UNIT 10 BAND THEORY OF SOLIDS

Structure
10.1 Introduction
Objectives
10.2 An Electron in a Periodic Potential
Kronig-Penney Model
Energy Bands and Motion of Electrons
Classification of Solids in Metal, Insulator and Semiconductor
10.3 Energy Bands and Brillouin Zones
10.4 Hall Effect
10.5 Summary
10.6 Terminal Questions
10.7 Solutions and Answers

10.1 INTRODUCTION
In Unit 9, you have studied the free electron theory of metals. One of the basic
assumptions of this theory is that positively charged ions in a metal crystal do not
influence the motion of free electrons. Obviously such an assumption over-simplifies
the actual situation in crystalline solids including metals. Still, the theory gives good
insight into wide range of metallic properties such as electrical conduction, heat
capacity and thermal conductivity. Some important questions, however, remained
unanswered. For example, free electron theory predicts that electrical conductivity is
directly proportional to the number density of free electrons implying thereby that
bivalent and trivalent metals should have larger electrical conductivities than those of
monovalent metals. But this inference is not supported by experimental observations.
Further, it also does not explain why some solids are conductors and some others are
not.
Most of such limitations of free electron theory were overcome by considering a more
realistic model of solid. It was proposed that in a solid, electrons move under the
influence of a periodic potential due to ions arranged along a periodic lattice. The
energy spectrum of such electrons consists of allowed and forbidden energy bands and
the theory developed on the basis of this model is called Band Theory of Solids. For
ease in mathematical analysis, Kronig and Penney proposed that one-dimensional
periodic potential of a crystal can be considered equivalent to a series of potential
wells. In Sec.10.2 you will learn the Kronig-Penney model and the consequences of
energy bands on the motion of electrons in a solid. You will also learn how band
theory leads to the concept of hole - a concept of utmost importance for
understanding electrical properties of a group of solids called semiconductors. The
existence of energy bands provides a basis to categorise solids as metal, insulator and
semiconductor. In Unit 2, you have learnt about Brillouin zone. In Sec. 10.3, you will
learn how is this geometrical concept useful in the study of band theory of solids.
Determination of electron transport parameters such as charge carrier concentration is
necessary for making devices. These parameters of metals as well as semiconductors
can be determined on the basis of the Hall effect about which you will learn in
Sec. 10.4.
One of the important successes of the band theory of solids was theoretical
understanding of semiconductors and their physical properties which you will learn in
the next unit.
d
 Band Theory of Solids
Objectives
After studying this unit, you should be able to:
I
list the limitations of the free electron theory;
state Bloch theorem;
discuss the results of Kronig-Penney model and formation of energy bands;
8 explain how Kronig-Penney model describes the motion of electrons in metals
and provides the basis for categorisation of solids;
relate Brillouin zone to energy bands; and
8 explain the significance of Hall effect experiment with respect to band theory and
obtain an expression for Hall coefficient.

10.2 AN ELECTRON IN A PERIODIC POTENTIAL

i You may recall from Unit 9 that, according to the free electron theory, a metal crystal
can be considered analogous to an empty box containing free electrons. The lattice
ions do not influence the motion of electrons in any way. That is, electrons move
i
completely randomly all around the volume of the crystal. However, this description
is not enough to understand various properties of solids. From Unit 9, you will recall
that the separation between neighbouring atoms in a metal is less than the sum of their
radii and it is reasonable to assume that orbitals of electrons belonging to these atoms
overlap. As a results, electrons find themselves in potential field of lattice ions. That
is, electrons experience a potential field.
To appreciate the nature of potential field experienced by electrons in a solid, let us
first understand the nature of potential field of an isolated positively charged ion for
an electron. Refer to Fig. 10.1a which depicts the variation of potential energy of a
conduction electron with respect to its separation from a positively charged ion. The
negative value of the potential energy indicates that the force between the ion and
electron is attractive. Fig. 10.1b shows the potential energy curve for an electron
under the influence of two ions. Note that the resultant curve in between the ions gets
modified. Now, recall that ions in a crystalline solid are arranged along a periodic
lattice. Therefore, in a one-dimensional (1-D) periodic crystal lattice, the potential
energy of electron will be due to a series of equidistant ions as shown in Fig.lO.1~.
The potential experienced by an electron in such an arrangement of ions is periodic in
nature.

Fig.lO.1: Potential energy of an electron due to a) an ion; b) two ions separated by a distance a;
and c) series of equidistant ions representing I-D periodic crystal

In the previous unit we solved the Schrodinger equation to investigate the behaviour
of free electrons in a metal where it was assumed that ions do not influence the motion
of electrons, i.e., electrons move in a constant (zero) potential. To understand the
behaviour of electrons in a periodic potential, we should solve the same equation
Electronic Properties incorporating periodic nature of potential. The 1-D Schrodinger equation for an
electron in a periodic potential is given by

where V (x) is the periodic potential experienced by an electron in the crystal and ~ ( x )
and E are respectively the wavefunction and the energy of an electron.
For simplicity we have assumed a 1-D periodic lattice with lattice constant a. Thus,
the periodicity of potential V (x) can be expressed as

Eq. (10.2)means that the V(x+a)=V(x). (10.2)

value of potential will be
same at any two points To proceed further, we must know how to solve Eq. (10.1) with this kind of potential.
along the x-axis separated
This problem was first tackled by F. Bloch who showed that the general solution of

I
by distance the
constant. This is because of Schrlidinger equation with a periodic potential is
the perlodic arrangement of
crystal ions along x-axis.
w (x) = uk( 4 e f l k x (10.3)

where uk(x) is a periodic function with periodicity of the lattice: 1

( 3
In Eq. (10.3), k = - is the wave number associated with deBroglie wavelength 3t
of electron. The wavefunction given by Eq. (10.3) is called Bloch function and the
result that ~ ( xcan
) be expressed as aplane wave modulated by aperiodicfinction is ,
called the Bloch theorem. Further, the Bloch function is modulated by the periodicity
of the lattice because

fika
= W(X) e (10.4)

using Eqs. (10.3) and (10.3a).

