Key questions

• Are carriers sitting still in thermal equilibrium?

• How do carriers move in an electric field? What are the key dependencies
of the drift velocity?

• How do the energy band diagrams represent the presence of an electric

field?

• How does a concentration gradient affect carriers?

Slide 2-1
 Chapter 2 Motion and Recombination
of Electrons and Holes
2.1 Thermal Motion
We can think of carriers as particles in an ideal gas.
At finite T, carriers have finite thermal energy. All this energy resides in
the kinetic energy of the particles.
Carriers move in random directions: no net velocity, but average carrier
velocity is thermal velocity:

3 1 2
Average electron or hole kinetic energy  kT  mvth
2 2
3kT 3  1.38  10 23 JK 1  300K
vth  
meff 0.26  9.1 10 31 kg

 2.3 105 m/s  2.3  107 cm/s

Slide 2-2
 2.1 Thermal Motion

But…. Semiconductor crystal is not perfect:

• the Si atoms themselves are vibrating around their equilibrium position in
the lattice
• Zig-zag motion is due to collisions or scattering with imperfections in
the crystal.
• Mean free path, Ɩc: average distance travelled between collisions [cm].
• Mean time between collisions is m ~ 0.1ps

Slide 2-3
Scattering mechanisms:
1.Lattice or phonon scattering: carriers collide with vibrating lattice atoms
(phonon absorption and emission)
 Some energy exchanged (~ tens of meV)

2.ionized impurity scattering: Coulombic interaction between charged

impurities and carriers
 no energy exchanged

3.Surface scattering in inversion layer

4.Neutral impurity scattering with neutral dopants, interstitials, vacancies,

etc

5.Carrier-carrier scattering

Slide 2-4
 Impurity (Dopant)-Ion Scattering or Coulombic Scattering

Boron Ion Electron

_
- -
Electron +
Arsenic
Ion

There is less change in the direction of travel if the electron zips by

the ion at a higher speed.
3 3/ 2
v T
impurity  th
Na  Nd Na  Nd

Slide 2-5
 2.2 Drift

2.2.1 Electron and Hole Mobilities

• Drift is the motion caused by an electric field.

• Nonzero velocity is called the drift velocity

Slide 2-6
 2.2.1 Electron and Hole Mobilities

m p v  qE mp

q E  mp
v
mp

v   pE v    nE
q mp q mn
p  n 
mp mn

• p is the hole mobility and n is the electron mobility

Slide 2-7
 2.2.1 Electron and Hole Mobilities

v = E ;  has the dimensions of v/E  cm/s  cm  .

2

 
 V/cm V s 

Electron and hole mobilities of selected

semiconductors
Si Ge GaAs InAs
 n (cm2/V∙s) 1400 3900 8500 30000
 p (cm2/V∙s) 470 1900 400 500

Based on the above table alone, which semiconductor and which carriers
(electrons or holes) are attractive for applications in high-speed devices?

Slide 2-8
 Drift Velocity, Mean Free Time, Mean Free Path
EXAMPLE: Given p = 470 cm2/V·s, what is the hole drift velocity at
E= 103 V/cm? What is mp and what is the distance traveled between
collisions (called the mean free path)? Hint: When in doubt, use the
MKS system of units.
Solution:  = pE = 470 cm2/V·s  103 V/cm = 4.7 105 cm/s
mp = pmp/q =470 cm2/V ·s  0.39 9.110-31 kg/1.610-19 C
= 0.047 m2/V ·s 2.210-12 kg/C = 110-13s = 0.1 ps
mean free path = mhth ~ 1 10-13 s  2.2107 cm/s
= 2.210-6 cm = 220 Å = 22 nm
This is smaller than the typical dimensions of devices, but getting close.

Slide 2-9
 Total Mobility
1600
1 1 1
1400  
  phonon  impurity
1200 Electrons
1 1 1
 
M obility (cmV s )
-1

1000

  phonon  impurity
-1

800
2

600

400
Holes
200

1E14 1E15 1E16 1E17 1E18 1E19 1E20

Na +Concenration
Total Impurity Nd (cm-3) -3
(atoms cm )

Slide 2-10
Temperature Effect on Mobility

10 1 5

Question:
What Nd will make
dn/dT = 0 at room
temperature?

Slide 2-11
 Velocity Saturation
• When the kinetic energy of a carrier exceeds a critical value, it
generates an optical phonon and loses the kinetic energy.
• Therefore, the kinetic energy is capped at large E, and the
velocity does not rise above a saturation velocity, vsat .
• Velocity saturation has a deleterious effect on device speed as
shown in Ch. 6.

