Chapter 2.

Homogeneous Semiconductors
Homogeneous semiconductor : Semiconductor materials that consist of
one uniform material. For example, pure (intrinsic) Si or Si with impurities
uniformly distributed. We will discuss the concepts of “extrinsic” and
“nonuniformly doped semiconductors”.

The point of this chapter is to derive the electron and hole density near
the bottom of the conduction band and near the top of the valence band,
respectively.

Approach to derive the

density of electron and hole:

top top

no = ∫ S ( E ) f ( E )dE = ∫ n( E )dE
EC EC
EV EV EV

po = ∫
bottom
S ( E )[1 − f ( E )]dE = ∫ S(E ) f
bottom
p ( E )dE = ∫ p( E )dE
bottom

Semiconductor Engineering 53
 2.2 Pseudo-Classical Mechanics for Electrons in 1-D Crystals (p.49)

We will introduce the concept of an effective mass m* which incorporates

the electron interaction with the periodic potential of the crystal. Therefore,
we can correctly predict the behavior of the electron in the crystal via the
classical equations. For example,

F = m*a (2.1)

The Free Electron : potential energy (Eo) is constant

Recall the free electron case in Fig. 1.13

h2 K 2
E = Eo + (2.2) : previously
2mo (1.44)
Kinetic Energy

h2 K 2
EK = (2.3)
2mo
Figure 1.13 The E-K diagram for the
free electron.
Semiconductor Engineering 54
 h
For the free-electron case, de Broglie’s relation is valid : moν = P =
λ
Therefore, moν 2 P2
EK = =
2 2mo
dν
F = mo
dt
In addition, the velocity of the electron in the crystal is given by the group
velocity :
1 dE
v = νg = (2.4)
h dK

From Eq. (2.2), the electron mass mo can be expressed as

−1
2 d E 
2
mo = h  2  (2.5)
 dK 
Since the E-K curve is parabolic, d2E/dK2 is a constant and thus mo is
constant in energy.
Semiconductor Engineering 55
 The Quasi-Free Electron : potential energy is periodic

Electron located in this situation is named as the quasi-free electron in

your text.

Effective Mass in this case? Refer to Fig. 2.1.

Figure 2.1 The E-K
diagram for (a) an
electron at the bottom of
the conduction band at
K = 0, where the velocity
and kinetic energy are
both zero and thus the
total energy is equal to
the potential energy; (b) K=0
an electron in a local
minimum, where it has a
different effective mass,
and (c) an electron near GaAs
the top of the band.
Semiconductor Engineering 56
 Expansion of the E-K relation in a power series :
For an electron near the bottom of the band where E=Ec, the expansion at
K=0
0 Neglect these terms
 dE   1 d 2E  2
E = EC +   K +  2
K + HOTs (2.6)
 dK   2 dK 

This is the expression for a parabola shown as broken line in Fig. 2.1(a).

The group velocity is given as [We already discussed this. See p.39 of
your handout]

dx 1 dE
νg = = (2.8) same as (1.43)
dt h dK

Semiconductor Engineering 57
 Free Electron vs. Quasi-free Electron

h2 K 2
E = Eo + (2.2) : free electron
2mo
 1 d 2E  2
E = EC +  2
K (2.7) : quasi-free electron
 2 dK 

Potential Kinetic
Energy Energy

 1 d2E  2 h2 2
Since  2
K ⇔ K
 2 dK  2mo
−1
d E
2
m* = h 2  2  (2.11)
 dK 

Semiconductor Engineering 58
 Consider the cases shown in Fig. 2.1(b) and (c) : also parabola
[1] Near the minimum (E1, K1), we can expand the energy about K=K1.

1 d2E h 2 ( K − K1 ) 2
EK =
2 dK 2
( K − K1 ) 2
⇒ EK =
2m*
(2.13)
K = K1

−1 m* is the electron effective mass

d E 2  at this minimum
m =h  2
* 2 
 dK 
 K = K1 

Same form, but m* is different since the curvature at K1 is different.

[2] Near the minimum (E2, K2), the curvature of the E-K curve is negative.

“Since the development here is based on neglecting the HOTs, the

results are valid only in the parabolic regions where m* is constant.”

Semiconductor Engineering 59
 The force on an electron :
dE P dEC
F=− ⇔F=− (2.15)
dx dx
Electrons will collide with atoms, defects, or impurities (we call this phenomenon as
“scattering”, and we will discuss this later).

