Lecture Notes - Solid State Physics

PHS 319
(SOLID STATE PHYSICS I)
BY
OLOPADE M.A.
1
TABLE OF CONTENT
PAGE
Unit 1 Crystal Structure of solids ……………………………………. 3
Unit 2 Crystal Binding ………………………………………………. 20
Unit 3 X-ray diffraction in crystals and Applications ……………….. 36
Unit 4 Thermal properties of the crystal lattice ……………………… 53
Unit 5 Elastic properties …………………………………………….. 69
Unit 6 Lattice Vibrations and Phonon ………………………………. 82
Unit 7 Free electron theory of Metals ……………………………….. 98
Unit 8 Motion of electrons in periodic fields ………………………… 110
Unit 9 Energy bands ………………………………………………….. 125
Unit 10 Semiconductors I………... ……………………………………. 144
Unit 11 Semiconductors II ……………………………………………… 159
Unit 12 Hall Effect ……………………………………………………. 173
Unit 13 Superconductivity ……………………………………………. 179
Solutions to Self Assessment questions (SAQ) ……………………………. 185
2
Study Session 1 Crystal Structure of Solids
Introduction
A crystal is a solid composed of atoms or other microscopic particles arranged in an orderly
repetitive array. That is about the shortest definition of a crystal. It may not be a complete
definition, yet it is a true description.
There are three general types of solids – amorphous, polycrystalline and single crystal which are
distinguished by the size of ordered regions within the materials. The arrangement of atoms in
amorphous solids is limited to a few molecular distances. In polycrystalline materials, the solid is
made up of grains which are highly ordered crystalline regions of irregular size and orientation.
Single crystals have long-range order. Many important properties of materials are found to depend
on the structure of crystals and on the electron states within the crystals. At the beginning of the
study of crystals it was their external form which was related to the physical properties. In this
way, only a limited success was achieved. In the middle of the last century, a deeper
understanding was developed regarding the correlation of the structure of crystals and mechanical,
thermal, electrical, and magnetic properties of solids. This is primarily due to the advances in the
band theory of electron states and in the theory of bonding in solids. This knowledge has led to the
development of newer and better materials for electrical, electronic and structural engineering.
The study of crystal physics aims to interpret the macroscopic properties in terms of the properties
of the microscopic particles of which the solid is composed. The study of the geometric form and
other physical properties of crystalline solids by using x-rays, electron beams and neutron beams
constitute the science of crystallography or crystal physics.
Learning Outcomes for Study Session 1

When you have studied this session, you should be able to:
1.1 Define crystal and understand the three types of solids.
1.2 Explain what is meant by a unit cell, Bravais lattice and non-bravais lattice.
1.3 Describe the different kinds of crystal symmetries.
3
1.1 Main Content
A solid is said to be a crystal if its atoms are arranged in such a way that their positions are
exactly periodic. This concept is illustrated in Figure 1.1 using a two-dimensional (2D) structure.
Figure 1.1 Crystal Structure showing the positions of the atoms
A perfect crystal maintains this periodicity in both the x and y directions from -to +.
Following from this periodicity, the atoms A, B, C, etc. are equivalent. In other words, for an
observer located at any of these atomic sites, the crystal appears exactly the same. The same idea
can be expressed by saying that a crystal possesses a translational symmetry. The translational
symmetry means that if the crystal is translated by any vector joining two atoms, say T in
Figure 1.1, the crystal appears exactly the same as it did before the translation. In other words the
crystal remains invariant under any such translation. The structure of all crystals can be described
in terms of a lattice, with a group of atoms attached to every lattice point. For example, in the case
of the structure shown in Fig.1.1, if we replace each atom by a geometrical point located at the
equilibrium position of that atom, we obtain a crystal lattice. The crystal lattice has the same
geometrical properties as the crystal, but it is devoid of any physical contents. There are two
classes of lattices: the Bravais and the non-Bravais. In a Bravais lattice all lattice points are
equivalent and hence by necessity all atoms in the crystal are of the same kind. On the other hand,
in a non-Bravais lattice, some of the lattice points are non-equivalent.
4
Figure 1.2 Representation of atomic lattice sites
In Figure 1.2 the lattice sites A, B, C are equivalent to each other. Also the sites A 1, B1, C1, are
equivalent among themselves. However, sites A and A1 are not equivalent: the lattice is not
invariant under translation AA1.
Non-Bravais lattices are often referred to as a lattice with a basis. The basis is a set of atoms
which is located near each site of a Bravais lattice. Thus, in Figure 1.2 the basis is represented by
the two atoms A and A1. In a general case, crystal structure can be considered as
Crystal structure = lattice + basis.
The lattice is defined by fundamental translation vectors. For example, the position vector of any
lattice site of the two dimensional lattice in Figure 1.3 can be written as
T=n1a1+n2a2 (1.1)
where a1 and a2 are the two vectors shown in Fig.3, and n1, n2 is a pair of integers whose values
depend on the lattice site.
Figure 1.3 Position vector of lattice sites in 2-D
So, the two non-collinear vectors a1 and a2 can be used to obtain the positions of all lattice points
5
which are expressed by Eq.(1.1). The set of all vectors T expressed by this equation is called the
lattice vectors. Therefore, the lattice has a translational symmetry under displacements specified
by the lattice vectors T. In this sense the vectors a1 and a2 can be called the primitive translation
vectors.
The choice of the primitive translations vectors is not unique. One could equally well take the
vectors a1 and a = a1+a2 as primitive translation vectors (see Figure 1.3). This choice is usually
dictated by convenience.
Unit cell: In the case of a rectangular two dimensional lattice the unit cell is the rectangle, whose
sides are the vectors a1 and a2. If the unit cell is translated by all the lattice vectors expressed by
Eq. (1.1), the area of the whole lattice is covered once and only once. A primitive unit cell is the
unit cell with the smallest area which produces this coverage. In the two dimensional case the area
of the unit cell is given by S=|a1a2|.
The choice of the unit cell is not unique. For example, the parallelogram formed by the vectors a1
and a in Figure 1.3 is also an acceptable unit cell. The choice is again dictated by convenience.
The area of the unit cell based on vectors a1 and a2 is the same as that based on vectors a1 and a.
In-text question
1. Q: Define the following terms in crystal structure study: (i) Lattice (ii) Primitive cell
(iii) Unit cell.
A: see pages 4-6.
Some important crystal structure terms are defined below:

Coordination number (N): The coordination number is defined as the number of equidistant
nearest neighbours that an atom has in the given structure. The greater the Coordination number,
the more closely packed up will be the structure.
Atomic packing factor: This the fraction of the space occupied by atoms in a unit cell.
6
Atomic radius (r): Atomic radius is defined as half the distance between nearest neighbours in a
crystalline solid without impurity.
1.1.1. Wigner-Seitz unit cell:
The primitive cell may be chosen as shown in Figure 1.4.
(i) Draw lines to connect a given lattice point to all nearby lattice points.
(ii) At the midpoint and normal to these lines, draw new lines (planes in 3D). The smallest volume
enclosed is the Wigner-Seitz primitive cell. All the space of the crystal may be filled by these
primitive cells, by translating the unit cell by the lattice vectors.
Figure 1.4 Wigner-Seitz primitive cell
The unit cell can be primitive and non-primitive (or conventional). The unit cell discussed above
is primitive. However, in some cases it is more convenient to deal with a unit cell which is larger;
however, as it exhibits the symmetry of the lattice more clearly.
Figure 1.5 Showing primitive translation vectors for a lattice site
Vectors a1 and a2 can be chosen as primitive translation vectors for the lattice shown in Fig.1.5. In
this case, the unit cell is a parallelogram. However, the lattice can also be regarded as adjacent
7
rectangles, where the vectors c1 and c2 can be considered as primitive translation vectors. The unit
cell in this case is larger; however it exhibits the rectangular symmetry more clearly. In the first
case, we have just one atom in a unit cell, whereas in the second case, we have a lattice with a
basis. The basis consists of the two atoms: one atom is located in the corner of the unit cell and
another atom in the center of the unit cell. The area of the conventional unit cell is larger by a
factor of two than the area of the primitive unit cell.
1.2 Crystal symmetry
Crystal lattices are classified according to their symmetry properties, such as inversion, reflection
and rotation.
Inversion center: A cell has an inversion center if there is a point at which the cell remains
invariant under transformation r → -r. All the Bravais lattices are inversion symmetric. Non-
Bravais lattices may or may not have an inversion center depending on the symmetry of the basis.
Reflection plane: A cell has a reflection plane if it remains invariant when a mirror reflection in
this plane is performed.
Rotation axis: This is an axis such that, if the cell rotated around the axis through some angle, the
cell remains invariant. The axis is called n-fold if the angle of rotation is 2 /n. Only 2-, 3-, 4-, and
6-fold axes are possible.
There are five Bravais lattice types in two dimensions shown in Figure 1.6. For each of them, the
rotation axes and/or mirror planes occur at the lattice points. However, there are other locations in
the unit cell with comparable or lower degrees of symmetry with respect to rotation and reflection.
For non-Bravais lattice, we have to take into account the symmetry of the basis which is referred
to as point-group symmetry. The point group symmetry includes all possible rotations, reflections
and inversion, which leave the basis invariant. Point groups are denoted by a numerical and “m”.
The numerical indicates how many positions within the basis are equivalent by rotation symmetry.
A single m shows that the basis has mirror plane symmetry. (In two dimensions – it is mirror
axis). E.g., 3m means that there are 3 equivalent sites within the unit cell and there is one mirror
plane. In two dimensions there are 10 point groups. When we combine the rotation symmetry of
the point group with the transnational symmetries, we obtain a space-group symmetry.
8
In-text question
1. Q: Explain the following terms in crystallography:
(i) Space-group symmetry (ii) Wigner-Seitz primitive cell (iii) Packing factor.
A: See pages 6-8 .
Figure 1.6a Bravais lattice types in 2-D
9
Figure 1.6b Bravais lattice types in 2-D
All the lattice properties we discussed for two dimensions can be extended to three dimensions.
The lattice vectors are in this case,
T=n1a1+n2a2+n3a3 (1.2)
where a1, a2 and a3 are the primitive translation vectors, and (n1,n2,n3) are a triplet of integers
whose values depend on a particular lattice site. The unit cell in three dimensions is a
parallelepiped; whose sides are the primitive translation vectors (see Figure 1.7). Here again the
choice of the unit cell is not unique, although all primitive unit cells have equal volumes. The unit
cell fills all space by the repetition of crystal translation operations. The volume of the unit cell
represented by a parallelepiped with sides a1, a2 and a3 is given by
V=|a1a2a3| (1.3)
Also, it is sometimes more convenient to deal with non-primitive or conventional cells, which
have additional lattice sites either inside the cell or on its surface.
10
Figure 1.7 The unit cell in 3-D
In three dimensions there are 14 different Bravais crystal lattices which belong to 7 crystal
systems. These systems are triclinic, monoclinic, orthorhombic, tetragonal, cubic, hexagonal and
trigonal. The crystal lattices are shown in Fig.1.8. In all the cases, the unit cell represents a
parallelepiped whose sides are a1, a2 and a3. The opposite angles are called , and . The
relationship between the sides and the angles determines the crystal system. A simple lattice has
sites only at the corners, a body-centered lattice has one additional point at the center of the cell,
and a face-centered lattice has six additional points, one on each side. Note that in all the non-
simple lattices the unit cells are non-primitive. The volume of the primitive unit cell is equal to the
volume of the conventional unit cell divided by the number of sites.
Each of the 14 lattices has one or more types of symmetry properties with respect to reflection and
rotation.
Reflection: The triclinic structure has no reflection plane; the monoclinic has one plane midway
between and parallel to the basis plane, and so forth. The cubic cell has nine reflection planes:
three parallel to the faces, and six others, each of which passes through two opposite edges.
Rotation: The triclinic structure has no axis of rotation (do not take into account 1-fold axis), the
monoclinic has a 2-fold axis normal to the base. The cubic cell has three 4-fold axis normal to the
faces and four 3-fold axis, each passing through two opposite corners.
11
Figure 1.8
12
1.3 Most common crystal structures
Body-centered cubic (bcc) lattice:
Figures showing the bcc lattice crystal structure
Primitive translation vectors of the bcc lattice (in units of lattice parameter a) are a1 = ½½-½; a2 =
-½½½; a3 = ½-½½. The primitive cell is the rhombohedron. The packing ratio is 0.68, defined as
the maximum volume which can be filled by touching hard spheres in atomic positions. Each
atom has 8 nearest neighbors. The conventional unit cell is a cube based on vectors a1 = 001; a2 =
010; a3 = 001. It is twice as big compared to the primitive unit cell and has two atoms in it with
coordinates r1 = 000 and r2 = ½½½. The alkali metals (Na, Li, K, Rb, Cs), magnetic metals (Cr
and Fe) and refractory metals (Nb, W, Mo, Ta) exhibit the bcc lattice structure.
Face-centered cubic (fcc) lattice:
Figures showing the fcc lattice crystal structure
13
Primitive translation vectors of the bcc lattice (in units of lattice parameter a) are a1 = ½½0; a2 =
0½½; a3 = ½ 0½. The primitive cell is the rhombohedron. The packing ratio is 0.74. Each atom
has 12 nearest neighbors.
The conventional unit cell is a cube based on vectors a1 = 001; a2 = 010; a3 = 001. It is 4 times
bigger than the primitive unit cell and has 4 atoms in it with coordinates r1 = 000; r2 = ½½0; r3 =
0½½; r4 = ½0½.
The fcc lattice has noble metals such as Cu, Ag, Au, common metals such as Al, Pb, Ni and inert
gas solids such as Ne, Ar, Kr, Xe.
Hexagonal closed-packed (hcp) lattice:
The hcp structure has a1=a2a3, ==90o and =120o with a basis of two atoms, one at 000 and
the other at . Along with the fcc structure, the hcp structure maximizes the packing ratio,
making it 0.74.
14
A closed-packed structure is created by placing a layer of spheres B on top of identical close-
packed layer of spheres A. There are two choices for a third layer. It can go in over A or C. If it
goes in over A the sequence is ABABAB. . . and the structure is hcp. If the third layer goes in
over C the sequence is ABCABCABC. . . and the structure is fcc.
In perfect hcp structure the ratio of the height of the cell to the nearest neighbor spacing is (8/3) ½.
In practice the (a3/a1) ratio is larger than 1.633 for most hexagonal crystals. Examples of
nominally hcp crystals include the elements from Column II of the Periodic Table: Be, Mg, Zn,
and Cd. Hcp is also the stable structure for several transition elements, such as Ti and Co.
Diamond structure is adopted by solids with four symmetrically placed covalent bonds. This is the
situation in silicon, germanium, and grey tin, as well as in diamond. Diamond has the translational
symmetry of fcc lattice with a basis of two atoms, one at 000 and the other at ¼¼¼.
Diamond structure represents two inter-penetrating fcc sublattices displaced from each other by
one quarter of the cube diagonal distance.
15
In-text question
1. Q: Define the term Coordination number.

A: This is the number of nearest neighbours to a given atom in a crystal
2. Q: Describe briefly the seven systems of crystals.
A: See pages 13-15.
1.3.1 Index system for crystal directions and planes
Crystal direction: Any lattice vector can be written as that given by Eq. (1.2). The direction is
then specified by the three integers [n1n2n3]. If the numbers n1n2n3 have a common factor, this
factor is removed. For example, [111] is used rather than [222], or [100], rather than [400]. When
we speak about directions, we mean a whole set of parallel lines, which are equivalent due to
transnational symmetry. Opposite orientation is denoted by the negative sign over a number. For
example:
16
Crystal planes: The orientation of a plane in a lattice is specified by Miller indices. They are
defined as follows. We find intercept of the plane with the axes along the primitive translation
vectors a1, a2 and a3. Let these intercepts be x, y, and z, so that x is a fractional multiple of a1, y a
fractional multiple of a2 and z a fractional multiple of a3. Therefore we can measure x, y, and z in
units a1, a2 and a3 respectively. We then have a triplet of integers (x y z). Then we invert it (1/x 1/y
1/z) and reduce this set to a similar one having the smallest integers by multiplying by a common
factor. This set is called Miller indices of the plane (hkl). For example, if the plane intercepts x, y,
and z in points 3, 1, and 3, the index of this plane will be (313).
The Miller indices specify not just one plane but an infinite set of equivalent planes. Note that for
cubic crystals the direction [hkl] is perpendicular to a plane (hkl) having the same indices, but this
is not generally true for other crystal systems. Examples of the planes in a cubic system:
In-text question
1. Q: What are Miller indices? Draw neat diagrams to indicate Miller indices of the important
plane systems in a simple cubic crystal.
A: Miller indices are used to specify the orientation of a plane in a lattice. See page 17.
17
Summary of Study Session 1
In this study session 1, you have learnt:
1. What is meant by a crystal structure, the definition of terms such as lattice points, space
lattice, basis, unit cell, lattice parameters, and primitive cell.
2. That there are 32 classes of crystal systems based on the geometrical considerations, that it is
common practice to divide all the crystal systems into 7 groups or basic systems (cubic,
tetragonal, orthorhombic, monoclinic, triclinic, trigonal and hexagonal).
3. About crystal symmetries and the different types of symmetry operations that can be performed
on crystals and the Wigner seitz cell, and miller indices.
Self- Assessment Questions for Study Session 1

Now that you have completed this study session, you can assess how well you have achieved its
learning outcomes by answering the following questions. Write your answers in your Study Diary
and discuss them with your Tutor at the next Study Support Meeting.
SAQ 1.1
The packing ratio is defined as the fraction of the total volume of the cell that is filled by atoms.
Determine the maximum values of this ratio for equal spheres located at the points of simple-cubic,
body-centered cubic (bcc), face-centered cubic (fcc) and diamond-structure crystals.
SAQ 1.2
With reference to crystals, state what you understand by the terms:

(i) Unit cell (ii) Single crystal (iii) coordination number (iv) packing factor
SAQ 1.3
What do you understand by Miller indices of crystal plane? Show that in a cubic crystal the
spacing between consecutive parallel planes of Miller indices (h k l) is given by:
𝑑 =
√
18
SAQ 1.4
Calculate the number of atoms per unit cell of a metal having a lattice parameter 0.29nm and
density of 7870 kgm-3. Take the atomic weight of the metal is 55.85.
References/Further Readings
Richard Turton (2000). The Physics of Solids: Oxford University Press.
19
Study Session 2 Crystal Binding
Introduction
Many solids are aggregates of atoms. The arrangement of atoms in any solid materials is
determined by the character, strength and directionality of the binding forces, cohesive forces or
chemical bonds. The bonds are made of attractive and repulsive forces that tend to hold the
adjacent atoms or atomic units at a particular spacing such that the opposite forces just balance;
and the process of holding them together is known as bonding. Since the particular type of
bonding within a material plays a major role in determining the physical, chemical and electrical
properties of materials, engineers must possess a sound working knowledge of the type of bonding
that exist in materials.

2.1 Explain interatomic forces.
2.2 Distinguish the different kinds of bonding between atoms of solids such as ionic, metallic,
covalent and van-der waals forces.
2.1 Main Content
2.1.1 Interatomic forces

Solids are stable structures, and therefore there exist interactions holding atoms in a crystal
together. For example a crystal of sodium chloride is more stable than a collection of free Na and
Cl atoms. This implies that the Na and Cl atoms attract each other, i.e. there exist an attractive
Interatomic force, which holds the atoms together. This also implies that the energy of the crystal
is lower than the energy of the free atoms. The amount of energy which is required to pull the
crystal apart into a set of free atoms is called the cohesive energy of the crystal.
Cohesive energy = energy of free atoms – crystal energy
The magnitude of the cohesive energy varies for different solids from 1 to 10 eV/atom, except
inert gases in which the cohesive energy is of the order of 0.1eV/atom (see table 1 below). The
cohesive energy controls the melting temperature (compare table 1 and table 2).
20
21
22
Figure 2.1 A typical binding energy curve
A typical curve for the potential energy (binding energy) representing the interaction between two
atoms is shown in Figure 2.1. It has a minimum at some distance R=R0. For R>R0 the potential
increases gradually, approaching 0 as R, while for R<R0 the potential increases very rapidly,
tending to infinity at R=0. Since the system tends to have the lowest possible energy, it is most
stable at R=R0, which is the equilibrium interatomic distance. The corresponding energy U0 is the
cohesive energy. A typical value of the equilibrium distance is of the order of a few angstroms
(e.g. 2-3Å), so that the forces under consideration are short range.
The interatomic force is determined by the gradient of the potential energy, so that
𝐹(𝑅) = − (2.1)
If we apply this to the curve in Fig.2.1, we see that F(R)<0 for R>R0. This means that for large
separations the force is attractive, tending to pull the atoms together. On the other, hand F(R)>0
for R<R0, i.e. the force becomes repulsive at small separations of the atoms, and tends to push the
atoms apart. The repulsive and attractive forces cancel each other exactly at the point R0, which is
the point of equilibrium.
The attractive interatomic forces reflect the presence of bonds between atoms in solids, which are
responsible for the stability of the crystal. There are several types of bonding, depending on the
physical origin and nature of the bonding force involved. The four main types are: Van der Waals
(or molecular) bonding, ionic bonding, covalent bonding and metallic bonding.
23
Although the nature of the attractive energy is different in different solids, the origin of the
repulsive energy is similar in all solids. The origin of the repulsive force is mainly due to the Pauli
exclusion principle. The elementary statement of this principle is that two electrons cannot occupy
the same orbital. As ions approach each other close enough, the orbits of the electrons begin to
overlap, i.e. some electrons attempt to occupy orbits already occupied by others. This is, however,
forbidden by the Pauli exclusion principle. As a result, electrons are excited to unoccupied higher
energy states of the atoms. Thus, the electron overlap increases the total energy of the system and
gives repulsive contribution to the interaction. The repulsive interaction is not easy to treat
analytically from first principles. In order to make some quantitative estimates it is often assumed
that this interaction can be described by a central field repulsive potential of the form exp(r /) ,
where and are some constants or of the form A/Rn, where n is sufficiently large and B is some
constant.
In-text question
1. Q: Discuss the term “cohesive energy”.

A: The amount of energy which is required to pull the crystal apart into a set of free atoms is
called the cohesive energy of the crystal.
2.1.2 Van der Waals (molecular) bonding:

First, we consider crystals of inert gases, which are characterized by van der Waals (or molecular)
bonding. The electron distribution in such crystals is very close to that in free atoms. The noble
gases such as neon (Ne), argon (Ar), krypton (Kr) and xenon (Xe) are characterized by filled
electron shells and a spherical distribution of electronic clouds in the free atoms. In the crystal, the
inert gas atoms are packed together within the cubic fcc structure.
What holds atoms in an inert gas crystal together? Consider two inert gas atoms (1 and 2)
separated by distance R. The average charge distribution in a single atom is spherically
symmetric, which implies that the average dipole moment of atom 1 is zero: <d1>=0. Here, the
brackets denote the time average of the dipole moment. However, at any moment of time, there
may be a non-zero dipole moment caused by fluctuations of the electronic charge distribution. We
denote this dipole moment by d1. According to electrostatics, this dipole moment produces an
24
electric field, which induces a dipole moment on atom 2. This dipole moment is proportional to
the electric field, which is in turn proportional to the d1/R3 so that
𝑑 ~𝐸~ (2.2)
The dipole moments of the two atoms interact with each other. The energy is therefore reduced
due to this interaction. The energy of the interaction is proportional to the product of the dipole
moments and inversely proportional to the cube of the distance between the atoms, so that
− ~− (2.3)
So, we see that the coupling between the two dipoles, one caused by a fluctuation, and the other
induced by the electric field produced by the first one, results in the attractive force, which is
called the van der Waals force. The time averaged potential is determined by the average value of
< 𝑑 > which does not vanish, even though <d1> is zero.
𝑈~ =− (2.4)
The respective potential decreases as R6 with the separation between the atoms.
Van der Waals bonding is relatively weak; the respective cohesive energy is of the order of
0.1eV/atom.
This attractive interaction, described by Eq.(2.4), holds only for a relatively large separation
between atoms. At small separations a very strong repulsive force caused by the overlap of the
inner electronic shells starts to dominate. It appears that for inert gases this repulsive interaction
can be fitted quite well by the potential of the form B/R12, where B is a positive constant.
Combining this with Eq. (2.4) we obtain the total potential energy of two atoms at separation R
which can be represented as
𝑈 = 4𝜀 − , (2.5)
25
where 4𝜀𝜎 ≡ 𝐴 𝑎𝑛𝑑 4𝜀𝜎 ≡ 𝐵 . This potential is known as the Lennard-Jones potential.
In-text question
1. Q: Briefly describe the Van der Waals forces of interaction in atoms.
A: See pages 22-24.
2.1.3 Ionic bonding

The ionic bond results from the electrostatic interaction of oppositely charged ions. Let us take
sodium chloride as an example. In the crystalline state, each Na atom loses its single valence
electron to a neighboring Cl atom, producing Na+ and Cl-ions which have filled electronic shells.
As a result an ionic crystal is formed containing positive and negative ions coupled by a strong
electrostatic interaction.
Na + 5.1eV (ionization energy) → Na + + e-
e-+ Cl → Cl- + 3.6eV (electron affinity)
Na+ + Cl- → NaCl + 7.9eV (electrostatic energy)
The cohesive energy with respect to neutral atoms can be calculated as 7.9eV - 5.1eV + 3.6eV, i.e.
Na + Cl NaCl + 6.4 eV (cohesive energy).
The structure of NaCl is two interpenetrating fcc lattices of Na + and Cl-ions as shown in Figure
2.2
Figure 2.2 Structure of the NaCl crystal
Thus, each Na+ ion is surrounded by 6 Cl-ions and vice versa. This structure suggests that there is
a strong attractive Coulomb interaction between nearest-neighbors ions, which is responsible for
the ionic bonding.
26
To calculate binding energy we need to include Coulomb interactions with all atoms in the solid.
Also, we need to take into account the repulsive energy, which we assume to be exponential.
Thus, the interaction between two atoms i and j in a lattice is given by
⁄
𝑈 = 𝜆𝑒 ±𝑞 𝑟 (2.6)
Here rij is the distance between the two atoms, q is the electric charge on the atom, the (+) sign is
taken for like charges and the (–) sign for unlike charges.
The total energy of the crystal is the sum over i and j so that
⁄
𝑈 = ∑ , 𝑈 = 𝑁 ∑ 𝜆𝑒 ±𝑞 𝑟 (2.7)
In this formula, ½ is due to the fact that each pair of interactions should be counted only once. The
second equality results from the fact in the NaCl structure the sum over j does not depend on
whether the reference ion i is positive or negative, which gives the total number of atoms. The
latter divided by two gives the number of molecules N, composed of a positive and a negative ion.
We assume, for simplicity, that the repulsive interaction is non-zero only for the nearest neighbors
(because it drops down very quickly with the distance between atoms). In this case we obtain
⁄
𝑈 = 𝑁 𝑧𝜆𝑒 − 𝛼 𝑞 ⁄𝑅 (2.8)
Here, R is the distance between the nearest neighbours; z is the number of the nearest neighbors,
and is the Madelung constant:
(± )
𝛼=∑ , (2.9)
where 𝑃 is defined by 𝑟 ≡ 𝑃 𝑅. The value of the Madelung constant plays an important role in
the theory of ionic crystals. In general it is not possible to compute the Madelung constant
analytically. A powerful method for calculation of lattice sums was developed by Ewald, which is
called Ewald summation.
27
Example:
A one-dimensional lattice of ions of alternating sign as shown in Figure 2.3.
Figure 2.3 One-dimensional lattice of ions
In this case
𝛼 = 2 1 − + − + + … = 2 ln 2 , (2.10)
where we took into account the logarithm expansion into series
In three dimensions, calculation of the series is much more difficult and cannot be performed so
easily. The values of the Madelung constants for various solids are calculated, tabulated and can
be found in various solid-state books.
Now we calculate the equilibrium distance between the nearest neighbors for the NaCl type lattice
using Eq.(3.8). At the equilibrium, the derivative dU/dR=0, so that
⁄
− 𝑒 + =0 (2.11)
and therefore
⁄
𝑅 𝑒 = (2.12)
This relationship determines the equilibrium separation R0 in terms of the parameters and of
the repulsive potential. Using Eq. (2.8) and Eq. (2.12) the cohesive energy per atom of the ionic
solid can be written as:
28
𝑈 = − =− 1− (2.13)
Let us estimate the magnitude of the cohesive energy in NaCl. The Madelung constant is =1.75.
The interatomic distance is R0=a/22.8Å. The charge q=e. The repulsive interaction of atoms has
a very short range of the order of =0.1R0. As follows from Eq.(2.13)
. .
≈ −( ⁄ )
1− ≈− 27 ∙ 0.9𝑒𝑉 ≈ −8𝑒𝑉 (2.14)
We see that the typical value of the binding energy per pair of atoms is about 8eV. This implies
that ionic bond is very strong. Experimentally, this strength is characterized by the relatively high
melting temperatures. For example, the melting temperature of NaCl is about 1100K, while the
melting temperature for the Na metal is about 400K.
2.1.4 Covalent bonding

