Eng.

Physics Notes
Chap4-Band Theory of Solids

Introduction
Metals have certain special properties as compared to other solids. For instance, they are
very good conductors of electricity and heat, and are malleable, ductile etc. Explanation of such
properties demands a good theoretical understanding. Hence the study of metals assumes
significance. It becomes important for another strong reason. That is, a large number of elements
of periodic table belong to the category of metals, as can be seen from Fig. (6.1). It implies that
most of the solids that we see in nature are likely to have one or more metal elements as their
constituents. Therefore, even to understand the behaviour of solids that are not purely metals, one
must first understand about metals.

The first ever attempt, to give a theoretical model, that explains the observed properties
of metals, was made by Drude, in 1900. Since then, several models on metals have appeared. In
the present chapter, we will study two important of those models;
1-Drude-Lorentz model or classical free electron theory.
2-Kronig- Penny model or band theory

We should note that these three are not completely separate models. As you will see
shortly, they can be treated as successive stages of a single big theory

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

CLASSICAL FREE ELECTRON THEORY
Drude has developed a model of metals in the lines of kinetic theory of gases by making
the following assumptions. These assumptions constitute the classical free electron (or Drude’)
model. For the sake of understanding we divide the assumptions into four points along the
following lines.


1. How free e s are generated in metal.

2. How and why detached e s move freely

3. What happens to free e s in the absence of external electric field, E

4. What happens to free e s in the presence of external E.

1(a). Metals atoms have one or more loosely bound valence electrons. Such
electrons get detached easily even with small thermal energy.

1(b).The detached electrons are neither shared nor acquired by any of the atoms of the metal.
Hence they are free and just form a common pool.

2a. The electrostatic interaction of detached electrons with the positive ion cores and with other
(detached) electrons is assumed to be negligible.

2b.A consequence of assumption 2(a) is that the detached electrons can

move freely1 every where within the confinements of the metal piece. Hence these electrons
are called as free electrons or conduction electrons.
3a. In the absence of an electric field, the free electrons make fast random motion making
collisions2 with the ion cores.
3b. An electron on the average travels for a timecalled mean free time
between successive collisions. The average distance travelled between two successive
collisions is called as mean free path.

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

4a.When an electric field is applied, free electrons will experience slow drift motion3 in the
Positive direction of the field and produce a current in the metal.

4b.In the presence of electric field electrons should be accelerating indefinitely. However, the
(assumed) collisions of electrons with ion cores bring them to an equilibrium state. This leads to
a steady

IMPORTANT TERMINOLOGY OF FREE ELECTRON THEORY

Free electron theory uses the following terms:

(a)Mean collision time,  c

(b) Mean free path
© Mean velocity
(d) Drift velocity or average velocity, vd
(e) Relaxation time,  r
(f) Relation between, vd and  r
(a) Mean collision time (  c )

 12   23  ...   nn 1
c 
n

Where  ii 1 is the time taken between ith and (i+1)th collisions and n is the total number of such
collisions.
(b) Mean free path:

12  23  ...  nn 1


n
where ii 1 is the distance traveled between ith and (i+1)th collisions and n is the total number of
such distances.

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

(c) Mean velocity

1 n
v  vi
n i 1

Where vi is the velocity of the ith electron. Mean velocity of electrons in the absence of
external electric field is zero. Why? Because….

(d) Drift velocity, vd

Drift velocity (of electrons in a metal) may be defined as the mean velocity of all the free
electrons in the presence of an external electric field, E.

(e) Relaxation time,  r

Def: Relaxation time is of the free electrons is defined as the time taken for their drift
velocity to decay to the value of (1 e)th of its initial value, since the external electric field is
switched off.

