Translated from Russian to English - www.onlinedoctranslator.

com

O.Yu.Shevchenko

Fundamentals of Solid State Physics

Tutorial

Saint Petersburg

2010
MINISTRY OF EDUCATION AND SCIENCE OF THE RUSSIAN FEDERATION

FEDERAL AGENCY FOR EDUCATION

ST. PETERSBURG STATE UNIVERSITY

INFORMATION TECHNOLOGY, MECHANICS AND OPTICS

O.Yu. Shevchenko

FUNDAMENTALS OF SOLID STATE PHYSICS

Tutorial

Saint Petersburg

2010

1
 O.Yu. Shevchenko “Fundamentals of solid state physics.” Tutorial. – St. Petersburg: St.
Petersburg State University ITMO, 2010. – 76 p.

As part of the general physics course, the fundamentals of solid state physics
are considered. The elements of crystallography, an introduction to the band
theory of semiconductors, as well as thermal, kinetic and thermoelectric, and
contact phenomena in solids are presented.

For technical students. Published by

decision of the JNF Academic Council.
Protocol No. 6 of March 2, 2010

In 2009, the University became the winner of a multi-stage

competition, as a result of which 12 leading universities in Russia
were identified, which were assigned the category “National Research
University”. The Ministry of Education and Science of the Russian
Federation approved the Development Program of the state
educational institution of higher professional education “St.
Petersburg State University of Information Technologies, Mechanics
and Optics” for 2009–2018.

© St. Petersburg State University information

technology, mechanics and optics, 2010

©O. Yu. Shevchenko, 2010

2
 Content
1. Crystal structure of a solid……………………...…5
1.1. Crystal cell. Short and long range order. Symmetry of
crystals……………………………………………………………...5
1.2. Crystal lattice defects………………………………….9 a. Point
defects……………………………………………………………...……..9 b.
Dislocations………………………………………………………...…..9 1.3. Chemical
bonding in crystals……………………………………………………… …10
1.4. Thermal vibrations of the crystal lattice…………………....12
A. Heat capacity………………………………………………………...12
b. Thermal conductivity……………………………………………………....14
1.5. Diffraction of waves in crystals……………………………………..15

2. Electronic structure of crystals. Elements of band theory of

solids……………………………………………………..18
2.1. Strong coupling approximation………………………..19
2.2 The weak coupling approximation………………………………………………………
20 A. Free electron gas in metals……………………………...20
b. Periodicity of the crystal. Bloch function. Boundary
Karman–Born conditions. Brillouin zones…………………………..…..21 V.
Rationale for using the weak coupling approximation………….....26
G. Effective mass of an electron in a crystal……………………………26 d.
Mechanical analogy of effective mass………………………....28

3. Statistics of electrons and holes in semiconductors……………….30

3.1. Density of states……………………………………………..30
3.2. Fermi function………………………………………………………..31
3.3. Electron gas concentration…………………………………….33
3.4. Charge carrier concentration and level position
Fermi in semiconductors……………………………………………...35 A.
Proprietary semiconductor……………………………………………………..36 b.
Non-degenerate impurity semiconductor…………………………38 V.
Degenerate semiconductor…………………………………...…….41

4.Kinetic phenomena in semiconductors………………………...42

4.1 Mobility……………………………………………………….….42
4.2. Temperature dependence of electrical conductivity of
semiconductors……………………………………………………….….44 4.3.
Photoconductivity……………………………………………………46

3
4.4. Hall effect………………………………………………………47
4.5. Quantum Hall effect……………………………………………………….51
4.6. Superconductivity……………………………………………………………...55

5. Contact phenomena…………………………………………………………….57
5.1. Metal-semiconductor contact……………………………………57
5.2. Electron-hole transition……………………………………...62

6. Thermoelectric phenomena in semiconductors………………..67

6.1. Seebeck effect……………………………………………………..67
6.2. Peltier effect……………………………………………………..70
6.3. Thomson effect…………………………………………………….…72

Literature………………………………………………………………...……..73

4
Chapter 1. Crystal structure of solids.
1.1. Crystal cell. Short and long range order.
Symmetry of crystals.
Many properties of solids are explained by the fact that most of
them have a crystalline structure, that is, they are characterized by a
regular periodicity in the distribution of elements from one or a
group of atoms (ions) in space at the microscopic level. Therefore,
such a concept ascrystal cell. A group of atoms (lattice nodes)
repeating in space forms basis.The crystal lattice is a collection of
geometric points representing the centers of these bases. Order in
the relative arrangement of atoms in solids can be characterized by
the presence of long- and short-range order caused by the
interaction between particles. Order at distances comparable to
interatomic distances is calledneighbors in orderand orderliness
repeating indefinitely long distances -in a long way.So, for example,
in an ideal gas, the location of an atom at any point in space does not
depend on the location of other atoms, i.e. there is no long-range or
short-range order, but already in liquids and amorphous bodies there
is short-range order - a certain pattern in the arrangement of
neighboring atoms.

In a periodic crystal lattice one can distinguish some primitive

unit cell, which repeats periodically throughout the crystal. By
identifying such a cell, it is possible to describe the positions of atoms
and ions in a substance and, therefore, its
help characterize the structure of crystals. Every
rrr
the unit cell can be specified using three vectorsa1,a2,a3,
called translation vectors, which may not be
orthogonal and have different lengths. Vector corresponding
at some specific lattice point, represents a translation of the form
rrr
R=n1a1+n2a2+n3a3, Wheren 1,n2,n 3 — arbitrary integers.
Translational repetition of these vectors in space allows us to obtain
the so-calledBravais lattices. The Bravais lattice is an infinite periodic
structure that has the same spatial order and orientation regardless
of which node we take as the origin. One of the simplest volumetric
Bravais lattices is the simple cubic lattice shown in

5
 rice. 1.1. Alkali halide compounds, such as
NaCl, KCl, LiF, etc., crystallize in the form of
a simple cubic lattice. At the same time,
alkali metal ions (for example, Na+)
alternate at lattice sites with halogen ions
(for example, Cl-). Another example is a
face-centered cubic lattice (Fig. 1.2left). It
differs from a simple cubic lattice by
additional nodes located in the center of
each face. This is how some metals
crystallize - copper, silver, gold, aluminum,
Fig.1.1. Simple lead.
cubic lattice Typical
Brave.
germanium semiconductors Andsilicon
crystallize in a latticediamond type
(Fig. 1.2 on right). Her can be represented as two
face-centered lattices, shifted relative to each other by a quarter of the
diagonal of the cube.

Fig.1.2. Face centered cubic lattice (left), diamond type grating (on right).

The unit cell can be defined in other ways. For example, in the
form of a cellWigner-Seitz,used when considering the movement of
particles in a crystal. To select a cell, connect a given lattice node with
the nodes closest to it and draw planes perpendicular to the
segments through the middles of the resulting segments. The
resulting polyhedron, reflecting the symmetry of the crystal and
containing one node of the crystal lattice, forms the Wigner-Seitz cell.
The Wigner–Seitz cell is built in the so-calledreciprocal space(reverse
grid), setting aside

6
along the coordinate axes are segments opposite to the translation
vectors in a given direction. For a simple cubic lattice, the Wigner–
Seitz cell is identical in shape and size to the elementary cell, that is, it
is a cube. For more complex lattices, the Wigner–Seitz cell has a more
complex shape. For example, a diamond-type lattice in this case is a
fourteen-sided structure, six of whose faces are square and eight are
hexagonal (Fig. 1.3).

Digital symbols in Figure 1.3.

are calledindexes
Miller[h,k,l]. This is the totality
three integers characterizing the
crystallographic direction. If any
atom of the crystal lattice is taken
as the origin in the Cartesian
coordinate system and connected
to other nearby lattice nodes, then
the corresponding lengths of the
segments, expressed in units of
lattice constant, are related to the
Miller indices by an inverse
relationship Fig.1.3. Wigner–Seitz cell of a
diamond-type crystal lattice.
1 1 1
h:k:l= : : (1.1)
x y z

To describe the symmetry of the crystal lattice, the concepts are

usedsymmetry transformationAndsymmetry elements. A symmetry
transformation is a generalized movement that combines a figure
with itself, leaving it invariant. From it follows the concept of
symmetry elements (axes, planes, centers of symmetry), the set of
which forms point and translation symmetry groups. Figure 1.4
shows, as an example, the symmetry elements of a simple cubic
lattice. The cubic lattice has three planes of symmetry parallel to the
faces (Fig. 1.4A), six diagonal planes of symmetry perpendicular to
the diagonals of the faces (Fig. 1.4b, c), three fourth-order axes (Fig.
1.4d), six second-order axes (Fig. 1.4G), four third-order axes (Fig. 1.4e
) and the center of symmetry in the center of the cube (Fig. 1.4Where
). There are a total of 14 types of different Bravais spatial lattices,
from which 32 point symmetry groups can be obtained. These are all
possible

7
combinations of symmetry elements for crystals. 32 types of point
symmetry, having similar symmetry elements, are combined into
system.

Fig.1.4. Symmetry elements of a simple cubic lattice.

There are seven systems in total (triclinic, monoclinic, rhombic,

tetragonal, trigonal, hexagonal, cubic). Depending on the order of the
axes of symmetry, systems are combined into categories. Thus, if we
subject Bravais lattices to symmetry transformations (translation,
reflection, inversion), then all known crystal structures can be
constructed. This is what crystallography does. The considered
classifications of crystal structure are accepted in group theory, which
allows one to mathematically describe the symmetry of any crystal.

There is a way to view crystals not only as lattices, but also as

dense packings of spheres. Atoms of different types correspond to
spheres of different sizes. At the same time, the concept of
coordination number is introduced - the number of atoms located in
the immediate environment of a given atom. The highest
coordination number is for crystals with the so-called close packing, it
is 12, the lowest is for covalent crystals (4). Due to the unequal
density of atoms in different directions of the crystal, different
properties are observed. This phenomenon is called anisotropy.

8
1.2. Crystal lattice defects.
Perfect crystalline lattice is yourself
repeated repetition of elementary cells. In a real crystal there are
always defects, violating periodicity
the arrangement of atoms in the crystal lattice and having a significant
impact on the properties of the material.

a).Point defects. Point defects include: a) free spaces at the nodes of

the crystal lattice - vacancies, called Schottky defects (Fig. 1.5A); b)
atoms that have shifted from the nodes of the crystal lattice into
interstitial spaces - dislocated atoms,

called Frenkel defects

(Fig. 1.5b); c) atoms of
other elements located
both at the nodes and
interstices of the crystal
lattice
- impurity atoms.

Spot defects
Fig.1.5. Point defects in crystals: a) vacancies,
are formed V process b) dislocated atoms.
crystallization under

exposure to thermal, mechanical, electrical influences, as well as

irradiation with neutrons, electrons, and x-rays.

b). Dislocations. Let's consider

linear, surface and volumetric
disturbances frequency
crystal structure, that is, dislocation.

Linear dislocations - This

violations crystalline
structures in one direction. There are
edge (Fig. 1.6) and screw (Fig. 1.7)
dislocations.

Regional dislocations connected With Fig.1.6. Edge dislocation.

presence "extra" atomic

9
half-planes in the crystal, and screw ones - with a partial shift of one
half-plane of the crystal lattice relative to the other.

Superficialdislocations are defects

that extend significantly in two
directions. These include
boundaries between grains,
interphase boundaries, packing
defects in the crystal lattice,
accumulation of dislocations in one
plane, etc.
Fig.1.7. Screw dislocation.
Volumetric defects have
length in all three
measurements. These defects include a set of point, linear and
surface defects that lead to distortion of the crystal lattice in large
volumes of the crystal, microinclusions of other phases, pores and
cracks.

1.3. Chemical bond in crystals.

Let's consider the question of what holds the atoms together in a
crystal. The connection between atoms is ensured by the Coulomb
interaction between charged particles: cations and anions, nuclei and
electrons. When atoms come together, attractive forces begin to act
between the nucleus of one atom and the electrons of another, as
well as repulsive forces between nuclei and between electrons. At
some distance, these forces balance each other, and a stable
chemical particle is formed. The existence of stable bonds between
atoms in a crystal suggests that the total energy of the crystal is less
than the total energy of individual atoms in the crystal by the amount
of the so-called chemical bond energy (or simply bond energy). Its
value differs greatly for different elements of the periodic table. For
example, for inert gases it is only hundredths of the binding energy
of typical semiconductors (Si, Ge). And crystals of alkali metals have
intermediate values of binding energy.

There are three types of chemical bonds in solids:ionic

(heteropolar),metalAndcovalent(homeopolar). At

10
In the formation of a chemical bond, a significant redistribution of the
electron density of atoms in the compound can occur in comparison
with free atoms. In the limiting case, this leads to the formation of
charged particles - ions. These ions form a crystal lattice due to the
fact that the Coulomb attraction between ions of opposite sign is
stronger than the Coulomb repulsion of like-charged ions. This
creates an ionic chemical bond. The most typical ionic crystals include
alkali halide compounds (for example, NaCl, KCl), which crystallize in
the form of a simple cubic lattice (Fig. 1.1).

Covalent bond- this is the main type of chemical bond,

characterizing the formation of a solid body. As a result of a covalent
bond, the electrons of neighboring atoms are shared, as a result of
which each atom has a completely filled eight-electron valence shell.
The covalent bond has a certain directionality and is characteristic of
most semiconductors (Si, Ge). Thus, each silicon atom lacks four
electrons before filling its electron shell, so atoms of neighboring
elements can be attracted by overlapping electron shells. As a result,
the atom is placed at the center of the tetrahedron and surrounded
by four nearest neighbors (Fig. 1.2.on right)

If, according to the type of chemical bond, crystals with covalent

and ionic bond types can be considered as limiting cases, then
between them there are crystals with intermediate bond types that
have a certain degree of bond ionicity. In this case, atoms with almost
filled shells (Na, Cl) tend to have ionic bonds, and elements of groups
III, IV and V (In, C, Ge, Si) tend to have covalent bonds.

