Band Theory of Solids

In the free electron model (in a constant or zero potential field) the energy of the electron is a
function of wave vector K given by,

ℏ2 k2 2𝜋 𝑃
E= ; where K = =ℏ
2𝑚 𝛌

Where, 𝛌 is the wavelength associated with the electron and P is the momentum. The potential
energy is assumed to be zero. There is no upper limit to the energy. The energy spectrum is
quasi-continuous because the limited dimensions of the box produce closely spaced but discrete
energy bands.

The plot of E Vs. K is shown in the figure which is a parabola for the free electron modal.

However, if one considers the motion of an electron in a periodic potential we arrive at the
following results. (a) There exists allowed energy band separated by Forbidden regions. (b)
The energy E(K) are periodic in K.
 The Bloch Theorem:

In a free electron theory, one assumes that an electron moves in a constant potential Vo leading
to the Schrodinger equation.

- ℏ2 𝑑 2 ψ
+ Eψ = Voψ
2𝑚 𝑑𝑥 2

𝑑2 ψ 2𝑚
⇒ + (E-Vo) ψ = 0
𝑑𝑥 2 ℏ2

𝑑2 ψ 2𝑚
⇒ + 𝑘2 ψ = 0; where 𝑘2 = (E-Vo)
𝑑𝑥 2 ℏ2

The solutions are plan wave in the form

Ψ= 𝑒 ±𝑖𝑘𝑥

This represents waves propagating along X-axis.

ℏ2 k2 p2
Ekin= E-Vo = = 2𝑚; where p = ℏk is the momentum. The physical meaning of k is that it
2𝑚

represents the momentum of an electron divided by ℏ.

𝐸
For time dependent wave function we have to multiply by 𝑒 𝑖𝜔𝑡 ; where 𝜔 = ℏ ; E = ℏ 𝜔

Let us consider the case when the electron is moving in a one-dimensional periodic potential.

V(x) = V(x+a); where a is the period.

𝑑2 ψ 2𝑚
Then, + (E-V(r)) ψ = 0
𝑑𝑥 2 ℏ2

The solutions of equation are of the form

Ψ= 𝑒 ±𝑖𝑘𝑥 uk(x)

Where uk(x) = uk(x+a) is a function of x has the same periodicity as the lattice. This theorem is
known as Bloch theorem.

Ψ(x) = 𝑒 ±𝑖𝑘𝑥 uk(x) are known as Bloch function.

Ψ(x+a) = 𝑒 ±𝑖𝑘(𝑥+𝑎) uk(x+a)

= 𝑒 ±𝑖𝑘𝑥 . 𝑒 ±𝑖𝑘𝑎 . uk(x)

 = Ψ(x) 𝑒 ±𝑖𝑘𝑎

Ψ(x+a) = QΨ(x); where Q = 𝑒 ±𝑖𝑘𝑎

Kronig-Penny Model

Consider one dimensional model of Kronig & Penny.

The potential energy of an electron has the form of a periodic array of square walls. The period
of the potential is (a + b).

In region 0 < x < a, V = 0

-b < x < 0, V = Vo

The Schrodinger equation is

ℏ2 𝑑 2 ψ
+ Eψ = Vψ
2𝑚 𝑑𝑥 2

𝑑2 ψ 2𝑚𝐸
⇒ + ψ=0 for 0 < x < a
𝑑𝑥 2 ℏ2

𝑑2 ψ 2𝑚(𝐸−𝑉0 )
⇒ + ψ = 0 for –b < x < 0
𝑑𝑥 2 ℏ2

Total energy of electron E is less than 𝑉0 and defining two real quantities 𝛼 and 𝛽.

2𝑚𝐸 2𝑚(𝑉0 −𝐸)

𝛼2 = ; 𝛽2 =
ℏ2 ℏ2
 𝑑2 ψ
We have, + 𝛼2ψ = 0 for 0 < x < a
𝑑𝑥 2

𝑑2 ψ
- 𝛽 2 ψ = 0 for –b < x < 0
𝑑𝑥 2

The solutions of these equations must be in the form of Bloch theorem.

