Introduction to the Quantum

Theory of Solids

Dr. Khun kimleang

1
 Objective
➢ Develop the concept of allowed and forbidden electron
energy bands in a single-crystal material, and describe
conduction and valence energy bands in a
semiconductor material.
➢ Discuss the concept of negatively charged electrons
and positively charged holes as two distinct charge
carriers in a semiconductor material.

2
 Objective
➢ Develop electron energy versus momentum curves in a
single-crystal material, which yields the concept of direct
and indirect bandgap semiconductor materials.
➢ Discuss the concept of effective mass of an electron and a
hole.
➢ Derive the density of quantum states in the allowed
energy bands.
➢ Develop the Fermi-Dirac probability function, which
describes the statistical distribution of electrons among the
3
allowed energy levels, and define the Fermi energy level.
 ALLOWED AND FORBIDDEN ENERGY
BANDS
➢ We can extrapolate these single-atom results to a crystal
and qualitatively derive the concepts of allowed and
forbidden energy bands.
➢ We will then apply quantum mechanics and
Schrodinger’s wave equation to the problem of an
electron in a single crystal.
➢ We find that the electronic energy states occur in bands of
allowed states that are separated by forbidden energy
bands. 4
 Formation of Energy Bands

Figure 2.11 | The radial probability density function for the one-
electron atom in the (a) lowest energy state an
5
 Formation of Energy Bands

(a) (b)

(c)

Figure 3.1 | (a) Probability density function of an isolated hydrogen

atom. (b) Overlapping probability density functions of two adjacent
hydrogen atoms. (c) The splitting of the n 1 state. 6
 Formation of Energy Bands

➢ Figure 3.1a shows the radial probability density function

for the lowest electron energy state of the single,
noninteracting hydrogen atom, and
➢ Figure 3.1b shows the same probability curves for two
atoms that are in close proximity to each other. The wave
functions of the electrons of the two atoms overlap,
which means that the two electrons will interact.

7
 Formation of Energy Bands
➢ This interaction or perturbation results in the discrete
quantized energy level splitting into two discrete energy
levels, schematically shown in Figure 3.1c .
➢ The splitting of the discrete state into two states is
consistent with the Pauli exclusion principle.

8
 Formation of Energy Bands
➢ Now, if we somehow start with a regular periodic
arrangement of hydrogen-type atoms that are initially
very far apart, and begin pushing the atoms together,
the initial quantized energy level will split into a band
of discrete energy levels.
➢ This effect is shown schematically in Figure 3.2 , where
the parameter r0 represents the equilibrium interatomic
distance in the crystal.
9
 Formation of Energy Bands

Figure 3.2 | The splitting of an energy state into a band of allowed

energies.
10
 Formation of Energy Bands

➢ At the equilibrium interatomic distance, there is a band

of allowed energies, but within the allowed band, the
energies are at discrete levels.
➢ The Pauli exclusion principle states that the joining of
atoms to form a system (crystal) does not alter the total
number of quantum states regardless of size.

11
 Formation of Energy Bands

Figure 3.3 | Schematic showing the splitting of three energy states

into allowed bands of energies. 12
 Formation of Energy Bands
➢ Suppose the atom in this imaginary crystal contains
electrons up through the n = 3 energy level. If the atoms
are initially very far apart, the electrons in adjacent atoms
will not interact and will occupy the discrete energy levels.
➢ If these atoms are brought closer together, the outermost
electrons in the n = 3 energy shell will begin to interact
initially, so that this discrete energy level will split into a
band of allowed energies.

13
 Formation of Energy Bands
➢ If the atoms continue to move closer together, the
electrons in the n=2 shell may begin to interact and will
also split into a band of allowed energies.
➢ Finally, if the atoms become sufficiently close together,
the innermost electrons in the n=1 level may interact, so
that this energy level may also split into a band of
allowed energies.
➢ The splitting of these discrete energy levels is
qualitatively shown in Figure 3.3 14
 Formation of Energy Bands
➢ If the equilibrium interatomic distance is r0 , then we
have bands of allowed energies that the electrons may
occupy separated by bands of forbidden energies.
➢ This energy-band splitting and the formation of allowed
and forbidden bands is the energy-band theory of single-
crystal materials.

15
 Formation of Energy Bands

➢ As the interatomic distance decreases, the 3s and 3p

states interact and overlap.
➢ At the equilibrium interatomic distance, the bands have
again split, but now four quantum states per atom are in
the lower band and four quantum states per atom are in
the upper band.

16
 Formation of Energy Bands
➢ At absolute zero degrees, electrons are in the lowest
energy state, so that all states in the lower band (the
valence band) will be full and all states in the upper band
(the conduction band) will be empty

➢ The bandgap energy Eg between the top of the valence

band and the bottom of the conduction band is the width
of the forbidden energy band.

17
 Formation of Energy Bands

(a) (b)

Figure 3.4 | (a) Schematic of an isolated silicon atom. (b) The

splitting of the 3s and 3p states of silicon into the allowed and
forbidden energy bands.
18
 The Kronig–Penney Model
➢ In the previous section, we discussed qualitatively the
splitting of allowed electron energies as atoms are brought
together to form a crystal.
➢ The concept of allowed and forbidden energy bands can be
developed more rigorously by considering quantum
mechanics and Schrodinger’s wave equation.
➢ It may be easy for the reader to “get lost” in the following
derivation, but the result forms the basis for the energy-
band theory of semiconductors.
19
 The Kronig–Penney Model
➢ The potential function of a single, noninteracting, one-
electron atom is shown in Figure 3.5a . Also indicated on
the figure are the discrete energy levels allowed for the
electron.
➢ Figure 3.5b shows the same type of potential function for
the case when several atoms in close proximity are
arranged in a one-dimensional array.
➢ The potential functions of adjacent atoms overlap, and the
net potential function for this case is shown in Figure 3.5c
20
The Kronig–Penney Model

21
 The Kronig–Penney Model

Figure 3.5 | (a) Potential function of a single isolated atom. (b)

Overlapping potential functions of adjacent atoms. (c) Net potential
function of a one-dimensional single crystal. 22
 The Kronig–Penney Model

Figure 3.6 | The one-dimensional periodic potential function of the

Kronig–Penney model.