Exist crystals, characterized by high

electrical conductivity caused by the presence of free electrons that
can move and participate in conduction. Such a chemical bond,
caused by the interaction of free electrons shared by the crystal with
the core of positively charged ions of the crystal lattice, is calledmetal.
The interaction of ionic cores with conduction electrons affects the
binding energy in metal crystals. This relationship is characterized by
a decrease in the kinetic energy of electrons in the crystal compared
to the kinetic energy of electrons in

eleven
free atom. Metal crystals tend to crystallize into close-packed
structures, for example, in the form of a face-centered cubic lattice
(Fig. 1.2left).

1.4. Thermal vibrations of the crystal lattice.

A. Heat capacity.
The thermal properties of crystals are mainly due to vibrations
of the crystal lattice. At a temperatureT=0 according to the classicaloK
theory, atoms can be considered motionless “fixed” in the nodes of
the crystal lattice. AtT>0oKThey oscillate around equilibrium positions.
Within the framework of classical theory, in an ideal monatomic gas,
each atom has
three degrees of freedom and has a kinetic energy of 3kT. If this
2
apply the theory to solids, then the internal energy of one mole of
crystal is:

E=3NAkT=3RT (1.2)

J
WhereR=8.31 is the gas constant, and the molar heat capacity
moleK
(the amount of heat required to heat one mole of a substance by 1K)
at constant volume is equal to:

dE
Cv= =3R (1.3)
dT
This relationship, called Dulong and Petit's law, is satisfied by
most solids at room temperature. However, as the temperature
decreases, the heat capacity decreases rapidly
and approaches zero at T→0oK. The reason for this is due to
quantum effects. According to quantum theory, vibrations of the
crystal lattice can be excited and absorbed only in portions (quanta).
By analogy with photons, quanta of vibrational energy are called
phonons.The energy of sound waves is quantized in the same way as
the energy of photons and electrons. Phonons mpossible
be considered as quasiparticles with energyhνand impulsehk, Where
k=ω,υ– speed of propagation of vibrations in the crystal. At
υ
low temperatureskT<<hν, i.e. there are still few phonons. WITH

12
As the temperature increases, many low-frequency phonons appear,
but so far few high-frequency phonons appear. And only at high
temperatures, whenkT>hνeven for high frequency waves, the
discreteness of energy no longer matters, and law (1.3) is satisfied.

A mathematical calculation of the total phonon energy as a

function of temperature, carried out by Debye in 1912, led to the
following expression for the heat capacity of any solid:
3θ
⎛T⎞TeXx4dx Cν=9
R⎜ ⎟∫ (1.4)
⎝θ⎠ 0 (ex−1)2

hνD
Whereθ= – characteristic Debye temperature, different for
k
different solids, D – frequency of the wave of the shortest length.
Dependence of heat capacityWITHvfromT/θshown in the figure (1.8), where
For comparison, experimental values for silver and aluminum are
given. For all solids, the theory agrees with experimental data.

Rice. 1.8. Heat capacity at constant volume as a function of temperature. The solid
curve is calculated using the Debye formula (1.4.3). Experimental points for silver
and aluminum are plotted on the curve to determine
Debye temperature, after which the graph was rebuilt into functionsT
Q.

From Figure (1.8) it is clear that law (1.3) is well satisfied when T>
θ, but not applicable whenT<θ. Thus, experimental dependence of
heat capacity on temperature for solids

13
confirm the applicability of quantum mechanics to the description of
acoustic processes in solids. In the case of metals, heat capacity due to
vibrations of the crystal lattice is added to the heat capacity due to free
electrons. However, this additive can only be detected at a sufficiently
low temperature. For the simplest metals, good agreement between
theory and experiment has been found.

b. Thermal conductivity.

Heat in all solids is transferred by lattice vibrations. Heat

transfer by lattice vibrations occurs as follows. Each atom oscillates
around its equilibrium position with an amplitude determined by
temperature. As the temperature increases, the amplitude of its
vibrations increases, therefore, the periodic force with which it acts
on neighboring atoms increases. If a temperature difference is
created in the crystal, then heat transfer will occur from the more
heated end to the less heated end, due to the diffusion of phonons.

Thermal energy flowQpassing through the cross section crystal per

unit time in the direction of the axisxis equal to:

dT
Q=λ (1.5)
dx
Whereλ– thermal conductivity coefficient, when calculating which it is
necessary to take into account the interaction between phonons, i.e.
phonon – phonon scattering due to anharmonicity of lattice
vibrations. That is, in an ideal crystal, atoms would have to oscillate
strictly periodically, emitting elastic waves that transfer heat. These
waves would not interact with each other, propagating at the speed
of sound. In a real crystal, there are density fluctuations and a wide
variety of defects, which we talked about earlier. This leads to
scattering of elastic waves both on lattice defects and when
encountering each other. By analogy with the kinetic theory of gases,
it can be shown that

u⋅l⋅c
λ= v
(1.6)
3

14
WhereAnd– phonon group velocity modulus,l– average length free
path of phonons between two successive collisions,cv– heat capacity
per unit volume.

For dielectrics at high temperatures, phonon–phonon scattering

predominates, and the mean free path should be inversely
proportional to the phonon concentration, which, according to Bose–
Einstein statistics
hν

n= (ekT−1)−1 (1.7)

at high temperaturesn~T. AtT>θas shown above,cv

1
practically independent of temperature, therefore,λ~ that
T
is in qualitative agreement with experiment.

AtT<θfree path determined by phonon phonon scattering,

increases sharply and is practically independent ofT. Thermal
conductivity in this region is entirely determined by the dependence
onTheat capacity of the crystal. In metals, heat transfer takes place
also conduction electrons. In the general case, for a metal the
thermal conductivity coefficient is equal to the sum of the lattice
phonon and electronic components:

λ=λ uh+λ f (1.8)

The contribution of electrons to the thermal conductivity of metals

is approximately twice the contribution of the lattice, so pure metals
conduct heat much better than non-metals. For the simplest metals,
good agreement between theory and experiment was found. The
phenomenon of heat transfer in semiconductors is more complex than
in dielectrics and metals, since both mechanisms of thermal conductivity
can be significant in them
uh f λ≈λ, which occurs at relatively high

charge concentration. It is also possible that the dielectric mechanism of

thermal conductivity predominates whenuh λ<λ. f

1.5. Wave diffraction in crystals.

Typically, to study the structure of crystals, X-ray diffraction is used,
for which the wavelength is comparable to

15
typical interatomic distances (about 10 times the study of crystals
− 8cm). For

requires X-ray radiation with a quantum energy of 10 to 50 KeV. In

1913, W. L. Bragg showed that in crystalline substances for some
certain wavelengths and directions of incident rays diffraction
maxima are observed when they are reflected from parallel planes of
ions separated from each other by a distanced(rice.1.8).

For the beams to interfere with gain, the path difference must be an
integer number of wavelengths, which leads to the conditionWulf -
Bragg
2dsinθ=mλ (1.9)

Whereθ– sliding angle,m=1,2,3... – diffraction order.

Let's explain meaning

introduced previously (clause 1.1)

concepts reverse grates.

Reverse grille Can
analyze in the same way as
crystalline grate,
look at the nodes in it
[h,k,l], important directions,
planes, coordinates of points,
etc. You could say What
Fig.1.8. Move rays at reciprocal lattice displays
observation diffraction V diffraction picture from
crystals. three-dimensional lattice. The
reciprocal lattice is rigidly
connected to the crystallinelattice
crystal, at turning
crystal together Withhim
turns and reverse
lattice. For observations
diffraction crystal
rotated so that the scattering
vector coincides with one of
Fig.1.9. Construction spheres the inverse nodes grates.
Ewald.
Predict And clearly
depict This Canrc
by constructingEwald(Fig. 1.9). Let us set aside the wave vectork 0

wave incident on the crystal, so that its end coincides with the node[0,0,0]

16
reverse grid. Since frequency and speedr scattered and
incident waves coincide, scattered wave vectork 1will have that
r
same length ask0, but an uncertain direction, then its convenient
r
depicted as a sphere (Ewald sphere) with the center at the beginning of the vectork.
rr 0

From Fig. 1.9. it is clear that between the lengths of the vectorsk0AndGexists
connection:
rr
G=k0⋅2sinθ (1.10)
r
One can find the set of all wave vectorsk 0 waves and particles,
corresponding to the condition of diffraction on the crystal. Equation (1.10) can be
rewrite as:
rr rr r2 rr
G=k0⋅2cos(Gk0) or G =2⋅(Gk 0) (1.10A)
r
The latter is an equation (relative to k0) for a plane
r
perpendicular to the vectorGand located at a distance from the origin of coordinates
r
G
distance . It also describes the possible coordinates of the vector
r 2
k0, satisfying the diffraction condition. Then there are many endings
r
vectorsk0, corresponding to the diffraction condition, lies on the planes
passing through the midpoints of all reciprocal lattice vectors and
perpendicular to them. It is in this way that the Wigner–Seitz unit cell
is constructed. The Wigner–Seitz cell constructed in reciprocal space
is usually called the firstBrillouin zone. It has an important property:
waves and particles whose wave vector is located on its boundary
satisfy the diffraction condition. Brillouin zones play an important role
in considering the movement of electrons, phonons and other
particles in a crystal and in analyzing energy bands in crystals.

17
 Chapter 2. Electronic structure of crystals.
Elements of band theory of solids.

Zone theory solid body allows Withsingle

quantum mechanical point of view explain existence
metals, semiconductors and dielectrics, their electrical properties. The
band theory is based on the formation and filling of allowed and
forbidden energy bands with electrons, taking into account the Pauli
principle and elements of statistics.
To determine the energy states of the electron within the framework of
the band theory of solids, a nonrelativistic solution is solved.the equation
Schrödinger

2m(E−V(r))ψ(r) h
Δψ(r) = 2
(2.0)

Where (r) is the electron wave function.

Accurate accounting of potential energy Coulomb

interaction of an electron with the electric field of a crystal latticeV(r)
leads to the fact that the solution to the Schrödinger equation
reduces to a many-particle problem. As a result, we obtain a system
of equations of order 1023, which is impossible to solve even with
modern computer methods, so various approximations are used to
solve this problem. For example,
the adiabatic approximation is based on the fact that mass of the nucleus in an atom

much greater than the mass of the electron, so the speed of the
nucleus is approximately two orders of magnitude lower than the
speed of the electrons. This allows, as a first approximation, to
consider only the motion of electrons in the field of stationary nuclei.
Another, more common approximation is the one-electron, or Harty–
Fock, method. In this model, the energy of pairwise interaction of
electrons in a crystal lattice is replaced by the interaction of one
electron with the averaged self-consistent field of all other electrons.

We will consider two approaches to considering the energy states of an electron

in a crystal - the approximationsstrong and weakcommunications.

18
2.1. Approaching strong connection.

Within the framework of the strong binding approximation, it is

assumed that electrons are in potential wells of their atoms (they are quasi-
bound). If the electron energy is less than the height of the potential barrier,
then from the classical point of view such a barrier is not permeable to the
electron, but from the point of view of quantum theory there is a probability
of an electron passing through a potential barrier of finite width due to the
tunneling effect. Let us consider the formation of a crystal within the
framework of the tight coupling approximation (Fig. 2.1).

At first N atoms,
are located far from each other,
and their energy levels are 2N-
fold degenerate. When atoms
come together into a crystalline
grille
degeneration energy
levels are removed. Discrete
energy levels atoms
interact with each other and
split into energy zones. The
interaction between atoms can
lead to the overlap of two Fig.2.1. Splitting of energy levels when
separate zones, in which case atoms approach each other.
one will get one
zone with the number of electronic states equal to the sum of these states in
the overlapping zones. And in this case, the number of electronic states will be
a multiple of 2N.
An energy zone or a set of energy zones formed as a result of
the splitting of one or more energy levels of isolated atoms when
combining them into crystals is calledpermitted zone. The allowed
bands include the valence band and the conduction band. On band
diagrams, the designation of the bottom of the conduction band is
also accepted -Eand ceiling C

valence band –EV(rice.2.2).

Within the framework of the tight-binding approximation,
bound electrons of the valence band are considered, which can,
having overcome the potential barrier, become free electrons of the
conduction band. This process is called generationfree
charge carriers (transition G in Fig. 2.2).

19
 As a result of such transitions,
free energy levels appear in the valence
band, the absence of electrons
on which should
interpret How presence on them
fictitious charges - holes. Transition of
electrons from the conduction band to the
valence band zone called
recombination(transitionR in Fig. 2.2).
The range of energy values that an
electron cannot possess is called
Fig.2.2. Zone scheme
prohibited zone. Width
forbidden zoneis the energy difference
semiconductor.
between the lower level of one
permitted zone and the upper level
neighboring prohibitedEg=EC−EV. The wider the band gap, the
fewer electrons are able to overcome it. The height of the potential
barrier corresponds to the width of the band gap. Based on the band
gap, solids can be classified into metals, dielectrics and
semiconductors. If the bandgap widthEg>2÷3 eV, then the solid is
calleddielectric. At
At absolute zero temperature, the dielectric has a completely filled
valence band and an empty conduction band. In the case when the
valence band is partially filled, or there is an overlap of bands, while
there is no band gapEg→0, such a rigid body
ismetal. Occupying an intermediate position semiconductors. The
difference between semiconductors and dielectrics is purely arbitrary.
We can say that semiconductors are dielectrics withEg<2 eV, for which
at low temperature all states are in
The valence band is filled with electrons.

2.2. Weak coupling approach.

A.Free electron gas in metals.
To describe the behavior of electrons in a solid within the
framework of the classical electronic theory of metals, Drude–
Lorentz, the concept of free electron gas is used, when each metal
atom gives up one valence electron for common use. These free
electrons are like an ideal gas

20
non-interacting and sizeless molecules from molecular physics. In
this case, the Coulomb repulsion by all other electrons of the metal is
balanced by the attraction by the nuclei (ions) of atoms. Free
electrons participate in chaotic (thermal) motion and collide with ions
of the crystal lattice, which leads to the establishment of thermal
equilibrium between the electron gas and the crystal lattice. In this
case, the average speed of thermal motion of electrons is similar to
the average speed of molecules of an ideal gas. Thus, it is possible to
consider an electron gas by analogy with an ideal one only if the
conditions of thermodynamic equilibrium are met.