Ψ(x) = 𝑒 𝑖𝑘𝑥 uk(x)

𝑑ψ 𝑑u
= 𝑒 𝑖𝑘𝑥 𝑑𝑥 + u (ik) 𝑒 𝑖𝑘𝑥 and
𝑑𝑥

𝑑2 ψ 𝑑2 u 𝑑u 𝑑u
= 𝑒 𝑖𝑘𝑥 + 𝑒 𝑖𝑘𝑥 (𝑖𝑘) 𝑑𝑥 + u (ik)2𝑒 𝑖𝑘𝑥 + (ik) 𝑒 𝑖𝑘𝑥 𝑑𝑥
𝑑𝑥 2 𝑑𝑥 2

𝑑2 u 𝑑u
= 𝑒 𝑖𝑘𝑥 + 2ik 𝑒 𝑖𝑘𝑥 𝑑𝑥 - k2u𝑒 𝑖𝑘𝑥
𝑑𝑥 2

Thus,

𝑑2 u 𝑑u
𝑒 𝑖𝑘𝑥 + 2ik 𝑒 𝑖𝑘𝑥 𝑑𝑥 - u k2𝑒 𝑖𝑘𝑥 + 𝛼 2 u𝑒 𝑖𝑘𝑥 = 0
𝑑𝑥 2

Divided both side by 𝑒 𝑖𝑘𝑥

𝑑2 u 𝑑u
⇒ 𝑑𝑥 2 + 2ik 𝑑𝑥 + (𝛼 2 − 𝑘 2 ) u = 0; for 0 < x < a

𝑑2 u 𝑑u
⇒ 𝑑𝑥 2 + 2ik 𝑑𝑥 - (𝛽 2 + 𝑘 2 ) u = 0; for –b < x < 0

𝑑2 u 𝑑u
⇒ 𝑑𝑥 2 + 2ik 𝑑𝑥 + (𝛼 2 − 𝑘 2 ) u = 0; for 0 < x < a

Let the trial solution is u = 𝑒 𝑚𝑥

m2 + 2ik𝑚 + (𝛼 2 − 𝑘 2 ) = 0

−2𝑖𝑘 ± √(4𝑖 2 𝑘 2 −4(𝛼2 −𝑘 2 ))

m= 2.1

−2𝑖𝑘 ±2𝑖𝛼
= = i (−𝑘 ± 𝛼)
2

U1 = A𝑒 𝑖(𝛼−𝑘)𝑥 + B𝑒 −𝑖(𝛼+𝑘)𝑥
Similarly,

𝑑2 u 𝑑u
+ 2ik 𝑑𝑥 - (𝛽 2 + 𝑘 2 ) u = 0
𝑑𝑥 2

u = 𝑒 𝑚𝑥 ; m2 + 2ik𝑚 + (𝛽 2 + 𝑘 2 ) = 0

−2𝑖𝑘 ± √(4𝑖 2 𝑘 2 −4(𝛽 2 +𝑘 2 ) )

m= = −𝑖𝑘 ± 𝛽
2.1

U2 = C𝑒 (𝛽−𝑖𝑘)𝑥 + D𝑒 −(𝛽+𝑖𝑘)𝑥 ; –b < x < 0

Where A, B, C, D are constants.

Use the boundary conditions:

U1(0) = U2 (0)

𝑑𝑈 𝑑𝑈
( 𝑑𝑥1 )x=0 = ( 𝑑𝑥2 )x=0; For continuity of wave function and their derivative.

U1(a) = U2(-b); Periodicity of the function U(x)

𝑑𝑈 𝑑𝑈
( 𝑑𝑥1 )x = a = ( 𝑑𝑥2 )x = -b

A+B=C+D

Ai (𝛼 − 𝑘) – Bi (𝛼 + 𝑘) = C (𝛽 − 𝑖𝑘) – D (𝛽 + 𝑖𝑘)

A𝑒 𝑖(𝛼−𝑘)𝑎 + B𝑒 −𝑖(𝛼+𝑘)𝑎 = C𝑒 −(𝛽−𝑖𝑘)𝑏 + D𝑒 +𝑖(𝛽+𝑖𝑘)𝑏

A𝑖(𝛼 − 𝑘)𝑒 𝑖(𝛼−𝑘)𝑎 - B𝑖(𝛼 + 𝑘)𝑒 −𝑖(𝛼+𝑘)𝑎 = C(𝛽 − 𝑖𝑘)𝑒 −(𝛽−𝑖𝑘)𝑏 - D(𝛽 + 𝑖𝑘)𝑒 +𝑖(𝛽+𝑖𝑘)𝑏