(f) Relation between, vd and  r

We can establish relation between vd and  r as follows. Let us assume that an electric field E is
applied on the free electron gas. The equation of motion of these electron under the influence of
this field can be written as,
dvd
m  Fext  Fretar
dt

Where, Fext  eE is the average external force acting on the electrons due to the field.
Fretar  mvd  r is the average retarding force acting on the electrons due to collisions. With these
values the above equation becomes:
dvd v
m  eE  m d
dt r

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

  dv v 
m d  d   eE
 dt  

When the field, E is switched off, the drift velocity starts decreasing. The time dependence of vd
can be obtained by setting, E  0 in the above equation and integrating the resultant eq. Thus,

 dv v 
m d  d 0
 dt  
dvd dt

vd 
 
dv 1 dx
0 vdd    0 dt  x
 ln x

This integration yields,

vd (t )  vd (0)et 

where vd (0) is the drift velocity when the field is switched off and vd (t ) is at time, t

CALCULATION OF ELECTRIC CONDUCTIVITY,  BASED ON CFET.

Let us assume that an external ele field, E applied to a metal.
Then equation of motion for free electrons can be written as:
dvd
m  Fext  Fretar
dt

F  ma

Where, Fext  eE is the average external force acting on the electrons due to the field.
Fretar  mvd  r is the average retarding force acting on the electrons due to collisions. With these
values the above equation becomes:

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

  dv v 
m  d  d   eE
 dt  
When steady state is reached,
dvd
0
dt

1 eE
vd 
 m
eE
Or vd  
m
Current density j in any conductor can be expressed as;
current
j
Area of cross section
j i/ A

Also current density in terms of density charge carriers, n is given by:

j  n(e)vd

j  nevd

Now substituting the value of vd we will have,

ne 2 r
j E
m
When we compare it with Ohm’s law ( i  V R ) in Tensor form,

j E

Comparing equations ( ) and ( ) conductivity,  becomes:

ne 2

m

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

Temperature dependence of 

At room temperature, the drift velocity imparted to electrons by the electric field is very
small as compared to random speed of electrons. The mean free time taken by the electrons in
traversing mean free distance will hence be decided not by the drift velocity, but by the so called
root mean square velocity (rms), caused by thermal motion. As per kinetic theory of gases rms
velocity is given by,
3kBT
c (10)
m
where kB is the Boltzman constant, T is the absolute temperature and m is the mass of the
electrons.
Relaxation time , mean free path  and rms velocity are related
by,
  c (11)

From Eq.( 10) and (11 )  can be expressed as,

m
  (12)
3k BT
Substituting, value of  from Eq.( 12 ) in ( 9a ) we have,

ne 2 m
 
m 3 k BT

ne2
or,  (13)
3mk BT
Eq. (13)indicates that  is proportional to 1 or resistivity  T
T
Success and failures of Drude’s model:
Successes
1. It explains the mechanism of electrical conductivity.
2. It explains the origin of electrical resistivity.

3. It explains the Weidmann-Frantz law  constant

4. Indicated that e-s are real charge carrier of electrical current
Failures:
1. Temperature dependence of electrical conductivity is not explained correctly.
2. Heat capacity and paramagnetism are not explained correctly.

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

INTRODUCTION TO BAND THEORY OF SOLIDS

Note: Why band theory?

The failure of the free electron models (both classical and quantum) is due to the over simplified
assumption that the interaction between a free or detached electrons and positive ion cores is negligible.
This assumption implies that potential due to a positive ion cores is zero or rather constant throughout the
crystal, as had been taken in the potential well. Which means the detached electrons can move freely within
the metal.

But in reality, the free electron-ion core interaction can not be neglected. As such the potential due
to the positive ion cores can not be taken as constant through out the crystal. And therefore detached
electrons can’t be completely free. The actual situation of potential in the crystal is roughly as follows. The
potential field of a single positive ion core -as seen by an electron- is as shown in Fig. (6.12a). The
corresponding potential due to that of a diatomic molecule is as in Fig. (6.12b). Extending the structure, the
potential field of a crystal (defined as a regular periodic arrangement of atoms) would be as shown in Fig.
(6.12c). That is the potential field now becomes a periodic function with a period ‘a’.
Thus, while the actual structure of the potential field is some thing like this or even more
complicated; free electron model takes just a constant potential ignoring the periodic variations. As
compared to this, the potential field assumed in free electron theory is a very crude approximation.