If an electron moves at a speedυ, then its impulse p=

mυ=hk. (2.1)

Electron de Broglie wavelength λ

=h mυ (2.1A)

Kinetic energy of a free electron

h2k 2
E(k) = (2.2)
2m
This formula expresses the dependence of kinetic energy on
the wave number, which is called the dispersion dependence
(dispersion law). For a free electron, the dispersion law is quadratic.
But the movement of an electron in a crystal cannot be considered
free. The electron in this case will bequasiparticle, having certain
values of the main parameters – quasi-pulse, quasi-wavenumbers
andeffective mass, different from the rest mass of the electron.

b.Periodicity of the crystal. Bloch function. Boundary

Karman-Born conditions. Brillouin zones.

Within the framework of the weak coupling approximation, the

state is described quasi-freeelectron moving in a periodic field crystal
lattice. Since in an ideal crystal the atoms are located at the nodes of
the crystal lattice, forming a strictly periodic structure, the potential
energy of the electronV(r)
also depends on spatial coordinates, i.e. has translational symmetry:

V(r) =V(r+a )
i (2.3)

21
Wherer– electron radius vector,a i(i=1,2,3) - translation vectors.
The solution to the Schrödinger equation (2.0) in the case of a
periodic potential field isBloch function k(r)

k (r) =u(rk )⋅eikr (2.4)

Where uk(r) reflects the periodicity of the lattice potential,k– wave

vector of an electron in a crystal. It can be noted that whenV(r) = 0
equation (2.0) has a solution in the form of a plane wave
k (r) =A⋅eikr (2.4A)

WhereA– normalization factor. Wave function (2.4A) describes the

motion of a free electron with momentum (2.1) and energy (2.2). Let us
consider in more detail the boundary conditions for solving the
Schrödinger equation, related to the periodicity of the arrangement of
atoms in the crystal lattice. Let the crystal have the shape of a
parallelepiped with dimensionsL x,Ly,L zalong the axesx,y,zrespectively.
For a cubic lattice with perioda Lx=aN
x,Ly=aNy,Lz=aNz (2.5)

WhereN x,Ny,N zis the number of atoms on the corresponding edges of the crystal.

A consequence of the periodicity of the arrangement of atoms in a crystal for

the wave function are cyclic boundary conditions
Karmana–Borna
(x+Lx,y+Ly,z+Lz) =ψ(x,y,z) (2.6)

and taking into account (2.4)

(x+Lx,y+Ly,z+Lz) =uk(r)e e
ikr ikxLx
eik L eik L
y y z z
(2.7)

to satisfy conditions (2.6) and (2.7) it is necessary that e

ikxLx=e ikyLy=e
ikzLz=1

this is only possible if kxLx=2πn1;k

L =2πn2;kzLz=2πn3
y y (2.8)

Wheren1,n2,n3- whole numbers.

Hence,
2π 2π 2π
kx= n1;ky= n2;kz= n3 (2.9)
Lx Ly Lz

22
 r
Thus, the components of the wave vectorkchange
discretely, therefore, the components of the impulse change discretely
rr
p=hk, and in accordance with this, the energy of electrons in the allowed band
turns outquantized. r
From formulas (2.5), (2.9) we can obtain the limits of changek

π π
− ≤k x< ( 2. 0)
1
a a

π π
− ≤k y < (2. eleven)
a a

π π
− ≤k < ( 2.12)
a a
z

Wherekx,ky,kz Nx,Ny,Nz values accordingly.

accept Then in
in the permitted zone there is onlyN=Nx⋅Ny⋅N energy
rz
states corresponding to different valuesk, i.e. equal to the number
atoms in a crystal. Taking into account the Pauli principle allows us to
conclude that in the allowed zone there can be 2Nquantum states
that confirms the earlier assumption. If similar to the usual
coordinate space (x,y,z) consider space
quasi-wave vectors(k x,ky,k z) or spacequasi-pulses

(px,py,pz), then from the formulas (2.10) – (2.12) it is clear that these spaces
arereversewith respect to coordinate space. The values of the quasi-
wave vector (quasi-momentum) in the intervals determined by
formulas (2.10) – (2.12) in the space of the quasi-wave vector (quasi-
momentum) correspond to a certain region called first Brillouin zone.
The first Brillouin zone is a Wigner–Seitz cell in reciprocal space. The
principle of its construction is discussed in paragraph 1.1. Figure 1.3
shows the first Brillouin zone for typical semiconductors - silicon and
germanium.
The second (next) Brillouin zone consists of two sections: fromπbefore
a
2π
a and from − 2π
a to −π. The
a subsequent
Brillouin zones. Each Brillouin zone contains all possible values of
quasi-momentum. All Brillouin zones give states,

23
physically equivalent to the states of the first Brillouin zone, and the
entire band diagram of the crystal can be represented within one -
givenBrillouin zone (Fig.2.3).
In the reduced Brillouin zone, mainly the two upper allowed zones
are used -valence band and conduction band,
since free charge carriers are located in
these zones. Near the bottom of the
permitted zonedispersion lawE(k)
may be close toquadratic(Fig. 2.4), however,
in most cases the dispersion dependences
have complex
character.
So way,
zone Brillouin
convenient For
characteristics
crystals, because
she reflects
Fig.2.3. E(k) For symmetry
different energy crystals and at
zones in the reduced this allows
Brillouin zone.
get All
possible quasi-momentum values for a
given crystal.
A graphical solution of equation (2.0) in the Fig. 2.4. Parabolic
one-dimensional case allows us to obtain law variances E(k)
the dispersion dependence E(kx), valence band and
shown by dark solid lines in Fig. 2.5. Unlike conduction band near
k=0 .
free movement, in
periodic field of the crystal latticeV(r)the equation (2.0)
does not have solutions for all energy values. Areas of allowed
energies alternate with zonesforbidden energies.In the weak
coupling model, this is explained by the Bragg reflection of electron
waves in the crystal. Let's consider this issue in more detail. Let the
electron wave be reflected from the systemmetal flatstey, lane
pendicular reciprocal lattice vectorG, in the opposite directionk0= −k1.

24
 Then amplitudes falling
(2.13) and reflected (2.14) waves
k(r) =uk(r)⋅e
ikr
(2.13)
(r) =u(−rk)⋅e−ikr
−k
(2.14)
will be the same
uk(r) =u−k(r) (2.15)
These waves give two solutions to the
Schrödinger equation:
1 (r) = 2uk(r)⋅cos(kr) (2.16)
2(r) = 2iuk(r)⋅sin(kr) (2.16A)
According to quantum
mechanics, probability density of
finding a particle at a pointr equal to
the square of the modulus of the wave
Rice. 2.5. AddictionE(kx) For
function. Hence,
permitted energy zones. charge distribution density in the
first case, according to the formula
(2.16) is equal to:

ρ1(r) = ψ 1(r)2= 4uk2(r)cos2(kr) (2.17)

This function describes the accumulation of a negative charge

1–smallest.
on positive ions, where the potential energy Similarly,Efrom
formula (2.16a) we obtain:

ρ(2r) =ψ(r)22= 4u2(r) sin (2.17A)

k 2(kr)

This function describes a distribution of electrons in which they are

located predominantly in areas corresponding to the middle of the
distances between ions. At the same time, potential energy E2there
will be more. Functions 2energy will correspondE 2>E 1.
r
Thus, alone and same meaningkon zone boundaries
Brillouin corresponds to two energy values, i.e. at the boundaries of the
zones the energy is disruptedEg=E2−E1, leading to the formation
forbidden zones widthEg. EnergyEdefines the top
1

the boundary of the first zone, and the energy2E– the lower boundary of the second zone.
This means that when electron waves propagate in crystals, energy
ranges arise for which there are no solutions to the Schrödinger
equation that have a wave nature. Thus, the free electron
approximation, like the approximation

25
strong connection, leads to the emergence of allowed and forbidden
energy bands.

V. Rationale for using the weak coupling approximation.

Let us consider how justified the described approximation is. As
is known from quantumrth mechanics, movement of a free electron
with wave vectorkcan be described using a wave packet, which is a
superposition of plane waves with continuously changing valuesk
ranging fromk− Δkbeforek+Δk
k+Δk
(x,t) =∫ai(kx−kωte)dk (2.18)
k−Δk

The motion of a wave packet is characterized by group velocity

dω
υgr= , equal to the speed of movement, for example, its maximum.
dk
dω
Let's find the coordinate of the maximum from the conditionx− t=0 :
dk
x dω
= =υ=υ gr (2.19)
t dk
Consequently, the speed of the electron is equal to the group speed
of the wave packet. We found that the motion of an electron in a
crystal is described by the wave function (2.4), which is a set of wave
functions with different valuesk. So the movement An electron in a
crystal can be described by a wave packet composed of Bloch wave
functions, the group velocity of which is equal to the average speed
of the electron in the crystal. This important conclusion gives
quantum mechanical meaning to the effective mass method
discussed below.

G. Effective mass of an electron in a crystal.

Movement of a free electron under the influence of an external forceF
obeys Newton's second law, however, the peculiarities of the
movement of electrons in a crystal are determined by their
interaction with the crystal lattice. The behavior of an individual
electron in a crystal can be described by the same equation as for a
free particle, if we apply the effective method masses. For free people
electrons
υ=hk m (2.20)

26
Let us determine from (2.2) by differentiation: dE
2m
dk=kh (2.21)

After this, substitutingkV (2.21), we get

1dE
υ= (2.22)
hdk
Let us further assume that the energy of the electron changes some
under external influence. Then we can write
dE=Fυdt (2.23)
WhereF– external force leading to a change in energy. Substituting
(2.22) into equation (2.23), we have
F dE
dE= dt (2.24)
hdk
From here we get
dk 1
= F (2.25)
dt h
From equation (2.22) we also have
dυ 1d 2E dk
= (2.26)
dt hdk 2dt
taking into account (2.25)

dυ 1d 2E
= F (2.27)
dt h2dk 2
If we introduce the notation for the effective electron massm*,
1
m* = (2.28)
1d 2E
h2dk 2
then equation (2.27) is Newton’s law connecting external force, mass and
acceleration. For free electrons, formula (2.28) taking into account
(2.2) givesm*=m.
Let's look at Fig. 2.4. Near the bottom of the allowed zone, the sign of
the second derivatived2E dk2is positive, hence the effective mass
electronm*> 0 and is comparable in order of magnitude to the mass of a
free electron. But near the middle part of the allowed zone there is an
inflection of the curveE(k) , Sod2E dk2goes to zero, and the quantity
m*→∞. Then the effective mass changes sign and becomes the quantity

27
negative. This can be understood if, instead of an electron, we
consider a hole (electron vacancy), which has a positive effective mass
and elementary charge, but moves in the opposite direction to the
electron. From Figure 2.4 it is clear that near the extremum points the
curveE(k) almost goes horizontal, and in
⎛ π π⎞
extremum points ⎜k= − , ⎟ valuedE
⎝ a a⎠ dk =0 . Therefore, with
taking into account formula (2.22), the electron velocity at the boundaries
of the Brillouin zone is zero. This corresponds to Bragg reflection of
electron waves in the crystal.
In the general case, the effective mass of an electron is a tensor of the
second rank
d 2E d 2E d 2E
, ,
dkx dkydkx dkzdkx
2

1 d 2E d 2E d 2E
(m ) =
* −1
, , (2.29)
h2 dkxdk y dky2 dkzdk y
d 2E d 2E d 2E
, ,
dk xdk z dk ydk z dkz2

The effective mass of an electron moving in a certain crystallographic

direction depends on the direction of motion of the electron in the
crystal and is determined through a specific component of the tensor
(2.29). Experimental determination of the effective mass of an
electron when moving in various crystallographic directions in a
crystal is carried out by various methods, for example, optical, in
studies of thermoelectric and galvanomagnetic phenomena, using
the field effect or using cyclotron resonance.

d. Mechanical analogy of effective mass.

As a mechanical analogy for the effective mass of an electron,

we can consider the effective mass of a solid ball located in a liquid
whose density increases from top to bottom. Consider the motion of
a ball with massmin a vessel filled with two liquids of different
densities (Fig. 2.6). Let the density of the liquid vary from
ρr at the top of the vesselρ 2below. Gravity
r
1

P=ρgV (2.30)

28
Whereρ– density of the ball material.
The powers of Archimedes
r
F=A ρ gVand (2.31)

Whereρand– liquid density. According to

Newton's second law

P−FA=ma (2.32)

If we consider only external forces, then P= Fig.2.6. Solid

m *
a (2.33) ball in liquid.
From formulas (2.32) and (2.33) we obtain
P
m*=m (2.34)
P−F A

or taking into account (2.30), (2.31)

ρ
m*=m (2.35)
ρ−ρ and

Consequently, the effective mass of the ball determines the action of the
internal force on itFand
A changes similarly to the effective mass
electron. Atρ>ρand, that is, in the upper part of the vessel,m*> 0, in the
middle partm→ ±∞ , and in the lower part, whereρ<ρand,m<0
*
.

29
 Chapter 3. Statistics of electrons and holes in
semiconductors

In a state of thermodynamic equilibrium, for a given crystal

sample at a given temperature, there is a certain energy distribution
of electrons and holes, as well as their concentration values.
Calculating the concentrations of majority and minority charge
carriers is the main task of statistics of electrons and holes in crystals.
The problem under consideration falls into two parts: a purely
quantum mechanical one - finding the number of possible quantum
states of electrons, and a statistical one - determining the actual
distribution of electrons over these quantum states at
thermodynamic equilibrium.

3.1. Density of states.

The Heisenberg uncertainty relation characterizes the region of

spatial localization of a microparticle (dx,dy,dz) at
given interval of pulse projections (dp x,dpy,dp z) and maybe

brought to mind
dxdydzdp xdp ydp z≥h 3
(3.1)

where is the minimum value of the resulting product

dx0dy0dz0dpxdp ydp z =h
0 0 0 3
(3.2)

characterizes elementaryphase cellin six-dimensional coordinate-

momentum spacedxdydz dpxdpydpzwhich can
occupy particle.
Let's find the number of cells in any volume of this spacedZ:
dVdpxdpydpz
dZ = (3.3)
h3
WheredV=dxdydz– volume in ordinary coordinate space.
Then the number elementary phase cells for unit ordinary
volumedz
dZ dpxdp ydp z
dz= = (3.4)
dV h3

thirty
Let's find the valuedzin the spherical
layer momentum space limited by
spheres with radii pAnd p+dp
(Fig.3.1):
dpx dpy dpz =4πp 2dp (3.5)
Let us take into account that the dispersion
law for free particles is
quadratic parabola, that is
1
⎛ m⎞ 2
p2= 2mE,dp=⎜ ⎟ dE. Then the number
⎝2E⎠
Fig.3.1. Spherical layer in
elementary phase cells V
momentum space
unit of normal volumedz

32
2π(2m) 1
dz= E 2dE (3.6)
h3
Density states N(E) called number elementary
phase cells in unit normal volume per
unit energy interval:
dz 2π(2m)
32
1
N(E) = = E 2
(3.7)
dE h 3

From the formula it is clear that the dependence

of the density of statesNfrom energyE
is a power function (Fig. 3.2). From
Figure 3.2 it is clear that for a fixed
interval dE
density of states grows with growth
energyE, which corresponds to the
conclusions quantum mechanics Fig.3.2 Addiction
about filling atomic levels (shells) with density of statesNfrom
electrons. energyE.

