CHAP #02 (PART 1)

SEMICONDUCTOR FUNDAMENTALS

In this lecture you will learn:

• What are semiconductors?

• Basic crystal structure of semiconductors

• Electrons and holes in semiconductors

• Intrinsic semiconductors

• Extrinsic semiconductors
n-doped and p-doped semiconductors

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Semiconductors in the Periodic Table

Atomic number

Group IV semiconductors

Each element in group IV has 4

electrons in its outer most
atomic shell

A Silicon atom has:

2 electrons in the first atomic shell

8 electrons in the second atomic shell
4 electrons in the third outermost atomic shell
A Silicon atom

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

1
 Semiconductors in the Periodic Table

Atomic number

Group IV semiconductors

Each element in group IV has 4

electrons in its outer most
atomic shell

• The outermost electrons are called valence core

electrons

• The inner electrons are called the core

electrons
A Silicon atom
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Covalent Bonding in Silicon

• A Silicon atom with 4 electrons in the valence shell is drawn in

a cartoon way as:

• Two Silicon atoms can come together to form a covalent bond by sharing two
electrons among themselves

also drawn as

shared electrons a covalent bond

• Covalent bonding is energetically favorable (i.e. Silicon atoms “like” to form

covalent bonds with each other)

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

2
 A Silicon Crystal Lattice (A Cartoon View)

In a Silicon crystal:

• Each Silicon atom is surrounded by 4 other Silicon atoms

• Each Silicon atom forms 4 covalent bonds with the neighboring Silicon atoms

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Actual 3D Structure of a Silicon Crystal Lattice

covalent bonds
• Each Silicon atom is surrounded by 4 other Silicon atoms in a tetrahedral
configuration

• Silicon atomic density = 5x1022 cm-3

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

3
 Electrons and Holes in Semiconductors - I

A perfect Silicon crystal lattice

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Electrons and Holes in Semiconductors - II

positively charged
hole +

negatively charged
free electron

A Silicon crystal lattice with one broken bond

• It requires energy to break a covalent bond

• The required energy is called the “bandgap” (bandgap of Silicon is ~1.12 eV)

• A broken bond results in one negatively charged “free electron” and one
positively charged “hole”

• The “free electron” can freely move around in the crystal

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

4
 Electrons and Holes in Semiconductors - III

positively charged
hole

+ negatively charged
free electron

A Silicon crystal lattice with one broken bond

• A hole can also move through the lattice !!

• A hole moves when an electron from a neighboring bond jumps over to fill that
“hole”

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Electrons and Holes in Semiconductors - III

positively charged
hole

negatively charged
free electron
+

A Silicon crystal lattice with one broken bond

• A hole can also move through the lattice !!

• A hole moves when an electron from a neighboring bond jumps over to fill that
“hole”

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

5
 Definitions and Notations Used in ECE 3150

• The word electron will usually mean a “free electron” (and not an electron
forming the covalent bond or a core electron)

• The electron density is denoted by: n (units: 1/cm3)

• The hole density is denoted by: p (units: 1/cm3)

• The charge of an electron is: q

q  1.6  10 19 Coulombs
• The charge of a hole is:  q

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Electrons and Holes at Near Zero Temperature

A perfect Silicon crystal lattice at temperature T0 K

• There are no broken bonds and no electrons and holes

( i.e. n = p = 0 )

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

6
 Electrons and Holes at Nonzero Temperature
holes
+

electrons
+

+
A Silicon crystal lattice at temperature T>0 K

• Thermal energy breaks the covalent bonds and electron-hole pairs are generated
(remember it takes energy to break a covalent bond)

• The number of electrons and holes generated are equal - for every electron
generated there is also a hole generated ( i.e. n = p )

• Question: what is the electron and hole density at room temperature?

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Thermal Energy (Radiative, Bias)

Thermal energy is typically measured in units of “ KT ”

“K” is the Boltzmann’s constant and equals ~1.38 x 10-23 Joules/Kelvin

Temperature “T” is measured in degrees Kelvin

Room temperature corresponds to T = 300 K

Room temperature corresponds to a “KT ” value of 4.14 x 10-21 Joules

or 25.8 meV

Energy in Joules
Energy in eV =
Electron charge in Coulombs

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

7
 Generation and Recombination in Semiconductors - I

Generation:
The breaking of a bond to generate an electron-hole pair is called generation

• Generation rate G(T ) is a function of temperature

• Units of G(T ) are: cm-3-s-1

Recombination:
An electron can also combine with a hole to form a bond. This process is called
recombination. It is the reverse of generation.