Slide 2-12
 2.2.3 Drift Current and Conductivity
E Jp

unit
+ area

+ 

Hole current density Jp = qpv A/cm2 or C/cm2·sec

EXAMPLE: If p = 1015cm-3 and v = 104 cm/s, then

Jp= 1.610-19C  1015cm-3  104cm/s
= 1.6 C/s  cm 2
 1.6 A/cm 2

Slide 2-13
 2.2.3 Drift Current and Conductivity

Jp,drift = qpv = qppE

Jn,drift = –qnv = qnnE

Jdrift = Jn,drift + Jp,drift =  E =(qnn+qpp)E

 conductivity (1/ohm-cm) of a semiconductor is

 = qnn + qpp

1/ = is resistivity (ohm-cm)

Slide 2-14
Relationship between Resistivity and Dopant Density

DOPANT DENSITY cm-3

P-type

N-type

RESISTIVITY
(
m) = 1/

c

Slide 2-15
EXAMPLE: Temperature Dependence of Resistance
(a) What is the resistivity () of silicon doped
with 1017cm-3 of arsenic?
(b) What is the resistance (R) of a piece of this
silicon material 1m long and 0.1 m2 in cross-
sectional area?
Solution:
(a) Using the N-type curve in the previous
figure, we find that  = 0.084 -cm.
(b) R = L/A = 0.084 -cm  1 m / 0.1 m2
= 0.084 -cm 10-4 cm/ 10-10 cm2
= 8.4  10-4 
Slide 2-16
EXAMPLE: Temperature Dependence of Resistance

By what factor will R increase or decrease from

T=300 K to T=400 K?

Solution: The temperature dependent factor in  (and

therefore ) is n. From the mobility vs. temperature
curve for 1017cm-3, we find that n decreases from 770
at 300K to 400 at 400K. As a result, R increases by
770
 1.93
400

Slide 2-17
Energy band diagram under electric field

Slide 2-18
Slide 2-19
 2.3 Diffusion Current

Particles diffuse from a higher-concentration location

to a lower-concentration location.

Slide 2-20
 2.3 Diffusion Current
dn dp
J n ,diffusion  qDn J p ,diffusion  qD p
dx dx
D is called the diffusion constant. Signs explained:
n p

x x

Slide 2-21
Total Current – Review of Four Current Components

JTOTAL = Jn + Jp

dn
Jn = Jn,drift + Jn,diffusion = qnnE + qDn
dx

dp
Jp = Jp,drift + Jp,diffusion = qppE – qD p
dx

Slide 2-22
 E
–
2.4 Relation Between the Energy +
Si

Diagram and V, E (a )

V (x )
0.7eV

0 .7 V

x
+ N- – 0
N type Si
(b )

E
Ec and Ev vary in the opposite - E c(x )
E E cE (x)
( )
f x
direction from the voltage. -
E f (x)
That
E v(x )
is, Ec and Ev are higher where E v (x)
0 .7 V
the voltagedV
is lower.
1 dEc 1 dEv + 0.7V
E(x)= dx q dx  q dx
  x
+

x (c )

Slide 2-23
2.5 Einstein Relationship between D and 
E f

E c( x )
Consider a piece of non-uniformly doped semiconductor.

 ( Ec  E f ) / kT
N n- -t tyy pp ee s es me mi c o inc d ou cn t do ru c t o r n  N ce
dn N ( E  E ) / kT dEc
D e c r e a s in g d o n o r c o n c e n t r a ti o n  ce c f
dx kT dx
Ec(x) n dEc

kT dx
Ef
n
  qE
Ev(x) kT

Slide 2-24
 2.5 Einstein Relationship between D and 
dn n
kT   qE
: Thermal voltage dx kT
q
dn
Jn qn nE qD n  0 at equilibrium.
dx
qD n
0 qn nE qn E
kT
kT kT
Dn  n Similarly, Dp  p
q q
These are known as the Einstein relationship.
Einstein relation does not depend on doping, only
on Temperature
Slide 2-25
 EXAMPLE: Diffusion Constant

What is the hole diffusion constant in a piece of

silicon with p = 410 cm2 V-1s-1 ?

Solution:
 kT 
D p     p  (26 mV)  410 cm 2 V 1s 1  11 cm 2 /s
 q 

Remember: kT/q = 26 mV at room temperature.

Slide 2-26
 2.6 Electron-Hole Recombination
•The equilibrium carrier concentrations are denoted with
n0 and p0.
•The total electron and hole concentrations can be
different from n0 and p0 .
•The differences are called the excess carrier
concentrations n’ and p’.

n  n0  n'
p  p0  p '

Slide 2-27
 Charge Neutrality
•Charge neutrality is satisfied at equilibrium (n’=
p’= 0).
• When a non-zero n’ is present, an equal p’ may
be assumed to be present to maintain charge
equality and vice-versa.
•If charge neutrality is not satisfied, the net charge
will attract or repel the (majority) carriers through
the drift current until neutrality is restored.

n'  p'

Slide 2-28
 Recombination Lifetime

•Assume light generates n’ and p’. If the light is

suddenly turned off, n’ and p’ decay with time
until they become zero.
•The process of decay is called recombination.
•The time constant of decay is the recombination
time or carrier lifetime,  .
•Recombination is nature’s way of restoring
equilibrium (n’= p’= 0).