3.2. Fermi function

In accordance with the Pauli principle, a state with the same set
of quantum numbers (in an elementary phase cell) can contain no
more than two electrons with antiparallel spins. But the cell can be
partially or completely occupied, or free. Probability of filling with
electrons with energyEelementary phase cell obeys quantumFermi-
Dirac statistics

31
 1
fF = E−EF
(3.8)
ekT
+1
WhereE F –Fermi level. At absolute zero T=0 oTOAll
energy levels below the Fermi level are occupied by electrons (fF=1),
and the levels lying above the Fermi level are free
(f F=0). AtT> 0TOprobability
o
of filling below (above) lying
1 1
levels are <f F <1 (f F < ) accordingly (Fig. 3.3).
2 2

Rice. 3.3 (a, b). Fermi–Dirac distribution atT= 0 T>0oTO(b). oTO(A), at

From Figure 3.3 it is clear that with increasing temperature, the “tail”
of the Fermi function shifts towards higher energies and can fall into
the conduction band. This means that there is a possibility of an
electron being thrown into the C-zone. Thus, the Fermi level is an
energy level maximally filled with electrons at absolute zero
temperatures, or half filled atT≠0oTO.
In thermodynamic terms, the Fermi level represents the
chemical potential of an electron gas per electron. This
thermodynamic interpretation of the Fermi level allows us to give
another definition of this concept. In a system of two contacting
bodies (metal - semiconductor, two semiconductors, etc.), between
which exchange of charge carriers is possible in a state of
thermodynamic equilibrium, their chemical potentials (Fermi levels)
are equalized.
An electron gas whose properties are described by the Fermi
function is calleddegenerate(quantum). However, if the number of
allowed energy states becomes significantly greater than the number
of electrons, the need for the Pauli ban on filling these states
disappears, and it becomes possible to use statistics

32
Maxwell-Boltzmann.In this case, the degeneracy of the electron gas is
removed and it becomesclassic.
An electron gas can be considered non-degenerate if the unity in
formula (3.8) can be neglected, that is, the condition is satisfied
E−EF

ekT >>1 (3.9)

In this case we get
E
−
f→
F fM−B=A*e
kT
(3.10)
EF
kT
WhereA=e
E−E F
Condition (3.9) is satisfied when <<2÷3 i.e. on the “tail” of the curve.
kT
In other words, if forEaccept the bottom of the conduction band, then
the Fermi level in this case should be located in the band gap at a
distance of severalkTfrom its edges, which is done for most
semiconductors. The exceptions are heavily doped (with a large
number of impurities) semiconductors, as well as narrow-gap and
gapless semiconductors. Fermi–Dirac statistics are used to describe
their electronic properties.

3.3. Electron gas concentration

Density of states N(E) determines the number of states

(elementary phase cells) per unit energy interval per unit volume. The
value of the Fermi function gives the probability of finding an
electron in a state with energyE.Let's define
concentration of free electrons, the energy of which lies in the range
fromEbeforeE+dE dn=dz⋅2fF=N⋅2fFdE
(3.11)
where the factor two is related to the Pauli principle. Then, taking into account formula (3.7)

32
4π(2m)
dn= E1 2f dE (3.12)
h3
F

dn
Function is shown in Fig. 3.4.
dE

33
Fig.3.4 (a, b). FunctiondnatT=0 oTO(a) and atT≠0 oTO(b).
dE

TOelectron concentration increases

From Fig. it is clear that whenT= 0 o

for higher levels and is maximum for the Fermi level. At T≠0oTOSome
electrons acquire energy sufficient for their
dn
thermal throw into the C-zone, and their energy distribution in
dE
the conduction zone is limited by the bottom of the C-band and the Fermi-
Dirac distribution (shaded in Fig. 3.4).
Integrating formula (3.12) over energy will allow us to estimate the
concentration of the degenerate electron gas
4π(2m) ∞ 1
32

n= ∫E 2f dE (3.13)
h
3 F
0

AtT= 0 TOwe get

o

32
8π(2m) 3
n0= E F02 ( 3.14)
3h3
It is typical for metalsn≈ 0 1028m −3 , then at absolute zero in
In the quantum case, degenerate electrons can have energies from 0
to EF0≈5eV, while for classical electrons the thermal
there is no movement, and their energy (kinetic and total) is zero.

Let's find the degeneracy limit. Integrating formula (3.13) taking into account
condition (3.12) allows us to obtain
32
2(2πmkT)
n =
gr
(3.15)
h3
0

28m−3
Then for metals withn≈10 0 , electronic transition temperature
gas to non-degenerate isT≈5⋅10 4K. Thus, in

In metals in the solid state, the electron gas is always degenerate.

34
3.4. Charge carrier concentration and position of the Fermi level in
semiconductors.

Semiconductors there areownAndimpurities.

Intrinsic semiconductors are semiconductors that do not contain
impurities -donors and acceptors(Fig. 3.16A).
Alloying– introduction of an impurity into a semiconductor, in this case a
semiconductor is called an impurity semiconductor. If an element of
group 5 is introduced as an impurity into a semiconductor consisting of
elements of group 4 (for example, silicon or germanium), we obtain a
donor semiconductor (it will have an electronic type of conductivity), or
a semiconductorn- sort of. The main charge carriers in it are electrons
(Fig. 3.16b). The minimum energy that must be imparted to an electron
located at the donor level in order to transfer it to the conduction band
is called the donor ionization energy ΔEd=Ec−Ed(rice.3.16V). If we
introduce element 3 as an impurity
group, then you get an acceptor semiconductor with hole
conductivity (p-type). The main charge carriers are holes.
The minimum energy that must be imparted to an electron in the
valence band in order to transfer it to the acceptor level is called
energy ionizationacceptor ΔEa=Ec−Eg+Ea(rice.3.17).

Fig.3.16. Tetrahedral lattice of silicon (A– own silicon,b

– silicon doped with phosphorus,V– donor level)

35
Fig.3.17. Acceptor semiconductor (A– silicon, doped phosphorus,b–
acceptor level)

Let us consider in more detail the concentration of charge carriers and the position of
the Fermi level depending on the temperature in these semiconductors.

A). Proprietary semiconductor.

Integrating the formula for a non-degenerate semiconductor
allows us to determine the equilibrium electron concentration in the
conduction band
Ec−EF
−
n0=N e c kT (3.16)
2(2πmnkT)3 2
WhereNc= (3.17)
h3
- effective density of statesin the conduction zone. The equilibrium
concentration of holes in the valence band is found similarly

EF−Eυ
−
p0=N υe kT
(3.18)
3
2(2πmpkT) 2

WhereN
υ = (3.19)
h3
- effective density of states in the valence band.
For calculationn0Andp0according to formulas (3.16) and (3.18) you need to know
Fermi level positionEF. However, the product of concentrations
electrons and holes for a non-degenerate semiconductor does not depend on the Fermi
level, although it does depend on temperature:
EC − Eυ Eg
− −
n.p.
0 0=NNc e υ
kT
=NcN eυ kT
(3.20)
Expression
n.p.
0 0
=n i2 (3.21)
calledlaw of mass action, and the valueni–own concentrationcharge
carriers in a semiconductor. Thus,
Eg
−
ni= NN
c e
υ
2kT
(3.22)

It should be noted that the carrier concentration increases exponentially

with temperature, while the coefficient in front of the exponential
exhibits a much weaker temperature dependence. Addictionni(T)

36
rectified in lg coordinatesn−103T, i therefore, the tangent of the angle
tiltαexperimental straight line allows you to determine the band gapE
own semiconductor at ga given
temperature (Fig. 3.18). So, for example, whenT=300oKForSi ni=1.5⋅10
10cm − 3,Eg=1.1eV.

Semiconductor will
electronic
(n-type), If performed
ratio n0>>n i>>p 0, And
holey (p-type), If
performedp 0>>n i>>n 0

Able thermodynamic
equilibrium semiconductor
electrically neutral, that is, the total
charge of all charged particles must
be equal to zero. In addition to free
carriers –
Fig.3.18. Addictionnifrom
electrons and holes motionless
temperatures in lg
carriers can find yourself in
coordinatesn−103T.
forbidden zone, which is reflected in
electroneutrality equation n0+Na=p0
+N d (3.23)
WhereNa– number of electrons at acceptor levels,N d– concentration
holes at donor levels. This equation is used to calculate p0with a
knownn0or, conversely, to calculaten0with a knownp 0.

In an intrinsic semiconductor, the electroneutrality equation (3.23) takes

the form
n0=p 0 (3.24)
Then, taking into account formulas (3.16) and (3.18) we obtain

Ec−EF EF−Eυ
− −
Nce = Nυe kT
kT
(3.25)
We will count the energy in the energy diagram from the bottom of
the conduction bandEc=0 thenEυ= −ΔE= −Eg, hence,
12
E ⎛N ⎞
E F= − g−kTln ⎜⎜ υ ⎟⎟ (3.26)
2 ⎝N c⎠
Or taking into account formulas (3.17) and (3.19)

37
 34
E ⎛m ⎞
E F = − g − kTln⎜ p⎟ (3. 27)
2 ⎜⎝m⎟n⎠

Eg
At T=0 oK EF = − =E. i At T≠0 oK Howrule mp≈m n,
2
therefore the second term is significantly less than the first, and we can
say that the Fermi level is near the
middle of the band gap (Fig. 3.19). If
mp>mn, then the value
EFlocated above the middleE And g.
vice versa, mp<m, n That E F
is located closer to the valence band.
Note also that with increasing
temperature atmp>m n level
Fermi is approaching To zone
Fig.3.19. Position level conductivity.
Fermi V own
semiconductor V dependencies
on temperature.

b) Non-degenerate impurity semiconductor.

Following the electroneutrality equation (3.23), the total charge
of all charged particles must be equal to zero. Let us find the
concentration and position of the Fermi level using the example of a
donor semiconductor. At low temperatures, casting is most likely
electrons into the conduction band
from donor levels rather than from
the valence band (Fig. 3.20).
Then the electron concentration in
the conduction band will be equal
to the hole concentration at donor
levelsnn=pd. If the concentration
free electrons is small, then the
electron concentration in the
Fig.3.20. Position level conduction band donor
Fermi V donor semiconductor calculated
semiconductor depending on similar to concentration For
temperature. own semiconductor (3.25), in this
case

38
 Ec−EF ΔEd−EF
−
Nce = Nde kT
kT
(3.28)
If we count the energy on the energy diagram from the bottom of the
conduction bandEc=0 , then after logarithm and transformation
not hard to get
12
ΔE d ⎛ Nd ⎞
EF= − − kTln⎜⎜ ⎟⎟ (3.29)
2 ⎝N c⎠
AtT=0 oK

ΔE d
E=F − (3.30)
2
i.e. at absolute zero, the Fermi level in the donor semiconductor is located
in the middle between the donor levels and the bottom of the band
conductivity (Fig. 3.20). As the temperature rises, untilNc<N ,
d

The Fermi
higher level rises towhen
temperatures, the bottom ofFermi
Nc>N, the the conduction
level dropsband,
to and at
d

in the middle of the band gap, since at high temperatures a noticeable

transition of electrons from the valence band to the conduction band
begins and the concentration of intrinsic carriers increasesni
(Fig. 3.18).
Becausen n=p p d , from relation (3.28) it follows
, then taking into accountn=n n
d

temperature dependence of the concentration of free electrons of a

donor semiconductor at relatively low temperatures and partial
ionization of impurity atoms:
ΔEd

n= (NcNd ) e 2kT
−
12
(3.31)
Where (NcNd) – magnitude, weak
12

temperature dependent versus

exponential factor.

From Figure 3.21 it can be seen that

at low temperatures incline
dependencies lgn0i−1T
determined by energy ionization
donors ΔEd for semiconductor Fig.3.21. Dependenciesn0 from
n-type. temperature V coordinates
At higher temperatures, when the lgn0−1T For donor
concentration of self-carriers is still semiconductor with increasing
lowni<<Nd, A degree doping With
all electrons from donor levels increasing ordinal
numbers.

39
moved to the conduction band, then the electron concentration in the
conduction bandnbecomes equal to the concentration of the donor impurity N
d, This temperature region is called the impurity depletion region

(horizontal section of the straight line in Fig. 3.21). With a further increase
in temperature, the probability of electron transition from the valence
band to the conduction band will increase. In this case, the conductivity of
the semiconductor turns from impurity into intrinsic conductivity (the
section of the graph that is the same for all curves is the straight line with
the maximum slope in Fig. 3.18). The transition to intrinsic conductivity
occurs at a higher temperature, the wider the band gap and the greater
the impurity concentration. The temperature region in which the
concentration of charge carriers is determined by the thermal generation
of electron-hole pairs (the formation of an electron in the conduction band
and a hole in the valence band upon excitation) is the region of intrinsic
electrical conductivity of semiconductors.
For non-degenerate acceptor semiconductors, the reasoning is
similar, and at low temperatures the main mechanism of hole
formation is the capture of electrons from the valence band by
acceptors due to thermal motion, whilep=n a.

The position of the Fermi level in the band gap is determined by the
formula
12
E ⎛N ⎞
E=F −g− ΔE − kTln⎜⎜ a ⎟⎟
a
(3.32)
2 ⎝N υ⎠
AtT=0 oK

E
E=F −g− ΔE
a
(3.33)
2
i.e. at absolute zero the Fermi
level is semiconductor,
containing acceptors,
located in the middle between
acceptor levels And
top valence zones
(Fig. 3.22). WITH increase
temperature, Bye Nυ<N, a
The Fermi level approaches the top of
the valence band. In the area of
exhaustion impurities All Fig.3.22. Position level
acceptor levels are filling Fermi V acceptor
semiconductor depending on
electrons, those who came from
temperature.