23
 The Kronig–Penney Model
➢ To obtain the solution to Schrodinger’s wave equation,
we make use of a mathematical theorem by Bloch.
➢ The theorem states that all one-electron wave functions,
for problems involving periodically varying potential
energy functions, must be of the form

24
 The Kronig–Penney Model
➢ We stated in Chapter 2 that the total solution to the wave
equation is the product of the time-independent solution
and the time-dependent solution, or

➢ This traveling-wave solution represents the motion of an

electron in a single-crystal material. The amplitude of the
traveling wave is a periodic function and the parameter k
is also referred to as a wave number. 25
 The Kronig–Penney Model
➢ We can now begin to determine a relation between the
parameter k , the total energy E , and the potential V0.
Region I in Figure 3.6

➢ The function u1(x) is the amplitude of the wave function

in region I and the parameter is defined as

26
 The Kronig–Penney Model
Consider now a specific region II,

where u2(x) is the amplitude of the wave function in region

II. We may define

so that Equation (3.6) may be written as

27
 The Kronig–Penney Model
❑ if E > V0 , the parameter 𝛽 is real, whereas if E < V0 ,
then 𝛽 is imaginary.

❑ The solution to Equation (3.4), for region I, is of the form

and the solution to Equation (3.8), for region II, is of the

form

28
 The Kronig–Penney Model
❑ This continuity condition implies that the wave amplitude
function u(x) and its first derivative 𝜕𝑢(𝑥)/𝜕𝑥 must also
be continuous.
❑ If we consider the boundary at 𝑥 = 0 and apply the
continuity condition to the wave amplitude, we have

29
 The Kronig–Penney Model
❑ Now applying the condition that

❑ We have considered region I as 0 < 𝑥 < 𝑎 and region II

as −𝑏 < 𝑥 < 0.

30
 The Kronig–Penney Model
❑ The last boundary condition is

❑ In order for there to be a nontrivial solution to the above

four equations for A, B, C, and D, the determinant of
their coefficients must vanish. This gives us the
determinantal equation
31
 The Kronig–Penney Model

❑ The evaluation of this determinant is extremely

laborious and will not be considered in detail. The
result is

32
 The Kronig–Penney Model
❑ The more interesting solutions occur for E<V0 , which
applies to the electron bound within the crystal. From
Equation (3.7), the parameter 𝛽 is then an imaginary
quantity. We may define
𝛽 = 𝑗𝛾
❑ where 𝛾 is a real quantity. Equation (3.19) can be written
in terms of as

33
 The Kronig–Penney Model
❑ To obtain an equation that is more susceptible to a
graphical solution and thus will illustrate the nature of
the results, let the potential barrier width b → 0 and the
barrier height V0 →∞ , but such that the product bV0
remains finite. Equation (3.21) then reduces to

34
 The Kronig–Penney Model

Eq. (3.24) again gives the relation between the parameter

k, total energy E (through the parameter 𝛼 ), and the
potential barrier bV0.

35
 The Kronig–Penney Model

❑ We may note that Eq.(3.24) is not a solution of

Schrodinger’s wave equation but gives the conditions for
which Schrodinger’s wave equation will have a solution.

❑ If we assume that the crystal is infinitely large, then k in

Eq.(3.24) can assume a continuum of values and must be
real.

36
 The k -Space Diagram
❑ To begin to understand the nature of the solution,
initially consider the special case for which V 0 = 0. In
this case P’ = 0, which corresponds to a free particle
since there are no potential barriers. From Equation
(3.24), we have that

37
 The k -Space Diagram

❑ where p is the particle momentum. The constant of the

motion parameter k is related to the particle momentum
for the free electron. The parameter k is also referred to
as a wave number. We can also relate the energy and
momentum as

38
 The k -Space Diagram

➢ Figure 3.7 shows the parabolic relation of Equation (3.28)

between the energy E and momentum p for the free
particle.
➢ Since the momentum and wave number are linearly
related, Figure 3.7 is also the E versus k curve for the free
particle.

39
 The k -Space Diagram

Figure 3.7 | The parabolic E versus k curve for the free electron.