The covalent bond is another important type of bond which exits in many solids. The covalent
bond between two atoms is usually formed by two electrons, one from each atom participating in
the bond. The electrons forming the bond tend to be partly localized in the region between the two
atoms joined by the bond. Normally the covalent bond is strong: for example, it is the bond which
couples carbon atoms in diamond. The covalent bond is also responsible for the binding of silicon
and germanium crystals.
We illustrate the appearance of the covalent bond by considering two atoms (e.g., hydrogen
atoms), which are described by orbitals 1 and 2. The molecular orbital of the two atoms is a
linear combination of the two orbitals. There are only two possibilities,
ѱ =ѱ +ѱ (2.15)
Or
ѱ =ѱ −ѱ , (2.16)
29
because symmetry considerations preclude any other linear combinations, since the distribution of
electron charge must be symmetric with respect to the two atoms.
Figure 2.4 Molecular orbital and charge distribution in covalent bonds
The molecular orbitals are sketched in Figure 2.4. This figure also shows the charge distribution
given by |ѱ | and |ѱ | . It can be seen that there is a sizable contribution to the charge density in
the region between the nuclei for the symmetric orbital, while there is a zero density between the
nuclei for the antisymmetric orbital. The two orbitals have different energies as is illustrated in
Figure 2.5, which shows the energy as a function of the interatomic distance.
Figure 2.5 Energy as a function of Interatomic distance in the two orbitals
We see that the symmetric orbital has a minimum of energy at certain distance and has a lower
energy than antisymmetric orbital. Thus, this is a bonding orbital which leads to a stable state of
the molecule. The other orbital is called anti-bonding orbital, which has a minimum of energy at
infinite separation of the atoms. This is a simple example of the covalent bonding between two
atoms.
30
Note that spins of the two electrons which participate in bonding are anti-parallel. This is a
consequence of the Pauli exclusion principle which requires the total wave function of the system
of electrons to be antisymmeteric with respect to any interchange of the coordinates of two
electrons. In the case of the bonding state the orbital wave function is symmetric and therefore the
spin contribution has to be asymmetric which means that the spins are anti-parallel. On the other
hand, the spins are parallel for the anti-bonding orbital. We see that The Pauli principle modifies
the distribution of charge depending on the spin orientation of electrons. This spin-dependent
contribution to the Coulomb energy is called exchange interaction. The simplest example of the
covalent bond is a hydrogen molecule.
The covalent bond in solids has strong directional properties. For example, carbon has four
valence electrons 1s22s22p2 and forms tetrahedral bonds with nearest neighbors, resulting in the
diamond type structure. The carbon atom is positioned in the center of the tetrahedron, the
neighboring carbon atoms being at the vertices of the tetrahedron (Figure 2.6). Since there are four
bonds joining the central atom to its neighbors, each C atom surrounds itself with eight valence
electrons, which is a stable structure because the second shell is now completely full. Such
tetrahedral coordination also occurs for the Si and Ge – those elements which can be found in the
fourth column of the periodic table.
Figure 2.6 Diamond Structure
To explain the tetrahedron arrangement in diamond, we note that each C atom has four electrons
in the second shell: two 2s electrons and two 2p electrons (2s 22p2). The s states are spherically
symmetric, whereas the p states represent charge distributions lying along x, y, and z coordinates.
The energy difference between these states is not very big. It appears that it is energetically
31
favorable to excite one of the s electrons to p states so that the electronic configuration becomes
2s2p3. We can now construct the linear combinations of atomic orbitals:
𝜳 = 1⁄2 𝑠+𝑝 +𝑝 +𝑝
𝜳 = 1⁄2 𝑠+𝑝 +𝑝 −𝑝
𝜳 = 1⁄2 𝑠+𝑝 −𝑝 −𝑝
𝜳 = 1⁄2 𝑠−𝑝 −𝑝 −𝑝 (2.17)
The densities corresponding to these orbitals are oriented along the tetrahedral directions.
These orbitals are therefore a better representation of the electron states i.e s, px, py, pz orbitals.
The mixing of the s and p states in Eq. (2.17) is referred to as the sp-hybridization. The particular
type of hybridization in diamond is known as sp3 hybridization. The sp3 hybridization occurs also
in Si and Ge. In Si one 3s and three 3p states hybridize to form tetrahedral bonds. In Ge the sp 3
hybridization involves one 4s and three 4p electrons.
Concluding the discussion about ionic and covalent bonds, we note that that there is a continuous
range of crystals between the ionic and covalent limits. In many cases, it is important to estimate
the extent a given bond is ionic or covalent. There are modern theoretical approaches that allow us
to quantify the degree of ionicity and covalence in many solids.
In-text question
1. Q: Describe ionic, covalent and Van der Waals bondings with examples.
A: See pages 24-27
2. Q: Calculate the binding (lattice energy) of the NaI for which the nearest neighbor distance is
0.324 nm. Express the energy in eV/molecule and also in KJ per kmol. Madelung constant
for NaI= 1.748 and n=9.5.
Solution
From the equation 𝑈 = − and substituting the values given in the question
above (i.e. A=1.748, n=9.5, ro=0.324nm and NA= Avogadro’s constant) gives the answer below:
A: 6.9eV or 667 x 103 kJ/kmol.
32
2.1.5 Metallic bonding
Metals are characterized by a high electrical conductivity, which implies that a large number of
electrons in a metal are free to move. The electrons capable to move throughout the crystal are
called the conduction electrons. Normally the valence electrons in atoms become the conduction
electrons in solids. The main feature of the metallic bond is the lowering of the energy of the
valence electrons in metal as compared to the free atoms. Below, some qualitative arguments are
given to explain this fact.
According to the Heisenberg uncertainty principle, the indefiniteness in coordinate and in the
momentum are related to each other such that ∆𝑥∆𝑝~ℏ . In a free atom the valence electrons are
restricted by a relatively small volume. Therefore, p is relatively large, which makes the kinetic
energy of the valence electrons in a free atom large. On the other hand in the crystalline state the
electrons are free to move throughout the whole crystal, the volume of which is large. Therefore,
the kinetic energy of the electrons is greatly reduced, which leads to diminishing the total energy
of the system in the solid. This mechanism is the source of the metallic bonding. Figuratively, the
negatively charged free electrons in a metal serve as glue that holds positively charged ions
together.
The metallic bond is somewhat weaker than the ionic and covalent bond. For instance the melting
temperature of metallic sodium is about 400K which is smaller than 1100K in NaCl and about
4000K in diamond. Nevertheless, this type of bond should be regarded as strong.
In transition metals like Fe, Ni, Ti, and Co, the mechanism of metallic bonding is more complex.
This is due to the fact that in addition to s electrons which behave like free electrons, there are 3d
electrons which are more localized. Hence, the d electrons tend to create covalent bonds with
nearest neighbors. The d electrons are normally strongly hybridized with s electrons, making the
picture of bonding much more complicated.
In-text question
1. Q: What are the distinguishing characteristics of metallic bonding?

A: See pages 31-32.
33
In this study session, you have learnt that:
1. Bonding between atoms is as a result of interatomic forces or bonds.
2. There are three strong principal types of primary bonds: Ionic, covalent and metallic. These
bonds are distinguished on the basis of the positions assumed by the bond electrons during the
formation of the bond. Van der Waals and hydrogen bonds are typical examples of secondary
bonds and they result from intermolecular attraction.
3. Generally, the stronger the bond, the higher the melting and boiling points.

SAQ 2.1
How is a covalent bond formed in a hydrogen molecule?
SAQ 2.2
Give two examples for: (i) ionic solid (ii) covalent solid.
SAQ 2.3
Describe the salient features of covalent and metallic bonded crystals
SAQ 2.4
If the ionic radius of Na decreases by 0.88 and that of Cl increases by 0.89, calculate the binding
energy of NaCl. Madelung constant for NaCl is 1.75 and n=9 for ionic crystals. Express your
result in kJ/kmol [rCl=0.0905 nm and rNa=0.186 nm].
34
SAQ 2.5
The ionic radii of Cs and Cl are 0.165 nm and 0.181 nm respectively, and their atomic weights
respectively 133 and 35.5. Calculate the density of CsCl.
Richard Turton, (2000). The Physics of Solids: Oxford University Press.
35
Study Session 3 X-ray Diffraction in Crystals, Applications.
Introduction
X-rays are electromagnetic waves like ordinary light; therefore, they should exhibit interference
and diffraction. The wavelength of X-rays is of the order of 0.1 nm, so that ordinary devices such
as ruled diffraction gratings do not produce observable effects with X-rays. In 1912, German
physicist Laue suggested that a crystal which consisted of a three-dimensional array of regularly
spaced atoms could serve the purpose of grating. The crystal differs from the ordinary grating in
the sense that the diffracting centres in the crystal are not in one plane. Hence the crystal acts as a
space grating rather than a plane grating.
On the suggestions of Laue, his associates, Friedrich and Knipping succeeded in diffracting X-
rays by passing them through a thin crystal of zinc blende. The diffraction pattern obtained
consists of a central spot and a series of spots arranged in a definite pattern around the central
spot. This symmetrical pattern of spots is known as Laue pattern, and it proves that X-rays are
electromagnetic radiation. A simple interpretation of the diffraction pattern was given by W.L.
Bragg (1912). According to him, the spots are produced due to the reflection of some of the
incident X-rays from the various sets of parallel crystal planes (called Bragg’s planes) which
contain a large number of atoms.

3.1 Explain the genesis of X-ray diffraction
3.2 Describe Brillouin zones and reciprocal lattices
3.3 Describe X-ray diffraction patterns and uses of x-ray diffraction.
3.1. Main Content
3.1.1 Bragg Law

Most methods for determining the atomic structure of crystals are based on the idea of scattering
of radiation. X-rays is one of the types of radiation which can be used for such a study. The
36
wavelength of the radiation should have a wavelength comparable to a typical interatomic
-8
distance which is in solids of a few angstroms (10 cm). The x-ray wavelength λ can be estimated
as follows
.
𝐸 = ℎ𝜈 = 𝜆(Å) = (3.1)
( )
Therefore, x-rays of energy 2-10 keV are suitable for studying the crystal structure.
X-rays interact with electronic shells of atoms in a solid. Electrons absorb and re-radiate x-rays
which can then be detected. Nuclei are too heavy to respond. The reflectivity of x-rays is of the
-3 -5
order of 10 - 10 , so that the penetration in the solid is deep. Therefore, x-rays serve as a bulk
probe.
In 1913 Bragg found that crystalline solids have remarkably characteristic patterns of reflected x-
ray radiation. In crystalline materials, for certain wavelengths and incident directions, intense
peaks of scattered radiation were observed. Bragg accounted for this by regarding a crystal as
made out of parallel planes of atoms, spaced by distance d apart. The conditions for a sharp peak
in the intensity of the scattered radiation were that:
(1) the x-rays should be specularly reflected by the atoms in one plane;
(2) the reflected rays from the successive planes interfere constructively.
Figure 3.1 X-ray reflected from planes
37
Figure 3.1 shows x-rays which are specularly reflected from adjacent planes. The path difference
between the two x-rays is equal to 2dsinθ. For the x-rays to interfere constructively this difference
must be an integral number of wavelengths. This leads to the Bragg condition:
2𝑑𝑠𝑖𝑛𝜃 = 𝑚𝜆 (3.2)
The integer m is known as the order of the corresponding reflection (or order of interference).
There are a number of various setups for studying crystal structure using x-ray diffraction. In most
cases, the wavelength of radiation is fixed, and the angle is varied to observe diffraction peaks
corresponding to reflections from different crystallographic planes. Using the Bragg law one can
then determine the distance between the planes.
The Bragg law is greatly oversimplified (but it works). It
(i) says nothing about intensity and width of x-ray diffraction peaks;
(ii) neglects differences in scattering from different atoms;
(iii) neglects the distribution of charge around atoms.
In-text question
1. Q: Explain Bragg’s law for x-ray diffraction in crystals.
A: See pages 35-36
3.1.2. Diffraction condition and reciprocal lattice

Later, Von Laue (1914) introduced a different approach for x-ray diffraction. He regarded a
crystal as composed of identical atoms placed at the lattice sites T and assumed that each atom
could reradiate the incident radiation in all directions. Sharp peaks were observed only in the
directions and at wavelengths for which the x-rays scattered from all lattice points interfered
constructively.
38
Figure 3.2 Constructive interference from reflected x-rays
To find the condition of constructive interference we consider two scatterers (Figure 3.2)
separated by a lattice vector T. Let x-rays be incident from infinity, along direction k with
wavelength λ and wave vector k=2πk/λ. We assume that the scattering is elastic, i.e. the x-rays are
scattered in 𝒌 direction with same wavelength λ, so that the wave vector (𝑘 = 2𝜋𝑘 /𝜆). The
path difference between the x-ray scattered from the two atoms should be an integral number of
wavelengths. Therefore, as is seen from Fig.3.2, the condition for constructive interference is
(𝑘 − 𝑘). 𝑇 = 𝑚𝜆 , (3.3)
where m is an integer. Multiplying both sides of Eq.(3.3) by 2π/λ leads to a condition on the
incident and scattered wave vectors:
(𝑘 ′ − 𝑘) ∙ 𝑇 = 2𝜋𝑚. (3.4)
Defining the scattering wave vector Δk = 𝑘 ′ − 𝑘, the diffraction condition can be written as
Δk=G, (3.5)
where G is, by definition, such a vector for which

G∙Τ = 2𝜋m. (3.6)
A set of vectors G which satisfies this condition form a reciprocal lattice. Vectors G are called
reciprocal lattice vectors.
A reciprocal lattice is defined with reference to a particular Bravais lattice which is determined by
a set of lattice vectors T. The Bravais lattice that determines a particular reciprocal lattice is
referred as the direct lattice, when viewed in relation to its reciprocal.
39
There is an algorithm for constricting the reciprocal lattice from the direct lattice. Let a , a , and
1 2
a be a set of primitive vectors of the direct lattice. Then the reciprocal lattice can be generated
3
using primitive vectors
𝒃 = 𝒂 ×𝒂 , 𝒃 = 𝒂 ×𝒂 , 𝒃 = 𝒂 ×𝒂 , (3.7)
where V = 𝒂 ∙ (𝒂 × 𝒂 ) is the volume of the unit cell, so that
G = 𝑚 𝒃𝟏 + 𝑚 𝒃𝟐 + 𝑚 𝒃𝟑 (3.8)
In order to prove that the vectors built in this way satisfy condition (3.6), we first note that the bi
satisfy the condition
𝒃 ∙ 𝒂 = 2πδij (i,j=1,2,3) (3.9)
where δ is the Kronecker symbol defined by

ij
δ =0, if i≠j (3.10)

ij
δ =1, if i=j .
ij
Indeed, since T = 𝑛 𝒂 + 𝑛 𝒂 + 𝑛 𝒂 we obtain
G∙T = 2𝜋(𝑛 𝑚 + 𝑛 𝑚 + 𝑛 𝑚 ), (3.11)

which proves Eq. (3.6).
Examples: reciprocal lattices for 1D and 2D-rectangular structures. Note: Eqs.(3.9) rather than
Eqs.(3.7) should be used in 1D and 2D cases.
40
Figure 3.3 Reciprocal lattices for 1D and 2D
Reciprocal lattice types for some 3D lattices:
Direct lattice Reciprocal lattice

sc sc
bcc fcc
fcc bcc
hcp hcp
Coming back to the diffraction condition (3.5), we can say that constructive interference occurs
provided that the scattering wave vector is a vector of the reciprocal lattice.
It is sometimes more convenient to give a different formulation of the diffraction conduction. In

elastic scattering the photon energy is conserved, so that the magnitudes of k and kˊ are equal, and
therefore 𝑘 = 𝑘ˊ . Therefore, it follows from Eq.(3.5) that
𝒌ˊ = (𝑮 + 𝒌) 0 = G + 2𝒌 ∙ 𝑮. (3.12)
By replacing G with -G, which is also a reciprocal lattice vector, we arrive at
𝟐𝒌 ∙ 𝑮 = 𝐺 , (3.13)
that can be considered as the diffraction condition.
Equation (3.13) is another statement of the Bragg law (3.1). We prove this in three steps.
41
(1) We show that the reciprocal lattice vector 𝑮 = ℎ𝒃𝟏 + 𝑘𝒃𝟐 + 𝑙𝒃𝟑 is orthogonal to the plane
represented by Miller indices (hkl).
Consider the plane (hkl) which intercepts axes at points x, y, and z given in units a1, a2 and a3:
Figure 3.4
By the definition of the Miller indices we can always find such interceptions that
(ℎ, 𝑘, 𝑙) = , , . (3.14)
As we know, any plane can be defined by two non-collinear vectors lying within this plane. We
can choose vectors u and v shown in Fig.3.4. They are given by 𝒖 = 𝑦𝒂 − 𝑥𝒂 and
𝐯 = y𝐚 − z𝐚 . To prove that the reciprocal vector G is normal to the plane (hkl), it is sufficient to
prove that this vector is orthogonal to u and v, i.e. 𝐮 ∙ 𝐆 = 0 and 𝐯 ∙ 𝐆 = 0. We have
𝐮 ∙ 𝐆 = (y𝐚 − 𝑥𝐚 ) ∙ (ℎ𝐛 + 𝑘𝐛 + 𝑙𝐛 ) = 2π(𝑦𝑘 − 𝑥ℎ) = 0, (3.15)
where the second equation follows from the orthogonality condition of the vectors of the direct
and reciprocal lattices (3.9) and the last equation follows from Eq.(3.14). In the same manner we
can show that G is orthogonal to v. We have proved, therefore, that vector G is orthogonal to the
plane (hkl).
(2) Now we prove that the distance between two adjacent parallel planes of the direct lattice is
d=2π/G.
42
First, we note that the nearest plane which is parallel to the plane (hkl) goes through the origin of
the Cartesian coordinates in Fig.3.4. Therefore, the interplanar distance is given by the projection
of one of the vectors xa , ya , za , to the direction normal to the (hkl) plane. This direction is
1 2 3
given by the unit vector G/G, since we have already established that G is normal to the plane.
Therefore
𝑑 = 𝑥𝒂 ∙ 𝐆⁄G = 2𝜋𝑥ℎ⁄𝐺 = 2𝜋⁄G. (3.16)
The connection between reciprocal vectors and crystal planes is now clear. The reciprocal vector
G(hkl) is associated with the crystal planes (hkl) and is normal to these planes. The separation
between these planes is 2π times the inverse of G.
(3) Now we are ready to show that the diffraction condition (3.13) is equivalent to the Bragg law (3.2).
It follows from Eqs. (3.13) and (3.16) that
2(2𝜋⁄𝜆) sin 𝜃 = 2𝜋⁄𝑑 , (3.17)
or 2𝑑 sin 𝜃 = 𝑚𝜆 , where θ is the angle between the incident beam and the crystal plane. The
integers hkl that define G are not necessarily identical with the indices of the actual plane, because
hkl may contain a common factor m, whereas in the definition of the Miller indices the common
factor has been eliminated. Therefore, we can substitute mG for G and obtain the Bragg result
2𝑑 sin 𝜃 = 𝑚𝜆. (3.18)
Brillouin zones. Brillouin gave another statement of the diffraction condition.
Consider a two-dimensional lattice in the reciprocal space (Fig.3.5a). Let O be the origin of this
lattice. Consider a reciprocal lattice vector, which connects points O and another reciprocal lattice
site. Now draw the line (in three dimensions it would be a plane), which is orthogonal to this
vector and intercepts it in the midpoint. The x-ray will be diffracted if its wavevector k has the
43
magnitude and direction that is required by the condition (3.13), which can be rewritten in the
following way
𝟐
𝐤∙ 𝐆 = 𝐆 , (3.19)
It is easy to see that any k vector connecting the origin and the plane will satisfy the diffraction
condition.
Figure 3.5a&b Two dimensional lattices in reciprocal lattice
In a similar way we can draw other lines (planes), which satisfy the diffraction condition. This is
shown in Fig.3.5b. So the Brillouin construction exhibits all the wave vectors k which can be
Bragg-reflected by the crystal.
The first Brillouin zone is the smallest volume entirely enclosed by the planes that are
perpendicular bisectors of the reciprocal lattice vectors. In this construction it is the rectangle
about the origin. The first Brillouin zone is the Wigner-Seitz primitive cell in the reciprocal
lattice.
44
Figure 3.6
First Brillouin zone of the bcc lattice (rhombic- First Brillouin zone of an oblique lattice
Dodecahedron). in two dimensions.
Another alternative expression for the diffraction conduction can be given in terms of Laue
equations. Taking a dot product of Eq. (3.5) with a , a , a we obtain:
1 2 3
𝒂 ∙ ∆𝐤 = 2𝜋𝑚 ,
𝒂 ∙ ∆𝐤 = 2𝜋𝑚 , (3.20)
𝒂 ∙ ∆𝐤 = 2𝜋𝑚 ,
These conditions state that the allowed scattering vectors, Δk, should lie at the intersections of
cones around each lattice vector.
In-text question
1. Q: Briefly explain the conditions necessary for diffraction to take place.
A: See pages 35-36
2. Q: Explain the following terms: (i) Brillouin zone (ii) reciprocal lattice.
A: (i) See page 37 (ii) See page 42
45
3.1.3. Diffraction amplitude:
So far we have not discussed the amplitude and the width of diffraction peaks, which play an
important role in the interpretation of x-ray diffraction data. This requires a more sophisticated
analysis which we outline now.
Figure 3.7 Scattering of x-ray in a solid
𝑹
We consider scattering of x-rays by a solid as is shown in Fig.3.7. An incident plane wave 𝑒
𝐤ˊ𝐑
(wave vector k) is scattered, and a scattered wave 𝑒 (wave vector ′k) is detected. Scattering
occurs due to the interaction of the incident x-rays with the electron charge distributed in a solid
with charge density n(r). The amplitude of scattering by an infinitesimal volume dV is
∆
proportional to the charge at this point, i.e. n(r)dV, and a phase factor 𝑒 acquired by the
scattered wave. The phase shift is equal to
∆𝜑 = 𝐊 ∙ 𝐫 − 𝐊ˊ ∙ 𝐫 = −(𝐊ˊ − 𝐊) ∙ 𝐫 = −∆𝐤 ∙ 𝐫. The field at the detector is the superposition of

the waves scattered from all points of the sample, and, therefore, the total amplitude of scattering
is given by
∆ .
𝐹=∫ 𝑑𝑉𝑛(𝑟)𝑒 (3.21)
The intensity of detected radiation is proportional to |𝐹 |, and therefore F determines the

intensity and the width of diffraction peaks.
46
3.1.4. Scattering from a lattice with basis.
If the crystal structure represents a lattice with a basis, then we should take into account scattering
by the atoms which have non-equivalent positions in a unit cell. The intensity of radiation
scattered in a given Bragg peak will depend on the extent to which the rays scattered from these
basis sites interfere with one another. To take into account basis atoms, first, let us rewrite Eq.
(3.21) for the scattering amplitude at the diffraction condition, in terms of the integral over a unit
cell:
∙( )
𝐹=∑ ∫ 𝑑𝑉𝑛(𝒓 + 𝑻)𝑒 , (3.22)
where the sum is taken over all the lattice vectors T. Taking into account that 𝑛(𝒓 + 𝑻) = 𝑛(𝒓) ,
∙
and that 𝑒 = 1,we obtain
(3.23)
where N is the number of cells in the solid, and we define the structure factor
(3.24)
Assuming that we have s atoms in a unit cell located at r , r , r ’……. it is convenient to write
1 2 3
charge density as the superposition of charge densities n associated with each atom j of the basis:
j
(3.25)
The structure factor may now be written as
(3.26)
where 𝝆 ≡ 𝒓 − 𝒓 . The integral in Eq.(3.26) is known as the atomic form factor :
47
(3.27)
which is determined by the charge density of atom j in the basis. The structure factor is then
(3.28)
Example: structure factor of bcc lattice. A conventional cell of the bcc lattice contains two
identical atoms with coordinates: 𝒓 = 0 and 𝒓 = (𝑎⁄2)(𝑥 + 𝑦 + 𝑧) . Since the atoms are
identical, the atomic form factors are same, i.e.𝑓 = 𝑓 = 𝑓. The reciprocal unit cell is cubic with
a cell side of 2π/a, and the reciprocal vector is given by
(3.29)
The structure factor is, then,
(3.30)
Therefore 𝑆 = 2𝑓, if 𝑚 + 𝑚 + 𝑚 is even, and 𝑆 = 0, if 𝑚 + 𝑚 + 𝑚 is odd.
Thus, diffraction peaks will be observed, e.g., from the (110), (200), (211) planes, but not from
the (100), (111), (210) planes. The later fact is due to the destructive interference from the basis
atoms which cancel some peaks. For example, as is seen from Fig.3.8 for the (100) plane, the
phase difference between successive planes is π, so that the reflected amplitude from two adjacent
planes is 1 + 𝑒 = 0.
48
Figure 3.8 Reflection of x-ray from two adjacent planes
Structure factor of fcc lattice. The basis of the fcc structure referred to the cubic cell has identical
atoms at r = 000; r = ½½0; r = 0½½; r = ½0½. Therefore,
1 2 3 4
(3.31)
which is non-zero only if all the indices are even or all the indices are odd. Allowed peaks are,
e.g., (111), (200), (222), (220), (131).
Atomic form factor: The atomic form factor is defined by Eq.(3.27). It depends on the number and
distribution of atomic electrons, and on the wavelength and angle of scattering of the radiation. It
measures the scattering power of the j-th atom in the unit cell. An example of the importance of the
atomic form factor is given below.
49
Fig. 3.9 Comparison of x-ray reflections from KCl and KBr powders
Both KCl and KBr have sodium chloride structure. In this structure the two types of atoms are
arranged alternatively at the lattice sites of a simple cubic lattice. The space lattice is fcc with a
basis of two non-equivalent atoms at 000 and ½½½.
In KCl, the numbers of electrons of K+ and Cl- ions are equal and the charge distribution is
+ -
similar. Therefore, the form factors for K and Cl are almost exactly equal, so that the crystal
looks to x-rays as if it were a monatomic simple cubic lattice of lattice constant a/2. Only even
integers occur in the reflection indices when these are based on a cubic lattice of lattice constant a.
- +
In KBr the form factor of Br is quite different than that of K , and all reflections of the fcc lattice
are present.
50
3.1.5 Uses of X-ray Diffraction
(i) Electron diffraction is particularly useful in exploring the structure of thin surface layers such
as oxide layers on metal surfaces.
(ii) Electron diffraction helps us to study orientation, lattice parameters and perfection of
evaporated thin films.
(iii) Electron diffraction helps us to see regularities of the atomic arrangement in a thin film which
varies over regions of crystal imperfection.
In-text question
1. Q: Discuss briefly the structure factor and atomic force factor.

