Metal is the most important form of solid.

Due to their mechanical strength,

conductivity and abundance in nature, they have found various applications in
engineering. Therefore these materials are being utilized in the development of
society. More investigation have been made from the beginning of twentieth
century. An atom is considered to be made up of electrons, protons and
neutrons. The electrons, revolving around the nucleus in various orbits play
important role in various engineering application. However, all electrons of
metal do not participate in various conduction properties. The valence shell
electrons of metals are responsible for overall behaviour of metal. These
electrons do not affect the motion of core electrons and nucleus of atoms.
Various models have been applied to investigate the electrons motion so that
controlling parameter for describing the properties of materials under external
influence can be defined. Modifications in the previous theory have been made
by applying quantum mechanics. Fundamental terms, definition are given in this
chapter, which are important tool to understand the functioning of devices. In
this chapter valence shell electrons have been treated as free electrons and their
motion is subjected under external effect. Fundamental aspects have been given
while describing the motion of free electrons through models consisting of
various assumptions. Further modifications made by scientist into the theory to
meet the experimental results have been discussed. These theoretical models
make us to equip with fundamental understanding of nanoelectronics. The
confined motion of electrons in certain boundaries has been dealt here, which is
helpful to understand the motion of electrons in nanostructure devices.

3.1 ELEMENTS OF FREE ELECTRON THEORY (CLASSICAL)

In this theory each metal contains a number of free electrons which behave like
the molecules of a perfect gas and are free to move about the whole volume of
metal. They thus form free electron gas in a container. In this model it is
assumed that the free electrons, i.e., those giving rise to the conductivity, find

© The Author(s) 2022 55

P. K. Basu and H. Dhasmana, Solid State Engineering Physics,
https://doi.org/10.1007/978-3-031-10940-9_3
56 SOLID STATE ENGINEERING PHYSICS

themselves in a potential which is constant everywhere inside the metal. Since

one does not observe electron emission from metals at room temperature, it
seems evident that potential energy of an electron’s at rest outside the metal.
Based on this model Lorentz and Drude in 1900 put forward their theory for the
calculation of conductivity of metals.
(i) Drude Theory (Lorentz – Drude)
The basic assumptions of Lorentz – Drude theory are listed below :
(i) The large no. of electrons in a metal is free to move about the whole
volume of the metal.
(ii) The motion of free electrons here is similar to the thermal agitation
of a perfect gas. The electrical or thermal conductivity of metals is
solely due to these electrons.
(iii) The free electron here move randomly in all possible directions with
widely different velocities. If vi is the random velocity of ith
n r
electron and n is the total number of electrons, then, S vi = 0. Also
i =1
the distribution of velocity is in accordance with Maxwell’s
distribution.
3
So, the average kinetic energy (KE) of a free electron is kT,
2
where, k is Boltzmann’s constant and T is absolute temperature of it.
(iv) The free electrons make collision with positive ions (fixed) in the
lattice and also among themselves. The collisions among electrons have
no practical contribution to the conductivity of metals.
(v) In absence of electric field random motion in all directions of
electrons occur. So current density in metal is zero.
(vi) When the external field is applied, the electrons drift slowly with
some average velocity known as average drift velocity, in the
direction opposite to that of electric field. The drift velocity is
superimposed over their random velocity.
(vii) Under external field free electrons being accelerated gain some
additional KE and by colliding with fixed positive ion loses its
additional KE. Thus there is a loss of KE in collision under electric
field (inelastic collision). Just after inelastic collision the electron
momentarily will have random motion. Again due to the presence of
electric field electron is again accelerated and gains drift velocity to
next collision. This process continues till the electric field is
acted on.
(viii) The average distance traversed by a free electron between two
successive collisions with positive ions is known as Mean Free
Path l.
(a) Determination of electrical conductivity, s, of the metal :
We consider a metal with n number of free electrons per cubic meter.
Under an external electric field of E volts/meter, the force experienced by a free
FREE ELECTRON THEORY 57
r r
electron is - eE Newton. The gain in momentum after a time t then is - eE ´ t
Newton-sec.
If m is the mass of the free electron, t i is time for ith electron, then, the
r r
momentum of ith electron at that instant is ( mvi - eEt i ).
The average momentum of electrons at that instant is
n r r
S ( mvi - eEt i )
r i =1
Pav = .
n
So, average drift velocity,
r r
r P 1 1 æç r eE ö÷
v d = av = S ç vi - ti …(3.1)
m n i =1 è m ÷ø
Now, the velocities of free electrons are randomly distributed,
n r
Þ S vi = 0 …(3.2)
i =1

From eqns. (3.1) and (3.2),

r
r 1 n æ eE ö
vd = S ç- ÷
n i =1 ç m ti ÷
è ø
n
r S ti
r eE i =1
Þ vd =-
m n
r
r eE
Þ vd =- t …(3.3)
m
1
Here, t= S ti …(3.4)
n i
This is the average time elapsed after collisions. This average time will be
equal to the average time from this instant to the next collision. So, if the
collisions occur at regular intervals the average time t between collisions will
be, t = t. Here t is the relaxation time.
So, from eqn. (3.3),
r
r eE
vd = - t
m
Hence, the current density,
r r
J = - nev d
N0 e
Here, current, I = - ; N 0 is total no. of electron within volume V & t
t
is time rate,
58 SOLID STATE ENGINEERING PHYSICS

e
I =-n´V ´
t
L´ A
Þ I =-n´ ´e
t
I L
Þ J = - = - n ´ ´ e = - nv d e]
A t
r
r eE ne 2 t r
So, J = - ne ´ - ´t= E
m m
r ne 2 t r
Þ |J | = E
r r m
But, J = s E ; s is electrical conductivity of the metal,
r
Hence, comparing above two expressions of J ,
ne 2 t
s= …(3.5)
m
From eqn. (3.5) we found that the differences in s for different metals are
due to the difference in n. A metal is pictured as composed of positive metal
ions whose valence electrons are free to move around in three dimension ionic
array. As there are 10 22 to 10 23 atom/cc, a large number of free electrons is
available moving in all directions inside the metal like the atoms or molecules of
a perfect gas.
(b) Determination of thermal conductivity, K, of the metal :
Based on the assumptions of this metal the thermal conductivity can also
be calculated. Let us consider a rectangular bar, as shown in Fig. 3.1 thermal

