Anda di halaman 1dari 22

SEMICONDUCTORS

 Semiconductors
 Semiconductor devices

Electronic Properties
Robert M Rose, Lawrence A Shepart, John Wulff
Wiley Eastern Limited, New Delhi (1987)
Energy gap in solids
 In the free electron theory a constant potential was assumed inside the solid.
 In reality the presence of the positive ion cores gives rise to a varying potential field. In a
simple model the potential as in Fig.1 can be assumed (‘a’ is the lattice spacing and ‘w’ is
the width of the potential). If ‘w’  0, we get ‘’ functions.
 The travelling electron wave interacts with this periodic potential (for a crystalline solid).
 The electron wave can be Bragg diffracted.

Bragg diffraction from a 1D solid n = 2d Sin 1D   = 90o n = 2a n2p/k = 2a

np � 2a � �p 2p 3p �
k= λCritical ��
2a, a, , ...� kCritical �
� , , , ...�
a � 3 � �a a a �
k →

E →
2 2
h k
E=
8p 2 m

Band gap

2p p p 2p
- - + +
a a a a
 The Velocity of electrons for the above values of k are zero.
 These values of k and the corresponding E are forbidden in the solid.
 The waveform of the electron wave is two standing waves.
 The standing waves have a periodic variation in amplitude and hence the electron
probability density in the crystal.
 The potential energy of the electron becomes a function of its position
(cannot be assumed to be constant (and zero) as was done in the free electron model).
 The magnitude of the Energy gap between two bands is the difference in the potential
energy of two electron locations.
 The effective mass of an electron (m*) in a solid could be different from the rest mass (m 0).
The effective mass can be larger or smaller than the rest mass.
 Collisions with atoms increases the rest mass, while enhanced propagation in the crystal
gives a reduced effective mass.
 From the concept of group velocity the effective mass can be deduced. It is related to the
curvature of the E-k curve. Close to band edges, m* can be negative.

K.E of the electron increasing h2 1


Decreasing velocity of the electron m = 2 2
*

-ve effective mass (m*) of the electron 4p ( d E / dk 2 )


E →

p
+
k → a
Within a band
If Fermi energy lies in this region
then the conductivity would be high
Effective energy gap → Forbidden gap → Band gap
 The effective gap for all directions of motion is called the forbidden gap.
 There is no forbidden gap if the maximum of a band for one direction of motion is higher
than the minimum for the higher band for another direction of motion  this happens if the
potential energy of the electron is not a strong function of the position in the crystal.
np
k=
d Sin
p p
k= k=
d Sin 90o d Sin 45o

[100] [110]

Effective gap
E →
E →

p k → 2p k →
+ +
a a
Energy band diagram: METALS

Divalent metals

Monovalent metals

 Monovalent metals: Ag, Cu, Au → 1 e- in the outermost orbital


 outermost energy band is only half filled
 Divalent metals: Mg, Be → overlapping conduction and valence bands
 they conduct even if the valence band is full
 Trivalent metals: Al → similar to monovalent metals!!!
 outermost energy band is only half filled !!!
Energy band diagram: SEMICONDUCTORS

Semiconductor

Conduction Band
2-3 eV
Valence Band

 Elements of the 4th column (C, Si, Ge, Sn, Pb) → valence band full but no overlap of
valence and conduction bands
 Diamond → PE as strong function of the position in the crystal
 Band gap is 5.4 eV
 Down the 4th column the outermost orbital is farther away from the nucleus and less bound
 the electron is less strong a function of the position in the crystal  reducing band gap
down the column
Energy band diagram: INSULATORS

Insulator

Conduction Band

> 3 eV

Valence Band
Intrinsic semiconductors
 At zero K very high field strengths (~ 1010 V/m) are required to move an electron from the
top of the valence band to the bottom of the conduction band
  Thermal excitation is an easier route