A: See pages 46-47
2. Q: Give three uses of X-ray diffraction.
A: See page 49

In this study session, you have learnt:
1. The basis of the study of x-ray diffraction is the Bragg’s reflection.
2. The brillouin zones have effect on the study of x-ray diffraction.
3. about the scattering of x-rays from a lattice with a basis.
4. about the uses of x-ray diffraction.

SAQ 3.1
What is the de Broglie wavelength of an electron moving with velocity of ?(take the value of
the rest mass of the electron as 9.11 x 10-31).
51
SAQ 3.2
A certain orthorhombic crystal has a ratio a: b: c =0.429: 1: 0.377. Find the Miller indices of the
faces whose intercepts are:
0.214:1:0.188
0.858:1:0.754
0.429:∞:0.126
SAQ 3.3
What is the de Broglie wavelength of neutrons at room temperature? Can they be used to study
crystal structure?
SAQ 3.4
Deduce Bragg’s law in X-ray diffraction. Describe Bragg’s spectrometer and explain how it is
used to determine the wavelength of x-rays.
SAQ 3.5
Derive Bragg’s law of x-ray diffraction in crystals. Give an account of powder method of crystal
structure analysis.
Ashcroft N. W. and Mermin N.D. (1976), Solid state Physics, Holt, Rine hart, Winston.
52
Study Session 4 Thermal Properties of Crystal Lattice
Introduction
Heat capacity per unit mass of a substance is known as specific heat. Specific heat is really a
measure of the number of degrees of freedom of a system. Since degrees of freedom imply
freedom to absorb potential or kinetic energy, the question to be answered is how many ways
energy can be given to a system. The system we would concern with is the oscillating lattice, but
its analysis is quite difficult, particularly in three dimensions. Instead, we would carry out the
development of the theory in a series of warming-up exercises, each of which bears a little more
to reality. Atoms vibrate about their mean equilibrium lattice sites in solids. These vibrations
occur at any temperature, even near absolute zero. They are almost entirely responsible for the
thermal properties- heat capacity, thermal conductivity, thermal expansion, etc., of insulators and
contribute the greater of the heat capacity of metals. (The conduction electrons contribute only a
small part of the heat capacity of metals but are almost entirely responsible for thermal
conductivity).

4.1 Describe the behavior of lattices of crystals to heat.
4.2 Explain the assumption made through classical and quantum theories.
4.3 Describe Einstein’s and Debye’s theories of the specific heat of crystal lattices
4.4 Discuss the successes and failures of these theories.
4.5 Explain thermal conductivity with respect to crystal lattices.
4.1 Main Content
4.1.1 Heat Capacity

There are two contributions to thermal properties of solids: one comes from phonons (or lattice
vibrations) and another from electrons. This section is devoted to the thermal properties of solids
due to lattice vibrations (the contribution from electrons in metals will be considered separately).
53
First, we consider the heat capacity of the specific heat. The heat capacity C is defined as the heat
ΔQ which is required to raise the temperature by ΔT, i.e.
(4.1)
If the process is carried out at constant volume V, then ΔQ = ΔE, where ΔE is the increase in
internal energy of the system. The heat capacity at constant volume C is therefore given by
V
(4.2)
The contribution of the phonons to the heat capacity of the crystal is called the lattice heat
capacity. The total energy of the phonons at temperature T in a crystal can be written as the sum
of the energies over all phonon modes, so that
(4.3)
Where 〈𝑛𝒒𝒑 〉 is the thermal equilibrium occupancy of phonons of wavevector q and mode p (p =
1…3s, where s is the number of atoms in a unit cell). The angular brackets denote the average in
thermal equilibrium. Note that we assume here that the zero-point energy is chosen as the origin
of the energy, so that the ground energy lies at zero. Now we calculate this average.
Consider a harmonic oscillator in a thermal bath. The probability of finding this oscillator in an
excited state, which is characterized by a particular energy E is given by the Boltzmann
n
distribution:
(4.4)
where the constant Po is determined from the normalization condition
54
(4.5)
So that
(4.6)
The average excitation number of the oscillator is given by
(4.7)
The summation in the enumerator can be performed using the known property of geometrical
progression:
(4.8)
Using this property we find:
(4.9)
ħ ⁄
Where 𝑥 = 𝑒 . Thus we obtain
(4.10)
The distribution given by Eq. (4.10) is known as the Planck distribution. Coming back to the
expression for the total energy of the phonons, we find that
(4.11)
Usually it is convenient to replace the summation over q by an integral over frequency. In order to
do this we need to introduce the density of modes or the density of states D (ω). D (ω)dω
p p
55
represents the number of modes of a given number s in the frequency range (ω, ω + dω). Then the
energy is
(4.12)
The lattice heat capacity can be found by differentiation of this equation with respect to
temperature, so that
(4.13)
We see that the central problem is to find the density of states D (ω), the number of modes per
p
unit frequency range.
In-text question
1. Q: In Aluminium, 𝜈 = 6.32 𝑋 10 𝑚/𝑠 and 𝜈 = 3.10 𝑋 10 𝑚/𝑠. The density of aluminium is

2.7 𝑋 10 𝑘𝑔/𝑚 and its atomic weight is 26.97. Calculate the Debye cut-off frequency for
aluminium from these data.
solution
From n = N/V= ρ/𝜇 = density of material/atomic weight.
Using the formula, 𝜈 = + and substituting the parameters given in the

question to have
A: debye frequency = 8.44 x 1012 Hz.

2. Q: Give the assumptions of classical theory of lattice specific heat.
A: see pages 5.3-56
4.1.2 Density of States

Consider the longitudinal waves in a long bar. The solution for the displacement of atoms is given
by
𝑢(𝑥) = 𝐴𝑒 (4.14)
56
where we omitted a time-dependent factor as irrelevant for the present discussion. We shall now
consider the effects of the boundary conditions on this solution. These boundary conditions are
determined by the external constraints applied to the ends of the bar. The most convenient type of
boundary condition is known as the periodic boundary condition. By this we mean that the right
end of the bar is constrained in such a way that it is always in the same state of oscillation as the
left end. It is as if the bar were deformed into a circular shape so that the right end joined the left.
Given that the length of the bar is L, if we take the origin as being at the left end, the periodic
condition means that
𝑢(𝑥 = 0) = 𝑢(𝑥 = 𝐿), (4.15)

where u is the solution given by Eq.(4.14). If we substitute (4.14) into (4.15), we find that
(4.16)
This equation imposes a condition on the admissible values of q i.e.
cos 𝑞𝐿 + 𝑖 sin 𝑞𝐿 = 1
Equating the real and imaginary parts,
cos qL=1, sin qL =0,
This implies 𝑞𝐿 = 2𝑛𝜋,
and 𝑞 = 𝑛 (4.17)
where n = 0, + 1, ±2, etc. When these values are plotted along a q-axis, they form a one-
dimensional mesh of regularly spaced points. The spacing between the points is 2π/L. When the
bar length is large, the spacing becomes small and the points form a quasi-continuous mesh.
Each q-value of (4.17) represents a mode of vibration. Let us choose an arbitrary interval dq in q-
space, and look for the number of modes whose q’s lie in this interval. We assume here that L is
large, so that the points are quasi-continuous, which is true for the macroscopic objects. Since the
spacing between the points is 2π/L, the number of modes is
= 𝑑𝑞 (4.18)
We are interested in the number of modes in the frequency range dω lying between (ω, ω + dω). The
density of states D(ω) is defined such that D(ω)dω gives this number. Comparing this definition with
57
(4.18), one may write D(ω)dω = (L/2π) dq, or D(ω) = (L/2π)/(dω/dq). We note from Fig. 4.1, however,
that in calculating D(ω) we must include the modes lying in the negative q-region as well as in the
positive region. The effect is to multiply the above expression for D(ω) by a factor of two. That is,
𝐷(⍵) = (4.19)
⍵⁄
We see that the density of states D(ω) is determined by the dispersion relation ⍵=⍵(q).
Figure 4.1 Density of states at different modes
Now we extend these results to the 3D case. The wave solution analogous to (4.14) is
(4.20)
where the propagation is described by the wave vector q = (q q q ), whose direction specifies the
x, y, z
direction of wave propagation. Here again we need to take into account the boundary conditions. For
simplicity, we assume a cubic sample whose edge is L. By imposing the periodic boundary conditions,
we find that the allowed values of q must satisfy the condition
(4.21)
Therefore, the values are given by
(4.22)
where l, m, n are some integers.
58
If we plot these values in a q-space, as in Fig. 4.2, we obtain a three-dimensional cubic mesh. The
3
volume assigned to each point in this q-space is (2π/L) .
Figure 4.2 Allowed values of q for a wave traveling in 3 dimensions.
Only the cross section in the (qx, qy) plane is shown. The shaded circular shell is used for counting
the modes.
Each point in Fig. 4.2 determines one mode. We now wish to find the number of modes lying in
the spherical shell between the radii q and q + dq, as shown in Fig.4.2. The volume of this shell is
3
4𝜋𝑞 𝑑𝑞 and since the volume per point is (2π/L) , it follows that the number we seek is
4𝜋𝑞 𝑑𝑞 = 4𝜋𝑞 𝑑𝑞 (4.23)

( )
3
where V = L is the volume of the sample. By definition of the density of modes, this quantity is
equal to D(ω)dω . Thus, we arrive at
59
(4.24)
We note that Eq. (4.24) is valid only for an isotropic solid, in which the vibrational frequency, ω,
does not depend on the direction of q. Also we note that in the above discussion we have
associated a single mode with each value of q. This is not quite true for the 3D case, because for
each q there are actually three different modes, one longitudinal and two transverse, associated
with the same value of q. In addition, in the case of non-Bravais lattice we have a few sites, so that
the number of modes is 3s, where s is the number of non-equivalent atoms. This should be taken
into account by index p = 1…3s in the density of states as was done before, because the dispersion
relations for the longitudinal and transverse waves are different, and acoustic and optical modes
are different.
In-text question
1. Q: Derive the density of modes of atoms in a crystal when subjected to thermal effects.
A: See pages 54-58
4.1.3 Debye Model

The Debye model assumes that the acoustic modes give the dominant contribution to the heat
capacity. Within the Debye approximation, the velocity of sound is taken as a constant
independent of polarization as it would be in a classical elastic continuum. The dispersion relation
is written as
𝜔 = 𝑣𝑞 (4.25)
Where 𝑣 is the velocity of sound.

In this approximation, the density of states is given by
(4.26)
i.e., the density of states increases quadratically with the frequency.
60
The normalization condition for the density of states determines the limits of integration over ω.
The lower limit is obviously ω=0. The upper limit can be found from the condition that the
number of vibrational modes in a crystal is finite and is equal to the number of degrees of freedom
of the lattice. Assuming that there are N unit cells is the crystal, and there is only one atom per cell
(so that there are N atoms in the crystal), the total number of phonon modes is 3N. Therefore, we
can write
(4.27)
where the cutoff frequency ω is known as Debye frequency. Assuming that the velocity of the
D
three acoustic modes is independent of polarization and substituting (4.26) in (4.27), we obtain:
(4.28)
The cutoff wavevector which corresponds to this frequency is given by
(4.29)
so that modes of wavevector larger than q are not allowed. This is due to the fact that the number
D
of modes with q≤q exhausts the number of degrees of freedom of the lattice.
D
The thermal energy is given by Eq. (4.12), so that
(4.30)
where a factor of 3 is due to the assumption that the phonon velocity is independent of
polarization. This leads to
61
(4.31)
Where 𝑥 ≡ ħ⍵⁄𝑘 𝑇 𝑎𝑛𝑑
𝑥 ≡ ħ⍵ ⁄𝑘 𝑇 ≡ 𝜃 ⁄𝑇 (4.32)
The latter expression defines the Debye temperature
(4.33)
The total phonon energy is then
(4.34)
where N is the number of atoms in the crystal and 𝑥 ≡ 𝜃 ⁄𝑇
The heat capacity is most easily found by differentiating the middle expression of (4.31) with
respect to the temperature (in Eq.(4.34) we will have to differentiate the upper limit) so that
(4.35)
In the limit T>>θ, we can expand the expression under the integral and obtain: 𝐶 = 3𝑁𝑘 . This is
exactly the classical value for the heat capacity, which is known from the elementary physics.
Recall that according to the elementary thermodynamics, the average thermal energy per degree
of freedom is equal to 𝑘 𝑇 . Therefore for a system of N atoms 𝐸 = 3𝑁𝑘 𝑇 which results in 𝐶 =
3𝑁𝑘 . This is known as the Dulong-Petit law.
62
Now consider an opposite limit, i.e. T<<θ. At very low temperatures we can approximate (4.34)
by letting the upper limit go to infinity. We obtain
(4.36)
and therefore
(4.37)
3
We see that within the Debye model at low temperatures, the heat capacity is proportional to T .
The cubic dependence may be understood from the following qualitative argument. At low
temperature, only a few modes are excited. These are the modes whose quantum energy ħ⍵ is less
than k T. The number of these modes may be estimated by drawing a sphere in the q-space whose
B
frequency ⍵ = 𝑘 𝑇⁄ħ , and counting the number of points inside, as shown in Fig. 4.3. This
sphere may be called the thermal sphere, in analogy with the Debye sphere discussed above. The
3 3 3
number of modes inside the thermal sphere is proportional to q ~ ω ~ T . Each mode is fully
excited and has an average energy equal to k T. Therefore the total energy of excitation is
B
4 3
proportional to T , which leads to a specific heat proportional to T , in agreement with (4.37).
63
Figure 4.3 The thermal sphere which is the frequency contour ⍵ = 𝒌𝑩 𝑻⁄ħ
To compare these predictions with experimental results one should know the Debye temperature.
This temperature is normally determined by fitting experimental data. Fig.4.4 shows the fitted
data versus the reduced temperature T/θ. You see that the curve is universal; it is the same for
different substances. The agreement between the calculated and experimental data is remarkable.
Figure 4.4. Specific heats versus reduced temperature for four substances .
64
In-text question
1. Q: At what temperature does the electronic contribution to the specific heat of silver become
identical with the Dulong-Petit value? E f = 5eV.
Solution
From 𝐸 = 2𝑘 𝑇 , where 𝑘 = 8.617 3324(78) × 10−5 eVK−1 ,
A: 3.125 x 104 K
2. Q: Write the expression for the classical value of molar lattice specific heat.
A: 3Ru
4.1.4 Einstein Model

Within the Einstein model the density of states is approximated by a delta function at some
frequency ωE, i.e
𝐷(𝜔) = 𝑁𝛿(𝜔 − 𝜔 ) (4.38)
where N is the total number of atoms (oscillators). ωE is known as the Einstein frequency. The
thermal energy of the system (4.12) is then
ħ
𝐸= ħ ⁄ (4.39)
where a factor of 3 reflects the fact that there are three degree of freedom for each oscillator. The
heat capacity is then
(4.40)
The high temperature limit for the Einstein model is the same as that for the Debye model, i.e.
𝐶 = 3𝑁𝑘 , which is the Dulong-Petit law. At low temperatures however (4.40) decreases
3
ħ⍵ ⁄
as 𝐶 ~ 𝑒 , which is different from the Debye T law. The reason for this disagreement is
65
that at low temperatures only acoustic phonons are populated and the Debye model is much better
approximation that the Einstein model. The Einstein model is often used to approximate the
optical phonon part of the phonon spectrum.
Concluding our discussion about the heat capacity of solids, we note that a real density of
vibrational modes could be much more complicated than those described by the Debye and
Einstein models. Fig.4.5 shows the density of states for Cu. The dash line is the Debye
approximation, which has the same area (under the curve) as the solid curve. The Einstein
approximation would have a delta peak at some frequency. At low frequencies the density of
states varies quadratically with the frequency, which is due to acoustic modes and similar to that
within the Debye approximation. At higher frequencies, there is a peak which is due to optical
modes. This density of states has to be included in order to obtain a quantitative description of
experimental data.
Figure 4.5. Total density of states for Cu, as deduced from data on neutron
scattering.
In-text question
1. Q: State Dulong-Petit’s law and show how the departure from this law at lower temperatures
has been explained by Einstein’s theory.
A: See page 62-65
66
2. Q: Discuss the variation of specific heat capacity of solids with temperature and use Einstein’s
theory to explain it.
A: See pages 63-64

1. At high temperatures, the experimental deduction on specific heat is in accordance with the
classical theory on specific heat (Dulong-Petit theory) while at low temperatures, the classical
theory fails.
2. Einstein’s approximation shows that lattice specific heat varies with temperature.
3. At high temperatures, Einstein’s result agrees with the classical and experimental results
because there is increase in the lattice vibration and also due to high quantum number
(correspondence principle).
4. Debye’s theory agrees with the experimental deductions both at low and high temperatures.
5. Debye’s theory gives a perfect insight into the lattice specific heat of solids.

SAQ 4.1
Show that the kinetic energy of a three-dimensional electron gas of N electrons at zero
temperature is U=3/5NE .
F
SAQ 4.2
Show that the density of states of a free-electron gas in two dimensions is independent of energy.
67
SAQ 4.3
Fcc Au (cubic lattice parameter a=4.08Å) has electrical resistivity ρ=2.2μΩcm at room
temperature. Using a free-electron model and assuming one valence electron per atom calculate:
a) the concentration of the conduction electrons n,

b) the Fermi energy,
c) the relaxation time τ,
d) the Fermi velocity v ,
F
e) the mean free path l,

f) the electronic heat capacity per atom at room temperature.
SAQ 4.4
The residual resistivity for 1 atomic percent of As impurities in Cu is 6.8μΩcm. Calculate the
cross section for the scattering of an electron by one As impurity in Cu. Use a free-electron model
assuming that Cu has the fcc structure with the cubic lattice parameter a=3.62Å and one valence
electron per atom.
Ashcroft N. W. and Mermin N.D., (1976), Solid state Physics, Holt, Rine hart, Winston.
68
Study Session 5 Elastic Properties
Introduction
All materials in nature are elastic. Thus, solids stretch in the direction in which they are pulled.
Any force or load applied on the material will result in stress and strain in the material. Stress
represents the intensity of the reaction force at any point in the body as imposed by service load,
assembly conditions, fabrication and thermal changes.

5.1 Explain the effect of external influences like pressure on the atoms of solid state
materials.
5.2 Describe what stress and strain depict in the elastic properties of materials.
5.3 Calculate the elastic constants of materials
5.1 Main Content

5.1.1 Elastic Constants
Elastic properties of solids are determined by interatomic forces acting on atoms when they are
displaced from the equilibrium positions. At small deformations these forces are proportional to
the displacements of atoms. As an example, consider a 1D solid. A typical binding curve has a
minimum at the equilibrium interatomic distance R 0 :
Figure 5.1 Graph of potential energy against interatomic distance

69
Expanding the energy at the minimum in the Taylor series we find:
𝑈(𝑅) = 𝑈 + │ (𝑅 − 𝑅 ) + │ (𝑅 − 𝑅 ) + ⋯ (5.1)
At equilibrium │ = 0, so that
𝑈(𝑅) = 𝑈 + 𝑘𝑢 , (5.2)
2
𝜕 𝑈
where we defined 𝑘= │𝑅 and 𝑢 = 𝑅 − 𝑅 is the displacement of an atom from
𝜕𝑅2 0
equilibrium position R0. Differentiating Eq.(5.2), 𝐹 = − , we obtain the force F acting on an
atom:
𝐹 = −𝑘𝑢 (5.3)
The constant k is an interatomic force constant. Eq.(5.3) represents the simplest expression for the
Hooke’s law showing that the force acting on an atom, F, is proportional to the displacement u.
This law is valid only for small displacements and characterizes a linear region in which the
restoring force is linear with respect to the displacement of atoms.
The elastic properties are described by considering a crystal as a homogeneous continuum

medium rather than a periodic array of atoms. In a general case the problem is formulated as
follows:
(i) Apply forces, which are described in terms of stress σ, and determine displacements of
atoms which are described in terms of strain ε.
(ii) Define elastic constants C relating stress σ and strain ε, so that σ = Cε.
Example: In 1D case, F =−ku , where u is a change in the crystal length under applied force F. We
can therefore write
70
𝜎= = = 𝐶𝜀 , (5.4)
where A is the area of the cross section, and L is the equilibrium length of the 1D crystal. The
stress σ is defined as the force per unit area and the strain ε is the dimensionless constant which
describes the relative displacement (deformation).
In a general case of a 3D crystal the stress and the strain are tensors which are defined as follows.
Stress has the meaning of locally applied “pressure”. It has components 𝜎 , showing that the
force can be applied along 3 directions “i” and 3 faces “j”. The stress is defined locally, so that
𝜎 = 𝜎 (𝒓).
Shear forces must come in pairs to conserve angular acceleration inside the crystal:
71
That makes the stress tensor diagonal, i.e.
𝜎 =𝜎 . (5.5)
Strain determines relative atomic displacement:
𝜀 (𝒓) = , (5.6)
where ui is displacement in “i” direction and xj is the direction along which ui may vary.
Compression strain(𝜀 𝜀 𝜀 ):
𝜀 = , (5.7)
In a homogeneous crystal 𝜀 is a constant𝜀 = , where u is the change in the crystal length L.
Shear strain(𝜀 𝜀 𝜀 𝜀 𝜀 𝜀 ):
(5.8)
Since 𝜎 and 𝜎 must always be applied together, we can define shear strains symmetrically:
𝜀 =𝜀 = + . (5.9)
Elastic constants C relate the strain and the stress in a linear fashion:
𝜎 = 𝑐 𝜀 (5.10)
Eq.(5.10) is a general form of the Hooke’s law. The matrix C in a most general form has
3x3x3x3=81 components. However, due to the symmetrical form of 𝜎 and 𝜀 - each of them
72
has 6 independent components, we need only 36 elastic constants. There is a convention to denote
these constants by Cmn, where indices m and n are defined as 1 = xx, 2 = yy, 3 = zz for the
compression components and as 4 = yz, 5 = zx, 6 = xy for the shear components. For example,
𝐶 =𝐶 ,𝐶 =𝐶 , 𝐶 =𝐶 .
Therefore, the general form of the Hooke’s law is given by
(5.11)
All 36 elastic constants are independent. However in crystals many of them are the same due to
symmetry. In particular, in cubic crystals C11 = C22 = C33 , 𝐶 =𝐶 =𝐶 =𝐶 =𝐶 =𝐶 ,
𝐶 =𝐶 =𝐶 due to the fact that x, y, and z axes are identical by symmetry. Also, the off
diagonal shear components are zero, i.e.𝐶 =𝐶 =𝐶 =𝐶 =𝐶 =𝐶 = 0, and mixed
compression/shear coupling does not occur, i.e.. 𝐶 =𝐶 =. . . = 0 .Therefore, the cubic
elasticity matrix has the form
73
(5.12)
We have only 3 independent constants.
74
In-text question
1. Q: Consider a longitudinal wave 𝑈 = 𝐴𝑐𝑜𝑠(𝑞𝑛𝑎 − ⍵𝑡) in a monoatomic linear lattice of

atoms of mass M, spacing a and nearest-neighbour interaction C.
(i)Show that the total energy of the wave is
,
Where n runs over all atoms.
A: The total energy of the wave is the sum of the kinetic energy E and the potential energy
kin
E . The kinetic energy is the sum of the kinetic energies of all atoms, i.e.
pot
1 𝑑𝑢
𝐸 = 𝑀
2 𝑑𝑡
where M is the mass of atoms and is the velocity of n-th atom. The potential energy is the
potential energy of all the “springs” connecting atoms. For two atoms, this energy is the same
as that for a harmonic oscillator, i.e. 𝐶𝑥 , where C is interatomic force constant and x is the
change in distance between the atoms from the equilibrium distance. Therefore, for atoms n
and n+1 having displacements 𝑢 and 𝑢 respectively this energy is 𝐶(𝑢 − 𝑢 ) . The
total potential energy is given by)
1
𝐸 = 𝐶 (𝑢 −𝑢 )
2
The total energy of the wave is therefore
𝐸=𝐸 +𝐸 = 𝑀 ∑ + 𝐶 ∑ (𝑢 −𝑢 ) .
75
5.1.2 Elastic waves
So far, we have assumed that atoms were at rest at their lattice sites. Atoms, however, are not
quite stationary, but can oscillate around their equilibrium positions (e.g., as a result of thermal
energy). This leads to lattice vibrations.
When considering lattice vibrations three major approximations are made:

