NATIONAL INSTITUTE OF TECHNOLOHY SIKKIM

Ravangla Campus
Barfung Block, Ravangla Sub Division, South Sikkim 737139

Subject: Semiconductor Devices (EC13103)

Module: 1

 Energy Bands:

• According to Quantum Mechanics, the energies of electrons in a free atom can not have
arbitrary values but only some definite (quantized) values.

• Example: Na11 ￫ 1s2 2s2 2p6 3s1

• The valance electrons are relatively weakly bounded, and involved in

chemical reaction.

• When two identical atoms far apart, the allowed energy level for a given quantum number
“n” consist of one doubly degenerate level.

• When two atoms brought closer, the doubly degenerate energy level splits into two level by
interaction between the atoms including forces of attraction and repulsion.

• When N atoms brought together to form solid, the result is an essentially continuous band.
1
 Energy Band in Semiconductor:

• Isolated Si atom:

Si14 ￫ 1s2 2s2 2p6 3s23p2

• Interaction of Si atom with Four shared

Si Atoms:

• N-Si atoms (Single Crystal Silicon):

2
1. r = OD (>> OA)

• Each of N atoms has its own energy levels. The energy levels are identical, sharp, discrete
and distinct.

• The outer two sub-shells (3s and 3p of M shell or n = 3 shell) of silicon atom contain two ‘s’
electrons and two ‘p’ electrons.

• There are 2N electrons completely filling 2N possible ‘s’ levels, all of which are at the same
energy.

• Of the 6N possible ‘p’ levels, only 2N are filled and all the filled p levels have the same
energy.

2. OC < r < OD

• There is no visible splitting of energy levels but there develops a tendency for the splitting
of energy levels.

3. r = OC

• The interaction between the outermost shell electrons of neighbouring silicon atoms
becomes appreciable and the splitting of the energy levels commences.

4. OB < r < OC

• The energy corresponding to the s and p levels of each atom gets slightly changed.
Corresponding to a single s level of an isolated atom, we get 2N levels. Similarly, there are
6N levels for a single p level of an isolated atom.

• Since N is a very large number (≈ 1029 atoms / m3) and the energy of each level is of a few
eV, therefore, the levels due to the spreading are very closely spaced. The spacing is ≈ 10 -23
eV for a 1 cm3 crystal.

• The collection of very closely spaced energy levels is called an energy band.

5. r = OB

• The energy gap disappears completely. 8N levels are distributed continuously. We can
only say that 4N levels are filled and 4N levels are empty.

6. r = OA

• The band of 4N filled energy levels is separated from the band of 4N unfilled energy levels
by an energy gap called forbidden gap or band gap.

3
 • The lower completely filled band (with valence electrons) is called the valence band.

• The upper unfilled band is called the conduction band.

Note:

1. The exact energy band picture depends on the relative orientation of atoms in a crystal.

2. If the bands in a solid are completely filled, the electrons are not permitted to move about,
because there are no vacant energy levels available.

 Review of Quantum mechanics:

• The idea of quantum mechanics was first introduced by Max Planck’s in the year 1900 in
an attempt to explain the observed the energy distribution of black body radiation.

• The idea was later used successfully by Einstein to explain photoelectric effect.

• Neils Bohr used a similar quantum concept to formulate a model for hydrogen atom and
explain the observed hydrogen spectra successfully.

• These developments led to following conclusions:

1. Radiant energy such as electromagnetic wave also exhibit particle nature.

2. Dynamical properties such as energy such as energy, mometum, angular momentum of

bound system has discrete values.

• The concept of dual nature of radiation was extended to particle by Louis de Broglie.

i.e. All the physical properties of a system can not be determined simultaneously with
ultimate accuracy.

• A probabilistic interpretation of the physical system in quantum mechanics was introduced

by ‘Max Born’.

i.e. Every system is characterized by a wave function Ψ, which describe the state of the
system completely.

• Wave mechanical formation of quantum theory was developed by Schrodinger, in which a

wave function satisfies a partial differential equation called ‘Schrodinger equation’.

• The matrix mechanics method to quantum mechanics was formulated by Heisenberg.

• P. A. M. Dirac formulated relativistic quantum mechanics to incorporate the effect of

special theory of relativity in quantum mechanics.
4
 This led to the development of quantum field theory which successfully describe the
interaction of radiation with matter.

 Energy of Quanta (Photoelectric Effect):

• Electrons are emitted from the surface of a

Incident Photoelect ron material by incidence of monochromatic light
monochroma tic kinetic energy under certain condition, and the emitted
light (T) electrons are called photoelectrons.

• According to classical physics, if the

intensity of light is large enough, the work
function of the material will be overcome and
Metal an electron will be emitted from the surface
independent of the incident frequency.
However, this result is not observed.

• The observed effect is that, at a constant

incident intensity, the maximum kinetic
energy of photoelectron varies linearly with
frequency with a limiting frequency  = 0 
below which no electron is produced. Tmax
• The minimum frequency of incident light
which cause photoelectric emission is called
‘threshold frequency (0)’.

• If the incident intensity varies at a constant 0 ν0 ν

frequency, the rate of photoelectron emission
changes, but the maximum kinetic energy
remains the same.

 Einstein theory of Photoelectric effect:

• Einstein in 1905 explained the photoelectric effect on the basis of Plank’s quantum
theory of light.

• Einstein interpreted the photoelectric results by suggesting that the energy in a light wave
is also contained in discrete packets or bundles called photon with energy E = h.

• According to Einstein, when light of frequency  is incident on a metallic surface, each

photon interacts with one electron and completely transfers its energy to the electron.

5
 • A photon with sufficient energy can knock an electron from the surface of the material.

• The minimum energy required to remove an electron is called ‘work function (W0)’of the
material and any excess photon energy goes into the kinetic energy of the photoelectron.

• The kinetic energy of the photoelectron is.

1 2 ………(1) Where,
Tmax  mv  hν  W 0
2
m mass of electron

v velocity of electron

 frequency of light

W0 work function

Or, Tmax  hν  hν 0 ..........(2) Where,

0 threshold frequency

This is Einstein’s photoelectric equations.

• The photoelectric effect demonstrate the particle like behavior of the photon.

Ex 1. Calculate the photon energy of X-ray of wavelength =0.708 10-8 cm.

8 10
Answer: Given; Wavelength of photon:   0.708  10 cm  0.708  10 m

Energy: E
hc

6.625  10  3  10   2.8110
34 8
15
J
 0.708  10
10

15
2.81  10
 19
1.6  10
 1.5  10 eV
4

Ex 2. Find the work function of sodium metal whose photoelectric threshold wavelength is 600
nm.
9
Answer: Given; Photoelectric threshold wavelength of light:   900nm  900  10 m
Photoelectric threshold energy of incident light:
34
6.626  10  3  10
8
hc 19
W0  E    2.21  10 J
 900  10
9
19
2.21  10
 19
eV  1.38eV
1.6  10
6
 Wave-particle duality:

In 1924 de-Broglie proposed that all the moving material particles have particle as well as wave
nature. The wave associated with the material particle is called matter wave or de-Broglie wave.

v 

m 
x

• The wavelength “” of matter wave associated with a moving particle of linear momentum
“p” is given by:
h
  ………. (3) Where,
p
h  Planck’s constant
Experimental verification:

Sample • In 1927, Davisson and Germer experimental

Incedent verified the wave nature of electron.
beam
 • The electrons from an heated filament were
θ0
Filament accelerated at normal incidence onto a single
crystal of nickel.

θ  45 • The density of scattered electrons as a function of
Galvanomet er angle was measured by a detector.

θ  90
Davisson - Germer Experiment
• The existence of a peak in the density of scattered electrons can be explained as a
constructive interference of waves scattered by the periodic atoms in the planes of nickel
crystal.

