ADVANCED MOSFET MODELING

Tutorial No:1

Q.1 Show that the effective density of states (N C) represents the density of
states in a strip 1.2 kT wide near the edge of the conduction band.
3 EC + ∆ E ( EC + ∆ E )
 2π kTme*  2
4π
Ans : NC = = ∫ = ∫ (2 me* ) 2 ( E − Ec ) 2 dE
3 1
2 ÷ g ( E ) dE
 h
2
 EC EC
h3

( Ec + ∆ E )
( EC + ∆ E )
4π  ( E − Ec ) 
3
4π 2

= ∫ =
3
( E − Ec )
3 1
(2 me* ) 2 2
dE (2me* ) 2  
h3 EC h3 
3
2  Ec

4π 2
= × (2 me* ) 2 × ( ∆E ) 2
3
×
3

h3 3

3
2  ∆E  2 3
Thus, = × (∆E) 2 Or,
3
2 (π kT ) 4π ×  ÷ =
3
2

3  π kT  4π

2
 3  3
Therefore, ∆E =  ÷ π kT
 4π 
=1.2kT
[Message : Only states which are very close to E C in the conduction band will be
occupied. Similarly holes will only occupy states very close to the top of valence
band]

Q:2 In a semiconductor sample, donor and acceptor levels are 0.3eV apart
from each other. If 80% of the acceptors are ionized at 300K, evaluate the
fraction of ionized donors. If the donor level is 2kT below the conduction band
edge, determine the position of Fermi level.
Ans:

1
ηA =  EA − EF 
1 + 2 exp  ÷
= 0.8
 kT 

 1 
 ÷−1
Or,  E A − EF   0.8 
exp  ÷=
= 0.125
 kT  2

(EA – EF ) = 0.026 ln (0.125) = - 0.054 eV

∴ (ED – EF) = (ED – EA) + (EA – EF) = 0.3 – 0.054 = 0.246 eV
1 1
× 100%
Therefore, ηD = ( E − ED )
1 + 2 exp F =  − 0.246 
1 + 2 exp  
= 99.99%
kT  0.026 

Position of Fermi level :

ED is 2kT = 0.052 eV below EC
∴ EC – EF = (EC – ED) + (ED – EF) = 0.052 + 0.246 = 0.298 eV

Q: 3 Silicon wafers are doped with (i) 1015 (ii) 1018 arsenic atoms / cm3. Show
whether the assumption of complete ionization is justified in each case at
temperatures of 100K and 300K. Arsenic introduces a donor level E D
0.049 eV below EC.

Ans: Eion = Ec – Ed = 0.049eV

 ND1 = 1015 atoms / cm3
ND2 = 1018 atoms / cm3

(a) At T = 100 K
Since NC α T3/2, at 100 K, NC = 2.8 x 10
100 3 2
19
×(
300
)
= 5.3886 x 1018 / cm3
Also, kT = 0.026/3 = 0.008667eV
For ND1 = 1015 atoms / cm3
2 2
= x 100%
ηD1= 1+ 1+
8 N D1  E 
exp  ion ÷ 1+ 1+
8 x 1015  0.049 
exp  ÷
= 91.19 %
NC  kT  5.3886 x 1018  0.008667 

For ND2 = 1018 atoms / cm3

2
= x 100%
ηD2 1+ 1+
8 x 1018  0.049 
exp  ÷
= 9.26%
5.3886 x 1018  0.008667 

(b) At T = 300K, For ND1 = 1015 atoms / cm3

2 2
= x 100%
ηD1 = 1+ 1+
8 N D1  E 
exp  ion ÷ 1+ 1+
8 x 10 15
 0.049 
exp  ÷
= 99.95%
NC  kT  2.8 x 1019  0.026 

For ND2 = 1018 atoms / cm3

2
= x 100%
ηD2 1+ 1+
8 x 10 18
 0.049 
exp  ÷
= 74.15%
2.8 x 1019  0.026 

[ Message : For lightly doped (~10 15 atoms / cm3) silicon, ionization is almost
complete at room temperature. However, for heavy doping (~1018 atoms / cm3), there
is only partial ionization at room temperature]

Q: 4 A silicon wafer is doped with 10 16 atoms /cm3 of Indium, which introduces

an accepter level EA 0.16eV above EV. Determine the temperature at which 50%
of Indium is ionized. If this wafer was doped with Boron instead of Indium, what
would have been the percentage of ionization at this temperature? For Boron, E A
– EV = 0.045 eV
2
(a) ηA = 1+ 1+
8N A  E 
exp  ion ÷
NV  kT 

