Part 3

PHM122S

By
Wael Fikry
Professor of Solid State Electronics
Eng. Physics Dept. – Faculty of Eng.
Ain Shams University
[email protected] - Room 223

5/29/2021 PHM122_Part 3 1
Outline
pn Junction at Thermal Equilibrium
 Introduction
 The built-in potential
 The depletion region (space charge region)
 The electric field in the space charge region
 Energy band diagram

pn Junction Diode
 Biasing of pn Junctions
 I-V Characteristics
 Junction Breakdown
 Zener Breakdown
 Avalanche Breakdown

 Junction Capacitances

5/29/2021 PHM122_Part 3 2
 pn Junction at Thermal Equilibrium
p-TYPE n-TYPE
Introduction
e e e e
The p-n junction is a metallurgical junction which can h
-
h
-
h
-
h
-
+
e
+
e
+
e
+
e
+ + + +
be formed by diffusion or ion implantation techniques h h h h +
e
+
e
+
e
+
e
- - - - e e e e
+ + + +
Imagine that the uniformly doped n-and p-type
semiconductor materials have just joined together h
NA- e ND+
 The larger carrier concentration gradients at the h h + +
e
+
e
+
e
- - - - e e e
junction cause carrier diffusion: + +
e
+
e
+
e
h h + + + +
 Holes will tend to diffuse right (Jpdiff ) leaving fixed - - - - + +
e
+
e
+
e

negative ions (NA-) near the junction 

 Electrons will tend to diffuse left (Jndiff ) leaving w
fixed positive ions (ND+) near the junction Fig.(1)
- Vbi +
 An electric field ( ) will be established from ND+ to
NA- across certain width (w) and tries to drift Jndiff
 Holes back into the p-region (Jpdrift ) Jndrift

 Electrons back into the n-type (Jndrift) Jpdiff

Jpdrift
pn_intro
 A built-in potential (Vbi) will appear across the width w

5/29/2021 PHM122_Part 3 3
At thermal equilibrium, the net current flow
Jt  Jn  J p  0
across the junction is Zero
For each type of carrier the drift dp ( x )
J P ( x )  q  p p ( x )  ( x )  qD p 0 (1)
current due to the electric field must dx
exactly cancel the diffusion current dn( x)
J n ( x)  q  n n( x)  ( x)  qDn 0 (2)
due to the concentration gradient dx

The Built-in Potential Vbi

To obtain the relation between Vbi and the doping concentration on each side of the
junction, we will use Eq. (1)
dp( x)
q p p( x)  ( x )  qD p 0
dx
p 1 dp( x)
 ( x)  (3)
Dp p ( x) dx
dV ( x ) p q
  ( x)   &  Einstein relation
dx Dp kT
q dV ( x ) 1 dp ( x ) q dp ( x )
Eq.(3) becomes    dV ( x )  (4)
kT dx p ( x ) dx kT p( x)

5/29/2021 PHM122_Part 3 4
 ppo NA nno ND

V bi p no npo= ni2/ppo pno= ni2/nno

q dp ( x )
By integrating Eq.(4) 
kT  dV ( x )   p( x)
= ni 2 / N A
p-type
= ni 2 / N D
n-type
0 p po p, n - Vbi +
q
 V bi  0   ln p no  ln p po 
kT
nno
p po ppo
  ln
p no

kT p po npo
 V bi  ln (5)
q p no w pno
x
Since ppo NA & pno= ni2/ ND
Fig.(2)
kT N A N D
V bi  ln 2 (6)
q ni
This relation will
Note, at thermal equilibrium condition ppo npo = pno nno = ni2 be very valuable
in calculation of
p po the I-V
n no
From Eq.(5)   e qVbi kT
(7) characteristics of
p no n po the junction

5/29/2021 PHM122_Part 3 5
The depletion region (space charge region)
p-type +++ n-type
------ +++
It is the region around the metallurgical junction where Neutral
region
------ +++
Neutral
region
recombination of electrons and holes depleted this from +++

its large number of equilibrium mobile majority carriers metallurgical junction

Fig.(3a)
• For ABRUPT junction, the doping impurities are assumed
to change suddenly from ND to NA as shown in Fig.(3b)
qND
• The width of the narrow regions between the two neutral +
regions and the depletion region can be neglected x
-
-qNA
• Neglecting mobile carrier densities in the space charge
region is well known by depletion region approximation. Fig.(3b)

pn0  xpo  x  xno Charge density

where xno and xpo denote the depletion layer widths of n and p sides qND
+
• From space charge neutrality : q N A x po A  q N D x no A -xpo
xno x
where A is the cross-section area of the sample -
-qNA
Depletion
 N A x po  N D x no (8) region
Fig.(3c)

Remarks: the depletion region extends farther into the side with lighter doping.

