sd_pn_3

Semiconductor Devices
p-n Junctions: Part 3
M. B. Patil
mbpatil@ee.iitb.ac.in
www.ee.iitb.ac.in/~sequel
Department of Electrical Engineering

Indian Institute of Technology Bombay
M. B. Patil, IIT Bombay

pn junction: derivation of I -V equation
Vpcontact Vpneutral Vj Vnneutral Vncontact
metal metal
Vp p+ p n n+ Vn
xj
p contact p neutral depletion n neutral n contact

region region region region region
∂p(x, t) 1 ∂Jp
Continuity equation for holes (x > xn ): =− − (R − G ) = 0 (in DC conditions).
∂t q ∂x

metal metal
Vp p+ p n n+ Vn
xj

∂p(x, t) 1 ∂Jp
∂t q ∂x
dp
In the neutral n-region, E is small. → Jpdrift = qpµp E is small. → Jp ≈ Jpdiff = −qDp .
dx

metal metal
Vp p+ p n n+ Vn
xj

∂p(x, t) 1 ∂Jp
∂t q ∂x
dp
dx
∆p p(x) − pn0
Also, assuming low-level injection, R − G ≈ = .
τp τp

metal metal
Vp p+ p n n+ Vn
xj

∂p(x, t) 1 ∂Jp
∂t q ∂x
dp
dx
∆p p(x) − pn0
Also, assuming low-level injection, R − G ≈ = .
τp τp
d 2p p − pn0 d 2 ∆p ∆p p
→ Dp − = 0 or − 2 = 0, where Lp = Dp τp is the hole diffusion length.
dx 2 τp dx 2 Lp

metal metal
Vp p+ p n n+ Vn
xj

Example: For an abrupt, uniformly doped silicon pn junction at T = 300 K, with Na = 1016 cm−3 ,
Nd = 1017 cm−3 , µp = 500 cm2 /V-s, calculate the diffusion length Lp for τp = 1 ns, 10 ns, 100 ns, 1 µs, and
10 µs, and compare it with the zero-bias value of W , the depletion width.

metal metal
Vp p+ p n n+ Vn
xj

!
1016 × 1017

Na Nd
Solution: Vbi = VT log = (0.0259 V) log = 0.75 V.
ni2 (1.5 × 1010 )2

metal metal
Vp p+ p n n+ Vn
xj

!
1016 × 1017

Na Nd
ni2 (1.5 × 1010 )2
s
2 Na + Nd
The depletion width is W = (Vbi − Va )
q Na Nd

metal metal
Vp p+ p n n+ Vn
xj

!
1016 × 1017

Na Nd
ni2 (1.5 × 1010 )2
s
2 Na + Nd
q Na Nd
s
2 × 11.8 × 8.85 × 10−14 1.1 × 1017
= × 0.75 cm = 0.33 µm.
1.6 × 10−19 1016 × 1017

metal metal
Vp p+ p n n+ Vn
xj

!
1016 × 1017

Na Nd
ni2 (1.5 × 1010 )2
s
2 Na + Nd
q Na Nd
s
2 × 11.8 × 8.85 × 10−14 1.1 × 1017
= × 0.75 cm = 0.33 µm.
1.6 × 10−19 1016 × 1017
Diffusion coefficient for holes is Dp = VT µp = 0.0259 × 500 = 12.9 cm2 /s.

metal metal
Vp p+ p n n+ Vn
xj

!
1016 × 1017

Na Nd
ni2 (1.5 × 1010 )2
s
2 Na + Nd
q Na Nd
s
2 × 11.8 × 8.85 × 10−14 1.1 × 1017
= × 0.75 cm = 0.33 µm.
1.6 × 10−19 1016 × 1017
Diffusion coefficient for holes is Dp = VT µp = 0.0259 × 500 = 12.9 cm2 /s.
r
cm2
For τp = 1 ns, Lp = 12.9 × (1 × 10−9 s) = 1.14 × 10−4 cm = 1.14 µm.
s
metal metal
Vp p+ p n n+ Vn
xj


metal metal
Vp p+ p n n+ Vn
xj

τp Lp (µm)
1 ns 1.14
10 ns 3.6
100 ns 11.4
1 µs 36.0
10 µs 113.8

metal metal
Vp p+ p n n+ Vn
xj

τp Lp (µm)
1 ns 1.14
10 ns 3.6
100 ns 11.4
1 µs 36.0
10 µs 113.8
Note that Lp W |0 V (0.33 µm), a typical situation.

Vp p+ p n n+ Vn
xp xj xn
d 2 ∆p ∆p
Hole continuity equation (x > xn ): − 2 = 0,.
dx 2 Lp

Va Va
Boundary conditions: ∆p(xn ) = pn0 exp − pn0 = pn0 exp −1
VT VT
∆p(x → ∞) = p(x → ∞) − pn0 = 0
x − xn

Va
→ ∆p(x) = pn0 exp − 1 exp − , x > xn ,
VT Lp
xp − x

Va
∆n(x) = np0 exp − 1 exp − , x < xp .
VT Ln
Va > 0 V p n Va < 0 V p n
p
n
∆p p
∆n pn0 n pn0
np0 np0
0
p
n
∆p p
∆n pn0 n pn0
np0 np0
0
x − xn xp − x

Va Va
∆p(x) = pn0 exp − 1 exp − , ∆n(x) = np0 exp − 1 exp − .
VT Lp VT Ln

p
n
∆p p
∆n pn0 n pn0
np0 np0
0
x − xn xp − x

Va Va
VT Lp VT Ln
* When x − xn = 5Lp , the exponential factor in ∆p(x) is e −5 = 0.0067 → In about five minority carrier
diffusion lengths, the disturbance caused by the applied bias vanishes.

p
n
∆p p
∆n pn0 n pn0
np0 np0
0
x − xn xp − x

Va Va
VT Lp VT Ln
* Consider the minority carrier concentrations at the depletion region edges.