Fig. 3.1 A rectangular metal bar showing three different slices at different

energy is flowing in positive X-direction. Let A, B, C are three slices of this bar
having uniform cross-section. Due to the existence of thermal gradient the
temperatures of the slices A, B, C are T1 T and T2 respectively in such a way that
T1 > T > T2 . Let l be the mean free path of the electrons separating slices
A, B, C, then, according to Joule’s law, the heat flux is
DT
DQ µ
Dx
DT éT - T2 ù
Þ DQ = K =K ê 1 …(3.6)
Dx ë 2l úû
Equation (3.6) gives the amount of energy per unit area per unit time
exchanged in the slice B.
FREE ELECTRON THEORY 59

Let n be the number of electrons per unit volume, then n/ 6 will be the
number of electrons per unit volume moving in positive X-direction. If v is the
average velocity of the electrons, then, nv/ 6 will represent the number of
electrons flowing in positive X-direction per unit area per unit time. Thus the
amount of energy transfer from the slice A to B per unit area per unit time is
3
Q1 = ( nv / 6) ´ kT1 .
2
Similarly the amount of energy transfer from the slice C to B per unit area
3
per unit time is Q2 = ( nv / 6) ´ kT2 . Here, k represents the Boltzmann’s
2
constant.
So net heat flux exchange in B is
3
DQ = Q1 - Q2 = ( nv / 6) ´ k (T1 - T2 ) = ( nv / 4) k (T1 - T2 ) …(3.7)
2
Hence by equating eqns. (3.6) and (3.7),
éT - T2 ù
Kê 1 = ( nv / 4) k (T1 - T2 )
ë 2l úû
1
Þ K= nvkl …(3.8)
2
From eqn. (3.8) we found that the differences in K for different metals are
due to the difference in n.
(ii) Wiedemann-Franz Law
Statement : The ratio of thermal conductivity, K, to electrical
conductivity, s, is proportional to the absolute temperature and this constant of
proportionality is known as ‘Lorentz Number’.
Proof : We have for a free electron gas from eqn. (3.8),
1
K = nk lv
2
l
For a particle with n and l, t = …(3.9)
v
Also from eqn. (I),
ne 2 ne 2 l
s= t= …(3.10)
m m v
1 1
Now, KE in three dimension is mv 2 and equals to 3 ´ kT.
2 2
1 2 1
\ mv = 3 ´ kT
2 2
3 kT
Þ mv =
v
3 kT
Þ mv = …(3.11)
v
60 SOLID STATE ENGINEERING PHYSICS

From eqns. (3.10) and (3.11),

ne 2 l ne 2 l 1 ne 2 lv
s= = ´ = …(3.12)
m v 1 3 kT / v 3 kT
1
nk lv
K 3 k2
Using eqns. (3.8) and (3.12), = 2 = T
s ne 2 lv / 3 kT 2e 2
K
Þ µ T = LT
s
K
Here, L = Þ This is known as Lorentz number and is given by
sT
3K 2
L=
2e 2

3.2 LIMITATIONS OF FREE ELECTRON THEORY (FET)

(1) Specific Heat : The electronic contribution of specific heat has serious
disagreement with the experimental results. The FET is unable to explain the
reason of dominance of low temperature specific heat evaluation by its
electronic contribution.
(2) Paramagnetism of Metals : Experimental fact that the paramagne-
tism of Metals is nearly independent of temperature, this is not at all explained
by Free Electron theory.
(3) Widemann-Franz Law : This is a consequence of Free Electron
k
Theory. It gives the value of , here, k is the thermal conductivity of the metal
sT
concerned, which are found to agree with theoretical value at ordinary
temperature, but deviations occur at low temperature.
(4) Hall Coefficient : Free electron theory predicts the value of hall
Coefficient being independent of temperature, relaxation time or the strength of
magnetic field applied. This is not at all found experimentally.
(5) The magneto resistance : FET predicts that the resistance of a wire
perpendicular to a uniform magnetic field should not dependent on the strength
of the field. Practically it does.
(6) Temperature dependence of Electrical Conductivity : FET does not
predict. However, it practically exists.
(7) Directional dependence of Electrical Conductivity : For some
metals s depends on the orientation of the specimen (if suitably prepared) with
r r
respect to the electric field. In such specimens J need not even be parallel to E.

3.3 QUANTUM THEORY OF CONDUCTION

The modification of FET was done by using quantum statistics (by
Sommerfield, 1928) on the basis of the following assumptions.
FREE ELECTRON THEORY 61

(i) The large numbers of conduction electrons in metals are not

completely free. They are bound to the metal as a whole. We now
arrive a physical model in which the interior of the metal is
represented by a potential energy (PE) box of depth E S as shown in
Fig. 3.2. E S is the energy difference between an electron at rest inside

Fig. 3.2 A PE box-depicting a metal.

the metal and in vacuum. At T = 0 K, all energy levels up to E F

(known as Fermi energy) are completely filled and all the higher ones
are empty. Here, j, known as the work function of the metal and is
the work required to be done to extract an electron from metal. This is
given by, f = E S - E F .
(ii) The forces between conduction electrons and ion cores are neglected,
so that the electron within the metal is treated as free. The total energy
of an electron is wholly the kinetic energy (KE) as PE is negligible
due to their smaller mass.
(iii) Due to their light mass and dense packing the electrons in a metal
may be considered as a gas under very high compression and hence to
a degenerate gas. Moreover, as the electron gas is charged, the free
electron gas in metal may be considered as dense Plasma.
(iv) The electrons are assumed to obey Pauli’s exclusion principle, so they
follow the Fermi-Dirac (FD) statistics, rather than the classical
Maxwell-Boltzmann (M-B) statistics.