Ex 3. Calculate the de Broglie wavelength of an electron traveling with velocity 107 cm/s.

v  10 cm / s  10 m / s
7 5
Answer: Given; Velocity of electron:

Linear momentum of electron: p  mv  9.1 10 31  105

26
 9.11  10
34
h 6.625  10 9 
De-Broglie of electron:    26
 7.27  10 m  72.7 A
p 9.11  10

7
 Heisenberg Uncertainty Principle:
The law states that the product of uncertainties in position and momentum of a particle at a given
instant of time is greater or equal to ħ/2.
Mathematically,
 ……….(4) Where, x uncertainty in position
 x p 
2
 p  uncertainty in momentum
ħ  h/2 = 1.054  10-34 Js
This is known as Uncertainty relation.
• In 1927 Werner Heisenberg enunciated that “one can not measure simultaneously both
position and momentum of a particle with absolute accuracy”.
i.e. if position is determined accuracy, the determination of momentum become inaccurate
and vice versa.
 when x = 0, then p = 
• The uncertainty relation for energy and time is:
 ………….(5)
E t 
2

• The uncertainty relation for angular momentum and angle is:

 …………(6)
J 
2

Ex 4. The position and momentum of a 1.00 keV electron are simultaneously measured. If the
position is located to an accuracy of 0.200 nm, what is the percentage of uncertainty in its
momentum?
Answer: Given: Energy of electron: E  1.00keV
and x  0.1nm
According to Heisenberg Uncertainty principle:
 1.054 10 34 Js
 x p   p 
2 2  0.2 10 9 m

 2.63 1025 kgm/ s

23
Momentum of electron: p  2mE  2  9.11031 11.6 1019  1.7110 kgm/ s

p 2.63 1025
Percentage of uncertainty: 100  100  1.5%
p 1.711023
8
 Wave function:
Wave functions are the mathematical representation of a particle which contains all the
information required for the probabilistic description of the particle. It is denoted by .
• It must be well behaved, that is, single valued and continuous everywhere.
• Mathematically,

r , t   Aei k r t 

 
………..(7)
Where,
 
r, t  wave function rPosition vector at time t

A  amplitude of wave   angular frequency

 2
k  propagation vector

 Basic properties of wave function:

1. It relates the particle and wave nature of matter statistically.

2. It is the function of space coordinate and time.


i.e. r, t   x, y, z, t 
3. It is as complex quantity and hence we can not measure it.

4. It is a single valued function.


i.e.  has only one value at a point “ ”r for time “t”.

5.  and its partial derivatives i.e.   

, , must be continuous everywhere.
x y z
6.  is square integrable function i.e.   d=v finite quantity.
2

7.  must satisfy Schrodinger’s equation.

 Physical significance of :

Square of absolute magnitude of  i.e.   2 at a particular point and time is directly

proportional to the probability of finding the system at that point and time.

• The possibility of finding a particle within a volume dv is:


……...(8)

2
P dv


9
 • As the total probability of finding the particle somewhere at all times is unity.



2
So, dv  1 ……..(9)

The wave function which obeys the above condition is said to be normalized.

 Dimension of :

 
  L-3/2 for 3-dimension

  L 
-1
for 2-dimension

 
  L-1/2 for 1-dimension

 Schrodinger’s wave equation:

• In 1926 Schrodinger provided a mathematical formulation, which incorporated the

principle of quanta introduced by Planck and wave particle duality introduced by de-
Broglie, and is known as Schrodinger’s wave equation.

• The time dependent Schrodinger wave equation is given by:

 x, t      x, t 
2 2

i   V  x  x............(10)
,t
t 2m x 2

Where, x, t   wave function

V x  Potential function

• Schrodinger wave equation is the fundamental equation of quantum mechanics, just as

Newton’s 2nd law of motion of classical mechanics.

10
 Time independent Schrodinger equation:

• Let the solution be product of two functions:

x, t    x f t  ……….(11) Where, (x)  function of x

f(t)  function of t

Differentiating w. r. t. x and t:
  x, t  df …………(12)

t dt

 2   x, t  d 2
 f ….………(13)
x 2 dx 2

Using equation (12) and (13) in (11):

df  2 d 2
i  f  Vf
dt 2m dx 2

Dividing both side by f:

1 df  2 1 d 2
i   V ………..(14)
f dt 2m  dx 2

The left side of equation (14) is function of t alone, and the right hand side is a function of x
alone.

Therefore, the only way it can possibly be true is if both side are in fact constant. Otherwise, by
varying t, it could change the left side without touching the right side and two would no longer
be equal. For this reason we shall call this separation constant E.

Then,  2 1 d 2
 V  E
2m  dx 2
 2 d 2
  V  E
2m dx 2
d 2 2m
 2  2 E  V   0 ……….(15)
dx 

11
 This is time independent Schrodinger equation in 1-D.

1 df
and, i E …………(16)
f dt
1 iE
 df   dt
f 
Integrating both sides:
1 iE iE
 f df     dt  ln f  

t

iE
 t
 f  e  ………..(17)

 Infinite Potential Well:

V(x)

m: mass of electron confined to potential barrier

V=  V= 
L: length of infinite potential barriers.

Potential Function:

V ( x)  0 0 x L
0 L X
………..(18)
 0 xL

Time independent Schrodinger Equation in 1-D:

d 2 n ( x) 2m
dx 2 
 
 2 En x  V  n ( x )  0 Where, Ψn(x) electron wave function

En  energy of electron
d  n ( x) 2mEnx
2
V ( x)  0
   n ( x)  0
dx 2 2

d 2 n ( x)
  kn2x n ( x)  0 2mEnx
dx 2 …….. (19) Where, k nx  ………..(20)
2

12
Solution of equation (19):

 n ( x)  A sin k n x  B cos k n x
x x
………..(21)

Where, A and B normalized constant

Boundary Condition:

 n (0)   n ( L)  0 ……….(22)

At x = 0,  n (0)  A sin 0  B cos 0  0  B  0

At x = L,  n ( L)  A sin k nx L  0  sin k nx L  sin n

n
 k n x L  n  k nx  ……….(23)
L
 n  ……….(24)
Solution:  n ( x)  A sin  x
 L 

From equation (20) and (23):

2mEnx n n 2 2 2
  En x 
2 L 2mL2
n 2 2 2
 En x 
2mL2 ……….(25)

 2 2
E1x 
For n = 1,
2mL2  2 2

Quantum Number (n)

2mL2 9 3
  2 2
E2 x  4 
Energy in units:

For n = 2,
2mL2
 4 E1x
 2 2 4 2
For n = 3, E3 x  9 
2mL2
 9 E1x
1 1
0 x L

13
Eigen function:

According to normalization condition:

 n   n 
L L

 x xdx  1   An sin  x   An sin  x dx  1

*

0 0  L   L 

 n 
L
 An  sin
 x dx  1
2 2

0  L 
 2 n 
L 1  cos x
 An 
2  L 
dx  1
0
2
L L
 2 n  
 An   dx   cos x dx   2
2

0 0  L  
   2 n   
L

  sin  L x   
 An x 0       2
2 L
   2 n    
   L   
 0
2  L  
 An L  0   sin 0  sin 2n   2
  2n  
2  L  
 An  L   0  0  2
  2n  

 An L  2
2

2
 An 
L………( 26) (normalized constant)
 Normalization wave function:

 n 
 n x  
2
sin  x  ………(2)
L  L 

14
  
 1 x  
2 2
sin  x 
For n = 1,
L L   L
3

Quantum Number (n)

Wave function (Ψ)
 2 
 2 x  
2 9E1
For n = 2, sin  x 3
L  L 
L
 3 
 3 x  
2
For n = 3, sin  x
L  L  4E1 2
  2L
E1 1
Probability of finding the electron: 0 x L

 3 (x)
2
2 2  n 
 n x   x  …….…(28)
2
sin 
L  L 

Quantum Number (n)

Probability( 2)

2 2   9E1 3
 1 x  
2
For n = 1, sin  x 
 2 (x)
2
L L 

2 2  2 
For n = 2,  2 x  
2
sin  x 4E1 2
 1 (x )
2
L  L 

2 2  3  E1 1
For n = 3,  3 x  
2
sin  x
L  L 
0 x L

Note: The probability distribution is not uniform. This feature contradicts the expectation of
classical physics, according to which the probability of finding the particle is uniform throughout
the well.

