1

Energy Bands
Energy

Insulator Metal Semiconductor Semiconductor

1
 Atomic & Molecular Wavefunctions

-e +e -e

+e
energy of -e

r
n=1
n=1 r
+e n=3
+e
n=3
-e +e +e n=2
Superposition of Coulomb r
potentials H2: n=2
energy of -e

Hydrogen Atoms
• As the two atoms
approach each other, ψ1s
wavefunctions overlap.

• The overlap of
wavefunctions can either
be in phase or out of
phase  two molecular
orbitals are formed with
different energies.

σ = 1𝑠 𝑟𝐴 + 1𝑠 𝑟𝐵
σ∗ = 1𝑠 𝑟𝐴 −
1𝑠(𝑟𝐵)

2
 Probability Distribution

• The first molecular orbital ψσ is symmetric and has considerable magnitude

between the nuclei, whereas the second ψσ*, is antisymmetric and has a node
between the nuclei.

Hydrogen Atoms
• As R decreases and the two H atoms
get closer, the energy of the ψσ
orbital state passes through a
minimum at R = a.
• The wavefunction ψσ corresponding
to the lowest electron energy is
called the bonding orbital, and ψσ* is
the antibonding orbital.

3
 Hydrogen Atoms
• The changes in the electron energy as two isolated H atoms, far left and far
right, come together to form a hydrogen molecule.

• E1s, splits into two, Eσ and Eσ*.

He Atoms
• Consider what happens when two He atoms come together.
• When He atoms come together, two of the electrons enter the Eσ level and
two the Eσ* level, so the overall energy is greater than two isolated He atoms.
• Quantum mechanical calculations show that the antibonding energy level Eσ*
shifts higher than the bonding level Eσ shifts lower.
• The overlap of full atomic orbital states does not lead to bonding.
• We need an overlap of half-occupied orbitals, as in the H2 molecule, to form
a bond between two atoms.

4
 3 Hydrogen Atoms
• Three separate molecular orbital states, ψa, ψb, and ψc, from three ψ1s atomic
states.
• Although H2 and H3 both have two electrons in the lowest energy level, H3
also has an extra electron at the higher energy level (Eb), which tends to
increase the net energy of the atom. Thus, the H3 molecule is much less
stable than the H2 molecule.

Li Atoms
• Consider N Li (lithium) atoms are brought close together.

10

5
 Bloch Theorem
• Bloch wave: A type of wavefunction for a particle in a periodically
repeating environment → consider an electron in a crystal.
• A wavefunction ψ is a Bloch wave if it has the form:
→→ →
 k→, r  e→ik r 
u
→
k,r

→→ 
u k , r : Periodic function with the same periodicity as the
crystal
 →

u k→, r  →
R u  
→
k , r→

A plane wave modulated by a function that
has periodicity of the potential.
11

Bloch Theorem

A Bloch wave (bottom) can be broken up into the product of a periodic function
(top) and a plane-wave (center). Blue is real part, red is imaginary part. The left
side and right side represent the same Bloch wave broken up in two different ways,
involving the wave vector k1 (left) or k2 (right). The difference (k1k2) is a
reciprocal lattice vector. 12

6
1

Kronig-Penny Model: Periodic Potential

• One electron approximation: Each valence electron is
considered independent and only acted upon by the periodic
positive ions.

7
 Schrodinger’s
Equation
V(x)
V0
c b
x
0
2
2 d ψ

2m dx2  V (x)ψ (x)  εψ
(x)
d 2ψ 2m V (x)ψ (x) 
dx2  2 ε 0

 be of Bloch form: ψ (x) 

• Wavefunction will
eikxu(x)
 u(x) has the periodicity of the lattice. 3

Schrodinger’s
Equation
• After
d 2 substitution
2m
ikx
e u(x) 
2 2
 ε V (x) ikxe u(x)
dx
0  du(x)  2m
d  ikx ikx ikx

ike u(x)  e  ε V (x)e u(x)

dx 
 0
 
2

dx 2 ikx ikx du(x) ikx du(x) ikx d

ik  e u(x)  ike 2u(x)  ike
dx

dx2
e
dx
2m
 ε V (x)eikxu(x)  0
 2

du(x) 2
d u(x) 2ik  k u(x)  2  ε V (x) u(x)
2
 2m
dx 2 0
dx 
d 2u(x) 2ik du(x)  2 2mV (x) 
α 2  2
dx  k  α 
 2
dx2  u(x) 
2 2mε 
0, 4