3.4 SOLUTION OF ONE DIMENSIONAL SCHRÖDINGER

EQUATION
(i) Single Particle Inside a Constant Potential Region
We now consider a single particle confined in a region of constant
potential V0 . It is one dimensional motion is governed by the region 0 < x < L as
shown in Fig. 3.3.
The probability of finding the particle outside this region is zero. The
wave function Y ( x, t) describes the behaviour of the particle.
62 SOLID STATE ENGINEERING PHYSICS

Fig. 3.3 A potential well of depth V0 .

So, Y ( x, t) = 0 for x excluding 0 < x < L.

The potential function V0 is independent of time t, we can then write one
dimensional Schrödinger equation of the particle in accordance with eqn. (2.22)
is,
d 2 Y( x) 2m
+( E - V0 ) Y( x) = 0 …(3.13)
2
dx h2
[as V0 is constant we are only considering Y to be a function of x and
independent of t]
The solution of eqn. (3.13) is
æ 2m ö æ 2m ö
Y( x) = A sin çç ( E - V0 ) ÷÷ x + B cos çç ( E - V0 ) ÷÷ x …(3.14)
2 2
è h ø è h ø
Here, A and B are integral constants.
Now, for this problem the boundary conditions are
Y ( x) = 0 at x = 0 ü
ý …(3.15)
and Y ( x) = 0 at x = L þ
Using the first part of eqn. (3.15) in eqn. (3.14) we get,
Þ 0 = 0 + B ´1 Þ B = 0
So, eqn. (3.14) now modifies into
Y( x) = A sin kx …(3.16)
æ 2m ö
where, k = çç ( E - V0 ) ÷÷ …(3.17)
2
è h ø
Also using the second part of eqn. (3) in eqn. (4) we get,
0 = A sin ( kL)
Þ A =0
This is not possible as both A and B cannot become zero.
Hence, sin ( kL) = 0 = sin ( np) (where, n = 0, ± 1, ± 2, ¼)
2
æ np ö
Þ k2 =ç ÷
è L ø
FREE ELECTRON THEORY 63

2
2m æ np ö
Þ ( E - V0 ) = ç ÷
h- 2 è L ø
n 2 p2 h -2
Þ E - V0 = …(3.18)
2mL2
In eqn. (3.18), n = 0
Þ No value for (E - V0 ), i. e., Y( x) does not exist Þ This is not possible.
Also, for n = - 1, - 2, ¼Þ same value of (E - V0 ) as for n = 1, 2, 3, ¼
Hence we finally get,
n 2 p2h 2
E - V0 = where n = 1, 2, 3, ¼ …(3.19)
2mL2
Thus the particle can only have only certain discrete energy values given
by
n 2 p2 h 2
E = V0 + where n = 1, 2, 3, ¼ …(3.20)
2mL2
These are known as energy eigen values.
np
Now, k = , so from eqn. (3.16),
L
np
Yn ( x) = A sin x …(3.21)
L
The value of A can be obtained from the normalization condition of
Yn ( x), i. e.,
+¥
ò-¥ Y * Y =1
+¥
Þ ò-¥ | Yn ( x) | 2 dx = 1
L
Þ ò0 | Yn ( x) | 2 dx = 1

(as Yn ( x) = 0 outside 0 < x < L)

2
L½ npx ½
Þ ò0 ½½ A sin L ½
dx = 1

L æ npx ö
Þ | A |2 ò0 sin 2 ç
è L ø
÷ dx = 1

L 2
Þ | A |2 ´ =1 Þ | A | =
2 L
2
Þ A=
L

2 np
Hence, Yn ( x) = sin x …(3.22)
L L
64 SOLID STATE ENGINEERING PHYSICS

Ground state wave function is

2 p
Y1 ( x) = sin x.
L L
2 2p
First excited state is Y2 ( x) = sin x, and so on. They are shown in
L L
adjoining Fig. (3.4). For three dimensional case the electrons are confined into a

Fig. 3.4 Energy states in potential well.

cube of volume L3 and so eigen functions are

8 æ p px ö æ n y py ö æ n pz ö
Y( x, y, z) = sin ç x ÷ sin çç ÷÷ sin ç z ÷ and energy eigen
L3 è L ø è L ø è L ø
values are
p2h 2
( n x 2 + n y 2 + nz 2 )
E n ( n x , n y , nz ) = …(3.23)
2mL2
n p ny p n p
where, kx = x , k y = , kz = z . …(3.24)
L L L
Here, n x , n y and nz are integers, having values 1, 2, 3, … and
2p 4p
k x = 0, ± ,± , ¼ etc. and also same for k y and kz .
L L
(ii) Particle in a Box
When a free particle is moving inside the infinitely high potentially barrier
particle is confined to move inside this
barrier as shown in figure 3.5. If we
consider particle’s motion in one
dimension then it may be treated as a free
particle moving in a one dimensional
box.
The probability of finding the
particle outside this region is zero as
V = ¥ for 0 > x > L
and V = 0 for 0 < x < L Fig. 3.5 Particle in a one
dimensional box.
FREE ELECTRON THEORY 65

The wave function Y ( x, t) describes the behaviour of the particle.

