EXCESS CARRIERS IN

SEMICONDUCTORS
Most semiconductor devices operate by the
creation of charge carriers in excess of the
thermal equilibrium values.

Optical absorption
Luminesecence
 These excess carriers can
be created by

Electrolum
inescence
due to
Optical Electron current-a
excitation bombardment forward
biased ‘pn’
Photolumin Cathodolumin junction
escence escence
OPTICAL ABSORPTION

* Good technique for measurement of band gap of a

material
•Two possible cases arise when light of some particular
wave length is shone on a sample.
hν ≥ Eg hν < Eg

Will be Will be transmitted

abosrbed in SC through SC.
OPTICAL ABSORPTION ( Contd.)

b
Ec
a c
Photon of Eg
energy  Eg
Ev
OPTICAL ABSORPTION ( Contd.)

If a beam of photons with h > Eg falls on a semiconductor,

then a predictable amount of absorption is going to occur.
Ratio of transmitted light to incident light will depend on
1. Photon wavelength
2. Thickness of sample
OPTICAL ABSORPTION ( Contd.)

Let us assume : -
Intensity of the incident photon beam directed on the
sample ( photons / cm2 –s ) = I0
Thickness of sample : l
Wavelength of light : 
OPTICAL ABSORPTION ( Contd.)
As the beam passes
through the sample, its
intensity at a distance ‘x’
from the surface can be
calculated by considering
the probability of absorption
within any increment ‘dx’.
Figure 4—4
Band gaps of some common semiconductors
relative to the optical spectrum.
Figure 4—6
Excitation and band-to-band recombination
leading to photoluminescence.
OPTICAL ABSORPTION ( Contd.)

Degradation of the intensity [–dI (x) / dx] is

proportional to the intensity remaining at ‘x’ :
-dI (x) / dx =  . I (x)

=> I (x) = I0 . e-  x
And hence the intensity of light transmitted
through the sample thickness ‘l’ is

It = I0 e-. l

‘ ’ is called the absorption coefficient (cm). –1

‘ ’ will depend upon photon wavelength and the

material.
OPTICAL ABSORPTION ( Contd.)

E = hC / λ
= 1.24 / λ
Luminescence

When electron hole pairs are generated in a

semiconductor, or when carriers are excited into
higher impurity levels, from which they fall to their
equilibrium states, light can be given off by the
material. Many of the semiconductors are well suited
for light emission, particularly the compound
semiconductors with direct band gaps. This general
property of light emission is called luminescence.
Luminescence (contd.)

Photoluminescence electroluminescence
* If carriers are cathodoluminescence *If the
excited by excitation
photon occurs by the
*If the excited introduction of
absorption,
carriers are a current into
created by a sample,
high energy
electron
bombardment
of the material,
Photoluminescence ( Contd.)

Fluorescence Phosphorescence
Fast process. Emission Persists for a longer
of photons stops after time . Main reason
approx. 10-8 sec after behind that
excitation is turned off. recombination occurs
through traps.
Reason behind : Direct
recombination. Applications : TV
Picture tube
Applications :
Fluorescent lamp Material : ZnS
Carrier recombination and generation
Explanation of phenomenon of
phosphorescence
Carrier life time and photoconductivity

Referred to as the resulting

increase in conductivity, when
the excess carriers arise from
optical luminescence
We are going to learn about
different types of recombination
kinetics
Carrier life time and photoconductivity

Direct recombination of electrons and holes

Spontaneous process
(Probability that an electron
and a hole will recombine is
constant in time.)
We actually wish to find out a mathematical relationship, which
can guide us about the minority and majority carrier life time in a
SC with injected excess carriers.
Direct recombination of electrons and holes
(Contd.)
Rate of decay of electrons at any time is
proportional to the number of electrons
remaining at ‘t’ and the number of holes with
some constant of proportionality for
recombination say ‘r ‘.
( Simple : Just proportional to the no. of
electrons which are still available for
recombination and no. of available holes with
which the electrons are supposed to eventually
recombine)
Direct recombination of electrons and
holes (Contd.)

So, recombination rate should be somehow

equal to ‘r. n(t) p(t)’.
Direct recombination of electrons
and holes ( Contd.)
The net rate of change in the conduction band
electron concentration is the thermal
generation rate r . ni2 minus the
recombination rate.