8
 Potential Well and
Barrier
d 2u(x) 2ik du(x)  2 2
2mV (x) 
dx2   k
  α  2  u(x)  0
dx
V(x)
0x
db 2:u (x) du
dx
1

 2ik dx  k α
2
2 2
u (x) 
1 c b

0 u2(x)
c  x  0 : u1(x)
x
d u2 (x)
2
du 0
 2ik  k 2  β 2
u2 (x)  a
dx 2
dx
0
β 2 
2m(ε 
V0 )
2
5

Solution
s
u1 (x)  ix (α k ) 0Be  i(α
x
x kb) :
Ae V(x)
c  xi ( β k0 ):x i ( β k ) x
u
2 (x)  Ce  De
c b
A, B, C, D are constants and must
be determined. u2(x)
u1(x)
x
Apply Boundary Conditions: 0
a
Wavefunctions and their derivatives must be continuous
at all points.
6

9
 Boundary Conditions
• At x = 0

u1 (0)  u2 (0)
V(x)
ABCD

u2(x) b
du1 (x) du2 (x) u1(x)
 x
dx x0 0
x0 dx a
i(α  k ) A  i(α  k )B  i(β  k )C  i(β
 k )D
7

Boundary Conditions
• At x = b
u1 (b)  u2 (c)
V(x)
Aei(α k )b
 Bei (α k )b
 Cei ( β k )c
 Dei ( β k )c

du1 (x) du2 (x) u2(x) b


dx xb xc
u1(x)
dx x
0
i(α  k ) Aei (α k )b
 i(α  k)Bei (α k )b
a
 i(β  k)Cei ( β k )c
 i(β  k)Dei( β
k )c

1
0
 Solution
s
ABCD
i(α  k ) A  i(α  k )B  i(β  k )C  i(β
 k)D Aei (α k )b  Bei (α k )b  Cei (β k )c 
k )c
Dei (β

i(α  k ) Aei (α k )b  i(α  k )Bei (α k )b 

i(β  k )Cei (β k )c  i(β  k )Dei ( β k )c

• For meaningful solutions to exist, the

1 1 1 1
determinant must be zero:
α k  (α  k) β k  (β  k )
ei (α k )b
ei (α k )b
ei (β k )c
ei ( β k )c 
0
(α  k )ei (α  (α  k )ei (α (β  k)ei (β k )c
 (β  k)ei (β
k )b k )b k )c

Dispersion
Relation
β 
2m(ε 2

V0 )
2

β  0  β is
2


α 2
β sin αb sin βc  cosαb cos βc  cos
 real.
2 2α ka
β

β 2
0  β is
imaginary.
sin αb sinh γ c  cosαb cosh γ c 
γ2α2  αcos2
ka
γ
β  iγ
10

10

1
1
 Dispersion
Relation
• Using trigonometry
2 1/ 2
 V
sin2 βc cos(αb  δ )  cos (ε  0V
1
0

ka,
 4 ε  ε V
0  )
  V
2 1/ 2

1 0
sinh2 γ c cos(αb  δ )  cos ka, (0  ε 0

V )
4 ε  V
0  ε 
w 
he α 2β γ 2 α
tan δ  2 tan βc and tan δ 2 tanh γ
re
 c
2αβ   2αγ

11

11

Energy
P Bands
• In an infinite lattice, the
states within any
ka  allowed band
2π would form a
ka  0 continuum.
1 • For a lattice of N atoms,
there are N discrete
0 states,
αa
however, there are 2N

states for spin
1 ka  • The
degeneracy.
energy gaps
π ka 
decrease as
π electron energy
increases  free electron
Forbidden energy behavior at high
range
α 2  energies.
2mε
2 12

12

1
2
 Dispersion
Relation
ε (k)

k
−π /a −π /a 0

π /a 2π /a

• Energies close to zero are forbidden.

• At k = nπ /a, there are two possible values of
13
energies.
13 • Bragg reflection at k =  nπ /a

Reduced Zone Representation

• Energy band diagram is plotted in the first Brillouin zone.

k′

• How the energy bands (E-k relation) will change for a 3-

D case?
14

14

1
3
 Dispersion Relation:
3D
ε
(k)

−π /a 0 kx, ky, kz
π /a
−π /b
π /b
−π a/c= b= c 15
π /c
15

Dispersion Relation: 3-D

• k-values at the zone boundaries along different crystal
orientations may be different  overlap of energy states at zone
boundaries.