(i) It is assumed that displacements of atoms are small, i.e.𝑢 ≪ 𝑎, where a is a lattice parameter.
(ii) Forces acting on atoms are assumed to be harmonic, i.e. proportional to the displacements:
F = -Cu . This is the same approximation which is used to describe a harmonic oscillator.
(iii) It is assumed that the adiabatic approximation is valid – electrons follow atoms, so that the
nature of bond is not affected by vibrations.
The discreteness of the lattice must be taken into account in the discussion of lattice vibrations.
However, when the wavelength is very long, i.e.. 𝜆 ≪ 𝑎, one may disregard the atomic nature and
treat the solid as a continuous medium. Such vibrations are referred to as elastic waves.
We consider an elastic wave in a long bar of cross-sectional area A and mass density M /V .
(1) First, we consider a longitudinal wave of compression/expansion.
We look at a segment of width dx at the point x and denote the elastic displacement by u.
According to the Newton’s second low
𝑚 = ∑ 𝐹, (5.13)
Which implies that
76
(𝜌𝐴𝑑𝑥) = 𝐹(𝑥 + 𝑑𝑥) − 𝐹(𝑥) , (5.14)
𝜌 = = , (5.15)
where we introduce the compression stress 𝜎 . Assuming that the wave propagates along the
[100] direction, we can write the Hooke’s law in the form
𝜎 =𝑐 𝜀 , (5.16)
where C11 is Young’s modulus. Since 𝜀 = , this leads to the wave equation
= . (5.17)
A solution of the wave equation has the form of a propagating longitudinal plane wave
( )
𝒖(𝑥, 𝑡) = 𝐴𝑒 𝒙, (5.18)
where q is the wave vector,
𝜔=𝑉 𝑞, (5.19)
is the frequency, and
𝑉 = (5.20)
is the longitudinal velocity of sound.
(2) Now we consider a transverse wave which is controlled by shear stress and strain.
77
In this case
𝜌 = , (5.21)
where the shear stress 𝜎 is determined by the shear modulus C44 and shear strain 𝜀 =
𝜎 =𝑐 𝜀 (5.22)
Therefore Eq.(5.21) takes the form
𝜌 =𝐶 =𝐶 , (5.23)
resulting in the wave equation
= (5.24)
This is the equation for the transverse plane wave, which has displacements in the y direction but
propagates in the x direction:
( )
𝒖(𝑥, 𝑡) = 𝐴𝑒 𝒚, (5.25)
where q is the wave vector,
𝜔=𝑉 𝑞,
78
in the frequency, and
𝐶44
𝑉 = (5.26)
is the transverse velocity of sound. Note that there are two linear independent transverse modes
characterized by the displacements in y and in z directions. For the [100] direction, by symmetry
the velocities of these modes are the same and given by Eq.(5.26).
Normally 𝐶 >𝐶 and therefore𝑉 > 𝑉 .
Waves we have considered are in [100] direction, i.e. q || [100]. In other directions, the sound
velocity depends on combinations of elastic constants:
𝑉= , (5.27)
where Ceff is an effective elastic constant which is given for cubic crystals in the table:
The relation connecting the frequency 𝜔 and the wave vector q is known as the dispersion
relation. For elastic waves, 𝜔 is proportional to q, and the ratio 𝜔/q gives a constant velocity. The
figure below shows the dispersion relation for elastic waves. There are three modes – one
longitudinal and two transverse, which represent straight lines whose slopes are equal to the
respective velocities of sound. For the [100] and [111] directions the two transverse modes are
degenerate, i.e. have the same vT.
79
In-text question
1. Q: List the major approximations considered when dealing with lattice vibrations.
A: See page 74

1. The general Hooke’s law has been transformed into 36 elastic constants.
2. All these 36 elastic constants are independent. However in crystals many of them are the same
due to symmetry. In particular, in cubic crystals C11 = C22 = C33 , 𝐶 =𝐶 =𝐶 =𝐶 =𝐶 =
𝐶 ,𝐶 =𝐶 =𝐶 due the fact that x, y, and z axes are identical by symmetry. Also the off
diagonal shear components are zero, i.e.𝐶 =𝐶 =𝐶 =𝐶 =𝐶 =𝐶 = 0, and mixed
compression/shear coupling does not occur, i.e., 𝐶 =𝐶 =. . . = 0 .
3. These 36 elastic constants have been reduced to 3 independent constants.
4. Approximation were made assuming the atoms were not at rest at their lattice sites and these
led to two major vibrational modes of the atoms (optical and acoustical mode).
80
Self Assessment Questions for Study Session 5
SAQ 5.1
Consider a longitudinal wave 𝑈 = 𝐴𝑐𝑜𝑠(𝑞𝑛𝑎 − ⍵𝑡) in a mono-atomic linear lattice of atoms of
mass M, spacing a and nearest-neighbour interaction C.
(a) Show that the time-averaged kinetic energy is equal to the time-averaged potential
energy.
(b) Show that the total time-averaged energy per atom is equal to ½MA2⍵2.
SAQ 5.2
Consider a linear chain in which alternative ions have masses M1 and M2 and only nearest
neighbours interact.
(a) Discuss the form of the dispersion relation and the nature of the vibrational
modes when M1 >> M2..
(b) Show that for M1=M2 the dispersion relation becomes identical to that for the
monoatomic lattice.
SAQ 5.3
Consider the normal modes of a linear chain in which the force constants between nearest-
neighbour atoms are alternatively C and 10C. Assuming that the masses are equal and the nearest
neighbour separation is a/2. Find ⍵(q) at q = 0 and q = 𝜋/a. Sketch the dispersion curve. This
problem simulates a crystal of diatomic molecules such as H 2.
Kittel C., (1963), Quantum theory of solids, Wiley.
81
Study Session 6 Lattice Vibration
Introduction
So far we have been discussing equilibrium properties of crystal lattices. When the lattice is at
equilibrium each atom is positioned exactly at its lattice site. Now suppose that an atom is
displaced from its equilibrium site by a small amount. Due to external force force acting on this
atom, it will tend to return to its equilibrium position. This results in lattice vibrations. Due to
interactions between atoms, various atoms move simultaneously, so we have to consider the
motion of the entire lattice.

6.1 Explain the classical approach to the lattice vibrations in terms of 1-d, 2-d, 3-d.
6.2 Discuss the treatment of lattice vibrations as phonons using the quantum approach.
6.1 Main Content

6.1.1 One-dimensional lattice
For simplicity we consider, first, a one-dimensional crystal lattice and assume that the forces
between the atoms in this lattice are proportional to relative displacements from the equilibrium
positions.
Figure 6.1 Monatomic Lattice
This is known as the harmonic approximation, which holds well provided that the displacements
are small. One might think about the atoms in the lattice as interconnected by elastic springs.
Therefore, the force exerted on n-th atom in the lattice is given as
82
𝐹 = 𝐶(𝑈 − 𝑈 ) + 𝐶(𝑈 −𝑈 ), (6.1)
where C is the interatomic force (elastic) constant. Applying Newton’s second law to the motion
of the n-th atom we obtain
𝑀 = 𝐹 = 𝐶(𝑈 − 𝑈 ) + 𝐶(𝑈 − 𝑈 ) = −𝐶(2𝑈 − 𝑈 −𝑈 ), (6.2)
where M is the mass of the atom. Note that we neglected here by the interaction of the n-th atom
with all but its nearest neighbours. A similar equation should be written for each atom in the
lattice, resulting in N coupled differential equations, which should be solved simultaneously (N is
the total number of atoms in the lattice). In addition, the boundary conditions applied to the end
atom in the lattice should be taken into account.
Now let us attempt a solution of the form
( )
𝑈 = 𝐴𝑒 (6.3)
where xn is the equilibrium position of the n-th atom so that xn=na. This equation represents a
travelling wave, in which all the atoms oscillate with the same frequency and the same
amplitude A and have wavevector q. Note that a solution of the form (6.3) is only possible because
of the transnational symmetry of the lattice.
Now substituting Eq.(6.3) into Eq.(6.2) and cancelling the common quantities (the amplitude and
the time-dependent factor) we obtain
( ) ( )
𝑀(−𝜔 )𝑒 = −𝐶 2𝑒 −𝑒 −𝑒 (6.4)
This equation can be further simplified by cancelling the common factor 𝑒 , which leads to
𝑀𝜔 = 𝐶 2 − 𝑒 −𝑒 = 2𝐶(1 − 𝑐𝑜𝑠 𝑞𝑎) = 4𝐶𝑠𝑖𝑛 . (6.5)
83
We find therefore the dispersion relation for the frequency
𝜔= sin , (6.6)
which is the relationship between the frequency of vibrations and the wavevector q. This
dispersion relation has a number of important properties.
(i) Reducing to the first Brillouin zone. The frequency (6.6) and the displacement of the atoms
(6.3) do not change when we change q by q+2/a. This means that these solutions are physically
identical. This allows us to set the range of independent values of q within the first Brillouin zone,
i.e.
− ≤𝑞 ≤ . (6.7)
Within this range of q the versus q is shown in Figure 6.2.
Figure 6.2 Dispersion curve in the first brillouin zone
The maximum frequency is 4𝐶 ⁄𝑀 . The frequency is symmetric with respect to the sign change
in q, i.e. (q) = (-q). This is not surprising because a mode with positive q corresponds to the
84
wave travelling in the lattice from the left to the right and a mode with a negative q corresponds to
the wave travelling from the right to the left. Since these two directions are equivalent in the
lattice the frequency does not change with the sign change in q.
At the boundaries of the Brillouin zone q = /a the solution represents a standing wave
𝑈 = 𝐴(−1) 𝑒 :
(ii) Phase and group velocity. The phase velocity is defined by
𝑣 = (6.8)
and the group velocity by
𝑣 = (6.9)
The physical distinction between the two velocities is that v p is the velocity of the propagation of
the plane wave, whereas the vg is the velocity of the propagation of the wave packet. The latter is
the velocity for the propagation of energy in the medium.
For the particular dispersion relation (6.6) the group velocity is given by
𝑣 𝑐𝑜𝑠 . (6.10)
As is seen from Eq. (6.10) the group velocity is zero at the edge of the zone where q=/a. Here
the wave is a standing wave and therefore the transmission velocity for the energy is zero.
(iii) Long wavelength limit. The long wavelength limit implies that >>a. In this limit qa<<1.
Expanding the sine in Eq.(6.6) to obtain the positive frequencies:
85
(𝑞𝑎)
sin 𝑞𝑎 − +⋯
3!
for 𝑞𝑎 ≪ 1, sin 𝑞𝑎 ≈ 𝑞𝑎, then
𝜔= 𝑞𝑎 . (6.11)
We see that the frequency of vibration is proportional to the wave vector. This is equivalent to the
statement that velocity is independent of frequency. In this case
𝑣 = = 𝑎 (6.12)
This is the velocity of sound for the one dimensional lattice which is consistent with the
expression we obtained earlier for elastic waves.
In-text question
1. Q: Briefly explain the harmonic approximation and use it to deduce the dispersion relation.
A: See page 80.
2. Q: Highlight the important properties of the dispersion relation.
A: See pages 82-84.
6.1.2 Diatomic 1D lattice

Now we consider a one-dimensional lattice with two non-equivalent atoms in a unit cell. It
appears that the diatomic lattice exhibits important features different from the monoatomic case.
Fig. 6.3 shows a diatomic lattice with the unit cell composed of two atoms of masses M1 and M2
with the distance between two neighbouring atoms a.
Figure 6.3 Diatomic lattice

86
We can treat the motion of this lattice in a similar fashion as for monoatomic lattice. However, in
this case because there are two different kinds of atoms, we write two equations of motion:
𝑑 𝑢
𝑀 = −𝐶(2𝑢 − 𝑢 −𝑢 )
𝑑𝑡
(6.13)
𝑑 𝑢
𝑀 = −𝐶(2𝑢 −𝑢 −𝑢 )
𝑑𝑡
In analogy with the monoatomic lattice we are looking for the solution in the form of travelling
mode for the two atoms:
𝑢 𝐴 𝑒
𝑢 = 𝑒 , (6.14)
𝐴 𝑒 ( )
which is written in the obvious matrix form. Substituting this solution into Eq.(6.13) we obtain
2𝐶 − 𝑀 𝜔 −2𝐶 cos 𝑞𝑎 𝐴
=0 . (6.15)
−2𝐶 cos 𝑞𝑎 2𝐶 − 𝑀 𝜔 𝐴
This is a system of linear homogeneous equations for the unknown quantities A 1 and A2. A
nontrivial solution exists only if the determinant of the matrix is zero. This leads to the secular
equation
(2𝐶 − 𝑀 𝜔 ) (2𝐶 − 𝑀 𝜔 ) − 4𝐶 𝑐𝑜𝑠 𝑞𝑎 = 0 . (6.16)
This is a quadratic equation, which can be readily solved:
𝜔 =𝐶 + ± + − . (6.17)
87
Depending on signs in this formula there are two different solutions corresponding to two
different dispersion curves, as is shown in Figure 6.4:
Figure 6.4 Dispersion curves
The lower curve is called the acoustic branch, while the upper curve is called the optical branch.
The optical branch begins at q = 0 and = 0. Then, with increasing q, the frequency increases in a
linear fashion. This is why this branch is called acoustic: it corresponds to elastic waves or sound.
Eventually this curve saturates at the edge of the Brillouin zone. On the other hand, the optical
branch has a non-zero frequency at zero q
𝜔 = 2𝐶 + . (6.18)
and does not change much with q.
The distinction between the acoustic and optical branches of lattice vibrations can be seen most
clearly by comparing them at q = 0 (infinite wavelength). This follows from Eq. (6.15), for the
acoustic branch = 0 and A1 = A2. So in this limit the two atoms in the cell have the same
amplitude and phase. Therefore, the molecule oscillates as a rigid body, as shown in Fig. 6.5 for
the acoustic mode.
88
Figure 6.5
On the other hand, for the optical vibrations, substituting eq.(6.18) to eq.(6.15), we obtain for q
= 0:
𝑀 𝐴 + 𝑀 𝐴 =0. (6.19)
It implies that the optical oscillation takes place in such a way that the centre of mass of a
molecule remains fixed. The two atoms move in out of phase as shown in Fig. 6.5. The frequency
of these vibrations lies in the infrared region which is the reason for referring to this branch as
optical.
In-text question
1. Q: Highlight the differences between the monoatomic and diatomic atoms as regards lattice
vibration.
A: See pages 80-87.
2. Q: Explain briefly the dispersion curves observed in a diatomic lattice.

A: See pages 84-86.
6.1.3 Three Dimensional Lattice

These considerations for monoatomic and diatomic lattice can be extended to the three-
dimensional lattice. To avoid mathematical details, we shall present only a qualitative discussion.
Consider, first, the monatomic Bravais lattice, in which each unit cell has a single atom. The
equation of motion of each atom can be written in a manner similar to that of Eq.(6.2). The
89
solution of this equation in three dimensions can be represented in terms of normal
modes,Type equation here.
𝐮 = 𝐀e (𝐪𝐫 )
(6.20)
where the wave vector q specifies both the wavelength and direction of propagation. The vector A
determines the amplitude as well as the direction of vibration of the atoms. Thus this vector
specifies the polarization of the wave, i.e., whether the wave is longitudinal (A parallel to q) or
transverse (A perpendicular to q).
When we substitute Eq.(6.20) into the equation of motion, we obtain three simultaneous equations
involving Ax, Ay. and Az, the components of A. These equations are coupled together and are
equivalent to a 3 x 3 matrix equation. The roots of this equation lead to three different dispersion
relations, or three dispersion curves, as shown in Fig. 6.6. All three branches pass through the
origin, which means all the branches are acoustic. This is of course to be expected, since we are
dealing with a monatomic Bravais lattice.
Figure 6.6 Dispersion curves
The three branches in Figure 6.6 differ in their polarization. When q lies along a direction of high
symmetry - for example, the [100] or [110] directions these waves may be classified as either
pure longitudinal or pure transverse waves. In that case, two of the branches are transverse and
one is longitudinal. One usually refers to these as the TA - transverse acoustic and LA
longitudinal acoustic branches, respectively. However, along non-symmetry directions the
waves may not be pure longitudinal or pure transverse, but have a mixed character.
90
Figure 6.7
Figure 6.7 The dispersion curves for Al in the [100] and [110] directions.
In certain high-symmetry directions, such as the [100] in Al, the two transverse branches coincide.
The branches are then said to be degenerate.
We turn our attention now to the non-Bravais three-dimensional lattice. Here the unit cell contains
two or more atoms. If there are s atoms per cell, then on the basis of our previous experience we
conclude that there are 3s dispersion curves. Of these, three branches are acoustic, and the
remaining (3s 3) are optical. The mathematical justification for this assertion is as follows: We
write the equation of motion for each atom in the cell, which results in s equations. Since these are
vector equations, they are equivalent to 3s scalar equations, which have 3s roots. It can be shown
that three of these roots always vanish at q = 0, which results in three acoustic branches. The
remaining (3s 3) roots, therefore, belong to the optical branches, as stated above.
91
Figure 6.8 The dispersion curves for Ge
The acoustic branches may be classified, as before, by their polarizations as TA 1, TA2, and LA.
The optical branches can also be classified as longitudinal or transverse when q lies along a high
symmetry direction, and one speaks of LO and TO branches. As in the one-dimensional case, one
can also show that, for an optical branch, the atoms in the unit cell vibrate out of phase relative to
each other. As an example of a non-Bravais lattice, the dispersion curves for Ge are shown in Fig.
6.8. Since there are two atoms per unit cell in germanium, there are six branches: three acoustic
and three optical. Note that the two transverse branches are degenerate along the [100] direction,
as indicated earlier.
6.1.4 Phonons
So far we discussed a classical approach to the lattice vibrations. As we know from quantum
mechanics, the energy levels of the harmonic oscillator are quantized. Similarly the energy levels
of lattice vibrations are quantized. The quantum of vibration is called a phonon in analogy with
the photon, which is the quantum of the electromagnetic wave.
We know that the allowed energy levels of the harmonic oscillator are,
92
𝐸 = (𝑛 + )ħ𝜔 (6.21)
where n is the quantum number. A normal vibration mode in a crystal of frequency is given by
Eq.(6.20). If the energy of this mode is given by Eq.(6.21) we can say that this mode is occupied
by n phonons of energy ħ. The term ½ħis the zero point energy of the mode.
Let us now make a comparison between the classical and quantum solutions in the one-
dimensional case. Consider a normal vibration
𝒒. 𝒓 = 𝑞𝑥 in this case
( )
𝑢 = 𝐴𝑒 (6.22)
where u is the displacement of an atom from its equilibrium position x and A is the amplitude. The
energy of this vibrational mode averaged over time is
𝐸 = 1 2 𝑀𝜔 𝐴 = 𝑛 + ħ𝜔 = 𝐸 (6.23)
We see that there is a relationship between the amplitude of vibration and the frequency and the
phonon occupation of the mode. In classical mechanics any amplitude of vibration is possible,
whereas in quantum mechanics only discrete values are allowed. This is illustrated in Figure 6.9
Figure 6.9 Amplitude of vibration in Quantum mechanics
93
The lattice with s atoms in a unit cell is described by 3s independent oscillators. The frequencies
of normal modes of these oscillators will be given by the solution of 3s linear equations as we
discussed before. They are 𝜔 (𝒒) , where p denotes a particular mode, i.e. p = 1,…3s. The energy
of this mode is given by
𝐸𝒒 = 𝑛𝒒 + 1 2 ħ 𝜔 (𝒒). (6.24)
where 𝑛𝒒 is the occupation number of the normal mode and is an integer. A vibrational state of
the entire crystal is specified by giving the occupation numbers for each of the 3s modes. The total
vibrational energy of the crystal is the sum of the energies of the individual modes, so that
𝐸 = ∑𝒒 𝐸𝒒 = ∑𝒒 𝑛𝒒 + 1 2 ħ 𝜔 (𝒒) . (6.25)
Phonons can interact with other particles such as photons, neutrons and electrons. This interaction
occurs such as if photon had a momentum ħq . However, a phonon does not carry real physical
momentum. The reason is that the center of mass of the crystal does not change it position under
vibrations (except at q = 0).
In crystals there exist selection rules for allowed transitions between quantum states. We saw that
the elastic scattering of an x-ray photon by a crystal is governed by the wave vector selection rule
kk G , where G is a vector in the reciprocal lattice, k is the wave vector of the incident
photon and kis the wave vector of the scattered photon. This equation can be considered as the
condition for the conservation of the momentum of the whole system, in which the lattice acquires
a momentum ħG .
If the scattering of photon is inelastic and is accompanied by the excitation or absorption of a

phonon the selection rule becomes
kk±𝐪 G , (6.26)
94
where sign (+) corresponds to creation of phonon and sign (–) corresponds to absorption of
phonon.
Phonon dispersion relations 𝜔 (𝒒) can be determined by the inelastic scattering of neutrons with
emission or absorption of phonons. In this case in addition to the condition of momentum
conservation, there is the requirement of conservation of energy. The latter condition can be
written as
ħ ħ ḱ
= ± ħ𝜔 (6.27)
where M is the mass of the neutron and ħ𝜔 and ħkare respectively the momenta of the incident
and scattered
neutron. Once we know in experiment the kinetic energy of the incident and scattered neutrons
from Eq.(6.27), we can determine the frequency of the emitted or absorbed phonon. Then
experimentally, we need to determine those directions, which are characterized by the highest
intensity of the scattered beam. For these directions the conditions (6.26) are satisfied and
therefore from Eq.(6.26), we can find the wave vector of the phonon. Therefore, this is the way to
obtain the dispersion conditions for the frequency of phonons, which we have discussed before.
In-text question
1. Q: Briefly explain what you understand by “degenerate branches” in three dimensional lattices.
A: See pages 88-89.
2. Q: Derive the phonon dispersion relations from the Neutron inelastic scattering method.
A: See pages 92-93.
95
In Study Session 6, you have learnt that:
1. The quantum unit of a lattice vibration is a phonon. If the angular frequency is 𝜔, the energy of
the phonon is ħ𝜔.
2. When a phonon of wave vector K is created by the inelastic scattering of a photon or neutron
from wave vector k to kthe wave vector selection rule that governs the process is k= kG,
where G is a reciprocal lattice vector.
3. All lattice waves can be described by wave vectors that lie within the first Brillouin zone in
the reciprocal space.
4. If there are p atoms in the primitive, the phonon dispersion relation will have 3 acoustical
phonon branches and 3p-3 optical phonon branches.
Self Assessment Questions for Study Session 6

SAQ 6.1
Using the dispersion relation for the monoatomic linear lattice of N atoms with nearest neighbour
interactions, show that the density of vibrational modes is given by
𝐷(𝜔) = , where ω is the maximum frequency.

m
SAQ 6.2
Consider a dielectric crystal made up of layers of atoms with rigid coupling between layers so that
the motion of atoms is restricted to the plane of the layer (i.e. 2D solid). Using the Debye
approximation, obtain the expression for the thermal energy and show that the phonon heat
capacity in the low temperature limit is proportional to T 2.
96
SAQ 6.3
In the Debye approximation, show that the mean square displacement of an atom at absolute zero
ħ
is 〈𝑅 〉 = , where 𝑣 is velocity of sound. Estimate this value for Cu
3
(𝜃 = ħ𝜔 ⁄𝑘 = 343 𝐾 , ρ = 8920 kg/m , 𝑣 = 3570 m/s).
References/ Further Readings

97
Study Session 7 Free-Electron Theory of Metals
Introduction
A free electron model is the simplest way to represent the electronic structure of metals. Although
the free electron model is a great oversimplification of the reality, surprisingly in many cases it
works pretty well, so that it is able to describe many important properties of metals.
According to this model, the valence electrons of the constituent atoms of the crystal become
conduction electrons and travel freely throughout the crystal. Therefore, within this model we
neglect the interaction of conduction electrons with ions of the lattice and the interaction between
the conduction electrons. In this sense we are talking about a free electron gas. However, there is
a principal difference between the free electron gas and ordinary gas of molecules. First, electrons
are charged particles. Therefore, in order to maintain the charge neutrality of the whole crystal,
we need to include positive ions. This is done within the jelly model, according to which the
positive charge of ions is smeared out uniformly throughout the crystal. This positive background
maintains the charge neutrality but does not exert any field on the electrons. Ions form a uniform
jelly in which electrons move.
The second important property of the free electron gas is that it should meet the Pauli exclusion
principle, which leads to important consequences.
Learning Outcome for Study Session 7

7.1 Distinguish between the classical and quantum approaches to the free electron theory.
7.2 Describe the treatment of electron dynamics as a free electron gas.
7.1 Main Contents
7.1.1 One dimension

We consider first a free electron gas in one dimension. We assume that an electron of mass m is
confined to a length L by infinite potential barriers. The wave function 𝛹 (𝑥) of the electron is a
98
solution of the Schrödinger equation 𝛹 (𝑥) = 𝐸 𝛹 (𝑥). , where En is the energy of electron
orbital. Since we can assume that the potential lies at zero, the Hamiltonian H includes only the
kinetic energy so that
ħ
𝐻𝛹 (𝑥) = 𝛹 (𝑥) = − 𝛹 (𝑥) = 𝐸 𝛹 (𝑥). (7.1)
Note that this is a one-electron equation, which means that we neglect the electron-electron
interactions. We use the term orbital to describe the solution of this equation.
Since the 𝛹 (𝑥) is a continuous function and is equal to zero beyond the length L, the boundary
conditions for the wave function are 𝛹 (0) = 𝛹 (𝐿) = 0. The solution of Eq.(7.1) is therefore
Ψ (𝑥) = 𝐴𝑠𝑖𝑛 𝑥
where A is a constant and n is an integer. Substituting (7.2) into (7.1) we obtain for the Eigen
values
ℏ 𝜋𝑛
𝐸 =
2𝑚 𝐿
These solutions correspond to standing waves with different number of nodes within the potential
well as is shown in Fig.7.1.
The energy levels are labeled according to the quantum number n which gives the number of half-
wavelengths in the wave function. The wavelengths are indicated on the wave functions.
Now we need to accommodate N valence electrons in these quantum states. According to the
Pauli Exclusion Principle, no two electrons can have their quantum number identical. That is, each
electronic quantum state can be occupied by at most one electron. The electronic state in a one-
dimensional solid is characterized by two quantum numbers n and ms, where n describes the
orbital Ψ (x), and ms describes the projection of the spin momentum on a quantization axis.
Electron spin is equal to s = 1/2, so that there (2s+1) = 2 possible spin states with ms = ±½.
Therefore, each orbital labelled by the quantum number n can accommodate two electrons, one
with spin up and the other with spin down orientation. Let nF denote the highest filled energy
99
level, where we start filling the levels from the bottom (n = 1) and continue filling higher levels
with electrons until all N electrons are accommodated.
Fig.7.1 First three energy levels and wave-functions of a free electron of mass m confined to a line of
length L
It is convenient to suppose that N is an even number. The condition 2nF = N determines nF, the
value of n for the uppermost filled level. The energy of the highest occupied level is called the
Fermi energy EF. For the one-dimensional system of N electrons we find, using Eq. (7.3),
ℏ
𝐸 = . (7.4)
In metals the value of the Fermi energy is of the order of 5 eV. The ground state of the N electron
system is illustrated in Fig.7.2a: All the electronic levels are filled up to the Fermi energy. All the
levels above are empty.
In-text question
1. Q: Highlight the assumptions made in the free electron gas theory and deduce the solution to
the Schrodinger’s equation.
A: See pages 96-99.
100
7.1.2. The Fermi distribution
This is defined as the ground state of the N electron system at absolute zero. What happens if the
temperature is increased? The kinetic energy of the electron gas increases with temperature.
Therefore, some energy levels become occupied which were vacant at zero temperature, and some
levels become vacant which were occupied at absolute zero. The distribution of electrons among
the levels is usually described by the distribution function, f(E), which is defined as the probability
that the level E is occupied by an electron. Thus if the level is certainly empty, then, f(E) = 0,
while if it is certainly full, then f(E) = 1. In general, f(E) has a value between zero and unity.
Fig. 7.2 (a) Occupation of energy levels according to the Pauli exclusion principle,
(b) The distribution function f(E), at T = 0°K and T> 0°K.
It follows from the preceding discussion that the distribution functions for electrons at T = 0°K
has the form
1, 𝐸 < 𝐸
𝑓(𝐸) = (7.5)
0, 𝐸 > 𝐸
That is, all levels below EF are completely filled, and all those above EF are completely empty.
This function is plotted in Fig. 7.2(b), which shows a discontinuity at the Fermi energy. When the
system is heated (T>0°K), thermal energy excites the electrons. However, all the electrons do not
share this energy equally, as would be the case in the classical treatment, because the electrons
lying well below the Fermi level EF cannot absorb energy. If they did, they would move to a
higher level, which would be already occupied, and hence the exclusion principle would be
violated.
Recall in this context that the energy which an electron may absorb thermally is of the order 𝐾 𝑇
(= 0.025 eV at room temperature), which is much smaller than 𝐸 , this being of the order of 5 eV.
101
Therefore, only those electrons close to the Fermi level can be excited, because the levels above
𝐸 are empty, and hence when those electrons move to a higher level there is no violation of the
exclusion principle. Thus, only these electrons, which are a small fraction of the total number, are
capable of being thermally excited.
The distribution function at non-zero temperature is given by the Fermi distribution function.
The Fermi distribution function determines the probability that an orbital of energy E is
occupied at thermal equilibrium
𝑓(𝐸) = ( )⁄ (7.6)
This function is also plotted in Fig. 7.2(b), which shows that it is substantially the same as the
distribution at T = 0°K, except very close to the Fermi level, where some of the electrons are
excited from below EF to above fermi level.
The quantity is called the chemical potential. The chemical potential can be determined in a way
that the total number of electrons in the system is equal to N. At absolute zero, 𝜇 = 𝐸 .
In-text question
1. Q: Explain briefly what happens to the physical properties of a metal when the temperature of
the metal is increased.
A: See pages 99-100.
2. Q: Write out the Fermi distribution function and explain what each of the term in the equation
represents.
102
7.1.3. Three dimensions
The Schrödinger equation in the three dimensions takes the form
If the electrons are confined to a cube of edge L, the solution is the standing wave
where nx, ny, and nz are positive integers.
In many cases, however, it is convenient to introduce periodic boundary conditions, as we did for
lattice vibrations. The advantage of this description is that we assume that our crystal is infinite
and disregard the influence of the outer boundaries of the crystal on the solution. We require then
that our wave function is periodic in x, y, and z directions with period L, so that
and similarly for the y and z coordinates. The solution of the Schrödinger equation (7.7) which
satisfies these boundary conditions has the form of the traveling plane wave:
provided that the component of the wave vector k are determined from
where nx, ny, and nz are positive or negative integers.
103
If we now substitute eq. (7.10) into Eq.(7.7), we will obtain for the energies of the orbital with the
wave vector k
The wavefunctions (7.10) are the eigenfunctions of the momentum 𝒑 = −𝑖ℏ∇, which can be
easily seen by differentiating (7.10) :
The eigenvalue of the momentum is ℏk . The velocity of the electron is defined by

v p /m ℏk /m.
In the ground state, a system of N electrons occupies states with lowest possible energies.
Therefore, all the occupied states lie inside the sphere of radius 𝐾 . The energy at the surface of
this sphere is the Fermi energy 𝐸 .The magnitude of the wave vector 𝐾 and the Fermi energy
are related by the following equation:
The Fermi energy and the Fermi wave vector (momentum) are determined by the number of
valence electrons in the system. In order to find the relationship between N and 𝐾 , we need to
count the total number of orbitals in a sphere of radius 𝐾 which should be equal to N. There are
two available spin states for a given set of kx, ky, and kz. The volume in the k space which is
occupied by this state is equal to (2𝑝 / 𝐿) . Thus, in the sphere of (4𝜋𝑘 ⁄3), the total number of
states is
104
where the factor 2 comes from the spin degeneracy. Then
which depends only on the electron concentration. We obtain, for the Fermi energy;
and the Fermi velocity
A few estimates for Na: Na has bcc structure with cubic lattice parameter a=4.2Å and one valence
electron per atom. Since there are 2 atoms in a unit cell, the electron concentration is N/V =
2/(4.2Å ) = 3.1022cm-3. Then, the Fermi momentum 𝐾 ≈ (3.10.3. 10 cm ) ⁄
≈ 10 𝑐𝑚 =
1Å .
ℏ ℏ
The Fermi energy is given by 𝐸 = = ≈ 13.6𝑒𝑉. 0.25 ≈ 3.5𝑒𝑉 . The Fermi
temperature is defined as 𝑇 = 𝐸 ⁄𝑘 ≈ 40000𝐾 (1eV corresponds to 1.16.104K). The Fermi