Without proving the Bloch theorem, we shall use its results to investigate the effect of
periodic potential on the energy of electrons in a solid. However, for potential field
shown in Fig. lO.lc, it is difficult to solve the Schrodinger equation even with the help
of Bloch theorem. A way out of this problem was suggested by Kronig and Penney.
They proposed a simplified version of the actual potential experienced by an electron
and solved the Schrijdinger equation. You will learn the details of Kronig-Penney
model now.

10.2.1 Kronig-Penney Model

Kronig and Penney proposed that the potential experienced by an electron in a 1-D
crystalline solid can be represented by a periodic array of potential wells as shown in
Fig. 10.2.
 Band Theury of Solids

v (4

-
C
-
-
C
.-
0
.-0
%? F
K I
Y
II
I

Fig.10.2: Potential well representation of the periodic potential experienced by an electron due to
ions in a 1-D crystal of solid

The nature of potential well representation can be understood by comparing Fig.(l0.2)

with Fig.(lO. lc) which represents potential of an electron in a I -D crystal. The
characteristic features of Fig.(lO.lc) are:

a) when electron is in the vicinity of an ion, its potential energy has a finite negative
value (say, - Yo); and
b) when electron is away from an ion (that is, when it is between two ions, its
potential energy is zero.

You know that taking zero of potential energy is a matter of convenience; important
thing is the relative value of potential energy between two point. In this context,

i Fig. (10.1~)shows that when an electron is in the vicinity of an ion, potential energy is
lower than its value when electron is in-between the ions. This feature must be

I incorporated in any representation of the potential of an electron in a periodic crystal.

Now, refer to Fig.lO.2. First of all, note that the origin of coordinate system has been
I shifted to - Vo along the potential axis. Each potential well (regions I) represents the
1 potential of an electron in the vicinity of an ion. Each of these regions is of width a
1 and the value of potential is zero. Two consecutive potential wells are separated by a
1 distance b (regions 11) and the value of the potential in each of these regions is Vo.
Region 11 in Fig. 10.2 corresponds to the situation when electron is in-between the
ions. Thus, in Kronig-Penney model, we find that when electron is in the vicinity of
an ion, its potential is lower than when it is in-between ions. There is complete
correspondence between electron potential in a periodic crystal (Fig. 10. lc) and its
potential wells representation (Fig. 10.2) except for the fact that, in the latter, the
lattice constant is (a + b) instead of a. The potential of electron for the two regions
can, therefore, be expressed as

Vo for region I1 (-6 < x < 0)

Thus, for the Kronig-Penney model, we write Schrodinger equation for two distinct
regions separately:
Electronic Properties

Now, let us assume that the energy E of electron is less than the height of the potential
barrier, Vo. Note that this assumption is quite justified because at room temperature,
electrons do not acquire sufficient thermal energy to escape from the crystal. Next, we
introduce two real quantities a and p by defining

and
p2 = 2m(Vo - E )
A

Since Vo> E, p will be a positive quantity. In terms of a and P, Eqs. (10.6) and (10.7)
respectively take the form

and

According to Bloch theorem, the solution of these differential equations is given by

w (XI= u k ( x ) ei k x (10.12)

You will not be asked To solve Eqs. (1 0.10) and (1 0.1 I), we should use the expression for ~ ( x as
) given by
questions in examination on Eq. (10.12) and apply appropriate boundary conditions. However, this involves
the contents presented in the complex mathematics. For details, see Appendix at the end of the unit. Here it is
Appendix.
sufficient to quote the result since we are interested in the physical significance of the
result obtained by solving these equations. It is found that solutions of these equations
can exist only if the following condition is satisfied:

p2 - a2
sinh p b sin aa + cosh pb cos aa = cos k ( a + b). (1 0.13)
2 4

To simplify above expression, Kronig-Penney made an assumption regarding the

nature of potential encountered by electrons: as Vo + m, b + 0 in such a way that Vo
times bywhich represents the area of the potential barrier, is finite. Physically, it
means that the electrons are not completely free. Further, this assumption implies that

sinh pb + pb

and
cosh Pb + 1.

Then Eq. (1 0.13) reduces to

p2 - a2b s i n a a + c o s a a = cos ka.

2a

42
Further, by combining Eqs. (10.8) and (10.9), we can write (see margin remark) Band Theory of Solids

where

Since Vobrepresents the strength of potential barrier for electron (Fig. l0.2),
Since Vo >> E, we can neglect the
Eq. (10.16) implies that P is a measure of the strength of attraction between electron second term the brackets to
and an ion. In other words, the value of P is a measure of how strongly an electron is obtain
bound in a crystal.

Using the result contained in Eq. (10.15) in Eq. (10.14), we get

sin a a
P- + cos a a = cos k a
a a

You may like to know: What does Eq. (10.17) signify? You should realise that the
condition expressed by Eq. (10.1 7) must be satisfied if the wavefunction y (given by
Eq. (10.12)) is to represent an electron in a periodic potential. Further, it can also be
used for obtaining useful information about the energy spectrum of electrons. Now,
before proceeding further, you should answer an SAQ to fix these ideas.

SAQ 1

a) What is the origin of the periodic potential field experienced by electrons in a Spend
crystalline solid? 3 min.

b) If the wavefunction y(x) is to represent an electron in a crystalline solid, what

should be its nature?

To further appreciate the significance of Eq. (1 0.1 7) for energy of an electron, it is

instructive to plot
sin aa
1 37t
+ cos a a vs, a a . Such a plot for P = -
2
is shown in

Fig.(l0.3). Note from Eq. (1 0.17) that irrespective of the value of ka, the right hand
side can take values between +I and -1 only. This imposes limits on values the LHS:
sin a a
only those values of are allowed which are in the range + 1 to -1 .

These limits are shown by the horizontal solid lines parallel to aa-axis in Fig. (10.3).

For plot shown in Fig.lO.3,

value of aa has been taken
as multiples of x because it
simplifies the tabulation of
data as well as gives better
insight about the energy
spectrum of electron.

aa for P = 3 ~ 1 2
I
I1 Electronic Properties Now, before we discuss inferences flowing from the plot in Fig.(] 0.3), you should
convince yourself about its nature by solving the following SAQ.