εz
ε
y
εx

kx, ky, kz
π/a π/b
π/c 16

16

1
4
1

Effective Mass
• A particle's effective mass (often denoted m*) is the mass that
it seems to have when responding to forces.
• The effective mass is usually stated in units of the true mass
of the electron me (9.11×10−31 kg). In these units it is usually
in the range 0.01 to 10, but can also be lower or higher.
• The electronic effective mass can be seen as an important
basic parameter that influences measurable properties of a
solid, including everything from the efficiency of a solar cell
to the speed of an integrated circuit.

1
5
 Parabolic Isotropic Dispersion
• At the highest energies of the valence band in many semiconductors (Ge, Si,
GaAs), and the lowest energies of the conduction band in some
semiconductors (GaAs), the band structure E(k) can be locally
approximated as:

→

E k  E0 
2
2
k
*

 2m
• It can be shown that the electrons placed in these bands behave as free
electrons except with a different mass, as long as their energy stays within
the range of validity of the approximation above. As a result,
the electron mass in models such as the Drude model must
be replaced with the effective mass.

Parabolic Non-isotropic
Dispersion
• In some important semiconductors (notably, silicon) the
lowest energies of the conduction band are not
symmetrical 2so that the band2 minimum can be
E k   E0 only by
approximated 2m k  k
x*  x 0, x 
 2 2

 2m*y kyy  k 0,
 2


2m* k  k
z 0,z

z  2

• The speed of an electron will depend on its direction, and it will accelerate
to a different degree depending on the direction of the force.
• For the purposes of calculating conductivity as in the Drude model, via
the harmonic mean
1
 1 1
 * 1 * 
m*conductivity
 3 m  m*y 
 x mz  
4

1
6
 General
Case
• The dispersion relation:
ε  ω
• The electron’s acceleration:
dvg d  dω  d  d ω  dk 1 d 2ε
  
d (k) dt dt  dk  dk  dk  dt 2 dk 2
dt
• If force Fe accelerates electron, will do work dWe in time dt,
while electron’s energy and momentum change by dε and dk.
dω
dWe  Fe dx  F e v dt e F
dk
g
dt
dε
dWe  dε  dk dk  
5
dω dk
dk
5

Effective
Mass
• Equating: d
Fe 
dt
(k)
• After dvg 1 d 2εF e
 e 
substituting: dt  dk
2
m
F2
*
m* 2  Effective mass
 d 2ε / dk
2
Material Electron Effective Mass Hole Effective
Mass
Group IV
Si (4.2 K) 1.08 0.56
Ge 0.555 0.37
Groups III-IV
GaAs 0.067 0.45
InSb 0.013 0.60
Groups II-VI
ZnO 0.,19 1.21
ZnSe 0.17 1.44 6

1
7
 Free
Electron
 2 2
2
2

2m dk 2
m
• Effective mass:

2
m*  
 2 / m m

Effective Mass
• In parabolic sections: Near the top and bottom of the
bands:
d2 
ε  C(k  nπ2 ) ε2 2  2C  m*
 dk 
2C
• In non-parabolic sections: d 2ε / dk2 depends on energy 
m* is not constant.
ε (k)

k 8

1
8
 Parabolic Approximation

Effective Mass

10

10

1
9
 Effective Mass

11

11

Negative Effective Mass

• Negative curvatures near the top of
dispersion relation  negative
effective mass.
• External field E in the positive x-direction
acting on a negatively charged
particle with a negative mass
produces a positive acceleration component
along the x-direction.
• The change in sign of m* can be thought of
the change of sign of charge q.

dvg


qE dt

m*
12
• Quasi-particle 
positive charge
and positive mass
 holes.
• Without lattice, 2
holes cannot 0
exist.
1

Density of States
• How many states per unit energy per unit volume?
• N atoms  N electron wavefunctions.

1 = 𝐴 + 𝐵 + 𝐶 + 𝐷 + ⋯

𝑁 = 𝐴 − 𝐵 + 𝐶 − 𝐷 + ⋯

• 𝑔 𝐸 : Density of states
• 𝑔 𝐸 𝑑𝐸  number of
states (i.e.,
wavefunctions) in the
energy interval E to
(E + dE) per unit
volume of the sample.

2
1
 Density of States
• The number of states per unit volume up to some energy E′

𝐸𝘍

𝑆𝑣 𝐸 ′ = ∫ 𝑔 𝐸 𝑑𝐸
0

• The energy of an electron in a cubic PE well of size L is given

by
𝐸= 𝑘22 + 𝑘2 + 𝑘2

= 𝑘′ 8𝑚ℎ𝑒
2 2 1 2
3
𝐿
• The spatial dimension L of the well now refers to the size of the entire
solid, as the electron is confined to be somewhere inside that solid. Thus,
L is very large compared to atomic dimensions, which means that the
separation between the energy levels is very small.