ℏ ℏ
velocity can be found from = 𝑘 𝑎 = 𝑘 𝑎 ≈ 0.5 ≈ 4. 10 , which
ℏ
results in 𝑉 ≈ 10 𝑐𝑚𝑠 . We see that the velocity of electrons is relatively large.
An important quantity which characterizes electronic properties of a solid is the density of states,
which is the number of electronic states per unit energy range. To find it we use Eq.(7.17) and
write the total number of orbitals of energy < E :
𝑁(𝐸) = (7.19)
ℏ
105
The density of states is then
3 1
𝑉 2𝑚 2
𝐷(𝐸) = =
2𝜋2 ℏ2
𝐸 2 (7.20)
or equivalently
𝐷(𝐸) = (7.21)
So within a factor of the order of unity, the number of states per unit energy interval at the Fermi
energy, 𝐷(𝐸 ) , is the total number of conduction electrons divided by the Fermi energy, just as
we would expect.
The density of states is normalized in such a way that the integral
𝑁=∫ 𝐷(𝐸)𝑑𝐸 (7.22)
gives the total number of electrons in the system. At non-zero temperature we should take into
account the Fermi distribution function so that
𝑁 = ∫ 𝐷(𝐸)𝑓(𝐸)𝑑𝐸, (7.23)
This expression also determines the chemical potential.
106
Figure 7.3 Density of single-particle states as a function of energy, for a free electron gas in
three dimensions
The dashed curve represents the density f(E,T)D(E) of filled orbitals at a finite temperature, but
such that kT is small in comparison with EF. The shaded area represents the filled orbitals at
absolute zero. The average energy is increased when the temperature is increased from 0 to T, for
electrons are thermally excited from region 1 to region 2.
In-text question
ℏ
1. Q: What is the degree of degeneracy of the energy level of the particle in a cubical
potential box of side ‘a’?
A: (6 1 1), (1 6 1), ( 1 1 6), (5 3 2), (5 2 3), (3 5 2), (3 2 5), (2 5 3), and (2 3 5) give the same
energy level; therefore, the level is nine-fold degenerate.
2. Q: Find the energies of the six lowest energy levels of a particle in cubical box. Which of the
levels are degenerate?
A: (1 1 1), (1 1 2), (1 2 2), (1 1 3), (2 2 2), (1 2 3)
The first and the fifth are non-degenerate; 2nd, 3rd and 4th are three-fold degenerate and the six-fold
degenerate. Their energy values are:
107
, , , , ,

1. The consideration of free electron in 1D and solutions to its Eigen functions and energies were
obtained.
2. About the changes observed in the electron gas when the temperature of the metal is increased
and its Fermi distribution.
3. treating the free electron in 3D and its eigenfunctions and eigenvalues were obtained using
Schrödinger equation.
Self Assessment Question for Study Session 7
SAQ 7.1
Consider the free electron energy bands of an fcc crystal lattice in the reduced zone scheme in
which all k's are transformed to lie in the first Brillouin zone. Plot roughly in the [111] direction
the energies of all bands up to six times the lowest band energy at the zone boundary at k =
(2π/a)(½,½,½). Explain what happens with these bands in the presence of a weak crystal potential.
SAQ 7.2
Suppose the crystal potential in a one-dimensional lattice of lattice constant a is composed of a

series of rectangular wells which surround the atom. Suppose the depth of each well is U0 and its
width a/5.
a. Calculate the values of the first three energy gaps. Compare the magnitudes of these gaps.
108
b. Evaluate these gaps for the case of U = 5 eV and a = 4Å.
0
SAQ 7.3
Using the solution for the energy bands near the zone boundary in the presence of a weak crystal
potential, show that the electron velocity is parallel to the Bragg plane. Since the gradient is
perpendicular to the surfaces on which a function is constant, this fact allows us to conclude that
the constant energy surfaces (such as the Fermi surface) at the Bragg plane are perpendicular to
that plane.
109
Study Session 8 Metals: Electron Dynamics and Fermi Surfaces
Introduction
The next important subject we address is electron dynamics in metals. Our consideration will be
based on a semi-classical model. The term “semi-classical” comes from the fact that within this
model the electronic structure is described quantum-mechanically but electron dynamics itself is
considered in a classical way, i.e., using classical equations of motion. Within the semi-classical
model we assume that we know the electronic structure of metal, which determines the energy
band as a function of the wave vector. The aim of the model is to relate the band structure to the
transport properties as a response to the applied electric field.

8.1 Determine the energy band as a function of wave vector using the semiclassical model.
8.2 Relate the band structure to the transport properties due to an applied electric field.
8.1 Main Content

8.1.1 Electron Dynamics
Given the functions 𝐸 (𝑘), the semi-classical model associates with each electron, a position, a
wave vector and a band index n. In the presence of applied fields, the position, the wave vector,
and the index are taken to evolve according to the following rules:
(1) The band index is a constant of the motion. The semiclassical model ignores the possibility of
interband transitions. This implies that within this model, it is assumed that the applied electric
field is small.
(2) The time evolution of the position and the wave vector of an electron with band index n are
determined by the equations of motion:
( )
𝑣 (𝑘) = (8.1)

 = 𝐹(𝑟, 𝑡) = −𝑒𝐸(𝑟, 𝑡) (8.2)
110
Clearly, Eq.(8.2) has to be proved. It is identical to the Newton’s second law if we assume that the
electron momentum is equal to k. The fact that electrons belong to particular bands makes their
movement in the applied electric field different from that of free electrons.
For example, if the applied electric field is independent of time, according to Eq. (8.2) the wave
vector of the electron increases uniformly with time.
𝑘(𝑡) = 𝑘(0) − (8.3)

Since velocity and energy are periodic in the reciprocal lattice, the will exhibit oscillation. This is
in striking contrast to the free electron case, where v is proportional to k and grows linearly with
time.
The k dependence (and, to within a scale factor, the t dependence) of the velocity is illustrated in
Fig.8.1, where both E(k) and v(k) are plotted in one dimension. Although the velocity is linear in k
near the band minimum, it reaches a maximum as the zone boundary is approached, and then
drops back down, going to zero at the zone edge. In the region between the maximum of v and the
zone edge, the velocity actually decreases with increasing k, so that the acceleration of the
electron is opposite the externally applied electric force.
This extraordinary behavior is a consequence of the additional force exerted by the periodic
potential, which is included in the functional form of E(k). As an electron approaches a Bragg
plane, the external electric field moves it in the opposite direction due to the Bragg-reflection.
Fig.8.1. E(k) and v(k) vs. k in one dimension (or three dimensions, in a direction parallel to a
reciprocal lattice vector that determines one of the first-zone faces)
111
Effective Mass
When discussing electron dynamics in solids, it is often convenient to introduce the concept of
effective mass. If we differentiate Eq. (8.1) with respect to time, we find that
= = , (8.4)
 
where the second derivative with respect to a vector should be understood as a tensor. Using Eq.
(8.2) we find that
= 𝑭, (8.5)

In the one dimensional case, this reduces to
= 𝑭. (8.6)

This has the same form as the Newton’s second law, provided that we define an effective mass by
the relation:
= . (8.7)
∗ 
The mass 𝑚∗ is inversely proportional to the curvature of the band; where the curvature is large -
that is, 𝑑 𝐸 𝑑𝑘 is large - the mass is small; a small curvature implies a large mass (Fig. 8.2).
Fig. 8.2 The inverse relationship between the mass and the curvature of the energy band
In a general case, the effective mass is a tensor which is defined by
= , (8.8)
∗ 
112
Where 𝑘 and 𝑘 are Cartesian coordinates.
The effective mass can be different depending on the directions in the crystal.
Current Density
The current density within a free electron model is defined as 𝒋 = −𝑒𝑛𝒗, where n is the number of
valence electrons per unit volume, and v is the velocity of electrons. This expression can be
generalized to the case of Bloch electrons. In this case, the velocity depends on the wave vector
and we need to sum up over k vectors for which there are occupied states available:
j= ∑ , 𝑉(𝑘) (8.9)
Here the sum is performed within the extended zone scheme and V is the volume of the solid. It is
often convenient to replace the summation by an integral. The volume of k-space per allowed k
value is ∆𝑘 = 8𝜋 𝑉 , we can therefore, write the sum over k as
∑ = ∫ 𝑑𝑘. (8.10)
Taking into account the spin degeneracy, we obtain for the current density:
j= −𝑒 ∫ 𝑉(𝑘). (8.11)
Using this expression we show now that completely filled bands do not contribute to the current.
For the filled bands Eq. (8.11) should be replaced by
( )
𝒋 = −𝑒 ∫ 𝑉(𝑘) , (8.12)
This vanishes as a consequence of the theorem that the integral over any primitive cell of the
gradient of a periodic function must vanish.
Proof:
Let f(r) be any function with the periodicity of the lattice. The integral over the primitive cell
𝑰(𝒓 ) = ∫ 𝑑𝑟𝑓(𝑟 + 𝑟 ) , (8.13)
is independent of 𝒓 . Therefore,
113
( ) ( ) ( )
0= =∫ 𝑑𝑟 =∫ 𝑑𝑟 . (8.14)
At point 𝑟 = 0 this results in
( )
∫ 𝑑𝑟 =0. (8.15)
Since the Brillouin zone is a primitive cell in the reciprocal space, the integral (8.12) vanishes.
This implies that filled bands do not contribute to the current. Only partially filled bands need be
considered in calculating the electronic properties of a solid. This explains why the Drude’s theory
assumption is often successful: in many cases, those bands derived from the atomic valence
electrons are the only ones that are partially filled.
In-text question
1. Q: State the rules that the position wave vector and index evolve through in the presence of an
applied field.
2. Q: Show that completely filled bands do not contribute to the current in the presence of an
applied field.
8.1.2 Hole
One of the most impressive achievements of the semiclassical model is its explanation for the
phenomena on that free electron theory can account for only if the carriers have a positive charge.
We now introduce the concept of a hole.
The contribution of all the electrons in a given band to the current density is given by Eq. (8.11),
where the integral is over all occupied levels in the band. By exploiting the fact that a completely
filled band carries no current,
0=∫ 𝑉(𝑘) = ∫ 𝑉(𝑘) + ∫ 𝑉(𝑘) . (8.16)
we can equally write Eq. (8.11), in the form:
j= +𝑒 ∫ 𝑉(𝑘) . (8.17)
114
Thus the current produced by electrons occupying a specified set of levels in a band is precisely
the same as the current that would be produced if the specified levels were unoccupied and all
other levels in the band were occupied but with particles of charge +e (opposite to the electronic
charge).
Therefore, even though the only charge carriers are electrons, we may, whenever it is convenient,
consider the current to be carried entirely by fictitious particles of positive charge that fill all those
levels in the band that are unoccupied by electrons. The fictitious particles are called holes. It
must be emphasized that particles cannot be mixed within a given band. If one wishes to regard
electrons as carrying the current, then the unoccupied levels make no contribution; if one wishes
to regard the holes as carrying the current, then the electrons make no contribution. One may,
however, regard some bands using the electron picture and other bands using the hole picture, at
one's convenience.
Normally, it is convenient to consider transport of the holes for the bands which are almost
occupied, so that only a few electrons are missing. This happens in semiconductors in which a few
electrons are excited from the valence to the conduction bands. Similar to electrons, we can
introduce the effective mass for the holes. It has a negative sign.
Fermi surface
The ground state of N Bloch electrons is constructed in a similar fashion as that for free electrons,
i.e., by occupying all one-electron energy levels with band energies 𝐸 𝐾 less than 𝐸 ,where 𝐸 is
determined by requiring the total number of levels with energies less than 𝐸 to be equal to the
total number of electrons. The wave vector k must be confined to a single primitive cell of the
reciprocal lattice. When the lowest of these levels are filled by a specified number of electrons,
two quite distinct types of configuration can result:
1. A certain number of bands may be completely filled, all others remaining empty. Because the
number of levels in a band is equal to the number of primitive cells in the crystal (and because
each level can accommodate two electrons (one of each spin), a configuration with a band gap can
arise only if the number of electrons per primitive cell is even.
115
2. A number of bands may be partially filled. When this occurs, the energy of the highest
occupied level, the Fermi energy𝐸 , lies within the energy range of one or more bands. For each
partially filled band, there will be a surface in k-space separating the occupied from the
unoccupied levels. The set of all such surfaces is known as the Fermi surface, and is the
generalization to Bloch electrons of the free electron Fermi sphere. The parts of the Fermi surface
arising from individual partially filled bands are known as branches of the Fermi surface.
Analytically, the branch of the Fermi surface in the n-th band is that surface in k-space determined
by
𝐸 𝐾=𝐸 (8.18)
Thus, the Fermi surface is a constant energy surface (surfaces) in k-space.

Since the 𝐸 𝐾 are periodic in the reciprocal lattice, the complete solution to Eq.(8.18) for each n
is a k-space surface with the periodicity of the reciprocal lattice. When a branch of the Fermi
surface is represented by the full periodic structure, it is said to be described in a repeated zone
scheme. Often, however, it is preferable to take just enough of each branch of the Fermi surface so
that every physically distinct level is represented by just one point of the surface. This is achieved
by representing each branch by that portion of the full periodic surface contained within a single
primitive cell of the reciprocal lattice. Such a representation is described as a reduced zone
scheme. The primitive cell chosen is often, but not always, the first Brillouin zone.
Brillouin zones
We consider now an example of building of a Fermi surface. We start from considering the Fermi
surface for free electrons and then investigating the influence of the crystal potential. The Fermi
surface for free electrons is a sphere centered at k = 0. To construct the Fermi surface in the
reduced-zone scheme, one can translate all the pieces of the sphere into the first zone through
reciprocal lattice vectors. This procedure is made systematic through the geometrical notion of the
higher Brillouin zones.
116
Figure 8.3 (a) Construction in k space of the first three Brillouin zones of a square lattice.
The three shortest forms of the reciprocal lattice vectors are indicated as G1, G2, and G3.
The lines drawn are perpendicular bisectors of these G's.
(b) On constructing all lines equivalent by symmetry to the three lines in (a) we obtain the
regions in k space which form the first three Brillouin zones. The numbers denote the zone
to which the regions belong; the numbers here are ordered according to the length of vector
G involved in the construction of the outer boundary of the region.
We illustrate this construction for the two dimensional cubic lattice shown in Fig.8.3. Recall that
the boundaries of the Brillouin zones are planes normal to G at the midpoint of G. The first
Brillouin zone of the square lattice is the area enclosed by the perpendicular bisectors of G1 and
of the three reciprocal lattice vectors equivalent by symmetry to G1 in Fig. 8.3a. These four
reciprocal lattice vectors are ± 2𝜋 𝑎 𝑘 and ± 2𝜋 𝑎 𝑘 .
The second zone is constructed from G2 and the three vectors equivalent to it by symmetry, and
similarly for the third zone. The pieces of the second and third zones are drawn in Fig. 8.3b.
In general, the first Brillouin zone is the set of points in k-space that can be reached from the
origin without crossing any Bragg plane. The second Brillouin zone is the set of points that can be
reached from the first zone by crossing only one Bragg plane. The (n+1)-th Brillouin zone is the
set of points not in the (n-l)-th zone that can be reached from the n-th zone by crossing only one
Bragg plane.
117
Figure 8.4 Brillouin zones of a square lattice in two dimensions.
The circle shown is a surface of constant energy for free electrons; it will be the Fermi surface for
some particular value of the electron concentration. The total area of the filled region in K space
depends only on the electron concentration and is independent of the interaction of the electrons
with the lattice. The shape of the Fermi surface depends on the lattice interaction and the shape
will not be an exact circle in an actual lattice.
The free electron Fermi surface for an arbitrary electron concentration is shown in Fig. 8.4. Now
we perform a transformation to the reduced zone scheme as is shown in Figs.8.5 and 8.6. We take
the triangle labeled 2a and move it by a reciprocal lattice vector 𝐺 = − 2𝜋 𝑎 𝑘 such that the
triangle reappears in the area of the first Brillouin zone (Figure 8.5). Other reciprocal lattice
vectors will shift the triangles 2b, 2c, 2d to other parts of the first zone, completing the mapping of
the second zone into the reduced zone scheme. The parts of the Fermi surface falling in the second
zone are now connected, as shown in Fig. 8.6.
118
Figure 8.5 Mapping of the first, second, and third Brillouin zones in the reduced zone
scheme.
The sections of the second zone in Fig. 8.5 are put together into a square by translation through an
appropriate reciprocal lattice vector.
A different G is needed for each piece of a zone.
Fig.8.6 The free electron Fermi surfaces of Fig. 8.4, as viewed in the reduced zone scheme.
The shaded areas represent occupied electron states. Parts of the Fermi surface fall into the
second, third and fourth zones. The fourth zone is not shown. The first zone is entirely
occupied.The construction of Brillouin zones and Fermi surfaces in three-dimensions is more
complicated. Fig.8.7 shows the first three Brillouin zones for bcc and fcc structures.
119
Figure 8.7 Surfaces of the first, second and third Brillouin zones for (a) body-centered cubic
and (b) face-centered cubic crystals. (Only the exterior surfaces are shown). The interior
surface of the nth zone is identical to the exterior surface of the (n-1) th zone). Evidently the
surfaces bounding the zones become increasingly complex as the zone number increases.
The free electron Fermi surfaces for fcc cubic metals of valence 2 and 3 are shown in Fig.8.8.
Fig.8.8 The free electron Fermi surfaces for face-centered cubic metals of valence 2 and 3.
For valence 1 the surface lies entirely within the interior of the first zone and therefore remains a
sphere to lowest order. All branches of the Fermi surface are shown. The primitive cells in which
they are displayed have the shape and orientation of the first Brillouin zone.
120
Effect of a crystal potential
How do we go from Fermi surfaces for free electrons to Fermi surfaces in the presence of a weak
crystal potential? We can make approximate constructions freehand by the use of four facts:
(i) The interaction of the electron with the periodic potential of the crystal causes energy gaps at
the zone boundaries.
(ii) Almost always, the Fermi surface will intersect zone boundaries perpendicularly. Using the

equation for the energy near the zone boundary, it is easy to show that = (𝑘 − 𝐺), which
implies that on the Bragg plane the gradient of energy is parallel to the Bragg plane. Since the
gradient is perpendicular to the surfaces on which function 𝐸 (𝑘) is constant, the constant energy
surfaces at the Bragg plane are perpendicular to the plane.
(iii) The crystal potential will round out sharp corners in the Fermi surfaces.
(iv) The total volume enclosed by the Fermi surface depends only on the electron concentration
and is independent of the details of the lattice interaction.
If a branch of the Fermi surface consists of very small pieces of surface (surrounding either
occupied or unoccupied levels, known as "pockets of electrons" or "pockets of holes"), then a
weak periodic potential may cause these to disappear. In addition, if the free electron Fermi
surface has parts with a very narrow cross section, a weak periodic potential may cause it to
become disconnected at such points.
In-text question:
1. Q: State the assumption made as regards the holes in conduction through metals.
2. Q: State the distinct type of configuration that can result when the lowest of the energy levels
are filled by a specified number of electrons on a Fermi surface.
A: See page 113-115
3. Q: How do you construct a Fermi surface in the reduced zone scheme?
A: See page 116-118.
Below are few examples for real metals.
121
8.1.3. Alkali metals
The radius of the Fermi sphere in bcc alkali metals is less than the shortest distance from the
center of the zone to a zone face and therefore the Fermi sphere lies entirely within the first
Brillouin zone. The crystal potential does not distort much the free electron Fermi surface and it
remains very similar to a sphere.
Figure 8.9 Fermi Surface of sodium
8.1.4. The noble metals

The Fermi surface for a single half-filled free electron band in an fcc Bravais lattice is a sphere
entirely contained within the first Brillouin zone, approaching the surface of the zone most closely
in the [111] directions, where it reaches 0.903 of the distance from the origin to the center of the
hexagonal face. For all three noble metals therefore, Fermi surfaces are closely related to the free
electron sphere. However, in the [111] direction contact is actually made with the zone faces, and
the measured Fermi surfaces have the shape shown in Fig.8.10. Eight "necks" reach out to touch
the eight hexagonal faces of the zone, but otherwise the surface is not grossly distorted from
spherical.
Figure 8.10 In the three noble metals the free electron sphere bulges out in the [111]
directions to make contact with the hexagonal zone faces
122
The Fermi surface of aluminium is close to that of the free electron surface for fcc cubic
monoatomic lattice with three conduction electrons per atom. The first Brillouin zone is filled and
the Fermi surface of free electrons is entirely contained in the second, third and fourth Brillouin
zones. When displayed in a reduced-zone scheme the second-zone surface is a closed structure
containing unoccupied levels, while the third-zone surface is a complex structure of narrow tubes
(Fig.8.8). The amount of surface in the fourth zone is very small, enclosing tiny pockets of
occupied levels.
The effect of a weak periodic potential is to eliminate the fourth-zone pockets of electrons, and
reduce the third-zone surface to a set of disconnected "rings" (Fig.8.11).
Aluminum provides a striking illustration of the theory of Hall coefficients. The high-field Hall
coefficient should be 𝑅 = −1⁄(𝑛 − 𝑛 ) 𝑒 , where 𝑛 and 𝑛 are the number of levels per unit
volume enclosed by the electron-like and hole-like branches of the Fermi surface. Since the first
zone of aluminum is completely filled and accommodates two electrons per atom, one of the three
valence electrons per atom remains to occupy second- and third-zone levels. Thus,
𝑛 +𝑛 = (8.19)
where n is the free electron carrier density appropriate to valence 3. On the other hand, since the
total number of levels in any zone is enough to hold two electrons per atom, we also have
𝑛 +𝑛 =2 (8.20)
Subtracting (8.20) from (8.19) gives

𝑛 −𝑛 =− (8.21)
Thus, the high-field Hall coefficient should have a positive sign and yield an effective density of
carriers a third of the free electron value. This is precisely what is observed.
Figure 8.11 Fermi surface of Aluminum
123
In-text question
1. Q: Briefly explain the Fermi sphere for Alkali metals.

A: See page 120

In this study session 8, you have learnt:
1. that conduction takes place through holes aside from electrons.
2. the formation and effect of brillouin zones.
3. the effect of crystal potentials on the Fermi surface.
Self-Assessment Questions for Study Session 8

SAQ 8.1
What are Brillouin zones? How are they related to the energy of an electron in a metal?
SAQ 8.2
Explain the significance of Brillouin zones with particular reference to any cubic lattice.
SAQ 8.3
Illustrate the three Brillouin zones for a two dimensional square lattice.

124
Study Session 9 Energy Bands
Introduction
The free electron model gives us a good insight into many properties of metals, such as the heat
capacity, thermal conductivity and electrical conductivity. However, this model fails to help us
with other important properties. For example, it does not predict the difference between metals,
semiconductors and insulators. It does not explain the occurrence of positive values of the Hall
coefficient. Also, the relation between conduction electrons in the metal and the number of
valence electrons in free atoms is not always correct. We need a more accurate theory, which
would be able to answer these questions.

