Free Electron Theory of Metals

The outstanding properties of metals:

1) Metals have high electrical and thermal conductivities.
2) In the steady state Ohm’s law is obeyed. i.e.,
𝑖∝𝐸
3) Resistivity above Debye’s temperature is proportional to absolute temperature. i.e.,
𝜌∝𝑇
4) At low temperature, the resistivity of the metal is proportional to the fifth power of
absolute temperature. i.e.,
𝜌 ∝ 𝑇5
5) For most metals, resistivity is inversely proportional to pressure. i.e.,
1
𝜌∝
𝑃
6) Wiedemann and Franz’s law: Above Debye’s temperature the ratio of thermal and
electrical conductivities is proportional to the absolute temperature. i.e.,
𝐾
∝𝑇
𝜎
7) Superconductivity: At absolute zero, the resistivity tends towards zero.
Drude-Lorentz free electron theory of metals

 In 1900 Drude made a very good proposal that metals are composed of
positive metal ions whose valence electrons are free to roam among the ionic
array, with the only restriction that they are confined to remain within the
boundaries of the metals.

 The metal ions are bounded to the electrons by an electrostatic attraction

between their positive charges and the negatively charged electron gas.

 The parent atoms of the metal are regarded as continuously dissociating into
negative electrons and these electrons are free to move in all directions inside
the metal like the atoms or molecules of a perfect gas.

 In general, their motion is random, but in an electric field, the negative

charged electrons stream in the positive field direction and produce a current
in the metal. In order to prevent the electrons from accelerating indefinitely it
was assumed that they collide elastically with metal ions. This leads to a
steady state current which is proportional to an applied voltage and explains
the origin of Ohm’s law.
 Since the conduction electrons move in a uniform electrostatic field of ion
cores, their potential energy remains constant and is normally taken as zero.
i.e., the existence of ion core is ignored. Thus the total energy of a
conduction electron is equal to its kinetic energy.

 Since the movement of conduction electrons is restricted to within the

crystal only, the potential energy of a stationary electron inside a metal is
less than the potential energy of an identical electron just outside it. This
energy difference, 𝑉0 , serves as a potential barrier and stops the inner
electrons from leaving the surface of the metal.

 Thus, in the free electron gas model, the movement of the electrons in a
metal is equivalent to the movement of a free electron gas inside a potential
energy box.
 Based on the Drude’s considerations that the electron gas behaves as a
perfect gas. H. A. Lorentz postulated in 1909 that the electrons of electron
gas obey Maxwell-Boltzmann statistics under equilibrium conditions. Since
the theory is based on the classical ideas of Maxwell-Boltzmann statistics, it
is called classical theory.

 Success:

This theory has been applied to explain the following properties of metals in a
quite satisfactory way:

1) Ohm’s law is obeyed, 𝑖 ∝ 𝐸.

2) For impure metals, electrical conductivity is inversely proportional to the
square root of absolute temperature,
1
𝜎∝
𝑇
1
3) For pure metals, 𝜎 ∝
𝑇
4) For most metals, resistivity is inversely proportional to pressure,
1
𝜌∝
𝑃
 𝐾
5) Wiedemann-Franz relation is obeyed, ∝ 𝑇.
𝜎

6) High lustre and complete opacity of metals.

 Limitations:

 In this theory, the mutual repulsion between the negatively charged electrons
was neglected and the potential field due to positive ions was assumed to be
everywhere constant.

 This theory failed to predict the temperature dependent resistivity accurately.

 This theory also failed to explain the heat capacity and paramagnetic
susceptibility of the conduction electrons.
Electrical Conductivity
If free time (i.e., time taken between two successive collisions) be 𝜏
and velocity along one direction be 𝑢, then
𝜆
𝜏 = … … … … … … … … … … … . (1)
𝑢
Where, 𝜆 is the mean free path.
If the applied field on the electron of charge –e be E then the equation of motion
is
𝑑2 𝑥
𝑚 2 = −𝑒𝐸
𝑑𝑡
2
𝑑 𝑥 𝑒
=− 𝐸
𝑑𝑡 2 𝑚
Integrating it we get
𝑒
𝑥 = − 𝐸𝑡 + 𝑐
𝑚
At 𝑡 = 0, 𝑥=0 ∴𝑐=0