15
 Tunneling Phenomenon:

Potential Barrier:

V0: potential barrier height

V ( x)
L: width of potential barrier.
V0
Ae ik1 x Feik1 x Potential Function:
Ce k 2 x
E x0
V ( x)  0
Be  ik1 x
De  k 2 x
III  V0 0 x L ..……(29)
I II
0 xL
0 L x

E: energy of the particle incident on the barrier.

Ψ1, Ψ2, Ψ3 : particle wave function in region I, II, and III.

Schrodinger Equations for the system:

Schrodinger Equations for three regions:

In region I (x<0):

d  1 ( x)
2 2...........(31)
mE
 k  1 ( x)  0
2 ...........(30) Where, k1  2
dx
2 1 

In region II (0<x<L):
d  2 ( x)
2
2mV0  E 
 k 2 2 ( x)  ............(32)
2
0 Where, k2  .........(33)
2
dx
2


In region I II (x>L):
d  3 ( x)
2
2mE
 k1  3 ( x)  ............(34)
2
0 Where, k1  2
dx
2 

16
General solution of equation (30), (32) and (34):
ik1x
 1 ( x)  Ae
ik1x
 Be ............(35)
 k2 x
 2 ( x)  Ce ............(36)
k2 x
 De
 ik1 x ............(37)
 3 ( x)  Fe
ik1 x
 Ge

There cannot be a wave propagating in the -x direction in region III.

G=0

 3 ( x)  Fe ………..(38)
ik1 x
Therefore,

Boundary condition at x = 0:

 1 (0)   2 (0) ............(39)

  1    2 
   
also,  x  x 0  x  x 0.............(40)

Boundary condition at x = L:

 2 ( L)   3 ( L) .............(41)

 d 2   d 3 
also,     ..............(42)
 dx  x  L  dx  x  L

Applying boundary conditions to equations (35), (36) and (38), we can find the values of
coefficients B, C, D, and F in terms of A.

Transmission co-efficient:
v  ( x)
2 2
J F E  E  2 k2 L
T t  1 t  2  16 1  e ..........(43)
v1  i ( x)
2
Ji A V0  V0 

2 k 2 L
T  e ...............(44)

17
In quantum world the particle the particle has finite probability to cross/penetrate the barrier
whatever small it may by. This process is called “quantum mechanical tunneling”.

Applications:

E The quantum mechanical tunneling phenomenon can

be applied to semiconductor device characteristics
0 L x such as in tunneling diode.

Ex 5: A beam of electrons with energy of 3 eV are incident on a potential barrier 20 eV in

height and having a width of 1 Å. Find the transmission probability of electrons.

Answer: Given: Energy of electron: E  3.0eV

Barrier height: V0  20eV

Barrier width: L  1A  0.100nm

0

E  E  2 k2 L
Transmission probability: T  16 1  e
V0  V0 

2mV0  E  2  9.110 31 20  3.01.6 10 19

Where, k2  
2 1.054 10 34

2  9.1 1031  17  1.6  1019


1.054  1034
22.25 10 25

1.054 10 34
 21.11109 m 1
Thus, transmission probability:

T  16
3  3 

1   exp  2  21.1110  0.110
20  20 
9 9

3.0  20  3 
 16   exp  4.22
20  20 
3  17 
 16     1.47 10  2
20  20 

 2.99 102
18
 Bloch Theorem:

Block has proved the important theorem that the solutions of the Schrödinger equation for a
periodic potential must be of a special form:
   
 k (r )  uk (r ) exp( ik  r ) ………(45) (Block function)

Where u k (r ) has the period of the crystal lattice with
   ………….(46)

u k (r  T )  u k (r ) : lattice
T translation vector

Bloch theorem: The eigen

  functions of the wave equation for a periodic potential are the product

of a plane wave exp( ik  r ) times a function k
u (r ) with the periodicity of the crystal lattice.

 Electrons in periodic potential: Kronig-Penny Model:

In a crystal each atom is represented by a finite square well of width a and depth V0. The atomic
spacing is a+b.

V x 
V0

P
 a  b  b 0 a ab x Q
PQ : A periodic potential of a crystal consisting of a periodic arrangement of potential wells
and potential barrier.

a : width of potential wells of periodic potential

b : width of potential barriers of periodic potential

V0 : barriers height of the periodic potential

m : Effective mass of the electron of energy E moving through the periodic potential.

19
Potential Function is given by:

V ( x)  0 0  x  a (region I)

……….(47)
 V0  b  x  0 (region II)

Time dependent Schrodinger Equation in region I and II:

Region I:

 2 ( x) 2mE
 2  ( x)  0
x 2 

 2 ( x) 2…….(49)
mE
  K 2 ( x) ………..(48)
0 Where, K
x 2
2
Region II:

 2 ( x) 2mE  V0 
  ( x)  0
x 2 2

 2 ( x) 2mV0  E 
   ( x)  0
x 2 2


d 2 ( x)
 Q 2 ( x)  ……..(50)
2mV0  E 
2
0 Where Q ………(51)
dx 2

General solution of equation (48), and (50) are:

 I ( x)  AeiKx  Be iKx …………(52) 0  x  a (region I)

 II ( x)  CeQx  De Qx …………(53)  b  x  0 (region II)

We want the complete solution to have Bloch form (50). Thus the solution in the region a < x <
a+b must be related to the solution in the region –b < x < 0 by Bloch theorem:

 (a  x  a  b)   (b  x  0)e ik a b  …….(54)

Which serves to define the wave vector k used as an index label the solutions.

The constants A, B, C, D are so chosen that  and d/dx are continuous at x = 0, and x = a.
20
At the boundary x = 0:
 I (0)   II (0)


 Ae iKx  Be iKx 
x 0 
 Ce Qx  De Qx x 0

 A B  C  D ……….(55)

  I    II 
Also,    
 x  x 0  x  x 0

 
  
 
AeiKx  Be iKx    CeQx  De Qx   
 x  x 0  x  x 0


 iKAe ik1x  iKBe ik1x 
x 0 
 QCe Qx  QDe Qx 
x 0

 iKA  iKB  QC  QD

 iK  A  B  QC  D ………..(56)

At the boundary x = a:

 I (a)   II (a)

  I (a)   II (b)e ik a b 


 Ae iKx  Be iKx xa 
 Ce Qx  De Qx  x b e ik a b 


 AeiKa  Be iKa  Ce Qb  De Qb e ik a b  ……….(57) 
  I    II 
Also,    
 x  x a  x  x  a
  I    II 
    e ik a b 
 x  x a  x  x b
 
  
 
AeiKx  Be iKx    CeQx  De Qx  e ik a b  
 x  xa  x  x b

 iKAe iKx  iKBe iKx   QCe  QDe  e  
xa
Qx  Qx
xb
ik a  b

 iKAe iKa
 iKBe iKa   QCe  QDe e  
 Qb Qb ik a b

 iK Ae iKa
 Be iKa   QCe  De e  ………..(57)
 Qb Qb ik a b

21
The four simultaneous equations (55), (56), (57), and (58) can be written compactly in matrix
form as:

 1 1 1 1   A
 iK  iK Q Q  B
    ………..(59)
0
 e iKa e iKa Qb ik ( a  b )
e e  e Qb e ik ( a b )  C 
 iKa   
iKe  iKe iKa  Qe Qb e ik ( a b ) Qe Qb e ik ( a b )   D

The equations (59) have a solution only if the determinant of coefficients A, B, C, and D
vanishes.