5/29/2021 PHM122_Part 3 6
The electric field distribution in the space charge region Charge density

The variation of the electric field across the depletion qND

of an abrupt pn junction is shown in Fig.(4b) +
-xpo
 The negative value of the electric field means it
xno x
is in –x direction -
-qNA Fig.(4a)
 The maximum value of the electric field (o) is
at the metallurgical junction (x=0) Depletion region
w
 To calculate the electric field distribution, we begin
with Poisson’s equation in conjunction with the 
depletion approximation
  qN D -xpo xno
  0  x  xno (9) x
x  
   qN A Fig.(4b)
   x po  x  0 (10)
x   o
By integrating Eqs.(9 & 10)
qN D x no
0
qN D
 d  x dx   x   x  xno  0  x  x no (11)
  x  
  x x
 qN A qN A
 d 
  dx   x  

x  x po  x po  x  0 (12)
0  x po

5/29/2021 PHM122_Part 3 7
At x=0, the two expressions (11&12) reach maximum value of the electric field
qN D qN A 
o   xno   x po (13)
 
From Eq.(13) o can be related to depletion width w
(w = xno+ xpo ) by -xpo xno
x
q ND N A
o  w
 ND  N A (14) Fig.(5a)
o
 Electric field and built-in potential relationship
x no
dV ( x )
  ( x)   or V bi     ( x ) dx
dx  x po
From Eqs.(11&12) V
0
qN A x no qN
V bi  [  

 x  x po  dx  

D
 x  xno  dx]
 x po 0 Vbi
qN A 2 qN D 2 1
x no   o  x po  xno 
xno
 x po  -xpo x
2 2 2 Fig.(5b)
1
  ow (15)
2

5/29/2021 PHM122_Part 3 8
 Depletion width and built-in potential relationship
1 q ND N A 2
From Eq.(15) and Eq.(14) V bi  w
2  ND  NA
1

 2  N  N   2

w
q  N N  
A D
V bi
(16)
    A D
Recall
 Depletion width and doping concentration relationship kT N N
Vbi  ln A 2 D
q ni
1

 2 kT  N  N   N N   2

w
q  N N   n
2 ln
A

D A
2
D
(17)
    A 
D i

NA xpo = ND xno
One sided abrupt (step) junction ( p+n or n+p)
Assume p+n as an example, this means NA >> ND then xno >> xpo or w  xno
All depletion width almost in the lighter doping region

5/29/2021 PHM122_Part 3 9
Energy Band Diagram
p-type n-type
• At thermal equilibrium, Fermi level must
remain constant across the junction Ecp
Efn
Ecn

• The electrostatic field due the fixed Eip Ein

charges in the space-charge region Efp

Evp Evn
causes the energy band to bend as shown
in the figure
• A built-in voltage Vbi thus exist 
across the depletion region. Ecp

Electron energy
qVbi
• Since the intrinsic Fermi level Ei always Eip
e
follow the energy band, we can write: Ef
Ecn

1 dEi Evp
 h Ein
q dx

hole energy
qVbi
 Electrons in the bottom of n-type semiconductors Evn
conduction band face a potential barrier of
height qVbi
 holes in the top of p-type semiconductors valance
band face a potential barrier of height qVbi

5/29/2021 PHM122_Part 3 10
 Example
Consider an abrupt silicon pn junction at room temperature (300 K) in which the acceptor
concentration is 1018 cm-3 and the donor concentration is 5 1015 cm-3. The junction has a
circular cross section with a diameter of 10 m.
(a) Calculate the Fermi level positions in the p and n neutral regions.
(b) Draw an equilibrium band diagram for the junction and determine the built-in potential
Vbi from the diagram.
(c) Compare the results of part (b) with built-in potential Vbi as calculated from Eq.(6).
(d) Find the depletion width in both sides of the junction.
(e) Calculate the uncompensated charge on either side of the junction.
(f) Find the maximum electric field.
(g) Sketch charge density ρ(x) and (x), Comment on these graphs.
relative permitivity of silicon ( r )  11.8  o  8.854  1014 F/ cm k  8.62  105 eV/ K