Hence dn (t) / dt = r . ni2 - r . n(t) . p(t)

Direct recombination of electrons and holes ( Contd.)
d n(t) / dt = r . ni2 - r . [ n0 +  n(t) ].[ p0 +  p(t) ]
= -r [ ( n0 + p0 )  n(t) +  n2 (t) ]

An equation which seems difficult to be solved, but

not that difficult for Low level injection

Low level injection refers to the case, when

majority carrier population remains essentially
unperturbed after injection of excess carriers,
while minority carrier population may get
affected. .
In case of high level injection, even the
majority carrier population may get perturbed.
Direct recombination of electrons and holes ( Contd.)

d n(t) / dt = -r [ ( n0 + p0 )  n(t) +  n2 (t) ]

Considering
Considering material to be low level
extrinsic (p type), equilibrium injection, this
minority carrier density can be term can be
ignored ( i.e. po >> no) ignored

We get the eqn. as

d  n(t) / dt = - r . p0 .  n(t)
Direct recombination of electrons and holes ( Contd.)
d  n(t) / dt = - r . p0 .  n(t)

The solution to this eqn. is an exponential decay from

the original excess carrier concentration  n :
 n ( t) =  n e -r p0. t =  n e-t / n
where n = (r . p0 )-1 called the recombination life time.
And as the calculation has been made in terms of
minority carriers, so n is called the minority carrier life
time.
Direct recombination of electrons and
holes ( Contd.)
Similarly, the decay of excess holes in n- type
material occurs with p = (r . n0 )-1
A more general expression for the carrier life time
is
n = 1 / r ( n0 + p0 )
This expression is valid for n or p – type material,
if the injection level is low.
Problem : Assume a sample of GaAs doped with
1015 acceptors/cm3. The intrinsic carrier
concentration of GaAs is approximately 106 cm-3.
If 1014 EHP/cm3 are created at t = 0, approximate
the decay of these carriers in time. Assume n =
p = 10-8 sec
Solution : po = 10 15, ni = 106
So, no = ni2 / po = 10-3
Hence no << po So, we should be able to use
relationship of minority carrier life time for
electrons
Total conc. Of electrons at t = 0 is 1014
and that of holes is 10 14 + 1015 ( Δ p + po )
Solution to problem :
 INDIRECT RECOMBINATION

# Occurs through traps

# Any impurity or lattice defect can serve as a
recombination center, if it is capable of receiving a
carrier of one type and subsequently capturing a
carrier of the other type , thereby annihilating the
pair.
INDIRECT RECOMBINATION
( Contd.)
EC
EF
ER
Ei

EV
INDIRECT RECOMBINATION ( Contd.)
Photoconductive decay experiment.
Used for measurement of effects of recombination and
trapping
Following general principle of
 (t) = q . [ n (t) n + p(t) p ]

These figures are going

to change due to
recombination
Ultimately , change in resistance has to be measured
Figure 4—9
Energy levels of impurities in Si. The energies are measured
from the nearest band edge (Ev or Ec); donor levels are
designated by a plus sign and acceptors by a minus sign.
INDIRECT RECOMBINATION ( Contd.)
 (When there is no
trapping)

n
=
n
gopis an Optical Generation Rate.
 Steady state ( Contd.)

Thus the excess carrier densities for

electrons and holes under steady
illumination will be given as :
δn = gop n
δp = gop p
 Quasi Fermi levels

* Quasi Fermi levels are indicators of relative

occupancy of bands, under conditions, when
excess carriers are present

* There are separate Quasi Fermi levels for

electrons and holes represented by Fn and Fp
respectively.
 Quasi Fermi levels ( Contd.)

The positions of Quasi Fermi levels

relative to the position of intrinsic Fermi
level are given by following two
equations :
n = ni e (Fn – Ei) / kT
p = ni e (Ei – Fp) / kT
 Quasi Fermi levels ( Contd.)

In summary When excess carriers are

present, the deviations of Fn and Fp from Ef
indicate how far the electron and hole
populations are from the equilibrium values no
and po.
The separation of Quasi Fermi levels Fn – Fp
is a direct measure of the deviation from
equilibrium
( At equilibrium Fn = Fp = Ef).
Problem : Suppose that in a sample 1013 EHP / cm3 are
created optically every microsecond in a silicon sample with
n0 = 1014 / cm3 and  n =  p = 2 micro seconds.
(a) Calculate the steady state excess electron ( or hole )
concentration.
(b) Find out the percentage change in the concentration of
electrons and holes as a result of induction of these excess
carriers.
(c) Find the position of Quasi Fermi levels for electrons and
holes w.r.t intrinsic level. Show these levels with the help of
an energy band diagram.
 Behavior when illuminated
Given Si :
no  1014 cm 3 .
ni  1.5  1010 cm 3 .
p o  _________
2.25 106 cm 3
Ef  Ei  0________
.228 eV
(at Equlibrium)
Fn  E f 1019 EHP/cc - sec.
 n   p  2 sec .
Ei  10 cm
n  2____
13 3