P(E) →

E →
EF Eg
Eg/2

0 0.5 1
T>0K 1
P( E ) =
 E - EF  Eg
1 + exp 
 kT 
 ( E - EF ) =
2

( E - E F ) Silicon
 Eg  E - EF
=   = 0.55 eV kT  0.026 eV  1
 2  Silicon kT
 Unity in denominator can be ignored

 Eg 
P( E )  exp - 
 2 kT 

 Eg   ne → Number of electrons promoted across the gap


ne = N exp -  (= no. of holes in the valence band)
 2 kT   N → Number of electrons available at the top of the
valance band for excitation
Conduction in an intrinsic semiconductor
 Under applied field the electrons (thermally excited into the conduction band) can move
using the vacant sites in the conduction band
 Holes move in the opposite direction in the valence band
 The conductivity of a semiconductor depends on the concentration of these charge carriers
(ne & nh)
 Similar to drift velocity of electrons under an applied field in metals in semiconductors the
concept of mobility is used to calculate conductivity
Mobility Mobility = drift velocity
field gradient
 m/s 
V / m  = m2
/V / s 
 = ne e e + nh e h Species Mobility (m2 / V / s) at RT
Si Ge
Electrons 0.14 0.39
Holes 0.05 0.19
Conductivity as a function of temperature
 = ne e e + nh e h
 Eg 
 = N e (  e +  h ) exp- 
 2 kT 

 Eg 
 = C exp - 
 2 kT 

Eg
ln  = C1 -
2kT Eg
-

Ln()→
2k

1/T (/K) →
Extrinsic semiconductors
 The addition of doping elements significantly increases the conductivity of a semiconductor.
In fact impurity is accidentally present in semiconductors (even in low concentrations like 1
atom in 1012 atoms), which make it extrinsic.

 Doping of Si
 V column element (P, As, Sb) → the extra unbonded electron is practically free (with a
radius of motion of ~ 80 Å)
 Energy level near the conduction band
 n- type semiconductor
 III column element (Al, Ga, In) → the extra electron for bonding supplied by a
neighbouring Si atom → leaves a hole in Si.
 Energy level near the valence band
 p- type semiconductor
n-type  Ionization Energy→
Energy required to promote an electron
from the Donor level to conduction band.
EF EIonization~0.01 eV
 EIonization < Eg
Eg ~1 eV  even at RT large fraction of the donor
electrons are exited into the conduction
Donor level (ED) band.

 Electrons in the conduction band are the majority charge carriers


 The fraction of the donor level electrons excited into the conduction band
is much larger than the number of electrons excited from the valence band
 Law of mass action: (ne)conduction band x (nh)valence band = Constant
 The number of holes is very small in an n-type semiconductor
  Number of electrons ≠ Number of holes
p-type

Acceptor level (EA)


Eg
EF EIonization~0.01 eV

 At zero K the holes are bound to the dopant atom


 As T↑ the holes gain thermal energy and break away from the dopant atom
 available for conduction
 The level of the bound holes are called the acceptor level (which can accept
and electron) and acceptor level is close to the valance band
 Holes are the majority charge carriers
 Intrinsically excited electrons are small in number
  Number of electrons ≠ Number of holes
Ionization energies for dopants in Si & Ge (eV)
Type Element In Si In Ge
P 0.044 0.012
n-type As 0.049 0.013
Sb 0.039 0.010
B 0.045 0.010
Al 0.057 0.010
p-type
Ga 0.065 0.011
In 0.16 0.011
10 4
Eg Intrinsic
- slope
2k
3
All dopant atoms have been excited
10 Exhaustion
R4 R2

10 2
R3
 (/ Ohm / K) →

Exponential
1 R1
10

Ex
function

t
rin
+ve slope due to

sic
Temperature dependent
mobility term
10 0 Slope can be used for the
calculation of EIonization

0.02 0.04 0.06 0.08 0.1


50 K 10 K
1/T (/K) →
 Semiconductor device  chose the flat region where the conductivity does
not change much with temperature
 Thermistor (for measuring temperature)  maximum sensitivity is
required