(a) (b) Ecp

 ND   5  1015  qVbi
EF  Ein  kT ln    0.0259 ln  10 
 0.329eV Eip
 ni  1.5  10  0.467 Ecn
EF
Evp 0.329
 NA   1018  Ein
Ei p  EF  kT ln    0.0259 ln  10 
 0.467 eV
 ni  1.5 10 
Evn
qVbi = 0.467 +0.329 = 0.796 eV

5/29/2021 PHM122_Part 3 11
  18 15 
kT  N A N D  10  5  10
(c) Vbi  ln    0.0259 ln    0.796 V the same value as (b )
 1.5  10  
q 2
 ni  10 2

1

 2  N  N   2
 2  11.8  8.854  1014  (1018  5  1015 )  0.796 
0.5

w
q  N N  
(d) A
V
D
 
bi
 1.6  1019  1018  5 1015 
   A D

 4.571 105 cm  0.457  m

x po N D x po x ND
 N A x po  N d x no     po 
x no N A x no  x po w NA ND

 ND   5  1015  4.571  10 5 
x po   w     2.274  10 7 cm  0.002 μm
 N A  N D   (10  5  10 ) 
18 15

 NA   1018  4.571  10 5  5
xno   w     4.548  10 cm  0.455 μm
 N A  N D   (10  5  10 ) 
18 15

5/29/2021 PHM122_Part 3 12
(e) From space charge neutrality
uncompensated charge on either side  q N D x no A  q N A x po A
 1.6 1019  5 1015  4.548 105  7.854 105
 2.868 1014 C

qN D xno 1.6  10 19  5  1015  4.548  10 5

(f) From Eq.(13) 
o     3.48  10 4 V/ cm
 r o 11.8  8.85  10 14

𝝆 (x) (x)
(g)
5  1015  q
-xpo 0.5
-xpo xno 0.5
x (𝝁m) x (𝝁m)
xno

1018  q 3.84 104

Note how in regions where the doping density is high the width of the depletion region
is small while when the doping density is low the width of the depletion region is large
[one sided abrupt (step) junction]

5/29/2021 PHM122_Part 3 13
Summary of pn Junction at
Thermal Equilibruim

qND
q N A x po A  q N D x no A
-xpo +
 N A x po  N D x no - xno x
-qNA

qN D
  x   x  xno  0  x  x no -xpo xno
 x
qN A
  x  

x x  po  xpo  x  0 o
V

kT N A N D
V bi  ln 2 Vbi
q ni
-xpo xno x
Ecp
1

 2 kT  N  N   N N   2
Eip
Ecn
w
q  N N  ln  n  
A D A D Ef
2
2 Evp
     A D i
Ein

Evn

5/29/2021 PHM122_Part 3 14
 pn Junction Diode
Biasing of pn Junctions
When an EXTERNAL voltage is applied across the pn junction this voltage is mainly
dropped across the width of the depletion region (neglect the resistance of neutral regions)

To understand the influence of the applied voltage on the electrostatics we simply replace
the built-in voltage with the NET VOLTAGE (Vj) in our electrostatic analysis
We distinguish TWO different regime of biasing
 In FORWARD bias mode the external voltage OPPOSES the built-in voltage
 In REVERSE bias mode the external voltage ADDS to the built-in voltage

FORWARD BIAS NO BIAS REVERSE BIAS

Thermal equilibrium
Vf VR

- Vbi +

p n
p n p n

Vj = Vbi
Vj = Vbi - Vf Vj = Vbi + VR

5/29/2021 PHM122_Part 3 15
 The WIDTH of the depletion field and the ELECTRIC FIELD that develops across it will
therefore be REDUCED by the forward biasing and vise versa by the reverse biasing
 Note: xn and xp mark the limits of the space-charge region when the diode is biased

Forward Bias Thermal equilibrium Reverse Bias

Metallurgical
junction

p n p n p n
-xp xn -xpo 0 xno -xp xn
w w w
  
-xp xn -xpo xno -xp xn
x x x
o
o
o
V
V V
Vbi + VR
Vbi - Vf Vbi
-xp xn x -xpo xno x -xp xn x