SAQ 2

Spend
5 min. [a
: 1
Tabulate values of -sin aa + cos aa for different values of a a between 0 and 3x

3x
to show allowed and forbidden values of LHS of Eq. (10.17). Take P = -.
2

57t
Solving SAQ 2, you must have noted that for some value of a a , such as 0 and -,
4

of Eq. (I 0.17).
1
a a + cos aa is greater than 1 1/ and cannot be equal to cos ka, RHS

From Fig. 10.3, you will also note that:

For certain values of a a , the curve crosses &1 along

implies that the corresponding energy values are not permissible to electrons.
Therefore, we can conclude that the energy spectrum of electron consists of
groups or bands of energy levels comprising allowed energy bands and
forbidden energy bands.

As the value of a a and hence the energy of an electron increases, the width of
the allowed energy bands increases. This is because the value of the first term in
LHS of Eq. (10.17) decreases as a a increases. You should convince yourself
about this conclusion by solving the following SAQ.

SAQ 3
Spend
5 min. Using Eq. (10.17), show that the width of allowed energy bands increases as energy of
electron is increased.

Fig. 10.3 has been drawn for a fixed value of P (= 3x12). You may ask: Does the
value of P affect band width? And if so, how? From Eq. (10.17), it can be seen
that as P increases, the width of a particular allowed band decreases. You should ,
convince yourself about this statement by calculating the value of
--sin a a + cosaa for the first allowed band for three increasing

3x 5x 7x
values of P such as -, - and -. Since P is a measure of the strength of
2 2 2
potential barrier, larger values of P implies that an electron is strongly bound to its
parent atom. That is, condition that P is large represents the physical situation that
an electron is in the constant potential; P small implies that electron is a
completely free particle. It is instructive to obtain expressions for energy of an
electron for these two cases. You will learn to do so now.

i) P + co : In this case, Eq. (10.17) can be rewritten as

aa(coska-cosaa)
sinaa=
P
For P + co, sin a a + 0. This implies that
 Band Theory of Solids

Comparing this expression for a2with that given by Eq. (10.8), we obtain

Do you recognise this expression? It is same as Eq. (9.13) of Unit 9 which

gives energy of an electron moving in a constant potential. This means that
the condition P + co leads us to the results of Sommerfeld free electron
model. It is understandable because very large value of P implies that the
strength of the potential barrier (P =rob)is so strong that electron cannot
escape from an isolated potential well. The energy spectrum for P + c~ is
shown in the right portion of Fig. 10.4. Note that it consists of discrete energy
levels which is consistent with the energy spectrum of electron as per the
Sommerfeld model. Thus, we may conclude that the Sommerfeld model is a
special case of the Kronig-Penney model, that is, fhe band theory of solids.

ii) P+ 0: In this case the first term on the left hand side of Eq. (10.17) will
vanish and the expression reduces to

cos aa = cos ka.

From this equality, it readily follows that

Fig.lO.4: Energy spectrum of

electrons for
Again, comparing this value of a*with that given in Eq. (10.8), we get different strengths of
the potential field in
a crystalline solid
(10.19)

wherep = A k is the momentum of electron.

This expression gives the energy of a free electron and all energies are
allowed as shown in the left portion of Fig. 10.4. This situation corresponds to
the Drude-Lorentz model. Thus, Drude-Lorentz and Sommerfeld models are
particular cases of Kronig-Penney model. You would indeed agree that this
feature of band theory is a significant improvement over the free electron
theory.

Yet another important conclusion relates to dispersion relation (E vs. k curve) for
electrons in a periodic potential. It is obvious from Eq. (10.17) that for a given
value of a a (and hence energy E), the RHS can take only one value. But, cosine
function is an even periodic function. Therefore, cos ka will have the same value
for f ka as well as for f ka f 2nrc; n = 1,2,3,. ... It suggests that the energy of an
electron is an even periodic function of k with period 2rc. Further, to obtain
dispersion curve, you should note that discontinuities in the energy spectrum will
 Electronic Properties c o s k a = f 1 = c o s ( f nsc)
or

k a = f nsc.

That is, energy spectrum will be discontinuous at

n = f l , +2, f3,... (10.20)

Let us pause for a while and ask: What does this result tell us? It suggests that the

(a) (b)
Fig.10.5: a ) Plot of energy against wave number of an electron in a 1 - D crystal; and
b ) corresponding energy bands

Energy States in a Band

At this point, you may ask: Are the energy states in an allowed energy band
v(X +L ) = U k ( X+ L ) ~ ~ ( X +=Lv)( X ) continuous? If not, why? It is observed that though very closely separated, the energy
levels in a band are discrete. To know the number of energy states in a band, let us
consider a 1-D crystal of length L having N unit cells. L and N are connected through
:. uk(x)elk = u k ( x + ~ ) e ' ~ ( ' + ~ )
the relation

L=N(a+b)
(': u k ( x )= u k ( x + L ) )
where (a + b) is the distance between two neighbouring ions. The cyclic boundary
This implies that condition on the wavefunction ~ ( x is)

and it holds for

w (x + L) = W (XI. (10.21)

k L a 2 n n ; n = 0,f 1, f 2 , s . . But ~ ( x is) Bloch function given by Eq. (10.3) as

w ( x ) = u ~ ( x e) i k x .