For Indium, Eion = 0.16 eV

Taking NA = 1016 atoms / cm3 and NV = 1 x 1019 / cm3 (value at 300K) and ηA = 0.5,
we have

2
0.5 = 8 × 1016  0.16 
1+ 1+
8 × 1016  0.16 
exp  ÷
Or, 1+
1 × 1019
exp 
 kT 
÷ =3
1 × 1019  kT 

 0.16  ( 0.16 ) = 0.02316

Or, exp 
 kT 
÷ = 10
3
Or, kT =
ln (103 )
 0.02316
Therefore, T = 8.667 × 10 − 5
= 267 K

(b) For Boron, Eion = 0.045 eV, T = 267 K, NA = 1016 /cm3, NV ≈ 1 × 1019 / cm3

2
× 100%
ηA = 1+ 1+
8 × 1016 
exp 
0.045 
÷
= 98.64%
−5
1 × 1019  8.667 × 10 × 267 

[ Message : Ionization is much less with increase in ionization energy]

NC N D  E 
Q.5 Assuming n = ND+, show that at very low temperature n=
2
exp  − ion ÷
 2 kT 
.
Hence show that is this temperature range, the Fermi level is an n-type
EC + E D kT  2NC 
semiconductor is given by EF =
2
−
2
ln 
 ND 
÷ .

Also find EF at 0K
 E − EF 
(a) n = N C exp  − C
 kT 
 (1)

1 1
ηD = ≈
 EF − ED 
1 + 2 exp 
 EF − E D 
÷ 2 exp  ÷
at very low temperatures
 kT   kT 

ND
∴ n = η D ND =
 E F − ED 
2 exp  ÷
(2)
 kT 

Multiplying (1) & (2)

NC N D  −( EC − E F ) − ( E F − E D )  NC N D  ( EC − ED ) 
n2 = exp   = exp − 
2  kT  2  kT 

NC N D  E 
Or, n=
2
exp  − ion ÷
 2 kT 

(b) Equating eqns. (1) & (2)

 E − EF  ND
N C exp  − C  =
 kT   EF − E D 
2 exp  ÷
 kT 

2 NC  − ( E F − ED ) −( EC − E F )  2 NC  E + ED − 2 E F 
Or, ND
= exp 
 kT
−
kT 
 Or, ND
= exp  C
 kT 


 E + ED  kT  2 NC 
∴ EF =  C ÷− ln  ÷
 2  2  ND 

(c)At T = 0K, second term becomes zero.

EC + ED
EF = 2 i.e., Fermi level lies in this middle of EC & ED.
Q : 6 At moderately high temperature, assuming ND+ = ND, show that
  2ni  
2
ND 
n=
2 
1+ 1+  ÷  Intrinsic temperature is defined as the temperature at
  ND  

which the instrinsic concentration ‘n i’ equals the doping concentration ‘N’.

Ti Eg
=
Show that the ratio Ti / To is given by To  NC NV  . Determine Ti for Si
2kT ln  ÷÷
 N
 

and Ge, assuming NC and NV to be independent of temperature because of the

logarithmic dependence [NC = 2.8 x 1019 / cm3, NV = 1 x 1019 / cm3 for Si, NC = 1 x
1019 / cm3 for Si, NC = 1 x 1019 / cm3, NV = 6 x 1018 / cm3 for Ge]
(N − N A− ) + (N − N A− ) + 4ni2
+ + 2

Ans (a) n=
D D

Taking N D+ = N D and N A− = 0

ND + + 4 ni2 ND  4ni2 
2 2
ND ND ND
n= = + + ni2 = 1 + 1+ 2 
2 2 4 2 
 ND 


  2ni  
2
ND 
Therefore n=
2 
1+ 1+  ÷ 
  ND  

(b) At T = Ti, N = ni
 − Eg  Eg
Or, N = ni = Nc Nv exp  ÷ Or, 2kTi
= ln ( Nc N v N )
 2kT 

Ti Eg
=
Therefore, To 2kTo ln Nc Nv N ( )
Eg
Ti =
(c) 2k ln ( N C NV N )
For Si, Eg = 1.1eV, Nc = 2.8 x 1019 / cm3, Nv = 1 x 1019 / cm3 and N = 1 x 1015 / cm3
1.1
Ti =
 2.8 × 1019 × 1019  = 653 K
2 × 8.667 × 10− 5 ln  ÷
 1015 ÷
 