So, Y ( x, t) = 0 for x having 0 > x > L.
We can then write one dimensional Schrödinger equation of the particle
within this box using eqn. (2.22) as
d 2 Y( x) 2mE
+ y( x) = 0 …(3.25)
2
dx h2
The solution of eqn. (3.25) is
æ 2mE ö æ 2mE ö
Y( x) = A sin çç ÷ x + B cos çç ÷x …(3.26)
2 ÷ 2 ÷
è h ø è h ø
Here, A and B are integral constants.
Now, for this problem the boundary conditions are
Y( x) = 0 at x = 0 and Y( x) = 0 at x = L …(3.27)
Using the first part of eqn. (3.27) in eqn. (3.26) we get,
Þ 0 = 0 + B ´1 Þ B = 0
So, eqn. (3.26) now modifies into
Y( x) = A sin kx …(3.28)
æ 2mE ö
where, k = çç ÷ …(3.29)
2 ÷
è h ø
Also using the second part of eqn. (3.27) in eqn. (3.28) we get,
0 = A sin ( kL)
Þ A =0
This is not possible as both A and B cannot become zero to have finite Y.
Hence, sin ( kL) = 0 = sin ( np) (where, n = 0 ± 1, ± 2, ¼)
2
æ np ö
Þ k2 =ç ÷
è L ø
2
2mE æ np ö
Þ =ç ÷
h2 è L ø
n 2 p2h 2
Þ E= …(3.30)
2mL2
In eqn. (3.30), n = 0.
Þ No value for E, i. e., Y( x) does not exist ® This is not Possible.
Also, for n = - 1, - 2, ¼Þ same value of E as for n = 1, 2, 3, ¼
Hence we finally get,
n 2 p2h 2
E= where n = 1, 2, 3, ¼ …(3.31)
2mL2
Thus the particle can only have only certain discrete energy values given
by
n 2 p2h 2
E= where n = 1, 2, 3, ¼ …(3.32)
2mL2
66 SOLID STATE ENGINEERING PHYSICS

3.5 DENSITY OF STATES

We have seen that the energy levels of a particle in a region of constant potential
are quantized. The number of energy levels per unit energy range per unit
volume at a given energy is called the Density of States. This is represented by
dn
DOS, or g( E) dE, and is given by .
dE n
Each energy band in a crystal accommodates a large numbers of electron
energy levels. Each energy level consists of two states and each state
accommodates only one electron. One energy level can be occupied by two
electrons only having opposite directions of spin. DOS represents the number of
states that can be occupied by charge carriers. However, all the available energy
states are not filled in an energy band. A particular energy level E is occupied or
not is determined by the probability f ( E) (this is known as the Fermi-Dirac
distribution function as the electrons are Fermions) that a carrier can have the
energy E.
If G ( E) dE is the number of available quantum states in the energy range
between E and E + dE, then the carrier concentration in energy range dE is
obtained by
dN = N( E) dE = f ( E) G ( E) dE …(3.33)
where, N( E) is the carrier distribution function.
In order to calculate the DOS
we consider a solid sphere in n
point space as shown in Fig. (3.6).
There any point (n x , n y , nz ) with
integer values of co-ordinates
represent an energy state. Thus, all
the points on the surface of the
sphere of radius n (where,
n 2 = n x 2 + n y 2 + nz 2 ) will have
the same energy. Since n x , n y , nz
can have only positive integral
value, so, the number of states of
energy E will be the octant of the Fig. 3.6 A point sphere of radius n.
1æ4 ö
sphere, i.e., G ( E) = ç pn 3 ÷ .
8 è3 ø
n 2 p2h 2 L
We know from earlier, E = Þ n= [ 2mE ]1/ 2 .
2
2mL ph
3/ 2
1 4p æç 2mL2 ö
÷
Hence, G ( E) = . . E 3/ 2 …(3.34)
8 3 çè p 2 h 2 ÷
ø
FREE ELECTRON THEORY 67

Now, on differentiation of eqn. (3.34)

3/ 2
1 4p æç 2mL2
ö
÷ 3 1/ 2
G ( E) dE = . . E . dE …(3.35)
8 3 çè p 2 h 2
÷
ø 2
Using Pauli’s exclusion principle, the number of energy states lying
between E and E + dE in volume V (= L3 ) by using eqn. (3.35) is
3/ 2
1 4p æç 2mL2 ö÷
G ( E) dE = 2 ´ . . E 1/ 2 . dE
8 3 çè p 2 h 2 ÷ø
4pV
= ( 2m) 3 / 2 E 1/ 2 . dE
3
h
So, DOS, the number of energy states lying between E and E + dE per
unit volume V is given by
4p
g( E) dE = ( 2m) 3 / 2 E 1/ 2 . dE …(3.36)
h3
Adjoining Fig. (3.36) shows the variation of DOS as a function of energy
in accordance to eqn. (3.34).
(i) Density of States (three Dimension)
r
For three dimension we consider the linear momentum, p , which
r r
corresponds to the operator, p = - ihÑ , in quantum mechanics.
Thus for the energy state,
r r r r h h 2p
pYk (r ) = - ihÑ y k (r ) = h k Yk (r ) [Q p = = = h k ]]
l 2p l
So, the plane wave Yk is an eigen
function of the linear operator p with the eigen
value hk. Hence the particle velocity in the
hk
energy state of orbital k is given by v = . In
m
the ground state of a system of N free
electrons the occupied orbitals may be
represented as points inside a sphere in
k-space. The energy at the surface of the
sphere is E F and the wave vectors at the
Fermi surface have a magnitude k F as in Fig. 3.7 Fermi surface in
Fig. 3.7. k-spaces.
h2 k 2
So, EF = …(3.37)
2m
3
æ 2p ö
Now the volume of each orbital is ç ÷ . Also the volume of the Fermi
è L ø
4p
surface or sphere of radius k F is ( k F ) 3 . Hence, the total number of
3
orbitals is
68 SOLID STATE ENGINEERING PHYSICS