5/29/2021 PHM122_Part 3 16
 Under forward bias and due to the reduction of the electric field we have an enhanced hole
diffusion from the p-side to the n-side and electron diffusion from the n-side to the p-side.
 minority carrier injections occur
 The drift current is reduced in comparison to the diffusion current

 Under reverse bias and due to the increase of the electric field minority carrier densities
decrease below their equilibrium value from the p-side to the n-side
 minority carrier extraction occur
 The diffusion current is reduced greatly resulting in a small reverse saturation current

Forward Bias Thermal equilibrium Reverse Bias

Vbi - Vf Low level

- + injection - Vbi + - Vbi + VR +
  
nn nno nn
pp ppo pp

np
pn npo npo
pno pno
w pno w
w np pn
x x x
The only current which flows is due to drift of minority
carriers across the junction. This current is fed by diffusion
of minority carriers toward junction

5/29/2021 PHM122_Part 3 17
 Forward Bias Thermal equilibrium Reverse Bias

Ecp Ecp Ecp
Ecn
Eip Eip Eip
Efn
Efp Ef Ecn
Ein Efp
Evp Evp
Ein Evp
Ecn
Evn Efn
Evn Ein

Evn
J f (  ) = J n ( ) + J p ( ) Jn = Jp = 0 JR () = Jn () + Jp ()
Large current No current Small current

 With a voltage applied across the junction it is no longer at thermal equilibrium and we
may NOT associate a SINGLE Fermi level with the junction

 Instead we introduce the concept of QUASI-FERMI LEVELS to help us determine

the carrier concentrations in the different regions of the device
 The choice of these Fermi levels is motivated by the requirement that the Fermi level
AWAY from the depletion region should not change
 The difference in the quasi-Fermi levels is therefore determined by the applied voltage

5/29/2021 PHM122_Part 3 18
Current –Voltage (I-V) Characteristics
The IDEAL I-V characteristics will be derived based on the following assumptions:
1. Abrupt junction under steady state conditions
2. Depletion region approximation (no mobile charges in depletion region)
3. Low level injection conditions (no change in majority carrier concentrations in neutral regions)
4. Neither generation nor recombination current exists in the depletion region
5. Neglect the resistance or voltage drop on neutral regions (no electric field in neutral region)

Rearranging Eq.(7), minority carrier concentrations in the neutral regions are given by:

pno  p po e  qVbi kT
(18) n po  nno e  qVbi kT
(19)
When a forward bias is applied, the potential difference is reduced to Vbi – Vf ; but when
a reverse bias is applied, the potential difference is increased to Vbi + VR.

Thus, Eq. (18) is modified to pn  p p e q (Vbi V ) kT (20)

where pn and np are the non-equilibrium hole densities AT THE BOUNDARIES of the
depletion region in the n- and p-sides, respectively, with V positive for forward bias and
negative for reverse bias.

5/29/2021 PHM122_Part 3 19
  q ( V bi  V )
Under low level injection, p p  p po , Eq. 20 becomes p n  p po e
kT
(21)

Using (18) and (21), the excess holes under bias in the n-neutral region at the edge of the
depletion layer is
 p  p n  p no  p po e  q (Vbi V ) kT
 p po e  qVbi kT
 p po e  qVbi kT
e qV kT
 1

 p  p n  p no  p no  e qV KT
 1 (22)

According to the ideal diode assumptions, the continuity equation for holes in the n-region
pn   x  p n  x   2 pn  x   pn  pno 
   p pn  x    p  x   Dp   G   nn  nno
t x x x 2  L p
 
pp  ppo
2 pn ( x) pn  pno
reduces to Dp  0
x 2
p
p-TYPE np n-TYPE

 2 pn ( x) pn  pno pn
or  0 (23)
x 2 2
Lp npo p
pn(x)

where Lp  Dp p is the diffusion length pno

-xp 0 xn x

5/29/2021 PHM122_Part 3 20
 nn  nno
The solution of Eq.(23) with boundary conditions
pn(xn) = pn = pno + p and pn( ) = pno is pp  ppo

pn  x   pno   pn  pno  e
 ( x  xn ) L p
(24)
p-neutral n-neutral
region np region
From Eq. (23) and Eq. (24)
pn  x   pno  pno  e qV  1 e
kT  ( x  xn ) L p pn
  x  xn (25)
pn(x)
npo p
The hole diffusion current density in the n-neutral region
pno
dpn ( x) qDp pno qV
J p  x   qDp
dx