Va
∆p = pn0 exp − 1 at x = xn ,
V
T
Va
∆n = np0 exp − 1 at x = xp .
VT

p
n
∆p p
∆n pn0 n pn0
np0 np0
0
x − xn xp − x

Va Va
VT Lp VT Ln

Va
∆p = pn0 exp − 1 at x = xn ,
V
T
Va
∆n = np0 exp − 1 at x = xp .
VT
For forward bias, ∆p(xn ) and ∆n(xp ) are positive.

p
n
∆p p
∆n pn0 n pn0
np0 np0
0
x − xn xp − x

Va Va
VT Lp VT Ln

Va
∆p = pn0 exp − 1 at x = xn ,
V
T
Va
∆n = np0 exp − 1 at x = xp .
VT
For forward bias, ∆p(xn ) and ∆n(xp ) are positive.
For reverse bias, ∆p(xn ) and ∆n(xp ) are negative.

p n Va > 0 V p n Va < 0 V
n
p
n
np0 ∆n ∆p np0 p
pn0 pn0
0
Consider an abrupt, uniformly doped silicon pn junction at T = 300 K, with Na = 5 × 1016 cm−3 and
Nd = 1018 cm−3 . Compute ∆n(xp ) and ∆p(xn ) for Va = 0.1, 0.2, 0.3, 0.6, 0.7, −0.1, −0.2, −0.5, −1, and
−2 V. (ni = 1.5 × 1010 cm−3 for silicon at T = 300 K.)
ni2 (1.5 × 1010 )2

Solution: pp0 ≈ Na = 5 × 1016 cm−3 → np0 = = = 4.5 × 103 cm−3 .
pp0 5 × 1016

n
p
n
np0 ∆n ∆p np0 p
pn0 pn0
0
ni2 (1.5 × 1010 )2

pp0 5 × 1016
ni2 (1.5 × 1010 )2
nn0 ≈ Nd = 1 × 1018 cm−3 → pn0 = = = 2.25 × 102 cm−3 .
nn0 1 × 1018

n
p
n
np0 ∆n ∆p np0 p
pn0 pn0
0
ni2 (1.5 × 1010 )2

pp0 5 × 1016
ni2 (1.5 × 1010 )2
nn0 ≈ Nd = 1 × 1018 cm−3 → pn0 = = = 2.25 × 102 cm−3 .
nn0 1 × 1018

Va Va
∆p(xn ) = pn0 exp − 1 , ∆n(xp ) = np0 exp −1 .
VT VT

n
p
n
np0 ∆n ∆p np0 p
pn0 pn0
0
Va ∆n(xp ) ∆p(xn ) Va ∆n(xp ) ∆p(xn )

(V) (cm−3 ) (cm−3 ) (V) (cm−3 ) (cm−3 )
0 0 0 0 0 0
0.1 2.09 × 105 1.05 × 104 −0.1 −4.41 × 103 −2.20 × 102
0.2 1.02 × 107 5.08 × 105 −0.2 −4.50 × 103 −2.25 × 102
0.3 4.83 × 108 2.41 × 107 −0.5 −4.50 × 103 −2.25 × 102
0.6 5.18 × 1013 2.59 × 1012 −1 −4.50 × 103 −2.25 × 102
0.7 2.46 × 1015 1.23 × 1014 −2 −4.50 × 103 −2.25 × 102

n
p
n
np0 ∆n ∆p np0 p
pn0 pn0
0
* Forward bias:
Va ∆n(xp ) ∆p(xn ) Va ∆n(xp ) ∆p(xn ) ∆p(xn ) and ∆n(xp ) increase by several
(V) (cm−3 ) (cm−3 ) (V) (cm−3 ) (cm−3 ) orders of magnitude as Va is increased.
0 0 0 0 0 0
0.1 2.09 × 105 1.05 × 104 −0.1 −4.41 × 103 −2.20 × 102
0.2 1.02 × 107 5.08 × 105 −0.2 −4.50 × 103 −2.25 × 102
0.3 4.83 × 108 2.41 × 107 −0.5 −4.50 × 103 −2.25 × 102
0.6 5.18 × 1013 2.59 × 1012 −1 −4.50 × 103 −2.25 × 102
0.7 2.46 × 1015 1.23 × 1014 −2 −4.50 × 103 −2.25 × 102

n
p
n
np0 ∆n ∆p np0 p
pn0 pn0
0
* Forward bias:
Va ∆n(xp ) ∆p(xn ) Va ∆n(xp ) ∆p(xn ) ∆p(xn ) and ∆n(xp ) increase by several
(V) (cm−3 ) (cm−3 ) (V) (cm−3 ) (cm−3 ) orders of magnitude as Va is increased.
* Reverse bias:
0 0 0 0 0 0
∆p(xn ) ≈ −pn0 , ∆n(xp ) ≈ −np0 .
0.1 2.09 × 105 1.05 × 104 −0.1 −4.41 × 103 −2.20 × 102
0.2 1.02 × 107 5.08 × 105 −0.2 −4.50 × 103 −2.25 × 102
0.3 4.83 × 108 2.41 × 107 −0.5 −4.50 × 103 −2.25 × 102
0.6 5.18 × 1013 2.59 × 1012 −1 −4.50 × 103 −2.25 × 102
0.7 2.46 × 1015 1.23 × 1014 −2 −4.50 × 103 −2.25 × 102

pn junction under forward bias: simulation results
n
p
n
np0 ∆n ∆p np0 p
pn0 pn0
0
xj xj
6 2
(in 104 cm−3 ) Forward bias (in 103 cm−3 ) Reverse bias
np0
n
n p pn0
p
0 0
0 10 20 30 40 0 10 20 30 40
x (µm) x (µm)