Figure 2.2 An external electric field is applied across a bar of semiconductor. (a) The
physical picture; (b) the energy band diagram. Electrons in the conduction band are
accelerated to the right; they travel at constant energy between collisions.

Semiconductor Engineering 60
 2.3 Conduction Band Structure (p.56)

For a semiconductor having a cubic unit cell structure and having an extremum at
K=0 (e.g., GaAs), the curvature is direction independent and
m* = mx* = my* = mz* (2.20)
or the effective mass is a scalar.

For Si and Ge, conduction band minima are not at K=0 (i.e., not a scalar).
Therefore, the effective mass depends on the direction in which the electron is
traveling. See Table 2.1 in p. 58.

Figure 2.3 The E-K

diagrams for three
common
semiconductors. The
crystallographic
direction is shown on
the K axis. The slope of
the E-K curve must be
zero at the Brillouin
zone edge, unless
multiple bands coincide
there.
Semiconductor Engineering 61
 2.4 Valence Band Structure (p.58)

The valence bands are qualitatively similar for most semiconductors: See Fig. 2.4.

If we consider the empty states to be the holes, then we can consider the holes to
have a positive charge along with a positive effective mass.

Figure 2.4 The E-K diagram for

the valence band in most
semiconductors, indicating (h)
the heavy hole band, (l) the light
hole band, and (s) the split-off
band.

Semiconductor Engineering 62
 2.5 Intrinsic Semiconductor (p.59)

A semiconductor with the relation no = po =ni (2.21)

Figure 2.5 In thermal generation, a valence electron

acquires some extra energy and moves into the
conduction band. (a) Physical picture or bond diagram;
(b) energy band diagram. In recombination (c) an
electron from the conduction band falls into a hole in
the valence band, and both the electron and the hole
disappear. It is also possible for an electron to lose
energy by colliding with something (c, right) in which
case it may remain in the conduction band. In the
collision case, the electron continues to exist.

Semiconductor Engineering 63
 2.6 Extrinsic Semiconductor (p.62)

By adding dopant atoms (i.e., donors or acceptors), the number of electrons and
holes in the conduction and valence band, respectively, can be controlled: n ≠ p
In this case, we call the semiconductor as extrinsic semiconductor.
n-type : no > po p-type : no < po

3 valence electrons

5 valence
electrons

Semiconductor Engineering 64
 Basic Diffusion Process : Add impurity using thermal
energy

Furnace
Carefully controlled quartz tube

Processing Temperature
800 – 1200oC for Si
Source 600 – 1000oC for GaAs
Gas : B2H6, AsH3, PH3
Liquid : BBr3, AsCl3, POCl3 Example
Solid : BN, As2O3, P2O5
4POCl3 + 3O2 → 2P2O5 + 6Cl2↑
2P2O5 +5Si → 4P + 5SiO2

Figure : Schematic diagram of a typical open-tube diffusion system.

Semiconductor Engineering 65
 Basic Ion Implantation Process : Add impurity using
kinetic energy

Figure: Ion
Implanter

Figure: Minimizing channeling. (a) Implantation through an amorphous oxide layer.

(b) Misorientation of the beam direction to all crystal axes. (c) Pre-damage on the crystal
surfaces.
Semiconductor Engineering 66
 2.6.1 Donors (p.62)

The 5th valence electron which did not participated in the covalent bonding is
loosely bounded, and can be easily separated and becomes the free electron.

Binding energy

Figure 2.6 Donors in a silicon crystal: (a) bond diagram of the crystal; (b) energy band
diagram of silicon doped with one phosphorus (donor) atom.

Semiconductor Engineering 67
 2.6.2 Acceptors (p.66)

An empty state will be introduced if B is forced to make covalent bond with Si, so
that nearby electron can be easily jump into the empty state, causing a hole in the
valence band.

Figure 2.8 Acceptors in a semiconductor. (a) bond diagram; (b) energy band
diagram. An electron is excited from the valence band to the acceptor state, leaving
behind a quasi-free hole.

Semiconductor Engineering 68
 2.7 The Concept of Holes (p.67)

2.7.1 Hole Charge (p.67)

Hole has to be treated as a particle (like an electron) that carry current with a
positive charge.

Consider an n-type semiconductor with an electric field applied : the charge density
in the conduction band is –qn. Therefore, the electron current density
(amperes/unit area) J is

J = − qn υ (2.32)

where <υ> is the average velocity of the electrons. Equation (2.32) can be also
written as
q
J=− ∑
volume i
υi (2.32)

where the volume is the volume of the crystal and the summation is taken over the
velocities of the individual electrons.

