Semiconductor Lecture Notes.

PRESBYTERIAN UNIVERSITY COLLEGE GHANA
OKWAHU CAMPUS
FACULTY OF SCIENCE AND TECHNOLOGY
DEPARTMENT OF INFORMATION &
COMMUNICATION TECHNOLOGY
2023/2024 ACADEMIC YEAR SECOND SEMESTER
Course Title: SEMICONDUCTOR DEVICES
Course Code: CPE 254
Contact Hours: 3 CREDITS HOURS
Course Pre-requisite: READING AHEAD OF THE
LECTURE NOTES
Semester: YEAR 2 SECOND SEMESTER
Name of Lecturer: ING. ISAAC EMPRAIM
Teaching Assistant: N/A
Contact Details:
Office Hours: Monday to Friday, 08.00—
17.00
Email: ikebaidooempraim@gmail.com
Telephone: 0244026082/0202306371
Course Objectives:
 To introduce students to basic understanding of microelectronics devices and study of the
circuits contained in such devices.
 To enable students to understand the basic functional characteristics of Bipolar Junction
Transistors (BJT) and Metal Oxide Silicon(MOS) devices in both digital and analog
applications.
 To provide students with basic techniques for modeling, analysis, and design of the
circuits.
Content:
PN Junction: Junction diodes and their static and dynamic properties at high frequencies, switching.
Metal-Semiconductor junctions: Energy band diagram of the Schottky barrier, MIS Schottky
diode, ohmic contact, applications of Schottky-Barrier diodes, heterojunctions. Solar Cell and
light-emitting diode: optical absorption in a Semiconductor, Photovoltaic effect, Schottky-barrier
and MIS solar cells, LEDs, eye sensitivity and brightness, Quantum efficiency.
MOS Devices MOS capacitor, MOS transistor (DC characteristics, depletion - MOST and JFET).
Polysilicon gate technology, Metal gate technology, comparison. Control of threshold voltage through iron
implantation. Frequency and small signal properties of MOST, special MOS devices (CMOS, VMOS,
DMOS, CCD). Examples of CCD in electro-optical applications and signal processing.
Bipolar Junction Devices: The BJT transistor action, Ebers-Molls model, Hybrid-pi and h equivalent
circuit, the BJT as a switch, breakdown voltages, lateral PNP transistor, PNPN devices, Noise source in
BJT.
Mode of Delivery
Lectures through Power Point Presentations; Class Exercises; Home works and group assignments; and
Tutorials.
Reading List
 Bapat, Y. N. (1992). Electronic Circuits and Systems : Analog and Digital,1e. Tata McGraw-Hill
Education.
 Jain, R. P. (1987). Digital Electronics and Microprocessors: Problems and Solutions. Tata
McGraw-Hill Education.
 Jain. (2010). Modern Digital Electronics 4E. Tata McGraw-Hill Education.
 Milnes, A. G. (2012). Semiconductor Devices and Integrated Electronics. Springer Science &
Business Media.
 Taub, H., & Schilling, D. L. (1977). Digital integrated electronics. McGraw-Hill.
Mode of Assessment:
State how you intend to grade the course. For example
 Attendance - 0%
 Assignments - 10%
 Mid Semester examination - 30%
 Final (End of semester exams) - 60%
FUNDAMENTALS OF SEMICONDUCTORS
SESSION 1-1:CRYSTAL STRUCTURE AND
ENERGY BANDS, AND ELECTRONIC
STRUCTURES
1-1.1 Introduction: Crystal Structure And Energy Bands