pn junction under forward bias: simulation results
n
p
n
np0 ∆n ∆p np0 p
pn0 pn0
0
xj xj
6 2
(in 104 cm−3 ) Forward bias (in 103 cm−3 ) Reverse bias
np0
n
n p pn0
p
0 0
0 10 20 30 40 0 10 20 30 40
x (µm) x (µm)
* As we have seen earlier, the minority carrier diffusion lengths (i.e., Ln on the p-side, Lp on the n-side) are
typically much larger than the depletion width.
xn
High-injection regime
1.2
(in 1016 cm−3 )
1.0 n
0.8
1016
1014
Va = 0.2 V
1012
1010
108
106
104 p
2
10
xn xn
1.2
(in 1016 cm−3 )
1.0 n n
0.8
1016
1014
Va = 0.2 V Va = 0.6 V
1012
1010
108
106
104 p p
2
10
xn xn xn
1.2
(in 1016 cm−3 )
1.0 nn0
n n n
0.8 x
1016
1014
Va = 0.2 V Va = 0.6 V Va = 0.8 V
1012
1010
108
106
104 pn0
p p p
102 x
xn xn xn
1.2
(in 1016 cm−3 )
1.0 nn0
n n n
0.8 x
1016
1014
Va = 0.2 V Va = 0.6 V Va = 0.8 V
1012
1010
108
106
104 pn0
p p p
102 x
* As the forward bias is increased, the minority carrier concentration increases rapidly, and at some point becomes
comparable to the majority carrier concentration. This regime is called the “high-injection” regime.

xn xn xn
1.2
(in 1016 cm−3 )
1.0 nn0
n n n
0.8 x
1016
1014
Va = 0.2 V Va = 0.6 V Va = 0.8 V
1012
1010
108
106
104 pn0
p p p
102 x
* As the forward bias is increased, the minority carrier concentration increases rapidly, and at some point becomes
comparable to the majority carrier concentration. This regime is called the “high-injection” regime.
* In the high-injection regime, the majority carrier concentration also increases appreciably (e.g., ∆n ≈ ∆p on the
n side), and the overall charge neutrality is maintained in the neutral regions.
pn junction: current flow under forward bias
x − xn

Va
∆p(x) = pn0 exp − 1 exp − , x > xn .
VT Lp Va > 0 V p n
xp − x

Va
∆n(x) = np0 exp − 1 exp − , x < xp . Fndiff Fpdiff
VT Ln
Jdiff
n Jdiff
p
n p
∆p pn0
∆n
np0
0
Jdiff
p
Jdiff
n
0

x − xn

Va
∆p(x) = pn0 exp − 1 exp − , x > xn .
VT Lp Va > 0 V p n
xp − x

Va
VT Ln
Jdiff Jdiff
Note that, although Fndiff (for x < xp ) and Fpdiff (for x > xn ) are in n p
n p
opposite directions, Jndiff (for x < xp ) and Jpdiff (for x > xn ) are in
the same direction. ∆p pn0
∆n
np0
0
Jdiff
p
Jdiff
n
0

x − xn

Va
∆p(x) = pn0 exp − 1 exp − , x > xn .
VT Lp Va > 0 V p n
xp − x

Va
VT Ln
Jdiff Jdiff
Note that, although Fndiff (for x < xp ) and Fpdiff (for x > xn ) are in n p
n p
the same direction. ∆p pn0
∆n
np0
In particular, we are interested in Jndiff (xp ) and Jpdiff (xn ). 0
qDn np0 Va /VT
Jndiff (xp ) = e −1 ,
Ln
diff qDp pn0 Va /VT
Jdiff
Jp (xn ) = e −1 . Jdiff
p
Lp n
0

pn junction: current flow under reverse bias
x − xn

Va
∆p(x) = pn0 exp − 1 exp − , x > xn .
VT Lp Va < 0 V p n
xp − x

Va
∆n(x) = np0 exp − 1 exp − , x < xp .
VT Ln Fndiff Fpdiff
Jdiff
n Jdiff
p
p
n pn0
np0
0
Jdiff
n
Jdiff
p
0

x − xn

Va
∆p(x) = pn0 exp − 1 exp − , x > xn .
VT Lp Va < 0 V p n
xp − x

Va
∆n(x) = np0 exp − 1 exp − , x < xp .
VT Ln Fndiff Fpdiff
Note that, although Fndiff (for x < xp ) and Fpdiff (for x > xn ) are in Jdiff
n Jdiff
p
p
the same direction. n pn0
np0
0
Jdiff
n
Jdiff
p
0

x − xn

Va
∆p(x) = pn0 exp − 1 exp − , x > xn .
VT Lp Va < 0 V p n
xp − x

Va
∆n(x) = np0 exp − 1 exp − , x < xp .
VT Ln Fndiff Fpdiff
n Jdiff
p
p
np0
In particular, we are interested in Jndiff (xp )
and Jpdiff (xn ). 0
diff qDn np0 Va /VT qDn np0
Jn (xp ) = e −1 ≈− ,
Ln Ln
qDp pn0 qDp pn0
Jpdiff (xn ) = e Va /VT − 1 ≈ − .
Lp Lp Jdiff
n
Jdiff
p
0

x − xn

Va
∆p(x) = pn0 exp − 1 exp − , x > xn .
VT Lp Va < 0 V p n
xp − x

Va
∆n(x) = np0 exp − 1 exp − , x < xp .
VT Ln Fndiff Fpdiff
n Jdiff
p
p
np0
In particular, we are interested in Jndiff (xp )
and Jpdiff (xn ). 0
diff qDn np0 Va /VT qDn np0
Jn (xp ) = e −1 ≈− ,
Ln Ln
qDp pn0 qDp pn0
Jpdiff (xn ) = e Va /VT − 1 ≈ − .
Lp Lp Jdiff
n
Jdiff
p
The currents are much smaller under reverse bias. 0

pn junction: What is happening inside the depletion region?
Consider x in the depletion region, i.e., xp < x < xn . p

n
1 ψ(xp ) − ψ(x)
Z Z
p(x)
Jpdiff ≈ −Jpdrift → dψ = −VT dp → = exp .
p p(xp ) VT
pn0
np0

n
1 ψ(xp ) − ψ(x)
Z Z
p(x)
p p(xp ) VT
1 ψ(x) − ψ(xp )
Z Z
n(x)
Jndiff ≈ −Jndrift → dψ = +VT dn → = exp .
n n(xp ) VT
pn0
np0

n
1 ψ(xp ) − ψ(x)
Z Z
p(x)
p p(xp ) VT
1 ψ(x) − ψ(xp )
Z Z
n(x)
n n(xp ) VT
pn0
→ p(x)n(x) = p(xp )n(xp ) = pp0 np0 e Va /VT = ni2 e Va /VT np0

n
1 ψ(xp ) − ψ(x)
Z Z
p(x)
p p(xp ) VT
1 ψ(x) − ψ(xp )
Z Z
n(x)
n n(xp ) VT
pn0
If Va > 0 V, pn > ni2 in the depletion region; else, pn < ni2 .


