• If they replace column V elements, then act as

acceptors
• Known as amphoteric dopants
• Carrier Concentrations:
– Needed for calculating current transport
– For this, we need:
• Distribution Function (DF) of particles as a
function of energy
• The Density of Available States [known as the
Density of States (DoS)] as a function of energy
• The carrier concentration is a product of these two,
integrated over all allowed energies
Aloke Dutta/EE/IIT Kanpur 1
• Distribution Functions (DF):
– Maxwell-Boltzmann (MB)
– Fermi-Dirac (FD)
– Bose-Einstein (BE)
• Maxwell-Boltzmann Distribution
Function (MBDF):
– For classical, distinguishable particles, without
spin (e.g., gas particles, dilute carrier
concentrations, etc.)

Aloke Dutta/EE/IIT Kanpur 2

– f(E)  exp[E/(kT)]
– n2/n1 = exp[E/(kT)]
E = E2 E1
n1,n2: Particles in energy levels E1,E2
E/(kT): Boltzmann factor
– Particle concentration drops off exponentially
with an increase in energy
– As T, the distribution tends to become
uniform

Aloke Dutta/EE/IIT Kanpur 3

Aloke Dutta/EE/IIT Kanpur 4
• Fermi-Dirac Distribution Function
(FDDF):
– Two basic differences with MBDF:
• Particles have spin and obey Pauli Exclusion
Principle
• Particles are indistinguishable (very heavy
concentration)
1
 f E 
 E  EF 
1  exp  
 kT 
f(E): Probability that a state at an energy E is
occupied by an electron
Aloke Dutta/EE/IIT Kanpur 5
f(E)
1
EF: Fermi Energy
(Average energy of
T1
electrons in the system)
1/2
T2 > T 1 > 0 K
T=0K

0
EF E
Aloke Dutta/EE/IIT Kanpur 6
– At T = 0, f(E) becomes a step function:
• For E < EF, f(E) = 1
• For E > EF, f(E) = 0
– f(E) gives the probability that a QS at energy
E will be occupied
– f(E) = 1/2 for E = EF at all temperatures
– As T, the edges of the step function gets
rounded, and the function spreads out
– Amount of this spread ~ (3-4)kT
– For very high T, the distribution tends to
diffuse and spread out
Aloke Dutta/EE/IIT Kanpur 7
• The Boltzmann Approximation:
– If (EEF) > (3-4)kT:
 EF   E   E 
f  E   exp   exp     exp   
 kT   kT   kT 
Exactly similar to MBDF
 High energy tail of FDDF is Maxwellian
in nature
– [(EEF) > (3-4)kT] is known as the Boltzmann
approximation  greatly simplifies maths
– Separates semiconductors into 2 categories:
degenerate and non-degenerate
Aloke Dutta/EE/IIT Kanpur 8
• Bose-Einstein Distribution Function
(BEDF):
– Applicable for particles:
• With integral spin
→ spin quantum # = 0,1,2,3,…
• Not subject to Pauli Exclusion
• Indistinguishable
– Example: Photons (Generic Name: Bosons)
– DF: 1
f  E   E  kT 
e 1

Aloke Dutta/EE/IIT Kanpur 9

• Conditions for Equilibrium:
– No external excitation except a constant
temperature
– Across any interface, no net:
• Transport of energy
• Carrier motion
• Charge transport
– There can’t be any gradient in the equilibrium
Fermi level EF

Aloke Dutta/EE/IIT Kanpur 10

– Proof:
• Let two layers (1 and 2) in contact with each other
• Let for a given specific energy E:
– The DoS in layers 1 and 2 be g1(E) and g2(E) respectively
– The DF in layers 1 and 2 be f1(E) and f2(E) respectively
• Carrier flow from layer 1 to layer 2:
 (# of filled states in layer 1)  (# of empty states
in layer 2)
 g1(E)f1(E)  g2(E)[1f2(E)] (1)
• Similarly, carrier flow from layer 2 to layer 1:
 g2(E)f2(E)  g1(E)[1f1(E)] (2)

Aloke Dutta/EE/IIT Kanpur 11

 • Based on the definition of equilibrium, (1) must equal
(2)
 f1(E) = f2(E)  EF1 = EF2  dEF/dx = 0
 No gradient in EF
An extremely important condition for equilibrium
– Physical Interpretation:
• EF is a measure of average energy of electrons in the
system
• Constant EF implies that electron energy is constant
across the system
 There can’t be any net motion
• When the system is out of equilibrium, then the
equilibrium Fermi level loses its physical significance

Aloke Dutta/EE/IIT Kanpur 12

• Movement of EF Within Eg:
– Intrinsic:
• n = p = ni
• # of electrons in CB = # of holes in VB
• Average energy of electrons in the system will
roughly be at the middle of the band gap
• Known as the intrinsic level Ei
• Ei is not exactly at the middle of the band gap,
however, the difference from the center is so small,
that it can be neglected