A great majority of semiconductor materials finding practical application in electronics are in the
form of crystalline solids like that of diamond and zincblende example, the crystalline structure of
silicon (Si) is that of diamond whilst that of the gallium arsenide (GaAs) is of the form of
zincblende. In the crystalline solid, atoms forming the primitive cell (lattice) are arranged in an
ordering manner, which. by repeating itself forms the crystal. It is often necessary to refer to
systems of planes within the crystal that run in certain directions, intersecting certain sets of
atoms. It is the orientation of such systems of planes that determine the properties of the crystal.
These sets of planes are specified by sets of three numbers called Miller indices.
The bond formation of crystalline silicon material is of the covalent -electron pair bond in which
an atom shares its four valence electrons with its four nearest neighbours. Thus, no free electrons
are available for conduction at absolute zero temperature. The conductivity is however, every
sensitive to tiny amounts of impurity atoms and increases with rising temperature. In the
semiconductor, the valence electrons group together to occupy a band of energy levels, called the
valence band. The next higher band of allowed energy levels called the conduction band is
separated from the former by a forbidden gap of energy, e.g., it is this energy gap E.g., that
classifies a material as a metal, insulator or semiconductor. In the case of metals, E.g., is zero
whilst e.g., is large in insulators for an electron to be successfully ejected from the valence band
to the conduction band. In a semiconductor the forbidden energy gap is not too large, for
example, E.g., =1.12 eV in silicon and 1.43 E g=1.12 eV
EE 271 SEMICONDUCTOR DEVICES
eV in GaAs. This allows electrons in the valence band to be promoted to thunder the
influence of an external electric field. For each successful transition of Thus, current
conduction in semiconductors is due to both electrons (negatively charged) and holes
(positively charged). In terms of quantum mechanics, the electron in a crystalline lattice is
considered as a particle in constant motion with momentum k and energy E. The energy-
momentum (E-K) relationship is usually obtained by solving the Schrodinger equation.
2
−h 2
[ 2
∇ +V (r)]Φ k (r )=E k Φ k (r ) (1.1)
8π m
Where V(r) is the periodic potential energy of the lattice, h is the Planck's constant and m
the mass of electron. ) is the solution of equation (1.1) which has the form:
(1.2)
jk ⋅r
ϕ K (r )=e U n (k , r)=B 10 cch
WhereU n (k ,r ) is periodic in r with the periodicity of the direct lattice and n is the band index.
function
The band structure resulting from equation (1.1) will be discuss latter in the course for Si and
GaAs. In either crystal the energy gap Eg is the difference between the lowest conduction-band
minimum and the valence band maximum.The mass of the carriers (which is different from the
electron mass m。 in free space) depends on the energy band they occupy, the momentum of the
carriers and the direction of the applied field.
1-1.2 Electronic Structure
1-1.2.1 Number of states in a band
The E(k) relationship tells us where allowable electronic states are, we also need to know how
many there are:
A crystal of size L is a finite potential well or “Box” for a “free"(we find that
1.
1
electrons in a crystal could move between the walls like free electrons) electron, Fig
13
Figure 1.1:“free” Bloch electrons.

We know that: the number of atoms in crystal and a is the atomic spacing. To account for the fact that
the crystal has finite size, we need to impose the following boundary condition on the wave function.
Ψ ( x )=Ψ ( x+ L)
(1.3)
This results in only certain discrete values of k being allowed. It can easily be verified that
the previous B.C. leads to k values of the form (apply this B.C. to the free electron solution):
(1.4)
2π 2π
k =n in other words, 4k between states in k-space is .So we have Fig. 1.2 Therefore the
L L
2π
spacing between allowed k valulog = due to finite crystal size L.A region of k-space of
L
Ω
volume Ω will contain allowed values of k.Or the number of k values per unit volume of
¿¿
Ω
k-space. density of k states) is where 3 V is the volume of the crystal (L3).
8π
Figure 1.2: Region of k-space
14
In one dimension, the density of k states in k-space is just L/2it as we have one state every
−π π
2π/ L. We also find that for a crystal electron K ranges from ¿ and we have that the
a a
number of states equals the length in k space (a/2πt) times the density of states ( 2π/L) for a
given energy band. Therefore, there are a total of
2π L L
⋅ = =M allowed Bloch wave functions or states in each band. The Pauli
a 2π a
Exclusion Principle says we can place 2 electrons in each Bloch state (2 because and each
atom will provide 2 electron states to each band.
Discrete states separated by
(1.5)
As M is very large the band is quasi continuous and size of crystal affects the state density,
Fig 1.3.
We can make the effective crystal very small and have only a few states per band. This is
used for band gap engineering with thin layers or small structures.
Figure 1.3:Statesin Bands
1-1.2.2 Insulators, metals and semiconductors