n
1 ψ(xp ) − ψ(x)
Z Z
p(x)
p p(xp ) VT
1 ψ(x) − ψ(xp )
Z Z
n(x)
n n(xp ) VT
pn0

pn − ni2
R −G =
τn (n + n1 ) + τp (p + p1 )
→ we have a net recombination inside the depletion region if Va > 0 V, and a
net generation if Va < 0 V.

n
1 ψ(xp ) − ψ(x)
Z Z
p(x)
p p(xp ) VT
1 ψ(x) − ψ(xp )
Z Z
n(x)
n n(xp ) VT
pn0

xp xn
pn − ni2
R −G =
τn (n + n1 ) + τp (p + p1 ) 1 1
Jp (xp ) holes Jp (xn )
→ we have a net recombination inside the depletion region if Va > 0 V, and a q q
net generation if Va < 0 V.
recombination
1 1
− Jn (xp ) electrons − Jn (xn )
q q

p
n
pn0
np0
xp xn
1 1
q q
recombination
1 1
q q
p
n
pn0
np0
xp xn
1 1
q q
recombination
1 1
q q
* To obtain a first-order I -V model, we ignore G-R in the depletion region.

p p
n n
pn0 pn0
np0 np0
xp xn xp xn
1 1 1 1
Jp (xp ) holes Jp (xn ) Jp (xp ) holes Jp (xn )
q q q q
recombination
1 1 1 1
− Jn (xp ) electrons − Jn (xn ) − Jn (xp ) electrons − Jn (xn )
q q q q

p p
n n
p n
pn0 pn0
np0 np0
xp xn xp xn
Jp
1 1 1 1 Jn
Jp (xp ) holes Jp (xn ) Jp (xp ) holes Jp (xn )
q q q q
recombination
1 1 1 1
− Jn (xp ) electrons − Jn (xn ) − Jn (xp ) electrons − Jn (xn )
q q q q

pn junction: total current density
p n
Jp
Jn
p n
Jp
Jn
* The total current density is the same throughout the device.

p n
Jp
Jn

* If there is no G-R in the depletion region, we have J = Jn (xp ) + Jp (xn ).
p n p n
J
Jp Jp
Jn Jn

p n p n
J
Jp Jp
Jn Jn

* Using our earlier results for Jp (xn ) and Jn (xp ), we get

qDp pn0 qDn np0
e Va /VT − 1 .

J = Jp (xn ) + Jn (xp ) = +
Lp Ln
p n p n
J
Jp Jp
Jn Jn


qDp pn0 qDn np0
e Va /VT − 1 .

J = Jp (xn ) + Jn (xp ) = +
Lp Ln
* We can now obtain Jn (x > xn ) and Jp (x < xp ) using Jn (x) + Jp (x) = J.
p n p n
J
Jp Jp
Jn Jn


qDp pn0 qDn np0
e Va /VT − 1 .

J = Jp (xn ) + Jn (xp ) = +
Lp Ln
* We can now obtain Jn (x > xn ) and Jp (x < xp ) using Jn (x) + Jp (x) = J.

p n
J
Jp
Jn
Consider the situation sufficiently far from the depletion region (i.e., about 5Ln on the p-side
and 5Lp on the n-side).

p n
J
Jp
Jn
* The current density is due to majority carriers (drift component).

p n
J
Jp
Jn
* The current density is due to majority carriers (drift component).
* Since the majority carrier concentration is large, a very small electric field suffices to
produce the required current density (Jndrift = qnµn E, Jpdrift = qpµp E).

pn junction under forward bias: numerical results
p n
xj
J (normalised)
1 Doping densities:
Jp
(1) Na = Nd = 1016 cm−3
Jn (xp )
Jp (xn )
0 Jn
40 60 80 100
x (µm)
Na = Nd
(Parameters: Va = 0.5 V, µn = 1400 cm2 /V-s, µp = 500 cm2 /V-s, τn = 10 ns, τp = 10 ns, T = 300 K)
p n p n
xj xj
J (normalised) J
1 Doping densities:
Jp Jn
(1) Na = Nd = 1016 cm−3
Jp (xn )
Jn (xp ) (2) Na = 1017 cm−3
Jp (xn ) Nd = 1016 cm−3
Jn (xp )
0 Jn Jp
40 60 80 100 40 60 80 100
x (µm) x (µm)
Na = Nd Na = 10 × Nd
p n p n p n
xj xj xj
J (normalised) J J
1 Doping densities:
Jp Jn Jp
Jn (xp ) (1) Na = Nd = 1016 cm−3
Jp (xn )
Jn (xp ) (2) Na = 1017 cm−3
Jp (xn ) Jn Nd = 1016 cm−3
Jn (xp ) ∆J
0 Jn Jp (3) Na = 1016 cm−3
Jp (xn )
Nd = 1017 cm−3
40 60 80 100 40 60 80 100 40 60 80 100
x (µm) x (µm) x (µm)
Na = Nd Na = 10 × Nd Nd = 10 × Na
p n p n p n
xj xj xj
J (normalised) J J
1 Doping densities:
Jp Jn Jp
Jn (xp ) (1) Na = Nd = 1016 cm−3
Jp (xn )
Jn (xp ) (2) Na = 1017 cm−3
Jp (xn ) Jn Nd = 1016 cm−3
Jn (xp ) ∆J
0 Jn Jp (3) Na = 1016 cm−3
Jp (xn )
Nd = 1017 cm−3
40 60 80 100 40 60 80 100 40 60 80 100
Na = Nd Na = 10 × Nd Nd = 10 × Na
(Parameters: Va = 0.5 V, µn = 1400 cm2 /V-s,