Semiconductor Engineering 69
 Does the Eq. (2.33) valid for electrons in the valence band? The answer is yes.
But…
Figure 2.9 States in the
valence band on the E-K
diagram. (a) All states are
full; (b) one state is empty
(hole), meaning there is an
electron with no opposing
electron at the same
energy but opposite K
vector and opposite
velocity.

1 dE
The velocity of an electron: υ = (2.34)
h dK
The velocity of an electron is proportional to the
curvature of the E-K diagram !
In Fig. 2.9(a), the E-K diagram is symmetrical (therefore, for each electron with
velocity υi there is an opposing electron with velocity - υi), and there is no empty
state. ⇒ J = 0

In Fig. 2.9(b), however, we have an empty state (i.e., hole).

Semiconductor Engineering 70
 Therefore, J for the unopposed electrons is

−q
J= ∑
volume i
υui (2.35) : The subscript u refers to the
unopposed electrons.
Equation (2.35) can be rewritten as

−q −q
J=
volume
∑ [ i hi ] volume
υ − υ = [∑ υ − ∑ υ ]
i hi (2.36)

where ∑υ ui = ∑ (υi − υhi )

Now, Equation (2.36) becomes

0
−q −q
J=
volume
[ ]
∑ υi − ∑ υhi = volume
∑ υi +
q
volume
∑ υhi (2.38)

It is reasonable to assign that the current is caused by the holes of charge +q.

Semiconductor Engineering 71
 2.7.2 Effective Mass of Holes (p.69)

Previously we showed that when an electron is near the top of the valence band,
its effective mass is negative, since the curvature of the band is negative in that
region.

Now, if we consider the vacant states to be holes with positive charge, they must
also be considered to have a positive effective mass, equal in magnitude to the
(negative) effective mass of the unopposed electron.

Figure 2.10 Lorentz force on an electron in a p-type sample. (a) The unopposed
electrons (negative charge, negative effective mass) are accelerated downward; (b)
holes, which have positive charge and positive effective mass, are accelerated upward.
In either case the net result is the same: the sample becomes more negatively charged
at the bottom with respect to the top.
Semiconductor Engineering 72
 2.8 Density-of-states Functions for Electrons in Bands (p.71)

To determine the current-voltage relations in the semiconductor devices, we need

to know the distribution of electrons and holes with energy.

Game Plan:
(1) Obtain the available energy states in the corresponding bands → S(E)
(2) Obtain the probability that a state at a given energy is occupied → f(E)

2.8.1 Density of States and Density-of-states Effective Mass (p.71)

The density-of-state function S(E) is the number of states per unit volume per unit
energy, # of available states/cm3-eV

For a free electron (EP = Eo and is constant) : The density-of-state function is

3/ 2
1  2mo 
S(E) = 2  2 
E − Eo (2.42) Refer to Appendix D
2π  h 

Semiconductor Engineering 73
 For the density-of-state function S(E) in semiconductor, derive the equation for the
electron near the bottom of the conduction band, where the effective mass is
constant.
3/ 2 3/ 2
1  2mdse
*
 1  2mdse
*

S(E) =   E − EC =   EK (2.43)
2π 2  h 2  2π 2  h 2 

where m*dse is referred to as the density-of-states effective mass for electrons. For
the description of m*dse, see p. 57. m*dse is the weighted average of mII and m⊥).

Similarly, in the parabolic region near the top of the valence band,

3/ 2
1  2mdsh
*

S(E) =   EV − E (2.44)
2π 2  h 2 

where m*dsh is some combination of the light hole and heavy hole effective masses.

Semiconductor Engineering 74
 Figure 2.11 The density of states functions for electrons in the conduction band and
the valence band. The density of states versus energy plot is superimposed on the
energy band diagram (energy versus position x).

Semiconductor Engineering 75
 2.9 Fermi-Dirac Statistics (p.73)

As we discussed previously, the total number of charged carrier (e.g., electrons)

per unit volume can be derived by

no = ∫ S ( E ) f ( E )dE
band
(2.48)

2.9.1 Fermi-Dirac Statistics for Electrons and Holes in Bands (p.73)

The probability that an electron occupies a given state at energy E in an allowed

band is given by

1
f (E ) = ( E − E f ) / kT (2.49)
1+ e

Where the energy Ef is a reference energy called the Fermi energy or Fermi level.