4p
(k F ) 3
N =2´ 3 …(3.38)
( 2p/ L) 3
Here ‘2’ comes from spin degeneracy.
3 3
4p k F ´ L V
So, N =2´ ´ = k F 3 [V = L3 ]
2 2
3 2p ´ 4p 3p
1/ 3
é3p 2 N ù
Þ kF =ê ú …(3.39)
ë V û
where, ‘N’ also denotes the total number of electrons. So, k F mainly depends
upon N / V or particle concentration and not on the mass.
2/ 3
h2 é3p 2 N ù
So, EF = ê ú
2m ë V û
3/ 2
V é 2mE F ù
Þ N= ê ú …(3.40)
3p 2 ë h 2 û
The electron velocity v F at the Fermi surface is
1/ 3
hk h é3p 2 N ù
vF = F = ê ú …(3.41)
m m êë V úû
The density of states now can be expressed as
EF
ò0 G ( E) dE = N
3/ 2
EF V é 2mE F ù
Þ ò0 G ( E) dE = ê ú
3p 2 ë h 2 û
[from (3.40)]

3/ 2
V é 2mE ù
Þ ò G ( E) dE = ê ú
3p 2 ë h 2 û
[expressing in indefinite integral form]
On differentiation,
3/ 2
V æ 2m ö
G ( E) dE = ç ÷ E 1/ 2 dE
2p 2 è h 2 ø
4p
So, g( E) dE = ( 2m) 3 / 2 . E 1/ 2 dE …(3.42)
3
h

3.6 FERMI-DIRAC (FD) DISTRIBUTION FUNCTION

We know, N( E) dE = f ( E) G ( E) dE
Here, f ( E) determines the carrier occupancy of the energy states (i. e., if
100 states has to be shared by 10 carriers then f ( E) = 10 /100 = 01. . The
expression that governs the distribution of electrons among the energy levels as
FREE ELECTRON THEORY 69

a function of temperature is known as Fermi-Dirac (FD) Distribution function

f ( E) and, is given by
1
f ( E) = …(3.43)
( E - E F )/ kT
e +1
· f ( E) generally indicates the probability of an energy level E to be
occupied by an electron.
· f ( E) is independent of DOS.
· f ( E) =1 if an energy state is filled by electrons.
· f ( E) = 0 if an energy state is empty.
E F is called the Fermi Energy or Fermi Level.

3.7 FERMI-DIRAC (FD) DISTRIBUTION FUNCTION

(I) Conductor
1
(a) At T = 0 : We have, f ( E) = .
( E - E F )/ kT
e +1
(i) When E < E F : E - E F = negative, and hence,
1 1
f ( E) = = = 1.
-¥
e +1 0 +1

Fig. 3.8 (a) Variation of energy with FD distribution function for metal at T = 0° K
(b) Variation of no. of energy states with energy.

So, f ( E) =1 Þ all the levels are occupied by electrons.

The variations of f ( E) and N ( E) with E are shown in Fig. 3.8(a) & 3.8(b).
(ii) When E > E F : E - E F = positive, and hence,
1 1
f ( E) = = =0
+¥
e +1 ¥
So, f ( E) = 0 ® all the levels above E F are vacant as shown in above Fig.
(iii) When E = E F : E - E F = 0, and hence, from eqn. (3.43),
1
f ( E) = Þ Indeterminable Þ Varying between 0 and 1 values.
0/ 0
e +1
70 SOLID STATE ENGINEERING PHYSICS

(b) At T > 0 K : At T > 0 K electrons are excited to vacant levels above

E F (in general E F >> kT ). Thus a few levels within the range kT below E F
become partially depleted and a few levels within the energy range kT above E F
are partially filled, but most of the deeper levels remain as they were at 0 K. As
a result of thermal excitation the probability of finding electrons at levels
E > E F becomes greater than zero and the probability at levels E < E F becomes
less than unity. The complete distribution function is shown in Fig. 3.9(a)
& 3.9(b).

Fig. 3.9 (a) Fermi distribution function at various energy levels for semiconductor
(b) No. of states variation with energy for semiconductor.

(i) At T > 0 K , E = E F : From eqn. (3.43)

1 1
f ( E) = = = 0.5
e +1 1 +1
0

· Hence, the probability of occupancy of E F at T > 0 K is always 0.5.

· E F is the average energy possessed by the electrons which participate
in conduction process in conductors at T > 0 K. The point where f ( E) =
0.5 is known as the Crossover point. At E >> E F ,
1 1
f ( E) = » = e - ( E - E F )/ kT
E - E F / kT E - EF
e +1 e / kT
Þ This is the Maxwell Boltzmann Distribution only.
(II) Insulator
The concept of E F can be extended to insulator also. As the energy levels
in the valence band (VB) are filled f ( E) is equal to unity throughout the valence
band. As there are no electrons in the conduction band (CB), f ( E) is zero
throughout the CB. Fig. 3.10(a) shows the diagram for an insulator for the
probability function.
(III) Semiconductor
A semiconductor is characterized by a CB and VB separated by a smaller
energy gap. At normal temperature a significant no. of electrons are thermally
excited from VB to CB and an equal numbers of holes are produced in the VB.
Fig. 3.10(b) aside depicts the FD distribution function of the semiconductor.
Since the probability of electron occupancy of the CB enhances at T > 0 K, f ( E)
is blurred and tapers off towards higher E values in CB. Same situation occurs
FREE ELECTRON THEORY 71

Fig. 3.10 (a) Variation of energy with f ( E ) for insulator at T = 0 K

(b) Variation of energy with f ( E ) for semiconductor at T > 0 K.

for holes near the top edge of VB. The extent of blurring of probability curve in
both the bands is equal, which indicates that the concentrations of electrons and
holes are equal. Also, f ( E) ® 0, when E increases. Electrons in the CB are
clustered very close to the bottom edge of
the band. Same is for holes in VB at the top
edge of the band. So the electrons are likely
to be located at the bottom edge of CB and
holes are at top edge of VB. The E F lies at
the middle of the energy gap as it
represents the average energy of carriers
participating in conduction.
(i) Variation of f (E) with
temperature : This variation is clearly
depicted in the Fig. 3.11 given aside. The Fig. 3.11 Variation of energy with
f ( E ) for semiconductor at different
more the T the more the electrons and holes temperatures.
will be free for conduction.