1 1 1 1
iK  iK Q Q
i.e.  0………..(60)
e iKa e iKa Qb ik ( a  b )
e e Qb ik ( a  b )
e e
iKe iKa  iKe iKa  Qe Qb e ik ( a b ) Qe Qb e ik ( a b )

The solution of the determinant is:

Q 2
 
 K 2 / 2QK sinh Qb sin Ka  cosh Qb cos Ka  cos k a………….(61)
 b

To simplify the equation, Kronig and Penny considered the possibility that V0b remains finite.
Such a function is called delta function.

i.e. Equation (58) is simplified if we represent the potential by periodic delta function obtained
by using the limit b0 and V0.

For b0, sinh Qb  Qb

and cosh Qb  1 ……….(62)

Using equation (62) in (61):

Q 2
 
 K 2 / 2QK Qb sin Ka  cos Ka  cos ka …………(63)

In the limit Q >> K (i.e. V0 >> E), equ (63) becomes:

Q ba / 2Kasin Ka  cos Ka  cos ka

2
…………(64)

22
  P / Kasin Ka  cos Ka  cos ka ………..(65)

Q 2 ab
Where, P   a finite quantity called scattering power
2
mV0 ab ……….(66)
P
2
The term V0b is called barrier strength.

In equation (65) there are two variables K and k. The right hand side of equation (65) is bound
since it can only assume values between -1 and +1.

 If we plot the left hand side of the equation against Ka, it will be possible to determine those
values of K (and hence energy) which are permissible.

i.e. (P/Ka)sinKa+cosKa takes values between -1 to +1

The plot of (P/Ka)sinKa+cosKa vs Ka is shown in the figure below for P = 3/2.

E
P / Kasin Ka  cos Ka

1
 3  0  3 Ka
 4  2 2 4
1
The allowed values of the energy E are given by those ranges of Ka = (2mE/ħ2)1/2a for which the
function lies between 1.

23
 • The plot of energy vs wavenumber for
the Kronig-Penney potential with P = 20
3/2 is shown .

• The energy gaps noticed at ka = , 2, 3

15

E(ħ2 2/2ma2)
…

• For Ka =  n, Equation (65) becomes:

10

cos n   cos ka
5
 ka  n
n ……..(77)
k  0
a  2 3 4
ka
These values of k are points of discontinuity in the E-k curves for electrons in the crystals.

Special Case:

1. If P (or V0b) is large, the curve proceeds more E

steeply, the allowed bands are narrow.

2. If P is smaller, the allowed bands become wider. E

24
 3. If P = 0, cos Ka  cos ka

 Ka  ka 
2mE
k
E
2

2mE
 2
 k2

2k 2
E ………..(68)
2m
 All energies are allowed, and approximate to completely free electron
sin Ka
4. If P = , 0
Ka
E
 sin Ka  0
2mE n
 Ka  n  
2 a
2mE n 2 2 n 2 2  2
 2  2 E
 a 2ma 2 ……….(69)

 Discrete allowed energies, and is the case for infinite potential well.

 Effective Mass and Bandgap

• Energy of Free electron:

p2 ……... (70)
E
2m0
Where, p: momentum

& m0: free electron mass

• E vs p plot is parabola.

• Presence of periodic potential makes equation (70) no longer valid, and replaced by:
p2 …..…..(71)
E
2mn

25
• Differentiating Equation (71) w.r.t. p:
dE 2 p p …..…..(72)
 
dp 2mn mn

Again differentiation w. r. t. p:
d 2E 1
2

dp mn
1
1  d 2E 
 mn  2  2 
 d E   dp  ……….(73)
 dp 2 
 

• Narrower the parabola, larger the second order derivative and smaller the effective mass.

• Hole effective mass (mp):

1
 d 2E  ………..(74)
mp   
2
 dp 

Where, p*: hole momentum

& E*: energy of hole

• For mn=0.25 m0 and mp=m0, E-p diagram is

shown.

• The spacing at p = 0 between these two

parabolas is the bandgap ‘Eg’.

• Effective mass concept is very useful because

it enables us to treat electrons and holes
essentially a classical charge particles.

26
Ex 6: In a crystalline solid, the band structure (E-K relation) for an electron of mass m is given
by:
 2 k 2k  3
E
2m

Determine the effective mass of the electron in the crystal.

 2 k 2k  3
Answer: Given: E
2m

Effective mass of electron:

1
2   k 2k  3 
 2

2  1
d E   1 d E 
2 1  d   
mn   2    2 
 2
1  2m  
2 
 dp    dk   dk 2 
 
 

 
1
  2 d  d 2k 2  3k 
 2  
 2m dk  dk 
1
 1 d  4k  3 
  
 2m dk  dk 
1
 1 
  4
 2m 
m

2
 Direct Bandgap semiconductor

• Semiconductors of which the lowest point of

conduction band occurred at same k value as the highest
point of the valence band are called direct bandgap
semiconductors.

• Electron transition from valence band to conduction

band occurred without a change in momentum.

• The threshold frequency of absorption by direct

transition determines the energy gap.

E g   g …..…..(75)

• Example: GaAs, InP

27
 Indirect Bandgap semiconductor

• Semiconductors of which the lowest point

of the conduction band not located at same
k value as the highest point of the valence
band are called indirect bandgap
semiconductors.

• Electron transition from valence band to

conduction band occurred through the
involvement of phonon.

• The threshold energy for indirect absorption is greater than true bandgap.
  E g   ……….(76)

• Example: Ge, Si

 Actual Band Diagram

• General Features:

1. There is a bandgap between bottom of conduction band and top of valence band.

2. Near minimum of conduction band or maximum of valence band, the E - p curves are
parabolic.

28
 • In Ge and Si:

1. Maximum of valence band occurrs at p = 0, minimum in conduction band occurs along

[100] direction at p - pc.

2. An electron transition from the maximum point in valence band to the minimum point in
conduction band requires not only an energy change (≥Eg) but also some momentum
change(≥ pc).

• In GaAs:

1. Maximum of valence band and minimum in conduction band occur at same momentum
(p = 0).

2. An electron transition from the valence band to the conduction band occurs without a
change in momentum.

Charge Carriers in Semiconductors:

• At T = 0K, each silicon atom is surrounded by eight valence electrons that are in their
lowest energy state and directly involved in covalent bonding.

i.e. All the valence electrons are in valence band and conduction band is completely
empty.

• As the temperature is raised from 0K, a few valence band electrons may gain enough
thermal energy to break the covalent bond and jump into conduction band.

29
 • The semiconductor is neutrally charged. This means that, as the negatively charged
electron breaks away from its covalent bonding, a positively charge “empty state” is
created in the original covalent bonding position in the valence band.

• As the temperature increases further, more covalent bonds are broken, and more electron
jump to conduction band and more positive “empty states” are created in the valence
band.

• If a valence electron gains a small amount of thermal energy, it may hop into the empty
state.

• Thus, the movement of a valence electron into the empty state is equivalent to the
movement of the positively charged empty state itself in the valence band.

• Hence, the semiconductor has a second equally important charge carrier called “hole”
that can give rise to a current.

 If the conduction band electron and the hole are created by excitation of valence band
electron to the conduction band, they are called an “electron-hole pair.”

• The drift current density due to electrons in the valence band is given by:

J  e  v ...........(77)
i  filled 
i

Where summation extends over all filled states.