• Recombination rate R(T ) is proportional to the product np

R T   np  R T   k T  np
(you need electrons as well as holes for recombination to happen)
• Units of R(T ) are also: cm-3-s-1

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Generation and Recombination in Semiconductors - II

Condition of Thermal Equilibrium:

• In thermal equilibrium a steady state exists in which the rate of electron-hole
generation is equal to the rate of electron-hole recombination,
R T   G T 
 k T no po  G T  thermal equilibrium electron and hole
G T  densities are usually denoted by no
 no po  and po
k T 
G T 
is written as ni T 
2
• By convention, the ratio
k T 
• Therefore, in thermal equilibrium, no po  ni2 T 
Điều kiện cân bằng nhiệt động
• Since equal number of electrons and holes are present in thermal
equilibrium, we have, no  po  ni T 

• ni is called the “intrinsic” carrier density. It equals the number of electrons

(or holes) present in a pure semiconductor in equilibrium at a given
temperature.
10 -3
• For Silicon, ni  1 10 cm at room temperature (i.e. at T = 300K)

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

8
 Doping in Semiconductors
Doping:
The introduction of certain impurity atoms in a pure semiconductor to control
its electronic properties is called doping

• Doping is done by two kinds of impurity atoms:

a)Donor atoms
b)Acceptor atoms

Donors: (Pentavalent)
Donor atoms are used to increase the electron density in a semiconductor

• Group V elements have 5 electrons in

their outermost atomic shell (one more
than group IV atoms)

• Group V elements can act as electron

“donors” in Silicon

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Doping by Donors in Silicon (n-doping)

positively charged
fixed donor atom
+
As
negatively charged
free electron

Doping Silicon with Arsenic Atoms

• Donor atom concentration is denoted by: Nd (units: 1/cm3)

• Each donor atom contributes one free electron to the crystal

• Donor atom after giving off an electron becomes positively charged

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

9
 Doping in Semiconductors

Acceptors: (Trivalent)
Acceptor atoms are used to increase the hole density in a semiconductor

• Group III elements have 3 electrons in

their outermost atomic shell (one less than
group IV atoms)

• Group III elements can act as electron

“Acceptors” in Silicon

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Doping by Acceptors in Silicon (p-doping)

negatively charged
fixed acceptor atom
-
B +
positively charged
hole

Doping Silicon with Boron Atoms

• Acceptor atom concentration is denoted by: Na (units: 1/cm3)

• Each acceptor atom contributes one hole to the crystal by “accepting” one
electron from a neighboring bond

• Acceptor atom after giving off a hole (or equivalently, after accepting an
electron) becomes negatively charged

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

10
 Electron-Hole Density in Doped Semiconductors

Consider a N-doped semiconductor in thermal equilibrium:

Doping density = Nd

• Use condition of charge neutrality: q  Nd  no  po   0

2
• Together with the relation: no po  ni ni= 10^10cm^3

• To obtain: 2 ni^2=BT^3.e^(-EG/kT)
Nd N 
no    d   ni2 BT(bandgap enery): 1,12eV
2  2  k(Boltmann's constant): 8,62.10^(-5)eV/K
2
Nd N  B(material dependent parametar): 1,08*10^31 k^(-3).cm^(-6)
po     d   ni2
2  2 

• If Nd  ni , which is usually the case for N-doping, then the above relations
simplify:
no  Nd
n-doping lets one make the electron density
n2 much greater than the intrinsic value ni
po  i
Nd
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Electron-Hole Density in Doped Semiconductors

Now consider a P-doped semiconductor in thermal equilibrium:

Doping density = Na

• Use condition of charge neutrality: q  Na  no  po   0

• Together with the relation: no po  ni2

• To obtain: 2
Na N 
po    a   ni2
2  2 
2
Na N 
no     a   ni2
2  2 

• If Na  ni , which is usually the case for P-doping, then the above relations
simplify:
po  Na
p-doping lets one make the hole density much
n2 greater than the intrinsic value ni
no  i
Na
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

11
 Electron-Hole Density Vs Doping Density

N-doped semiconductors
• With increasing N-doping the electron
density increases above the intrinsic
value and the hole density decreases
below the intrinsic value
no