3.8 CHARACTERISTIC OF FERMI LEVEL (EF )

(i) E F is the uppermost occupied energy level in a conductor at 0 K (or
maximum energy of a metallic electron at 0 K).
(ii) E F is the average energy possessed by electrons participating in the
electrical conduction.
(iii) E F is a reference level. It is totally occupied in metals.
(iv) When the temperature of a metal is increased only those levels in the
vicinity of E F is disturbed. Some of the levels lower to E F are
vacated and some of the levels above to it are occupied by electrons.
(v) E F lies in the forbidden gap in case of semiconductors and insulators.
This is not occupied here. For semiconductor, fermi level occupancy
is 1/2.
72 SOLID STATE ENGINEERING PHYSICS

3.9 FERMI ENERGY AND MEAN ENERGY AT ABSOLUTE

ZERO TEMPERATURE
At T = 0 K, when E < E F : E - E F = negative, and hence,
1 1
f ( E) = = =1
-¥ +1
e +1 0
Then, the total number of electrons is
4pV 3/ 2 E F
ò N( E) dE = h 3 ( 2m) ò0 E dE
1/ 2
…(3.44)

(by combining the expressions of f ( E) and G ( E) dE).

So, from eqn. (3.44),
4pV 2
N= ( 2m) 3 / 2 . . ( E F ) 3 / 2
3 3
h
The number of electrons per unit volume is then given by
N 4p 2
n= = ( 2m) 3 / 2 . . ( E F ) 3 / 2
V h 3 3
2/ 3
h 2 æ 3n ö
Þ EF = ç ÷ …(3.45)
2m è 8p ø
The mean energy at absolute zero temperature is given by
1 E
E = ò F EN( E) dE
N 0
4pV E
So, E= ( 2m) 3 / 2 × ò F E 3 / 2 dE
3 0
Nh
2 4pV
E= ´ ( 2m) 3 / 2 . ( E F ) 3 / 2
5 Nh 3
2 4pV
´ ( 2m) 3 / 2 . ( E F ) 5 / 2
5 h 3
Þ E=
4pV 2
( 2m) 3 / 2 . . ( E F ) 3 / 2
3 3
h
3
Þ E = EF
5

3.10 EFFECT OF TEMPERATURE ON FERMI ENERGY

DISTRIBUTION FUNCTION
With the increase in temperature the electrons below Fermi level gains energy
and get excited. They then occupy the higher levels which were vacant earlier at
absolute zero. The number of free electrons lying in the energy interval ‘dE’ at
any temperature more than the absolute zero is given by
¥ ¥
N =ò N( E) dE = ò G ( E) f ( E) dE
0 0
4pV ¥ E 1/ 2 dE
Þ N=
h3
( 2m) 3 / 2 ò0 éE - EF ù
…(3.46)
exp ê +1
ë kT úû
FREE ELECTRON THEORY 73

Evaluating this integral by integration by parts, we have

¥ E 1/ 2 dE
I =ò
0 éE - EF ù
exp ê +1
ë kT úû
¥
½ ½ éE - EF ù
E 3 / 2 exp ê dE
½2E 3 / 2 dE ½ 2 ¥ ë kT úû
Þ I =½
3
´
é E - EF ù
½ +
3 ò0 2
½ exp ê + 1½ é éE - EF ù ù
½ ë kT úû ½0 êexp ê kT ú + 1ú × kT
ë ë û û
…(3.47)
The probability of finding an electron for both (zero and infinite energies)
are zero, so, the first term of RHS of eqn. (3.47) is zero for both the limits. The
second term can be evaluated by using Taylor’s series. According to it any
function G ( E) in the neighborhood of E = E F can be expanded in powers of
( E - E F ) as
(E - E F ) 2
G ( E) = G ( E F ) + ( E - E F ) G ¢ ( E F ) + G ¢¢( E F ) + ¼
2!
(where G¢ and G ¢¢ represent first and second differentiation)
This gives us,
3 (E - E F ) 2 3
E 3 / 2 = E F 3 / 2 + ( E - E F ) . E F 1/ 2 + E F - 1/ 2 + ¼
2 2! 4
…(3.48)
Substuting eqn. (3.48) into (3.47), we have
éE - EF ù
exp ê dE
2 ¥ ë kT úû
3kT ò0 é
I=
éE - EF ù ù
êexp ê kT ú + 1ú
ë ë û û
é 3/ 2 3 3 ù
E + ( E - E F ) . E F 1/ 2 + ( E - E F ) 2 . E F - 1/ 2 + ¼ dE…(3.49)
êë F 2 8 úû
E - EF
Let we put, x = , so that, dE = kT . dx. Also at low temperature
kT
kT << E F , the derivative f ¢ ( E) is large only at energies near E = E F as shown
in Fig. 3.12.
1
Here, f ( E) =
éE - EF ù
exp ê +1
ë kT úû
df ( E)
And f ¢ ( E) =
dE
-2
é éE - EF ù ù
êexp ê kT ú + 1ú
ë ë û û éE - EF ù
Þ f ¢ ( E) = - . exp ê
kT ë kT úû
74 SOLID STATE ENGINEERING PHYSICS