These four equations have a definite solution only if the determinant formed by the coefficient
of A, B, C, D vanishes. This led to the condition

𝛽 2 −𝛼2
sinh 𝛽b sin 𝛼a + cosh 𝛽b cos 𝛼a = cos k(a+b)
2𝛼𝛽

Kronig and Penny considered the case where the potential barrier becomes delta function i.e.,
Vo tends to infinity and b approaches zero. But the product Vob remains finite.

2𝑚(𝑉0 −𝐸) 2𝑚𝐸

[ − 2 ]
ℏ2 ℏ
sinh 𝛽b sin 𝛼a + cosh 𝛽b cos 𝛼a = cos k(a+b)
2𝛼𝛽
As b → 0 and 𝑉0 → ∞

2𝑚𝑉0
⇒ ℏ2 2𝛼𝛽 𝛽b sin𝛼a +cos𝛼a = coska

𝑚𝑉0 𝑏
⇒ sin𝛼a +cos𝛼a = coska
ℏ2 𝛼

𝑚𝑉0 𝑎𝑏
Let us define p = ℏ2

sin𝛼a
p +cos𝛼a = coska
𝛼𝑎

Here p (Vob) is the measure of area of the potential barrier. The theoretical meaning of p is:
increase of p means, the binding of the electron becomes more stronger to a particular potential
well.

3𝜋
The left-hand side of the equation is represented as a function of 𝛼a for the value of p = .
2

𝛼 2 is proportional to energy and hence α is a measure of energy. The right-hand side accept
only the values between +1and -1 because cosine can lie between +1 and -1.

Salient features:

(i) The energy spectrum of the electron consists of a number of allowed energy bands
separated by the forbidden regions.
(ii) The width of the energy band increases with the increasing values of 𝛼a i.e., with
the increasing energy.
(iii) As p increases, width of the allowed bands decreases and with p → ∞, the allowed
regions become narrow and it becomes line spectrum. In that case the only solution
exits if sin 𝛼a = 0, i.e.

𝛼a = n𝜋

(𝛼a)2 = (n𝜋)2

2𝑚𝐸
a2 = n2𝜋2;
ℏ2

𝜋 2 ℏ2
E = 2𝑚𝑎2 n2

n𝜋
(iv) The variation of energy E with wave vector k shows discontinuity at K= ; where
𝑎

n = 1, 2, 3 …. The k values define the boundaries of first, second Brillouin zone etc.
𝜋 𝜋
The first Brillouin zone is - to + 𝑎. The second Brillouin zone consists of two
𝑎
𝜋 2𝜋 𝜋 2𝜋
parts, - 𝑎 to - and + 𝑎 to + .
𝑎 𝑎
 Tight Binding method for Energy Band

Consider two hydrogen Atoms, each with an electron in the 1s ground state. ψA and ψB are the
wavefunctions of the separated atoms. As they are brought together, their wavefunctions
overlap. The contributions are ψA ± ψB. The electron in the state ψA+ ψB will have somewhat
lower energy than ψA- ψB.

In the state ψA+ ψB, the electron spends part of the time in the region midway between the two
protons and in this region, it is in the attractive potential of both protons thereby increasing the
binding energy. In the state ψA- ψB, the probability density vanishes midway between the
nuclei, extra binding does not appear. As two atoms are brought together, two separated energy
levels are formed for each level of the isolated atom. For N atoms brought together, N orbitals
are formed for each orbital of the isolated atom.

The approximation that starts out from the wavefunctions of the free atoms is known as the
tight binding energy approximation or the LCAO (Linear combination of Atomic Orbitals).
The tight binding approximation is ideally suited to deal with inner core electrons. It has been
successfully applied to d-electron in transition metals and to the valence electrons in diamond
like and in each gas crystals. Simplest case is s-electron.
Consider an s-electron in an isolated atom positioned at rn with the ground state wavefunction
φ(r-rn) where r determines the electrons position in space.
𝐻𝑜 𝜑(𝑟 − 𝑟𝑛 ) = 𝜀𝑜 𝜑(𝑟 − 𝑟𝑛 )
Where, Ho and εo are the Hamiltonian and ground state energy of the electron in a free atom.