Kronig and Penny had tried to rectify this draw back by taking the periodic variation of
the potential into account. But, they did not take the potential field of Fig. (6.12c) as it is would
be too complex for mathematical analysis. Rather they have taken a rectangular type of periodic
potential of Fig. (6.3 ). Though, even this is not an exact representation of the potential given in
Fig. ( 6.12c), this is much closer to the actual situation than the previous models. In fact this model
lead to marvelous conclusions and gave clue to several of the properties of solids that are not
explained by the previous models. In the following we proceed for the study of Kronig and Penny
model.

The Kronig-Penny Model

The periodic potential assumed by Kronig and Penny is as shown in Fig. 2 i.e. a series of
rectangular wells of width a and are placed at a separation of b. Thus periodicity of the potential
function is a  b . In regions where 0  x  a , the potential energy is zero and in regions such as
b  x  0 , the potential energy is V0 . The detailed mathematical study of this model is out of the
scope of our discussion. Therefore here we will just look at the main features of the model and its
predictions qualitatively.

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

A. Main features of the model features:
Schroedinger equation
The dynamical behaviour of electrons in the Kronig-Penny model is represented by the
following Schroedinger equation,

d 2  8 2m 
   E  V0   0, for  b  x  0 (1a)
dx 2  h2 
And

d 2  8 2m 
  E  0, for 0  x  a (1b)
dx 2  h 2 

Let us assume that the total energy E of the electron under consideration is less than Vo.
Further let us substitute
8 2mE 8 2 m V0  E 
 
2
and  
2
(2)
h2 h2
where  and  are real quantities. On substitution of conditions given in
Eq. (2) we will have,

d 2
and
2
  2  0, for  b  x  0 (3a)
dx

d 2
2
  2  0, for 0  x  a (3b)
dx
These eqs (3) can be solved based on Bloch theorem. The final solution of this equation
is given is the form of the following condition.
sin  a
P  cos  a  cos ka (4)
a

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

 where, P  8 2 mV0ba h2

. That means, Eq. (3) will have a solution only when the condition ( 4) is satisfied. The physical
significance of solutions (of Eqs.3) is discussed, below, in a graphical approach.

sin  a
B. Graph of a vs P  cos  a
a

Values of cos ka, on RHS of Eq. (4), lies between +1 and -1. Therefore a on LHS should
take only those values for which the total LHS value lies between +1 and -1. Other values are not
allowed. Fig shows a plot of the condition (4 ) i.e. LHS versus a [Note: here p = 3/2].

The vertical axis lying between -1 and +1 represents the values acceptable to the LHS. Horizontal
axis represents the value of a, which is a measure of energy.

C. P versus energy relationship

To have a comprehensive idea we look at eq. (4) in two extreme conditions, i.e., P  
and P  0 as follows. Note that since P  mV0ba 2 variation in P is linked to potential V0 .

(i) Tight binding limit ( P   )

When P   i.e., increasing the barrier strength , V0   , electrons can’t escape the
potential and are tightly bound to the atoms. In this case Eq. (4) will have solution only if sin  a  0
. This condition implies,
 a  n (5)
where n is a positive integer. Substitution of value of  in Eq. (2) i.e.,  2  8 mE
2

h2
yields,
8 2 mE 2 h2 2
a  n 
2 2
or, E  n (6)
h2 8ma 2

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

 The energy level spectrum given by Eq. (6 ) is same as that of an electron confined to a
potential well of width a. That implies that the width of allowed bands decreases and the band
structure reduces to line spectrum. Hence the limit, P   is also known as the tight binding limit.