• Enumerate all possible choices of integers for n1, n2, and n3 that satisfy
𝑘12 + 𝑘22 + 𝑘
3
2 ≤ 𝑘

′2 .
3

2D
• 𝑘12 + 𝑘22 ≤ 𝑘 Material
′2

1
Area = 4
𝜋𝑘2

2
2
3D Material

• 𝑘12 + 𝑘2 2 + 𝑘3
2

≤ 𝑘′ 2
• The number of orbital states Sorb(n
′) within this volume is given by

1 4 1
𝜋𝑘′
� = 𝑘′ 3 =
orb 𝑘 ′ 3
� 8 3 6
• The number of states, including

𝑆 𝑘 ′ = orb 𝑘 ′ = 1 𝜋𝑘′3
spin:
3
2𝑆
• Substitute 𝑘′ 2 :

ℎ2
8𝑚 𝑒 𝐿 2
𝐸= 𝑘 ′2  𝑘 ′2 = 𝐸 ℎ2
8𝑚𝑒 2

𝐿
5

3D
Material
• 𝜋𝐿of
The number 3 (8𝑚states 𝑒 𝐸with 𝐸 ≤ 𝐸′:
𝑆 𝐸′ =
′ ) 3 / 2 3ℎ 3

• The number of states per unit

𝜋(8𝑚 𝑒 𝐸
volume:
′
𝑆𝑣 𝐸 =
′) 3 / 2
3ℎper
• Density of states
3
unit
volume:

= (8 2𝜋) 2
𝑆𝑣 𝑚𝑒
𝑔 𝐸 =𝑑
3/2 ℎ
𝐸 1 / 2𝐸

2
3
 Carrier Statistics
• If E is the electron energy and f (E) is the probability that a state with energy E
is occupied, then

𝑘 = ∫ 𝑓 𝐸 𝑔 𝐸 𝑑𝐸
Ba nd

• We need to know 𝑓 𝐸 .

Boltzmann Probability Function

• If E is the electron energy and f (E) is the probability that a state with
energy E
is occupied, then
𝑓 𝐸 = 𝐴 exp
𝐸 𝑘
− 𝑇
• 𝑓 ( 𝐸 ) decreases exponentially with
energy.

• Any number of particles may have a

given energy E. There is no restriction to
permit only one particle per state at an
energy E, as in the Pauli exclusion
principle.
𝑁2 𝐸2 − 𝐸1
= exp −
𝑁1 𝑘𝑇

• If 𝐸2 ≫ 𝐸1, then 𝑁2
« 𝑁1 . 8

2
4
 Boltzmann Probability Function
• Classical particles obey the
Boltzmann statistics.
• If there are many more states than the
number of particles, the likelihood of
two particles having the same set of
quantum numbers is negligible and we
do not have to worry about the Pauli
exclusion principle
 We can use the Boltzmann statistics.
• Example: Electrons in the conduction
band of a semiconductor, where
usually there are many more states than
electrons.

Fermi-Dirac Statistics
• If E is the electron energy and f (E) is the probability that a state with energy E
is occupied, then

• 𝐸 − 𝐸𝐹 ≫ 𝑘𝑇:

10

10

2
5
 Carrier Density
𝑘𝐸 = 𝑔 𝐸 𝑓 ( 𝐸 )

11

11

Carrier
Density
𝑘𝐸 = 𝑔 𝐸 𝑓 ( 𝐸 )

• In the small energy range

E to (E + dE), there are
nE dE
electrons per unit volume.
• Total electrons in the
To p of To p of b a n d
band: 𝑘 𝐸 𝑑𝐸 = ∫
𝑘 = ∫b a n d 𝑔 𝐸 𝑓
0 𝐸 𝑑𝐸
0

• f → 0 when E ≫
EF
8 ∞
𝐸 1/ 2 𝑑𝐸
𝑘 = 2𝜋𝑚 3 /𝑒2 ∫
3 ℎ 0
1 + exp 𝐸𝑘−
𝐸𝐹 𝑇
12

12

2
6
Fermi Energy

• At T = 0 K
𝑘𝐸 = 𝑔 𝐸
𝑓(𝐸)

8 𝐸𝐹0
𝐸 1/ 2 𝑑𝐸
𝑘 = 2𝜋𝑚 3 /𝑒2 ∫
3ℎ 0
1 + exp 𝐸𝑘− 𝐸𝐹0
𝑇
ℎ2 2/ 3
𝐸𝐹0 = 3𝑘

𝜋
8𝑚 𝑒

13

13

2
7