For Ge, Eg = 0.67eV, Nc = 1 x 1019 / cm3, Nv = 6 x 1018 / cm3 and N = 1 x 1015 / cm3
0.67
Ti =
 1× 1019 × 6 × 1018  = 357 K
2 × 8.667 × 10 − 5 ln  ÷
 1015 ÷
 

[Message : The intrinsic temperature is higher for a wider bandgap semiconductor]

Q.7 Draw the energy band diagram of silicon doped with 10 15 arsenic
atoms/cm3 at 77K, 300K and 600K showing the Fermi level?

0.026
(a) At 77K, kT = 300
× 77 = 6.67 × 10 −3 eV

77 3 2
N C = 2.8 × 1019 × (
300
) = 3.64×1018/cm3
 2 2
ηD = =
1+ 1+
8N D  E 
exp  ion ÷ 1+ 1+
8 × 1015  0.049 
exp  ÷
= 0.3226
NC  kT  3.64 × 1018  0.00667 

 E − EF 
n= ηD × N D = 3.226×1014/cm3 and n = N C exp  − C
 kT 


 NC   3.64 × 1018 
Therefore EC − E F = kT ln 
 n 
÷ = 0.00667 ln  14 ÷
 3.226 × 10 
= 0.0622eV

Or, EF – Ei = (EC – Ei) – (EC – EF) = 0.55 – 0.0622 = 0.4878 eV

(b) At 300K, kT = 0.026 eV, Nc = 2.8 x 1019 / cm3, ηD ≈ 1, or n ≈ ND = 1015/cm3
 NC   2.8 × 1019 
Therefore EC − EF = kT ln 
 n 
÷ = 0.026 ln 
 1015
÷ = 0.2662eV


Or, EF – Ei = (EC – Ei) – (EC – EF) = 0.55 – 0.2662 = 0.2838 eV

0.026
(c) At 600K, kT = 300
× 600 = 0.052 eV
600 3 2 600 3 2
N C = 2.8 × 1019 × (
300
)
= 7.92×1019/cm3, NV = 1019 × (
300
) = 2.828 ×1019 / cm3

 − Eg   −1.1 
ni = Nc Nv exp  ÷= 7.92 × 2.828 × 1019 exp  ÷ = 1.21× 10 / cm
15 3

 2kT   0.104 

  2ni   1015   2 × 1.21 × 1015  

2 2
ND  1 +
n= 1+ 1+  ÷ = 1+  ÷  = 1.81× 10 / cm
15 3

2   N D   2   10 15
 
 

 n   1.81 
Therefore EF − Ei = kT ln  ÷ = 0.052 ln 
 ni 
÷ = 0.021eV
 1.21 

[ Message : At very low temperature the Fermi level is close to the band edge. With
increase in temperature, the Fermi level moves towards the middle of the bandgap]

Q.8 A silicon wafer is doped with 2 x 1016 boron and 1016 phosphorus atoms/
cm3. Calculate n, p, EF at room temperature, assuming complete
ionization. Repeat the same for 8 x 1015 boron atoms / cm3?

(a) ND+ = 1016 / cm3 and NA- = 2 x 1016 / cm3

− + − + 2
(N A − N D ) + ( N A − N D ) 2 + 4 ni
p= ≈ (NA – ND) = 1 x 1016 / cm3
2

ni 2
(1.5 × 1010 ) 2
n= = (1 × 1016 ) = 2.25 x 104 / cm3
p
 p   1016 
EF = Ei − kT ln  ÷ = Ei − 0.026 ln  10 ÷
= Ei − 0.349eV
 ni   1.5 × 10 

(b) NA- = 8 x 1015 / cm3 and ND+ = 1016 / cm3

 + − + − 2
( ND − NA )+ ( N D − N A ) 2 + 4n i
n= ≈ (ND – NA) = 2 x 1015 / cm3
2

(1.5 × 1010 )
2
ni2
p= = = 1.125 × 105 / cm 3
n 2 × 1015

 n   2 × 1015 
EF = Ei + kT ln  ÷ = Ei + 0.026 ln  10 ÷
= Ei + 0.307eV
n
 i   1.5 × 10 