40
 The k -Space Diagram
❑ We now want to consider the relation between E and k
from Equation (3.24) for the particle in the single-crystal
lattice.
❑ As the parameter P increases, the particle becomes more
tightly bound to the potential well or atom.
❑ We may define the left side of Equation (3.24) to be a
function f(a𝛼), so that

41
 The k -Space Diagram
➢ Figure 3.8a is a plot of the first term of Equation (3.29)
versus a𝛼.
➢ Figure 3.8b shows a plot of the cosa𝛼 term and Figure
3.8c is the sum of the two terms, or f (a𝛼).
➢ Now from Equation (3.24), we also have that

For Equation (3.30) to be valid, the allowed values of the f

(a𝛼) function must be bounded between +1 and -1. Figure
3.8c shows the allowed values of f (a𝛼) and the allowed
42
values of a𝛼 in the shaded areas
 The k -Space Diagram

Figure 3.8 | A plot of (a) the first term in Equation (3.29), (b) the
second term in Equation (3.29), and (c) the entire f (a𝛼) function.
43
 The k -Space Diagram

The shaded areas show the allowed values of (a) corresponding to real
values of k .
44
 The k -Space Diagram
❑ The parameter 𝛼 is related to the total energy E of the
particle through Equation (3.5), which is 𝛼2 = 2𝑚𝐸/ħ2 .
❑ A plot of the energy E of the particle as a function of the
wave number k can be generated from Figure 3.8c .
❑ Figure 3.9 shows this plot and thus shows the concept of
allowed energy bands for the particle propagating in the
crystal lattice.
❑ Since the energy E has discontinuities, we also have the
concept of forbidden energies for the particles in the crystal.
45
 The k -Space Diagram

Figure 3.9 | The E versus k diagram generated from Figure 3.8. The
allowed energy bands and forbidden energy bandgaps are indicated.
46
 The k -Space Diagram
Objective: Determine the width (in eV) of a forbidden
energy band. Determine the width of the forbidden bandgap
that exists at ka=𝜋 (see Figure 3.9 ). Assume that the
coefficient P = 8 and the potential width is a = 4.5 Å .

47
 The k -Space Diagram
❑ Consider again the right side of Equation (3.24), which
is the function cos ka. The cosine function is periodic
so that

where n is a positive integer.

❑ We may consider Figure 3.9 and displace portions of the
curve by 2. Mathematically, Equation (3.24) is still
satisfied. Figure 3.10 shows how various segments of
the curve can be displaced by the 2 factor.
48
 The k -Space Diagram
❑ Figure 3.11 shows the case in which the entire E versus
k plot is contained within −π/𝑎 < 𝑘 < 𝜋/𝑎 . This plot
is referred to as a reduced k -space diagram, or a
reduced-zone representation.

49
 The k -Space Diagram

Figure 3.10 | The E versus k diagram showing Figure 3.11 | The E versus k diagram
2 displacements of several sections of allowed in the reduced-zone representation.
energy bands.
50
 The k -Space Diagram
❑ We noted in Equation (3.27) that for a free electron, the
particle momentum and the wave number k are related
by 𝑝 = 𝑘ħ .
❑ The similarity between the free electron solution and the
results of the single crystal shown in Figure 3.9 , the
parameter 𝑘ħ in a single crystal is referred to as the
crystal momentum.
❑ This parameter is not the actual momentum of the
electron in the crystal but is a constant of the motion that
51
includes the crystal interaction.
 The k -Space Diagram

❑ We have been considering the Kronig–Penney model,

which is a one- dimensional periodic potential function
used to model a single-crystal lattice.
❑ The principal result of this analysis, so far, is that
electrons in the crystal occupy certain allowed energy
bands and are excluded from the forbidden energy
bands.

52
 The k -Space Diagram

❑ For real three dimensional single-crystal materials, a

similar energy-band theory exists.
❑ We will obtain additional electron properties from the
Kronig–Penney model in the next sections

53
TYU

54
 ELECTRICAL CONDUCTION IN
SOLIDS
❑ Again, we are eventually interested in determining the
current–voltage characteristics of semiconductor devices.
❑ We will need to consider electrical conduction in solids
as it relates to the band theory we have just developed.
❑ Let us begin by considering the motion of electrons in
the various allowed energy bands.

55
 The Energy Band and the Bond
Model
❑ Figure 3.12 shows a two dimensional representation of
the covalent bonding in a single-crystal silicon lattice.
❑ This figure represents silicon at T= 0 K in which each
silicon atom is surrounded by eight valence electrons that
are in their lowest energy state and are directly involved
in the covalent bonding.
❑ Figure 3.4b represented the splitting of the discrete
silicon energy states into bands of allowed energies as
the silicon crystal is formed.
56
 The Energy Band and the Bond
Model
❑ At 𝑇 = 0 𝐾, the 4 N states in the lower band, the valence
band, are filled with the valence electrons.
❑ All of the valence electrons schematically shown in
Figure 3.12 are in the valence band. The upper energy
band, the conduction band, is completely empty at T = 0
K.

57
 The Energy Band and the Bond
Model

Figure 3.12 | Two-dimensional representation of the covalent

bonding in a semiconductor at T = 0 K. 58
 The Energy Band and the Bond
Model

Figure 3.13 | (a) Two-dimensional representation of the breaking of a covalent bond. (b)
Corresponding line representation of the energy band and the generation of a negative and
positive charge with the breaking of a covalent bond. 59
 The Energy Band and the Bond
Model
❑ We can also relate this bond breaking to the E versus k
energy bands.
❑ Figure 3.14a shows the E versus k diagram of the
conduction and valence bands at T = 0 K.
❑ The energy states in the valence band are completely full
and the states in the conduction band are empty.

60
 The Energy Band and the Bond
Model

❑Figure 3.14b shows these same bands for

𝑇 > 0𝐾, in which some electrons have
gained enough energy to jump to the
conduction band and have left empty
states in the valence band.
61
 The Energy Band and the Bond
Model

Figure 3.14 | The E versus k diagram of the conduction and valence

bands of a semiconductor at (a) T = 0 K and (b) T > 0 K. 62
 Drift Current
❑ Current is due to the net flow of charge. If
we had a collection of positively charged
ions with a volume density N (cm-3 ) and an
average drift velocity 𝑣𝑑 (cm/s), then the
drift current density would be

63
 Drift Current

❑ If, instead of considering the average drift

velocity, we considered the individual ion
velocities, then we could write the drift current
density as

64
 Drift Current
where vi is the velocity of the ith ion. The summation
in Equation (3.33) is taken over a unit volume so that
the current density J is still in units of A/cm2.
❑ Since electrons are charged particles, a net drift of
electrons in the conduction band will give rise to a
current.