µp = 500 cm2 /V-s,
τn = 10 ns, τp = 10 ns, T = 300 K)
qDp pn0 Va /VT qDn np0 Va /VT n2 n2 n2 n2
Jp (xn ) = e − 1 , Jn (xp ) = e − 1 , pn0 = i ≈ i , np0 = i ≈ i .
Lp Ln nn0 Nd pp0 Na
√
Jp (xn ) Dp Ln pn0 Dp Dn τn Na µp τn Na
r
* The ratio is = p = .
Jn (xp ) Dn Lp np0 Dn Dp τp Nd µ n τp N d

p n p n p n
xj xj xj
J (normalised) J J
1 Doping densities:
Jp Jn Jp
Jn (xp ) (1) Na = Nd = 1016 cm−3
Jp (xn )
Jn (xp ) (2) Na = 1017 cm−3
Jp (xn ) Jn Nd = 1016 cm−3
Jn (xp ) ∆J
0 Jn Jp (3) Na = 1016 cm−3
Jp (xn )
Nd = 1017 cm−3
40 60 80 100 40 60 80 100 40 60 80 100
Na = Nd Na = 10 × Nd Nd = 10 × Na
(Parameters: Va = 0.5 V, µn = 1400 cm2 /V-s,

µp = 500 cm2 /V-s,
τn = 10 ns, τp = 10 ns, T = 300 K)
qDp pn0 Va /VT qDn np0 Va /VT n2 n2 n2 n2
Jp (xn ) = e − 1 , Jn (xp ) = e − 1 , pn0 = i ≈ i , np0 = i ≈ i .
Lp Ln nn0 Nd pp0 Na
√
Jp (xn ) Dp Ln pn0 Dp Dn τn Na µp τn Na
r
* The ratio is = p = .
Jn (xp ) Dn Lp np0 Dn Dp τp Nd µ n τp N d
* Because of recombination, there is a change in Jp and Jn across the depletion region (which has been
ignored in our analysis). This change is seen as vertical lines in the figure since the depletion width is
much smaller than the diffusion lengths.
Diode I -V equation
I 100
area = A
10−2
I
p n
p n 10−4
I (A)
Va
10−6
Va 10−8

qDp pn0 qDn np0 10−10
e Va /VT − 1 .

J= + 0 0.2 0.4 0.6 0.8 1
Lp Ln Va (V)
100
80
60
I (mA)
40
20
0
−5 −4 −3 −2 −1 0 1
Va (V)
Diode I -V equation
I 100
area = A
10−2
I
p n
p n 10−4
I (A)
Va
10−6
Va 10−8

e Va /VT − 1 .

J= + 0 0.2 0.4 0.6 0.8 1
Lp Ln Va (V)
qDp pn0 qDn np0
→ I = A × J = Is e Va /VT

− 1 , with Is = A + . 100
Lp Ln
80
60
I (mA)
40
20
0
−5 −4 −3 −2 −1 0 1
Va (V)
Diode I -V equation
I 100
area = A
10−2
I
p n
p n 10−4
I (A)
Va
10−6
Va 10−8

e Va /VT − 1 .

J= + 0 0.2 0.4 0.6 0.8 1
Lp Ln Va (V)
qDp pn0 qDn np0

− 1 , with Is = A + . 100
Lp Ln
* This equation is known as the “Shockley diode equation.” 80
60
I (mA)
40
20
0
−5 −4 −3 −2 −1 0 1
Va (V)
Diode I -V equation
I 100
area = A
10−2
I
p n
p n 10−4
I (A)
Va
10−6
Va 10−8

e Va /VT − 1 .

J= + 0 0.2 0.4 0.6 0.8 1
Lp Ln Va (V)
qDp pn0 qDn np0

− 1 , with Is = A + . 100
Lp Ln
* This equation is known as the “Shockley diode equation.” 80
* Under reverse bias, with VR equal to a few VT or larger, 60
I (mA)
e Va /VT = e −VR /VT ≈ 0, and I ≈ −Is , i.e., the diode current
40
“saturates” (at −Is ). Is is therefore called the “reverse saturation
current.” 20
0
−5 −4 −3 −2 −1 0 1
Va (V)
Diode I -V equation
I 100
p n
10−2
Va

qDp pn0 qDn np0
e Va /VT − 1 . 10−4

J= +
Lp Ln
I (A)
10−6

qDp pn0 qDn np0
→ I = Is e Va /VT − 1 , with Is = A

+ .
Lp Ln
10−8
10−10
0 0.2 0.4 0.6 0.8 1
Va (V)
100
80
60
I (mA)
40
20
0
−5 −4 −3 −2 −1 0 1
Va (V)
Diode I -V equation
I 100
p n
10−2
Va

qDp pn0 qDn np0
e Va /VT − 1 . 10−4

J= +
Lp Ln
I (A)
10−6

qDp pn0 qDn np0

+ .
Lp Ln
10−8
* In a real diode, other factors often dominate in reverse bias,
including generation in the depletion region and surface leakage. 10−10
0 0.2 0.4 0.6 0.8 1
Also, as we will see, a real diode cannot withstand indefinitely large Va (V)
reverse voltages and will “break down” at some point.
100
80
60
I (mA)
40
20
0
−5 −4 −3 −2 −1 0 1
Va (V)
Diode I -V equation
I 100
p n
10−2
Va

qDp pn0 qDn np0
e Va /VT − 1 . 10−4

J= +
Lp Ln
I (A)
10−6

qDp pn0 qDn np0

+ .
Lp Ln
10−8
* In a real diode, other factors often dominate in reverse bias,
including generation in the depletion region and surface leakage. 10−10
0 0.2 0.4 0.6 0.8 1
Also, as we will see, a real diode cannot withstand indefinitely large Va (V)
reverse voltages and will “break down” at some point.
100
* Recombination in the depletion region under forward bias can be
80
incorporated in the Shockley equation with an “ideality factor” η
(1 < η < 2): 60
I (mA)