Semiconductor Engineering 76
 1
f (E ) = ( E − E f ) / kT
1+ e

Ef

Difficult @ T=0 Increase as T increases

Figure 2.12 The Fermi-Dirac distribution function gives the probability of occupancy of
an energy state E if the state exists.
Semiconductor Engineering 77
 The probability of a state being occupied by a hole is
1
f P (E ) = 1 − ( E − E f ) / kT
1+ e
or
1
f P (E ) = ( E f − E ) / kT
1+ e
Example 2.2 (p.76)
Estimate the probability of occupancy of a state at the bottom of the conduction
band in intrinsic Si at room temperature.

(sol.) From given information, we get E = Ec, @ room temperature, and EC-Ef ≈
Eg/2 = 0.56 eV. Now we have Eq. 2.49

1 1 1
f (E) = ( E − E f ) / kT
= ( E C − E f ) / kT
= = 4.4 × 10−19
1+ e 1+ e 1+ e ( 0.56 eV ) / ( 0.026 eV )

Therefore, under this condition, about one state in two billion available states is
occupied.

Semiconductor Engineering 78
 2.10 Electron and Hole Distributions with Energy (p.76)

As we discussed previously, the total number of charged carrier (e.g., electrons)

per unit volume can be derived by

Figure 2.13 The distribution of the electrons near the bottom of the conduction band,
n(E), is the product of the density of states distribution S(E) times the probability of
occupancy of states f(E) at a particular energy. The distribution of holes near the top of
the valence band p(E) is the product of the density of states distribution times the
probability of vacancy of states at a particular energy. (a) n type, (b) p type.
Semiconductor Engineering 79
 The total number of electrons and holes in the conduction and valence band,
respectively, at equilibrium are
top top top

no = ∫ S ( E ) f ( E )dE = ∫ n( E )dE = ∫ S ( E ) f ( E )dE

EC EC EC
(2.55)

EV EV EV

po = ∫ S(E) f
bottom
p ( E ) dE = ∫ p( E )dE = ∫ S ( E )[1 − f ( E )]dE
bottom bottom
(2.56)

Recall that most of the carriers are concentrated near the extrema of their
respective bands where effective mass is constant. Therefore, we use the Eqs.
(2.43) and (2.44) for calculating no and po.
3/ 2 3/ 2
1  2mdse
*
 1  2mdse
*

S(E) =   E − EC =   EK (2.43)
2π 2  h 2  2π 2  h 2 
3/ 2
1  2m  *
S(E ) =   dsh
EV − E (2.44)
2π 2  h  2

Semiconductor Engineering 80
 Solve the Equations (2.55) and (2.56) :

Assumptions
1. Nondegenerate Semiconductor
2. top → ∞ and bottom → -∞

Ec
3kT

Ef exists here
3kT
Ev

What happens here?

Degenerated !

Semiconductor Engineering 81
 Equations (2.55) and (2.56) become

3/ 2 ∞
1  2mdse
*

E − EC e [
− ( E − E f ) / kT ]
no = 2  2 
2π  h  ∫
EC
dE (2.57)

3 / 2 EV
1  2m  *
EV − E e [ f
− ( E − E ) / kT ]
po =  
2π 2  h 
dsh
2 ∫
−∞
dE (2.58)

3/ 2
 m kT 
*
Let N C = 2 dse
 (2.61) : Effective density of states in the
 2πh 2  conduction band
3/ 2
 mdsh
*
kT 
NV = 2 2 
(2.62) : Effective density of states in the
 2πh  valence band

no = N C e [ C f ]
− ( E − E ) / kT
Then (2.59)

po = NV e [ f V ]
− ( E − E ) / kT
(2.60)

Semiconductor Engineering 82
 Useful relationship between no and po :

For example, for intrinsic semiconductor, no = po =ni and Ef=Ei

Therefore, Equations (2.59) and (2.60) become

ni = N C e −[ ( E C − Ei ) / kT ] N C = ni e ( E C − E i ) / kT
−[ ( E i − EV ) / kT ]
⇒
ni = NV e NV = ni e ( Ei − EV ) / kT
Substitute into Equations (2.59) and (2.60), then:

( E f − E i ) / kT
no = ni e
(2.75)
( E i − E f ) / kT
po = ni e
Therefore

no po = ni2 (2.66) : for nondegenerate semiconductors in equilibrium

no po = N C NV e [ C f ]e [ f V ] = N C NV e −[ ( E C − EV ) / kT ] = N C NV e [ g ] = ni2
− ( E − E ) / kT − ( E − E ) / kT − E / kT

(2.65)
Semiconductor Engineering 83
 Example 2.3 (p.80)

Find the intrinsic concentration ni for the case of Si at 300 K. The effective
masses are m*dse =1.09mo and m*dsh = 1.15mo.