𝑑𝑥 𝑒
Hence =𝑥=− 𝐸𝑡
𝑑𝑡 𝑚
Average velocity between two collisions
𝜏
1 𝑒𝐸
𝑥= − 𝑡𝑑𝑡
𝜏 𝑚
0
1 𝑒𝐸 𝜏 2
=− . .
𝜏 𝑚 2
𝑒𝐸 𝜏
𝑥 = − . … … … … … … … … … (2)
𝑚 2
If 𝑖 is the current density and 𝑛 the number of electrons per unit volume, then we
have,
𝑖 = −𝑛𝑒𝑥
𝑛𝑒 2 𝐸 𝜏
= .
𝑚 2
2
1 𝑛𝑒 𝐸 𝜆
= . . … … … … … … … … … … … . (3)
2 𝑚 𝑢
Multiplying and dividing by 𝑢, we get
1 𝑛𝑒 2 𝐸 𝜆𝑢
𝑖= . .
2 𝑚 𝑢. 𝑢
1 3
Since we know that 𝑚𝑢2 = 𝑘𝑇
2 2
⇒ 𝑚𝑢2 = 3𝑘𝑇
 Substituting the value in (3) we get
1 𝑛𝑒 2 𝐸. 𝜆𝑢
𝑖= .
2 3𝑘𝑇
2
𝑛𝑒 𝜆𝑢
= .𝐸
6𝑘𝑇
⇒ 𝑖 = 𝜎𝐸
𝑛𝑒 2 𝜆𝑢
Where σ = is known as electrical conductivity.
6𝑘𝑇

Thermal Conductivity

On the basis of kinetic theory, if two

temperatures are equal, i.e., if 𝑇1 = 𝑇2 ,
there is no transfer of energy.

If 𝑇1 > 𝑇2 , there is transfer of energy

from 𝐸 → 𝐹. The number of electrons
1
per unit area per unit time is 𝑛𝑢 and
6
1 2.
each electron has energy 𝑚𝑢1
2
Hence total energy transferred from E to F
1 1
= 𝑛𝑢 . 𝑚𝑢1 2
6 2
1 3
= 𝑛𝑢 . 𝑘𝑇1
6 2
1
= 𝑛𝑢𝑘𝑇1 … … … … … … … … . (1)
4
Similarly the energy transferred from F to E
1
= 𝑛𝑢𝑘𝑇2 … … … … … … … … … … . (2)
4
Hence, net energy transferred from E to F per unit area per unit time
1
= 𝑛𝑢𝑘 𝑇1 − 𝑇2 … … … … … … … … . (3)
4
If K be the thermal conductivity, then the transfer of energy per unit area per
unit time
𝑇1 − 𝑇2
=𝐾. … … … … … … … … . (4)
2𝜆
From (3) and (4) we get
𝑇1 − 𝑇2 1
𝐾. = 𝑛𝑢𝑘 𝑇1 − 𝑇2
2𝜆 4
1
⇒ 𝐾 = 𝑛𝑢𝜆𝑘 … … … … … … … … … (5)
4
Weidemann-Franz relation: The ratio between thermal and electrical
conductivities is known as Weidemann-Franz ratio. Now,
𝐾 1 6𝑘𝑇
= 𝑛𝑢𝜆𝑘 ×
𝜎 2 𝑛𝑒 2 𝜆𝑢
2
𝑘
=3 𝑇
𝑒
𝑘 2
Here 3 = 𝐿 is known as Lorentz number.
𝑒

𝐾
∴ = 𝐿𝑇
𝜎

𝐾
⟹ ∝ 𝑇
𝜎
i.e., Weidemann-Franz ratio is directly proportional to temperature.
Sommerfeld’s free electron theory of metals

Considering the free nature of valence electrons, Sommerfeld treated the

problem quantum mechanically using the Fermi-Derac statistics.

For mathematical simplicity, consider first an electron of mass ‘m’ is bound

to move in a one dimensional crystal of length L. The electron is prevented
from leaving the crystal by the presence of a large potential energy barrier
at it surfaces.

[Although the barriers extend over a few atomic layers near the surface,
these are taken infinitely large for the sake of simplicity.]

The problem is identical to that of an electron moving in a one-dimensional

potential box as shown in the Fig. (1).
The potential energy everywhere within the crystal
∞ ∞
or box is constant and is equal to zero.