• This summation is inconvenient since it extends over nearly full valence band and takes
into account of large number of states. Therefore, we can rewrite equation (77) in the
form:
 
J  e  v i    e  vi 

i total  i empty 

 J  e v
i total
i e  v...........(78)
i empty
i

• For completely filled band: e v

i total
i  0……….(79)

• Therefore, the drift current density is:

J  e v
i empty 
i …………(80)
 1  dE 
Where, v in the summation is: vE      ...........(81)
   dk 
30
 Fermi-Dirac Statistics:

• The probability that an electron occupies an electronic state with energy E is given by the
Fermi-Dirac distribution function, which is also called the Fermi distribution function
‘F(E)’.
1
i.e. F (E)  ...........(82)
1  exp E  E F  / kT 

Where, k  Boltzmann constant

T Absolute temperature in K

E F  Energy of Fermi level

 Density of States:

The number of allowed energy states per unit volume is known as ‘density of states’.

• The density of states for electron is given by:

3/ 2
1  2me 
De    ………..(83)
1/ 2
 2 
2   
2

Where,   Energy of the electron

• Density of states for electron at ε in conduction band is:

3/ 2
1  2me 
De     2    EC …………(84)
1 / 2
2   
2

Where, me Effective mass of the electron

• Density of states for holes at ε in conduction band is:

3/ 2
1  2m h 
Dh     2  EV   …...……(85)
1/ 2

2   
2

Where, mh Effective mass of the hole

31
 Classification of Semiconductors:

Semiconductors

Based on intensional
doping
No Yes

Extrinsic
Intrinsic
Semiconductors
Semiconductors

Based on dopant valency

compared to Host
> <

n-type p-type

 Intrinsic semiconductors:

Semiconductor in pure form is known as Intrinsic Semiconductor.

Example: Pure Si and Pure Ge

• Each Si atom has four valence electrons which are shared to form covalent bonds with
the four surrounding Si atoms.
32
• At T = 0K: No electrons are available for current conduction, i.e. valence band is
completely filled and conduction band is completely empty, and the material behaves as
insulator.

• At Room Temperature:

1. On receiving energy, one of the electrons from a covalent band breaks and is free to
move in the crystal lattice, and contribute to current conduction.

2. While coming out of the covalent bond, electron leaves behind a vacancy named ‘hole’.

3. An electron from the neighbouring atom can break away and come to the place of the
missing electron (or hole) for completing the covalent bond, which creates a hole at
another place, resulting random movement of holes.

4. The completion of a bond may not be necessarily due to an electron from a bond of a
neighbouring atom. The bond may be completed by a conduction band electron (free
electron), which is referred as ‘electron -hole recombination’.

5. Thermal excitation results number of electron in conduction band is equal to the number
of holes in valence band.

ni  pi ………(86)

where ni and pi are the intrinsic carrier concentration

• Equilibrium concentration of electrons in conduction band is:

 E  EF 
n  N C exp   C 
...........(87)
 kT 

3/ 2
 m kT 
Where, N C  2 e 2  :effective density of states in CB.
 2 
………(88)

NC: 2.86 × 1019 cm-3 for Si at 300K

NC: 4.7 × 1017 cm-3 for GaAs at 300K

33
 • Equilibrium concentration of hole in valence band is:
 E  E………..(89)
V 
p  N V exp   F 
 kT 
3/ 2
 m kT 
Where, NV  2 h 2  :effective density of states in VB.
 2 
..........(90)

NV: 2.66 × 1019 cm-3 for Si at 300K

NV: 7.0 × 1018 cm-3 for GaAs at 300K

Graphical Representation

E E E E
C.B.
EC EC

Eg EF EF

EV EV
V.B.

N(E) 0 ½ 1 F(E) n(E), p(E)

Band Diagram Density of Fermi distribution Carrier
States Function Concentration

 Intrinsic Carrier Concentration

• For Intrinsic Semiconductor:

Electron Concentration in CB = hole concentration in VB

n  p  ni …………(91)

Where, ni is intrinsic carrier concentration.

Electron Concentration in CB:

 E  EF 
n  ni  N C exp   C  …..…..(92)
 kT  34
 Hole Concentration:
 E  EV 
p  pi  NV exp   F  ………(93)
 kT 
Product of Equation (92) and (93) gives:
 E  EF   E  EV 
np  ni  N C NV exp   C  exp   F 
2

 kT   kT 

 E  EF  EF  EV 
 ni  N C NV exp   C 
2

 kT 
 E  EV 
 N C N V exp   C 
 kT 
 Eg 
 ni  N C NV exp    ………..(94)
 2kT 
For a given semiconductor at a constant temperature the intrinsic carrier concentration (ni)
does not depend on the Fermi energy.

 Temperature dependence of ni

T(˚C)
1000 500 200 100 27 0 -50
At room temperature (300K):
Intrinsic Carrier Density ni(cm-3)

1018

ni: 9.65 × 109 cm-3 for Si

1016

Si ni: 2.25 × 106 cm-3 for GaAs

1014

1012

1010 9.65×109
GaAs
108

106 2.25×106
0.5 1.5 2.5 3.5 4.5
1000/T(K-1)

35
 Ex 7: The effective density of states at the conduction band edge of Ge is 1.04×10 19 cm-3 at
room temperature 300 K. Ge has an optical band gap of 0.66 eV. Calculate the intrinsic
carrier concentration (in cm-3) in Ge at room temperature (300 K).

Answer: Given: Density of states at Conduction band Edge:

N  1.04  1019 cm 3

Energy gap of Ge: E g  0.66eV

Temperature: T  300K

For intrinsic semiconductors:

 Eg   E 
ni  N C NV exp     N exp   g 
 2kT 
 2kT   
 0.66  1.6  10 19 
 1.04  1019 exp    23

 2  1.38  10  300 
 1.04  1019 exp  12.75

 1.04  1019  2.89  10 6

 3.006  1013
 Intrinsic Fermi Level Position

• For Intrinsic Semiconductor:

Electron Concentration in CB = hole concentration in VB

ni  pi ……….(95)

 EC  EFi   E  EV 
 NC exp     NV exp   Fi 
 kT   kT 
 EC  EFi 
exp   
  kT   NV
 EF  EV  NC
exp   i 
 kT 

  EC  EFi  EFi  EV  NV
 exp   
 kT  NC

36
   EC  EV   2 EFi  NV
 exp   
 kT  NC
   EC  EV   2 EFi  N 
 ln exp    ln  V 
  kT   NC 

 EC  EV   2 EFi N 
  ln  V 
kT  NC 


2 EFi

EC  EV   ln  NV 
 
kT kT  NC 

 EFi 
EC  EV   kT ln  NV 
  ………..(96)
2  NC 2
  mh kT 3 / 2 
 2  
 EFi 
 EC  EV  kT   2 2  
 ln
2 2   me kT 3 / 2 
 2 2  
  2  
E  EV   kT ln  mh 
3/ 2

 EFi  C  
2 2  me 

 EFi 
EC  EV   3kT ln  mh 
  …………(97)
2 4  me 
For mh = me:

EFi 
EC  EV   3kT ln 1
2 4

 E Fi 
EC  EV  ……….(98) (EFi at the centre of Bandgap)
2
For mh > me:

EFi 
EC  EV   3kT  x (EFi above the centre of Bandgap)
2 4
……….(99)
m 
Where, x  ln  h  is non zero value.
 me 

37
 E E

C.B.
EC

EF
EV
V.B.

0 ½ 1 F(E)
• Intrinsic Fermi level lies at the center of the bandgap where probability of occupancy is
½.

 N-type semiconductors

N-type Semiconductor is formed when pentavalent impurity (As or P) is doped to pure

semiconductor of tetravalent element (Si or Ge).

• When Silicon (Si) is doped with arsenic (As), the four valence electrons of As form
covalent bonds with four Si atoms and the fifth electron of As atom is loosely bound.

• The energy required to detach the fifth loosely bound electron is only of the order of
0.054 eV for Si.
38
• The energy required to detach the fifth loosely bound electron is only of the order of
0.054 eV for Si, which is provided by thermal agitation.

• The force of attraction between this mobile electron and the positively charged (+5)
impurity ion is weakened by the dielectric constant of the medium.

• Such electrons from impurity atoms will have energies slightly less than the energies of
the electrons in the conduction band.