FP p  2____
 1013 cm 3
 1013 cm 3
p  2_____
.2  1014 cm 3
n  1_____
Fn  Ei  0_____
.233 eV
Ei  F p  0_____
.186 eV

Before Photoluminescence
no  1014 cm 3 , po  2.25  10 6 , E f  E i  0.228eV At equilibrium
Figure 4—11
Quasi-Fermi levels Fn and Fp for a Si sample with n0 = 1014
cm23, tp = 2 ms, and gop = 1019 EHP/cm3-s (Example 4–4).
Diffusion Process

Figure 4—12
Spreading of a pulse of electrons by diffusion.
Figure 4—13
An arbitrary electron concentration gradient in one dimension: (a)
division of n(x) into segments of length equal to a mean free path for the
electrons; (b) expanded view of two of the segments centered at x0.
Figure 4—14
Drift and diffusion directions for electrons and holes in a carrier gradient
and an electric field. Particle flow directions are indicated by dashed
arrows, and the resulting currents are indicated by solid arrows.
Figure 4—15
Energy band diagram of a semiconductor in an electric field %(x).
Problem : An intrinsic Si sample is
doped with donors from one side
such that Nd = N0 exp (-ax) .
(a)Find an expression for E(x) at
equilibrium over the range for which
Nd >> ni .
(b)Evaluate E(x), when a = 1(m )-1 .
(c)Sketch a resulting band diagram and
indicate the direction of ‘E’.
(a) E(x) at equilibrium can be found from
Jn = + q. Dn ( n /  x ) + q un n(x) E(x)
E(x) = –Dn dn / dx = kT . a
un n q

(b) E (x) = .0259 ( 104 ) = 259 V / cm.

Resulting band diagram

Ex

Ec

Ei

Ev
X
 Diffusion and recombination
If we consider a differential length x of a
semiconductor sample with area A. Now ,
the hole current density leaving the
volume Jp ( x + x ) can be larger or
smaller than the current density entering
Jp ( x ) depending on the generation and
recombination of carriers taking place with
in the volume.
This aspect is taken care of in Continuity equation,
which is basically a charge conservation eqn.
 Continuity equation( Contd.)

The net increase in hole p

concentration per unit time, t =
Difference between the hole flux per unit volume
entering and leaving , minus the recombination rate.
 Continuity equation (Contd.)
p 1 J p ( x)  J p ( x  x) p
 
t x  x  x q x p

Rate of increase
holes of holes Recombi
build up conc.in nation
Δx.A / Rate
unit time
 Continuity equation (Contd.)
• In the limit as Δx -> 0 hole continuity eqn.
can be written as
p 1 J p p
 
t q x  p
Similarly
• Continuity eqn. for electrons may be
written as
n 1 J n n
 
t q x  n
 The Diffusion equation
• If the current flow is entirely due to diffusion,
then continuity eqn for holes can be written as :

p  p p 2
 Dp 
t x 2
p
p
As J p ( Diff )  qD p
x
 The Diffusion equation
Similarly, we can write the diffusion
equation for electrons.
n  n n
2
 Dn 
t x 2
n
Both these equations i.e. diffusion equation
for electrons and holes are useful in solving
transient problems of diffusion with
recombination.
Steady state carrier injection :
Diffusion Length
When steady state distribution of excess
carriers is maintained, then in diffusion
eqn. of electrons and holes

n p
 0
t t
 Steady state carrier injection(
Contd.)
Notice d n 2
n n
the   2
change dx 2
Dn n Ln
here.
d p2
p p
  2
dx 2
D p p L p

Ln  Dn n and L p  D p p
are diffusion lengths for electrons and holes
respectively.
 Steady state carrier injection(
Contd.)
Take a semi infinite semiconductor bar and inject
some holes at position x = 0.
p ( x = 0 ) = p
Somehow, steady state hole injection maintains
constant excess hole concentration at the
injection point.
What will happen now?
 Steady state carrier injection(
Contd.)
The injected holes diffuse along the bar ,
recombining with a characteristic life time
p .
In steady state, we expect the distribution of
excess holes to decay to zero for large
values of ‘x’, because of recombination.
For this problem, we shall use the steady
state diffusion equation for holes
 Solution to steady state Diffusion equation