Example:

Suppose
Nd  1017 cm- 3 and ni  1010 cm-3
then
po
no  1017 cm- 3  Since Nd  ni 
and
ni2
po   103 cm- 3
no
Nd

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Electron-Hole Density Vs Doping Density

P-doped semiconductors
• With increasing P-doping the hole
density increases above the intrinsic
value and the electron density
decreases below the intrinsic value
po

Example:

Suppose
Na  1017 cm- 3 and ni  1010 cm-3
then
no
po  1017 cm-3  Since Na  ni 
and
ni2
no   103 cm-3
po
Na

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

12
 Compound Semiconductors

III-V semiconductors:
Elements in group III can be combined with
elements in group V to give compound
semiconductors (as opposed to elemental
semiconductors of group IV)

*One can also have II-VI semiconductors

Ga atoms
Crystal lattice of
the III-V compound
semiconductor
As atoms GaAs

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

13
 Dẫn -> e CHAP #02 (PART 2)
Bán dẫn -> e, holes

Electron and Hole Transport in Semiconductors

In this lecture you will learn:

• How electrons and holes move in semiconductors

• Thermal motion of electrons and holes

• Electric current via drift

• Electric current via diffusion

• Semiconductor resistors

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Review: Electrons and Holes in Semiconductors

holes
+

electrons
+
+
As

A Silicon crystal lattice +

● There are two types of mobile charges in semiconductors: electrons and holes

In an intrinsic (or undoped) semiconductor electron density equals hole density

● Semiconductors can be doped in two ways:

N-doping: to increase the electron density

P-doping: to increase the hole density
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

1
 Thermal Motion of Electrons and Holes
In thermal equilibrium carriers (i.e. electrons or holes) are not standing still but
are moving around in the crystal lattice and undergoing collisions with:

• vibrating Silicon atoms

• with other electrons and holes
• with dopant atoms (donors or acceptors) and other impurity atoms
(Thời gian trung bình giữa các lần va chạm)
Mean time between collisions =  c

In between two successive collisions electrons (or

holes) move with an average velocity which is (Vận tốc trung bình
called the thermal velocity = Vth của hạt chuyển động
giữa 2 lần va chạm)
In pure Silicon,  c  0.1 10 12 s  0.1 ps
v th  107 cm s
(Quãng đường trung bình
Brownian Motion Mean distance traveled between collisions is giữa 2 va chạm)
called the mean free path    v th  c -> Tính chất của dòng ngõ ra
7 12
In pure Silicon,   10  0.1 10  10- 6 cm  0.01m

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Drift: Motion of Electrons Under an Applied Electric Field

Silicon slab

L
V
E
+ V - L

• Force on an electron because of the electric field = Fn = -qE

• The electron moves in the direction opposite to the applied field with a
constant drift velocity equal to vdn

• The electron drift velocity vdn is proportional to the electric field strength
v dn   E  v dn    n E

• The constant n is called the electron mobility. It has units: cm2

cm2 V-s
• In pure Silicon, n  1500
V-s
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

2
 Drift: Motion of Holes Under an Applied Electric Field

Silicon slab

L
V
E
+ V - L

• Force on a hole because of the electric field = Fp = qE

• The hole moves in the direction of the applied field with a constant drift
velocity equal to vdp

• The hole drift velocity vdp is proportional to the electric field strength
v dp  E  v dp   p E
2
• The constant p is called the hole mobility. It has units: cm
cm2 V-s
• In pure Silicon,  p  500
V-s
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Derivation of Expressions for Mobility

Electrons:
Force on an electron because of the electric field  Fn  qE
Fn qE
Acceleration of the electron  a  
mn mn
Since the mean time between collisions is  c , the acceleration lasts only for a
time period of  c before a collision completely destroys electron’s velocity
q c
So in time  c electron’s velocity reaches a value  a  c   E
mn
q c
This is the average drift velocity of the electron, i.e. v dn   E
mn
q c
Comparing with v dn   n E we get, n 
mn
Holes: q c
Similarly for holes one gets,  p 
mp

Special note: Masses of electrons and holes (mn and mp) in Solids are not the
same as the mass of electrons in free space which equals 9.1 10  31 kg