• The energy state corresponding to the fifth electron is in the forbidden gap and slightly
below the lower level of the conduction band, and the energy level is referred as donor
level.

• Carrier Concentration :

1. When intrinsic semiconductor is doped with donor impurities, not only does the number
of electrons increase, but also the number of holes decreases compared to that of
available in the intrinsic semiconductor.

2. The decrease of number of holes is caused by the enhance electron hole recombination,
due to the presence of the larger number of electrons.

3. Consequently, free electrons are the majority charge carriers and holes are the minority
charge carriers in N-type semiconductors.

4. If n and p represent the electron and hole concentrations respectively in N-type

semiconductor, then

np  ni pi  ni
2
…………(100)

Where, ni and pi are intrinsic carrier concentration.

When the concentration of electrons is increased above the intrinsic value by the addition of
donor impurities, the concentration of holes falls below its intrinsic value, making the
product np a constant.

• Position of Fermi level with donor concentration:

N 
EC  E F  kT ln  C……….(101)

 ND 

Where, ND Donor concentration

39
 • Concentration of electron in n-type semiconductor:
 E F  E Fi 
n  ni exp  
………..(102)

 kT 
Where EF Fermi level of donor

E E E E
C.B.
EC EC

Eg EF EF

EV EV
V.B.

N(E) 0 ½ 1 F(E) n(E), p(E)

Band Diagram Density of Fermi distribution Carrier
States Function Concentration

• EC - EF decrease with increasing donor impurities i.e. Fermi level moves towards CB.

 P-type semiconductors

P-type Semiconductor is formed when trivalent impurity (In or Ga) is doped to pure
semiconductor of tetravalent element (Si or Ge).

40
• When Silicon (Si) is doped with indium (In), the three valence electrons of In form three
covalent bonds with three Si atoms. The vacancy that exists with the fourth covalent
bond with fourth Si atom constitutes a hole.

• The hole which is deliberately created may be filled with an electron from neighbouring
atom, creating a hole in that position from where the electron jumped.

• Since the hole is associated with a positive charge moving from one position to another,
the semiconductor is called p – type semiconductor.

• The trivalent impurity atom is called ‘acceptor, which produces an energy level just
above the valence band, and is called ‘acceptor level’.

• The energy difference between the acceptor energy level and the top of the valence band
is of the order of meV, which is much smaller than the band gap.

• Electrons from the valence band can easily move into the acceptor level by thermal
agitation.

• In a p–type semiconductor, holes are the majority charge carriers and the electrons are the
minority charge carriers.

• If n and p represent the electron and hole concentrations respectively in p-type

semiconductor, then

np  ni pi  ni
2 ………….(103)

Where, ni and pi are intrinsic carrier concentration

• Position of Fermi level with acceptor concentration:

 NV 
E F  EV  kT ln  ……….(104)

NA 

Where, NA Acceptor concentration

• Concentration of hole in p-type semiconductor:

 EF  E Fi 
...........(105)
p  ni exp   
 kT 

Where EF Fermi level of acceptor

41
 E E E E
C.B.
EC EC

Eg
EA NA EF EF
EV EV
V.B.

N(E) 0 ½ 1F(E) n(E), p(E)

• EF-EV decrease with increasing acceptor impurities i.e. Fermi level moves towards VB.

 Law of Mass Action:

• Product of Equation (102) and (105) gives:

 E F  E Fi   E  E Fi 
np  ni exp    ni exp   F
 


 kT   kT 
 np  ni
2
...........(106) (Mass Action Law)

42
 Mobility:

 vd
A

P Q
L
I + -
E
PQ : An n-type semiconductor sample of length L, and area of cross-section A.

n: Electron concentration of semiconductor.

E: Applied electric field

v d : Drift velocity of electron

Mobility is defined as the drift velocity per unit electric field.

v
i.e. n   n ……….(107)
E
Where,  n  Electron mobility in units of cm2/V-s

v n  Drift velocity of electron

E  Applied electric field

 q 
Electron mobility:  n   c  ………(108)
 mn 

Where , m n  Effective mass of electron

q  Charge of electron

c  Mean free time

43
 vp
Mobility of hole:  p  ………..(109)
E
Where, v p  Drift velocity of hole

The negative sign is removed because the hole drifts in the same direction as the electric
field.

Note:

 The drift velocity of electron/hole is proportional to the applied electric field.

 Electron/hole mobility depends on the mean free time and the effective mass.

• Mobility is related directly to the mean free time between collisions.

 Mobility is determined by various scattering mechanism, in particular lattice scattering

and impurity scattering.

Lattice Scattering:

• Lattice scattering results from thermal vibrations of the lattice atoms at any temperature
above absolute temperature.

• These vibrations disturb the lattice periodic potential and allow energy to be transferred
between the carriers and lattice.

• Lattice vibration increases with increasing temperature.

 Lattice scattering becomes dominant at high temperatures.

 Mobility decreases with increasing temperature

• Mobility due to lattice scattering (L) is proportional to T-3/2.

i.e.  LT 3 / 2 ………..(109)

Impurity Scattering:

• Impurity scattering results when a charge carrier travels past an ionized dopant impurities
(donor and acceptor), owing to their Coulombic interaction causing deflection of charge
carrier path.

44
• The probability of impurity scattering depends on the total concentration of the ionized
impurities i.e. sum of concentration of negatively and positively charged ions.

• At higher temperatures, the carrier moves move faster, and therefore remain shorted time
near impurity atoms.

 Impurity scattering is less effective at higher temperature.

• Mobility due to impurity scattering (I) is proportional to T3/2/NT.

i.e. T T 3 / 2 / N T ………(110) Where, NImpurity

T conc.

• The probability of collision taking place per unit time is 1/c.

The sum of the probabilities of collision due to the various of scattering mechanism is:
1 1 1 ………(111)
  10 4
c  c, Lattice  c,Im purity ND=1014cm-3

T3/2 T -3/2

LOGn
1 1 1
or,   ...…….(112)
 L I n(cm2-V - s) Impurity Lattice
Scattering Scattering
• For Lightly doped samples ~ 10 cm : 14 -3 1016
LOGT
10 3
1017
- The lattice scattering dominates and mobility
decreases with increase in temperature. 1018

• For highly doped samples ~ 1019 cm-3:

1019
10 2
- The effect of impurity scattering is more
pronounced at low temperature, and the 50
100 200 500 100
mobility increases with increase in
temperature. T(K)
- For a given temperature mobility decreases with increasing impurity concentration.

45
 Mobility (cm2/V-s) Mobility (cm2/V-s) 2000

Diffusivity (cm2/s) Diffusivity (cm2/s)

50
1000 Si • Mobility reaches a maximum value at
n,Dn 20
500 low impurity concentration.
10
200 5
100 • Both electron and hole mobilities
2 decreases with increasing impurity
50 p,Dp
1
concentration and eventually
20
approach a minimum value at high
10000
200 concentrations.
5000 GaAs
100
2000 50 Note: The mobility of electrons is
n,Dn
1000 20 greater than that of holes. Greater
500 electron mobility is mainly due to
p,Dp 10
200 5 smaller effective mass of electrons.
100
1014 1015 1016 1017 1018 1019 1020
Impurity concentration(cm-3)

Ex 8: Calculate the mean free time of an electron having a mobility of 1000 cm2/V-s at 300 K;
also calculate the mean free path. Assume mn=0.26 m0 in these calculations. (given vth = 107
cm/s)

Answer: Given: Electron mobility:  n  1000cm 2 / V  s

Effective mass of electron: mn  0.26m0  0.26  9.1 10 31 kg

 2.366  10 31 kg
mn  n 2.366  10 31  1000
Mean free time:  c  
q 1.6  10 19
 1.48  10 13 s  0.148 ps
Mean free path: l  v th c  10 7  1.48  10 13 cm
 v th  10 7 cm / s 
 1.48  10 6 cm
 14.8nm

46
 Conductivity and Resistivity:

P Q
L
I
+ E -

PQ : An n-type semiconductor sample of length L, and area of cross-section A.

n: Electron concentration of semiconductor.