On applying the boundary condition (Eq. (10.21)) on ~ ( x )you

, will obtain (see
margin remark)
 Band Theory of Solids

What does this result signify? It indicates that k can take discrete set of values
+
0, +2xIL, 4n:lL, + 6xlL, .. . ., etc, and, therefore, energy E is quantized. If length L of
the I-D crystal is large, separation between energy levels in a band becomes small,
and they form a quasi-continuous band. For macroscopic specimens, this is indeed the
case. Now suppose we take an arbitratry interval dk in k-space and wish to know the
number of energy states whose k-values lie in this interval. Under the assumption that
L is large and energy levels in a band is quasi-continuous, we can write the number of
possible energy states in the interval dk as

Hence, total number of energy states in the first allowed energy band (for k values in
range - nla to + xla) can be obtained by integrating Eq. (10.23);

Thus,

If the separation between the potential wells (Fig.] 0.2) is very-very small, that is, You should note that width of
each band is equal to 2da.
b = 0, we can write (a + b) = a. Thus, we get Therefore, the result obtained
for first allowed band has been
N(a+b)=Na=L (1 0.25) generalised for all allowed
bands in a crystal.
Comparing Eqs. (10.24) and (10.25), we obtain

That is, the total number of energy states (n) for an electron in an allowed band
of a crystalline solid is equal to the number of unit cells. Further, since each energy
state can accommodate two electrons - one with spin up and another with spin
down - maximum number of electrons which can occupy an allowed energy band is
2N. So, when the number of electrons in an energy band is equal to twice the number
of unit cells in the crystal, the band is completelyfilled. This conclusion has far
reaching consequences. You will learn later in this unit that it enables us to distinguish
metal from insulator and semiconductor.
So far, you have studied about the allowed and forbidden energy bands for an electron
in the periodic potential of a crystalline solid. You can also explain how many
electrons can occupy an allowed energy band and at what values of the wave number
disconti~iuitywould occur in the energy spectrum. In addition, the band theory also
provides a completely new perspective on some of the physical parameters such as
velocity and Inass of an electron in a crystalline solid. You will learn it now.

10.2.2 Energy Bands and Motion of Electrons

You may recall from Unit 9 that energy of free electrons is proportional to the square
of the wave number. But the dispersion curve (E vs. k curve) predicted by the band
 Electronic Properties theory do not show parabolic relationship between E and k, particularly near the
boundaries of bands. You may note in Fig. 10.5 that at the boundaries, the E vs. k
curve deviates from the parabolic nature as predicted by free electron theory. This
deviation has important implications for the microscopic parameters of electrons. In
particular, the velocity and mass of an electron changes significantly. Let us see how.

Velocity
You know that the group velocity of a wave is defined as

where o and k respectively denote the angular frequency and wave number of the
wave. In this case we shall consider de Broglie wave associated with an electron.
1
I Since energy of electron E = ha,the expression for group velocity in terms of E takes
the form

(a) v = -d=o- d o -=--

dE 1 dE
(10.26)
dk d E ' d k A dk

I
I
I . do 1
since -= -.
I b" I dE A

(b) For a completely free electron, E = A2 k2 I 2m. Using this relation in Eq. (10.26) we
find that

v = ~ ~ [ E ) = ~ = ~ . (10.27)
A dk 2m m m

(c) This result shows that the velocity of a completely free electron is a linear function of
k. To see how this relation is influenced by the band theory, let us consider the region

(1 1 r-1
I
I bound by - nla < k < nla of the E vs. k curve (Fig. 10.5). For this region, the E vs. k
and v vs. k curves for an electron are shown in Figs. l0.6a and b respectively.

-n / a 0 n/a
' On comparing these figures, you will notice that
Flg.10.6: Variation of a) energy; the velocity of an electron becomes zero for k = f nla of the band. This is
b) velocity; and c) understandable because at these points, the energy is independent of the wave
mass of an electron number, that is, (dE/dk) = 0.
with respect to its
wave number in the as the energy increasesfrom k = 0, the velocity increases atfirst linearly, attains
first allowed energy a maximum valuefor k = ko and then decreases to zero at top of the band. The
hand
point at which k = ko is called point of inflection of the E vs. k curve.

These features of the velocity of electrons in a crystalline solid are unique predictions
of the band theory. As per the free electron theory, velocity of an electron increases as
second power of energy.

Effective Mass
You know that in an electric field, electrons experience a force and get accelerated;
stronger the applied field, higher is the acceleration. You may now ask the question:
Does an electron in a crystalline solid also behave similarly? It is not so because, as
you have seen above, the v vs. k curve of an electron in a solid is different from that of
a completely free electron. This is because the (effective) mass of an electron in a
crystalline solidplaced in an electric field is different from its free mass. To
appreciate the concept of effective mass, let us consider an electron of mass m moving
 Band T d r y eC.Solids
under the influence of a periodic potential. The velocity of such an electron is given
/
by Eq. (10.26). Thus, the acceleration can be written as

Using the E vs. k curve, we can obtain (d2 ~ l d2k) . To find variation of the wave
number with time, let us assume that the electric field E acts on an electron for time dt
and during this time interval, it acquires a velocity v. Thus, the work done on the
electron increases its kinetic energy:
I
dE = ( e E ) x ( v dt)

By re-arranging terms, we obtain

Using this result in Eq. (10.28), we get

1 d 2 ~e c
a= (X) [z )
(T) *

On inverting the expression, we find that

According to free electron

theory, the energy of a
completely free electron is
If we compare this result with the equation of motion, force = mass x acceleration, we given by Eq. (10.19):
find that the effective mass, m* of an electron in a crystalline solid placed in an'
electric field is
Thus,
dE
-=- h2k '

dk m
d
---.
2 = ~-
ti2
Note that effective mass of an electron in a crystalline solid depends on the variation
dk2 m
of energy with the wave number. For example, if you use Eq. (10.19) which expresses
relation between energy and wave number of a completely free electron, you will find Substituting this expression for
that m* is equal to m, the rest mass of the electron. The effective mass comes into play (d ' ~ l d k ' )in Eq. (10.30),we
pet,
because we assume that electron is acted upon by the external field & only. h2 fi2
m = =-=m.
The variation of effective mass with wave number is shown Fig.10.6~.You may note
that in the lower half of the energy band, m' is positive. However, in the upper half of
the band, i.e., beyond the point of inflection, m is negative: the electron has a
negative effective mass!You may ask: What is the physical significance of the
P 49
Electronic Properties negative effective mass? You should note that, beyond the point of inflection, velocity
of electron decreases as k increases. That is, the electric field which accelerates the
electron for k < ko, decelerates it for k > h.It is as if the deceleration is caused by the
negative effective mass of the electron. In other words, beyond the point of inflection,
the electron behaves as ifit is apositively chargedparticle (called hole)!The concept
of hole - electron with negative effective mass - has made significant contribution in
our understanding about semiconductors. You will study about these in Unit 11.
Before proceeding further, you should solve the following SAQ.