9.1 Differentiate metals, semiconductors and insulators from one another.
9.2 Distinguish between the free electron model and the many electron problem.
9.3 Calculate current density and life-time of holes and electrons.
9.1 Main Content
9.1.1 Electron in Periodic Field of Crystal

The problem of electrons in a solid is in general a many-electron problem. The full Hamiltonian of
the solid contains not only the one-electron potentials describing the interactions of the electrons
with atomic nuclei, but also pair potentials describing the electron-electron interactions. The
many-electron problem is impossible to solve exactly and therefore we need simplified
assumptions. The simplest approach we have already considered is a free electron model. The
next step in building the complexity is to consider an independent electron approximation,
assuming that all the interactions are described by an effective potential. One of the most
important properties of this potential is that it is periodic on a lattice
𝑈(𝑟) = 𝑈(𝑟 + 𝑻) , (9.1)
125
where 𝑻 is a lattice vector. Qualitatively, a typical crystalline potential might be expected to have
a form shown in Fig.9.1, resembling the individual atomic potentials as the ion is approached
closely and flattening off in the region between ions.
Fig. 9.1 The crystal potential seen by the electron.
Within the approximation of non-interacting electrons the electronic properties of a solid can be
examined with the Schrödingeħr equation,
ħ
− ∇ + 𝑈(𝑟) (𝑟) = 𝐸(𝑟), (9.2)
in which(𝑟) is a wave function for one electron. Independent electrons, which obey a one-
electron Schrödinger equation (9.2) with a periodic potential, are known as Bloch electrons, in
contrast to "free electrons," to which Bloch electrons reduce when the periodic potential is
identically zero.
Now we discuss general properties of the solution of the Schrödinger equation (9.2) taking into
account the periodicity of the effective potential (9.1) and discuss main properties of Bloch
electrons, which follow from this solution.
We represent the solution as an expansion over plane waves:
(𝑟) = ∑ 𝑐 𝑒 . (9.3)
This expansion in a Fourier series is a natural generalization of the free-electron solution for a
zero potential. The summation in (9.3) is performed over all k vectors, which are permitted by the
periodic boundary conditions. According to these conditions the wave function (9.3) should
satisfy
126
(𝑥, 𝑦, 𝑧) = (𝑥 + 𝐿, 𝑦, 𝑧) = (𝑥, 𝑦 + 𝐿, 𝑧) = (𝑥, 𝑦, 𝑧 + 𝐿) , (9.4)
so that
𝑘 = ; 𝑘 = ; 𝑘 = , (9.5)
where 𝑛 , 𝑛 , and 𝑛 are positive or negative integers. Note that in general, (𝑟) is not periodic
in the lattice translation vectors. On the other hand, according to Eq.(9.1), the potential energy is
periodic, i.e., it is invariant under a crystal lattice translation. Therefore, its plane wave expansion
will only contain plane waves with the periodicity of the lattice. Therefore, only reciprocal lattice
vectors are left in the Fourier expansion for the potential:
𝑈(𝑟) = ∑ 𝑈 𝑒 , (9.6)
where the Fourier coefficients 𝑈 are related to 𝑈(𝑟) by
𝑈 = ∫ 𝑒 𝑈(𝑟)𝑑𝑟 , (9.7)
where 𝑉 is the volume of the unit cell. It is easy to see that, indeed, the potential energy
represented by (9.6) is periodic in the lattice:
( )
𝑈(𝑟 + 𝑇) = ∑ 𝑈 𝑒 =𝑒 ∑ 𝑈 𝑒 = 𝑈(𝑟) , (9.8)
where the last equation comes from the definition of the reciprocal lattice vectors 𝑒 = 1. The
values of Fourier components 𝑈 for actual crystal potentials tend to decrease rapidly with
increasing magnitude of G. For example, for a Coulomb potential, 𝑈 decreases as 1 𝐺 . Note
that since the potential energy is real, the Fourier components should satisfy 𝑈 = 𝑈∗ .
We now substitute (9.3) and (9.6) in Eq.(9.2) and obtain:
ħ ( )
∑ 𝑘 𝑐 𝑒 +∑ ∑ 𝑈 𝑐 𝑒 =𝐸∑ 𝑐 𝑒 . (9.9)
127
Changing the summation index in the second sum on the left from k to k+G, this equation can be
rewritten in form:
ħ
∑ 𝑒 𝑘 − 𝐸 𝑐 + ∑ 𝑈 𝑐( ) =0. (9.10)
Since this equation must be satisfied for any r the Fourier coefficients in each separate term of
(9.10) must vanish and therefore

𝑘 − 𝐸 𝑐 + ∑ 𝑈 𝑐( ) =0. (9.11)
This is a set of linear equations for the coefficients 𝑐 . These equations are nothing but a
restatement of the original Schrödinger equation in the momentum space, simplified by the fact
that the potential is periodic. This set of equations does not look very pleasant because, in
principle, an infinite number of coefficients should be determined. However, a careful
examination of Eq.(9.11) leads to important consequences.
First, we see that for a fixed value of k the set of equations (9.11) couples only those coefficients,
whose wave vectors differ from k by a reciprocal lattice vector. In the one-dimensional case these
are k, 𝑘 ± 2𝜋 𝑎, 𝑘 ± 4𝜋 𝑎, and so on. We can therefore assume that the k vector belongs to the
first Brillouin zone. The original problem is decoupled to N independent problems (N is the total
number of atoms in a lattice) for each allowed value of k in the first Brillouin zone. Each such
problem has solutions that are a superposition of plane waves containing only the wave vector k
and wave vectors differing from k by the reciprocal lattice vector.
Putting this information back into the expansion (9.3) of the wave function (𝑟), we see that the
wave function will be of the form
 (𝑟) = ∑ 𝑐 𝑒 ( )
, (9.12)
128
where the summation is performed over the reciprocal lattice vectors and we introduce index k for
the wave function. We can rearrange this so that
 (𝑟) = 𝑒 ∑ 𝑐 𝑒 , (9.13)
or
 (𝑟) = 𝑒 𝑈 (𝑟) , (9.14)
where 𝑈 (𝑟) = 𝑈 (𝑟 + 𝑇) is a periodic function which is defined by
𝑈 (𝑟) = ∑ 𝑐 𝑒 . (9.15)
Equation (9.14) is known as Bloch theorem, which plays an important role in electronic band
structure theory. Now we discuss a number of important conclusions which follow from the Bloch
theorem.
1. Bloch's theorem introduces a wave vector k, which plays the same fundamental role in the
general problem of motion in a periodic potential that the free electron wave vector k plays in the
free-electron theory. Note, however, that although the free electron wave vector is simply
𝒑
, where p is the momentum of the electron, in the Bloch case k is not proportional to the
ħ
electronic momentum. This is clear on general grounds, since the Hamiltonian does not have
complete translational invariance in the presence of a non-constant potential, and therefore its
eigenstates will not be simultaneous eigenstates of the momentum operator. This conclusion is
confirmed by the fact that the momentum operator, 𝒑 = −𝑖ħ∇, when acting on  (𝑟) gives
−𝑖ħ∇ (𝑟) = −𝑖ħ∇𝑒 𝑈 (𝑟)] = ħ𝑘  (𝑟) − 𝑖ħe ∇𝑈 (𝑟) , (9.16)

which is not, in general, just a constant times  (𝑟); i.e  (𝑟) is not a momentum eigenstate.
Nevertheless, in many ways ħ𝑘 is a natural extension of p to the case of a periodic potential. It is

known as the crystal momentum or quasi momentum of the electron, to emphasize this similarity,
but one should not be misled by the name into thinking that ħ𝑘 is a momentum.
129
2. The wave vector k appearing in Bloch's theorem can always be confined to the first Brillouin
zone (or to any other convenient primitive cell of the reciprocal lattice). This is because any k' not
in the first Brillouin zone can be written as
𝑘 = 𝑘 + 𝐺, (9.17)
where G is a reciprocal lattice vector and k does lie in the first zone. Since 𝑒 = 1 for any
reciprocal lattice vector, if the Bloch form (9.14) holds for k', it will also hold for k. An example
is given below for a nearly free electron model.
The energy E of free electrons which is plotted versus k in Fig.9.2a exhibits a curve in the familiar
parabolic shape. Figure 9.2b shows the result of translations. Segments of the parabola of Fig.9.2a
are cut at the edges of the various zones, and are translated by multiples of 𝐺 = 2𝜋 𝑎 in order to
ensure that the energy is the same at any two equivalent points. Fig.9.2c displays the shape of the
energy spectrum when we confine our consideration to the first Brillouin zone only.
The type of representation used in Fig.9.2c is referred to as the reduced-zone scheme. Because it
specifies all the needed information, it is the one we shall find most convenient. The
representation of Fig.9.1, known as the extended-zone scheme, is convenient when we wish to
emphasize the close connection between a crystalline and a free electron. Fig.9.2b employs the
periodic-zone scheme, and is sometimes useful in topological considerations involving the k-
space. All these representations are strictly equivalent; the use of any particular one is dictated by
convenience, and not by any intrinsic advantages it has over the others.
130
Fig.9.2 Free electron bands within reduced- (a), extended- (b) and periodic-zone (c) scheme.
3. An important consequence of the Bloch theorem is the appearance of the energy bands.
All solutions to the Schrodinger equation (9.2) have the Bloch form  (𝒓) = 𝑒 𝑈 (𝒓) where k
is fixed and 𝑈 (𝒓) has the periodicity of the Bravais lattice. Substituting this into the Schrodinger
equation, we find that 𝑈 (𝒓) is determined by the eigenvalue problem
ħ
𝐻(𝒌)𝑈 (𝒓) = − (𝑖 + 𝒌∇) + 𝑈(𝒓) 𝑈 (𝒓) = 𝐸(𝒌) 𝑈 (𝒓) , (9.18)
with boundary condition
𝑈 (𝒓) = 𝑈 (𝒓 + 𝑻) , (9.19)
131
Due to the periodic boundary condition we can regard (9.18) as an eigenvalue problem restricted
to a single primitive cell of the crystal. The eigenvalue problem is set in a fixed finite volume. We
expect on general grounds to find an infinite family of solutions with discretely spaced
eigenvalues, which we label with the band index n. The Bloch function can therefore be denoted
by  (𝒓) which indicates that each value of the band index n and the vector k specifies an
electron state, or orbital with energy 𝐸 (𝒌).
Note that in terms of the eigenvalue problem specified by (9.18) and (9.19), the wave vector k
appears only as a parameter in the Hamiltonian H(k). We therefore expect each of the energy
levels, for given k, to vary continuously as k varies. In this way we arrive at a description of the
levels of an electron in a periodic potential in terms of a family of continuous functions 𝐸 (𝒌).
For each n, the set of electronic levels specified by 𝐸 (𝒌) is called an energy band. The
information contained in these functions for different n and k is referred to as the band structure
of the solid.
1. Number of states in a band.

The number of orbitals in a band within the first Brillouin zone is equal to the number of unit cells
N in the crystal. This is much the same as the statement made in connection with the number of
lattice vibrational modes, and is proved in a like manner, by appealing to the boundary conditions.
Consider first the one-dimensional case. The allowed values of k form a uniform mesh whose unit
spacing is 2𝜋 𝐿. The number of states inside the first zone, whose length is 2𝜋 𝑎, is therefore
equal to (2𝜋 𝑎)/( 2𝜋 𝐿) = 𝐿 𝑎 = 𝑁, where N is the number of unit cells, in agreement with the
assertion made earlier.
A similar argument may be used to establish the validity of the statement in two- and three-
dimensional lattices. It has been shown that each band has N states inside the first zone. Since
each such state can accommodate at most two electrons of opposite spins in accordance with Pauli
Exclusion principle, it follows that the maximum number of electrons that may occupy a single
band is 2N. This result is significant, as it will be used in a later section to establish the criterion
for predicting whether a solid is going to behave as a metal or as an insulator.
132
5. Now we show that an electron in a level specified by band index n and wave vector k has a
non-vanishing mean velocity, given by
𝒏 (𝒌)
𝑉 (𝒌) = . (9.20)
 𝒌
To show this we calculate the expectation value of the derivative of the Hamiltonian H(k) in
Eq.(9.18) with respect to k:
(𝒌) ħ ħ
𝑈 𝑈 = 𝑈 −𝑖 (𝑖𝒌 + ∇) 𝑈 =  ħ − ∇  . (9.21)
𝒌
Since 𝒗 = (−𝑖ħ 𝑚)∇ is the velocity operator, this establishes (9.20).
This is a remarkable fact. It asserts that there are stationary levels for an electron in a periodic
potential in which, in spite of the interaction of the electron with the fixed lattice of ions, it moves
forever without any degradation of its mean velocity. This is in striking contrast to the idea of
Drude, that collisions were simply encounters between the electron and a static ion.
In-text questions
1. Q: Describe how the electronic properties of non-interacting electrons can be examined.
2. Q: What are the main conclusions resulting from the Bloch theorem.
A: See page 129
3. Q: Differentiate the reduced, extended and periodic zone scheme in band structure formation.
A: See page 128
9.1.2 Weak Potential

When the potential is zero the solutions of the Schrödinger equation (9.11) are plane waves:
133
ħ 𝒌
𝐸 (𝒌) = , (9.22)
𝒌 = 𝑒 , (9.23)
where the wave function is normalized to the volume of unit cell Vc. In the reduced-zone
representation shown in Fig.9.3, for each k there is an infinite number of solutions which
correspond to different G (and can be labeled by index n), as we have already discussed. Each
band in Fig.9.3 corresponds to a different value of G in the extended scheme.
Fig.9.3 Only those states which have the same k in the First Brillouin zone are coupled by
perturbation
Suppose now that a weak potential is switched on. According to the Schrödinger equation (9.11)
only those states, which differ by G, are coupled by a perturbation. In the reduced zone scheme
those states have same k and different n (see Fig.9.3). As you know from quantum mechanics, if
the perturbation is small compared to the energy difference between the states, which are coupled
by the perturbation, we can use the perturbation theory to calculate wave functions and energy
levels. Assuming for simplicity that we are looking for the correction to the energy of the lowest
band 𝐸 (𝒌), the condition for using the perturbation theory is
|𝐸 (𝒌) − 𝐸 (𝒌 − 𝑮)| ≫ 𝑈 , (9.24)

for any 𝑮 ≠ 0. According to the perturbation theory the energy is given by
134
𝒌 𝑈 𝒌
𝐸(𝒌) = 𝐸 (𝒌) + 𝒌 𝑈 𝒌 + ∑𝑮 (𝒌)
𝑮
(𝒌 𝑮)
, (9.25)
The first term in Eq.(9.25) is the undisturbed free-electron value for the energy. The second term
is the mean value of the potential in the state 𝒌 (𝒓):
𝒌 𝑈 𝒌 = ∫ 𝑈(𝒓)𝑑𝑟 = 𝑈 . (9.26)
This term gives us a constant independent of k. Its effect on the spectrum is a rigid shift by a
constant value without causing any change in the shape of the energy spectrum. This term can be
set equal to zero. The third term can be rewritten as
𝒌 𝑈 𝒌 𝑮
= ∫ 𝑒 𝑈(𝒓)𝑒 (𝒌 𝑮)
𝑑𝑟 = ∫ 𝑒 𝑑𝑟 = 𝑈 . (9.27)
Finally, we obtain for the energy:
| |
𝐸(𝒌) = 𝐸 (𝒌) + ∑𝑮 (𝒌) (𝒌 𝑮)
. (9.28)
The perturbation theory breaks down, however, in those cases when the potential cannot be
considered as a small perturbation. This happens when the magnitude of the potential becomes
comparable with the energy separation between the bands, i.e.,
|𝐸 (𝒌) − 𝐸 (𝒌 − 𝑮)| ≤ 𝑈 , (9.29)
In this case we have to include these levels in the Schrödinger equation and solve explicitly.
There are special k points for which the energy levels become degenerate and the relationship
(9.29) holds for any non-zero value of the potential. For these k points
𝐸 (𝒌) = 𝐸 (𝒌 − 𝑮) , (9.30)
and consequently
135
|𝒌| = |𝒌 − 𝑮| . (9.31)
The latter condition implies that k must lie on a Bragg plane bisecting the line joining the origin
of k space and the reciprocal lattice point G, as is shown in Fig.9.4.
Fig. 9.4 |𝒌| = |𝒌 − 𝑮| , then the point k must lie in the Bragg plane determined by G.
Therefore, a weak periodic potential has its major effect on those free electron levels whose wave
vectors are close to ones at which the Bragg reflection can occur. In order to find the energy levels
and the wave functions near these points we will include in equation (9.11) only two levels: one
which corresponds to k and the other which corresponds to k-G, assuming that k lies near the
Bragg plane:
(𝐸 (𝒌) − 𝐸)𝑐 + 𝑈 𝑐 =0,

(𝐸 (𝒌 − 𝑮) − 𝐸)𝑐 +𝑈 𝑐 =0. (9.32)
These equations have the solution when the determinant is equal to zero, i.e.
𝐸 (𝒌) − 𝐸 𝑈
=0, (9.33)
𝑈 ∗ 𝐸 (𝒌 − 𝑮) − 𝐸
which leads to the quadratic equation
(𝐸 (𝒌) − 𝐸)(𝐸 (𝒌 − 𝑮) − 𝐸) − |𝑈 | = 0 . (9.34)
The two roots are
𝐸= 𝐸 (𝒌) + 𝐸 (𝒌 − 𝑮) ± 𝐸 (𝒌) − 𝐸 (𝒌 − 𝑮) + |𝑈 | . (9.35)
136
Figure 9.5 Plot of the energy bands given by eq.(9.35) for k parallel to G.
the lower band corresponds to the choice of the minus sign in eq. (9.35) and the upper band to the
plus sign. When 𝑘 = 𝐺, the two bands are separated by a band gap of magnitude 2|UG|. When k
is far removed from the Bragg plane, the levels (to leading order) are indistinguishable from their
free electron values (denoted by dotted lines).
This results is particularly simple for points lying on the Bragg plane, since in this case
𝐸 (𝒌) = 𝐸 (𝒌 − 𝑮) . We find from (9.35) then that
𝐸 = 𝐸 (𝒌) ± |𝑈 | . (9.36)
Thus, at all points in the Bragg plane, one level is uniformly raised by |𝑈 | and the other is
uniformly lowered by the same amount. This means that there are no states in the energy interval
between 𝐸 = 𝐸 (𝒌) − |𝑈 | and 𝐸 = 𝐸 (𝒌) − |𝑈 | , which implies the creation of the band gap.
The magnitude of the band gap is equal to twice the Fourier component of the crystal potential.
We illustrate this behavior using a one-dimensional lattice shown in Fig.9.6. We see the splitting
of the bands at each Bragg plane in the extended-zone scheme (Fig.9.6b). This results in the
splitting of the bands both at the boundaries and at the centre of the first Brillouin zone (Fig.9.6a).
There are two important points to note. First, since the energy there increases as 𝑘 , the higher the
band, the greater its width. Second, the higher the energy, the narrower the gap; this follows from
the fact that the gap is proportional to a Fourier component of the crystal potential and that the
137
order of the component increases as the energy rises. Since the Fourier components of the
potential decrease rapidly as the order increases, this leads to a decrease in the energy gap. It
follows therefore that, as we move up the energy scale, the bands become wider and the gaps
narrower; i.e., the electron behaves more and more like a free particle.
Fig. 9.6 (a) Dispersion curves in the nearly-free-electron model, in the reduced-zone scheme;
(b) The same dispersion curves in the extended-zone scheme.
Now we discuss the origin of the appearance of the band gaps at the Bragg planes. When k lies on
a Bragg plane we can easily find the form of the wave function corresponding to the two solutions
(9.36). Assuming for simplicity that the potential is real, we obtain from Eq.(9.32),
𝑐 = ±𝑐 . (9.37)
For simplicity we consider a one-dimensional lattice, for which the Bragg reflection occurs at
𝒌 = 1 2 𝑮 . We have then
± = 𝑒 ±𝑒 . (9.38)
We see that at the zone edge, the scattering is so strong that the reflected wave has the same
amplitude as the incident wave. The electron is represented there by a standing wave, very unlike
a free particle.
138
The distribution of the charge density is proportional to || , so that
 ∝ cos(𝐺 ∙ 𝑟 2),
 ∝ sin(𝐺 ∙ 𝑟 2), (9.39)
Since the origin lies at the ion, the  state distributes the electron so that it is piled
predominantly at the nuclei (see Fig.9.7). Since the potential is most negative there, this
distribution has a low energy. The function  therefore corresponds to the energy at the top of
band 1, that is, point A1 in Fig. 9.6a.
Figure 9.7 Spatial distributions of the charge density described by the functions 𝚿 and 𝚿
By contrast, the function  deposits its electrons mostly between the ions (as shown in Fig.9.7),
and corresponds to the bottom of band 2 in Fig.9.6a, that is, point A2. The gap arises, therefore,
because of the two different distributions for the same value of k, the distributions having
different energies.
In-text question
1. Q: Highlight how the wave functions and energy levels can be calculated when there is a weak
potential.
9.1.3 Metals and Insulators

Solids are divided into two major classes: metals and insulators. A metal – or a conductor – is a
solid in which an electric current flows under the application of electric field. By contrast,
application of an electric field produces no electric current in an insulator. There is a simple
139
criterion for distinguishing between the two classes on the basis of the band structure. If the
valence electrons exactly fill one or more bands, leaving others empty, the crystal will be an
insulator. An external electric field will not cause current flow in an insulator. Provided that a
filled band is separated by an energy gap from the next higher band, there is no continuous way to
change the total momentum of the electrons if every accessible state is filled. Nothing changes
when the field is applied.
Fig. 9.8 Occupied states and band structures giving (a) an insulator, (b) a metal or a
semimetal because of band overlap, and (c) a metal because of electron concentration.
In (b) the overlap need not occur along the same directions in the Brillouin zone. If the overlap is
small, with relatively few states involved, we speak of a semimetal. On the contrary, if the valence
band is not completely filled, the solid is a metal. In a metal there are empty states available above
the Fermi level as in a free electron gas. An application of an external electric field results in
current flow. It is possible to determine whether a solid is a metal of an insulator by considering
the number of valence electrons. A crystal can be an insulator only if the number of valence
electrons in a primitive cell of the crystal is an even integer. This is because each band can
accommodate only two electrons per primitive cell. For example, diamond has two atoms of
valence four, so that there are eight valence electrons per primitive cell. The band gap in diamond
is 7eV and this crystal is a good insulator.
However, if a crystal has an even number of valence electrons per primitive cell, it is not
necessarily an insulator. It may happen that the bands overlap in energy. If the bands overlap in
140
energy, then instead of one filled band giving an insulator, we can have two partly filled bands
giving a metal (Fig.9.8b). For example, the divalent metals, such as Mg or Zn, have two valence
electrons per cell. However, they are metals, although poor ones – their conductivity is small. If
this overlap is very small, we deal with semimetals. The best known example of a semimetal is
bismuth (Bi).
If the number of valence electrons per cell is odd, the solid is a metal. For example, the alkali
metals and the noble metals have one valence electron per primitive cell, so that they have to be
metals. The alkaline earth metals have two valence electrons per primitive cell; they could be
insulators, but the bands overlap in energy to give metals, but not very good metals. Diamond,
silicon, and germanium each have two atoms of valence four, so that there are eight valence
electrons per primitive cell; the bands do not overlap, and the pure crystals are insulators at
absolute zero. There are substances which fall in an intermediate position between metals and
insulators. If the gap between the valence band and the band immediately above it is small, then
electrons are readily excitable thermally from the former to the latter band. Both bands become
only partially filled and both contribute to the electric condition. Such a substance is known as a
semiconductor. Examples are Si and Ge, in which the gaps are about 1 and 0.7 eV, respectively.
Roughly speaking, a substance behaves as a semiconductor at room temperature whenever the gap
is less than 2 eV. The conductivity of a typical semiconductor is very small compared to that of a
metal, but it is still many orders of magnitude larger than that of an insulator. It is justifiable,
therefore, to classify semiconductors as a new class of substance, although they are, strictly
speaking, insulators at very low temperatures.
In-text questions
1. Q: Differentiate between Metals and Insulators in terms of energy band gap and occupied
states.
2. Q: Briefly describe a semiconductor.

A: See page 139.
141
In this study session, you have learnt the:
1. Calculation of wave functions and energy levels in both strong and weak potential.
2. Distinguishing features of metals, semiconductors and insulators.

SAQ 9.1
Compute the concentration of electrons and holes in an intrinsic semiconductor InSb at room
temperature (Eg = 0.2eV, me = 0.01m and mh = 0.018 m). Determine the position of the Fermi energy.
SAQ 9.2
Indium antimonide has Eg = 0.23 eV; dielectric constant ε = 18; electron effective mass me = 0.015 m.
Calculate the donor ionization energy and the radius of the ground state orbit. At what minimum donor
concentration will appreciable overlap effects between the orbits of adjacent impurity atoms occur?
This overlap tends to produce an impurity band - a band of energy levels which permit conductivity
presumably by a hopping mechanism in which electrons move from one impurity site to a neighboring
ionized impurity site.
SAQ 9.3
2 2
Given the data for Si: μe = 1350 cm /V⋅s, μh = 475 cm /V⋅s, me = 0.19m, mh = 0.16m and Eg = 1.1 eV,
calculate
a) The lifetimes of electrons and holes.

b) The intrinsic conductivity, σ, at room temperature.
142
Study Session 10 Semiconductors I
Introduction
Semiconductors are materials whose electronic properties are intermediate between those of metal
and insulators. These intermediate properties are determined by the crystal structure, bonding
143
characteristics, and electronic energy bands and also by the fact that, unlike metals, a
semiconductor has both positive (hole) and negative (electron) carriers of electricity whose
densities can be controlled by doping the pure semiconductor with chemical impurities during
crystal growth. The climax of investigations into these materials was the invention and
development of the transistor by J. Bardeen, W. Brattain and W. Shockley in 1948 for which they
received the Nobel Prize in 1956.