Lp
 e kT
1 e
( x  xn ) Lp
(26)
Jn,p
-xp 0 xn
x

diffusion due to
The hole diffusion current density at (x = xn) recombination as the
injected holes get deeper
in the n part
qD p pno
J p  xn  
Lp
e qV kT
 1 (27) Jn(-xp)

Jp(xn)
Similarly, the electron diffusion current density
at (x = -xp) Jn(x)
Jp(x) [Eq.(26)]
qDn n po qV kT
Jn   xp  
Ln
 e  1 (28) x
-xp 0 xn

5/29/2021 PHM122_Part 3 21
The depletion region was assumed to have no generation or recombination. Therefore,
the current through it is a constant-"what goes in must come out."
 the minority carrier diffusion currents are known at the edges of the depletion region,
then the constant depletion currents are known.
J n (depl.)  J n   x p  J p (depl.)  J p  xn  Jn,p

J  constant  J p  x   J n  x 
Since the device has only two terminals, the
total current (J) through the diode must be a Jn(-xp)
Jp(x) Jn(x)
constant at each point (x). {majority} {majority}

J  constant  J p  x   J n  x 
p-neutral n-neutral
(29) region region
Jp(xn)

At x  xn J  J p  xn   J n  xn   J p  xn   J n   x p  Jn(x)
{minority} Jp(x) {minority}
x
The constant total current density throughout the -xp 0 xn
device is the sum of Eqs.(27) and (28)
As the minority carrier current density Jp(x) is
known in the n-neutral region, the majority
 qD p pno qDn n po  qV
 e  1
carrier current density Jn(x) is also known
J   kT
(30) from Eq. (29) where Jn(x)= J - Jp(x)
 L p Ln  Similarly, for majority Jp(x) in p-neutral
region.

5/29/2021 PHM122_Part 3 22
The total current passing through the junction of area (A) is equal to
D p Dn 
I  AJ  Aq  p no  n po   e qV  1
kT ppo NA nno ND

 Lp Ln  npo= ni2/ppo & pno= ni2/nno

= n i 2/ N A = ni2/ ND
I  I o  eqV kT
 1 (31) Ideal diode equation

 D p pno Dn n po   D p n i2 Dn n i2  Io is the reverse

I o  Aq     Aq    (32)
L
 p L n  L
 p DN Ln N A saturation current

The major features of the predicted I - V I

characteristics are summarized in the figure

 For forward biasing greater than a few kT/q. the

exponential term dominates and I ~ Io exp(qV /kT).
Reflecting the expected exponential dependence. I ~ Io e qV /kT.

 For reverse biases greater than a few kT/q, a few

tenths of a volt at room temperature, the V
exponential voltage term in the ideal diode
equation becomes negligible and I ~ - Io I ~ - Io

5/29/2021 PHM122_Part 3 23
 ln(I)
 Reflecting the expecting exponential dependence, the
ideal forward bias characteristics are often plotted as
illustrated in the figure.
I  I o  e qV kT
 1  I o  e qV kT
 Slope = q/kT

ln I  ln I o  qV kT
ln(Io)
V
 Effect of temperature on I-V characteristics
 The reverse saturation current increases with increasing
operating temperature or decreasing band gap
 Increase of intrinsic carrier concentration
3/2
 2 kT   E g 2 kT
ni  2  2  (mn* m*p )3/4 e
 h 
 The built-in potential decreases with increasing
temperature

Recall
kT N A N D  D p n i2 Dn n i2 
V bi  ln Io  Aq   
q 2
ni  p N D Ln N A 
L

5/29/2021 PHM122_Part 3 24
 Real Diode I-V Characteristics (Deviation From Ideal Characteristics)
Forward Bias
1. Recombination in depletion region
Recombination
• Under forward bias, significant recombination can current

take place because the depletion region contains

excess carriers of both types, which are in transit
from one side of the junction to the other.
• An accurate analysis of the recombination current (Irec)
shows that it increases with forward bias according to
approximately exp(qV/2kT). =1

• The recombination current component is added to the

ideal diode diffusion currents and dominates in silicon

Forward Current (A)

diodes at very low forward bias.
One can rewrite the forward bias current as:

I  I o\  e qV  kT  1
=1
where  is called the ideality factor. Its value is 1 when Recombination
=2 current
diffusion current dominates and 2 when recombination
current dominates.
Forward Bias (Volt)
5/29/2021 PHM122_Part 3 25
2. High level injection effect
• Under high forward bias, the excess minority concentration approaches the majority
concentration (high level injection, which is beyond the level of this course)
• However, the net result of high-level injection is that the ideality factor changes to about
𝜂 = 2 at larger current levels.
=2
3. Series resistance effect
=1
• The neutral n and p Series
regions have an Resistance
inherent resistance Ffect

combined to form the

Forward Current (A)

series resistance Rs High Level
Injection
• The voltage drop across the junction
Vj = V - IRs
• At low currents, IRs is totally negligible and =1
Vj ≅ V as assumed in ideal diode.
Recombination
=2 current
• At high currents, IRs becomes comparable
to V and Vj reduced.
 Due to the reduction in Vj the current increases Forward Bias (Volt)
more slowly as shown in the figure.

5/29/2021 PHM122_Part 3 26
 
Reverse Bias
Ideal diode current

• In the reverse-bias, generated carriers by any Generation current

way in the transition (depletion) region are
swept out before recombination can occur, and
net generation current results.
Generation current
• Thermal generation depends on the temperature
Ideal diode current
and the nature of the recombination centers
(traps) and naturally increases as w increases.
Recall 1
• As a result, the generation current increases  2  N  N   2

almost linearly with w, or with the square root Under Bias w  

q (V V ) 
a d
bi R
of reverse-bias voltage.   N N a d 
• The aside figure shows: I
 The ideal reverse characteristic (dashed line) which is Ideal current
voltage-independent current = -Io. equation

 Actual (real) characteristics (solid line) are higher due to -Io V

generation in the depletion region.
 The generation current increases with increasing reverse Real current
bias until breakdown occurs at high voltages.
Breakdown

5/29/2021 PHM122_Part 3 27
 I
Junction Breakdown
• When a sufficiently large reverse voltage is qV
applied to a p-n junction, the junction breaks e  kT

down and conducts a very large current

VBD
• The reverse voltage at which the current
V
increases sharply is called the BREAKDOWN
VOLTAGE, VBD
• Reverse breakdown can occur by two mechanisms:
1- Zener Breakdown
2- Avalanche Breakdown
Potential
barrier
1. Zener Breakdown (Band to Band Tunneling)

Energy
As we studied, there are two major requirements for
quantum tunneling to occur and be significant: Electron

I. The width of the potential energy barrier, d,

must be very thin. Quantum mechanical
tunneling becomes significant only if d < 10 nm d

II. There must be filled states in one side of the barrier and empty states on the other
side of the barrier at the same energy.

5/29/2021 PHM122_Part 3 28
• Band to band occurs when valence electrons in p w
side pass through the depletion layer and reach Ecp
Efp
the n-side conduction band. Empty Filled Evp
state in state in
n side p side
• This mechanism is possible if the depletion layer
 VR
is very thin and the electric field is high.
Ecn
For silicon Efn
Evn
 High impurity concentration on both the p and n
sides (> 5 1017 cm-3).
 Zener Breakdown field is about 106 V/cm
 the breakdown field is reached approximately
with a reverse bias less than 6V.

• Since the energy bandgap Eg decreases with increasing

temperature, the Zener breakdown voltage decreases
with increasing temperature.

5/29/2021 PHM122_Part 3 29
2. Avalanche Breakdown
• For lightly doped junctions, the breakdown mechanism involves the impact ionization
of host atoms by energetic carriers.

• When the electric field () in the depletion n out

M  multiplication factor 
region is increased, the carriers gain energy n in
(E) equal to: 
nout nin
l Ecp
E  q    x Efp
0 Evp
where l is the mean free path
VR
• If carriers have a kinetic energy at least Ecn
Efn
3/2 Eg they can excite electron-hole
Evn
pairs by impact ionization.

• The degree of MULTIPLICATION could become very high if carriers generated

within the depletion region also have ionizing collisions with the lattice.
 This is an Avalanche Process, since each incoming carrier can initiate
the creation of a large number of new carriers.