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

3
 (Độ linh động)
Mobility Vs Doping
More doping (n-type of p-type) means more frequent collisions with charged donor
and acceptor impurity atoms and this lowers the carrier mobility
e>>p
-> Thể tích p>=2e

Mobility (cm2/V-s)

Dopant concentration (1/cm3)

Note: Doping in the above figure can either be n-type or p-type

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

(Dòng trôi)
Drift Current Density of Electrons
e ngược chiều E
Consider electrons moving under an applied electric field: lỗ trống cùng chiều E
E

v dn    n E

(Thông lượng)
Flux Density:
Flux density is the number of particles crossing a unit area surface per second
It has units cm-2-s-1

Unit area surface

Density: n
Velocity: vdn

Flux density: nvdn Area Time

Volume = 1 x (vdn x 1)
Mật độ.Vận tốc

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

4
 Drift Current Density of Electrons
Electrons Drift Current Density: Check directions
Electron flux density from drift  n v dn
E
v dn
Electron drift current density J ndrift is,
J ndrift

J ndrift  q  electron flux density v dn    n E

 q n v dn  q n  n E

Jndrift has units: Coulombs  Amps

cm2 - s cm2

e: drift ngược chiều flux

holes: drift cùng chiều flux

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Drift Current Density of Holes

Holes Drift Current Density: Check directions

drift ,
The hole drift current density is J p
E
J pdrift   q  hole flux  v dp
  q p v dp  q p  p E J pdrift
v dp   p E

J pdrift has units: Coulombs  Amps

cm2 - s cm2

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

5
 Conductivity and Resistivity

drift current cùng chiều điện trường

Total Drift Current Density:
drift
The total drift current density J drift is the sum of J n drift
and J p
drift current phụ thuộc khả năng di chuyển của hạt->
J drift  J ndrift  J pdrift độ linh động-> loại hạt dẫn, vật liệu

 q n n  p  p  E
 E
The quantity  is the conductivity of the semiconductor:
độ dẫn điện
  q n n  p  p 

Conductivity describes how much current flows when an electric field

is applied. Another related quantity is the resistivity  which is the
inverse of the conductivity, điện trở suất
1


Units of conductivity are: Ohm-1-cm-1 or -1-cm-1 or S-cm-1
Units of resistivity are: Ohm-cm or -cm or S-1-cm

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Example: A Semiconductor Resistor

Silicon slab
For a resistor we know that,
V
I 1 L
R
-
We also know that, V
+
Area  A
I  J drift A
 E A
I
V  A
 A V 2
L L

1  2   R 
L

L where  
1


 q n n  p  p 
 A A
Lessons:
• Knowing electron and hole densities and mobilities, one can calculate the
electrical resistance of semiconductors Biết mật độ và độ linh động của e, holes
->tính được điện trở của chất bán dẫn
• n-doping or p-doping can be used to change the conductivity of
semiconductors by orders of magnitudes

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

6
 (Khếch tán)
Diffusion

Diffusion of ink in a glass beaker

Why does diffusion happen?

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Diffusion and Diffusivity

There is another mechanism by which current flows in semiconductors …….

• Suppose the electron density inside a semiconductor is not uniform in space,

as shown below
n(x)
electron flux in +x direction
electron flux in -x direction
d n x 
slope 
dx
x
• Since the electrons move about randomly in all directions (Brownian motion),
as time goes on more electrons will move from regions of higher electron
density to regions of lower electron density than the electrons that move from
regions of lower electron density to regions of higher electron density
d n x 
• Net electron flux density in +x direction  
dx
d n x 
  Dn
dx
• The constant Dn is called the diffusivity of electrons (units: cm2-s-1 )

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

7
 Đối với e: Dòng sinh ra từ sự khếch tán
ngược chiều hướng chuyển động

Đối với holes: Dòng sinh ra từ sự khếch tán

Diffusion Current Density cùng chiều hướng chuyển động

Electrons Diffusion Current Density:

d n x  Check directions
Electron flux density from diffusion   Dn
dx n x 

Electron diffusion current density J ndiff is,

J ndiff  q  electron flux density
x

d nx  Elec. flux

 q Dn e: diff ngược chiều flux
dx holes: driff cùng chiều flux J ndiff
Holes Diffusion Current Density:
d p x  Check directions
Hole flux density from diffusion   Dp
dx p x 
diff
Hole diffusion current density J p is,
diff

J p   q  hole flux  x
d p x  Hole flux
 q Dp
dx J pdiff

J ndiff and J pdiff has units Coulombs  Amps

cm2 - s cm2
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Einstein Relations
Einstein worked on other things besides the theory of relativity……..