(sol.) The effective density of states NC and NV are

3/ 2 3/ 2
19  mdse 
*
 T 
N C = 2.54 × 10     (2.63)
 mo   300 
3/ 2 3/ 2
19  mdsh 
*
 T 
NV = 2.54 × 10     (2.64)
 mo   300 

Therefore, NC = 2.86x1019/cm3 and NV = 3.1x1019/cm3.

At 300 K, Eg = 1.1242 eV and kT = 0.02586 eV.

So, using the Equation (2.65)

ni = 1.08 x 1010 electrons(or holes)/cm3

Semiconductor Engineering 84
 The Fermi energy level for an intrinsic semiconductor (see pp. 80-81), Ei :

 EC + EV  3  mdsh
*

Ei =   + kT ln *  (2.72)
 2  4  mdse 

Example 2.4 (p.81)

Find the energy by which Ei is offset from the midgap for Si at room temperature.

(sol.) Using the Equation (2.72)

 EC + EV  3  mdsh
*

Ei =   + kT ln  *  EC
 2  4  mdse 
3  mdsh
*
 Ei
⇔ Ei − Emidgap = kT ln *  = 105
. meV
4  dse 
m Emidgap
EV

Semiconductor Engineering 85
 Fermi energy level as a function of the doping density : Using Equation (2.75)

n 
E f − Ei = kT ln o  (2.76)
 ni 
no ≈ N D + ni

po < ni

Figure 2.14 Electrons and holes in intrinsic and doped n-type material. In the doped case,
some of the electrons from the donors recombine with holes.

Typically,
no ≈ ND when ND >> ni

po ≈ NA when NA >> ni

Semiconductor Engineering 86
 Example 2.5 (p.83)

Find the electron and hole concentration in GaAs doped with NA = 1016 cm-3, and
locate the Fermi level. The band gap of GaAs is 1.43 eV.

(sol.) From Equations (2.63) and (2.64), we obtain

NC = 4.4 x 1017 / cm3 and NV = 8.3 x 1018 / cm3
Therefore, from Equation (2.67), ni = 2.2 x 106 / cm3

NA >> ni → po ≈ NA = 1016 cm-3

Now, from Equation (2.60)

Ef – EV = 0.17 eV

In addition, using Equation

(2.66)

no = 4.8 x 10-4 /cm3

Figure 2.15 Energy band diagram

for the p-type GaAs of Example 2.5.

Semiconductor Engineering 87
 Property change due to semiconductor doping :

Doping is done routinely to form the junction; for example, pn junction

Assuming all donors and acceptors are ionized,

If ND > NA, then no = ND – NA and the material is n type. (2.79)

If NA > ND, then po = NA – ND and the material is p type. (2.80)

Both of these equations apply only when ND - NA or NA – ND >> ni.

Semiconductor Engineering 88
 Example 2.5 (p.83)

A sample of Si is doped everywhere with a background concentration of NA

= 4 x 1016 / cm3. Then 1 x 1017 / cm3 donors are doped. Find the room
temperature concentrations of electrons and holes in the original and final
materials.

(sol.) Based on the given problem, we know that the starting material is p-type. In
addition, NA >> ni (1.08 x 1010/cm3).

Therefore, po = NA = 4 x 1016 / cm3

Ef – EV = 0.17 eV [from Eg. (2.60)]
no = ni2 / po = 2.9 x 103 / cm3

For the n-type material,

no = ND – NA = 1017 – 4 x 1016 = 6 x 1016 / cm3
EC – Ef = 0.16 eV [from Eg. (2.59)]
po = ni2 / no = 1.9 x 103 / cm3

Semiconductor Engineering 89
 Figure 2.16 The energy band
diagrams for Example 2.6: (a)
uncompensated p-type material;
(b) compensating donors are
added to make the material net
n type.

Semiconductor Engineering 90
 Electron concentration as a function of energy :

If we let Ek = E - Ec, then (refer to the derivation process shown in pp. 86-87)

n( E K ) = C E K e − E K / kT (2.84)

“With increasing Ek, the electron

concentration n(Ek) increases, reaches a
maximum, and then decreases
approximately exponentially.”