The Hamiltonian of an electron in the crystal is:

𝐻 = 𝐻𝑜 + 𝑉(𝑟 − 𝑟𝑛 )
ћ2 2
⇒ 𝐻 = [− ▽ + 𝑉𝑜 (𝑟 − 𝑟𝑛 ) + 𝑉(𝑟 − 𝑟𝑛 )]
2𝑚
With 𝑉(𝑟 − 𝑟𝑛 ) = ∑𝑚≠𝑛 𝑉𝑜 (𝑟 − 𝑟𝑚 ); the potential energy on the electron due to the atoms in
the vicinity of rn.

𝑉𝑜 (𝑟 − 𝑟𝑛 ) is the potential of an electron when localised at the isolated atom positioned at rn.
The influence of atoms in the vicinity of rn is treated as perturbation on Ho and is represented
by 𝑉(𝑟 − 𝑟𝑛 ).
𝐻𝛹𝑘 (𝑟) = 𝜀𝑘 𝛹𝑘 (𝑟)
Where, 𝜀𝑘 is the electron energy in the crystal and 𝛹𝑘 (𝑟) is the Bloch function.
Let the solution is

𝛹𝑘 (𝑟) = ∑ 𝑒𝑥𝑝 (𝑖𝑘. 𝑟𝑛 ) 𝜑(𝑟 − 𝑟𝑛 )

Where, 𝜑(𝑟 − 𝑟𝑛 ) is the wavefunction of electron for free atom.

First order energy is given by:

∫ 𝛹𝑘∗ 𝐻 𝛹𝑘 𝑑𝑉
𝜀𝑘 =
∫ 𝛹𝑘∗ 𝛹𝑘 𝑑𝑉

∫ 𝛹𝑘∗ (𝑟) 𝛹𝑘 (𝑟) 𝑑𝑉 = ∑ 𝑒𝑥𝑝[𝑖𝑘. (𝑟𝑛 − 𝑟𝑚 )] ∫ 𝜑 ∗ (𝑟 − 𝑟𝑚 ) ∗ 𝜑 (𝑟 − 𝑟𝑛 ) 𝑑𝑉

𝑛,𝑚

For strongly localised electron 𝜑 (𝑟 − 𝑟𝑚 ) is significant in the proximity of rm. Putting m = n,

∫ 𝛹𝑘∗ (𝑟) 𝛹𝑘 (𝑟) 𝑑𝑉 = ∑ 1 = 𝑁

𝑛
 As, ∫ 𝜑 ∗ (𝑟 − 𝑟𝑚 )𝜑(𝑟 − 𝑟𝑛 )𝑑𝑟 = 𝛿𝑚𝑛

1
𝜀𝑘 = ∑ 𝑒𝑥𝑝[𝑖𝑘. (𝑟𝑛 − 𝑟𝑚 )] ∫ 𝜑 ∗ (𝑟 − 𝑟𝑚 ) [𝜀𝑜 + 𝑉(𝑟 − 𝑟𝑛 )] ∗ 𝜑 (𝑟 − 𝑟𝑛 ) 𝑑𝑉
𝑁
𝑛,𝑚

For the term containing εo, we neglect the overlap between the nearest neighbours, so we have

𝜀𝑜
∑ 𝑒𝑥𝑝[𝑖𝑘. (𝑟𝑛 − 𝑟𝑚 )] ∫ 𝜑 ∗ (𝑟 − 𝑟𝑚 ) ∗ 𝜑 (𝑟 − 𝑟𝑛 ) 𝑑𝑉 = 𝜀𝑜
𝑁
𝑛,𝑚

Including the overlap up to nearest neighbours for the perturbation term, we write