Aloke Dutta/EE/IIT Kanpur 13

– n-type:
• n >> p
• # of electrons in CB >> # of holes in VB
• Average electron energy of the system 
• EF moves towards EC
– p-type:
• p >> n
• # of holes in VB >> # of electrons in CB
• Average electron energy of the system 
• EF moves towards EV

Aloke Dutta/EE/IIT Kanpur 14

 EC EC EC
EF

Ei = E F Ei Ei

EF
EV EV EV

Intrinsic n-type p-type

Aloke Dutta/EE/IIT Kanpur 15

• Carrier Concentrations:
– Depend not only on f(E), but also on the # of
states available at E
– If there are no available states, then there can’t
be any carriers, irrespective of the value of f(E)
– For example, within the band gap, there is no
available state
 No carrier can stay within the band gap
unless a state is created there somehow,
e.g., dopant/trap levels, recombination
centers, etc.
Aloke Dutta/EE/IIT Kanpur 1
 30-60 meV ( Shallow Donor States)
EC
Traps
Eg
Recombination
Ei
Centers
Traps
EV
30-60 meV ( Shallow Acceptor States)

Aloke Dutta/EE/IIT Kanpur 2

• Electron and Hole Concentrations in
Equilibrium:
– Equilibrium electron concentration in CB:

n0   f (E)N(E)dE
EC

– N(E)dE: Density of available states (DoS) per

unit volume in the CB within an
energy range dE
– Note the upper limit (why  ?)

Aloke Dutta/EE/IIT Kanpur 3

– N(E) is calculated using SE under PBC:
3/2
1  2m  *
N(E)  2  2  E1/2
2   
Applicable for both CB and VB by replacing:
m* by m*n and m*p , and E by  E  E C  and
 E V  E  , respectively
– Note:
• At EC and EV , DoS = 0
• Square root dependence on E
Aloke Dutta/EE/IIT Kanpur 4
E
Conduction
Band
1/2
 (E – EC)
EC
0

0
EV
1/2
Valence  (EV – E)
Band

N(E)

Aloke Dutta/EE/IIT Kanpur 5

– Within the band gap, DoS = 0
– In CB, f(E) as E goes above EC
 f(E)N(E)
 Most of the electrons occupy states near
the bottom of CB (~3kT)
– In VB, [1f(E)] as E goes below EV
 Most of the holes occupy states near the
top of the VB (~3kT)
– Can assume that all available states in CB are
located at EC and use FDDF at EC under the
Boltzmann approximation
Aloke Dutta/EE/IIT Kanpur 6
– Thus, electron concentration in CB:
 EC  EF 
n 0  N C f (E C )  N C exp    (show)
 kT 
NC: Effective DoS at EC (note that actual DoS
at EC is zero)
3/2
 2m kT 
*
 2 n
2 
 h 

– Note: N C   m 
32
*
n and  T 3 2
– Care: It is valid only if (ECEF) > 3kT

Aloke Dutta/EE/IIT Kanpur 7

– Similarly, hole concentration in VB:
 EF  EV 
p 0  N V [1  f (E V )]  N V exp   
 kT 

NV: Effective DoS at EV (actual DoS at EV is

zero)
3/2
 2m kT 
*

 2 
p
 h 2
 3 2
*
 
– Note: N V  m p and  T 3 2
– Care: It is valid only if (EFEV) > 3kT

Aloke Dutta/EE/IIT Kanpur 8

– In equilibrium:

 Eg 
n 0 p0  n  N C N V exp  
2
i 
 kT 
 Eg 
 n i  N C N V exp   
 2kT 

– For intrinsic material, EF lies at Ei (intrinsic

level)
– Care: Ei is NOT exactly at the center of the
band gap, since m*n  m*p

Aloke Dutta/EE/IIT Kanpur 9

– Fundamental expressions for carrier
concentrations:
 EC  Ei   Ei  E v 
n i  N C exp    and pi  N v exp   
 kT   kT 
and
 E F  Ei   Ei  E F 
n 0  n i exp   and p0  n i exp  
 kT   kT 

– These expressions are immensely important

– They express n0 and p0 in terms of EC, EV, Ei,
EF, and ni
Aloke Dutta/EE/IIT Kanpur 10
• Temperature Dependence of Carrier
Concentration:
– Explicit temperature dependence through the
term kT
– Implicit dependence through ni
 2kT   Eg 
3/2

n i (T)  2  2  *
(m m )
n
* 3/4
p exp   
 h   2kT 
 Eg 
= AT 32
exp   
 2kT 

Aloke Dutta/EE/IIT Kanpur 11

• Temperature Dependence of Carrier
Concentration:
– Explicit temperature dependence through the
term kT
– Implicit dependence through ni
 2kT   Eg 
3/2

n i (T)  2  2  *
(m m )
n
* 3/4
p exp   
 h   2kT 
 Eg 
= AT 32
exp   
 2kT 

Aloke Dutta/EE/IIT Kanpur 1

– A: Temperature independent constant
– Temperature dependent terms: T3/2 and
exp[Eg/(2kT)]
– Obviously, the latter one would dominate
– Also Eg has a very weak temperature
dependence (Varshini Model):
T 2
E g  T   E g0 
T
E g0  E g at T = 0 