Fig. 3.12 Energy derivative of fermi distribution function and

fermi distribution at T > 0 K.
éE - EF ù
exp ê
æ 1 ö ë kT úû
Þ f ¢ ( E) = - ç ÷
è kT ø é éE - EF ù ù
2

êexp ê kT ú + 1ú
ë ë û û
For other values of E, especially for negative E, f ¢ ( E) is negligible.
Therefore the lower limit of the integral may be taken as ‘–¥’ instead of
( - E F / kT ). The above integral now becomes
2 ¥ ex é 3/ 2 3 3 ù
I= ò-¥ E + kTxE F2:/2 + ( kTx) 2 × E F-1/2 + ¼ kT dx
x 2 êë F úû
3kT ( e + 1) 2 8

2 é 3/ 2 ¥ e x dx 3 ¥ xe x dx
Þ I= êE F
3 êë ò0 ( e x + 1) 2
+
2
kTE F1/ 2 ò-¥ ( e x + 1) 2
+

3 ¥ x 2 e x dx
( kT ) 2 E F-1/ 2 ò-¥ + ¼ …(3.50)
8 ( e x + 1) 2
Using standard integrals as
¥ e x dx
ò-¥ ( e x + 1) 2
=1

¥ xe x dx
ò-¥ ( e x + 1) 2
=0

¥ x 2 e x dx p2
and ò-¥ ( e x + 1) 2
=
3
.

Eqn. (3.50) becomes

2é 1 3 p2 ù
I = êE F 3 / 2 . 2 . + 0 + ( kT ) 2 E F - 1/ 2 . 2 . + ¼ú
3ë 2 8 3 û

2 é p 2 æ kT ö
2 ù
Þ I= E F 3 / 2 ê1 + ç ÷÷ + ¼ú …(3.51)
3 êë 8 çè E F ø úû
FREE ELECTRON THEORY 75

Taking into account only upto second term of this integral as in (3.51) into
eqn. (3.46), we have
4pV 2 é p 2 æ kT ö 2 ù
N= ( 2m) 2 / 3 E F 3 / 2 ê1 + ç ÷ ú …(3.52)
h3 3 êë 8 çè E F ÷ø ú
û
At absolute zero, EF = EF 0
So eqn. (3.52) reduces to
4pV 2
N= ( 2m) 2 / 3 ´ E F 0 3 / 2 …(3.53)
3 3
h
Now substituting eqn. (3.53) into (3.52), we have
é p 2 æ kT ö
2ù
E F 0 3 / 2 = E F 3 / 2 ê1 + ç ÷÷ ú
êë 8 çè E F 0 ø úû

é p2 æ kT ö
2ù
Þ E F = E F 0 ê1 - çç ÷÷ ú …(3.54)
êë 12 è EF 0 ø úû

Eqn. (3.54) indicates that the Fermi energy is not constant but decreases
slightly with an increase in temperature. However, the value of the factor
( kT / E F 0 ) 2 is very small at room temperature and the Fermi energy is
considered to be a constant. Hence, the subscript ‘0’ is dropped from
eqn. (3.53).

3.11 WORK FUNCTION

Any metal contains a large no. of free electrons (typically electron concentration
in a metal »10 29 per meter 3 ). At normal temperature most electrons occupy the
levels up to E F . A conduction electron may move freely within the interior of
the metal. But it can not escape from the metal. The electron has to surmount the
potential barrier on the metal surface to escape from the metal. If an electron
attempts to escape it will induce a
positive charge on the surface of the
metal. So unless the electron
possesses enough energy to move
away from the region of influence of
the induced positive charge it will be
compelled to fall back the potential
well. So energy is to be supplied to it
to remove it from the metal. An
electron at the lowest level has
energy E 0 and an electron at Fermi Fig. 3.13 PE diagram of metal showing
work function.
level has energy (E F ) as in Fig. 3.13.
· The minimum energy required for an electron to be just emitted from
the metal surface is called the Work Function (F) of that particular
metal. Hence, F = E 0 - E F .
76 SOLID STATE ENGINEERING PHYSICS

3.12 ELECTRON EMISSION

· The liberation of electrons from the metal surface is known as Electron
Emission.
It is possible in the following four ways :
(a) Photoelectric Emission : Here metal is illuminated with a high
frequency light.
(b) Field Emission : Here metal is subjected with a high electric field.
(c) Secondary Emission : Here metal is bombarded with a high energy
electron.
(d) Thermoionic Emission : Here metal is heated to a high temperature,
the electrons thus emitted are also called ‘Thermions’.

3.13 THERMOIONIC EMISSION

When a metal is heated the electrons gain thermal energy (kT) and when it is
more that F the electrons are emitted from the surface. The thermal energy, in
excess of F will be converted into the KE of the electrons.
Requirements for Thermoionic Emitters : (The emitters are normally at
zero potential or at negative potential with respect to an Anode Þ Cathodes)
(a) Low work function.
(b) High melting point (m.p.), (m.p. > 1500°C) of the material.
(c) High mechanical strength. This is to withstand ion bombardment
arises by the collisions between impurity gas molecules present (even
in high vacuum) and energized electrons.
Materials of cathodes normally used :
1. Tungsten :
· F = 4.52 eV (high);
· m.p. = 3300°C;
· high mechanical strength consistent emission.
Problem : Very high F, so its operating temperature 2300°C
2. Thoriated Tungsten : Tungsten filament coated with a layer of
Thorium.
· F = 2.63 eV (not high) operating temperature > 1700°C.
3. Oxide coated cathode : Commonly used.
· F =11 . eV (not high) operating temperature » 700 – 900°C;
· A nickel ribbon or cylinder coated with a thin layer of Barium or
strontium oxides.