SAQ 4 j

The energy of an electron in a crystalline solid is related to the wave number k by the i
relation
1
Spend 10 h2k2
3 min. E=
m

Calculate its velocity and effective mass.

10.2.3 Classification of Solids in Metal, Insulator and Semiconductor

You must have observed that electric appliances have ebonite handles but aluminium
or steel base or body. In a physics laboratory, copper wires are used for electrical
connections and tungsten wires are used in resistance boxes and rheostats. In a
transistor, germanium, silicon or their alloys are used. This is because of the varied
responses of these materials to electric field. The ability of a solid to conduct electric
current may be explained in terms of availability of free electrons in it. It means that
free electrons are not readily available in insulators and pure semiconductors. To

t'
address this question, let us once again consider the energy spectrum of electrons in a
I I '~nstallinesolid consisting of allowed and forbidden energy bands.
Refer to Fig.lO.7 which shows the lowest allowed band filled only upto a given value,
say kl (< xla) of k. Let us define a parameterfk, which gives the extent to which an
electron with a given k is free to move under the influence of an applied field, as:

(10.31)
a

"led
You would definitely like to know: What doesfk signify? Note from Eq. (10.31) that if
lowest allowed
band of a solid the effective mass of an electron in a band is large,& will be small. In this situation,
electron is relatively heavy and hence may not participate in electrical conduction at
low field values. That is, when an electricfield is applied, the electrical conduction
will depend upon the number of effectivelypee electrons rather than on the total
number o f f r e electrons in a band. Further, we can express the number of effectively
free electrons as

where the summation extends over all the occupied energy states of the band. If
separation between the energy levels is small, we can replace summation over discrete
levels by a continuous distribution and write

Using Eqs. (10.23) and (10.31) we get

 N eff ==, nh2
kl
~[$]dk
Band Theory of Solids .

I-[= 2mL d E
n h 2 dk k=k,
(10.33)

The factor of 2 has been included because each energy state can accommodate two
electrons. For a completely filled band, we have

k = f nrcla; n = l , 2, 3...

and

dE nn
-=O atk=+_-.
dk a

So, Eq. (10.33) reduces to

Neff = 0 (10.34)

That is, in a completely filled band, the effective number of free electrons is zero.
Physically, you can appreciate this result by referring to v vs. k diagram in Fig.lO.6b.
Corresponding to any velocity v of an electron, there is another electron with velocity
- v so that the net velocity in the energy band is zero. Further, Eq. (10.33) shows that
Nen will be maximum for k value where dEldk has the maximum. From Fig. (10.6),it
is evident that (dEldk) is maximum at the point of inflection. The point a f inflection
refers to nearly half-filled band. Therefore, the effective number of free electrons in a
band will be maximum if it is filled upto this point.
Now, refer to Fig.lO.8a. It shows allowed and forbidden energy bands. You will note
that some of the lower energy kinds are completely filled and Nen is zero. The highest
partially or completely filled band is known as valence band and the next allowed
unfilled band is known as conduction band. Note that only these two bands are o f
interest to us. It is because all other bands remain unaffected at ordinary temperatures
and electrical fields; you should, however, keep in mind that other bands also exist.
'on the basis of these energy band diagrams we differentiate between metal, insulator
and semiconductor.

Fig.10.8: Different configurations of energy spectrum * ,

Suppose that thepth allowed band is partially filled and the @-l)'h and @ +l)thbands
are,respectively completely filled and completely empty.
Electronic Properties According to the band theory, thepth allowed band consists of a large number of
energy levels. Since this band is partially filled, there are empty energy levels in
which electrons can be accommodated. The energy difference between the levels
within a band is very small. Electrons in filled levels can move to empty levels within
the band even if a very small amount of energy is supplied to them. Thus, when an
electric field is applied to a solid specimen having this type of energy spectrum, the
electrons in the partially filled band will gain energy, move to empty higher energy
levels in the same band and are, therefore, available for electrical conduction. Such
solids are called metals. The partially filled band is known as the conduction band.
Now let us consider another type of energy spectrum depicted in Fig.lO.8b. In this
case, thepth band is completely filled but the (p +l)thband is completely empty. The
energy gap (the forbidden band) between these two bands is, say E,. Electrons in the
pt"band can move to the empty ( p + l)thband only when they receive energy more
than or equal to E,. Therefore, if the applied electric field is not strong enough to
supply this much energy, electrical conduction will not take place. The energy gained
by electrons at ordinarily applied electric fields is much less than E,. Thus, solids with
energy spectrum as shown in Fig.lO.8b correspond to insulators.
Lastly, let us consider the third type of energy spectrum, which is shown in Fig.lO.8~.
Comparing this spectrum with the one shown in Fig.l0.8b, you will observe that the
only difference between the two lies in the value of the energy gap E,. You may now
like to know: How does the value of E, influence the behaviour of solids? At absolute
zero, solids with energy spectrum as in Fig.lO.8~also behave as insulators. But, at
temperature T, electrons in the valence band gain energy equal to kBTand if kBT> E,,
it is quite possible that some electrons in the valence band may get excited to the
conduction band. Therefore, at temperatures above absolute zero, the energy spectrum
shown in Fig. 1 0 . 8 ~will contain empty energy levels at the top of the valence band as
well as filled energy levels at the bottom of the conduction band. The availability of
empty levels in the allowed bands facilitates electrical conduction in such solid. Solids
with small energy gap between their valence and conduction bands tend to conduct
electricity as temperature is raised from zero. Such solids are called semiconductors.
Note that at T = 0 K, there is no difference between an insulator and a
semiconductor in so far as the electrical conduction is concerned.
On the basis of above discussions, we can understand a few experimental observations
on electrical conduction in metals. Recall from Unit 9 that conductivity of a metal is
proportional to the concentration of electrons. It implies that divalent and trivalent
metals should have larger values of conductivity than that of monovalent metals. This
prediction does not conform to experimental observation. The explanation of this
variance is provided by band theory which predicts that feasibility of electrical
conduction is guided by a parameter called number of effectively free electrons (NeE)
available for electrical conduction and not on the total number of valence (free)
electrons. From Eq. (10.33), it is obvious that Neffdepends on the nature of energy
bands and which is more or less same for all metals irrespective of their valencies and
hence conductivities of metals are of the same order of magnitude.
You now know that Kronig-Penney model introduced the concept of allowed and
forbidden energy bands which affect the dynamical behaviour of electrons
significantly. One of the salient features of band theory is that it provides a basis for
constructing Brillouin zone pattern of the crystals and thereby helps in the
determination of crystal structure. In the following, you will learn how band theory
predicts the Brillouin zones.