40
valence band. At higher temperatures, whenN υ>N a,

the most probable process is the generation of electron-hole pairs, with

the Fermi level moving towards the middle of the band gap into the
region of its own conductivity. The relationship for the temperature
dependence of the concentration of free holes at low temperatures has
the following form
ΔEa
−
p0= (NυN) a
12
e 2kT
(3.34)

Addictionp 0 on temperature in lg coordinatesp 0−1TFor

of an acceptor semiconductor with increasing doping degree with

increasing serial number is similar to Figure 3.21 for a donor
semiconductor. Therefore, as in the case of a donor semiconductor at
low temperatures, the slope of the log dependencep0−1Tdetermined
by ionization energy acceptors ΔEa For
semiconductorp-type.

V). Degenerate semiconductor.

If the impurity concentration in a semiconductor is high, then
the average distance between impurity atoms becomes so small that
electrons can move from one atom to another due to the tunneling
effect. In this case, the impurity levels combine into a band that
merges with the allowed band (for donor levels - with the conduction
band, for acceptor levels - with the valence band), and the Fermi level
falls into the allowed band. Such a semiconductor is called
degenerate, and the charge carriers in it obey Fermi statistics
- Dirac. The carrier concentration in degenerate semiconductors is
higher due to the high impurity concentration. For example, in non-
degenerate germanium the carrier concentration is of the order of
1020−1022m − 3, and in degenerate germanium – 1024−1026m−3.

41
Chapter 4. Kinetic phenomena in semiconductors.
4.1 Mobility.
oTO
In the absence of an external electric field atT> 0 free electrons
and holes in a semiconductor undergo chaotic motion. When the
electric field is turned onE electrons accelerate in the direction
opposite to the field, and holes
– in the direction of the field. Due to the thermal motion of atoms,
electrons and holes are scattered on atoms and defects
crystal lattice. Average speed acquired
r r r
electronsυn(holesυp) under the influence of an electric fieldE,
equal to
r
r qE
υn = τ n (4.1)
mn
r
r qE
υp = τ p (4.2)
mp
Where τ n,τp – average free travel time of electrons and holes
respectively.
Mean free travel time for electrons and holes
ln
τn = (4.3)
υn
l
τp = p (4.4)
υp
Wherenl,l p– the average free path of an electron and a hole;
υ,nυ p – module of the average speed of thermal movement of electrons and
holes respectively. Substituting the valuesτnAndτpinto expressions (4.1) and
(4.2), we obtain equations for the charge carrier mobilities: υn
qln
un = = (4.5)
E mnnυ
υ qlp
up = p= (4.6)
E mpυp
Whereu
n ,u p – mobility of electrons and holes, respectively.

Mobilityu n And u
pdifferent in size due to differences
in effective masses and free path times of electrons and
holes, which depend on the scattering mechanism of electrons and holes

42
in the crystal lattice.
There are several mainscattering mechanismscharge carriers:

1) on thermal vibrations of the structural elements of the crystal

lattice;
2) on ionized impurities (impurity ions);
3) on neutral impurities (impurity atoms);
4) on lattice defects (vacancies, point defects, dislocations, crystallite
boundaries, cleavage planes, etc.);
5) on charge carriers.
Due to the usually low concentration of defects and charge carriers, the
fourth and fifth types of scattering are usually neglected in the theory of
semiconductors.
1.Scattering of carriers by thermal vibrations of the atomic lattice(Ge, Si, Te,
etc.). From general considerations it follows that the larger the cross-
section of the volume occupied by the vibrating atom, the greater the
probability of carrier scattering. This cross section, like the energy of
thermal vibrations of the lattice atoms, is directly proportional to the
square of the amplitude of vibrations of the lattice atoms. And the lattice
energy increases with temperature according to a linear law. Then the
scattering probability is directly proportional to the temperature, and
therefore the mean free path is inversely proportional to the absolute
temperature
1
l~ (4.7)
T
Average speed thermal movement carriers charge V
non-degenerate semiconductors depends on temperature in the same way as the
speed of movement of ideal gas molecules:

υ~T (4.8)
Combining mobility dependencies (4.5), (4.6), (4.7) we get for
and (4.8)
u~T − 3 2 (4.9)
i.e., with increasing temperature, the mobility due to scattering by
thermal vibrations of the lattice decreases. Scattering from thermal
vibrations of the lattice plays a predominant role at high
temperatures, while scattering from lattice defects and impurities
mainly affects itself at low temperatures.
2.Scattering of carriers by ionized impurities. This scattering mechanism
is characterized by a decrease in the efficiency of scattering of a moving
charge with increasing speed of movement, since with

43
An increase in speed reduces the time the scattering centers act on
moving charges. As a result, the free path of the carrier increases
with increasing speed according to the law:l~υ4, and mobilityu~v3.
Then the temperature dependence of the mobility during scattering
by ionized impurities
u~T 3 2 (4.10)
This scattering mechanism dominates over the first at low
temperatures.
If both considered scattering mechanisms are involved in carrier
scattering and they are independent, then the temperature dependence
of carrier mobility can be represented according to formulas (4.9) and
(4.10) by the relation
1 −3
=aT3+2bT 2
(4.11)
u
WhereaAndb– proportionality coefficients.
In general, the dependence of charge carrier mobility on temperature
turns out to be very complex, and formula (4.11) is valid only in a
relatively narrow temperature range. This is explained by the fact that
in real semiconductors at a given temperature other scattering
mechanisms also take place, such as, for example, scattering by
optical vibrations and charge carriers.

4.2. Temperature addiction electrical conductivity

semiconductors

semiconductors
Electrical conductivity σ depends from
charge carrier concentrations (n,p) and their mobility (u ,Andp)
n

σn=qnu n (4.12)
σp=qpu p (4.13)
And in general it is the sum of the electrical conductivities of all
charge carriers, for semiconductors - electrons and holes σ
=σn+σ p (4.14)
Taking into account the dependence of the concentration and mobility of
charge carriers on temperature (see formulas (3.16) and (4.9)), the specific
electrical conductivity of the intrinsic semiconductorσican be written in
form
Eg
−3 −
σ=i cT 2e 2kT
(4.15)

44
 −3
where is the multiplier cT 2changes slowly with temperature, whereas
Eg
−
factore2kT strongly depends on temperature atE g >>kT. By entering
−3
designationσ0i≡cT2, we get
Eg
−
σ=i σe2kT0i (4.15a)

In an impurity semiconductor at sufficiently high temperatures,

the conductivity of the semiconductor isown, and at low temperatures –
impurity. In the region of low temperatures for the specific electrical
conductivity of impurity conductivity, taking into account formulas
(3.31), (3.34) specific electrical conductivity of the donor (or
acceptor) semiconductor
e− 2EkT
g
σ n =σ 0n (4.16)
Eg
−
σ p =σ 0p e 2kT
(4.17)
From formulas (4.16), (4.17) it is obvious that the dependence of the
specific electrical conductivityσfrom absolute temperatureTV
semiconductor containing donors (acceptors) of varying
concentrationsNd(N), ais determined mainly by temperature
dependence of the concentration of charge carriers and has a form similar
to Figure 3.21. From Figure 3.21 it is clear that in the region of intrinsic
conductivity of a given semiconductor the slope of the straight line
(tgα) is determined by the band gap of the semiconductorE g, and in

103
region of impurity conductivity, slope of log dependencesσ− By
T
(tgα d) And (tgαa) allows one to estimate the activation energy of donor and
acceptor levels for semiconductors nAnd p– type
respectively.
In the region of impurity depletion, the concentration of the
majority carriers remains constant and the conductivity changes due
to changes in mobility with temperature. If the main mechanism of
carrier scattering in the region of impurity depletion is scattering by
thermal vibrations of the lattice, then the conductivity decreases with
increasing temperature. If the main scattering mechanism is
scattering by ionized impurities, then the conductivity will increase
with increasing temperature.

45
Practically at research temperature dependencies
conductivity of semiconductors, they often use not the conductivity, but the
resistance of the semiconductor
1
R= (4.18)
σ
then for the intrinsic semiconductor
Eg

R=i R e 0i
2kT
(4.19)
for semiconductorn– type
ΔEd
Rn=R0ne 2kT
(4.20)
for semiconductorp– type
ΔEa

R p=R 0pe
2kT
(4.21)

By measuring the temperature variation of the resistance of a

semiconductor in a certain temperature range, it is possible to determine
the band gap from expression (4.19)Eg, from formulas (4.20) and (4.21) –
donor ionization energy ΔEdor acceptor ΔEaimpurities.

4.3. Photoconductivity

The cause of conductivity in semiconductors, in addition to

thermal effects, can be other processes, for example, the action of an
electric field, ionization under the influence of various types of
radiation (light, X-rays,γ- rays). Let's take a closer look
on light exposure. When light
quanta are absorbed, transitions
are possible leading to the
appearance of free
charge carriers (Fig. 4.1):
1) transition1corresponds
own absorption of light (by the
main atoms of the crystal), which
leads to the formation of free
Fig.4.1. Possible transitions electron And
electrons in a semiconductor upon free hole for each absorbed
absorption quanta Sveta, photon (intrinsic
leading to the appearance of free
photoconductivity);
media:1– transition fromυ- zones in s
- zone;2, 3– transitions from impurities
2) transition 2 corresponds
centers. absorption of light by local

46
imperfections in the crystal (donors), while for each absorbed photon
a free electron and a hole are formed, associated with the
corresponding center;
3) transition3corresponds to the case when each absorbed
a photon excites an electron from the valence band to an unoccupied
local level (acceptor), resulting in the formation of a free hole and a
bound electron.
Existsred borderphotoconductivity - the maximum wavelength
at which light is still photoelectrically active, that is, it creates free
charge carriers. For intrinsic photoconductivity, the red boundary is
determined by the relation
ch ch
λ max i = = (4.22)
ΔE E g
for impurity photoconductivity
ch
λ Max d = (4.23)
ΔE d
ch
λ Max a = (4.24)
ΔE a

4.4. Hall effect

EHall effect- galvanomagnetic phenomenon observed in

semiconductors and is widely used in practice, for example, to study
the electrical properties of semiconductor materials; the action of a
Hall sensor designed to measure the strength (induction) of a
magnetic field is based on this effect. Galvanomagneticare called
kinetic phenomena that arise under the simultaneous action of
electric and magnetic fields. In particular, the Hall effect consists of
the appearance of an electromotive force in the direction
perpendicular to the current when a sample is placed in a transverse
magnetic field.
Note that the use of the formulas of classical physics within the framework
of a quantum mechanical consideration of the Hall effect (and the previous
theory of electrical conductivity) is justified from the point of view of the
approximation of weakly bound electrons. Thus, following the Heisenberg
uncertainty relation,
ΔrΔp~h (4.25)
uncertainty in the coordinate for a free electron Δr→ ∞ , which allows
you to accurately determine the momentum and speed of the electron
and use the classical laws of dynamics.

47
Let us present the elementary theory of the Hall effect for charge carriers of
the same type. To do this, consider a semiconductor sample in the form of a
rectangular parallelepiped (Fig. 4.2).

Let currentleaks left

right. If the current is created by
r
holes, then the drift speedυ etc.

has the same direction as the

current I (Fig. 4.4); if by electrons,
then it is directed in the opposite
direction. Note that the drift speed
is the average speed of the
movement of charge carriers
Fig.4.2. Deflection of electrons directed by the electric field. Let's
and holes moving in a
magnetic field in the Hall
place the semiconductor in an
effect. external magnetthread field to
inductionBwas sent perpendicular
to the current (towards us).
Lorentz force
[ r
]
F=l q υ etc.×IN (4.26)
r
for electrons and holes Lorentz forceFl as a result of vector
the product will be directed downward, therefore, both holes and
electrons will be deflected by the magnetic field to the lower face, and the
upper face will be depleted in the first case by holes, in the second - by
electrons.
Thus, in a semiconductorp– like the bottom edge is charging
putrthen, and the upper bound is negative, and the Hall effect arises field
qE, directed from bottom to top. In a semiconductorn- bottom type
side (with the same direction of current) is charged negatively, the top
– positive, and the Hall field is directed from top to bottom. If both
holes and electrons participate in charge transfer (mixed
conductivity), then the picture becomes more complicated. If the
mobility and concentration of carriers of both signs were the same,
then due to the mutual compensation of electrons and holes at the
side faces of the sample, the total charge would be zero. As a rule,
the concentration and mobility of carriers of one sign are greater
than that of the other, and at the faces of the plate partial mutual
compensation of charges of opposite signs occurs, and a transverse
electric field strength of uncompensated charges arises and,
consequently, a transverse potential difference (EMFHall).

48
For one type of charge carriers, the condition for the stationary
distribution of charge carriers, which determines the Hallssome field,
is the equality of the absolute values of the Lorentz forceF and strength
r
l

emerging Hall fieldqE: Fl=qE

(4.27)
or
qυWedIN=qE (4.28)
In expression (4.28) underυ means average speed module
Wed

drift, since most charge carriers move at speeds close to the average.
Multiplying both sides of (4.28) by the electron concentrationn, we
get for the semiconductorn– type nqυWed=qE
nn (4.28A)
Since the conduction current density is determined by the formula

I
j n=nqυ cp= (4.29)
S
WhereS=ab, aAndb– geometric dimensions of the sample (Fig.4.2), then with

taking into account (4.28A) we get

1 I.B. 1
E=n ⋅ n = jB (4.30)
qn ab qnn
For Hall potential between the upper and lower edges (EMF
difference)Vn=Enbshould
1 InB 1
Vn= ⋅ = jnBb (4.31)
qn a qn
Formula (4.31) does not take into account the statistical nature of the
distribution of charge carriers over velocities and may not be satisfied for
charge carriers having different velocities in magnitude and direction. In
addition, in more detail, the condition for the equilibrium distribution of
electrons and holes is satisfied not at the moment when the Lorentz force
balances the force of the electric field (4.27), but when the charge stops
accumulating on the side faces of the sample.
Taking into account the speed distribution of charge carriers over speeds for
non-degenerate semiconductors allows us to rewrite formula (4.31) in the
form
A I nB A
Vn= ⋅= jBb (4.32)
qn a qn
n

where is the coefficientAcalledhall mechanism

– factorand depends on
of charge carrier scattering. Such semiconductors like

49
silicon, germanium, InSb predominates at high temperatures
3π
scattering by thermal vibrations of the lattice, and the constantA= =1.18.
8
If scattering occurs mainly on impurity ions, A=1.93. In degenerate
semiconductors and metals, high-energy charge carriers participate
in electrical conductivity; in this case, the velocity distribution of
electrons and holes can be ignored andA=1.