,: Constants (material dependent)

– For rough calculations, this dependence of Eg
on temperature is generally ignored
Aloke Dutta/EE/IIT Kanpur 2
Material
Eg0  
(eV) (eV/K) (K)
–4
Ge 0.74 4.7710 235
–4
Si 1.17 4.7310 636
–4
GaAs 1.52 5.4110 204

Aloke Dutta/EE/IIT Kanpur 3

Aloke Dutta/EE/IIT Kanpur 4
• Concentration as a Function of Temperature:

N D = 1015 cm 3

Aloke Dutta/EE/IIT Kanpur 5

– Consider an n-type material with donor doping
ND cm3
– Free electron concentration = n0
– At very low T, all dopants are not ionized
 n0 small
– As T , n0, since more dopants start to get
ionized
– Known as the Ionization Zone
– After all dopants get ionized, n0 = ND, and stays
constant w.r.t. T
– Known as the Extrinsic Zone
Aloke Dutta/EE/IIT Kanpur 6
– In the background, ni keeps on growing
exponentially with T
– A point comes when n0 approaches ni
– Beyond this point, n0 follows ni, and the
semiconductor is no more extrinsic
 End of the Extrinsic Zone and start of the
Intrinsic Zone
– Definition of Extrinsic Zone:
The range of temperature over which the
majority carrier concentration remains equal to
the doping concentration and is sufficiently
greater than ni (about 2-5 times)
Aloke Dutta/EE/IIT Kanpur 7
– The extrinsic range is a function of the doping
concentration
• As the doping concentration, the extrinsic range
also
– For a given material, the dopant concentration
should be chosen such that it gives a
sufficiently large extrinsic range
– Obvious that Ge has the least extrinsic range
(why?), followed by Si, and GaAs has the
largest extrinsic range, for any value of doping

Aloke Dutta/EE/IIT Kanpur 8

• Space Charge Neutrality:
– Intrinsic material is charge neutral
– Extrinsic material also is charge neutral
– Total four types of charges present in a
semiconductor:
• Electrons n0 (negative)
• Holes p0 (positive)
 Ionized Donors N D (positive)
 Ionized Acceptors N A (negative)
– Total positive charges must equal total
negative charges

Aloke Dutta/EE/IIT Kanpur 9

– Leads to the expression for the space charge
neutrality:
 
p0  N D  n 0  N A
– This, along with the law of mass action
uniquely determines n0 and p0 for given values
of ND and NA
– In general, 100% ionization approximation is
valid for shallow dopants:
 p0 + ND = n0 + NA
– Interesting scenario:
If ND = NA  n0 = p0 = ni

Aloke Dutta/EE/IIT Kanpur 10

– Sample has turned intrinsic by doping of both
types of equal magnitude:
Known as Full Compensation
Can also be partially compensated
– Note:
• If NA = 0 and ND >> ni, n0  ND and 0
p  n 2
i ND

• If ND = 0 and NA >> ni, p0  NA and n 0  n i N A

2

• If (NDNA) >> ni, n0  (NDNA) and p 0  n i2  N D  N A 

• If (NAND) >> ni, p0  (NAND) and n 0  n i2  N A  N D 

Aloke Dutta/EE/IIT Kanpur 11

• Resistivity and Conductivity:
– General Expression for Extrinsic Conductivity:
 = 1/ = q(nn + pp)
n,p: Electron,hole mobility
– Mobility:
• Measure of ease of carrier motion under an applied
electric field
• Given by:  = vd/E
vd: drift velocity
E : electric field
• Defined as velocity per unit electric field

Aloke Dutta/EE/IIT Kanpur 12

– For intrinsic material, intrinsic conductivity:
i = 1/i = qni(n+p)
– Care: i is not the same as max
– Exercise: Find the condition to obtain max
– In intrinsic Si:
n ~ 1300-1400 cm2/V-s
p ~ 400-500 cm2/V-s
– In intrinsic GaAs:
n ~ 8500-9000 cm2/V-s
p ~ 400-500 cm2/V-s
Aloke Dutta/EE/IIT Kanpur 13
– With doping, these values drop
– Doping classifications:
• 1012-1014 cm3: Low
• 1015-1017 cm3: Moderate
• 1018-1019 cm3: High
• > 1019 cm3: Very high (degenerate)
– Under degenerate doping, n and p can drop to
as low as 500 and 200 cm2/V-s respectively
– Conductivity can be increased greatly by:
• Light (photo-conductivity)
• Temperature
• Doping
• High energy carrier bombardment

Aloke Dutta/EE/IIT Kanpur 14