energy bands, separated by band gaps in which there are no allowed energies. We also know:
Each band can hold 2M electrons (M is the number of atoms in crystal)
14
·Bands fill up from bottom (lowest energy). As each atom has N electrons:
- If N is even then, N/2 bands are completely full and the material is most likely an insulator.
-If N is odd, the top band is half full and (N-1)/2 bands are completely full and the material is
a conductor.
Consider carbon in the form of a diamond crystal (16 electrons), where the band structure
depends on atomic spacing and therefore the crystal structure, Fig 1.4.Energy Band
Structure: (we only show the top two bands). For carbon, the band gap between the firs
empty band and the last filled band is 5.4eV.
conduction-band (allowed)
forbidden gap
(filled band →> no conduction)
valence-band (allowed)
Figure 1.4:Energy band of carbon
Carbon has just enough electrons to fill the valence band (electrons tend to occupy
lowest available levels) and seven lower bands, and the conduction band is empty. We
have assumed that all electrons go to the lowest available E level. However, heat
(temperature) is kinetic energy and some electrons will be raised to higher E states for
T > Ok (absolute 0). At room temperature, T=0.0259eV and is the average kinetic
energy of electrons (k is Boltzmann's constant = 1.38x10-23 J/K). Electron energies
are described by an energy distribution function that is a function of T. The larger T
the hotter (more energetic) the electrons are. As kT at room temperature <<E.g., of
carbon, we have that it is almost impossible for electron to acquire enough thermal
(kinetic) energy to be activated into the conduction band and carbon is an insulator.
Next consider silicon (14 electrons) Fig 1.5; same crystal structure as carbon, but
smaller band gap.
14
EMPTY
1.12eV
conduction-band
(allowed)
forbidden gap
valence-band full
(allowed)
˙
I >0T =0 Figure 1.5:Energy band of silicon
Now there is a small (but finite) chance for a few electrons to be thermally excited from
valence band to conduction band.
Finally consider an alkali metal Fig 1.6: Odd number of electrons: 1/2 filled band
filled band
Figure 1.6:Energy band of an alkali metal
1-1.2.3 Conduction
We know. that conduction is the movement, i.e., acceleration of an electron. To be
accelerated an electron must gain energy and change its state/momentum. If a band is
completely filled an electron cannot change state as all states are filled (grid in the
conduction band are free to move through crystal in response to electric field electrons move
in the conduction band.
14
Consider electron moving under influence of crystal potential Φ, and externally applied electric field
E.
Example:
Figure 1.7: electrons moving under influence of crystal potential o, and externally applied
electric field E.
−dϕ
Φ-electro-static potential and ϵ =
dx
We model the electron as a wave packet of Bloch waves (free electrons almost) cantered round wave
vector k. We expect an external field to accelerate he electron, increasing E and k and change the
electron's state, Figure 1.8.
Figure 1.8: E-k curve
We can write:
dE dϕ Δ x dω
E=−qϕ , =−q , , andw V s = =
dt dt Δ t dk
Rewriting we have:
dE dϕ dx dϕ
=−q ⋅ =−q ⋅V ,V is the group velocity
dt dx dt dx s s
dE dk dϕ dω 1 dE
: =−q ⋅V g , but V R= =
dk dt dx clk h dk
18
d dϕ d dϕ
Vk⋅ hk=−q ⋅ V g ,∨ hk =−q
dt dx dt dx
dp d dϕ
For a free electron,hk =min cot ⁡cot ⁡tan= p sin ⁡d = (hk)=−q =−qϵ=F , Fis
dt dt dt
the "crystal momentum" which changes in response to external forces. We can also dcrive:
d Vk 2
1 d E d
2
1 t E
Buta= =( 2 2
) hk=( 2 2
)(−qϵ)
dt h c k dt h dk
F ¿
This has the forma= ¿ where m =¿ which relates the curvature of
∈
the band to mobility and to “effective mass". We can show that a frce elcctron “band” gives
an effectve mass equal to that of an electron. Thercfore, clectrons in a crystal are accelcrated
in response to an external forcejust as though they were free clectrons with cffective mass
m*.
Consider the following E-k curve:
Suppose there are