∫ 𝜑 ∗ (𝑟 − 𝑟𝑚 ) 𝑉(𝑟 − 𝑟𝑛 ) ∗ 𝜑 (𝑟 − 𝑟𝑛 ) 𝑑𝑉 = −𝛼 ; 𝑜𝑛 𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑎𝑡𝑜𝑚 𝑚 = 𝑛

∫ 𝜑 ∗ (𝑟 − 𝑟𝑚 ) 𝑉(𝑟 − 𝑟𝑛 ) ∗ 𝜑 (𝑟 − 𝑟𝑛 ) 𝑑𝑉 = −𝛾 ; 𝑏𝑒𝑡𝑤𝑒𝑒𝑛 𝑛𝑒𝑎𝑟𝑒𝑠𝑡 𝑛𝑒𝑖𝑔ℎ𝑏𝑜𝑢𝑟𝑠 𝑚 ≠ 𝑛

All terms in the summation over n, each of which is evaluated over all m, are equal in
magnitude. Since the summation over n runs over all atoms in the crystal, the sum is simply N
times the value of single term.
(𝑟𝑛 −𝑟𝑚 )
𝜀𝑘 = 𝜀𝑜 − 𝛼 − 𝛾 ∑ 𝑒 𝑖𝑘.
𝑚

The sum is carried only over the nearest neighbours.

In Simple Cubic with lattice constant ‘a’;

𝑟𝑛 − 𝑟𝑚 = (±𝑎, 0,0), (0, ±𝑎, 0) 𝑎𝑛𝑑 (0,0, ±𝑎)
For s-state energy is
𝜀𝑘 = 𝜀𝑜 − 𝛼 − 2𝛾 (𝑐𝑜𝑠 𝑘𝑥 𝑎 + 𝑐𝑜𝑠 𝑘𝑦 𝑎 + 𝑐𝑜𝑠 𝑘𝑧 𝑎)
At the bottom; k = 0. So, energy is
𝜀𝑏𝑜𝑡 = 𝜀𝑜 − 𝛼 − 6𝛾
At the top; k = ± π/a. the energy is
𝜀𝑡𝑜𝑝 = 𝜀𝑜 − 𝛼 + 6𝛾
And the bandwidth = 12γ
As γ determines the bandwidth, α is interpreted as the lowering of energy of the free atoms
level on forming the solid. As one proceeds from inner to outer shells, the width of the repulsive
energy bands goes on increasing because of more and more overlap.
 Wigner-Seitz Method (Cellular Method)

The first systematic calculation of energy bands appeared in the form of this model. The
technique of calculation, referred to as cellular method, is based on the symmetry properties of
a certain primitive cell designed by Wigner and Seitz.
The alkali metals have a BCC structure for which Wigner-Seitz cell is polyhedron. The initial
volume of the crystal is imagined to have been filled up with identical polyhedra, assuming
there is only one electron in a given polyhedron at a time, together with the positively charged
metal ion at the centre. The electron is moving in a spherically symmetric potential field of the
ion.
Consider an extreme case when k = 0. The Bloch wave function ψk(r) is given by: ψk(r) = uo(r).
The wave function is periodic in crystal, i.e., the wave function remains unchanged when
𝜕𝜓
translated from one face to the opposite face of the cell. It requires that 𝜕𝑛 = 0 at the boundary

of the polyhedron where ‘n’ is normal to any of faces of the polyhedron. The above condition
in the crystal replaces the free atom.
𝜕𝜓
Boundary condition: 𝜓(𝑟) → 0 𝑎𝑠 𝑟 → ∞ by ( ) = 0 where, rs is the radius of the sphere
𝜕𝑟 𝑟=𝑟𝑠

to which polyhedron is approximated.

Wigner-Seitz gave an estimate of uo(r).

The radial Schrodinger wave equation is:

ћ2
[− 2𝑚 ▽2 + 𝑉𝑜 (𝑟)] 𝛹(𝑟) = 𝜀𝑜 𝛹(𝑟)
1 𝜕 2 𝜕 2𝑚
⇒[ 2
(𝑟 ) + 2 {𝜀𝑜 − 𝑉𝑜 (𝑟)}] 𝛹(𝑟) = 0
𝑟 𝜕𝑟 𝜕𝑟 ћ

Vo(r) is the potential energy if s-electron and εo is the energy value in the field of crystalline
ion within one of polyhedral.

