Transcript Slide 1

Chem 140a
Lecture Notes #8: Class #10
Recombination-Generation Statistics (R-G)
(Shockley-Read-Hall statistics)
<Review>
Photocurrent: due to minority carrier current generated from photoinduced holes, p
h  n  p
n  ND
n
p
ns
nb
p
If band bending is steep enough, all p cross the interface
and J ph is a constant
J   J 0 [exp(  qV / AkT )  1]  J ph
Δp
at Voc, J  0 So,
J 0 [exp(  qV / AkT )  1]  J ph
if V 
3kT
, we can ignore "1"
q
J 0 [exp(  qV / AkT )]  J ph
 Voc 
AkT J ph
ln
q
J0
 Name of this game is
to minimize J 0 to get large Voc
<Review>
SC/Metal interface,
qb
J 0  A T exp( 
)
kT
*
2
: we want to have large b to minimize J 0
SC/Solution interface,
J 0  qket [ A]ns 0
J 0, so ln  J 0,m
MIS,
insulator induces a less than unity probability for the e- to corss the SC interface
Limit of b ,
Max b can only be as big as band gap energy, E g
Five Mechanisms
Majority carrier process
1
Ecb
2
2 Tunneling
5
Minority carrier process
3
ET
1 Thermionic Emission
3 Bulk Recombination
4
Evb
4 Depletion Region Recombination
5 Surface Recombination
Under light or a bias, np>ni2. So there is a driving force to return to the equilibrium.
Extra holes will recombine with extra electrons  Recombination
If somehow we were sucking carriers out of a region of the SC, np<ni2 and carriers
would be generated  Generation
Shockley-Read-Hall statistics: Good model for all of these minority carrier processes
of recombination in the SC
Bulk Recombination
For recombination (loss of carriers),
e1
kn
2
kn’
Ecb
dn
dn
dn


dt R dt 1 dt 2
ET
kp
3
kp’
4
p+
1
dp
dt

R
dp
dp

dt 3 dt 4
fT  fraction of filled trap states
Evb
1  fT  fraction of empty trap states
dn
  kn  nb  (1  fT )  NT
dt 1
2
dn
 kn ' fT  NT
dt 2
[cm-3]
3
dp
  k p  pb  fT  NT
dt 3
[cm-3sec-1]
[cm-3]
(unitless)
[cm3/sec] : has an analogy with  Vth
where  is cross-section and Vth is velocity
dp
 k p ' (1  fT )  NT
4
dt 4
Bulk Recombination
e1
kn
2
kn’
Ecb
ET
kp
3
kp’
4
p+
Evb
rn =
dn
dn
dn


  kn  nb  (1  fT )  N T  kn ' fT  N T
dt R dt 1 dt 2
rp =
dp
dp
dp


  k p  pb  fT  N T  k p ' (1  fT )  N T
dt R dt 3 dt 4
Equilibrium conditions: Principle of detailed balance
 All fundamental processes and their inverse processes must self balance
rn  rp  0
electrons,
kn nb ,0 (1  fT ) N T  kn ' fT N T
kn ' 
1  fT
kn nb ,0
fT
Bulk Recombination
For electrons,
kn nb ,0 (1  fT ) N T  kn ' fT N T
kn ' 
1  fT
kn nb ,0
fT
What is the ratio,
1  fT
?
fT
Probability that a trap is filled is given by Fermi-Dirac Statistics.
fT 
1
exp[( EF  ET ) / kT ]  1
1  fT 
1
exp[( EF  ET ) / kT ]  1
1
fT
exp[( EF  ET ) / kT ]  1
exp[( EF  ET ) / kT ]  1
1



1
1  fT
exp[( EF  ET ) / kT ]  1 exp[( EF  ET ) / kT ]  1  1
1
exp[( EF  ET ) / kT ]  1

fT
1

1  fT exp[( EF  ET ) / kT ]
Bulk Recombination
1  fT
 exp[( E F  ET ) / kT ]
fT
 kn '  kn (
1  fT
)nb ,0  kn nb ,0 exp[( EF  ET ) / kT ]= kn N C exp[ ( EC  ET ) / kT ]
fT
(
nb ,0 =NC exp[ ( EC  E F ) / kT ])
kn ' doesn't depend on the Fermi level and only depends on the trap energy level
For a given trap energy, NC exp[( EC  ET ) / kT ] is a constant.
Let n1 =NC exp[( EC  ET ) / kT ]
Then, kn '  kn n1
Similarly, k p '  k p p1
where p1 =N V exp[( ET  EV ) / kT ]
By plugging in kn ' and k p ' ,
rn =
dn
  kn  nb  (1  fT )  N T  kn  n1  fT  N T
dt R
rp =
dp
  k p  pb  fT  N T  k p  p1  (1  fT )  N T
dt R
Bulk Recombination
Steady state relationship: Major assumption in device problems and analysis
At steady state, rn  rp
( Different from the equilibrium)
U bulk  
dn
dp

 recombination rate
dt R
dt R
Plug in our values for rn & rp
kn  nb  (1  fT )  N T  kn  n1  fT  N T  k p  pb  fT  N T  k p  p1  (1  fT )  N T
N T s cancel and solve for fT
 kn  nb  (1  fT )  kn  n1  fT  k p  pb  fT  k p  p1  (1  fT )
 kn  nb  kn  nb  fT  kn  n1  fT  k p  pb  fT  k p  p1  k p  p1  fT
fT 
kn  nb  k p  p1
kn  nb  kn  n1  k p  pb  k p  p1
1  fT 
k p  pb  kn  n1
kn  nb  kn  n1  k p  pb  k p  p1
Now that we have fT and 1-fT we can use them to calculate U bulk .
Bulk Recombination
U bulk  
dn
dp

dt R
dt
R
=  kn  nb  (1  fT )  N T  kn  n1  fT  N T
U bulk  N T
U bulk  N T
kn  nb  ( k p  pb  kn  n1 )  kn  n1  ( k n  nb  k p  p1 )
kn  ( nb  n1 )  k p  ( pb  p1 )
kn  k p  ( nb  pb  n1  p1 )
kn  ( nb  n1 )  k p  ( pb  p1 )
n1  p1  N C exp[( EC  ET ) / kT ]  N V exp[ ( ET  EV ) / kT ]
 N C  N V  exp[( EC  EV ) / kT ]  ni 2
So the recombination rate in bulk,
U bulk  N T
kn  k p  ( nb  pb  ni 2 )
kn  ( nb  n1 )  k p  ( pb  p1 )