electrons in these states
2
d E
2
<0
dk
Figure 1,9; E-k curves
18
18
These electrons have negative mass, move backwards in response to force.We interpret this as a +ve
mass particle with a +ve charge and it's called a hole!
1-1.2.4 Electrons and holes
The energy band structure of real semiconductors can be very complex. However to understand many
properties it is often enough to considera band structure as shown in fig 1.10 below. We might regard
this band representation as being typical for a semiconductor:
conduction-band
valence-ban aud
Figure 1.10: Band structure of a semiconductor
At OK in a pure material, conduction band is completely empty, valence band

completely filled, and no conduction (insulation). At 300K, a few electrons are
promoted to the conduction band, leaving a few empty states near the top of the
valence band.
At T'OK the charge carriers in a solid are in constant thermal-motion. In the absence
of external fields, this motion is quite random resulting from carriers scattered by
collisions. As the motion is random there is no net displacement of charge in any
direction and no net current flow (I=0). Clearly, with a field the electrons excited into
the conduction band can contribute to current flow. How do the empty states in the
valence band contribute to current flow?
As there are now empty states in the valence band, the electrons in this band can also
occupying the vacant states,,
18
2
1 d E
2 2
<0 (1.6)
h dk
(Negative!) and the electrons will move in sume direction as clectric field (wrong
way!)
Therefore,the electrons contribute to current as though thcy were positively charged, positive
effective mass particles. Il is therefore convenient to describe valence band as containing imaginary
positive charge, positive mass particles called holes.
In a semi-conductor, wetherefore have two charge carriers
- Electrons +ive mass, -ive charge (conduction band)
- Holes +ive mass, +ive charge (valence band)
1-1.2.5 Occupancy of available energy levels and filling the bands

We have determined E(k) and a density of states. We also nced a way of determining if a state is
filled or not at a given temperature. The distribution of the energies of a large number of particles
and its change with tcmperature.can be calculated by means of statistical considerations. The
probability of electrons occupying a level at energy E is given by;
Fermi-Dirac distribution function(1.7)

1
Where EF is a parameter called the Fermi energy, which represents the highest f (E)= (E −E F )/k x T
1+ e
energy that the electrons assume at T=(T≠0) the probability (f(E)) of occupation of the energy level
at E=Er is 1/2 (note there may not OK. At any temperature above absolute zero
be an actual state at the Fermi energy!) The function is plotted in fig 1.11 bclow.The higher the T the
more "smearing".
21
Figure 1.11:Probability (f(E)) of occupation of the energy level at 1 E=E F .
We can see from this figure that at T= 0 all levels that have an energy smaller than Thus if an
energy level is completely occupied by electrons, the Fermi distribution function
F(E) equals 1 (certainty); for an empty energy level one obtains F =0. The figure also shows that
the probability of occupation of states above EF is finite for T T > 0 K and that there is a
Corresponding probability thatstates below EF are empty.
1-1.2.6 Calculating n
To calculate the carrier concentrations in energy bands we need to know the following parameters:
(i) The distribution of energy states or levels as a function of energy within the energy band
(g(E)).
(ii) The probabiliry of each of these states being occupied by an electron (f(E)).So we can
write:
∞
n=dens electrons= ∫ ❑ f (E)g(E) dE
E=0
sity of ele (1.8)
where:f(E)-probability of occupancy at E and g(E) -density of states in energy at E

The hole density is given by:
p=den sin ⁡yofelectrons= ∬ ❑[1−f ( E)]g (E)dE (1.9)
ϵ →0
22
If the occupation probability of a level is given byf(E),the probability that this level will
not be occupicd by an electron, i.c., the probability of occupation by a hole is:(I -f(E)] -
probability of a missing electron at E (non occupancy)
1-1.2.7 Conduction band density of states
(Previously we calculated total number of states, not distribution.) Consider a