When you have studied this session, you should be able to explain:
10.1 The crystal structure, band structure and crystal bonding in semiconductors.
10.2 The different classes of semiconductors.
10.3 Statistical distribution of carriers in semiconductors.
10.4 Electron conductivity and mobility in semiconductors.
10.1 Main Content

10.1.1 Crystal Structure and Bonding
Semiconductors include a large number of substances of widely different chemical and physical
properties. These materials are grouped into several classes of similar behavior, the classification
being based on the position in the periodic table of the elements.
The best-known class is the Group IV semiconductors - C (diamond), Si, Ge, - all of which lie in
the fourth column of the periodic table. They have been studied intensively, particularly Si and
Ge, which have found many applications in electronic devices. The elemental semiconductors all
crystallize in the diamond structure. The diamond structure has an fcc lattice with a basis
composed of two identical atoms, and is such that each atom is surrounded by four neighboring
atoms, forming a regular tetrahedron. Group IV semiconductors are covalent crystals, i.e., the
atoms are held together by covalent bonds. These bonds consist of two electrons of opposite spins
distributed along the line joining the two atoms. The covalent electrons forming the bonds are
hybrid sp3 atomic orbitals.
Another important group of semiconductors is the Group III-V compounds, so named because
each contains two elements, one from the third and the other from the fifth column of the periodic
144
table. The best-known members of this group are GaAs and InSb (indium antimonite), but the list
also contains compounds such as GaP, InAs, GaSb, and many others.
These substances crystallize in the zincblende structure, which is the same as the diamond
structure, except that the two atoms forming the basis of the lattice are now different. Thus, in
GaAs, the basis of the fcc lattice consists of two atoms Ga and As. Due to this structure, each
atom is surrounded by four others of the opposite kind, and these latter atoms form a regular
tetrahedron, just as in the diamond structure.
The bonding in the III-V compounds is also primarily covalent. The eight electrons required for
the four tetrahedral covalent bonds are supplied by the two types of atoms, the trivalent atom
contributing its three valence electrons, and the pentavalent atom five electrons. The bonding in
this group is not entirely covalent. The two elements in the compound are different; the
distribution of the electrons along the bond is not symmetric, but displaced toward one of the
atoms. As a result, one of the atoms acquires a net electric charge. Such a bond is called
heteropolar, in contrast to the purely covalent bond in the elemental semiconductors, which is
called homopolar.
The distribution of electrons in the bond is displaced toward the atom of higher electronegativity.
In GaAs, for instance, the As atom has a higher electronegativity than the Ga, and consequently
the As atom acquires a net negative charge, whose value is -0.46e per atom (a typical value in
Group III-V compounds). The Ga atom correspondingly acquires a net positive charge of 0.46e.
Charge transfer leads to an ionic contribution to the bonding in Group III-V compounds. Their
bonding is therefore actually a mixture of covalent and ionic components, although covalent ones
predominate in most of these substances.
10.1.2. Band Structure

A semiconductor is a solid in which the highest occupied energy band, the valence band, is
completely full at 𝑇 = 0 𝑘 , but in which the gap above this band is also small, so that electrons
may be excited thermally at room temperature from the valence band to the next-higher band,
which is known as the conduction band. Generally, the number of excited electrons is appreciable
(at room temperature) whenever the energy gap 𝐸 is less than 2 eV. The substance may then be
classified as a semiconductor. When the gap is larger, the number of electrons is negligible, and
the substance is an insulator.
145
When electrons are excited across the gap, the bottom of the conduction band (CB) is populated
by electrons, and the top of the valence band (VB) by holes. As a result, both bands are only
partially full, and would carry a current if an electric field were applied. The conductivity of the
semiconductor is small compared with the conductivities of metals because of the small number
of electrons and holes involved, but this conductivity is nonetheless sufficiently large for practical
purposes.
Fig. 10.1 Band structure in a semiconductor.
The simplest band structure of a semiconductor is indicated in Fig. 10.1. Since we are interested
only in the regions which lie close to the band gap, where electrons and holes lie, we can ignore a
more complex variation of the energy bands far away from the gap.
The energy of the conduction band has the form
ℏ
𝐸 (𝑘) = 𝐸 + (10.1)
where k is the wave vector and 𝑚 the effective mass of the electron. The energy 𝐸 represents
the energy gap. The zero-energy level is chosen to lie at the top of the valence band.
The energy of the VB (Fig. 10.1) may be written as
ℏ
𝐸 (𝑘) = 𝐸 − (10.2)
146
where 𝑚 is the effective mass of the hole, which is positive. (Because of the inverted shape of
the VB, the mass of an electron at the top of the VB is negative, but the mass of a hole is positive.)
Within this simple picture of the semiconductor, the primary band-structure parameters are thus
the electron and hole masses 𝑚 and 𝑚 and the band gap 𝐸 . Table 10.1 gives the parameters
for various semiconductors. Note that the masses differ considerably from the free-electron mass.
In many cases they are much smaller than the free-electron mass. The energy gaps range from
0.18 eV in InSb to 3.7 eV in ZnS. The table also shows that the wider the gap, the greater the mass
of the electron.
Table 10.1. Band Structure parameters of Semiconductors
Group Crystal Eg(eV) me/m mh/m

IV C 5.3
IV Si 1.1 0.19 0.16
IV Ge 0.7 0.08 0.04
III-V GaAs 1.4 0.07 0.09
III-V GaP 2.3 0.12 0.50
III-V InSb 0.2 0.01 0.18
II-VI ZnS 3.6 0.40 5.41
II-VI ZnSe 2.7 0.10 0.60
II-VI CdSe 1.7 0.13 0.45
The energy gap for a semiconductor varies with temperature, but the variation is usually slight.
That variation with temperature should exist at all can be appreciated from the fact that the crystal,
when it is heated, experiences a volume expansion, and hence a change in its lattice constant.
This, in turn, affects the band structure, which is a sensitive function of the lattice constant.
The band structure in Fig.10.1 is the simplest possible structure. Band structures of real
semiconductors are somewhat more complicated, as we shall see, but for the present the simple
structure will suffice for our purposes.
In-text questions
1. Q: Differentiate between homopolar and heteropolar bonding in semiconductors.

A: See page 143.
2. Q: Briefly explain how semiconductors are formed.
147
3. Q: How many free electrons does one donor atom contribute?
A: See page 143.
4. Q: Give the graphical representation of the energy band model of silicon.
A: See page 144 (Figure 10.1)
10.1.3. Types of Semiconductors
Intrinsic Semiconductors
In the field of semiconductors, electrons and holes are usually referred to as free carriers, or
simply carriers, because it is these particles which are responsible for carrying the electric current.
The number of carriers is an important property of a semiconductor, as this determines its
electrical conductivity. Intrinsic semiconductors are semiconductors in which the number of
carriers and the conductivity are not influenced by impurities. Intrinsic conductivity is typical at
relatively high temperatures in highly purified specimens.
In order to determine the number of carriers, we need some of the basic results of statistical
mechanics. The most important result in this regard is the Fermi-Dirac (FD) distribution function
𝑓(𝐸) = ( )⁄ (10.3)
This function gives the probability that an energy level E is occupied by an electron when the
system is at temperature T.
The function is plotted versus E in Fig.10.2. Here, we see that as the temperature rises, the
unoccupied region below the Fermi level 𝐸 becomes longer, which implies that the occupation of
high energy states increases as the temperature is raised, a conclusion which is most plausible,
since increasing the temperature raises the overall energy of the system.
148
Fig. 10.2 The Fermi-Dirac distribution function
We shall see later that the Fermi level in intrinsic semiconductors lies close to the middle of the
band gap. Therefore, we can represent the distribution function and the conduction and valence
bands of the semiconductor by Fig.10.3.
Fig. 10.3 (a) Conduction and valence bands, (b) The distribution function, (c) Density of
states for electrons and holes.
149
First we calculate the concentration of electrons in the CB. The number of states in the energy
range (E, E + dE) is equal to 𝐷 (E)dE, where 𝐷 (E) is the density of electron states. Since each of
these states has an occupation probability f(E), the number of electrons actually found in this
energy range is equal to f(E) 𝐷 (E)dE. The concentration of electrons throughout the CB is thus
given by the integral over the conduction band
∞
𝑛 = ∫ 𝑓 (𝐸)𝐷 (𝐸)𝑑𝐸 (10.4)
Where 𝐸 is the bottom, the conduction band, as shown in Fig.10.3. The band gap in
semiconductors is of the order of 1eV, which is much larger than kT. Therefore, (𝐸 − 𝜇) ≫ 𝑘 𝑇
and we can neglect the unity term in the denominator of the distribution function (10.3), so that
( )⁄
𝑓 (𝐸) ≈ 𝑒 . (10.5)
The density of states for the conduction band is given by
⁄
⁄
𝐷 (𝐸) = (𝐸 − 𝐸 ) . (10.6)
ℏ
Note that 𝐷 (𝐸) vanishes for 𝐸 < 𝐸 , and is finite only for > 𝐸 , as shown in Fig. 10.3.
When we substitute equations for 𝑓 (𝐸) and 𝐷 (𝐸) into Eq. (10.4), we obtain
⁄ ∞
⁄ ⁄ ⁄
𝑛= 𝑒 ∫ (𝐸 − 𝐸 ) 𝑒 𝑑𝐸 . (10.7)
ℏ
By changing the variable, and using the result
∞ ⁄ √
∫ 𝑥 𝑒 𝑑𝐸 = (10.8)
one can readily evaluate the integral in (10.7). The electron concentration then reduces to the
expression
⁄
)⁄
𝑛=2 𝑒( (10.9)
πℏ
The electron concentration is still not known explicitly because the Fermi energy 𝜇 is so far
unknown.
150
Essentially the same ideas employed above may also be used to evaluate the number of holes in
the VB. The probability that a hole occupies a level E in this band is equal to 1 − 𝑓 (𝐸), since
𝑓 (𝐸) is the probability of electron occupation. Assuming that the Fermi level lies close to the
middle of the band gap, i.e. (𝜇 − 𝐸 ) ≫ 𝑘 𝑇 for the valence band, we find for the distribution
function of holes
( )⁄
𝑓 (𝐸) = 1 − ( )⁄ = ( )⁄ ≈𝑒 . (10.10)
The density of states for the holes is
⁄
⁄
𝐷 (𝐸) = (𝐸 − 𝐸) , (10.11)
ℏ
where 𝐸 is the energy of the valence band edge. Proceeding in a similar fashion as we did for
electrons, we find the concentration of holes in the valence band
⁄
)⁄
𝑝=∫ 𝑓 (𝐸)𝐷 (𝐸)𝑑𝐸 = 2 𝑒( . (10.12)
πℏ
The electron and hole concentrations have thus far been treated as independent quantities. For
intrinsic semiconductors, the two concentrations are, in fact, equal, because the electrons in the
CB are due to excitations from the VB across the energy gap, and for each electron thus excited, a
hole is created in the VB. Therefore,
𝑛=𝑝 (10.13)
and
⁄ )⁄ ⁄ )⁄
(𝑚 ) 𝑒( = (𝑚 ) 𝑒( . (10.14)
We obtain, then, for the Fermi energy
𝜇= + 𝑘 𝑇 ln (10.15)
The second term on the right of (10.15) is very small compared with the first, and the energy level
is close to the middle of the energy gap. This is consistent with earlier assertions that both the
bottom of the CB and the top of the VB are far from the Fermi level.
151
Fig. 10.4 Electron concentration n versus 1/T in Ge (in units 10 -3K-1)
The concentration of electrons may now be evaluated explicitly by using the above value of 𝜇.
Substitution of Eq. (10.15) into Eq. (10.9) yields
⁄
⁄ ⁄
𝑛=2 (𝑚 𝑚 ) 𝑒 (10.16)
πℏ
where 𝐸 = 𝐸 − 𝐸 is the band gap. The important feature of this expression is that n increases
very rapidly - exponentially - with temperature, particularly by virtue of the exponential factor.
Thus, as the temperature is raised, a vastly greater number of electrons is excited across the gap.
Figure 10.4. is a plot of log n versus 1/T. The curve is a straight line of slope equal to (−𝐸 ⁄2𝑘 ).
⁄
[The 𝑇 - dependence in (10.16) is so weak in comparison with the exponential dependence that
the former may be disregarded for the purpose of this discussion.)
For Ge at room temperature ~10 𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑛𝑠⁄𝑐𝑚 , a typical value of carrier concentration in

semiconductors.
Note that the expression (10.16) also gives the hole concentration, since n = p. Our discussion of
carrier concentration in this section is based on the premise of a pure semiconductor. When the
substance is impure, additional electrons or holes are provided by the impurities. In that case, the
152
concentrations of electrons and holes may no longer be equal, and the amount of each depends on
the concentration and type of impurity present. When the substance is sufficiently pure so that the
concentrations of electrons and holes are equal, we speak of an intrinsic semiconductor. That is,
the concentrations are determined by the intrinsic properties of the semiconductor itself. On the
other hand, when a substance contains a large number of impurities which supply most of the
carriers, it is referred to as an extrinsic semiconductor.
Impurity States (extrinsic semiconductor)

A pure semiconductor has equal numbers of both types of carriers, electrons and holes. In most
applications, however, one needs specimens which have one type of carrier only, and none of the
other. By doping the semiconductor with appropriate impurities, one can obtain samples which
contain either electrons only or holes only.
Fig 10.5 An As impurity in a Si crystal.
The extra electron migrates through the crystal. Consider, for instance, a specimen of Si which has
been doped by As. The As atoms (the impurities) occupy some of the lattice sites formerly
occupied by the Si host atoms. The distribution of the impurities is random throughout the lattice.
But their presence affects the solid in one very important respect: The As atom has valence 5
while Si has valence 4. Of the five electrons of As, four participate in the tetrahedral bond of Si,
as shown in Fig. 10.5. The fifth electron cannot enter the bond, which is now saturated, and hence
this electron detaches from the impurity and is free to migrate through the crystal as a conduction
electron, i.e., the electron enters the CB. The impurity is now actually a positive ion, As + (since it
has lost one of its electrons), and it tends to capture the free electron. But we shall show shortly
that the attraction force is very weak, and not enough to capture the electron in most
circumstances. The net result is that the As impurities contribute electrons to the CB of the
153
semiconductors, and for this reason these impurities are called donors. Note that the electrons
have been created without the generation of holes.
When an electron is captured by an ionized donor, it orbits around the donor much like the
situation in hydrogen. We can calculate the binding energy by using the familiar Bohr model.
However, we must take into account the fact that the coulomb interaction here is weakened by the
screening due to the presence of the semiconductor crystal, which serves as a medium in which
both the donor and ion reside. Thus, the coulomb potential is now given by
𝑉(𝑟) = − , (10.17)
where ε is the reduced dielectric constant of the medium. The dielectric constant ε = 11.7 in Si, for
example, shows a substantial decrease in the interaction force. It is this screening which is
responsible for the small binding energy of the electron at the donor site.
Using this potential in the Bohr model, we find the binding energy corresponding to the ground
state of the donor to be
𝐸 =− . (10.18)
ℏ
Note that the effective mass 𝑚 has been used rather than the free mass m. The binding energy of
the hydrogen atom is equal to 13.6 eV. The binding energy of the donor is reduced by the factor
1⁄𝜀 , and also by the mass factor 𝑚 ⁄𝑚, which is usually smaller than unity. Using the typical
values 𝜀~10 and 𝑚 ⁄𝑚 ~0.1 we find that the binding energy of the donor is about 10 -3 of the
hydrogen energy, i.e., about 0.01 eV. This is indeed the order of the observed values.
154
Fig. 10.6 The donor level in a semiconductor.
The donor level lies in the energy gap, very slightly below the conduction band, as shown in
Fig.10.6. The level is so close to the CB, almost all the donors are ionized at room temperature,
their electrons having been excited into the CB.
It is instructive to evaluate the Bohr radius of the donor electron. Straightforward adaptation of the
Bohr result leads to
𝑟 =𝜀 𝑎 . (10.19)
where 𝑎 is the Bohr radius, equal to 0.53 Å. The radius of the orbit is thus much larger than 𝑎 ,
by a factor of 100, if we use the previous values for 𝜀 and 𝑚 . A typical radius is thus of the order
of 50Å. Since this is much greater than the interatomic spacing, the orbit of the electron encloses a
great many host atoms, and our picture of the lattice acting as a continuous, polarisable dielectric
is thus a plausible one.
Since the donors are almost all ionized, the concentration of electrons is nearly equal to that of the
donors. Typical concentrations are about 10 𝑐𝑚 . But sometimes much higher concentrations
are obtained by heavy doping of the sample, for example, 10 𝑐𝑚 or even more.
Acceptors
An appropriate choice of impurity may produce holes instead of electrons. Suppose that the Si
crystal is doped with Ga impurity atoms. The Ga impurity resides at a site previously occupied by
155
a Si atom, but since Ga is trivalent, one of the electron bonds remains vacant (Fig. 10.7). This
vacancy may be filled by an electron moving in from another bond, resulting in a vacancy (or
hole) at this latter bond. The hole is then free to migrate throughout the crystal. In this manner, by
introducing a large number of trivalent impurities, one creates an appreciable concentration of
holes, which lack electrons.
The trivalent impurity is called an acceptor, because it accepts an electron to complete its
tetrahedral bond. The acceptor is negatively charged, by virtue of the additional electron it has
entrapped. Since the resulting hole has a positive charge, it is attracted by the acceptor. We can
evaluate the binding energy of the hole at the acceptor in the same manner followed above in the
case of the donor. Again, this energy is very small, of the order of 0.01 eV. Thus essentially all the
acceptors are ionized at room temperature.
Fig. 10.7 A Ga impurity in a Si crystal (The extra hole migrates through the crystal).
The acceptor level lies in the energy gap, slightly above the edge of the VB, as shown in Fig. 10.8.
This level corresponds to the hole being captured by the acceptor. When an acceptor is ionized (an
electron excited from the top of the VB to fill this hole), the hole falls to the top of the VB, and is
now a free carrier. Thus, the ionization process, indicated by upward transition of the electron on
the energy scale, may be represented by a downward transition of the hole on this scale.
156
Fig. 10.8 The acceptor level in a semiconductor.
In-text questions
1. Q: How are n-type and p-type semiconductor produced?
2. Q: Write down as many as you can of the ways in which the extrinsic semiconductors differ
from an intrinsic semiconductor.
3. Q: Mark the Fermi-level for (i) an intrinsic semiconductor (ii) n-type semiconductor and (iii) p-
type semiconductor.

1. The definition of semiconductors and understand band structures of semiconductors.

2. The different types of semiconductors and their properties.
3. The distribution of carriers in semiconductor using Fermi-Dirac statistics.
157
Self –Assessment Questions for Study Session 10
SAQ 10.1
Enumerate some of the properties of semiconductors.
SAQ 10.2
Explain with suitable diagrams the conduction band, valence band and the forbidden band and
hence explain the contribution of electrons and holes to electrical conduction.
SAQ 10.3
What is a hole in a semiconductor and how is it formed?
SAQ 10.4
Distinguish between intrinsic and impurity semiconductors. Give an example for each class of
semiconductors.

158
Study Session 11 Semiconductors II
Introduction
As previously stated, at infinite temperature some electrons will acquire sufficient thermal energy
to raise them from the valence band to the conduction band. The actual number depends on the
number of permissible electron energy levels and the probability of these levels being occupied.

When you have studied this session, you should be able to explain:
11.1 The various electronic processes in semiconductors.
11.2 The Statistical distribution of carriers in semiconductors.
11.3 The characteristics of excitons.
11.1 Main Content

11.1.1. Semiconductor Statistics
Semiconductors usually contain both donors and acceptors. Electrons in the CB can be created
either by thermal excitation or by thermal ionization of the donors. Holes in the VB may be
generated by interband excitation, or by thermal excitation of electrons from the VB into the
acceptor level. In addition, electrons may fall from the donor levels to the acceptor level. Figure
11.1 indicates these various processes.
Fig.11.1 The various electronic processes in a semiconductor.
159
Finding the concentrations of carriers, both electrons and holes, taking all these processes into
account, is quite complicated. We shall treat a few special cases, which are often encountered in
practice. Two regions may be distinguished, depending on the physical parameters involved: The
intrinsic and the extrinsic regions.
The intrinsic region

The concentration of carriers in the intrinsic region is determined primarily by thermally induced
interband transitions. In this region n = p. The intrinsic region obtains when the impurity doping is
small. When we denote the concentrations of donors and acceptors by N d and Na, the requirement
for the validity of the intrinsic condition is
𝑛 ≫ 𝑁 ,𝑁 (11.1)
Since n increases rapidly with temperature, the intrinsic condition becomes more favorable at
higher temperatures. All semiconductors, in fact, become intrinsic at sufficiently high
temperatures (unless the doping is unusually high).
The extrinsic region

Quite often, the intrinsic condition is not satisfied. For the common dopings encountered, about
1015cm-3, the number of carriers supplied by the impurities is large enough to change the intrinsic
concentration appreciably at room temperature. The contribution of impurities, in fact, frequently
exceeds those carriers that are supplied by interband excitation. When this is so, the sample is in
the extrinsic region. Two different types of extrinsic regions may be distinguished. The first
occurs when the donor concentration greatly exceeds the acceptor concentration, that is, when
𝑁 ≫ 𝑁 . In this case, the concentration of electrons may be evaluated quite readily. Since the
donor's ionization energy (i.e. the binding energy) is quite small, all the donors are essentially
ionized, their electrons going into the conduction band. Therefore, to a good approximation,
𝑛=𝑁 (11.2)
A semiconductor in which n >> p is called an n-type semiconductor (n for negative). Such a
sample is characterized, as we have seen, by a great concentration of electrons. The other type of
extrinsic region occurs when Na >> Nd, that is, the doping is primarily by acceptors. Using an
argument similar to the above, one then has
160
𝑝=𝑁 (11.3)
i.e., all the acceptors are ionized. Such a material is called a p-type semiconductor. It is
characterized by a preponderance of holes.
In discussing ionization of donors (and acceptors), we assumed that the temperature is sufficiently
high so that all of these are ionized. This is certainly true at room temperature. But if the
temperature is progressively lowered, a point is reached at which the thermal energy becomes too
small to cause electron excitation. In that case, the electrons fall from the conduction band into the
donor level, and the conductivity of the sample diminishes dramatically. This is referred to as
freeze-out, in that the electrons are now "frozen" at their impurity sites. The temperature at which
freeze-out takes place is Ed ~ kT, which gives a temperature of about 100°K.
The variation of the electron concentration with temperature in an n-type sample is indicated
schematically in Fig. 11.2.
Fig. 11.2 Variation of electron concentration n with temperature in an n-type semiconductor.
11.1.2. Electrical conductivity; Mobility

Electrical conductivity is, of course, the quantity of primary interest in semiconductors. Both
electrons and holes contribute to electric current. Assume first that a sample is strongly n-type and
contains only one type of carrier: electrons. The conductivity can be treated according to the free-
electron model:
161
𝜎 = (11.4)
where 𝑚 is an effective mass and 𝜏 is the lifetime of the electron. To estimate the value for 𝜎 ,
we substitute 𝑛 = 10 𝑐𝑚 , which is eight orders in magnitude less than that in metals, and
𝑚 = 0.1𝑚 This leads to 𝜎 ~10 (𝜇𝑜ℎ𝑚. 𝑐𝑚) , which is a typical figure in semiconductors.
Although this is many orders of magnitude smaller than the value in a typical metal, where
𝜎 ~1(𝜇𝑜ℎ𝑚. 𝑐𝑚) , the conductivity in a semiconductor is still sufficiently large for practical
applications.
Semiconductor physicists often use another transport coefficient: mobility. The mobility 𝜇 is
defined as the proportionality coefficient between the electron drift velocity and the applied
electric field, i.e.
𝑣 =𝜇 𝐸, (11.5)
where |𝑣 | is the absolute value of the velocity. Taking into account that 𝒋 = −𝑒𝑛 𝑣 and 𝒋 =
𝜎 𝐸, we find that
𝜇 = . (11.6)
As defined, mobility is a measure of the rapidity of the motion of the electron in the field. The
longer the lifetime of the electron and the smaller its mass, the higher the mobility.
We can now express electrical conductivity in terms of mobility. We can write
𝜎 = 𝑛𝑒𝜇 (11.7)
indicating that 𝜎 is proportional to 𝜇 . A typical value for 𝜇 may be obtained by substituting

𝜎 ~10 (𝜇𝑜ℎ𝑚. 𝑐𝑚) and 𝑛 = 10 𝑐𝑚 in Eq.(11.7). This yields
𝜇 = 10 𝑐𝑚 𝑉 𝑠 (11.8)
What we have said about electrons in a strongly n-type substance can be carried over to a
discussion of holes in a strongly p-type substance. The conductivity of the holes is given by
𝜎 = = 𝑝𝑒𝜇 (11.9)
where 𝜇 is the hole mobility.
162
Let us now treat the general case in which both electrons and holes are present. When a field is
applied, electrons drift opposite to the field and holes drift in the same direction as the field. The
currents of the two carriers are additive, however, and consequently the conductivities are too.
Therefore,
𝜎 =𝜎 +𝜎 (11.10)
i.e., both electrons and holes contribute to the current. In terms of the mobilities, one may write
𝜎 = 𝑛𝑒𝜇 + 𝑝𝑒𝜇 (11.11)
The carriers' concentrations n and p may be different if the sample is doped, as discussed before.
Also, one or the other of the carriers may dominate, depending on whether the semiconductor is n-
or p-type. When the substance is in the intrinsic region, however, n = p, and Eq. (10.30) becomes
𝜎 = 𝑛𝑒(𝜇 + 𝜇 ) (11.12)
where n is the intrinsic concentration. Even now the two carriers do not contribute equally to the
current. The carrier with the greater mobility - usually the electron - contributes the larger share.
Dependence on Temperature
Conductivity depends on temperature, and this dependence is often pronounced. Consider a
semiconductor in the intrinsic region. Its conductivity is expressed by (11.12). But in this
situation, the concentration 𝑛 increases exponentially with temperature, as may be recalled from
(10.16). We may write the conductivity in the form
⁄
𝜎 = 𝐹(𝑇)𝑒 (11.13)
where F(T) is a function which depends only weakly on the temperature. (This function depends
on the mobilities and effective masses of the carriers.) Thus, conductivity increases exponentially
with temperature from 0.0022 K-1 as shown in Fig.11.3.
163
Fig. 11.3 Conductivity of Si versus 1/T in the intrinsic range.
This result can be used to determine the energy gaps in semiconductors. In the early days of
semiconductors this was the standard procedure for finding the energy gap. Nowadays, however,
the gap is often measured by optical methods. When the substance is not in the intrinsic region, its
conductivity is given by the general expression (11.11). In that case, the temperature dependence
of the conductivity on T is not usually as strong as indicated above. To see the reason for this,
suppose that the substance is extrinsic and strongly n-type. The conductivity is
𝜎 = 𝑛𝑒𝜇 (11.14)
But the electron concentration n is now a constant equal to 𝑁 , the donor (hole) concentration.
Also, any temperature dependence present must be due to the mobility of electrons or holes.
Mobility versus Temperature: scattering mechanisms

Mobility of electrons (or holes) varies with temperature. In n-type semiconductor, it is expressed
as :
164
𝜇 = = . (11.15)
Since the lifetime of the electron, or its collision time, varies with temperature, its mobility also
varies with temperature. Normally, both lifetime and mobility diminish as the temperature rises.
The relaxation time is given by 𝜏 = 𝑙 ⁄𝑉 , where 𝑙 is the mean free path of the electron and 𝑉 is
the drift velocity. The velocity of electrons is different depending on their location in the
conduction band. Electrons at the bottom of the conduction band in a semiconductor obey the
classical statistics and not the highly degenerate Fermi statistics prevailing in metals. The higher
electrons are in the band, the greater their velocity.
We can evaluate the conductivity by assuming that 𝑉 is the average velocity. The average
velocity can be estimated using the procedure of the kinetic theory of gases:
𝑚 𝑉 = 𝑘𝑇 (11.16)
⁄
This introduces a factor of 𝑇 dependence in the mobility:
𝜇 = ⁄ . (11.17)
( ) ⁄
The mean free path 𝑙 also depends on the temperature, and in much the same way as it does in
metals. 𝑙 is determined by the various collision mechanisms acting on the electrons. These
mechanisms are the collisions of electrons with thermally excited phonons and collisions with
impurities. At high temperatures, at which collisions with phonons is the dominant factor, 𝑙 is
⁄
inversely proportional to temperature, that is, 𝑙 ~𝑇 . In that case, mobility varies as 𝜇 ~𝑇 .
Figure 11.4. shows this for Ge.
Another important scattering mechanism in semiconductors is that of ionized impurities. When a

substance is doped, the donors (or acceptors) lose their electrons (or holes) to the conduction
band.
The impurities are thus ionized and are quite effective in scattering the electrons (holes). At high
temperatures, this scattering is masked by the much stronger phonon mechanism, but at low
temperatures, this latter mechanism becomes weak and the ionized-impurity scattering gradually
takes over.
165
Fig. 11.4 Electron mobility versus T in Ge. The dashed curve represents pure phonon
scattering; numbers in parentheses refer to donor concentrations.
11.1.3 Band Structure of Real Semiconductors