65
 Drift Current

❑ If a force is applied to a particle and the particle

moves, it must gain energy. This effect is
expressed as

where F is the applied force, dx is the differential distance

the particle moves, v is the velocity, and dE is the increase
in energy. 66
 Drift Current

❑ If an external force is applied to the

electrons in the conduction band, there are
empty energy states into which the electrons
can move; therefore, because of the external
force, electrons can gain energy and a net
momentum show in the figure 3.15.
67
 Drift Current

Figure 3.15 | The asymmetric distribution of electrons in

the E versus k diagram when an external force is applied.
68
 Drift Current

where e is the magnitude of the electronic charge

and n is the number of electrons per unit volume in
the conduction band. Again, the summation is taken
over a unit volume so that the current density is still
in units of A/cm2 .

69
 Electron Effective Mass
❖The movement of an electron in a lattice will, in general,
be different from that of an electron in free space.

❖In addition to an externally applied force, there are

internal forces in the crystal due to positively charged
ions or protons and negatively charged electrons, which
will influence the motion of electrons in the lattice. We
can write

𝐹𝑡𝑜𝑡𝑎𝑙 = 𝐹𝑒𝑥𝑡 + 𝐹𝑖𝑛𝑡 = 𝑚𝑎

70
 Electron Effective Mass
❖ where 𝐹𝑡𝑜𝑡𝑎𝑙 , 𝐹𝑒𝑥𝑡 , and 𝐹𝑖𝑛𝑡 are the total force, the
externally applied force, and the internal forces,
respectively, acting on a particle in a crystal.
❖The parameter a is the acceleration and m is the rest
mass of the particle.

𝐹𝑒𝑥𝑡 = 𝑚∗ 𝑎
71
 Electron Effective Mass
❖ where the acceleration a is now directly related to the
external force. The parameter m *, called the effective mass,
takes into account the particle mass and also takes into
account the effect of the internal forces.
❖ We can also relate the effective mass of an electron in a
crystal to the E versus k curves, such as is shown in Figure
3.11 . In a semiconductor material, we will be dealing with
allowed energy bands that are almost empty of electrons and
other energy bands that are almost full of electrons.
72
 Electron Effective Mass
❖To begin, consider the case of a free electron whose E
versus k curve is shown in Figure 3.7 .
❖Recalling Equation (3.28), the energy and momentum
𝑝2 ħ2 𝑘 2
are related by 𝐸 = = , where m is the mass of
2𝑚 2𝑚
the electron. The momentum and wave number k are
related by 𝑝 = ħ𝑘 . If we take the derivative of
Equation (3.28) with respect to k , we obtain
𝑑𝐸 ħ2 𝑘 ħ𝑝
= = (3.38)
𝑑𝑘 𝑚 𝑚 73
 Electron Effective Mass
❖ Relating momentum to velocity, Equation (3.38) can be
written as 1 𝑑𝐸 𝑝
= =𝑣 (3.39)
ħ 𝑑𝑘 𝑚

where v is the velocity of the particle. The first derivative of E

with respect to k is related to the velocity of the particle.
If we now take the second derivative of E with respect to k, we
have
𝑑 2 𝐸 ħ2
2
= (3.40)
𝑑𝑘 𝑚 74
 Electron Effective Mass
1 𝑑2 𝐸 1
= 3.41
ħ2 𝑑𝑘 2 𝑚
The second derivative of E with respect to k is inversely proportional to the
mass of the particle. For the case of a free electron, the mass is a constant
(nonrelativistic effect), so the second derivative function is a constant. We
𝑑2 𝐸
may also note from Figure 3.7 that is a positive quantity, which implies
𝑑𝑘 2
that the mass of the electron is also a positive quantity.
If we apply an electric field to the free electron and use Newton’s classical
equation of motion, we can write
𝑒𝐸
𝐹 = 𝑚𝑎 = −𝑒𝐸 𝑎=−
𝑚
75
 Electron Effective Mass
❖ The motion of the free electron is in the opposite
direction to the applied electric field because of the
negative charge.

❖We may now apply the results to the electron in the

bottom of an allowed energy band. Consider the
allowed energy band in Figure 3.16a .

76
 Electron Effective Mass

Figure 3.16 | (a) The conduction band in reduced k space, and the parabolic
approximation. (b) The valence band in reduced k space, and the parabolic
approximation. 77
 Electron Effective Mass
The energy near the bottom of this energy band may be
approximated by a parabola, just as that of a free particle.
We may write
𝐸 − 𝐸𝑐 = 𝐶1 𝑘 2 (3.44)
𝑑2 𝐸
2
= 2𝐶1 (3.45)
𝑑𝑘
1 𝑑2 𝐸 2𝐶1
2 2
= 2 (3.46)
ħ 𝑑𝑘 ħ
78
 Electron Effective Mass
❖Comparing Equation (3.46) with Equation (3.41), we
2𝐶1
may equate to the mass of the particle.
ħ2
❖However, the curvature of the curve in Figure 3.16a will
not, in general, be the same as the curvature of the free-
particle curve. We may write