Va Va 40
I = Is1 exp + Is2 exp
η1 VT η2 VT
20
V a
≈ Iseff exp
ηVT 0
−5 −4 −3 −2 −1 0 1
Va (V)
Example
For an abrupt, uniformly doped silicon pn junction diode, Na = 1017 cm−3 , Nd = 2 × 1016 cm−3 ,
µn = 1500 cm2 /V-s, µp = 500 cm2 /V-s, τn = 2 µs, τp = 5 µs, A = 10−3 cm2 . Compute the
following for a forward bias of 0.65 V at T = 300 K:

Example
(1) depletion width W and the maximum electric field Em ,

Example
(2) n(xp ) and p(xn ),

Example
(3) Jn (xp ) and Jp (xn ),

Example
(4) the diode current I ,

Example
(5) electric fields in the neutral p and n regions, more than five diffusion lengths away from
the depletion region,

Example
(5) electric fields in the neutral p and n regions, more than five diffusion lengths away from
the depletion region,
(6) the reverse saturation current Is .

Example (continued)
xj
p n
!
Na Nd (1017 )(2 × 1016 ) q (Vbi − Va )
Vbi = VT log = (0.0259 V) log = 0.77 V . Ec ∆n
ni2 (1.5 × 1010 )2 EFn
EFp
∆p
Ev
qNd
−qNa
0
E
−Em
ψ (Vbi − Va )

Example (continued)
xj
p n
!
Na Nd (1017 )(2 × 1016 ) q (Vbi − Va )
ni2 (1.5 × 1010 )2 EFn
EFp
∆p
Ev
s
2 Na + Nd
W = (Vbi − Va )
q Na Nd qNd
−qNa
0
E
−Em
ψ (Vbi − Va )

Example (continued)
xj
p n
!
Na Nd (1017 )(2 × 1016 ) q (Vbi − Va )
ni2 (1.5 × 1010 )2 EFn
EFp
∆p
Ev
s
2 Na + Nd
W = (Vbi − Va )
q Na Nd qNd
s
2 × 11.8 × 8.85 × 10−14 1.2 × 1017 ρ
= × 0.12 cm = 0.097 µm.
1.6 × 10−19 2 × 1033
0
−qNa
0
E
−Em
ψ (Vbi − Va )

Example (continued)
xj
p n
!
Na Nd (1017 )(2 × 1016 ) q (Vbi − Va )
ni2 (1.5 × 1010 )2 EFn
EFp
∆p
Ev
s
2 Na + Nd
W = (Vbi − Va )
q Na Nd qNd
s
2 × 11.8 × 8.85 × 10−14 1.2 × 1017 ρ
= × 0.12 cm = 0.097 µm.
1.6 × 10−19 2 × 1033
0
1
(Vbi − Va ) = Em W −qNa
2
0
E
−Em
ψ (Vbi − Va )

Example (continued)
xj
p n
!
Na Nd (1017 )(2 × 1016 ) q (Vbi − Va )
ni2 (1.5 × 1010 )2 EFn
EFp
∆p
Ev
s
2 Na + Nd
W = (Vbi − Va )
q Na Nd qNd
s
2 × 11.8 × 8.85 × 10−14 1.2 × 1017 ρ
= × 0.12 cm = 0.097 µm.
1.6 × 10−19 2 × 1033
0
1
(Vbi − Va ) = Em W −qNa
2
2 (Vbi − Va ) 2 × 0.12 V 0
→ Em = = = 25 kV/cm.
W 0.097 × 10−4 cm E
−Em
ψ (Vbi − Va )

Example (continued)
xp xn
p n
p
pp0 n
nn0
p
n
pn0
np0

Example (continued)
xp xn
p n
The equilibrium minority carrier densities are
p
ni2 ni2 (1.5 ×1010 )2 pp0 n
pn0 = ≈ = = 1.125 × 104 cm−3 , nn0
nn0 Nd 2 × 1016
p
n
pn0
np0

Example (continued)
xp xn
p n
p
ni2 ni2 (1.5 ×1010 )2 pp0 n
pn0 = ≈ = = 1.125 × 104 cm−3 , nn0
nn0 Nd 2 × 1016
ni2 ni2 (1.5 ×1010 )2
np0 = ≈ = = 2.25 × 103 cm−3 .
pp0 Na 1 × 1017
p
n
pn0
np0

Example (continued)
xp xn
p n
p
ni2 ni2 (1.5 ×1010 )2 pp0 n
pn0 = ≈ = = 1.125 × 104 cm−3 , nn0
nn0 Nd 2 × 1016
ni2 ni2 (1.5 ×1010 )2
np0 = ≈ = = 2.25 × 103 cm−3 .
pp0 Na 1 × 1017
The minority carrier densities at xp and xn are
n(xp ) = np0 e Va /VT = 2.25 × 103 × e 0.65/0.0259 = 1.8 × 1014 cm−3 , p

n
pn0
np0

Example (continued)
xp xn
p n
p
ni2 ni2 (1.5 ×1010 )2 pp0 n
pn0 = ≈ = = 1.125 × 104 cm−3 , nn0
nn0 Nd 2 × 1016
ni2 ni2 (1.5 ×1010 )2
np0 = ≈ = = 2.25 × 103 cm−3 .
pp0 Na 1 × 1017
The minority carrier densities at xp and xn are
n(xp ) = np0 e Va /VT = 2.25 × 103 × e 0.65/0.0259 = 1.8 × 1014 cm−3 , p

n
p(xn ) = pn0 e Va /VT = 1.125 × 104 × e 0.65/0.0259 = 8.9 × 1014 cm−3 . pn0
np0