10.3 ENERGY BANDS AND BRILLOUIN ZONES

Refer to Fig.10.5 again. You will note that discontinuities in the energy spectrum of 1
electrons in a crystalline solid occurs at the k values given by iI
so that

2a = ti h. (10.35) You may recall from Unit 4

that Brillouin zone
If a denotes the separation between the planes normal to the direction of propagation boundariesare the reflecting
of electron wave, then Eq. (1 0.35) represents Bragg's law (2d sin 8 = nh) for electrons planes forX-rays. Thus, we
for 8 = n/2. The similarity between discontinuity in energy spectrum and Bragg's can obtain the X-ray
diffraction condition is a consequence of the wave nature of electron. Therefore, difiaction pattern and
Eq. (10.35) implies that for this relation between a and h, electrons are reflected back crystal structure of a solid if
we know its Brillouin zone
and cannot propagate in the crystal giving rise to the discontinuity in the E vs.k curve.

7c
In view of this, we can say that the region bound between k = -- and k = + - on the
n
a a

E vs. k curve (Fig. 10.4) represents the first Brillouin zone. The regions between
n
--
a
271 71 2n
and - - and - and - represents other Brillouin zone and so on. These values of
a a a
k represent the Brillouin zone boundaries because the corresponding energy values are
not permissible to electrons in a periodic potential. Fig.lO.9 shows a few Brillouin
zones in one dimension.

Fig.10.9: First few Brillouin zones for a I-D crystal

In order to construct Brillouin zone patterns using the results of band theory, let us
consider the motion of an electron in a 2-D square crystal. It is convenient to use the
wave numbers kx and k, as coordinate axes for the construction of Brillouin zones.
The space defined by these axes is called k-space, which is equivalent to reciprocal
space because distances in k-space are reciprocal to the distances in crystal. The
condition for energy discontinuity is given by Eq. (10.20):

, , ,

I
a

Thus, the boundaries of the first Brillouin zone for 2-D square lattice crystal will be

where n, and n2 are integers corresponding to kx and k, axes. I

Electronic Properties Thus, to draw the first zone, we must choose nl and n2 such that it leads to
X
k.. + k.. = f -. This is ~ossiblefor two different sets of values of n~and n?:

n l = _ + l , and nz=0.

With these values of nl and n?, Eq. (10.36) gives

ii) nl = O a n d n 2 = f I,weget,

The first Brillouin zone with these values of the coordinates k, and k, will be a square
PQRS as shown in Fig. 10.10.

Fig.lO.lO: First Brillouin zone for a 2-Dsquare lattice crystal

In order to fix your ideas about Brillouin zones, you should answer an SAQ.

Spend SAQ 5
4 min.
Sketch the second Brillouin zone for a 2-D sauare crvstal.

Solving SAQ 5, you have seen that the second Brillouin zone of a 2-D crystal lattice
consists of four separate parts. However, note that the total area of the second zone is
equal to that of the first Brillouin zone. Further, it is also possible to construct
Brillouin zones for real 3-D crystal lattices. We do not include it in this unit to avoid
mathematical complexities. You may recallfrom Unit 3 thatjrst Brillouin zone can
be taken as a unit cell of a crystal lattice in reciprocal space. Therefore, band theory
calculations leading to construction of Brillouin zones provide an analytical tool to
determine crystal structure of a solid.
In Sec. 10.2, you learnt that the effective mass of an electron in a band acquires
negative value if it is near the top of the band, that is, if its wave number is greater '
than ko - wave number at the point of inflection. An electron with negative effective
mass behaves like a positively charged particle having same amount of charge. You
may like to know: Is there any experiment to check this prediction of band theory?
The Hall effect measurement enables us to determine the sign of charge carriers and
check this prediction. Further, the study of Hall effect is also used for determining
carrier concentration in solids which is of crucial importance for designing solid state
devices. In addition, this experiment can be used for determining whether an extrinsic
semiconductor is n-type orp-type. However, in the following, we confine our
discussion of Hall effect for metals only.

10.4 HALL EFFECT

In Hall effect experiment, an electric field Ex is applied to a rectangular parallelepiped
shaped metallic specimen. We take this as the positive x-axis. Let the current density
due to this field be j, (Fig. 10.1la). A magnetic field B, is applied normal to the
specimen. We take this as the positive z-direction.

.(a)

Fig.lO.11: Schematic diagram of Hall effect experiment

You know from the course on Electric and Magnetic Phenomena (PHE-07) that a
moving charge experiences a force due to the applied magnetic field. This force acts
in a direction perpendicular to both the applied magnetic field and the direction of
motion of the charge and deflects it. For the directions of magnetic field and moving
charges (sincej, is along positive x direction, motion of electrons will be along
negative x direction), this deflection will be along the negative y-direction as shown in
Fig. 10.1lb. Electrons, however, cannot move very far and accumulate on one surface,
A of the specimen (Fig. 10.1 lb). On the opposite surface, B, positive charges
(deficiency of electrons) begins to accumulate simultaneously. Due to accumulation of
opposite charges on the opposite surfaces of the specimen, an electric field Ey
develops along negative y-direction which opposes further accumulation of electrons
on the lower surface. This field is called Hall field and, in equilibrium, it counter-
balances the force on moving electrons due to magnetic field.
To obtain an expression for the Hall field, you should recall that the force experienced
by a charge q moving with a velocity v in a magnetic field B is given by

F=q(vxB) (10.37)

To appreciate the direction of this force, note that if the charge carrier is electron, q
becomes - e in Eq. (10.37) and direction of F is opposite to (v x B). However, for the
current flowing along positive x-direction, the direction of v for electron is along
negative x-direction. So, (v x B) is along positive y-direction and, therefore, F is along
negative y-direction. So, for electrons as charge carriers, we can write

Fy=-ev,B,

In steady state, Fyis balanced by the force (- e EY)due to Hall field. Thus, we obtain

-eEy=-ev,B,
 Eleetrmie Preperties
or

This relation gives the Hall field and can be written in terms of measurable quantities
if we use the expression for current density along the x-axis:

where n is concentration of electrons.