(ii) Free electron limit ( P  0 )

In the limit of, P  0 , Eq. (4) becomes,

cos a  cos ka or,  k

k 2 2 p2
E 
2m 2m
(7)
The energy spectrum given by Eq. (7) is same as that of a free electron. That means all
values of energy are allowed for the electron like in the case of a classical free particle. Hence this
limit is called free electron limit.

The conditions given by Eqs. (6 and 4) are re-plotted in the form of E vs P in the following Fig.
(2) below.

Conclusions of the graph

1. The energy spectrum (of metals) consists of alternative regions of allowed and forbidden
(vacancy) bands. Forbidden band imply that the energy levels that lie in this region are not
occupied by the free electrons of the metals.

2. The width of an allowed energy band increases with a or the energy. That is bands at higher
energy values are wider than those at lower energy values (move vertically in the graph)

3.(a) For P  0 (i.e., on the extreme left) the hole energy spectrum is quasi-continuous. That is all
allowed bands are joined together forming an almost continuum. However, the width of a particular
allowed energy band decreases with increase in the value of P. As P   , the allowed energy
bands are compressed into simple energy levels and thus result in a line spectrum.

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

ENERGY vs WAVE VECTOR (W - K) RELATIONSHIP: ORIGIN OF ENERGY BANDS

From Eq. (4) we have obtained E as a function of k i.e., keeping the value P as a constant.
However note that cos ka of Eq. (4) is an even periodic function of k. Hence it remains unchanged
if k is replaced by k  k  (2 n) / a , where n = 1, 2, 3, etc. It implies that the energy E is also an
even periodic function of k with period of 2 / a . Fig.3 gives the variation of E with k. As can be
seen from the figure, E is not continuous with respect to k, but shows discontinuity at ka  n
or k  n / a . These discontinuities divide the whole energy range into regions of acceptable and
forbidden values. The acceptable and forbidden regions are referred to as acceptable and
forbidden band.

EFFECTIVE MASS

Consider the effect of an applied electric field E on the motion of an electron in a periodic
potential. If v is the velocity of the electron and the field E acts on it for a time dt, the increase in
energy of the electron is given by
dE  eEvd dt (1)

k2 2
We can obtain value of vd from Eq. E  as
2m
1  dE 
vd    (2)
 dk 

Differentiating Eq. (2) w.r.t, t we will have,

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

 dv 1 d  dE  1 d 2 E dk
  
dt dt  dk  dk 2 dt
dv 1 d 2 E dk
 (3)
dt dk 2 dt

Substituting value of vd from Eq.(2) in Eq. (1) yields,

1  dE 
dE  eE   dt (4a)
 dk 
Left side of Eq. (4a ) can be expressed as,
dE
dE  dk (4b)
dk
Then,
dE 1  dE 
dk  eE   dt (5)
dk  dk 
We can get dk dt after simplifying eq. (5) as,

dk 1
 eE (6)
dt
Now substituting this value of dk dt in Eq. ( 3) we get,
dv 1 d 2 E
 2 eE (7)
dt dk 2

Let us rewrite (7) as follows,

2
dv
2
 eE 2
(d E dk ) dt

2
dv
2
F 2
[ F  eE ] (8)
(d E dk ) dt
Notice that Eq. (8) resembles Newton’s second law expressed in the form of,
dv
m F (9)
dt
2
Comparison of Eqs. (8 ) and (9 ) reveals that the quantity plays the role of
( d 2 E dk 2 )
mass of the electron. For this reason, this quantity is usually called as effective mass of the electron
in the crystal and is denoted by m* .

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090

Def:

Thus, the effective mass, m* of the electron in crystal field may be defined as a fictitious
mass that greatly differs from its actual mass m of the electron. Depending upon the value of,
(d 2 E dk 2 ) m* can be more or less than m. Further it may take on negative values also.

The k versus effective mass, m* of an electron in a periodic potential is shown in the Fig. 3

Dr. G V Rao, H&S Dept., VNRVJIET, Bachupally, Hyderabad-500 090