So far, we have assumed the simplest possible band structure, namely, a conduction band of a
standard form, centered at the origin, k = 0, and a valence band of a standard inverted form, also
centered at the origin. Such a simple structure is applicable for elucidating many observed
phenomena, but it does not represent the actual band structures of many common semiconductors.
Only when one uses the actual band structure is it possible to obtain a quantitative agreement
between experiments and theoretical analysis.
A material whose band structure comes close to the ideal structure is GaAs (Fig. 11.5). The
conduction band has a minimum at the origin k = 0, and the region close to the origin is well
ħ
represented by a quadratic energy dependence, (𝑘) = , where 𝑚 = 0.072𝑚. Since the
electrons are most likely to populate this region, one can represent this band by a single effective
166
mass. Note however that, as k increases, the energy E(k) is no longer quadratic in k, and those
states may no longer by represented by a single, unique effective mass. The dependence of energy
on k in the neighborhood of this secondary minimum is quadratic, and hence an effective mass
may be defined locally, but its value is much greater than that of the primary minimum (at the
center). The actual value is 0.36m. Due to cubic symmetry, there are six equivalent secondary
minima, or valleys, in all, along the [100] directions.
Fig. 11.5 Band structure of GaAs plotted along the [100] and [111] directions.
These secondary valleys do not play any role under most circumstances, since the electrons
usually occupy only the central or primary valley. In such situations, these secondary valleys may
be disregarded altogether. There are also other secondary valleys in the [111] directions, as shown
in Fig. 11.5. These are higher than the [100] valleys, and hence are even less likely to be
populated by electrons.
The valence band is also illustrated in Fig.11.5. Here, it is composed of three closely spaced sub-
bands. The curvatures of the bands are different, so are the effective masses of the corresponding
holes. One speaks of light holes and heavy holes.
Other III-V semiconductors have band structures quite similar to that of GaAs.
167
Fig.11.6 (a) Band structure of Si plotted along the [100] and [111] directions (b) Ellipsoidal
energy surfaces corresponding to primary valleys along the [100] directions.
Figure 11.6(a) shows the band structure of Si. An interesting feature is that the conduction band
has its lowest (primary) minimum not at k=0. The minimum lies along the [100] direction, at
about 0.85 the distance from the center to the edge of the zone. Note that the bottom of the
conduction does not lie directly above the top of the valence band. This type of semiconductors is
known as indirect gap semiconductors. These should be distinguished from direct gap
semiconductors such as GaAs.
Due to the nature of the cubic symmetry, there are actually six equivalent primary valleys located
along the [100] directions. These are illustrated in Fig. 11.6(b). The energy surfaces at these
valleys are composed of elongated ellipsoidal surfaces of revolution, whose axes of symmetry are
along the [100] directions. There are two different effective masses which correspond to these
surfaces: the longitudinal and the transverse effective masses. The longitudinal mass is 𝑚 =
168
0.97m, while the two identical transverse masses are 𝑚 = 0.19m. The mass anisotropy ratio is
about 5.
The valence band in silicon is represented by three different holes (Fig.11.6a). One of the holes is
heavy (𝑚 = 0.5m), and the other two are light. The energy gap in Si, from the top of the valence
band to the bottom of the conduction band, is equal to 1.08 eV. The fact that the bottom of the
conduction does not lie directly above the top of the valence band, is irrelevant to the definition of
the band gap.
11.1.4. Excitons
An electron and a hole may be bound together by their attractive coulomb interaction, just as an
electron is bound to a proton to form a neutral hydrogen atom. The bound electron-hole pair is
called an exciton, Fig.11.7. An exciton can move through the crystal and transport energy; it does
not transport charge because it is electrically neutral. It is similar to positronium, which is formed
from an electron and a positron.
Excitons can be formed in every insulating crystal. All excitons are unstable with respect to the
ultimate recombination process in which the electron drops into the hole.
Fig.11.7 The exciton shown is weakly bound, with an average electron-hole distance large in
comparison with a lattice constant.
An exciton is a bound electron-hole pair, usually free to move together through the crystal.
The binding energy of the exciton can be measured by optical transitions from the valence band,
by the difference between the energy required to create an exciton and the energy to create a free
electron and free hole, Fig.11.8.
169
Fig.11.8 Energy levels of an exciton.
Optical transitions from the top of the valence band are shown by the arrows; the longest arrow
corresponds to the energy gap. The binding energy of the exciton is E , referred to a free electron
and free hole.
Energy levels of an exciton can be calculated as follows. Consider an electron in the conduction
band and a hole in the valence band. The electron and hole attract each other by the Coulomb
potential
𝑉(𝑟) = − , (11.18)
where r is the distance between the particles and ε is the appropriate dielectric constant. There will
be bound states of the exciton system having total energies lower than the bottom of the
conduction band.
The problem is the hydrogen atom problem if the energy surfaces for the electron and hole are
spherical and non-degenerate. The energy levels are given by
𝐸 =𝐸 − . (11.19)
ℏ
Here 𝑛 is the principal quantum number and 𝜇 the reduced mass:
= + (11.20)
formed from the effective masses of the electron and hole.

170
The exciton ground state energy is obtained on setting 𝑛 = 1 . This is the ionization energy of the
exciton.
In-text questions:
1. Q: What are Excitons? Explain briefly how the energy levels of excitons can be calculated.
2. Q: Calculate the position of Fermi level Ef and the conductivity at 300k for a germanium crystal
containing 5 x 1022 arsenic atoms/m3. Also calculate the conductivity if the mobility of the
electron is 0.39 m2V-1s-1.
A: Ec - Ef =0.16eV ; 𝜎 = 3120𝛺 𝑚
3. Q: Obtain the equation for the conductivity of an intrinsic semiconductor in terms of carrier
concentration and carrier mobilities. Suggest a method for evaluating the energy gap of
semiconductors.

1. The distribution of carriers in semiconductors using the Fermi-Dirac statistics.

2. About excitons and its characteristics.
Self –Assessment Questions for Study Session 11

SAQ 11.1
Consider a sample of n-type silicon with Nd = 1021/m3. Find n and p at 300K. The number of intrinsic
carriers at 300K is 9.8 x 1015.
171
SAQ 11.2
In a semiconductor the effective mass of the electron is 0.07m 0 and that of a hole is 0.4m0, where m0 is
the free electron mass. Assuming that the average relaxation time for the holes is half that for
electrons, calculate the mobility of the holes when the mobility of the electrons is 0.8m 2V-1s-1.
SAQ 11.3
Compare the density of charge carriers in a pure silicon crystal at the two temperatures 27 0C and 570C.
Eg for Si is 1.1eV.
SAQ 11.4
Intrinsic semiconductor material A has an energy gap 0.36eV while material B has an energy gap of
0.72 eV. Compare the intrinsic density of carriers in these two semiconductors at 300K. Assume that
the effective masses of all the electrons and holes are equal to the free electron mass.
Richard Turton (2000), The Physics of Solids, Oxford University Press.
172
Study Session 12 Hall Effect
Introduction
It is often necessary to determine whether a material is n-type or p-type. Measurement of the

conductivity of a specimen will not give this information since it cannot distinguish between
positive hole and electron conduction. The Hall Effect can be utilized in distinguishing between
the two types of carriers, and it also allows the density of the charge carriers to be determined.
12.1 Understand the Hall Effect.
12.2 Distinguish between the two types of carriers using Hall Effect.
12.3 Determine the density of the charge carriers using Hall Effect.
12.1 Main Content

12.1.1 Hall Voltage
If a piece of conductor (metal or semiconductor) carrying a current is placed in a transverse
magnetic field, an electric field is produced inside the conductor in a direction normal to both the
current and magnetic field. This phenomenon is known as the Hall Effect, and the generated
voltage is called the Hall voltage. See Fig. 12.1 below
Fig.12.1. Illustration of the Hall Effect

173
Fig.12.2. Motion of electrons in a n-type semiconductor
Hall Effect may be explained by reference to Fig. 12.2 which shows the front face of the slab
only.
If v is the velocity of electrons at right angles to the magnetic field, there is a downward force on
each electron of magnitude Bev. This causes the electron current to be deflected in a downward
direction and causes a negative charge to accumulate on the bottom face of the slab (face 1). A
potential difference is therefore established from top to bottom of the specimen with bottom face
negative. This potential difference causes a field EH in the negative y-direction, and so there is a
force of eEH acting in the upward direction on the electron. Equilibrium occurs when
𝑒𝐸 = 𝑒𝐵𝑣 (12.1)
𝐸 = 𝐵𝑣 (12.2)
If 𝐽 is the current density in the x-direction, then
𝐽 = 𝑛𝑒𝑣
Where n is the concentration of current carriers.
Thus, 𝐸 = (12.3)
The Hall Effect is described by means of the Hall coefficient RH, defined in terms of the current
density 𝐽 by the relation
𝐸 =𝑅 𝐽 𝐵
174
Or 𝑅 = (12.4)
𝐽𝑥 𝐵
i.e. 𝑅 = (12.5)
In this case 𝑅 = − (12.6)
Negative sign is used because the electric field developed is in the negative y-direction
1
𝑅 =− =− (12.7)
𝐽𝑥 𝐵 𝑛𝑒
All the three quantities EH, B and Jx can be measured, and so the Hall coefficient and carrier
density n can be found,
1
𝑅 = = (12.8)
𝐽𝑥 𝐵 𝑝𝑒
Where p is the positive hole density.
In-text questions:
1. Q: Explain the term Hall Effect.

2. Q: Show that for a p-type semiconductor the Hall co-efficient R H is given by
1
𝑅 =
𝑝𝑒
.
A: See page 176.
3. Q: Briefly discuss the physical origin of Hall Effect.
A: See page 176.
11.1.2. Determination of the Hall Co-efficient

The Hall coefficient is determined by measuring the Hall voltage that generates the Hall field. If
VH is the Hall voltage across the sample of thickness t, then
𝑉 =𝐸 𝑡
Substituting for 𝐸 from equation (12.4), one gets
𝑉 = 𝐵𝑡𝑅 𝐽 (12.9)
If b is the width of the sample, then its cross-section will be bt and the current density
175
𝐽 = 𝐼 ⁄𝑏𝑡
Thus 𝑉 = =
Hence, 𝑅 = (12.10)
Note that the polarity of 𝑉 will be opposite for n-type and p-type semiconductors.
In-text question:
1. Q: Briefly explain how the Hall co-efficient can be determined.
12.1.3 Applications of Hall Effect

(i) Determination of semiconductor type:
For an n-type semiconductor the Hall coefficient is negative whereas for a p-type
semiconductor it is positive. Thus, the sign of the Hall coefficient can be used to
determine whether a given semiconductor is n or p-type.
(ii) Calculation of Carrier Concentration:

The Hall voltage VH is measured as usual by placing the two probes at the centres of
the top and bottom faces of the sample. If the magnetic flux density is B Wb/m2 , then
1
𝑛=
𝑒𝑅
1 𝑉 𝑏
𝑤ℎ𝑒𝑟𝑒 𝑅 = =
𝑛𝑒 𝐼 𝐵
(iii) Determination of Mobility:

If the conduction is due to one type of carriers, e.g., electrons, we have
𝜎 = 𝑛𝑒𝜇
𝜇 = 𝜎⁄𝑛𝑒 = 𝜎𝑅
𝜇 = 𝜎(𝑉 𝐵⁄𝐼 𝐵)
i.e., knowing 𝜎, the mobility 𝜇 can be determined.
176
(iv) Measurement of Magnetic Flux Density:
Since Hall voltage 𝑉 is proportional to the magnetic flux density B for a given current
𝐼 through a sample, the Hall effect can be used as the basis for the design of a
magnetic flux density meter.
(v) Measurement of Power in an Electromagnetic Wave:

In an electromagnetic wave in free space the magnetic field H and the electric field E
are at right angles. Thus, if a semiconductor sample is placed parallel to E it will derive
a current I in the semiconductor. The semiconductor is subjected simultaneously to
transverse magnetic field H producing a Hall voltage across the sample. The Hall
voltage will be proportional to the product E and H, i.e. to the magnitude of the
poynting vector of the electromagnetic wave. Thus, the Hall effect can be used to
determine the power flow in an electromagnetic wave.
(vi) Hall Effect Multiplier:

If the magnetic flux density B is produced by passing a current I’ through an air-core
coil, B will be proportional to I’. The Hall voltage is thus proportional to the product of
I and I’. this forms the basis of the mutiplier.
In-text questions:
1. Q: Explain how Hall Effect is used to determine the mobility of Charge carriers.
A: See page 176.
2. Q: Give the principle of Hall Effect multiplier.
A: See page 177.
3. Q: List the applications of Hall Effect.
177
1. The definition of Hall Effect and how to determine Hall voltage and Hall coefficient.
2. That the Hall coefficient is related to charge mobility and conductivity of semiconductors.
3. The different applications of the Hall Effect.

SAQ 12.1
A semiconducting crystal of length 0.012m, width 0.0005m, and thickness 0.001m is placed in
magnetic field, B=1Wb.m2 and the current passed through it is 20mA and Ey = 7.4mV.Calculate
the Hall Co-efficient.
SAQ 12.2
A semiconducting crystal in the form of a thin rectangular wafer is placed with its plane normal to
a magnetic induction, B while a current is made to flow along its length under e.m.f, E x. Show that
an e.m.f. Ey develops across its width.
SAQ 12.3
The Hall coefficient of a certain specimen of silicon was found to be -7.35 x 10- 5 m3C-1 from 100 to
400K. Is this semiconductor intrinsic or extrinsic at room temperature, and is it n-type or p-type? The
electrical conductivity at room temperature was found to be 200Ω-1m-1. Calculate the density and
mobility of charge carriers at room temperature.
SAQ 12.4
The resistivity of a doped silicon sample is 8.9 x 10 -3Ω-m. The Hall coefficient was measured to be 3.6
x 10-4m3/C. Assuming single carrier conduction, find the mobility and density of the charge carriers.
Richard Turton, (2000), The Physics of Solids, Oxford University Press.
178
Study Session 13 Superconductivity
Introduction
The electrical resistivity of all metals and alloys decreases when they are cooled. When the
temperature is lowered, the thermal vibrations of the atoms decrease and the conduction electrons
are less frequently scattered. The decrease of resistance is linear down to a temperature equal to
one-third of the characteristic Debye temperature of the material, but below this, the resistance
decreases less rapidly as the temperature falls .

13.1 Explain Superconductivity.
13.2 Explain Meissner effect and the BCS theory.
13.3 Explain the effect of a magnetic field on superconductors.
13.1 Main Content

13.1.1 A survey of superconductivity
To understand anything about superconductivity, we have to know something about ordinary
conductivity. All ordinary materials resist the flow of electricity to a lesser or greater extent.
Electricity or electrical current consists of the movement of electrons through a material. In order
to make electrons flow, we need to supply energy. Electrons in a metal behave like water in a
hose. The water would not flow unless we supply some pressure. The amount of pressure it takes
depends on the kind of hose we use.
Superconductors are extraordinary because they alone are immune to the effects of joule heating.
It takes no energy at all to make current flow in conductor, and no energy is lost to friction to
sustain the current either. Its electrical resistance is precisely zero. Superconductivity was first
discovered by the Dutch physicist Heike Kamerlingh Onnes in 1911. Not all pure metals have
been found to be superconductors; for example; copper, iron, sodium have not shown
superconductivity down to the lowest temperature to which they have been so far cooled.
Superconductivity can be exhibited in a large number of alloys e.g Bi-Pd. Also, superconductivity
can be shown by conductors which are not metals in the ordinary sense; e.g the semiconducting
mixed oxide of barium, lead, bismuth, and conducting polymer.
179
13.2 An account of the mechanism of superconductivity
The theory of superconduction has been developed over many years, but Bardeen, Cooper, and
Schrieffer (BCS) theory in 1957, published the first really satisfactory account of the mechanism.
They show that superconductivity occurs when a special state of affairs exists between the
conduction electrons.
Two electrons in free space will be mutually repelled by the Coulomb force between, but, in the
solid state, the force between the two electrons will be modified by the interaction of the electrons
within the crystal lattice. In certain substances, the lattice interaction is so great that simple
repulsive force become modified into an attractive force binding certain electrons together into
what are called Cooper pairs. For two electrons to become bound into a Cooper pair, they must be
in equilibrium condition, i.e., with no current flowing, have equal and opposite momenta.
Some of the important properties of superconductors are the following:
1. The current in the superconconductors persists for a very long time.
2. Superconductivity has normally been observed only for those metallic substances for which the
number of valence electrons Z lies between 2 and 8.
3. In all cases involving transition metals, the variation of T c with the number of valence electrons
shows sharp maxima for Z = 3, 5, 7.
4. For a given value of Z, certain crystal structures seem more favourable than others. For
example, β-tungsten and α-manganese structure are conductive to the phenomenon of
superconductivity.
5. The magnetic field does not penetrate into the body of the superconductor. The property known
as the Meissner effect, is the fundamental characterization of superconductivity. However,
when the magnetic field B is greater than critical value Bc(T), the superconductor becomes a
normal conductor [Bc(T) is zero at T = Tc ] and has the largest value at T = 0.
6. Superconductivity occurs in materials having high normal resisitivities.
7. When the current through the superconductor is increased beyond a critical value Ic(T), the
superconductor again becomes a normal conductor.
180
8. The specific heat of the material shows an abrupt change at T = Tc, jumping to a large value for
T < Tc .
If one observes the total list of superconducting materials, the general features to be noted are:
 Monovalent metals are not generally superconductors.

 Ferromagnetic and antiferromagnetic materials are not superconductors.
 Good conductors at room temperature are not superconductors, and superconducting materials
are not very good conductors at room temperature as the normal metals.
 Amorphous thin films of Be, Bi, and Fe show superconductivity.
 Bismuth, antimony and tellurium become superconducting under high pressure. However,
there are exceptions.
In-text questions:
1. Q: Define Superconductivity.
2. Q: List some of the important properties of superconductors.

3. Q: What are Cooper pairs?

A: See page 180.
13. 3 Effects of Magnetic Field

The superconducting state of a metal exists only in a particular range of temperature and field
strength. The condition for superconducting state to exist in the metal is that some combination of
temperature and field strength should be less than a critical value. Superconductivity will
disappear if the temperature of the specimen is raised above its Tc , or if a sufficiently strong
magnetic field is employed. The curves are nearly parabolic and can be represented by the
relation.
𝐻 =𝐻 1− (13.1)
181
where Hc is the maximum critical field strength at the temperature, T, H0 is the maximum critical
field strength occurring at absolute zero, and Tc is the critical temperature, the highest
temperature for superconductivity. Thus the above equation defines a curve which divides the
normal region of the field temperature diagram of the metal from the superconducting region.
13.4 Critical Current
The magnetic field which causes a superconductor to become normal from a superconducting
state need not necessarily be an external applied field. It may arise as a result of electric current
flow in the conductor. The minimum current that can be passed in a sample without destroying its
superconductivity is called critical current Ic. If a wire of radius r of a type I superconductor
carries a current I, there is a surface magnetic field, 𝐻 = 1⁄2𝜋𝑟 associated with the current. If HI
exceeds Hc, the material will go normal. If in addition, a transverse magnetic field H is applied to
the wire, the condition for the transition to the normal state at the surface is that the sum of the
applied field and the field due to the current should be equal to the critical field. Thus, we have
𝐻 = 𝐻 + 2𝐻
𝐼
𝐻 = = 𝐻 − 2𝐻
2𝜋𝑟
𝐼 = 2𝜋𝑟(𝐻 − 2𝐻) (13.2)
This is called Silsbee’s rule. The critical current Ic will decrease linearly with increase of applied
field until it reaches zero at 𝐻 = 𝐻 ⁄2 . If the applied field is zero, 𝐼 = 2𝜋𝑟𝐻 .
In-text questions:
1. Q: Briefly explain the effect of magnetic field on superconductors.
A: See page 181.
2. Q: What is meant by the term critical current?
A: See page 182.
3. Q: Describe the Silsbee rule.
A: See page 180.

182
13.5 Flux Exclusion: The Meissner Effect
It was assumed that the effect of a magnetic field on a superconductor would be as that in a metal.
However, in 1933 Meissner and Ochsenfeld measured the flux distribution outside tin and lead
specimens which had been cooled below their transition temperatures while in a magnetic field.
They found that at their transition temperatures, the specimens spontaneously became perfectly
diamagnetic, cancelling all flux inside even though they had been cooled in a magnetic field.
This experiment was the first to demonstrate that superconductors are something more than
materials which are perfectly conducting; they have an additional property that are merely
resistanceless metal would not possess: a metal in the superconducting state never allows a
magnetic flux density to exist in its interior. That is to say, inside a superconducting metal we
always have
B=0
whereas inside a merely resistanceless metal there may or may not be a flux density, depending on
the circumstances. When a superconductor is cooled in a weak magnetic field, at the transition
temperature, persistent currents arise on the surface and circulate so as to cancel the flux density
inside, the same way as a magnetic field is applied after the metal has been cooled. This effect,
whereby a superconductor never has a flux density even when in applied magnetic field, is called
Meissner effect.

1. The definition and characteristics of superconductors.
2. The Meissner effect.
3. The BCS theory.

183
SAQ 13.1
What are superconductors? Mention the important property changes that occur in materials when
they change from normal to superconducting state. Give some examples of practical uses that
exploit the above property change.
SAQ 13.2
Name two superconducting materials. Define the critical magnetic field and derive the
thermodynamic relation connecting the critical magnetic field and the transition temperature.
SAQ 13.3
What is BCS theory of superconductivity? How is superconductivity affected by magnetic field?
SAQ 13.4
Calculate the critical current density for 1mm diameter wire of lead at
(a) 4.2K
(b) 7K.
A parabolic dependence of Hc upon T may be assumed. Given: Tc for lead is 7.18 K and H0 for lead is
6.5 x 104 ampere/metre.
Richard Turton, (2000), The Physics of Solids, Oxford University Press.
184
SOLUTIONS TO SELF ASSESSMENT QUESTIONS
STUDY SESSION 1
SELF ASSESMENT QUESTIONS
1.1. Please see text.
STUDY SESSION 2
2.4. 6.9eV or 667 x 103 kJ/kmol.
STUDY SESSION 3
3.3 Please see text.

STUDY SESSION 4

185
STUDY SESSION 5
5.1 The total energy of the wave is the sum of the kinetic energy E and the potential energy E .
kin pot
The kinetic energy is the sum of the kinetic energies of all atoms, i.e.
1 𝑑𝑢
𝐸 = 𝑀
2 𝑑𝑡
where M is the mass of atoms and is the velocity of n-th atom. The potential energy is the
potential energy of all the “springs” connecting atoms. For two atoms this energy is the same as
that for a harmonic oscillator, i.e. 𝐶𝑥 , where C is interatomic force constant and x is the change
in distance between the atoms from the equilibrium distance. Therefore, for atoms n and n+1
having displacements 𝑢 and 𝑢 respectively this energy is 𝐶(𝑢 − 𝑢 ) . The total potential
energy is given by)
1
𝐸 = 𝐶 (𝑢 −𝑢 )
2
The total energy of the wave is therefore
𝐸=𝐸 +𝐸 = 𝑀 ∑ + 𝐶 ∑ (𝑢 −𝑢 ) .

STUDY SESSION 6

6.3.Please see text.
186
STUDY SESSION 7

7.2.(a) Please see text.
(b) Please see text.
STUDY SESSION 8
8.1. Please see text

8.3 Please see text
STUDY SESSION 9

STUDY SESSION 10
1.(i) Please see text

(ii) Please see text
(iii) Please see text
(iv) Please see text
2. (i) Please see text
(ii) Please see text
187
3. (i) Please see text
(ii) Ec – E’f =0.16eV ; 𝜎 = 3120𝛺 𝑚

STUDY SESSION 11
11.1 𝑛 = 10 /𝑚 , 𝑝 = 9.6 × 10 /𝑚
11.2 0.07𝑚 𝑉 𝑠
11.3 0.128
11.4 1015
STUDY SESSION 12
12.2 Please see text
12.3 8.5 × 10 /𝑚 , 1.47 × 10 𝑚 𝑉 𝑠
12.4 0.0344 𝑚 𝑉 𝑠 , 2.04 × 10 /𝑚
STUDY SESSION 13

13.4 1.71 × 10 𝐴/𝑚
188

Lecture Notes - Solid State Physics

Uploaded by

Copyright:

Available Formats

Lecture Notes - Solid State Physics

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lecture Notes - Solid State Physics

Uploaded by

Copyright:

Available Formats

PHS 319

(SOLID STATE PHYSICS I)

Learning Outcomes for Study Session 1

Figure 1.1 Crystal Structure showing the positions of the atoms

Figure 1.3 Position vector of lattice sites in 2-D

Some important crystal structure terms are defined below:

1.1.1. Wigner-Seitz unit cell:

The primitive cell may be chosen as shown in Figure 1.4.

Figure 1.4 Wigner-Seitz primitive cell

Figure 1.5 Showing primitive translation vectors for a lattice site

1.2 Crystal symmetry

Figure 1.6a Bravais lattice types in 2-D

Figures showing the bcc lattice crystal structure

Face-centered cubic (fcc) lattice:

Figures showing the fcc lattice crystal structure

Hexagonal closed-packed (hcp) lattice:

1. Q: Define the term Coordination number.

1.3.1 Index system for crystal directions and planes

Self- Assessment Questions for Study Session 1

With reference to crystals, state what you understand by the terms:

Richard Turton (2000). The Physics of Solids: Oxford University Press.

Learning Outcomes for Study Session 2

2.1 Main Content

2.1.1 Interatomic forces

1. Q: Discuss the term “cohesive energy”.

2.1.2 Van der Waals (molecular) bonding:

2.1.3 Ionic bonding

Figure 2.2 Structure of the NaCl crystal

Figure 2.3 One-dimensional lattice of ions

where we took into account the logarithm expansion into series

2.1.4 Covalent bonding

Figure 2.4 Molecular orbital and charge distribution in covalent bonds

Figure 2.5 Energy as a function of Interatomic distance in the two orbitals

Figure 2.6 Diamond Structure

A: 6.9eV or 667 x 103 kJ/kmol.

1. Q: What are the distinguishing characteristics of metallic bonding?

Self- Assessment Questions for Study Session 2

How is a covalent bond formed in a hydrogen molecule?

Describe the salient features of covalent and metallic bonded crystals

Learning Outcomes for Study Session 2

3.1. Main Content

3.1.1 Bragg Law

Figure 3.1 X-ray reflected from planes

The Bragg law is greatly oversimplified (but it works). It

(ii) neglects differences in scattering from different atoms;

(iii) neglects the distribution of charge around atoms.

A: See pages 35-36

3.1.2. Diffraction condition and reciprocal lattice

where G is, by definition, such a vector for which

using primitive vectors

where V = 𝒂 ∙ (𝒂 × 𝒂 ) is the volume of the unit cell, so that

𝒃 ∙ 𝒂 = 2πδij (i,j=1,2,3) (3.9)

where δ is the Kronecker symbol defined by

δ =0, if i≠j (3.10)

Indeed, since T = 𝑛 𝒂 + 𝑛 𝒂 + 𝑛 𝒂 we obtain

G∙T = 2𝜋(𝑛 𝑚 + 𝑛 𝑚 + 𝑛 𝑚 ), (3.11)

Reciprocal lattice types for some 3D lattices:

Direct lattice Reciprocal lattice

It is sometimes more convenient to give a different formulation of the diffraction conduction. In