1 𝑑 2 𝐸 2𝐶1 1
2 2
= 2 = ∗ (3.47)
ħ 𝑑𝑘 ħ 𝑚
79
 Electron Effective Mass
❖where m* is called the effective mass. Since C1>0, we have
that m* > 0 also.
❖ The effective mass is a parameter that relates the quantum
mechanical results to the classical force equations. In most
instances, the electron in the bottom of the conduction band can
be thought of as a classical particle whose motion can be
modeled by Newtonian mechanics, provided that the internal
forces and quantum mechanical properties are considered
through the effective mass.
80
 Electron Effective Mass

❖ If we apply an electric field to the electron in the

bottom of the allowed energy band, we may write the
acceleration as
𝑒𝐸
𝑎=− ∗
𝑚
❖ where 𝑚𝑛∗ is the effective mass of the electron. The
effective mass 𝑚𝑛∗ of the electron near the bottom of
the conduction band is a constant.
81
 Concept of the Hole
In considering the two-dimensional representation of the

covalent bonding shown in Figure 3.13a , a positively charged

“empty state” was created when a valence electron was

elevated into the conduction band. For T > 0 K, all valence

electrons may gain thermal energy; if a valence electron gains

a small amount of thermal energy, it may hop into the empty

state. The movement of a valence electron into the empty state

is equivalent to the movement of the positively charged empty

state itself. 82
 Concept of the Hole
❖ Figure 3.17 shows the movement of valence electrons in the
crystal, alternately filling one empty state and creating a new
empty state—a motion equivalent to a positive charge moving in
the valence band. The crystal now has a second equally important
charge carrier that can give rise to a current.

❖This charge carrier is called a hole and, as we will see, can also be
thought of as a classical particle whose motion can be modeled
using Newtonian mechanics.

83
Figure 3.17 | Visualization of the movement of a hole in a semiconductor. 84
 Concept of the Hole
❖ The drift current density due to electrons in the valence
band, such as shown in Figure 3.14b , can be written as

𝐽 = −𝑒 ෍ 𝑣𝑖 (3.49)
𝑖 𝑓𝑖𝑙𝑙𝑒𝑑

❖where the summation extends over all fi lled states. This

summation is inconvenient since it extends over a nearly full
valence band and takes into account a very large number of
states. We may rewrite Equation (3.49) in the form

𝐽 = −𝑒 ෍ 𝑣𝑖 + 𝑒 ෍ 𝑣𝑖 (3.50)
𝑖 𝑡𝑜𝑡𝑎𝑙 𝑖 𝑒𝑚𝑝𝑡𝑦

85
 Concept of the Hole
❖ If we consider a band that is totally full, all available states
are occupied by electrons. The individual electrons can be
thought of as moving with a velocity as given by Equation
(3.39) 1 𝑑𝐸
𝑣(𝑘) = (3.39)
ħ 𝑑𝑘
The band is symmetric in k and each state is occupied so that, for every electron
with a velocity 𝑣 , there is a corresponding electron with a velocity − 𝑣 . Since
the band is full, the distribution of electrons with respect to k cannot be changed
with an externally applied force. The net drift current density generated from a
completely full band, then, is zero, or

−𝑒 ෍ 𝑣𝑖 = 0 (3.51)
𝑖 𝑡𝑜𝑡𝑎𝑙
86
 Concept of the Hole
❖ We can now write the drift current density from
Equation (3.50) for an almost full band as
𝐽 = +𝑒 ෍ 𝑣𝑖 (3.52)
𝑖 𝑒𝑚𝑝𝑡𝑦

❖ where the 𝑣𝑖 in the summation is the associated with the

empty state

1 𝑑𝐸
𝑣(𝑘) =
ħ 𝑑𝑘

87
 Concept of the Hole
Equation (3.52) is entirely equivalent to placing a positively
charged particle in the empty states and assuming all other states
in the band are empty, or neutrally charged. This concept is shown
in Figure 3.18. Figure 3.18a shows the valence band with the
conventional electron-filled states and empty states, whereas
Figure 3.18b shows the new concept of positive charges
occupying the original empty states. This concept is consistent
with the discussion of the positively charged “empty state” in the
valence band, as shown in Figure 3.17 .

88
 Concept of the Hole

Figure 3.18 | (a) Valence band with conventional electron-filled states and
empty states. (b) Concept of positive charges occupying the original empty
states.

89
 Concept of the Hole
❖ The 𝑣𝑖 in the summation of Equation (3.52) is
related to how well this positively charged particle
moves in the semiconductor.

❖ Now consider an electron near the top of the allowed

energy band shown in Figure 3.16b . The energy near
the top of the allowed energy band may again be
approximated by a parabola so that we may write
2
(𝐸 − 𝐸𝑣 ) = − 𝐶2 𝑘 (3.53)
90
 Concept of the Hole
The energy 𝐸𝑣 is the energy at the top of the energy
band. Since E < 𝐸𝑣 for electrons in this band, the
parameter 𝐶2 must be a positive quantity. Taking the
second derivative of E with respect to k from
Equation (3.53), we obtain
𝑑2 𝐸
2
= −2𝐶2 (3.54)
𝑑𝑘
1 𝑑2 𝐸 −2𝐶2
2 2
= 2 (3.54)
ħ 𝑑𝑘 ħ 91
 Concept of the Hole
Comparing Equation (3.55) with Equation (3.41), we may write