Using boundary conditions

δp = Δp at x = 0 and δp = 0 at x = α
and solving the Diffusion eqn, we get
 x / Lp
p( x)  p.e
From this equation, it is clear that the injected
excess hole concentration dies out
exponentially in ‘x’ due to recombination.
Diffusion length Lp represents the distance at
which the excess hole distribution is reduced to 1/e
of its value at the point of injection. It can be shown
that Lp is the average distance , a hole diffuses
before recombining.
 Solution to steady state
Diffusion equation
To calculate an average diffusion length, we
must obtain an expression for the probability
that an injected hole recombines in a particular
interval dx.
The probability that a hole injected at x = 0
survives to ‘x’ without recombination is
p ( x) / p = exp ( -x /Lp )

the ratio of the steady state concentrations at ‘x’

and 0.
 Solution to steady state
Diffusion equation
• On the other hand, the probability that a
hole at ‘x’ will recombine in the subsequent
interval dx is

p( x)  p( x  dx)  (dp( x) / dx)dx 1

  dx
p ( x ) p ( x ) Lp
 Solution to steady state
Diffusion equation
• Thus the total probability that a hole
injected at x = 0 will recombine in a
given dx is the product of the two
probabilities
1  x / Lp
.e dx
Lp
From this equation, we can use the
averaging technique so as to find the
average distance a hole diffuses
before recombining is

  x / Lp
e
x  x dx  L p
0
Lp
Resulting hole current
For excess hole concentration :
p(x) = p e –x / L
p

For diffusion current density due to holes

Jp (diff) = - q Dp dp(x) / dx
So, we get :

dp dp Dp  x / Lp Dp
J p ( x)  qD p  qD p q pe q p ( x )
dx dx Lp Lp
PROBLEM : In a very long p – type
silicon bar with cross sectional area =
0.5 cm2 and Na = 1017 cm-3 , we inject
holes such that the steady state
excess hole concentration is 5 x 1016
cm-3 at x = 0. What is the steady state
separation between Fp and Ei at x =
1000 A0 ? What is the hole current
there ? Assume mobility of holes = 500
cm2 / V – sec and p = 10-10 sec.
 Haynes Shockley
Experiment

First time performed by J.R.Haynes and

W.Shockley in 1951 at Bell Telephone
Laboratories.
Basically used to measure the minority
carrier mobility and diffusion coefficient
(D).
 Haynes Shockley
Experiment ( Contd.)
Light Electric field (E)
pulse

n - type

X=0 X=L
•Assume ( no >> po ) .
•Assume that the excess carriers have a
negligible effect on the electron
concentration, but change the hole
concentration significantly.
Figure 4—18
Drift and diffusion of a hole pulse in an n-type bar:
(a) sample geometry; (b) position and shape of the
pulse for several times during its drift down the bar.
 Haynes Shockley
Experiment ( Contd.)
• Say excess holes travel a distance of ‘L’
• Measure the drift time td
• And, Calculate the drift velocity ‘vd’
vd = L / td
• and hence the hole mobility

p = vd / E
 Haynes Shockley
Experiment ( Contd.)
• Most conveniently, Dp can be expressed
as
Dp = (x)2 / 16 td
Where x is the
distance, where the
value of excess td is the
carrier Corresponding
concentration is
down by 1 / e of its time
peak value
 Haynes Shockley
Experiment ( Contd.)
• There is no direct way to find x, So it is
found with help of relation :

x = t . vd = t . L / td
are separated by 0.95 cms. Battery voltage
Eo is 2 V. A pulse arrives at point (2) 0.25
ms after injection at (1); the width of pulse
t is 117 usec.
Calculate the hole mobility and diffusion
coefficient.
p = vd / E=1900 cm2/vs
Dp = (x)2 / 16 td
x = t . vd = t . L / td
Dp=49.4
 Gradients in the Quasi Fermi
levels
dn( x)
J n ( x)  q n n( x) E ( x)  qDn
dx

Electron current density Gradient in

in case in case of non Electron
equilibrium electron concentration
concentration
Gradients in the Quasi Fermi
levels (Contd.)

Differentiating
 Gradients in the Quasi Fermi
levels (Contd.)
• Using Einstein’s relationships for Dn , we get

dFn dEi
Jn ( X )  q  nn( X ) E ( X )   nn( X )[  ]
dX dX
After Simplification, Because

dEi
  q.E ( x)
dx
Gradients in the Quasi Fermi
levels (Contd.)

From this equation, physically, we can

write that Current flows due to rate of
change of Quasi Fermi level with
position
 Gradients in the Quasi Fermi
levels (Contd.)
IN TERMS OF MODIFIED Ohm’s LAW
d ( Fn / q) d ( Fn / q )
Jn ( X )  q  nn( X )   n( X )
dX dX
d ( Fp / q) d ( Fp / q )
Jp ( X )  q  pp( X )   p( X )
dX dX
Therefore, any drift, diffusion or combination of the
two in a Semiconductor results in Currents
proportional to the gradients of the two Quasi-Fermi
Levels.