Taking into account the sign of the charge, formula (4.32) can be rewritten in the following
form
I.B.
Vn=R⋅ n
(4.33)
a
n

Where

A
Rn= − (4.34)
qn
– Hall coefficient (orHall constant) for an electronic semiconductor,
and similarly for a hole semiconductor we obtain
I.B.
Vp=R⋅
p
(4.35)
a
p

Where

A
Rp= − (4.36)
qp
– Hall coefficient for a semiconductorp- sort of.

From formulas (4.33) – (4.36) it is clear that according to the sign of the Hall emfVn,Vp, A
therefore, according to the sign of the Hall constantR n ,Rp can be determined
sign of charge carriers or type of conductivity. Measuring quantities
experimentallyVpn),(VIN,In(I p),a, you can calculate the Hall coefficient from
formulas (4.33) or (4.35), then using formulas (4.34) or (4.36) find the
electron concentrationnor holespin the studied impurity
semiconductor in the region of impurity conductivity. By independently
determining the electrical conductivity of an impurity semiconductor σn
orσ p, we can estimate the mobility of charge carriersunorAnd,
p

thus obtaining all the basic information about charge carriers in an

impurity semiconductor.
When strictly calculating the effect parameters Hall's case
mixed conductivity taking into account the Maxwellian distribution

50
velocities of charge carriers for non-degenerate semiconductors for
Hall constant obtained
A pu2p−nu n
2

R= ⋅ (4.37)
q (pu p + nun)2
Ifn=p(intrinsic semiconductor) andun=up, then in case
own conductivity, the Hall constant is zero. Ifn=pAnd un≠up, That

A up−u
R= ⋅
n
(4.38)
qn up+u n

Thus, in the region of intrinsic conductivity, the sign of the Hall emf is
determined by the sign of the carriers whose mobility is greater, that is, the
sign of the electrons. From formula (4.37) it is clear that in the region of
mixed conductivity, measurements of the Hall coefficient and electrical
conductivity are not enough to determine the mobilities and
concentrations of electrons and holes.

4.5. Quantum Hall effect.

The quantum Hall effect and the phenomenon of superconductivity are

interesting primarily because they allow us to observe quantum
phenomena at the macroscopic level. The quantum Hall effect was
discovered by Claus von Klitzing in 1985, for which he It was obtained
Nobel Prize. This effect is a two- can be carried out in the
dimensional system in which movement of an electron along one of
directions are quantized. To observe
the quantum Hall effect, a two-
dimensional metal film with a
thickness of several angstroms is
used, which can be implemented in
a field-effect transistor (Fig. 4.3).
Sometimes it is also called a MOS
transistor (metal-oxide-
semiconductor transistor). On a
Fig.4.3. Schematic diagram of a
silicon wafer having a hole field-effect transistor device.

conductivity, an oxide layer is appliedSiO2(the surface is oxidized).

Electrodes are burned in on both sides, which are called source and
drain. A layer of metal is sprayed on top of the oxide and welded

51
electrode. This part of the device is called the shutter. If a positive
potential is now applied to the gate, then the holes located in the
semiconductor wafer will go as far as possible from the gate, and the
electrons (which are few in the hole material) will, on the contrary, be
pulled towards the dielectric layer, creating a conducting electronic
channel between the source and drain . This channel is a two-
dimensional metal. The electron concentration in the two-
dimensional layer is determined by the gate voltage and is practically
independent of temperature.

The discovery of the quantum Hall effect was preceded by the discovery of
another interesting effect - the disappearance of the resistance of a two-
dimensional metal in a strong magnetic field.
For a classical electron in a magnetic field
υ2
m =qυB (4.39)
r

Wherem,q– mass and charge of the electron,υ– electron velocity modulus,B–

module of the magnetic field induction vector. Then for the cyclical
electron frequency
υ qB
ω= = (4.40)
r m
In classical mechanics, the energy of an electron in a magnetic
field can be anything and is determined only by the speed valueυ. In
quantum mechanics, rotational motion in a plane can be obtained by
the superposition of two oscillations occurring in two mutually
perpendicular directions with the energy of a quantum harmonic
oscillatorEand frequencyω, coinciding with the previously calculated
classical cyclic frequency of revolution of an electron in a magnetic
field:
⎛ 1⎞
E=hω⎜k+ ⎟,k=0,1,2... (4.41)
⎝ 2⎠
In the quantum case, any electron is localized in a certain region
of space in a two-dimensional channel with a characteristic size
which is the magnetic lengthl. So, from the relation uncertainties
(4.25) at Δr~r≡l, taking into account the formula (4.40) we get

h
l= (4.42)
qB

52
Thus, in the quantum case, the energy of electrons runs through a
discrete series of values (quantized), and electrons can be at certain
energy levels -Landau levels. Let's find the number of electrons at the
Landau leveln. Taking into account the Pauli principle, in assuming that
the area of the two-dimensional channel is equal to unity, we obtain

1 qB
n= = (4.43)
πl2 πh
For B=102T calculation using formula (4.43) shows that at each
Landau level can accommodate approximately 1012electrons per
square centimeter of channel area. Taking into account the spin
splitting of the Landau level into two sublevels in a magnetic field
⎛ 1⎞
⎜s=± ⎟ for the electron concentration at the Landau sublevel we obtain:
⎝ 2⎠
1 qB qB
n= 2 = = (4.43A)
πl 2πh h

Let us now consider the conductivity and Hall effect of a two-

dimensional metal in a quantizing magnetic field. The experimental design
is similar to that shown in Fig. 4.2. Taking into account the formulas
(4.40), (4.41) distance between Landau levels ΔE=hω<<EF~B,
those. As the magnetic field decreases, the distance between the
Landau levels decreases. Then the electrons that were located on the
first and second will occupy the third Landau level, and the Fermi level
will alternately cross the Landau levels. If below the Fermi levelEF

locatedifilled Landau level, then the concentration electrons

qB
n=i (4.43b)
h
Then the so-calledhall resistance,expressed through the Hall constant
(4.34), expressed only through fundamental
constants
h
RH=R.B.= , i=1,2,3... (4.44)
iq2
Thus, the dependence graphRH(B) should have the form of a straight line
extending from the origin, the tangent of the angle of inclination of which
to the abscissa axis is equal to the Hall constant. For three-dimensional
metal at room temperature and not too high magnetic values

53
field (1–5 T), the experiment shows exactly this result (Hall effect).
However, in a two-dimensional MOS structure at a temperature of 1K,
when classical consideration is not applicable, and it is necessary to
take into account the quantization of electron motion, the Hall
resistance exhibits a number of pronounced steps, and the value of
the resistance for these steps is strictly determined by expression
(4.44). These experimental results are presented in Figure
4.4. The curve with sharp peaks is the dependence of the ohmic
resistance of the sample on the magnetic field. Ohmic resistance is
the ratio of the longitudinal potential difference to the current passed
through the sample. As follows from the graph, the resistance goes
to zero each time the quantum Hall effect reaches a plateau. As soon
as the Fermi level, due to the movement of Landau levels as the
magnetic field increases, enters the region of delocalized electrons,
the ohmic resistance immediately goes to zero and remains so until
the Fermi level enters the region
localized states. Constanth q approximately equal to 25 kOhm.
2

Fig.4.4. Dependence of Hall resistance on the magnitude of the

applied magnetic field.
Hall resistance value turns out so
stable (independent of sample parameters and temperature), which
made it possible to use it as a national standard for electrical
resistance in a number of developed countries of the world.

In 1982 Tsui (Nobel Prize 1998), Stoermer and Gossard fractional

opened quantum Hall effect. They discovered that,
if you place an unadulterated pure sample at very low

54
temperature (about 0.1 K) in a magnetic field of strength 15 – 20 T,
then Hall plateaus and deep dips in longitudinal resistance arise with
fractional fillings of the lowest Landau level (i=1/3,i=2/3). This may be
due to the interaction of electrons in ultrastrong magnetic fields,
when the localization area of electrons becomes very small and their
density becomes high. There was an idea that this effect was the first
case of experimental observation of quasiparticles with a fractional
electric charge. One way or another, the nature of the fractional
quantum Hall effect is still not fully explained.

4.6. Superconductivity.

Phenomenonsuperconductivity was opened in 1911

Dutch physicist H. Kamerlingh-Onnes, who discovered that at a
temperature of 4.15 K, called the critical transition temperatureT,The
resistance of k
mercury dropped almost abruptly to zero
(Fig. 4.5). Later, superconductivity was
discovered in some metals: tin, lead,
indium, aluminum, niobium and many
alloys. This is a phenomenon in which
quantum mechanical
effects appear V
macroscopic scale. His
the theory is very complex and was
created only in 1957 by American
physicists Bardeen, Cooper and Fig.4.5. Addiction
Schrieffer and independently of them resistance R from
in 1958. N.N. Bogolyubov. temperature T for mercury.
Superconductivity is not exhausted
only by turning the electrical resistance of the conductor to zero
(ideal conductivity). The superconducting state is also characterized
by the fact that the magnetic field does not penetrate the thickness of
the superconductor, since in an external magnetic field induced
currents arise in it, which, according to Lenz’s law, create magnetic
induction that compensates for the external induction. As a result,
the magnetic induction inside the body will be zero. Induction lines
are pushed out of the body and go around it (Fig. 4.6.). Such a
superconductor has zero magnetic permeabilityμ= 0, i.e. is an ideal
diamagnetic. This phenomenon is called Meissner effect.The
emergence of the phenomenon of superconductivity

55
due to interaction electronic subsystems With
crystal lattice, due to which the electrons in the metal are subject to
mutual attractive forces, which in the superconducting state exceed
the Coulomb repulsion forces. Attraction occurs due to deformation
(polarization) of the crystal lattice, as a result of which the electron is
surrounded by a cloud
positive charge, k
to whom will
attract another
electron, forming So
called Cooper's
a couple.

On quantum
language

mechanics such an
How result
attraction arises
exchange quanta
excitement gratings –
Fig.4.6. Magnetic field distribution around phonons. Distance
a superconducting ball and around a ball between electrons couples
made of an ideal semiconductor:A) atT>Tk, can reach about 10−4cm,
b) atT<Tk in the external
which is four orders of
magnetic field,V) atT<TkV magnitude exceeds
absence of external magnetic field. interatomic distances in
crystal, and therefore
inside the region occupied by any pair there will be centers of very
many (from 105 to 1011) pairs, strongly overlapping, which is very
important for the stability of the state. Cooper pairs, having come
into coordinated motion (under the influence, for example, of an
electric field), remain in this state indefinitely, creating a conduction
current. Not all conduction electrons are combined into Cooper pairs,
since in a degenerate gas of fermions only those particles that have
free states near their levels can interact with each other, i.e. Fermi
electrons. AtT>0oKthere is some probability of pair destruction. That's
why Along with such pairs, there are “normal” electrons. The closer T
ToT
k, the larger the proportion of normal electrons becomes. AtT=Tk

all electrons become normal and the superconducting state is

destroyed.

56
Chapter 5. Contact phenomena.

5.1. Metal-semiconductor contact.

The electron gas is held in the crystal by the field of positive ions
in the lattice. The potential energy of an electron in such a field is less
than the potential energy of an electron located outside the crystal,
and some work must be done to remove it from the body. In metal at
T=0oKelectrons fill all energy levels up to the Fermi level. The
minimum work (energy) that must be expended to remove an
electron with energy corresponding to the Fermi level to such a
distance from the body where its (body’s) effect on the electron can
be neglected (vacuum level), is called (thermodynamic)work function
of output Am(rice.5.1A).

Fig.5.1. Energy diagram of metal-semiconductor contactn- type (Am>AP):

A- until contact,b– metal-semiconductor contactn- kind of able thermodynamic
equilibrium.

AtT>0 oKelectrons appear in the metal at levels higher

Fermi level. Removal of electrons with energy less thanE F, violates

M

equilibrium in the crystal and causes the transition of fast electrons to the
released levels with the release of energy and heating of the crystal. The
removal of electrons with an energy greaterEF, causes,M on the contrary,
transition of the remaining electrons to higher levels with the extraction of
energy from the lattice and its cooling. Only the removal of an electron
from the Fermi level does not upset the equilibrium in the crystal, so the
work function is always measured from the Fermi level. Similar picture

57
It is also observed in semiconductors, where, as a rule, electrons are
located either above or below the Fermi level and their removal
causes either heating or cooling of the crystal. Therefore, in
semiconductors, the work function is usually measured from the
Fermi level, although there may be no electrons at all. For impurity
semiconductors, the work function value strongly depends on
temperature due to changes in the position of the Fermi level. Let us
consider the contact of an electronic semiconductor with a metal in
the absence of surface states and assume that the work function of
electrons from the semiconductorAPless than work function
electrons from metalA(Fig.5.1
m A). Then when brought into contact
metal and semiconductor, electrons will move mainly from the
semiconductor to the metal, since the Fermi level in the
semiconductor before contact was higher than in the metal. In this
case, the metal is charged negatively, and the semiconductor is
charged positively. The transition of electrons will occur until the
Fermi levels level out and dynamic equilibrium is established (Fig. 5.1
b). Some excess charge (in this case, positive) accumulates in the
near-surface region of the semiconductor. This area is calledarea of
spatial chargein a semiconductor (SCR). From the drawing5.1bit is
clear that in the HMO semiconductor occurs
bendzones by an amount called
contact potential difference V To. It is determined through
work function difference

qVTo
=A−Am P (5.1.)
The electron concentration in a metal is much greater than in a
semiconductor; therefore, the extent of the space charge region in it
is very thin compared to the semiconductor. Therefore, almost all of
the voltage drop occurs in the semiconductor. IfAm>APfor
semiconductorn- like, the zones will bend
up (Fig. 5.1b). In this case, a layer of thickness depleted of the main
carriers will appear on the surfacedn, calleddepth
field penetrationinto a semiconductorn- sort of.