Example (continued)
The diffusion coefficients are

Dp = VT µp = 0.0259 × 500 = 12.9 cm2 /s,
Dn = VT µn = 0.0259 × 1500 = 38.7 cm2 /s. Va > 0 V p n
The minority carrier diffusion lengths in the neutral regions are

p √
Lp = Dp τp = 12.9 × 5 × 10−6 cm = 80.3 µm,
p
√ √ n
Ln = Dn τn = 38.7 × 2 × 10−6 cm = 88 µm.
∆p pn0
∆n
np0

Example (continued)
The minority carrier current densities at xn and xp are

qDp pn0 Va /VT
Jp (xn ) = e −1 p n
Lp
J
Jp
Jn

Example (continued)

qDp pn0 Va /VT
Jp (xn ) = e −1 p n
Lp
1.6 × 10−19× 12.9 × 1.125 × 104 J
= × 8.12 × 1010
80.3 × 10−4
Jp
Jn

Example (continued)

qDp pn0 Va /VT
Jp (xn ) = e −1 p n
Lp
1.6 × 10−19× 12.9 × 1.125 × 104 J
= × 8.12 × 1010
80.3 × 10−4
= 0.235 A/cm2 .
Jp
Jn

Example (continued)

qDp pn0 Va /VT
Jp (xn ) = e −1 p n
Lp
1.6 × 10−19× 12.9 × 1.125 × 104 J
= × 8.12 × 1010
80.3 × 10−4
= 0.235 A/cm2 .
qDn np0 Va /VT

Jn (xp ) = e −1 Jp
Ln Jn
1.6 × 10−19 × 38.7 × 2.25 × 103
= × 8.12 × 1010
88 × 10−4
= 0.13 A/cm2 .

Example (continued)

qDp pn0 Va /VT
Jp (xn ) = e −1 p n
Lp
1.6 × 10−19× 12.9 × 1.125 × 104 J
= × 8.12 × 1010
80.3 × 10−4
= 0.235 A/cm2 .
qDn np0 Va /VT

Jn (xp ) = e −1 Jp
Ln Jn
1.6 × 10−19 × 38.7 × 2.25 × 103
= × 8.12 × 1010
88 × 10−4
= 0.13 A/cm2 .
The diode current I is
I = A (Jp (xn ) + Jn (xp ))

Example (continued)

qDp pn0 Va /VT
Jp (xn ) = e −1 p n
Lp
1.6 × 10−19× 12.9 × 1.125 × 104 J
= × 8.12 × 1010
80.3 × 10−4
= 0.235 A/cm2 .
qDn np0 Va /VT

Jn (xp ) = e −1 Jp
Ln Jn
1.6 × 10−19 × 38.7 × 2.25 × 103
= × 8.12 × 1010
88 × 10−4
= 0.13 A/cm2 .
I = A (Jp (xn ) + Jn (xp ))

= 10−3 cm2 × (0.235 + 0.13) A/cm2

Example (continued)

qDp pn0 Va /VT
Jp (xn ) = e −1 p n
Lp
1.6 × 10−19× 12.9 × 1.125 × 104 J
= × 8.12 × 1010
80.3 × 10−4
= 0.235 A/cm2 .
qDn np0 Va /VT

Jn (xp ) = e −1 Jp
Ln Jn
1.6 × 10−19 × 38.7 × 2.25 × 103
= × 8.12 × 1010
88 × 10−4
= 0.13 A/cm2 .
I = A (Jp (xn ) + Jn (xp ))

= 10−3 cm2 × (0.235 + 0.13) A/cm2
= 0.365 mA.

Example (continued)
In the neutral n region more than 5Lp away from the depletion region,
n
J ≈ Jn = qµn Eneutral nn0 , leading to
p n
A

0.365
n J cm2 J
Eneutral = = 2
cm 1

qµn nn0
1.6 × 10−19 [C] × 1500 × 2 × 1016
V-s cm3
Jp
Jn

Example (continued)
n
p n
A

0.365
n J cm2 J
Eneutral = = 2
cm 1

qµn nn0
1.6 × 10−19 [C] × 1500 × 2 × 1016
V-s cm3
= 0.076 V/cm.
Jp
Jn

Example (continued)
n
p n
A

0.365
n J cm2 J
Eneutral = = 2
cm 1

qµn nn0
1.6 × 10−19 [C] × 1500 × 2 × 1016
V-s cm3
= 0.076 V/cm.
Similarly, Jp
Jn
A

0.365
p J cm2
Eneutral = = 2
cm 1

qµp pp0
1.6 × 10−19 [C] × 500 × 1 × 1017
V-s cm3

Example (continued)
n
p n
A

0.365
n J cm2 J
Eneutral = = 2
cm 1

qµn nn0
1.6 × 10−19 [C] × 1500 × 2 × 1016
V-s cm3
= 0.076 V/cm.
Similarly, Jp
Jn
A

0.365
p J cm2
Eneutral = = 2
cm 1

qµp pp0
1.6 × 10−19 [C] × 500 × 1 × 1017
V-s cm3
= 0.045 V/cm.

Example (continued)
n
p n
A

0.365
n J cm2 J
Eneutral = = 2
cm 1

qµn nn0
1.6 × 10−19 [C] × 1500 × 2 × 1016
V-s cm3
= 0.076 V/cm.
Similarly, Jp
Jn
A

0.365
p J cm2
Eneutral = = 2
cm 1

qµp pp0
1.6 × 10−19 [C] × 500 × 1 × 1017
V-s cm3
= 0.045 V/cm.
Note that these values are much smaller than Em in the depletion region
(25 kV/cm).