For a general wavevector,

ћ2 2
[− ▽ + 𝑉𝑜 (𝑟)] 𝑢𝑘 (𝑟) exp(𝑖𝑘. 𝑟) = 𝜀𝑘 𝑢𝑘 (𝑟) exp(𝑖𝑘. 𝑟)
2𝑚
⇒ ▽2 [𝑢𝑘 (𝑟) exp(𝑖𝑘. 𝑟)] =▽ [𝑖𝑘 𝑢𝑘 (𝑟) exp(𝑖𝑘. 𝑟) + exp(𝑖𝑘. 𝑟) ▽ 𝑢𝑘 (𝑟)]
⇒ ▽2 [𝑢𝑘 (𝑟) exp(𝑖𝑘. 𝑟)] = exp(𝑖𝑘. 𝑟) [▽2 𝑢𝑘 (𝑟) − 𝑘 2 𝑢𝑘 (𝑟) + 2𝑖𝑘.▽ 𝑢𝑘 (𝑟)]

Substituting in the equation:

ћ2 ћ2 𝑘 2
[− (▽2 + 2𝑖𝑘 ▽) + 𝑉𝑜 (𝑟)] 𝑢𝑘 (𝑟) = (𝜀𝑘 − ) 𝑢𝑘 (𝑟)
2𝑚 2𝑚

k.▽ term is a perturbation and uk(r) obey the boundary condition. This gives the energies εk
ћ2 𝑘 2
in the form. 𝜀𝑘 = 𝜀𝑜 + which gives energies in the shape of a band as measured from the
2𝑚

level εo.

For 3s electron in Na-atom:

For free atom in Schrodinger wave equation, using boundary condition Ψ(r) → 0 as r →∞,
the energy eigenvalue is 5.15 eV.
𝜕𝛹
For metal at k = 0, subjected to Wigner-Seitz boundary condition, that ( 𝜕𝑟 ) = 0 where,
𝑟=𝑟𝑠

‘r’ is the mid-way between the atoms, the energy εo is -8.2 eV, considerably lower than the
free atoms.

Fig.: Variation of radial wavefunction for 3s electron in sodium.

The calculated energy of 3s orbitals at the zone boundary is 2.7 eV. These orbitals are empty
because 3s atom is half filled.
Estimation of Cohesive Energy:
The ground state energy of an electron in a crystal is less than that in a free atom. This lowering
of energy is the measure of cohesive energy.
The ground state energy of the valence electron in a crystal is
ћ2 𝑘 2 3
𝜀𝑘 = 𝜀𝑜 + = 𝜀𝑜 + 𝜀𝐹
2𝑚 5
Where, εo is the energy of the electron in a crystal at k = 0 and the second term is average
kinetic energy per electron.
For sodium,
𝜀𝑜 = -8.2 eV
𝜀𝐹 = 3.1 eV from tables

3
𝜀𝑘 = 𝜀𝑜 + 𝜀𝐹
5
3
𝜀𝑘 = −8.2 + × 3.1
5
𝜀𝑘 = −8.2 + 1.86
𝜀𝑘 = −6.34 𝑒𝑉

The energy of an electron in a free atom = -5.15 eV

So, cohesive energy = -5.15 – (-6.34) = 1.19 eV
Experimental value is 1.13 eV which shows that success of Wigner-Seitz approximately.
 Pseudopotentials
The orbitals of conduction electrons in metals are simple and smooth in the region between the
ion cores but are complicated by a nodal structure in the region of the core as shown in figure
(earlier theory). In the outer region the potential energy of the conduction electron is relatively
weak. In the outer regions, the wave functions are like plane waves.
If the conduction orbitals are like plane waves, the energy must depend on the wavevector.
ћ2 𝑘 2
𝜀𝑘 = , as for free electron.
2𝑚

We can replace the actual potential energy in the core by an effective potential energy, that
gives same wave functions outside the core as does the actual potential. These are
pseudopotentials. These are not unique. In the empty core model, we can take the unscreened
pseudopotential to be zero inside of some radius Re.