Consider the stateenrichmentmajority carriers in a non-

degenerate semiconductor, for example,n- like, that is, when on
electrons accumulate on the surface. SolutionPoisson equations
(Gauss's theorem for the electric field in differential form)

58
d 2V ρ(z)
=− (5.2)
dz 2
ε 0ε sc
– electrical constant,ε
Whereε 0 sc- the dielectric constant

semiconductor,ρ(z) – space charge density, depending on

coordinatesz,
will allow us to estimate the thickness in this case of the enriched layer on the
surface of the semiconductor in a state of equilibrium
2VToεε 0
dn= (5.3)
qNd

If the work function of an electron from a semiconductorn- like more

work function of the metal, then electrons mainly move from the metal
to the semiconductor and the SCR of the semiconductor is charged
negatively (Fig. 5.2).

Fig.5.2. Energy diagram of a metal-semiconductor contact:A–n– type (

Am<AP),b-p-type (Am<AP),V-p-type (Am>AP)
In this case, the energy bands of the semiconductor bend downward, and
the SCR in the semiconductorn– like it turns out to be enriched with basic
carriers - electrons. This is the areaaccumulation,when the concentration of
majority carriers on the surface of the semiconductor is greater than their
concentration in the bulk (anti-blocking layer).
When metal comes into contact with a semiconductorp– typeleanlayer in
semiconductor is formed whenAm<AP. State
depletion of major carriers on the surface is characterized by a
decrease in their number in the near-surface region compared to the
bulk value (barrier layer). With severe curvature

59
zones when the concentration of the majority carriers in the SCR
becomes much less than their concentration in the volume, and the
concentration of minority carriers on the surface becomes
comparable to the concentration of the majority carriers in the
volume, ainversion layer. There is also a state that characterizes the
absence of a space charge region in a semiconductor. When zones
near the surface are not curved, this is called a conditionflat areas.
Let's apply an external fieldVto metal-semiconductor contactn– type (
Am>AP). With forward biasV>0 (the “–” sign is attached to
electronic semiconductor, the “+” sign is for the metal), the Fermi level in
the semiconductor shifts upward by the amountqV, and bending zones
decreases (Fig. 5.3V). When connecting an external source, the plus is to the
electronic semiconductor, and the minus is to the metal (V<0) the Fermi level in
the semiconductor shifts down, and the band bending increases (Fig. 5.3b).

Fig.5.3. Energy diagram of metal-semiconductor contactn- type (Am<AP)

A- in balance,b– with reverse biasV<0 ,V–
at forward biasV>0 .

Let us estimate the depth of field penetration into a semiconductor in the

presence of external voltage. Formula (5.3) in this case will have the
following form whenV>0
2(VTo−V)εε
dn V>0= 0
(5.3A)
qNd
and with reverse voltageV<0
2(VTo+V)εε 0
dn V<0= (5.3b)
qNd

60
From formula (5.3A) it is clear that the upward bending of bands in the semiconductor
under consideration exists only whileV<VTo. From expressions (5.3A) And
(5.3b) we can also conclude that whenV>0, the depth of field
penetration into the semiconductor decreases, and atV<0 - increases.
Depleted layer changing its thickness according to formula (5.3A), is
called a physical barrier layer or Schottky barrier. The thickness of
this layer depends on the dopant concentration: the higher the
doping level, the thinner the Schottky layer. So, for example, for
germaniumn-type
ε=16,V=0.3
To
eV,N d =1014cm-3atV=0d=2.3⋅10 n − 4cm .

Let us consider the currents flowing at the metal-semiconductor

contact, which is associated with the passage of charge carriers through
the potential barriers that appear at the interface. There are two
possible approaches to consideration: quantum mechanicaltunnel Effect
Andover-barrier transfercharge carriers.
If the barrier is thin (the mean free path of electrons or holes is comparable to
the thickness of the barrier), then the problem of calculating the current is
reduced to the problem of calculating the thermionic emission current from
the surface of the semiconductor to the metal and back from the metal to the
semiconductor, taking into account the applied external voltage. Basically, the
current calculation is carried out in the same way as for a vacuum diode, and
this theory is calleddioderectification theory. It is valid for semiconductors
with high mobility (Ge, Si, InSb, etc.). Equation of the current-voltage
characteristic for a metal-semiconductor rectifying contact within the
framework of diode theory Richardson–Dashman) has the form (formula

⎛ ±
−1⎟ ⎞
qV

j=js⎜e kT
(5.4)
⎝ ⎠
Wherejs– saturation current density
Am
−
j=s AT 2e kT
(5.5)

In formula (5.5) Am– work function of electrons leaving the metal,

4πm* To2e A
A= =12⋅10 5
– Richardson constant.
h3 m2hail 2

The “+” sign in expression (5.4) corresponds to direct (access) in the

direction of current, and the “–” sign is in the opposite direction (
constipation). Assuming thatV>0 when connected directly, andV<0 –
otherwise, you can rewrite formula (5.4)

61
 ⎛ qV
⎞
j=j⎜ ekT − 1⎟ (5.4A)
s⎝ ⎠
View volt-ampere characteristics,
obtained within the framework of diode theory
is shown in Figure 5.3.

If thickness locking layer

dnsignificantly longer than the free length
electron path, then such a layer is called
thick and to estimate the currents flowing
through the metal-semiconductor
interface, in addition to drift currents
Fig.5.3. CVC contact caused by an external field, as in the case
metal - semiconductor in of a thin blocking layer, it is necessary to
diode theories take into account diffusion currents. In this
straightening. case, the volt-ampere characteristic
described by the formulaDavydova
⎛ ⎞qV

j=qu nEn0⎜e kT−1⎟ (5.6)

⎝ ⎠
WhereE– field strength. Difference from formula (5.5) is that now the
saturation current depends on the external voltage. Graphically this
dependence is presented in Fig. 5.4.

5.2. Electron-hole
transition

IN single crystal Germany,

silicon or another
semiconductor With with help
diffusion or ion implantation of
impurities can create such a
distribution that one crystal will be
a semiconductorn-type, but the
other is a semiconductorp-type. Fig.5.4. Current-voltage characteristics of the
Thus, in some thin area metal-semiconductor contact in the diffusion

crystal will theory of rectification.

there is a transition from n-type k

p-type. Electron-hole transition (p−n- transition)
called the semiconductor layer located on both sides of

62
interfacesnAndp– areas depleted of main carriers
charge. At the same time, the concentrations of minority carriers (holes in
the semiconductorn-typeNd and electrons in a semiconductorp-type
Na) are equal to the equilibrium concentrations of the main carriers
outside regionp−n– transition Nd=n0,Na=p
0 (5.7)
In this case they say that the impuritiesfully ionized. The concentration
of majority carriers (electrons in a semiconductorn– typennand holes in
the semiconductorp-typepp) significantly exceeds
minority carrier concentrationnn>>p n,pp>>p n, which leads
to the emergenceconcentration gradientdnAnddiffusionelectrons in
dx
dp
p–area as well as gradient and diffusion of holes inn-region.
dx
The corresponding diffusion currents can be expressed by the formulas
dn
I nD=qD nS (5.8)
dx
dp
I pD=qD pS (5.9)
dx
WherenDAndDp – coefficients diffusion of electrons and holes,S–
squarep−n– transition.
Vn– areas up to
As a result of the diffusion of electrons and holes
distancesdnuncompensated ionized donors are created and
nonequilibrium holes, and this region is charged positively, and inp–
region, uncompensated ionized acceptors and nonequilibrium
electrons arise, andp– the area is charged negatively. At
this in the areap−n– transition occurselectrical double layer
and electric fieldE, preventing diffusion transitions. This field leads to
the appearance of drift currents fromp–regions inn-region

InE=qnpEu n S (5.10)
fromn–regions in p-region
IpE=qpnEu pS (5.10A)
In steady dynamic equilibrium, when the Fermi level is the same forn
Andp–regions of the semiconductor diffusion currents
the majority charge carriers are balanced by drift currents

63
minority charge carriers, and the total current throughp−n– the transition is equal to
zero
I pD−I nE+I nD −I pE =0 (5.11)
p-typeE C takes
Bottom of the conduction band in a semiconductor
the highest position, so that for electrons in the conduction band of
the semiconductorn-like there is a potential barrier to transitionp–
area determined by the contact potential difference
equilibriump−n– transitionVTo(rice.5.5).

From Figure 5.5 it can be seen that

qVTo=E ip −E
= in
(5.12)
= (E ip−E F) + (EF−E in)
Then, assuming What V
depletion region in a state of
thermodynamic equilibrium
fair ratios
Boltzmann
EF−Ein
n0=nie kT
(5.13)
−
EF−Eip Fig.5.5. Zone diagram p−n
p0=nei kT
(5.14) - transition in thermodynamic
taking into account formula (5.7) we obtain equilibrium.
Na
qVTo=kTln (5.15)
Nd
Consequently, the more heavily both regions of the semiconductor are
doped, the greater the contact potential difference, the greater the
potential barrier for the transition of electrons fromnVp-region.
If top−n– apply external voltage to the junctionV<0, i.e., connect
the plus of the current source ton– area, and minus – withp–
area (Fig. 5.6A), then the height of the potential barrier for electrons
moving fromnVp–area will increase by an amountqV. Wherein
the equilibrium is disrupted and the Fermi level shifts upward by an
amount qV. In case of imbalance, formulas for equilibrium
concentrations (5.13), (5.14) and the concept of the Fermi level not applicable. IN
in this case introduce the conceptquasi-Fermi level(E
*
Fn E*FpFornAndp–
And
regions respectively), thennonequilibriumThe concentration of charge carriers is
expressed by similar formulas as the equilibrium concentration.

64
Fig.5.6. Band diagram for direct (b) and reverse inclusion (A)p−n– transition.

When applying reverse voltageV<0 in the area p−n–

transition, the following physical processes occur. Electrons are
pulled fromp–regions inn-area, and holes - fromn–regions in
p– area, i.e. there is a process of carrier transition from the area where
they were non-major, to the area where they are the main carriers,
calledextraction. It should be noted that every time a physical system
leaves a state of thermodynamic equilibrium, kinetic processes begin
to appear, tending to return the system to an equilibrium state. This
is the law of conservation of energy. In the case of extraction of
charge carriers, for example, an electron inton–area (Fig.5.6A), excess
energy appears, given to another electron from the valence band, the
process occursgenerationelectron-hole pair.

If top−n– apply external voltage to the junctionV>0, t.

e. plus the source connect top– area, and minus – withn-region, then
the potential barrier will decrease by the amountqVfor basic
charge carriers. In this case, the equilibrium is disrupted and the Fermi
level shifts by an amountqVdown (Fig.5.6b). Potential barrier to
transfer of electrons fromnVp-the area decreases and their injection
intop–region, and, similarly, injection of holes inton-region. Injection
is the transition of charge carriers throughp−n– transition (or
metal-semiconductor contact) to a region where they are not the
majority carriers. During injection, the concentration of minority
charge carriers innAndp– areas, however this
promotes an increase in the concentration of the majority carriers, since
when equilibrium is disturbed, Coulomb attractive forces arise between

65
charges of the opposite sign (minority carriers attract majority).
Happeningrecombination(a process inverse to generation), that is,
the disappearance of the electron-hole pair, and only part of the
diffusion current reaches the contact boundary.
When calculating the flux of charge carriers through a potential
p−n– transition is possible in two cases:
barrier
A) a thin transition where charge carriers hardly collide
with grill (l>>d) , while scattering in p−n– you can cross
neglect, and the current throughp−n– the transition is determined only by the number
carriers with sufficient energy to overcome the potential barrier. Then
the current throughp−n– transition is possible
considered as a thermocurrent using the Richardson–Dashman formula
(5.4). In this case, recombination-generation processes and the type of
ideal current-voltage characteristic are not taken into account p−n–
the transition is similar to the metal-semiconductor contact (Fig. 5.3);
b) thick transition, when the movement of charge carriers in p−n–
transition has a diffusion character (l<<d) . In this case, when
When calculating the current, its recombination-
generation components are taken into account,
and the type of current-voltage characteristic is
different from curve for
ideal p−n – transition
(Fig. 5.7).
Recombination increases With
increase direct voltage,
reducing the flow of charge carriers
throughp−n– transition, therefore, how
can be seen from Figure 5.7, the direct branch of the
Fig.5.7. CVC: 1 – ideal p−n-
current-voltage characteristic of an idealp−n– transition
transition),2 – taking into account
recombination-generation higher than the real one.
processes. Generational processes,
happening on contact
semiconductors Withvarious
type of conductivity, on the contrary, appear with reverse voltage. They
contribute to an increase in the number of charge carriers at the contact and,
consequently, an increase in the reverse current (Fig. 5.7).
Straightening propertyp−n– the transition has become practical
application in the production of semiconductor rectifier diodes.

66
 Chapter 6. Thermoelectric phenomena in
semiconductors.
Let's consider thermoelectric effects in semiconductors -
Seebeck, Peltier and Thomson effects.

6.1. Seebeck effect.

The effect is the occurrence of electromotive force EMF in
electrical circuit series-connected dissimilar
semiconductors or semiconductor and metal if the contact
temperatures are different.
Consider a non-degenerate semiconductorn– the type in which
dT
there is a temperature gradient (T2> T1) . Hot end
dx
semiconductor, the concentration of electrons and their kinetic energy are
higher than in a cold semiconductor, therefore, electron diffusion in the
direction of the temperature gradient will prevail. In addition, electrons can
move in the direction of the temperature gradient due to their entrainment
by phonons. When there is a temperature gradient at low temperatures in
a semiconductor, the number of phonons moving from the hot end to the
cold end is greater than at the opposite end. Phonons, colliding with
electrons, drag them in the direction of their movement. This phenomenon
plays a significant role only at low temperatures.

