5.4 Current flow through a non degenerated semiconductor

We investigate the charge flow in an electrical field for a non degenerated semiconductor. Combining the equations (5.18) and (5.21) we get

j = - q 4π3V kτ(k)v(k)f0 E (qE -rμ - (E -μ)r ln (T)) ,vd3k.

(5.24)

We will neglect gradients in the temperature, so (E -μ)r ln (T) vanishes. Since the semiconductor is not degenerated we can simplify the Fermi statistics by the Boltzmann statistics, i.e.

f0 E = -f0 kT.

(5.25)

We take the band energies of the free electron gas

E = E0 + m* 2 |v(k)| 2, sov i = m*ki.

(5.26)

In addition we assume τ to be independent of k. Summing up all approximations we find for the particle current

j = - qτ2 4π3kT(m*)2 [eE -μ]M~,

(5.27)

and M~ is a matrix with the components

M~ij =V kkikjf0(E(k))d3k =kdkk4f 0(E(k))SkdΩkikj k2 .

(5.28)

Using k2 = k x2 + k y2 + k z2 and integrating over the surface of a sphere we get

SkdΩkikj k2 = 4π 3 δij.

(5.29)

Using

dE = 2 m*kdk,

(5.30)

the current density is written as

j = - qτ 3π2kTm* [qE -μ] (2m* 2 )E0dE(E -E 0)3 2 exp (-E -μ kT ).

(5.31)

After partial integration we get

j = - qτ 3π2kTm* [qE -μ] (2m* 2 ) - 3kT 2 E0dE(E -E 0)1 2 exp (-E -μ kT ).

(5.32)

Taking into account the density of state of free electrons

D(E) = 1 2π2 (2m* 2 ) 3 2 (E -E0)1 2 ,

(5.33)

we finally get

j = qτ m* [qE -μ]E0dED(E)f 0(E) = qτne m* [qE -μ].

(5.34)