𝑈(𝑟) = 0 𝑓𝑜𝑟 𝑟 < 𝑅𝑒

𝑒2
𝑈(𝑟) = − 𝑓𝑜𝑟 𝑟 > 𝑅𝑒
𝑟
Where, Re =1.66 ao, ao is Bohr’s radius.
This potential is to be screened, that is each Fourier component U(k) of U(r) must be divided
by the dielectric function ε(k) of the electron gas.
The pseudopotential is much weaker than the true potential, but the wave functions in the outer
regions are nearly identical. The actual potential is much larger than the pseudopotential over
200 times larger at r = 0.15.
 Fermi Surface

The Fermi surface is the surface of constant energy εF in k-space. The fermi surface separates
the unfilled orbitals from the filled orbitals at absolute zero. The electrical properties of the
metal are determined by the shape of the Fermi surface, because the current is due to changes
in the occupancy of states near the Fermi surface.

Fig.: Brillouin zones of a square lattice in two dimensions.

The constant energy surface (the Fermi surface) of free electron, for a certain electron
concentration of the square lattice with its first three Brillouin zones.
Circle is a surface of constant energy for free electrons. It will be the fermi surface which
depends on the electron concentration.

The Fermi surface enclosed only the first zone completely. The parts of a zone are given by the
symbols a, b, c and d.

Mapping of first zone, second zone and third zone in reduced zone scheme. Parts of second
and third zones are put together in a square on translating them by reciprocal lattice vectors of
appropriate size.
 Fig.: Mapping of the 1st, 2nd and 3rd Brillouin zone schemes.

The shaded regions are the occupied electron states. Only the first zone is completely occupied.
 Quantization of Orbits in a Magnetic field (Landau Levels)

Consider free electrons in a cubical box of length L parallel to x. y, z axes. When a magnetic
field B is applied along the z-direction, the classical equation of motion of an electron is
𝑑2𝑟 𝑑𝑟
𝑚 2
= 𝑒 ( × 𝐵)
𝑑𝑡 𝑑𝑡
Where the Lorentz force is perpendicular to velocity and B. the force component along z-
𝑑2 𝑟
direction is zero, i.e., 𝑚 𝑑𝑡 2 = 0 that describes a free and steady motion.

The energy along z-direction remains unchanged. The electron moves in x-y plane on a circular
𝑒𝐵
path. The Lorentz force equals the centripetal force, giving 𝑤𝑐 = , where wc is the cyclotron
𝑚
1
frequency. The electron wave function along z-direction is exp (𝑖𝑘𝑧 𝑧) which has the
√𝐿
ћ2 𝑘𝑧2 𝜋𝑛𝑧
eigenvalue 𝜀𝑧 = ⇒ 𝑘𝑧 = ; nz being an integer.
2𝑚 𝐿

But the solution of one-electron Schrodinger wave equation in xy-plane yield an energy that is
quantized in steps of ћwc.
The electron energy in the xy-plane is
1
𝜀1 = (𝑛 + 2) ћ𝑤𝑐 ; where n is a positive integer.

Total energy is
𝜀𝑛 (𝑘𝑧 ) = 𝜀𝑧 + 𝜀1
ћ2 𝑘𝑧2 1
⇒ 𝜀𝑛 (𝑘𝑧 ) = + (𝑛 + 2) ћ𝑤𝑐
2𝑚

This above relation expresses orbit quantization.

Each group of degenerate energy levels as defined by a pair of quantum numbers n and kz is
called a Landau levels.
ћ𝑑𝑘 𝑑𝑟
We know = 𝑒 ( 𝑑𝑡 × 𝐵) assuming there is no scattering effect on electron.
𝑑𝑡

The tip of the wave vector k1 (k-perpendicular) in the x-y plane travers out a circle whose area
in k-space is πk12. According to quantum mechanics, the energy ε1 in the x-y plane can be
ћ2 𝑘12
expressed as .
2𝑚
 ћ2 𝑘12 1 1 𝑒𝐵
= (𝑛 + ) ћ𝑤𝑐 = (𝑛 + ) ћ
2𝑚 2 2 𝑚
1 2𝑒𝐵
⇒ 𝑘12 = (𝑛 + )
2 ћ

The area of the orbit in k-space is

1 2𝜋𝑒𝐵
𝑆 = 𝜋𝑘12 = (𝑛 + )
2 ћ

The difference between areas of consecutive orbits in k-space is

2𝜋𝑒𝐵
𝑆𝑛+1 − 𝑆𝑛 = ( )
ћ
2𝜋𝑒𝐵
Area enclosed by the electron path in a magnetic field are quantized in units of . The
ћ

quantized areas are cylindrical in shape. They are known as Landau tubes.