3.14 RICHARDSON’S EQUATION

Richardson determined the emission current density of the emitted electrons. In
the Free Electron model of the metal the electrons are free to move throughout
FREE ELECTRON THEORY 77

the volume (V) of the metal. According to FD statistics the no. of

electrons/volume having the momentum in the energy range E and E + dE is
dn = n( E) dE = f ( E) g( E) dE …(3.55)
We know, the DOS is given by,
4p
g( E) dE = ( 2m) 3 / 2 E 1/ 2 . dE …(3.56)
3
h
p2 p
However, E= Þ dE = dp
2m m
Using these into eqn.(3.56) we have,
4p p2 p 8pp 2 dp
g( E) dE = . ( 2m) 3 / 2 . . . dE = …(3.57)
h3 2m m h3
We know from eqn. (2.22)
1
f ( E) =
( E - E F )/ kT
e +1
from eqn. (3.55) and (3.57), we get
8pp 2 dp 1
dn =
3 ( E - E F )/ kT
h e +1
This is shown in Fig. 3.14 to be of a spherical shell of radius p and
thickness dp.
2 4pp 2 dp
Hence, dn = …(3.58)
h 3 e( E - E F )/ kT + 1
Now, p x , p y and pz are the
components of momentum of the
electron along X , Y and Z axes
such that p 2 = p x 2 + p y 2 + pz 2 .
Then the volume of the
momentum space lying between
px and p x + dp x , p y and
p y + dp y , dz and pz + dpz is
dp x dp y dpz = 4pp 2 dp.
Hence, the no. of electrons
per unit volume having momentum
components lie between p x and
Fig. 3.14 A solid sphere of radius p in
p x + dp x , p y and p y + dp y , pz
momentum space.
and pz + dpz is given by (using
eqn. (3.58)).
2 dp x dp y dpz
dn =
h 3 e( E - E F )/kT + 1
2 d( mv x ) d( mv y ) d( mvz )
Þ dn =
h3 e( E - E F )/kT + 1
78 SOLID STATE ENGINEERING PHYSICS

Here, v x , v y and vz are the components of velocity of the electrons along

X , Y and Z axes.
2m 3 dv x dv y dvz
Þ dn = …(3.59)
h 3 e( E - E F )/kT + 1
Above equation gives the no. of electrons per unit volume having velocity
components in the ranges between v x and v x + dv x , v y and v y + dv y , vz and
vz + dvz .
As kT » 0.3 eV, so, ( E - E F ) >> kT ; hence, from eqn. (3.59),
2m 3 dv x dv y dvz
Þ dn =
h 3 e( E - E F )/kT
2m 3
e E F / kT e - mv
2
/ 2 kT
Þ dn = dv x dv y dvz
3
h
1
- m ( vx 2 + v y 2 + v z 2 )
2m 3 E F / kT 2 kT
Þ dn = e e dv x dv y dvz
h3
2 2
1 mvx 2 1 mv y 1 mv z
2m 3 - - -
Þ dn = e E F /kT edv x e 2 kT
dv y e 2 kT
dvz …(3.60) 2 kT
3
h
Now we suppose that the surface of the metal lies in Y-Z plane and X-axis
is normal to this surface, so the no. of electrons per unit volume in velocity
range v x and v x + dv x is given by
2 2
1 mvx2 1 mv y 1 mv z
2m 3 E F /kTe - ¥ - ¥ -
Þ dn x =
h3
e 2 kT dv x ò-¥ e 2 kT dy y ò-¥ e 2 kT dvz

1 mvx 2
2m 3 - æ 2pkT ö æ 2pkT ö
Þ dn x = e E F /kT e 2 kT dv x çç ÷÷ çç ÷÷
3 m m
h è øè ø
1 mvx 2
4pm 2 kT -
Þ dn x = e E F / kT e dv x 2 kT …(3.61)
h3
When the metal is heated only those electrons can escape from the surface
along X-axis which have
1
mv x 2 ³ E
2
2E
Þ vx 2 ³ .
m
2E
Þ vx ³
m
Hence the total number of electrons leaving unit area of the surface in unit
time is
¥
nx = ò v x dn x
Ö 2E / m
FREE ELECTRON THEORY 79

1 mvx2
4pm 2 kT ¥ -
E F / kT
Þ nx =
h3
e òÖ 2 E / m e 2 kT v x dv x

We know,
é -
1 mvx2 ù
ê ¥ kT - E / kT ú
êòÖ 2 E / m
e 2 kT v x dv x = ×e
m ú
êë úû
4pmk 2 T 2
Hence, nx = e E F / kT e - E / kT
3
h
4pmk 2 T 2
Þ nx = e - f / kT …(3.62)
3
h
The emission current density ‘J’ is obtained by multiplying the no. of
electrons leaving unit area of the surface in unit time with the electronic charge.
4pmek 2 T 2
Hence, J = nx e = e - f / kT
h3
Þ J = AT 2 e - f /kT …(3.63)
4pmek 2
Here, A = Þ Constant, independent of the nature of the metal.
h3
Equation (6) is the well-known Richardson’s equation for thermionic
emission.
Experimental Verification
The emission current varies exponentially on work function and from eqn.
(3.63) we have,
J
= A . e - f /kT
2
T
æ J ö f
Þ log ç ÷ = log A -
2
èT ø kT
æ J ö fæ1 ö
Þ log ç ÷ = log A - ç ÷ …(3.64)
2
èT ø k èT ø
From the eqn. (3.64) it is evident that the plot
æ J ö
of log ç ÷ versus 1/T should be a straight line with
èT 2 ø
f
a slope - and an intercept log A. This is shown in
k Fig. 3.15 Theoretical
Fig. 3.15 aside and totally in agreement with the curve of 103 /T vs
experimental results. log ( J /T 2 ).

qqq