Note that most of the conduction band

electrons are existed at very near the 2.5 kT = 0.065 eV
conduction band edge.

Figure 2.17 The electron distribution

function n(E) as a function of energy
(energy on the vertical axis).

Semiconductor Engineering 91
 2.11 Temperature Dependence of Carrier Concentrations in
Nondegenerate Semiconductors (p.89)

2.11.1 Carrier Concentrations at High Temperatures

Using Equation (2.67),

− E g / 2 kT
ni = N C NV e (2.85)

We have to consider the effect of temperature on ni (therefore, Nc, Nv, and Eg as

well).

We already discussed
3/ 2 3/ 2
19  mdse 
*
 T 
N C = 2.54 × 10     (2.63)
 o 
m  300 
3/ 2 3/ 2
19  mdsh 
*
 T 
NV = 2.54 × 10     (2.64)
 mo   300 

Semiconductor Engineering 92
 In addition, the temperature dependence of Eg for Si is

4.73 × 10−4 T 2
E g ( T ) = 117
. − eV (2.86)
T + 636

Figure 2.18 Energy band-

gap dependence of
silicon on temperature.

Semiconductor Engineering 93
 Now let us look at what happens in no in the conduction band at high temperature.

Assumptions :
(1) Uniformly doped and nondegenerated semiconductor
(2) Equilibrium
(3) 100% ionization
(4) Space charge neutrality is satisfied

Then po − no + N D+ − N A− = 0 ⇒ po − no + N D − N A = 0 (2.90)

For the case of nondegenerated semiconductor,

no po = ni2 (2.66)

Therefore, Equation (2.90) becomes

ni2
− no + N D − N A = 0 ⇒ no2 − no ( N D − N A ) − ni2 = 0 (2.92)
no
The solution is then
1/ 2
N D − N A  N D − N A  
2

no = +   + n 2
i  (2.93)
2  2  
Semiconductor Engineering 94
 Figure 2.19 Plot of
electron concentration
n0 as a function of
temperature in n-type
silicon for four values
of net doping. Also
indicated is the
temperature
dependence of ni.
At high temperature,
Si becomes intrinsic !

Similarly,
1/ 2
N A − N D  N A − N D 
2
 ni2
po = +   + ni 
2
(2.95) no =
2  2   po

Semiconductor Engineering 95
 Intrinsic T-region

Extrinsic T-region
Figure 2.20
Normalized plot
of n0/ND as a
function of
temperature.
This plot is for
ND = 1016 cm-3.

Semiconductor Engineering 96
 2.12 Degenerate Semiconductors (p.94)
2.12.1 Impurity-Induced Band-Gap Narrowing

Degenerated semiconductor is the semiconductor doped with very high value.

Very closely spaced !

Very remotely spaced ! Form miniband, causing band-gap narrowing

Figure 2.21 Under high doping concentrations, the formerly discrete donor levels
smear into a band, effectively narrowing the band gap by an amount ∆Eg.

Semiconductor Engineering 97
 Figure 2.22 The states for the
higher donor levels can overlap if
the doping concentration is high
enough (dopant atoms close
enough together).

As the donor concentration increases, the shift in the conduction band also
increases : ∆EC = ECo – EC

“The increase in ∆EC with increasing ND is largely due to an increase in the

number of spatially overlapping equi-energy impurity states and also from the
broadening in energy of these impurity bands. At sufficiently high doping level,
even the ground states overlap.”

Semiconductor Engineering 98
 ∆E g = E go − E g ( N D ) : impurity-induced band-gap reduction

Empirical equation for the band-gap narrowing for n-type Si is

[( ]
−1 / 4
∆E g = 4.372 × 10 − 11
N D )
0.5 − 4
(
+ 1272
. −6
× 10 N D )
0.25 − 4
... eV

Negligible

Figure 2.23 Reduction of

room-temperature band gap
∆Eg as a function of donor
density in phosphorus-doped
silicon.

Semiconductor Engineering 99
 We will skip 2.12.2 and 2.12.3.

See Fig. P2.4 in p.106

Semiconductor Engineering 100

 Review Chap 2. By answering the question 2.16 (Review
Questions)

Homework #2

[1] Problem 2.6

[2] Problem 2.26
[3] Problem 2.37

Semiconductor Engineering 101