
Uncorrected Proof
BookID 160928 ChapID 05 Proof# 1 - 29/07/09
5.7 Energy Bands of Common Semiconductors 155
5.7 Energy Bands of Common Semiconductors 463
Many common semiconductors which crystallize in the cubic zincblende struc- 464
ture have valence–conduction band structures that are quite similar in gross 465
features. This results from the fact that each atom (or ion) has four electrons 466
outside a closed shell and there are two atoms per unit cell. For example, 467
silicon has the electron configuration [Ne]3s
2
3p
2
, i.e., two 3s electrons and 468
two 3p electrons outside a closed neon core. With two silicon atoms per 469
unit cell, this gives eight electrons per unit cell. The empty lattice has a 470
single Γ
1
band at E
Γ
= 0 and 8-fold degenerate bands at E
Γ
=3.The471
eightfold degeneracy is lifted by the periodic potential, so the valence and 472
conduction bands at Γ will arise from these eight bands. Germanium has 473
the electron configuration of [Ar]3d
10
4s
2
4p
2
, and III–V compounds like GaAs 474
-
Ga
[Ar]3d
10
4s
2
4p
1
As
[Ar]3d
10
4s
2
4p
3
.
look just like Ge if one 4p electron
475
transfersfromAstoGaleavingasomewhationicGa
−
As
+
molecule in the 476
unit cell instead of two Ge atoms. The same is true if any III–V elements 477
replace a pair of Si atoms or Ge atoms in a zincblende structure. 478
A nice example of the use of group concepts in studying energy band 479
structure is a simple nearly free electron type model used to give a rather good 480
description of the valence–conduction band semiconductors with zincblende 481
structures. We will give a rough sketch of the calculation, referring the reader 482
to an article by D. Brust
2
. To describe the band structures of Si and Ge, Brust 483
use the following 15 plane-wave wave functions corresponding to the 15 bands 484
at Γ which have energy E ≤ 4. (See the 15 bands at Γ in Fig. 5.9.) We can 485
write these 15 plane waves as w
i
,withi =1, 2, 3,...,15 defined by 486
w
0
=1E
0
(Γ) = 0,
w
1
= w
∗
5
=e
2πi
a
(x+y+z)
E
1
(Γ) = E
5
(Γ) = 3,
w
2
= w
∗
6
=e
2πi
a
(x−y−z)
E
2
(Γ) = E
6
(Γ) = 3,
w
3
= w
∗
7
=e
2πi
a
(−x+y−z)
E
3
(Γ) = E
7
(Γ) = 3,
w
4
= w
∗
8
=e
2πi
a
(−x−y+z)
E
4
(Γ) = E
8
(Γ) = 3,
w
9
= w
∗
12
=e
2πi
a
x
E
9
(Γ) = E
12
(Γ) = 4,
w
10
= w
∗
13
=e
2πi
a
y
E
10
(Γ) = E
13
(Γ) = 4,
w
11
= w
∗
14
=e
2πi
a
z
E
11
(Γ) = E
14
(Γ) = 4,
From these 15 functions w
0
,w
1
,...,w
14
, one can construct linear superposi- 487
tions belonging to IRs of the group of the wave vector Γ, X, L, etc. Some 488
examples are 489
2
D. Brust, Phys. Rev. 134, A1337 (1964).