1 𝑑2 𝐸 −2𝐶2 1
2 2
= 2 = ∗ (3.47)
ħ 𝑑𝑘 ħ 𝑚

where m* is again an effective mass. We have argued that

𝐶2 is a positive quantity, which now implies that m* is a
negative quantity. An electron moving near the top
of an allowed energy band behaves as if it has a negative
mass.
92
 Concept of the Hole
❖ We must keep in mind that the effective mass parameter
is used to relate quantum mechanics and classical
mechanics.
❖The attempt to relate these two theories leads to this
strange result of a negative effective mass. However, we
must recall that solutions to Schrodinger’s wave equation
also led to results that contradicted classical mechanics.
The negative effective mass is another such example.
93
 Concept of the Hole
❖In discussing the concept of effective mass in the previous section,
we used an analogy of marbles moving through two liquids. Now
consider placing an ice cube in the center of a container filled with
water: the ice cube will move upward toward the surface in a
direction opposite to the gravitational force. The ice cube appears to
have a negative effective mass since its acceleration is opposite to
the external force. The effective mass parameter takes into account
all internal forces acting on the particle.
❖ If we again consider an electron near the top of an allowed energy
band and use Newton’s force equation for an applied electric field,
we will have
94
 Concept of the Hole
𝐹 = 𝑚∗ 𝑎 = −𝑒𝐸
However, m* is now a negative quantity, so we may
write

𝑒𝐸 + 𝑒𝐸
𝑎=− ∗
=
−𝑚 𝑚∗

An electron moving near the top of an allowed energy band

moves in the same direction as the applied electric field.

95
 Concept of the Hole
❖ The net motion of electrons in a nearly full band can be described by
considering just the empty states, provided that a positive electronic charge
is associated with each state and that the negative of m * from Equation
(3.56) is associated with each state.
❖ We now can model this band as having particles with a positive electronic
charge and a positive effective mass. The density of these particles in the
valence band is the same as the density of empty electronic energy states.
This new particle is the hole. The hole, then, has a positive effective mass
denoted by 𝑚𝑝∗ and a positive electronic charge, so it will move in the same
direction as an applied field.

96
 Metals, Insulators, and
Semiconductors

3.5 to 6 eV or larger

Figure 3.19 | Allowed energy bands showing (a) an empty band,

(b) a completely full band, and (c) the bandgap energy between the
two allowed bands. 97
 Metals, Insulators, and
Semiconductors

1 eV

Figure 3.20 | Allowed energy bands showing (a) an almost empty band,
(b) an almost full band, and (c) the bandgap energy between the two
allowed bands. 98
 Metals, Insulators, and
Semiconductors

Figure 3.21 | Two possible energy bands of a metal showing (a) a

partially filled band and (b) overlapping allowed energy bands.
99
Metals, Insulators, and
Semiconductors

100
Test your understanding

101
 The k -Space Diagrams of Si and GaAs

indirect bandgap
direct bandgap

Figure 3.25 | Energy-band structures of (a) GaAs and (b) Si.

102
(From Sze [12].)
 The k -Space Diagrams of Si and GaAs

❑ Germanium is also an indirect bandgap material;

whose valence band maximum occurs at k = 0
and whose conduction band minimum occurs
along the [111] direction.
❑ GaAs is a direct bandgap semiconductor, but
other compound semiconductors, such as GaP and
AlAs, have indirect bandgaps.
103
 DENSITY OF STATES FUNCTION

❖ As we have stated, we eventually wish to describe the

current–voltage characteristics of semiconductor

devices.

❖ Since current is due to the flow of charge, an

important step in the process is to determine the

number of electrons and holes in the semiconductor

that will be available for conduction.

104
 DENSITY OF STATES FUNCTION
❖ The number of carriers that can contribute to the conduction
process is a function of the number of available energy or
quantum states since, by the Pauli exclusion principle, only one
electron can occupy a given quantum state.
❖ When we discussed the splitting of energy levels into bands of
allowed and forbidden energies, we indicated that the band of
allowed energies was made up of discrete energy levels. We must
determine the density of these allowed energy states as a
function of energy in order to calculate the electron and hole
concentrations.
105
 Mathematical Derivation

❖ To determine the density of allowed quantum states as a function

of energy, we need to consider an appropriate mathematical model.
Electrons are allowed to move relatively freely in the conduction
band of a semiconductor but are confined to the crystal.
❖ As a first step, we will consider a free electron confined to a
three-dimensional infinite potential well, where the potential well
represents the crystal. The potential of the infinite potential well is
defined as

106
 Mathematical Derivation

where the crystal is assumed to be a cube with length a .

❖ Schrodinger’s wave equation in three dimensions can be solved by
using the separation of variables technique. Extrapolating the results
from the one-dimensional infinite potential well, we can show (see
Problem 3.23) that

107
 Mathematical Derivation

where nx , ny , and nz are positive integers. (Negative values of nx , ny

, and nz yield the same wave function, except for the sign, as the
positive integer values, resulting in the same probability function and
energy; therefore the negative integers do not represent a different
quantum state.)
❖ We can schematically plot the allowed quantum states in k space.
Figure 3.26a shows a two-dimensional plot as a function of kx and
ky . Each point represents an allowed quantum state corresponding
to various integral values of nx and ny .
❖ Positive and negative values of kx , ky , or kz have the same energy
and represent the same energy state.
108
 Mathematical Derivation

Figure 3.26 | (a) A two-dimensional array of allowed quantum states in k space.

(b) The positive one-eighth of the spherical k space.

109
 Mathematical Derivation
❖Since negative values of kx , ky , or kz do no represent additional quantum
states, the density of quantum states will be determined by considering
only the positive one-eighth of the spherical k space as shown in Figure
3.26b .