Slide 2-29
 Recombination Lifetime
 ranges from 1ns to 1ms in Si and depends on
the density of metal impurities (contaminants)
such as Au and Pt.
•These deep traps capture electrons and holes to
facilitate recombination and are called
recombination centers.
Ec
Direct
Recombination Recombination
is unfavorable in centers
silicon
Ev

Slide 2-30
 Direct and Indirect Band Gap

Trap

Direct band gap Indirect band gap

Example: GaAs Example: Si
Direct recombination is efficient Direct recombination is rare as k
as k conservation is satisfied. conservation is not satisfied
 Rate of recombination (s-1cm-3)

dn n

dt 

n  p 

dn n p dp
  
dt   dt

Slide 2-32
 EXAMPLE: Photoconductors

A bar of Si is doped with boron at 1015cm-3. It is

exposed to light such that electron-hole pairs are
generated throughout the volume of the bar at the
rate of 1020/s·cm3. The recombination lifetime is
10s. What are (a) p0 , (b) n0 , (c) p’, (d) n’, (e) p ,
(f) n, and (g) the np product?

Slide 2-33
 EXAMPLE: Photoconductors

Solution:

(a) What is p0?

p0 = Na = 1015 cm-3

(b) What is n0 ?
n0 = ni2/p0 = 105 cm-3

(c) What is p’?

In steady-state, the rate of generation is equal to the
rate of recombination.
1020/s-cm3 = p’/
 p’= 1020/s-cm3 · 10-5s = 1015 cm-3
Slide 2-34
 EXAMPLE: Photoconductors

(d) What is n’?

n’= p’= 1015 cm-3

(e) What is p?
p = p0 + p’= 1015cm-3 + 1015cm-3 = 2×1015cm-3

(f) What is n?
n = n0 + n’= 105cm-3 + 1015cm-3 ~ 1015cm-3 since n0 << n’

(g) What is np?

np ~ 21015cm-3 ·1015cm-3 = 21030 cm-6 >> ni2 = 1020 cm-6.
The np product can be very different from ni2.

Slide 2-35
 2.7 Thermal Generation

If n’ is negative, there are fewer

electrons than the equilibrium value.

As a result, there is a net rate of

thermal generation at the rate of |n|/ .

Slide 2-36
 2.8 Quasi-equilibrium and Quasi-Fermi Levels
• Whenever n’ = p’  0, np  ni2. We would like to preserve
and use the simple relations:  ( Ec  E f ) / kT
n  Nce
 ( E f  Ev ) / kT
p  Nve

• But these equations lead to np = ni2. The solution is to introduce

two quasi-Fermi levels Efn and Efp such that ( Ec  E fn ) / kT
n  Nce
 ( E fp  Ev ) / kT
p  Nve
Even when electrons and holes are not at equilibrium, within
each group the carriers can be at equilibrium. Electrons are
closely linked to other electrons but only loosely to holes.

Slide 2-37
EXAMPLE: Quasi-Fermi Levels and Low-Level Injection

Consider a Si sample with Nd=1017cm-3 and n’=p’=1015cm-3.

(a) Find Ef .
n = Nd = 1017 cm-3 = Ncexp[–(Ec– Ef)/kT]
 Ec– Ef = 0.15 eV. (Ef is below Ec by 0.15 eV.)

Note: n’ and p’ are much less than the majority carrier

concentration. This condition is called low-level
injection.

Slide 2-38
 EXAMPLE: Quasi-Fermi Levels and Low-Level Injection

Now assume n = p = 1015 cm-3.

(b) Find Efn and Efp .
 ( Ec  E fn ) / kT
n = 1.0110 cm =
17 -3 Nce

 Ec–Efn = kT ln(Nc/1.011017cm-3)
= 26 meV ln(2.81019cm-3/1.011017cm-3)
= 0.15 eV

Efn is nearly identical to Ef because n  n0 .

Slide 2-39
 EXAMPLE: Quasi-Fermi Levels
 ( E fp  Ev ) / kT
p = 10 cm =
15 -3 Nve

 Efp–Ev = kT ln(Nv/1015cm-3)
= 26 meV ln(1.041019cm-3/1015cm-3)
= 0.24 eV

Ec

Ef Efn

Efp
Ev

Slide 2-40
 2.9 Chapter Summary

vp   pE
dn
J n ,diffusion  qDn
dx
vn  -  nE
dp
J p ,diffusion   qD p
J p ,drift qp pE dx
Jn ,drift qn nE
kT
Dn  n
q
kT
Dp  p
q

Slide 2-41
 2.9 Chapter Summary
 is the recombination lifetime.
n’ and p’ are the excess carrier concentrations.
n = n0+ n’
p = p0+ p’
Charge neutrality requires n’= p’.

rate of recombination = n’/ = p’/

Efn and Efp are the quasi-Fermi levels of electrons and

 ( E  E ) / kT
holes. n  Nce c fn

 ( E fp  Ev ) / kT
p  Nve

Slide 2-42