The diffusion flow of electrons from the hot end to the cold creates a
potential difference between them (thermo-EMF), obstructing this
flow. In a state of dynamic equilibrium (T2=T1) diffusion and drift flows
are balanced, and
the resulting current is zero.
Let us consider the contact between a semiconductor and a metal. If the
temperatures of the contacts between the semiconductor and the metal are
the same, then the contact potential differences are equal in magnitude and
directed in opposite directions. In the presence of a temperature gradient
dT
(T2> T1 ) , contact potential differenceVk contributes to
dx
thermo-EMFsemiconductor.
⎛dV k+ dVabout⎞
thermo - emf=⎜ ⎟ΔT (6.1)
⎝dT dT⎠
WhereVabout – volumetric potential difference arising due to the diffusion
of electrons from the hot end to the cold end.

67
The diffusion flow of charge carriers in the direction of decreasing
temperature gradient is caused by two reasons - the gradient
dn
free carrier concentrations at a constant coefficient
dx
dn
electron diffusion (D n):Dndx and changing the coefficientD With
n

dDn.
temperature at constant concentration:n
dx
In this case, the diffusion coefficientDndepends on the average kinetic
energyET, which in turn depends on the scattering mechanism charge
carriers and temperature Dn=C.E.
r+1 2,E
T= (r+2)kT (6.2)
T

Wherer— coefficient determined by the type of scattering; WITH-

proportionality factor.
In this case, the condition for dynamic equilibrium of free charge
carriers (electrons) in an isolated semiconductor can be written as
dn dD
formDndx + n n=Eu n (6.3)
dx
n

Where E– field strength createdthermo-EMF.

At Dn=constthe equation (6.3) will take the form
dn dn dT
Dndx =D n =Eu nn (6.4)
dT dx
Using Einstein's relation
kT
D= u (6.5)
q
From (6.4) we obtain an expression for the coefficientthermo-EMF
Edx k ∂lnn
αn= = T (6.6)
dT q ∂T
At n=const equation (6.3) will be written
dD ∂Dn∂E T
n n =n =Eu nn (6.7)
dx ∂ET ∂x
taking into account (6.2)

Edx k⎛ 1⎞
α=D = ⎜r+ ⎟ (6.8)
dT q⎝ 2⎠
Oddsα Andα Ddetermine the volumetric valuethermo-EMF.
n

Since the contact potential difference is determined by the difference in the work
functions of the semiconductor and metal

68
 Am−AP= EF−mEFn (6.9)
VTo=
q q
then the contact coefficientthermo-EMF
∂V 1∂E Fn
α To= To = − (6.10)
∂T q∂T
Taking into account formulas (3.16), (3.17)
k⎛3E l n⎞
∂n
α=To − ⎜− + Fn
+T ⎟ (6.11)
q⎝2kT ∂T⎠
Total coefficientthermo-EMFsemiconductorn-type
k⎛ N⎞
α=1 α+αn+α= D ⎜ 2 + r+ ln c⎟ (6.12)
q⎝ n⎠
To

Similarly for a non-degenerate hole semiconductor

k⎛ N⎞
α2=αp+αD+α= 2 +r+ln υ (6.13)
q⎜⎜ p ⎟⎟
To
⎝ ⎠
At high temperatures corresponding to the region of impurity
depletion (n=N d,p=Na), expressions for coefficientsα1Andα2have
view:
k⎛ Nc⎞
α=1 α+αn+α= D⎜⎜2 +rTo+ln (6.12A)
q⎝ N⎟⎟
d⎠

k⎛ Nυ ⎞
α 2=α +α
+ D= ⎜⎜α+To+ 2rln ⎟⎟ (6.13A)
q⎝ Na⎠
p

Thus, the valueαfor impurity semiconductors in the depletion

region increases with increasing temperature and decreases with
increasing impurity contentNaorN d.

In the region of intrinsic conductivity for the coefficientαThe theory gives the
following expression:
αu2 pp− α 1unn
α= (6.14)
upp+u nn
In this case, the electric current is carried by electrons and holes, and
theirthermo-EMFare mutually opposite in sign, and the resulting
thermo-EMFsignificantly lower than in the impurity region, i.e. the value
α upon transition from the impurity depletion region to the
intrinsic conductivity decreases. In the region of low temperatures
there is a large increasethermo-EMFdue to phonon drag of electrons.

Thus, the nature of the dependenceα(T) is quite complex and not

has a universal character for most semiconductors.

69
However, in a small temperature rangethermo-EThe DS can be considered
with sufficient accuracy for practical purposes to be proportional to the
temperature difference
thermo - emf=αΔT (6.15)
When connecting electron and hole semiconductors in series, they
thermo-EMF thermoEMF = (α1+α
2)ΔT =α12ΔT (6.16)
Where α 12 - coefficient differential thermo-EMF,
characterizing a pair of semiconductors.
Since in metals the concentration of free charges is high, it depends
and not

on temperature; the kinetic energy of a highly degenerate electron

gas of a metal practically does not change with temperature, as does
the contact potential of the metal at its boundaries. In
semiconductors, with a change in temperature, the concentration
and kinetic energy of free charge carriers changes sharply. In
addition, the contact potential and the associated Fermi level in
semiconductors are also functions of temperature. That's why
thermo-EMFsemiconductor thermocouples are hundreds of
microvolts per degree, which is hundreds of times higherthermo-EMF
metal thermocouples. Therefore, allthermo-EMF, arising in a metal-
semiconductor-metal circuit can be classified as a semiconductor, and
in the directionthermo-EMFsemiconductor determine the type of
conductivity or sign of charge carriers.
6.2. Peltier effect.

Peltier effect– release or absorption of heat in contact between

two dissimilar semiconductors or a semiconductor and a metal when
electric current flows through the contact. Amount of heat released
or absorbedQPat the point of contact
proportional to the amount of electricity passing through the contact
QP=±ПIt (6.17)
WhereP– Peltier coefficient.
The reason for the Peltier effect is that an external electric field
transfers electrons or holes from one material to another, and the
equilibrium energy of electrons (or holes) in both materials is
different and electrons coming from another material through the
junction have an excess or lack of energy according to compared to
others in this material. This excess (or lack) of energy is given to the
lattice (or replenished by

70
grid), as a result, the temperature of the junction rises or falls
accordingly.
Let us consider the contact of an electronic semiconductor with a
metal atAm>AP in a state of equilibrium (Fig. 6.1). In case of hole
semiconductor reasoning is similar. If the external voltage is such that
“plus” is applied to the metal, and “minus” is applied to the semiconductor,
then electrons move from the semiconductor to the metal, and each
electron drops to the Fermi levelEand givesFmback in collisions
energy in the formPeltier heatΔQPcontact
ΔQP=ET−EFn (6.18)
With the opposite direction of current (“plus” - on the semiconductor,
and “minus” - on the metal)
electrons move from the metal to
the semiconductor from energy
levels located above the bottom of
the zone conductivity. This
additional energy
electrons get behind check
lattice atoms located near the
contact. In this case, Peltier heat
(6.18) is absorbed and the contact
Fig.6.1. Energy scheme is cooled.
semiconductor contactn- type In the case of a non-degenerate
with metal atAm>AP. semiconductor, taking (6.2) into
account, we have

P=
ΔQP ET−EFn
= =
[(r+2)kT−E Fn ] (6.19)
q q q
then, taking into account equality (6.12), we obtain
P=αT 1
(6.20)
When current passes through the contact of two semiconductors
with different types of conductivity (p−n– transition) is also highlighted
or Peltier heat is absorbed, due to the fact that the average energy of the
electrons participating in the current on both sides of the contact is
different: the energy of free electrons of the semiconductorn-like more
energy bound electrons of a semiconductorp-type (Fig.6.2).
If throughp−n– the junction passes current in the forward
direction (“plus” source is attached top–area, and “minus” – ton–regions)
(Fig.6.2A), then electrons from the semiconductorn-type and holes from
semiconductorp-type move towards the contact and recombine, when

71
This releases energy, which is transferred to the crystal lattice and
goes to heat the junction.
When passing throughp−n– current transition in the shut-off valve
direction (Fig.6.2b) electrons move from the conduction band of the
semiconductorp-type inn–area, and the generation process occurs
electron-hole pairs, while energy borrowed from the lattice is
expended, the junction is cooled.

Fig.6.2. Energy diagram explaining heating (A) and cooling (b) at the contact of
two semiconductors with different types of conductivity.

The electric method of cooling when current passes through the

contact of two semiconductors has received practical application in
various cooling devices: microrefrigerators, thermostats,
thermoelectric hygrometers, etc. It is also promising to use the
Peltier effect for electric heating of rooms when current is passed in
the direction in which Peltier heat is released at the junction.

6.3. Thomson effect.

Thomson effectcalled excretion or absorption heat in a
homogeneous semiconductor when an electric current flows through
it in the presence of a longitudinal temperature gradient in it. The
amount of Thomson heat is proportional to the amount of electricity
passing through the semiconductorqAnd
dT
temperature gradient along the semiconductor
dx

72
 dT dT
Qτ=τ⋅It =τ⋅q (6.21)
dx dx
Where τ – Thomson coefficient, depending on the material
semiconductor.
The Thomson effect is explained similarly to the Peltier effect. When
current flows in an electronic semiconductor in the direction of
increasing temperature, electrons will move in the opposite direction,
that is, from places with a higher temperature (energy) to places with
a lower temperature. In this case, the electrons give up their excess
energy to the lattice, i.e., Thomson heat is released. When current
passes in the opposite direction, Thomson's heat will be absorbed.

According to theory Thomson coefficient Thomson connected With

thermo-EMF coefficient ratio
dα
τ=T (6.22)
dT
Wheredα – change in the thermo-EMF coefficient when changing
temperature atdT,T– average temperature of the semiconductor.
All three thermoelectric phenomena are mutually related to each other. Based
on the laws of thermodynamics for reversible processes, a connection has
been established between the coefficientsα,P,τ. Considering
closed circuit of two materials at a temperature differencedTbetween
contacts, based on the law of conservation of energy, we can write the
expression
∂P12dTdI+ (τ ∂
T 1−τ 2)dTdI=α 12 dTdI (6.23)

from which follows the relationship between the coefficientsα 12 ,P12,τ 1,τ2
for contact of two semiconductor materials
∂P
12+ (τ 1−τ 2) =α1 (6.24)
∂T
2

Literature.
1. Kittel Ch. Introduction to Solid State Physics. M., Nauka, 1978.
2. Oreshkin P.G. Physics of semiconductors and dielectrics. M.,
Higher School, 1977.
3. V.L. Bonch-Bruevich, S.G. Kalashnikov. Physics of semiconductors.
M., Nauka, 1990.
4. A.I. Anselm. Introduction to the theory of semiconductors. M., Nauka, 1978.

5. N. Ashcroft. Solid state physics. T.1,2. World, 1979.

73
In 2009, the University became the winner of a multi-stage
competition, as a result of which 12 leading universities in Russia
were identified, which were assigned the category “National Research
University”. The Ministry of Education and Science of the Russian
Federation approved the Development Program of the state
educational institution of higher professional education “St.
Petersburg State University of Information Technologies, Mechanics
and Optics” for 2009–2018.

DEPARTMENT OF PHYSICS

The Department of Physics as one of the general education departments has existed since the
organization of the Leningrad Institute of Precision Mechanics and Optics. In the pre- and post-
war years, the department was headed by A.P. Yushchenko, then Professor V.F. Troyanovsky, L.S.
Pollak, I.V. Poroikov, K.K. Aglintsev, D.B. Gogoberidze, N.A. Tolstoy, S.V. Andreev, A.Ya. Vyatskin,
founder of the LITMO thermophysical school, Professor G.M. Kondratyev and his students
Associate Professor A.F. Begunkova and Professor N.A. Yaryshev. Currently, the staff of the
department consists of graduates of our university, the Leningrad Polytechnic Institute, and the
Faculty of Physics of the Leningrad State University. 7 professors, doctors of sciences successfully
work at the department; 20 associate professors, candidates of sciences; 2 senior teachers,
candidates of sciences; 1 senior teacher and 4 assistants.

Since the establishment of the department, its employees have been paying great
attention to improving the teaching methods of physics, as one of the basic disciplines in
the training of future engineers and a subject that forms intelligence. Employees of the
department have written more than thirty textbooks for students on various sections of
the engineering physics course, on conducting laboratory work, processing the results of
physical experiments and solving physical problems. Currently, comprehensive work is
being carried out to improve the entire educational process, including the creation of
front-line computerized educational laboratories, banks for monitoring and testing
knowledge acquisition, and the preparation of software and methodological support for
distance learning of students through computer networks RUNNET and INTERNET. In
close cooperation with the Rosuchpribor association. The Ministry of Education of Russia
is developing laboratory training stands and workshops.

74
 The staff of the department conducts active scientific work. In 1957-1973, a scientific
direction was formed to study the physics of the interaction of electron beams with
matter. Since 1973, scientific research in the field of thermophysics, in particular, non-
stationary thermal conductivity and thermometry, has been developed. Since 1979,
scientific developments have been systematically carried out in the field of spectroscopy
of disordered condensed systems, since 1987 in the physics of wave processes, nonlinear
optics and radiophysics of anisotropic media, since 1994 - in optical and X-ray scattering
by supramolecular, in particular, fractal structures, since 1999 – on photonic crystals. The
scientific developments of the department have repeatedly received grants from the
Ministry of Education, Russian and International Scientific Foundations.

Since 1993, the department received the status of graduating in specialty 03.05.33
“Vocational training. Computer technologies”, and since 2003 – in the specialty 07.19.00
“Information systems and technologies”. The volume of scientific, methodological and
research work with students, bachelors and graduate students has increased
significantly. Dozens of students and graduate students, supervised by teachers of the
department, were awarded scholarships from the President of Russia, the Government of
Russia, the Administration of St. Petersburg, and international organizations (ISSEP, SPIE,
etc.).

75
 Olga Yurievna Shevchenko

Fundamentals of Solid State Physics

Tutorial

In the author's edition

Layout, design O.Yu.Shevchenko

Editorial and Publishing Department of the St. Petersburg State

University of Information Technologies, Mechanics and Optics

Head RIO N.F. Gusarova

License ID No. 00408 dated 05.11.09
Signed for printing 16.03.10 Order No.

Circulation

Printed on a risograph

76
Editorial and publishing department St.
Petersburg State University of Information
Technologies, Mechanics and Optics

197101, St. Petersburg, Kronverksky pr., 49

77
78