Thus,
𝑉 𝑥 = 0 𝑓𝑜𝑟 0 < 𝑥 < 𝐿

V(x)
𝑉 𝑥 = ∞ 𝑓𝑜𝑟 𝑥 ≤ 0 𝑎𝑛𝑑 𝑥 ≥ 𝐿

The wavefunction 𝜓𝑛 of the electron occupying the x

0 L
n-th state is determined from the solution of the
Schrödinger equation, i.e., Fig. (1)

𝑑 2 𝜓𝑛 2𝑚
+ 2 𝐸𝑛 − 𝑉 𝜓𝑛 = 0 … … … … … . . (1)
𝑑𝑥 2 ℏ

where 𝐸𝑛 represents the kinetic energy of the

electron in the n-th state and V is its potential
energy.
Since V=0 inside the box, equation (1) becomes

𝑑 2 𝜓𝑛 2𝑚
+ 2 𝐸𝑛 𝜓𝑛 = 0 … … … … … … … … . 2
𝑑𝑥 2 ℏ

The general solution of this equation is

𝜓𝑛 𝑥 = 𝐴 sin 𝑘𝑥 + 𝐵 cos 𝑘𝑥 … … … … … … … . (3)

where A and B are arbitrary constants to be determined from the boundary

conditions and k is given by
2𝑚𝐸𝑛
𝑘= … … … … … … … … … … … … … . (4)
ℏ2

The boundary conditions are-

𝜓𝑛 0 = 0 and 𝜓𝑛 𝐿 = 0

These conditions are based on the fact that at x=0 and L, 𝑉 → ∞

For x=0, equation (3) gives B=0 and the solution becomes
𝜓𝑛 𝑥 = 𝐴 sin 𝑘𝑥 … … … … … … … 5

Also, since 𝜓𝑛 𝐿 = 0 equation (5) yields

sin 𝑘𝐿 = 0

⇒ 𝑘𝐿 = 𝑛𝜋
𝑛𝜋
⇒𝑘= ; 𝑛 = 0, 1, 2, 3 … … … . .
𝐿
Thus the expression (5) becomes
𝑛𝜋
𝜓𝑛 𝑥 = 𝐴 sin 𝑥 … … … … … … … … (6)
𝐿

The allowed energy values can be obtained

ℏ2 𝑛𝜋 2
𝐸𝑛 =
2𝑚 𝐿
2
𝑛 ℎ 2
= 2
… … … … … … … … … (7)
8𝑚𝐿

⇒ 𝐸𝑛 ∝ 𝑛2
It is apparent from (6) and (7), the allowed wavefunctions 𝜓𝑛 𝑥 and
allowed energy values 𝐸𝑛 exist only for integral values of n. The number n
is called the quantum number. Hence the energy spectrum consists of
discrete energy levels. [The spacing between the levels is determined by
the values of n and L. If L is of the order of few cms, the energy levels
form almost a continuum.] The plot of 𝐸𝑛 vs n is shown in the Fig (2).

Fig. (2): 𝐸𝑛 versus n for a one-

dimensional crystal
The constant A in equation (6) is determined by using the condition that the
probability of finding an electron somewhere on the line is unity, i.e.,
𝐿

𝜓𝑛∗ 𝑥 𝜓𝑛 𝑥 dx = 1
0
𝐿
𝑛𝜋
⇒ 𝐴2 𝑠𝑖𝑛2 𝑥 dx = 1
𝐿
0
𝐿
𝐴2 2𝑛𝜋
⇒ 1 − 𝑐𝑜𝑠 𝑥 dx = 1
2 𝐿
0
𝐿
2
𝐴
⇒ dx = 1
2
0

2
⇒ 𝐴=
𝐿
Substituting the value of A in equation (6), we get the normalized wavefunction as
2 𝑛𝜋
𝜓𝑛 𝑥 = sin 𝑥 … … … … … … … … (8)
𝐿 𝐿
In three dimension a crystal can be approximated by a cube of edge L inside
which the potential energy is zero. Then the solution of equation (2) is just as
an extension of the one-dimensional normalized wavefunction i.e.,

8 𝑛𝑥 𝜋 𝑛𝑦 𝜋 𝑛𝑧 𝜋
𝜓𝑛 𝑟 = 3
𝑠𝑖𝑛 𝑥 𝑠𝑖𝑛 𝑦 𝑠𝑖𝑛 𝑧
𝐿 𝐿 𝐿 𝐿
8
Where 𝑛𝑥 , 𝑛𝑦 and 𝑛𝑧 are positive integers and is the normalizing
𝐿3
constant.