Substituting this expression for v, in Eq. (10.38), we get

Table 10.1: Hall co-efficient of

some metals at room Or
temperature

Metals RH x lo-''
(vm3 A-' ~ e s l a - ' )
Lithium - 1.70
(Li) where RHis called Hall codefficient or Hall constant. You may note that the Hall
field developed in metallic conductor is proportional to both: current flowing in
Copper - 0.55 conductor and the magnetic field in which it is placed.
(Cu)
This is a very useful result because it shows that Hall coefficient depends only on
Zinc (Zn) + 0.33 concentration of electrons. Measurement of RHis, in fact, a standard technique for
experimental determination of electron concentration in metals. In addition, it also
provides a test of the free electron theory of metals which assumes that valence
Aluminium - 0.30 electrons become free electrons participating in transport phenomenon in metals.
(All
Experimental values of the Hall co-efficients of some metals is given in Table 10.1.
Note that the values of RHfor Zn and Be are positive. From Eq. (10.40), it is obvious
Berelium + 2.70 that positive RHmeans electrical conduction by positively charged carriers.
(Be)
'This is an interesting observation because free electron theory assumes that electrons
are the only charge carriers in metals!
Sodium
(Na) The positive Hall co-efficient can be explained on the basis of band theory. The band
structure of these metals is such that the conduction bands are more than half-filled.
Thus, electrons at the top half of the band participate in electrical conduction. These
electrons have negative effective mass and they behave as positively charged carriers.
This feature of the Hall coefficient - its sign depends on the sign of the charge
carriers - becomes very useful in the study of semiconductors where both types of
charge carriers - electrons and holes - play important role simultaneously.
Now let us summarise what you have learnt in this unit.

10.5 SUMMARY
An electron in a crystalline solid move under the influence of a periodic potential
due to ions. According to Kronig-Penney model, the periodic potential due to
ions in a 1-D crystal can be represented by a series of equidistant potential wells.
 Band Theory of Sol
According to Bloch theorem, the wavefunction representing an electron in a
periodic potential can be expressed as a plane wave modulated by a periodic
function; that is,

~ ( x=) uk (x) . e* lkr

where uk(x) = uk (X+ a); a being the lattice constant.

Solutions of the Schrodinger equation for an electron in a periodic potential can

exist only if the following relation holds:

sin a a
P- + cos a a = cos k a
aa

ma
where P = -T Vob is a measure of the strength of attraction between electron and
A
ion. This relation shows that a) energy spectrum of an electron consists of allowed
and forbidden energy bands; b) as energy of electron increases, width of allowed
I
bands increases; and c) as P increases, width of the allowed bands decreases.

Free electron theory of metals proposed by Drude-Lorentz and ~ommerfeldare

particular cases of band theory of solids. ~ h e s two
e particulapases respectively
correspond to electron as a completely free particle and in a constant potential
field.

The dispersion relation of an electron in a periodic potential shows that the

discontinuities in energy spectrum occurs at the Brillouin zone boundaries i.e. at
k = f nnla; n = 1,2,3, ...

Number of energy states in an allowed band in a crystalline specimen is equal to

the number of unit cells in it. This means that total number of electrons that can be
accommodated in a band is twice the number of unit cells in a specimen.

According to band theory, the velocity and mass of an electron in a crystalline

solid is significantly different from those of a free electron. Velocity of an
electron near the top of the band decreases with increasing energy. Similarly,
mass of electron is different from its rest mass. The (effective) mass of an
electron in a crystalline solid is given by

1 tt2
m =
(d2~ l d k ' )

Electron near the top of the band has negative effective mass which means that
electron behaves like a positively charged particle (called hole).

For a completely filled band, the effective number of free electrons is zero and
when the band is half-filled, maximum number of effectively free electrons are
available for electric conduction. This conclusion of the band theory enables us to
differentiate metals from insulators and semiconductors.

According to band theory, metals have partially filled band whereas insulators
and semiconductors have completely filled conduction and completely empty
valence bands separated from each other by an energy gap, E,. However,
energy gap in semiconductors is much less than that of an insulator; therefore, as
temperature is raised from absolute zero, electrical conduction may take place in
semiconductors.
Electronic Properties The value of wave number for which discontinuities appear in the E vs, k curve of
an electron represent the Brillouin zone boundaries because the corresponding
energy values are not allowed to electrons. Therefore, it is possible to construct
Brillouin zone patterns of solids on the basis of their E vs. k curves. Since first
Brillouin zone is a unit cell of the crystal in reciprocal space, we can determine
the crystal structure of the solid on the basis of its Brillouin zone pattern.

Hall effect experiment is used for determining the concentration and sign of
carriers in metals as well as in semiconductors. Positive value of Hall coefficient
observed for some metals cannot be explained on the basis of free-electron theory.
Band theory, however, do account for positive Hall co-efficient of metals in
terms of negative effective mass of electrons at the top of the band. This
experiment also enables us to test the predictions of free electron theory regarding
concentration of charge carriers in metals. The Hall co-efficient is given by

R, =7=- Y 1
-
J BZ

10.6 TERMINAL QUESTIONS Spend 35 min.

1. A 2-D square lattice has side 2.5 A . What will be the momentum of an electron
whose wave terminates at the boundary of first Brillouin zone? Also calculate
energy of the electron.
2. When 90 mA current is passed through a sodium specimen under the magnetic
field 2.0 Weber m-2, the Hall voltage is 0.09 mV. The width-of the specimen is
0.04 mm. Calculate the carrier concentration.
0

3. For a 2-D square lattice of side 0.02 A , calculate the momentum of electron
corresponding to the boundary of the first Brillouin zone.

