton-nucleon 168–177 Spectator quark model 90
elastic 14 Spectrometer 139
410 INDEX

Spin dependence of nuclear forces 218 Tandem van de Graaff 113–115

Spin-lattice relaxation 291–292 Tau lepton 71
Spinor 8 Thermal neutron 61, 67
Spin-orbit interaction 226 Thermal nuclear reactor
Spins of nuclei 228 control rods 262
Spin-spin interaction 105 efficiency 263–264
Spin-spin relaxation 292 fuel elements 260–262
Spontaneous fission 209 moderator 261–262
Spontaneous symmetry breaking 209 radioactive waste disposal problem 265
Squarks 305 role of delayed neutrons 262–263
Standard model 5–7 Time-of-flight method 140
STAR detector at RHIC, BNL 148 Time-projection chamber 135–136
Stellar fusion Tokamak 276–277
CNO chain 268–269 Track chamber 134
production of heavy elements 269–270 Transmission coefficient 332
proton-proton cycle 267–268 Truth quantum number 89
solar neutrino problem 269 Tunelling (see Barrier penetration)
Storage rings 119
Strangeness quantum number 89 Unification condition 210
Streamer chamber 136 Units 29
Strings 307 Universality of lepton interactions
Stripping reaction 63 84–85
Strong nuclear force Universe
charge independence 93, 218 critical density 312–313
charge symmetry 93, 218 inflationary big bang theory 312
definition 3
isospin symmetry 93 Vacuum polarization 164
saturation 51, 218 Valence quarks 96, 172
short-range repulsion 217 V-A theory 189
Structure functions 169–177 Van de Graaff accelerator 113–115
Superdeformed nuclei 318–319 Van der Waals force 20, 219
Superheavy elements 319 Vibrational states 236
Supernova 308–310, 316 Virtual process 18
Superparticles 305 Volume term in SEMF 50
Supersymmetry (SUSY)
branes 307 W-boson
detection of superparticles 306 discovery 192–193
Minimal Supersymmetric Standard Model exchanged 183
(MSSM) 305 vertices 193–198
neutralino 306 Weak hypercharge 209
Planck mass 307 Weak interactions
proton lifetime 305 charge conjugation violation 183–185
strings 307 charged currents 183
superparticle properties 305 CP violation in 202–206
unification point 305 Fermi coupling 22
Surface term in SEMF 51 hadron decays 194–198
SUSY (see Supersymmetry) lepton decays 72–73
Synchrotron 116–119 lepton-quark symmetry 195
Synchrotron radiation 117 low-energy limit 22, 182
 INDEX 411

neutral currents 183 X-bosons 302–304

parity violation 183
V-A interaction 189 Y-bosons 302–304
Weak isospin 209 Yukawa potential
Weinberg angle 210, 303, 324 coupling constants 19
Weisskopf approximation 250 due to particle exchange 19–20
WIMPs 313
Wino 305 Z boson
Wire chambers 134 discovery 192–193
W-lepton vertices 193–194 exchange 183
Woods-Saxon potential 226 vertices 210–211
W-quark vertices 196–198 Zino 305