Landau levels
 Cyclotron Resonance

The phenomenon of cyclotron resonance in metals was first predicted by Azbel-Kaner in 1956
and the experimental verification was made on tin by Fawcett in the same year. The most
suitable experimental arrangement employed to observe cyclotron resonance is known as
Azbel-Kaner geometry where the applied alternating electric field may be parallel or
perpendicular to the static magnetic field B1 but both E and B1 must be parallel to the surface
of the specimen as shown in the figure.

Fig.: Azbel-Kaner resonance

Azbel-Kaner showed that the cyclotron resonance could be observed under the following
conditions:

𝜆 > 𝑟𝐵 , 𝑎𝑛𝑑 𝑟𝐵 ≫ 𝛿 (1)

i.e., the cyclotron resonance is observed in the extreme anomalous region, where λ, rB and δ
are mean free path of electron, radius of the circular path of electrons and skin depth
respectively. The first inequality in equation (1) is a well-known condition of cyclic motion
according to which the electrons must complete more than one rotation around a closed orbit
before being scattered. Further, since under the above experimental geometry, the electrons can
either describe closed orbits in planes perpendicular to the surface or move along helices whose
axes are parallel to the direction of the magnetic field. The second inequality in equation (1)
implies that the electron can interact with the electric field within the skin layer only during a
small fraction of the period of their motion. They absorb a little energy each time when they
pass through the skin layer. The resonant absorption of energy can occur only if the electrons
experience an electric field of the same phase each time when they enter the skin layer. This is
possible when the frequency of the oscillating electric field is an integral multiple of the
cyclotron frequency, i.e.,
𝜔 = 𝑛𝜔𝑐 (2)

This is called Azbel-Kaner resonance.

As we already know the cyclotron frequency is given by:

𝑒𝐵
𝜔𝑐 = 𝑚 (3)
𝑐

Substituting the value of cyclotron frequency 𝜔c from (3), we obtain:

𝑒𝐵 1 𝑛𝑒
𝜔 = 𝑛 𝑚 𝑜𝑟 = 𝜔𝑚 (4)
𝑐 𝐵 𝑐

Where mc is the cyclotron effective mass and is given by:

ћ2 𝑑𝐴𝑘
𝑚𝑐 = 2𝜋 (5)
𝑑𝐸

Where, Ak is the area of the electron orbit in k-space.

Fig.: Orbits of an electron in k-space

Substituting equation (5) in equation (4), we obtain the reciprocal value of magnetic field when
nth harmonic is observed as:

1 2𝜋𝑛𝑒 𝑑𝐸
= (6)
𝐵(𝑛) 𝜔ћ2 𝑑𝐴𝑘

In practice, one keeps the frequency 𝜔 constant and varies the magnetic field, then the expected
periodicity with which the successive resonant peaks appear can be given by the interval of
reciprocal magnetic fields between two consecutive resonances [e.g., the nth and (n+1)th] as:
 1 1 1 2𝜋𝑒 𝑑𝐸
𝛿 (𝐵) = 𝐵(𝑛+1) − 𝐵(𝑛) = 𝜔ћ2 𝑑𝐴 (7)
𝑘

An experimental measurement of a quantity which is proportional to derivative of the real

component of the impedance (Z=R+iX) over the magnetic field, dR/dB for copper at 24 GHz
is depicted in figure below:

Fig.: Field derivative of the surface impedance (dR/dB) of copper as a function of the field.

This cyclotron resonance curve makes it possible to determine very accurately the magnitude
of the cyclotron mass and the relaxation time τ of the corresponding group of carriers. However,
the observation of cyclotron resonance requires a complicated experimental technique, a highly
pure and perfect single crystal accuracy in the orientation of the crystal relative to the electric
and magnetic field vectors.