E: Applied electric field

The electron current density Jn flowing through the sample is sum of the product of charge
(q) on each electron times the electron velocity over all electrons per unit volume (n).
n
I
i.e. Jn     qv i Where, I Electron current
A i 1
 J n  qnv n
 J n  qn n E ……….(114)  v n   n E
Similar argument applies to holes, and hole current density is given by:

J p  qp p E ……….(115) where, p hole concentration

Total current density: J  J n  J p  qn n  p p E

 J  E ………(116)

The quantity  is called the conductivity of the semiconductor and given as:

  q n n  p p  ………..(117)

The resistivity () of a semiconductor is defined as the reciprocal of conductivity.

1 1
 
qn n  p p 
Thus, .………(118)


47
 1
For n-type semiconductor n >> p:    ……….(119)
qn n

1
For p-type semiconductor p >> n:    ………..(120)
qp p

Ex 9: Statements for linked Answer Questions (i) and (ii): The atomic density of a solid is
5.85×1028 m-3. Its electrical resistivity is 1.6×10-8 Ωm. Assume that electrical conduction is
described by the Drude model (classical theory), and that each atom contributes one electron.

Q (i): The drift mobility (in m2/V-s) of the conduction electron is: ______.

Q (ii): The relaxation time (mean free time), in second of the conduction electron is: ______.

Answer: (i) Given: Atomic density of solid:

1
Electrical resistivity:    1.6  10 m
8


1
  1.6  10 8 m
ne n
1 1
 n  8 
ne  1.6  10 5.85  10  1.6  10 19  1.6  10 8
28

1

5.85  10  1.6  10 19  1.6  10 8
28

 6.67  10 3 m 2 / V  s
 n m0 6.67  10 3  9.1  10 31
(ii) Mean free time:  
e 1.6  10 19
 3.79  10 14 s

48
  Diffusion:

In a semiconductor with a spatial variation of carrier concentration, the moment of carriers

from a region of high concentration to a region of low concentration is called diffusion.

X=a
Current

A B

Electron
nl 
Electron density n(x)

n0

n l 

l 0 l
Distance x

AB : An n-type semiconductor at uniform temperature in which the electron concentration n(x)

varies in x-direction.

n(0), n(l ), n(l ) : Electron concentrations at x = 0, x = -l, and x = l.

l : Mean free path of the electron  v th c

49
v th : Thermal velocity

 c : Mean free time

The electrons at x = -l, (one mean free path away on left side of x = 0), have equal chances of
moving left or right.

 In a mean free time c, one half of them will move across the plane x = 0.

 The average rate of electron flow per unit area F1 of electrons crossing the plane x = 0
from the left is:

The average rate of electron flow per unit area F1 of electrons at x = - l crossing the plane x = 0
from the left is:
n l   l
1
………(121)
 n l v th
F1  2 1
c 2

The average rate of electron flow per unit area F2 of electrons at x = l crossing the plane x = 0
from the right is:
nl   l
1
 nl v th
2 1 ………(122)
F1 
c 2

The net rate of carrier flow from left to right is:

F  F1  F2  1 v th n l   nl 
2
1  dn   dn  
 v th n0  l   n0  l  
2  dx   dx   (By Taylor series expansion)

1  dn 
v th n0   n0   l
dn
 l 
2  dx dx 
1  dn 

v th  2l 
2  dx 
dn
  v th l
dx
dn
  Dn ………(123)
dx
Where, Dn  v thlDiffusion coefficient or diffusivity

50
Since each electron carriers has charge –q, the carrier flow gives rise to a current:
dn
J n   qF  qDn …………(124)
dx
 The diffusion current is proportional to the spatial derivative of the electron density.

Ex 10: Assume that, in an n-type semiconductor at T = 300 K, the electron concentration varies
linearly from 11018 to 71017 cm-3 over a distance of 0.1 cm. Calculate the diffusion current
density if the electron diffusion coefficient is Dn = 22.5 cm2/s.

Answer: Given: Variation of Electron concentration over a length 0.1 cm :

 
n  1 1018  7  1017 cm 3
 10  1017

 7  1017 cm 3
 3  1017 cm 3

Change in length: x  0.1cm

Diffusion coefficient: Dn  22.5cm 2 / s

Diffusion current density:

dn n
J n  qDn  qDn
dx x
 3  1017 
 
 1.6  10 19  22.5   
 0.1 
 10.8 A / cm 2

 Einstein Relation:

From equipartition theorem for the kinetic energy of the particle in one-dimension is:
1 1
m n v th  kT ………(125)
2

2 2
kT ……….(126)
 v th 
2

mn

51
 Thus, diffusion coefficient or diffusivity becomes:
Dn  v thl  v th v th c   v 2th c

 kT   n mn ………(127)

 Dn    
 mn  q 

 kT 
 Dn     n ………..(128)
 q 
This is known as ‘Einstein Relation’.

Ex 11: Minority carriers (holes) are injected into a homogeneous n-type semiconductor sample
at one point. An electric field of 50 V/cm is applied across the sample and the field moves these
minority carriers a distance of 1 cm in 100 s. Find the drift velocity and the diffusivity of the
minority carriers.

Answer: Given; Mean free length: l  1cm

Mean free time:  c  100s

Applied field: E  50V / cm
1cm
Drift velocity: v p   10 4 cm / s
100  10 6 s

vp 10 4
Mobility: μp    200cm 2 / V  s
E 50

23
 p  1.38  10 19300  200
kT
Diffusion coefficient: Dp 
q 1.6  10
 0.0259  200
 5.18cm 2 / s

52
 Current density Equation:

• When an electric field is present in addition to a concentration gradient, both drift current
and diffusion current will flow.

 The total current density at any point is sum of the drift and diffusion components.
dn
For Electrons: J n  q n nE  qDn ………(129)
dx
where, E Electric field in the x-direction.
dp
For holes: J p  q p pE  qD p ……….(130)
dx
The -ve sign in equation (164) is because for a positive hole gradient the holes will diffuse in
the negative x-direction.

Thus, the total conduction current density is given by:

J cond  J n  J p ……….(131)

• The three equations (129), (130), and (131) constitute current density equations.

• The current density equations are important for analyzing device operations under low
electric fields.

 Generation and Recombination process:

• In thermal equilibrium: pn = ni2

• In nonequilibrium situation: pn > ni2

i.e. Excess carriers are introduced into a semiconductors.

The process of introducing excess carriers into a semiconductor is called ‘carrier injection’.

• Most semiconductor devices operates by the creation of charge carriers in excess of the
thermal equilibrium values.

• When the thermal-equilibrium condition is disturbed (i.e. pn  ni2), processes exist to

restore the system to equilibrium (i.e. pn = ni2) through recombination of injected
minority carriers with majority carriers.

 Recombination is a mechanism to restore equilibrium through which the injected

minority carriers recombine with the majority carriers, in the case of the injection of
excess carriers.

53
  Recombination process is classified in to two types depending on whether energy is
released in the form of a photon or dissipated as heat to lattice after recombination.

i. Radiative recombination: The process in which a photon is emitted after recombination

of injected minority carriers with majority carriers.

- Dominates in direct bandgap semiconductors

ii. Nonradiative recombination: The processed in which energy is dissipated as heat to

lattice after recombination.

- Dominates in indirect bandgap semiconductors

 Direct recombination:

EC EC

hν
Gth Rth GL Gth R

EV EV

• In thermal equilibrium, the continuous thermal vibration of lattice atoms causes some
bonds between neighbouring atoms to be broken and, therefore an electron-hole pair is
generated.