10.7 SOLUTIONS AND ANSWERS

Self-Assessment Questions (SAQs)

1. a) In a crystalline solid, atoms are arranged along a periodic lattice. In those

solids in which the valence electrons are loosely bound to the constituent
atoms, such as metals, the valence electrons becomepee leaving positively
charged ions at lattice sites. The free (conduction) electrons experience
potential field due to the positively charged ions. Since ions are arranged
along a periodic lattice, the potential field due to them is also periodic.

b) Since an electron experience a periodic potential in a crystalline solid, the

wave function describing such an electron must have the form of Bloch
function given by

where uk (x) = uk (X+ a), a being the lattice constant.

2. Values of -siniaa' 'i
aa + cos aa for different values of aa and P = 3n/2 is
Band Theory of Solids

tabulated below:

-sin aa + cos aa
5.7 1

3x/4 lStallowed band (width = 0 . 5 ~ )

3. Refer to the answer of SAQ 2 which gives -sin aa + cos aa

(:a
values of aa for fixed value of P (= 3~12).Note that these values have been
.tabulated at the intervals of ni4 in aa. The second allowed band covers the range
3n:
- 5 aa 5 2n and the width of the band is 0.5 IT.However, the third allowed

19n
band covers the range -I a a 5 3n giving width of the band equal to 0.66 n.
8
You can show that the allowed band width will further increase for increasing
values of aa. Thus, as energy of electron increases, allowed energy band width

From Eq. (10.26)

From Eq. (10.30), effective mass of an electron is

1 Electronic Properties We have

5. We know that discontinuities in dispersion curve appear at (Eq. (10.20))

2x '
For second Brillouin zone, we must take n = 52. So, k = +_ - . For a 2-D s q u ~
a
lattice, this condition transforms to

with n,2 + n22 = 2.

I Following combinations of values of n, and n2 will satisfy above condition:

2x
nl = + 1, n2 = - 1 which gives k, - ky= -

I
a
2n
n,=-1, n2=+1 whichgives-kx+ky=-
a

which gives -k,- ky= -

L I L
nl = - 1, n2 = - 1
a

I These are equations of straight lines and taking k, and k, as axes, we get the
following figure for the second Brillouin zone:

I Fig.lO.12: Second Brillouin zone (shown by shaded regions) of a 2-Dsquare lattice

 Band Theory of Solids
Terminal Questions (TQs)
1. You know that the boundaries of first Brillouin zone of a square lattice is given by
(Eq. (10.20))

k=+- nx
a

where 'a' is the lattice constant. We have,

a = 2.5 A = 2.5 x 10-''m.

Momentum of an electron at the boundary of first Brillouin zone can be written as

(taking n = 1 in above expression for the wave number),

n
p = h k = h . - = ( 1 . 0 5 ~ 1 0 - ~ ~ ~ s ) x3.14
a 2.5 x 10-'Om

= 1.38'~ kg.ms7]

Corresponding value of the energy of electron is

p2
E=-= (1.38 x kg.ms-')2
= 1.04 x 10-l8 J = 10.4 eV.
2m 2~(9.1x10-~'k~)

2. Hall co-efficient, RH is defined as the Hall electric field per unit current density
per unit magnetic field. So, if VH is the Hall voltage and b and d a r e the lengths of
the edges of a parallelepiped parallel toy- and z-axes respectively (refer to
Fig. 10.1I), we can write

E, =-VH
b

and

I
j =A
bd

where I, is current.

So, we can write Hall co-efficient as

v,
RH =4=- V H ~
Ix I , Bz
Bz ii
We have,

VH= 0.09 mV = 0.09 x V; I, = 90 mA = 90 x ~ o - ~ A ;

d = 0.04 m m = 0.04 x m; and B, = 2 Wb m-2

Electronic Properties Thus,

But, from Eq. (1 0.40), the magnitude of Hall co-efficient is given by

n: n:
3. We know that the first Brillouin zone lies in the range - - < k < -. So, we have
a a
to calculate the momentum of an electron whose wave number, k = + -.a
iL
And
momentum of an electron is given by
and

The solution of these differential equations was suggested by Bloch as under,

y(x) = u k( x )e'" (A.3)

 Electronic Properties
Thus, we have

and

Thus, we can write the general solution as

where A and B are constants.

Similarly, substituting the solution u2 = ePX in Eq. (A.6) and solving as above, we
can write the general solution as
- C e(P-ik)x + D e- ( b i k ) x
'42 - ('4.8)
where C and D are constants.

To obtain the values of constants A, B, C and D, we apply the following boundary

conditions on Eqs. (A.7) and (A.8):

iii)

Applying (i) on Eqs. (A.7) and (A.8), we get

(A+B) = (C+D) (A.9)

Applying (ii) on Eqs. (A.7) and (A.8), we get

I i(a - k)A - i(a + k)B = (P - ik)C - (P + ik)D (A. 10)

Applying (iii) on Eqs. (A.7) and (A.8), we get

A el(a-k)a + B -4a+k)a -
- e-(P-ik)b + D (P+lk)b
(A.ll)

Applying (iv) on Eqs. (A.7) and (A.8), we get

A i(a - k)e'(a-k)a - B i(a + k)e-(a+k)a = C (p - ik)e-(p-'k)b - D (p +ik)e(P+~~)~

Eqs. (A.9)to (A.12)will have non-vanishing solutions if and only if the determinant
of the coefficients of A, B, C and D vanishes. That is,

1 1 1 1
i(a -k ) - i(a k )+ (P - ik) - ( P + ik)
-;(a+k)a =0
ei(a-k)a
e -(P-ik)b
e (P+;k)b

( A .13)
To solve this determinant, let us write it as

Expanding and rearranging terms we get

( e - g ) ( f -h) (bd+ac)+(g- f ) (e-h) (bc+ad)

Substituting the values of a, b, c, d, e,f; g and h from Eq. ( A .13), in Eq. ( A .14) and
simplifLing we get,

Dividing throughout by 2, we get

Now, using the following relations I

Electronic Properties we can write

p2 -a2
2 sin aa sinh pb = 2 cos k ( a + b ) - 2 cos aa cosh pb
2aP
or

2
p 2 - a sinh~bsinaa+cosh~bcosaa=cosk(a+b) (A.15)
2aB

This is the desired relation expressed as Eq. (10.13).