• We introduced two material constants related to carrier transport:

1) Mobility (Độ linh động)
2) Diffusivity (Độ khếch tán)

•Both are connected with the transport of carriers (electrons or holes)

•It turns out that their values are related by the Einstein relationships

Einstein Relation for Electrons:

Example:
Dn KT
 In pure Silicon, n  1500 cm
2
V-s
n q
2
 p  500 cm V - s
Einstein Relation for Holes:
Dp KT This implies, Dn  37.5 cm2 s
 Dp  12.5 cm2 s
p q

• K is the Boltzman constant and its value is: 1.38x10-23 Joules  K

•
K T has a value equal to 0.0258 Volts at room temperature (at 300 oK)
q

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

8
 Total Electron and Hole Current Densities

Total electron and hole current densities is the sum of drift and diffusive
components
Electrons:

J n  x   J ndrift  x   J ndiff  x 
d n x 
 q n x  n E  x   q Dn
dx

Holes:

J p  x   J pdrift  x   J pdiff  x 
d p x 
 q p  x   p E  x   q Dp
dx

Electric currents are driven by electric fields and also by carrier density
gradients

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Thermal Equilibrium - I
There cannot be any net electron current or net hole current in thermal
equilibrium ……… what does this imply ??

Consider electrons first:

J n  x   Jndrift  x   J ndiff  x   0
d no  x 
 q no  x  n E  x   q Dn 0 1
dx

1 d logno  x  q
can also be written as:  E x 
dx KT
d  x 
Since the electric field is minus the gradient of the potential: E  x   
dx
We have: d logno  x   q d  x 
dx KT dx q  x 
The solution of the above differential equation is: no  x   constant  e KT

But what is that “constant” in the above equation ???

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

9
 Thermal Equilibrium - II
q  x 
We have: no  x   constant  e KT

Note: one can only measure potential differences and not the absolute values
of potentials
Convention: The potential of pure intrinsic Silicon is used as the reference
value and assumed to be equal to zero.
q x 
So for intrinsic Silicon, no  x   constant  e KT  constant
But we already know that in intrinsic Silicon, no  x   ni

So it must be that, constant  ni

q  x 
And we get the final answer, no  x   ni e KT

Consider Holes Now: q x 


One can repeat the above analysis for holes and obtain: po  x   ni e
KT

Check: no  x  po  x   ni2
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Potential of Doped Semiconductors

What are the values of potentials in N-doped and P-doped semiconductors ??

N-doped Semiconductors (doping density is Nd ):

The potential in n-doped semiconductors is denoted by: n

no  x   Nd Example:
q n  x  Suppose,
 Nd  ni e KT Nd  1017 cm- 3 and ni  1010 cm-3
KT N  KT N 
 n  log d   n  log d    0.4 Volts
q  ni  q  ni 

P-doped Semiconductors (doping density is Na ):

The potential in p-doped semiconductors is denoted by: p

po  x   Na Example:
q p  x  Suppose,

 Na  ni e KT Na  1017 cm- 3 and ni  1010 cm-3
KT N  KT N 
 p   log a   p   log a    0.4 Volts
q  ni  q  ni 
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

10
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

11
 CHAP #02 (PART 3)

Electrons and Holes in Semiconductors

In this lecture you will learn:

• Generation-recombination in semiconductors in more detail

• The basic set of equations governing the behavior of electrons and holes
in semiconductors

• Shockley Equations

• Quasi-neutrality in conductive materials

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Majority and Minority Carriers

In N-doped Semiconductors:
Electrons are the majority carriers
Holes are the minority carriers

In P-doped Semiconductors:
Holes are the majority carriers
Electrons are the minority carriers

Golden Rule of Thumb:

When trying to understand semiconductor devices, always

first see what the minority carriers are doing

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

1
 Generation and Recombination in Semiconductors
Generation process:

We can write the following

+
equations for the carrier
densities:
Rate = G n t 
GR
t
p t 
Recombination process: GR
t
These equations tell how the
electron and hole densities
change in time as a result of
recombination and
+ generation processes

Rate = R

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Generation and Recombination in Thermal Equilibrium

• From the first lecture, in thermal equilibrium:

The recombination rate  Ro  k no po
equals the generation rate  Go

i.e. Go  Ro

• Then in thermal equilibrium:

no t 
 Go  Ro  0
t
po t 
 Go  Ro  0
t

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

2
 Light Absorption in Semiconductors
Generation of electrons and holes by photons in semiconductors:


Photon

  E g
A perfect silicon crystal
lattice

Negatively charged
free electron

Positively charged
hole +

A silicon crystal lattice with one broken bond

(one electron and one hole)
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Generation and Recombination Out of Thermal Equilibrium

1) Consider a P-doped slab of Silicon:

 no  po 
Electron-hole recombination rate in thermal equilibrium  Ro  k no po
equals the generation rate  Go  k ni2

2) Now turn light on at time t = 0:

light

• Light breaks the Si-Si covalent bonds and generates excess electron-hole pairs

• The net generation rate now becomes: G  Go  GL

extra part
3) Mathematical model of the above situation:
• n’(t ) and p’(t) are the excess electron and hole densities
n  no  n' t 
• It must be that: n' t   p' t 
p  po  p' t 
• We also assume that: n' t , p' t   po
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

3
 Generation and Recombination Out of Thermal Equilibrium
• We can use the equations:

nt  pt  n  no  n' t 

GR GR p  po  p' t 
t t

• Generation rate:
G  Go  GL
• Recombination rate:
Rknp Assumptions: n' , p'  po

 k no  n'  po  p'

1  n is the lifetime of the
 k no  n' po  k po
n minority carriers (i.e. electrons)
 k no po  k n' po
n' The excess recombination rate is proportional
 Ro 
n to the excess MINORITY carrier density

• The equation for excess minority carriers (i.e. electrons) becomes:

n' t  n' t 
 GL 
t n
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Generation and Recombination Out of Thermal Equilibrium

n' t  n' t 
 GL 
t n
• Solution with the boundary condition, n' t  0   0 , is:

n’(t ) light turned-on at t = 0

  
t
  GL  n
n' t   GL  n  1  e  n 
 
 
t 0 t

• Excess hole density is, of course :

p' t   n' t 

• As t   the excess electron and hole densities reach a steady state value

n' t     GL  n nt     no  GL  n
and
p' t     GL  n pt     po  GL  n

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

4
 Generation and Recombination Out of Thermal Equilibrium

Now suppose that light had been turned-on for a very very long time and it was
turned-off at time t = 0

At time t = 0 : n'  GL  n n  no  GL  n
and
p'  GL  n p  po  GL  n

n’(t ) light turned-off at t = 0

??

t 0 t

• Since n  no , and p  po , the carrier densities are not equal to their thermal
equilibrium values. Thermal equilibrium must get restored since the light has been
turned-off

Question: How does thermal equilibrium gets restored??

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Generation and Recombination Out of Thermal Equilibrium

• We can use the equations:

nt  pt  n  no  n' t 

GR GR
t t p  po  p' t 
• Generation rate:
G  Go
• Recombination rate:
Rknp Assumptions: n' , p'  po

 k no  n'  po  p'

 k no  n' po
 k no po  k n' po
n' The excess recombination rate is proportional
 Ro  to the excess MINORITY carrier density
n
• The equation for excess minority carriers (i.e. electrons) becomes:

n' t  n' t 

t n

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

5
 Generation and Recombination Out of Thermal Equilibrium
n' t  n' t 

t n
• Solution is:
t

n' t   n' t  0  e  n 
Excess electron density decays exponentially
to zero from its initial value

n’(t ) light turned-off at t = 0

The excess carrier densities
decay with time and thermal
equilibrium values for carrier
densities are restored
t 0 t
t

• The excess hole density will also decay in the same way: p' t   p' t  0  e  n

• As t   the electron and hole densities reach their equilibrium values:

n' t     0 n t     no
and
p' t     0 p t     po
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Generation and Recombination in Doped Semiconductors

Whenever you have to find an expression for R use the following recipe:

• If it is a p-doped semiconductor:
n'  x , t   n is the minority carrier lifetime
R  Ro 
n

• If it is a n-doped semiconductor:
p'  x , t 
R  Ro   p is the minority carrier lifetime
p

The excess recombination rate (i.e. R - Ro ) is always proportional to the excess

MINORITY carrier density

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

6
 Electron and Hole Current Density Equations

From last lecture…………

Shockley: Mối quan hệ giữa ngõ ra, vào

d n x 
J n  x   q n x  n E  x   q Dn 1
dx

d p x  2
J p  x   q p  x   p E  x   q Dp
dx

These are two of Shockley’s equations !