❖ The distance between two quantum states in the kx direction, for

example, is given by

❖ Generalizing this result to three dimensions, the volume V k of a

single quantum state is

110
 Mathematical Derivation
❖ We can now determine the density of quantum states in k space. A
differential volume in k space is shown in Figure 3.26b and is given by
4𝜋𝑘 2 𝑑𝑘 , so the differential density of quantum states in k space can be
written as

The first factor, 2, takes into account the two spin states allowed for each
quantum state; the next factor, 1/8 , takes into account that we are
considering only the quantum states for positive values of kx , ky , and kz .
The factor 4𝜋k2dk is again the differential volume and the factor (𝜋/𝑎 )3 is
the volume of one quantum state. Equation (3.63) may be simplified to 111
 Mathematical Derivation

❖ Equation (3.64) gives the density of quantum states as a function

of momentum, through the parameter k . We can now determine the
density of quantum states as a function of energy E . For a free
electron, the parameters E and k are related by

112
 Mathematical Derivation

❖ Equation (3.68) gives the total number of quantum states between the energy E and
E+ dE in the crystal space volume of a 3 . If we divide by the volume a3 , then we
will obtain the density of quantum states per unit volume of the crystal. Equation
(3.68) then becomes

113
Mathematical Derivation

114
Mathematical Derivation

115
 Extension to Semiconductors

❖ In the previous section, we derived a general expression for the

density of allowed electron quantum states using the model of a
free electron with mass m bounded in a three-dimensional
infinite potential well.
❖ We can extend this same general model to a semiconductor to
determine the density of quantum states in the conduction band
and the density of quantum states in the valence band.
❖ Electrons and holes are confined within the semiconductor
crystal, so we will again use the basic model of the infinite
potential well.

116
 Extension to Semiconductors

Equation (3.72) is valid for E > Ec . As the energy of the electron in

the conduction band decreases, the number of available quantum
states also decreases.
117
 Extension to Semiconductors
❖ The density of quantum states in the valence band can be obtained by
using the same infinite potential well model, since the hole is also
confined in the semiconductor crystal and can be treated as a “free”
particle. The density of states effective mass of the hole is 𝑚𝑝∗ . Figure
3.16b shows the valence energy band in the reduced k space. We may
also approximate the E versus k curve near k = 0 by a parabola for a
“free” hole, so that

118
Extension to Semiconductors

119
 Extension to Semiconductors
❖ Again, the right side of Equation (3.74) is of the same form
used in the general derivation of the density of states
function. We may then generalize the density of states
function from Equation (3.69) to apply to the valence band,
so that

Equation (3.75) is valid for E < Ev.

120
 Extension to Semiconductors

❖ We have argued that quantum states do not exist within the

forbidden energy band, so g(E) = 0 for Ev < E < Ec. Figure 3.27
shows the plot of the density of quantum states as a function of
energy. If the electron and hole effective masses were equal, then
the functions gc(E) and gv(E) would be symmetrical about the
energy midway between Ec and Ev , or the midgap energy, E midgap .

121
Extension to Semiconductors

Figure 3.27 | The density of energy states in

the conduction band and the density of energy
states in the valence band as a function of
energy.

122
Extension to Semiconductors

123
Extension to Semiconductors

124
 STATISTICAL MECHANICS
• In dealing with large numbers of particles, we are interested
only in the statistical behavior of the group as a whole
rather than in the behavior of each individual particle.
• For example, gas within a container will exert an average
pressure on the walls of the vessel. The pressure is actually
due to the collisions of the individual gas molecules with
the walls, but we do not follow each individual molecule as
it collides with the wall. Likewise in a crystal, the electrical
characteristics will be determined by the statistical behavior
of a large number of electrons. 125
 Statistical Laws
❖ In determining the statistical behavior of particles, we must
consider the laws that the particles obey. There are three
distribution laws determining the distribution of particles
among available energy states.
❖ One distribution law is the Maxwell–Boltzmann probability
function. In this case, the particles are considered to be
distinguishable by being numbered, for example, from 1 to N,
with no limit to the number of particles allowed in each
energy state. The behavior of gas molecules in a container at
fairly low pressure is an example of this distribution.
126
 Statistical Laws
❖ A second distribution law is the Bose–Einstein function. The
particles in this case are indistinguishable and, again, there is no
limit to the number of particles permitted in each quantum state.
The behavior of photons, or black body radiation, is an example
of this law.
❖The third distribution law is the Fermi–Dirac probability
function. In this case, the particles are again indistinguishable,
but now only one particle is permitted in each quantum state.
Electrons in a crystal obey this law. In each case, the particles
are assumed to be noninteracting.
127
 The Fermi–Dirac Probability
Function
❖Figure 3.28 shows the ith energy level with gi quantum states. A
maximum of one particle is allowed in each quantum state by the
Pauli exclusion principle.
❖There are gi ways of choosing where to place the first particle, (gi -
1) ways of choosing where to place the second particle, ( gi - 2) ways
of choosing where to place the third particle, and so on.

Figure 3.28 | The ith energy level with gi quantum states. 128
 The Fermi–Dirac Probability
Function
❖ Then the total number of ways of arranging 𝑁𝑖
particles in the ith energy level (where 𝑁𝑖 ≤ 𝑔𝑖 ) is

❖ This expression includes all permutations of the Ni

particles among themselves.