 The thermal energy enables a valence electron to make an upward transition to conduction
band, leaving a hole in the valence band. The process is called carrier generation.

• The carrier generation is represented by generation rate Gth (i.e. number of e-h pairs
generated per cm3 per second).

• When an electron makes a transition down ward from the conduction band to the valence
band, an electron hole pair is annihilated. This reverse process is called recombination,
and is represented by the recombination rate Rth.

• Under thermal equilibrium condition, the generation rate must be equal to the
recombination rate, so that the carrier concentration remains constant and the condition
pn = ni2 is maintained.
54
• In a direct bandgap semiconductor, the bottom of the conduction band is lined up with
the top of the valence band.

• No additional momentum is required for transition across the bandgap.

• When excess carriers are introduced to a direct bandgap semiconductor, the probability
is high that electrons and holes will recombine directly.

• The rate of direct recombination R is proportional to the number of electrons available in

conduction band and the number of holes available in valence band.

i.e. R   np …………(132) Where,  Proportionality const.

In thermal equilibrium, the recombination rate must be balanced by generation rate, and
therefore:

Gth  Rth  nno pno……….(133) (For n-type semiconductor)

Where, nno , p no  Electron and hole densities of n-type semiconductor in thermal

equilibrium.

When light is incident on the semiconductor to produce electron-hole pair at a rate GL,
the carrier concentrations are above their equilibrium values.

 Recombination rate: R  nn pn   nn 0  n  pno  p………..(134)


and Generation rate: G  GL  Gth ……….(135)

Where, n, p  Excess carrier concentration and given by:

n  nn  nno ………(136)

p  pn  pno ………(137)

and, n  p to maintain overall charge neutrality.

The net rate of change of hole concentration is given by:

dpn
 G  R  GL  Gth  R ………(138)
dt
In steady state dpn/dt = 0. Thus from equation (138):

GL  Gth  R  0

 GL  R  Gth  U …………(139)

Where, U  net recombination rate

55
Substituting equation (135) and (136) in (139) yields:

U   nn 0  n  pno  p   nn 0 pn 0

  nn 0 pn0  nn 0 p  pno n  np   nn0 pn 0

  nn 0 p  pno n  np 
 U   nn 0  pno  p p ………(140)  n  p
For low level injection, p,pno << nno, thus equation (140) becomes:

U  nn0 p
p n  p no ………..(141)
U 
1
nn 0
Therefore, the net recombination rate is proportional to the excess minority carrier concentration.

The proportionality constant 1/nno is called the lifetime p of the excess minority carriers.
p n  p no 1
Thus, U ……..(142) Where,  p ……….(143)
p  n no

The physical meaning of lifetime can best be illustrated by the transient response of a device
after the sudden removal of the light source.

hν

A B
x

pn t 
p n 0
p GL

pno
0 P t
AB : An n-type semiconductor illuminated with light, which generated e - h pairs uniformly
throughout the sample with generation rate GL.

56
In steady state, from equation (139), and (142):
p n  p no
GL  U  ………..(144)
p
 p n  p no  G L p ………..(145)

If at an arbitrary time (say t=0), the light is suddenly turned off.

Thus, p n t  0  p no  G L p ……….(146)

and pn t     pno …………(147)

The net rate of change of hole concentration:

dpn
 Gth  R  U
dt
dp p  p no
 n  n
dt p ……….(148)
dp n 1
   dt …………(149)
p n  p no p

Integrating both sides from t = 0 to t = t:

t t
dp n 1
   dt  ln  pn  pno t0   t t0
1
0
p n  p no p 0 p

 ln
 pn (t )  pno    t
 pn (0)  pno   p
 p (t )  p no   e 
t

 n p

 p n (0)  p no 
t

 p n (t )  p no   p n (0)  p no e
p

t

p
 pn (t )  pno   p GL e ………..(150)

The minority carriers recombine with the majority carrier and decay exponentially with time
constant p.

57
 Ex 12: A Si sample with nno = 1014 cm-3 is illuminated with light and 1013 electron-hole
pairs/cm3 are created every microsecond. If n=p = 2 s, find the change in minority carrier
concentration.

Answer: Given, Equilibrium electron conc.: nno  1014 cm 3

Electron hole pair generation rate: GL  1013 cm 3

Equilibrium hole conc. before illumination: 

p no  ni2 / nno  9.65  10 9 
2
/ 1014

 9.31  10 5 cm 3

Lifetime:  n   p  2s
13
After illumination: p n  p no   p G L  9.31  10 5  2  10 6  10
1  10 6

 2  1013 cm 3
 Continuity Equation:

The fundamental law governing the flow of charge is called the ’continuity equation’.

• The continuity equation describes how the carrier concentration in a given elemental
volume of the crystal varies with time and distance.

• The variation in density is attribute to two basic causes:

(i)The rate of generation and the loss by recombination of carrier within the element.

(ii) Drift of carriers into out of the element.

• Thus the overall rate of electron increase is the algebraic of four components:

P Q Area A
dx

J n x 
Rn J n x  dx 

Gn
x x  dx

PQ : An infinitesimal slice of a semiconductor of thickness dx and cross-sectional area A.

58
J n x  : Current density of electrons entering into the slice at distance x.

J n x  dx  : Current density of electrons coming out of the slice at distance x+dx.

Gn , Rn : Generation and recombination rate of electrons and holes in the slice.

• The overall rate of electron increase is the algebraic of four components: the number of
electrons flowing into the slice at x, minus the number of electrons flowing out at x+dx,
plus the rate at which electrons are generated, minus the rate at which they are
recombined with holes in the slice.

n  J x A J n x  dx A 
Thus,
t
Adx   n    Gn  Rn Adx ..........(151)
 -q -q 
n  J x A A  J x  
 Adx   n  J n x   n dx  Gn  Rn Adx
t  q q x 
n J n x 
   Gn  Rn  ..........(152) (Continue equation for electrons)
t x

p J p x 
Similarly,   G p  R p  ……….(153) (Continue equation for holes)
t x

• Now substituting current expressions from equations (129) and (130) and recombination
expression from equation (142) in equation (152) and (153) we get:
 n p 
 n p  p E  D n 
n p  x  n p  n p0
  Gn 
t x 
n p n p  np n p  n p0
2
E ………..(154)
  n p n  nE  Dn  Gn 
t x x x
2
n

p n p  pn p  pn0
2
E
and,  pn  p   pE n  Dp  Gp  n ……….(155)
t x x x
2
p

• Equations (154) and (155) are called continuity equations for electrons and holes.

59
Quasi-Fermi Energy Level:

• If excess carriers are created in a semiconductor, we are no longer in thermal equilibrium

and the Fermi energy is strictly no longer defined.

• However, we may define a quasi-Fermi level for electrons and a quasi-Fermi level for
holes that apply for nonequilibrium.

• If n and p are the excess electron and hole concentrations respectively, then we may
write:
 E Fn  E Fi 
n  n  ni exp  

...........(156)
 kT 
 E Fp  E Fi 
and, p  p  ni exp    ...........(157)

 kT 
Where, E Fn , E Fp Quasi-Fermi energy levels for electrons and holes.

0.2982eV 0.2982eV 0.2982eV

EC EC
E Fn
EF EF
E Fi E Fi
E Fp
0.2982eV
EV EV

Equilibriu m Energy Band Nonequilib rium Energy Band

diagram diagram

• Since the majority carrier electron does not change significantly for this low injection
condition, the quasi-Fermi level for electrons is not much different from the thermal
equilibrium Fermi level.
• The quasi-Fermi energy level for the minority carrier holes is significantly different from
the Fermi level and illustrate the fact that we have divided the thermal equilibrium
significantly.
• Since the electron concentration has increased, the quasi-Fermi levels for electrons has
move slightly closer to the conduction band.
• Similarly, the hole concentration has increased significantly so that the quasi Fermi level
for holes has moved much closer to the valence band.

60