5/29/2021 PHM122_Part 3 30
• Physically, we expect that carrier
multiplication factor (M) increases
with increasing the electric field,
and therefore to depend on the
reverse bias
Avalanche
• Measurements of M lead to an
EMPIRICAL RELATION
1
M
 
n
1  V R V BD
where
n varies from about 3 to 6 depending on the material type
VBD is the breakdown voltage which depends on doping and energy gap of the material
as shown in the figure.

• The temperature effect is generally used to distinguish the band to band tunneling
mechanism from the avalanche mechanism. Increasing temperature increases the
probability of zener breakdown while decreases the probability of avalanche breakdown.

5/29/2021 PHM122_Part 3 31
pn Junction Capacitors
There are basically two types of capacitance associated with a junction:
(1) the depletion or transition capacitance (junction capacitance)
(2) the diffusion capacitance (charge storage capacitance)

We must use the more general definition of capacitance : C  Q

V
 Depletion Capacitance (Cj) Charge density

• For a diode, the space-charge region contains qND

equal and opposite fixed charges Qj -dQj  dV
+dQj  dV
-xp +
• If the junction reverse bias increases with dV, x
- xn
additional charge |dQj| is added at the edges -qNA
Depletion region
of the depletion region as shown in the Fig.
w
dQ j
thus the associated capacitance is C j 
dV Recall
1

  N N   2 ND
xp  w
NA ND
Q j  AqN D xn  AqN A x p  A 2q
 N  N (V V ) 
A D 1
bi
 2  N  N   2

    w (V V ) 
q  N N 
A D A D
bi

where A is the cross-sectional area of the junction    A 

D

5/29/2021 PHM122_Part 3 32
 45
Depletion capacitance equals to
1/ 2 40

dQ j A  2q ND NA  A

CAPACITANCE (pF)
Cj   
2  (V V ) ( N  N ) 
(33) 35
dV w
 bi D A
30

Note : C j  1 V
25

For a one-sided abrupt junction 20

0 5 10 15 20 25

Assume the case of p+n or NA >> ND , Eq.(33) will be REVERSE BIAS VOLTAGE (VOLTS)

1/2

A  2q N D  1 2(V  V )
Cj 
2  (V V ) 
 2 bi (34) 1/Cj2
Cj 2
A q N D
 bi 
• It is clear from Eq. 34 that a plot of 1/Cj2 versus
V produces a straight line for a one-sided Slope=2/A2qND𝜺
abrupt junction.
Capacitance data
 The slope gives the impurity
V
concentration ND of the substrate, and Vbi

the intercept (at 1/Cj2 = 0) gives Vbi.

5/29/2021 PHM122_Part 3 33
VARACTOR
The term varactor is a shortened form of variable reactor, referring to the voltage-variable
capacitance of a reversed pn junction

Varactors are used in the tuning stage of

radio receivers to replace more bulky
tunable parallel-plate capacitors. Other
application of varactors include use in Varactor

harmonic generation and active filters

Tunable parallel-plate capacitor

As previously derived, the reverse biased depletion capacitance is given by:

   
m m
C j  V bi V R or Cj  V R for V R  V bi

where m=1/2 for abrupt junction. (Eq.34)

m=1/3 for a linear graded abrupt junction.

We can increase the voltage sensitivity by using

hyperabrupt junction which having m >1/2
5/29/2021 PHM122_Part 3 34
 Diffusion Capacitance (Cd)
• Under forward bias, electrons are injected from the n-region into the p-region and holes
are injected from the p-region into the n-region.
• Hatching areas in the Fig. represent the charge
p-neutral n-neutral
of injected minority carriers (Qp and Qn) that region np region
are stored in the n and p neutral regions.
n pn
• The variation of stored charges with changing pn  x  pno pe
( xxn ) Lp
Qn
applied forward bias is presented by a diffusion npo p
capacitance (Cd)
Qp
d (Qp  Qn ) pno
Cd 
dV -xp xn
x
0
• If A is the diode cross section, then
 
Qp  qA   pn  pno  dx  qA  pno  e qV  1 e dx  qALp pno  e qV  1
kT  ( x  xn ) L p kT

xn xn

Similarly, Qn  qALn n po e 
qV kT
 1

dQn q A( Lp pno  Ln n po ) qV
2
dQ p
So, Cd    e kT

dV dV kT
• For p+n one-sided abrupt junction q 2 AL p pno
Cd  e qV kT

NA >> ND i.e. ppo >> nno or npo << pno kT

5/29/2021 PHM122_Part 3 35