Example (continued)
The reverse saturation current Is is given by

qDp pn0 qDn np0
Is = A × +
Lp Ln

Example (continued)

qDp pn0 qDn np0
Is = A × +
Lp Ln
1.6 × 10−19 × 12.9 × 1.125 × 104 1.6 × 10−19 × 38.7 × 2.25 × 103

= 10−3 +
80.3 × 10−4 88 × 10−4

Example (continued)

qDp pn0 qDn np0
Is = A × +
Lp Ln
1.6 × 10−19 × 12.9 × 1.125 × 104 1.6 × 10−19 × 38.7 × 2.25 × 103

= 10−3 +
80.3 × 10−4 88 × 10−4
= 4.5 × 10−15 A (i.e., 4.5 fA).

Example (continued)

qDp pn0 qDn np0
Is = A × +
Lp Ln
1.6 × 10−19 × 12.9 × 1.125 × 104 1.6 × 10−19 × 38.7 × 2.25 × 103

= 10−3 +
80.3 × 10−4 88 × 10−4
= 4.5 × 10−15 A (i.e., 4.5 fA).
Note how small Is is. The only reason we can get significant currents (∼ mA)
in forward bias is the huge exponential factor (e Va /VT ) in the Shockley
equation.

Forward bias: summary
xj
xj
xp xn 0.8
p Jtotal
pp0 n q (Vbi − Va )
nn0 Ec
Jp EFn
EFp
Ev
p n
2
0 (in A/cm )
p 0.8 ρ
n Jtotal
pn0 0
np0
(Note : log scale)

Jn
0
0 (in A/cm2 )
E
5 10 15 20 25 30 −Em
x (µm)
ψ (Vbi − Va )

xj
xj
xp xn 0.8
p Jtotal
nn0 Ec
Jp EFn
EFp
Ev
p n
2
0 (in A/cm )
p 0.8 ρ
n Jtotal
pn0 0
np0
(Note : log scale)
* On the p side, 5 Ln away from xp , J ≈ Jpdrift Jn

0
resulting from a small (positive) electric 0 (in A/cm2 )
E
field. 5 10 15 20 25 30 −Em
x (µm)
ψ (Vbi − Va )

xj
xj
xp xn 0.8
p Jtotal
nn0 Ec
Jp EFn
EFp
Ev
p n
2
0 (in A/cm )
p 0.8 ρ
n Jtotal
pn0 0
np0
(Note : log scale)

0
E
field. 5 10 15 20 25 30 −Em
x (µm)
* Holes continuously injected from the p side
recombine with electrons on the n side.
ψ (Vbi − Va )

xj
xj
xp xn 0.8
p Jtotal
nn0 Ec
Jp EFn
EFp
Ev
p n
2
0 (in A/cm )
p 0.8 ρ
n Jtotal
pn0 0
np0
(Note : log scale)

0
E
field. 5 10 15 20 25 30 −Em
x (µm)
* Holes continuously injected from the p side
recombine with electrons on the n side.
* 5 Lp away from xn , there are no excess holes, ψ (Vbi − Va )
and p ≈ pn0 .

xj
xj
xp xn 0.8
p Jtotal
nn0 Ec
Jp EFn
EFp
Ev
p n
2
0 (in A/cm )
p 0.8 ρ
n Jtotal
pn0 0
np0
(Note : log scale)

Jn
0
0 (in A/cm2 )
E
5 10 15 20 25 30 −Em
x (µm)
ψ (Vbi − Va )

xj
xj
xp xn 0.8
p Jtotal
nn0 Ec
Jp EFn
EFp
Ev
p n
2
0 (in A/cm )
p 0.8 ρ
n Jtotal
pn0 0
np0
(Note : log scale)
* On the n side, 5 Lp away from xn , J ≈ Jndrift Jn

0
E
field. 5 10 15 20 25 30 −Em
x (µm)
ψ (Vbi − Va )

xj
xj
xp xn 0.8
p Jtotal
nn0 Ec
Jp EFn
EFp
Ev
p n
2
0 (in A/cm )
p 0.8 ρ
n Jtotal
pn0 0
np0
(Note : log scale)

0
E
field. 5 10 15 20 25 30 −Em
x (µm)
* Electrons continuously injected from the n
side recombine with holes on the p side.
ψ (Vbi − Va )

xj
xj
xp xn 0.8
p Jtotal
nn0 Ec
Jp EFn
EFp
Ev
p n
2
0 (in A/cm )
p 0.8 ρ
n Jtotal
pn0 0
np0
(Note : log scale)

0
E
field. 5 10 15 20 25 30 −Em
x (µm)
* Electrons continuously injected from the n
side recombine with holes on the p side.
* 5 Ln away from xp , there are no excess ψ (Vbi − Va )
electrons, and n ≈ np0 .

sd_pn_3

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

sd_pn_3

Uploaded by

Copyright:

Available Formats

Semiconductor Devices

p-n Junctions: Part 3

Department of Electrical Engineering

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

M. B. Patil, IIT Bombay

p contact p neutral depletion n neutral n contact

Note that Lp  W |0 V (0.33 µm), a typical situation.

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

ni2 (1.5 × 1010 )2

M. B. Patil, IIT Bombay

ni2 (1.5 × 1010 )2

M. B. Patil, IIT Bombay

ni2 (1.5 × 1010 )2

M. B. Patil, IIT Bombay

Va ∆n(xp ) ∆p(xn ) Va ∆n(xp ) ∆p(xn )

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

M. B. Patil, IIT Bombay

Consider x in the depletion region, i.e., xp < x < xn . p

Consider x in the depletion region, i.e., xp < x < xn . p

Consider x in the depletion region, i.e., xp < x < xn . p

Consider x in the depletion region, i.e., xp < x < xn . p

If Va > 0 V, pn > ni2 in the depletion region; else, pn < ni2 .

Consider x in the depletion region, i.e., xp < x < xn . p

If Va > 0 V, pn > ni2 in the depletion region; else, pn < ni2 .

Consider x in the depletion region, i.e., xp < x < xn . p

If Va > 0 V, pn > ni2 in the depletion region; else, pn < ni2 .

M. B. Patil, IIT Bombay

Note that Lp W |0 V (0.33 µm), a typical situation.

qDn np0 Va /VT

qDn np0 Va /VT

qDn np0 Va /VT

qDn np0 Va /VT