Shockley, Bardeen, and Brattain

from Bell Labs were awarded the
Nobel Prize for inventing the
semiconductor transistor

William John Walter

Shockley Bardeen Brattain
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Electron and Hole Current Continuity Equations

• You have already seen the equations:

n  x , t 
GR
t
p  x , t 
GR
t
These equations tell how the electron and hole densities change in time as a
result of recombination and generation processes.

• Carrier densities can also change in time if the current densities change in
space !!!

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

7
 Electron and Hole Current Continuity Equations
Consider the infinitesimal strip between x and x+x
x

Jp x  x x  x J p  x  x 
The difference in hole fluxes at x and x+x must result in piling up of holes
in the infinitesimal strip …….
 p  x , t  x Note that q p  x , t  is the hole
J p  x , t   J p  x  x , t   q charge density
t
J p  x  x , t   J p  x , t   p x , t 
 q
x t
J p  x , t   p x , t   p x , t  1 J p  x , t 
 q 
x t t q x
Now add recombination and generation to the above equation:

 p x , t  1 J p  x , t 
GR
t q x

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Electron and Hole Current Continuity Equations - III

One can do the same for electrons as well……
x

Jn  x  x x  x J n  x  x 

 n x , t  1 J n  x , t 
GR
t q x

So now we have two new equations,

 p x , t  1 J p  x , t  3 
GR
t q x

 n x , t  1 J n  x , t  4 
GR
t q x

These are two more of Shockley’s equations !

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

8
 Gauss’s Law and Electrostatics

The net charge density in a semiconductor is,

  x , t   q   Nd  x   Na  x   p x , t   n x , t  

Gauss’s Law in differential form:

 E x, t   x , t 

x s

 E  x , t  q  Nd  x   N a  x   p  x , t   n  x , t  
 5 
x s

This is the fifth and the last of the Shockley’s equations !

 o  8.85  10 12 Farads/m

For Silicon:  s  11.7 o
 8.85  10 14 Farads/cm

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

The Five Shockley Equations

d n x , t 
J n  x , t   q n x , t  n E  x , t   q Dn 1
dx

d p x , t 
J p  x , t   q p x , t   p E  x , t   q Dp 2
dx

 p x , t  1 J p  x , t  3 
GR
t q x

 n x , t  1 J n  x , t  4 
GR
t q x

 E  x , t  q  Nd  x   N a  x   p  x , t   n  x , t  
 5 
x s

Using these equations one can understand the behavior of semiconductor

microelectronic devices !!

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

9
 Quasi-Neutrality
Materials with large conductivities are “quasi-neutral”

“Quasi-neutrality” implies that there cannot be large charge densities or electric fields
inside a conductive material

Lets see why this is true…..and how deviations from quasi-neutrality disappear…….

Consider an infinite and conductive N-doped semiconductor with a net charge

density at time t=0:
N-doped  
s Charge density

The charge density will generate electric fields (by Gauss’ law):

N-doped  
E
s

ECE 315 – Spring 2005 – Farhan Rana – Cornell University

Quasi-Neutrality
The electric field will generate electrical currents:
J n  x , t   q n  x , t   n E  x , t   E  x , t 

N-doped  E

s
Jn

The electrical currents will pile electrons on top of the charge density and neutralize it
and then there is no charge density left in the medium…………

N-doped 

s

This whole process takes a time of the order of the dielectric relaxation time  d :
s
d  ~ 10 15  10 13 Seconds


ECE 315 – Spring 2005 – Farhan Rana – Cornell University

10
 Appendix: Restoration of Quasi-Neutrality

N-doped  E

s
Jn

From Gauss’ law:


   r , t 
.E r , t  
s
Current equation:
   
J r , t   E r , t 

Use the continuity equation for charge:


 r , t       
 .J r , t   .E r , t     r , t 
t s
 
 r , t   r , t   s
  0  d 
t d   Charge density in a
t
conductive medium
 disappears on a time
  d
 r , t    r , t  0 e
Solution:
scale of d
ECE 315 – Spring 2005 – Farhan Rana – Cornell University

11