The corresponding form of the energy is

ℎ2 2 2 2
𝐸𝑛 = 𝑛 𝑥 + 𝑛 𝑦 + 𝑛 𝑧
8𝑚𝐿2
Momentum Space
𝑃2
The kinetic energy of an electron can be expressed to its momentum by 𝐸 =
2𝑚
and the potential energy is assumed to be equal to zero everywhere inside the
metal. Consequently the total energy of the electron equals the kinetic energy and
can be written as
𝑃2 ℎ2 2 2 2
𝐸= = 𝑛 𝑥 + 𝑛 𝑦 + 𝑛 𝑧 … … … … … … … (1)
2𝑚 8𝑚𝐿2
Rearranging,
2
2𝐿
𝑃 2 = 𝑛𝑥 2 + 𝑛𝑦 2 + 𝑛𝑧 2 … … … … … . . 2
ℎ
From this it follows that the momentum of an electron can be expressed similarly
in terms of its three quantum numbers.

The form of (2) suggests that all possible values of the momentum can be
represented in momentum space by a lattice of points corresponding to the three
quantum numbers.
Fermi-Derac Distribution

The electrons are distributed among the various possible energy levels in accordance
with the Pauli’s exclusion principle which states that no two electrons can have all
their quantum numbers identical i.e., each quantum state can be occupied by at the
most one electron.

Each energy level defined by the quantum number n can have two quantum states
(based on the two possible values of 𝑚𝑠 , + 1 2 𝑜𝑟 − 1 2 ) and hence can
accommodate a maximum of two electrons; one with spin up and the other with spin
down.

𝑁
Thus a total of N non-interacting electrons at 0K can be filled in energy levels. The
2
𝑁
topmost filled level is the ( ) th level and all the levels lying above it are empty.
2

The topmost filled energy level at 0K is known as Fermi level and the energy
corresponding to this level is called Fermi energy, is denoted as 𝐸𝐹0 . Thus, according
to quantum theory, even at absolute zero energy, the electrons are distributed among
the discrete energy levels having energies ranging from zero to 𝐸𝐹0 .
Effect of temperature on the occupancy of energy levels:

It is apparent that, for temperatures, greater than 0K, the Fermi level may not be
the topmost filled level since some of the electrons from the filled energy levels
may be excited to the higher energy levels.

Thus some of the levels below 𝐸𝐹0 would be empty while some above it would be
occupied.

The probability that a particular quantum state of energy E is occupied at a

temperature T is given by the Fermi function
1
𝑓 𝐸 = … … … … … … … … . (1)
𝐸 − 𝐸𝐹
𝑒𝑥𝑝 +1
𝑘𝐵 𝑇
Where 𝐸𝐹 is the Fermi energy.
At absolute zero, the equation (1) gives

𝑓 𝐸 =1 𝑓𝑜𝑟 𝐸 ≤ 𝐸𝐹0

=0 𝑓𝑜𝑟 𝐸 > 𝐸𝐹0

 Fig.: Fermi distribution function for three different temperatures. The variation of the Fermi energy with
temperature is shown.

This indicates that all the energy states below 𝐸𝐹0 are occupied and all the states
above it are empty, i.e., Fermi distribution function is a step function. Thus at 0K,
the Fermi level 𝐸𝐹0 represents the highest filled energy level.

For temperatures greater than 0K but less than the melting point of the metal such
that 𝑘𝐵 𝑇 ≪ 𝐸𝐹 , the distribution function loses its step character. The probability
of occupation, 𝑓(𝐸) decreases gradually from 1 to 0 near 𝐸𝐹 .
This indicates that some of the states below 𝐸𝐹 are empty while some others
above it are filled. This is because some of the electrons from the energy states
below 𝐸𝐹 gain thermal energy and get excited to the states above 𝐸𝐹 . At
𝐸 = 𝐸𝐹 , equation (1) gives
1
𝑓 𝐸 =
2

Thus, for temperatures greater than 0K, the Fermi level may be defined as the
1
level where the probability of occupation is . It is not the topmost filled level;
2
instead, it lies between the filled levels and empty levels. Apparently, the
position of the Fermi level is not fixed, but changes with temperature. An
approximate relationship between 𝐸𝐹 and 𝐸𝐹0 is

2
𝜋 2 𝑘𝐵 𝑇
𝐸𝐹 ≅ 𝐸𝐹0 1− … … … … … … … … … … … . . (2)
12 𝐸𝐹0
However, in case of metals, since the spacing between the levels is quite small
(~10-19 eV), the highest filled energy level is usually taken as the Fermi level.