129
 The Fermi–Dirac Probability
Function
❖ However, since the particles are indistinguishable, the Ni!
number of permutations that the particles have among
themselves in any given arrangement do not count as separate
arrangements. The interchange of any two electrons, for
example, does not produce a new arrangement.
❖ Therefore, the actual number of independent ways of
realizing a distribution of Ni particles in the ith level is

130
Example

131
The Fermi–Dirac Probability Function

132
 The Fermi–Dirac Probability Function

❖ Equation (3.77) gives the number of independent ways of

realizing a distribution of Ni particles in the i th level. The total
number of ways of arranging ( N1 , N2 , N3 , . . . , Nn )
indistinguishable particles among n energy levels is the product
of all distributions, or

❖The parameter W is the total number of ways in which N

electrons can be arranged in this system, where 𝑁 = σ𝑛𝑖=1 𝑁𝑖 is
the total number of electrons in the system.
133
 The Fermi–Dirac Probability
Function
❖We want to find the most probable distribution, which
means that we want to find the maximum W. The maximum
W is found by varying Ni among the Ei levels, which varies
the distribution, but at the same time, we will keep the total
number of particles and total energy constant.
❖We may write the most probable distribution function as

134
 The Fermi–Dirac Probability
Function
❖ The number density N(E) is the number of particles per unit
volume per unit energy and the function g(E) is the number
of quantum states per unit volume per unit energy. The
function fF(E) is called the Fermi–Dirac distribution or
probability function and gives the probability that a quantum
state at the energy E will be occupied by an electron.
❖ The energy EF is called the Fermi energy. Another
interpretation of the distribution function is that fF(E) is the
ratio of filled to total quantum states at any energy E .

135
 The Distribution Function and the
Fermi Energy
❖ The Fermi–Dirac distribution function for T = 0K is plotted in Figure
3.29 . This result shows that, for T= 0K, the electrons are in their
lowest possible energy states.
❖The probability of a quantum state being occupied is unity for E<EF and
the probability of a state being occupied is zero for E > EF . All
electrons have energies below the Fermi energy at T = 0 K.

Figure 3.29 | The Fermi probability function versus energy for T = 0 K. 136
The Distribution Function and the Fermi Energy

137
 The Distribution Function and the Fermi
Energy
❖Figure 3.30 shows discrete energy levels of a particular system as well as
the number of available quantum states at each energy. If we assume, for
this case, that the system contains 13 electrons, then Figure 3.30 shows
how these electrons are distributed among the various quantum states at
T = 0 K. The electrons will be in the lowest possible energy state, so the
probability of a quantum state being occupied in energy levels E1
through E4 is unity, and the probability of a quantum state being
occupied in energy level E5 is zero. The Fermi energy, for this case, must
be above E4 but less than E5. The Fermi energy determines the statistical
distribution of electrons and does not have to correspond to an allowed
energy level.

138
 The Distribution Function and the Fermi Energy
❖ Now consider a case in which the density of quantum
states g (E) is a continuous function of energy, as shown
in Figure 3.31. If we have N0 electrons in this system then
the distribution of these electrons among the quantum
states at T = 0K is shown by the dashed line.
❖ The electrons are in the lowest possible energy state so
that all states below EF are filled and all states above EF
are empty.
❖ If g(E) and N0 are known for this particular system, then
the Fermi energy EF can be determined.
139
The Distribution Function and the Fermi
Energy

140
 The Distribution Function and the
Fermi Energy
❖ Consider the situation when the temperature increases above T = 0K.
Electrons gain a certain amount of thermal energy so that some electrons
can jump to higher energy levels, which means that the distribution of
electrons among the available energy states will change.
❖ Figure 3.32 shows the same discrete energy levels and quantum states as
in Figure 3.30 . The distribution of electrons among the quantum states
has changed from the T = 0K case. Two electrons from the E4 level have
gained enough energy to jump to E5 , and one electron from E3 has
jumped to E4. As the temperature changes, the distribution of electrons
versus energy changes.

141
The Distribution Function and the
Fermi Energy

142
 The Distribution Function and the
Fermi Energy
❖The change in the electron distribution among energy levels for T > 0K
can be seen by plotting the Fermi–Dirac distribution function. If we let E
=EF and T > 0 K, then Equation (3.79) becomes

❖ The probability of a state being occupied at E= EF is 1/2. Figure

3.33 shows the Fermi–Dirac distribution function plotted for
several temperatures, assuming that the Fermi energy is
independent of temperature.
143
The Distribution Function and the
Fermi Energy

144
 The Distribution Function and the
Fermi Energy

❖ We can see that for temperatures above absolute

zero, there is a nonzero probability that some energy
states above EF will be occupied by electrons and
some energy states below EF will be empty.
❖ This result again means that some electrons have
jumped to higher energy levels with increasing
thermal energy.

145
The Distribution Function and the
Fermi Energy

146
 Exercise

❖ We can see from Figure 3.33 that the probability of an energy

above EF being occupied increases as the temperature increases
and the probability of a state below EF being empty increases as
the temperature increases.

147
The Distribution Function and the Fermi
Energy

148
Exercise

149
The Distribution Function and the Fermi
Energy

150
 The Distribution Function and the
Fermi Energy
❖ Equation (3.80) is known as the Maxwell Boltzmann
approximation, or simply the Boltzmann approximation, to the
Fermi–Dirac distribution function. Figure 3.35 shows the
Fermi–Dirac probability function and the Boltzmann
approximation. This figure gives an indication of the range of
energies over which the approximation is valid.

151
The Distribution Function and the
Fermi Energy

152
153
 Exercise

❖ The actual Boltzmann approximation is valid when exp[(E

EF)kT]>> 1. However, it is still common practice to use the E
EF >> kT notation when applying the Boltzmann
approximation. We will use this Boltzmann approximation in
our discussion of semiconductors in the next chapter.

154
Testing your understanding

155
exercise

156
157