For energies below 𝐸𝐹 such that (𝐸𝐹 −𝐸) ≫ 𝑘𝐵 𝑇, 𝑓 𝐸 = 1 , provided T

does not differ much from 0K. Thus it is only in the vicinity of 𝐸𝐹 minus a
few 𝑘𝐵 𝑇 that 𝑓 𝐸 becomes less than unity, i.e., some of the states below 𝐸𝐹
would be empty and some above 𝐸𝐹 would be filled.

It therefore, follows that as the temperature is raised above 0K, all the
electrons would not gain energy as expected classically, but only those which
lie within an energy range 𝑘𝐵 𝑇 below the Fermi level can do so. This is
because the electrons present near the Fermi level can jump to the vacant
higher energy states after acquiring energy of the order of 𝑘𝐵 𝑇 , but the
electrons present well below the Fermi level can not do due to the non-
availability of empty states within the range 𝑘𝐵 𝑇 .
 Thus according to quantum mechanics, only a small fraction of the electrons
can gain thermal energy and get excited to the higher energy states.

𝑘𝐵 𝑇
The value of this fraction is ≅ 0.01 at room temperature for 𝐸𝐹 =
𝐸𝐹
3.0 𝑒𝑉.

 Density of available electron states, D(E)

The density of states is defined as the total number of available electronic

states per unit energy range. It is denoted by D(E). i.e.,

𝑑𝑛
𝐷 𝐸 = …………………. 1
𝑑𝐸

where 𝑑𝑛 represents the number of quantum states present in the energy

interval E and E+dE.
Now we consider the linear momentum 𝑃 which is represented by the operator-

𝑃 = −𝑖ℏ𝛻
𝑃 𝜓𝑘 𝑟 = −𝑖ℏ𝛻 𝜓𝑘 𝑟 = ℏ𝑘 𝛻𝜓𝑘 𝑟

This indicates that the plane wavefunction 𝜓𝑘 is an eigen function of the linear
momentum with the eigen value ℏ𝑘. The possible velocity is given by

ℏ𝑘
𝑣= … … … … … (2)
𝑚

In a system of N free electrons, the occupied states or orbitals in the ground state
may be represented by points inside a sphere in the 𝑘-space. The energy
corresponding to the surface of the sphere then represents the Fermi energy.

Let 𝑘𝐹 be the wave vector from the origin of the 𝑘-space to the surface of the
sphere.
Then the Fermi energy is
ℏ2 2
𝐸𝐹 = 𝑘 … … … … … … (3)
2𝑚 𝐹

One allowed wave vector or one distinct

triplet of quantum numbers 𝑘𝑥 , 𝑘𝑦 and
𝑘𝑧 which corresponds to the volume
2𝜋 3
element of 𝑘-space.
𝐿
Fig.: A sphere of radius 𝑘𝐹 in the 𝑘-space containing
points represent the occupied states of a system of N
Thus, in the Fermi sphere of volume free electrons.
4 3
𝜋𝑘𝐹 , the total number of electronic
3
states are
3
4
3 𝜋𝑘𝐹 𝐿3 3 𝑉 3
=2. 3 = 2
𝑘 𝐹 = 2
𝑘 𝐹 = 𝑁 … … … … … … … … … … (4)
2𝜋 3𝜋 3𝜋
𝐿
where the factor 2 appears because there are two allowed values of 𝑚𝑠 and
hence two electronic states, corresponding to each value of k.

The number of states has been set equal to the number of electrons, N. From
equation (4)-
1
3𝜋 2 𝑁 3
𝑘𝐹 = … … … … … … … … (5)
𝑉
Now the Fermi energy from equation (3)
2
2 2 3
ℏ 3𝜋 𝑁
𝐸𝐹 = ………………. 6
2𝑚 𝑉

i.e., 𝐸𝐹 depends on both the electronic concentration and mass. The total
number of electrons is therefore,

3
𝑉 2𝑚𝐸𝐹 2
𝑁= 2
3𝜋 ℏ2
As all the orbital below the Fermi level are filled, as a general case we have
for a value of energy 𝐸 < 𝐸𝐹
3
𝑉 2𝑚𝐸 2
𝑁= 2
3𝜋 ℏ2

Where 𝐸 refers to the highest filled level energy.

Now, the density of states

3
𝑑𝑁 𝑑 𝑉 2𝑚𝐸 2
𝐷 𝐸 = =
𝑑𝐸 𝑑𝐸 3𝜋 2 ℏ2

3
𝑉 2𝑚 2 3 1
= 2 𝐸 2
3𝜋 ℏ2 2

3
𝑉 2𝑚 2
1
𝐷(𝐸) = 2 𝐸 2
2𝜋 ℏ2