semiconductor for which the bottom of the conduction band has spherical constant
energy surfaces for free elecirons with a mass m*figure1.12.
(Spherical is 3D version of parabolic (k).)
E=
2
hk
2
k=
√2 m E Or k 2=k 2 =k 2 =k 2
¿
x y z
2 m∗¿ ,¿ h
Figure 1.12: Semiconductor for which the bottom of the conduction band has spherical constant energy
4 3
A sphere of radius k in k-space contains ( π x )⋅¿ allowed states (which are uniformly
3
distributed in 3D k-space). Therefore the number of energy levels per unit volume (divide
by L3) is:
(1.10)
4π 1
N ( E)=2. ⋅ The density of states in energy is therefore:
3 ¿¿
23
d 1
g(E)= N (E)= 2 ⋅¿ ¿ (1.11)
dE 2π
Note that the density of states (g(E)) may be defined as the number of cnergy stutcs
per unit energy per unit volume. Returning to the calculation of carriers
concentrations:.
We have that the density of elccuons in conduction band is:
∫ ❑E=0 f (E) g( E)dE

∞
ForE≫ EFW we use the following approximation for f(E) (Boltzmann distribution):
1 −(E− EF )/ kT −( E−EC )/kT −(E C−E F )/kT
f (E)= (E −F F )/kT
=e =e e
1+ e
which is valid for the tail of the distribution when Eis well away from EF.we get:
∞
∴ n=e−(E −E c F )/ kT
∫ ❑ g(E)e−(E−E )/kT dE c
(1.12)
E=0
Define:
−(E −E c )/ kT
N C = ∫ g( E)e dE (1.13)
The solution of this integral can be shown to be
1
N C= ¿ (1.14)
4
with effective density of states in conduction band and thus giving:
−(E c −E F )/ kT
n=N C e (1.15a)
−(E F −E v )/kT
We can also get for p: p=N V e (1.15b)
With
1
N c= ¿ (1.16)
4
24
SESSION 2-1: INTRINSIC SEMICONDUCTORS AND

EXTRINSIC SEMICONDUCTORS
2-1.1 Intrinsic Semiconductors

A perfect semiconductor crystal containing no impurities or lattice defects is called an
intrinsic semiconductor. As the carriers are generated in pairs, the concentration n.of
electrons in the conduction band equals the concentration p of holes in the valence band,
n=p=n i , ¿- intrinsic carrier concentration) we can use
−(E c −EF )/ kT −(EF − EV )/ kT
and,∴ N C e =N V e (1.17)
NC
giving,kT ln ⁡( )=EC −E F−E F + E V
NV
E C + E V kT NC
EF= − ln ⁡( ) (1.18)
2 2 NV
the second term on the right hand side is usually very small and Epi=1/2Eg.
Figure 1.13:Energy band diagram
we also have:
2
np=n1 (1.19)
called the mass action law.
ni = √ N C N V e
−(EC −EF )/ kT
And (1.20a)
Or
(1.20b)
ni = √ N C N V e
− E /2 kT
It is important in devices to control n and p
G
concentrations and suppress: the upper limits in semiconductor operating temperature. We

generallyrequire ni<<
25.
(minimum doping density in device) and, practically, this means we need doping
concentrations above 1014 cm-3
2-1.2 Extrinsic Semiconductors
To engineer semiconductors we modify n and p by adding impurities. Suppose an Si atom is

replaced with a phosphorus atom (an example of “doping”) We frequently
We see that the phosphorus ion core has an extra +ve charge - this means that the periodic
potential is disrupted and we get a localized energy level, Ep. The phosphorus also
provides an extra electron to fill this level. We find that E EC−E D ≈ 45meV that is very
small, figure 1.15.
Figure 1.15:Energy levels of a doped silicon crystal.
This electron is easily thermally donated to the conduction band at room temperature as
Ec-Ep is small compared with kT. This means that at room temperaturé n =ND-we call this
the complete ionization (only true if ni l i ≪N D ¿ .. Therefore by doping Si crystal with
phosphorus, we increase the free electron concentration -this is “n-type doping” Er moves
closer to Ec with increases in Np, figure 1.16 (Ep concapt for metals extended to other
materials, probability of occupancy =cy =0.5 ¿We write
−(E c −E F )/ kT
n=N C e
26
Ev
Figure 1.16: Energy band structure of a phosphorus doped silicon.crystal.
Still have np=n2i for the material without any applied bias. At constant temperature, p
is reduced as n increases andnp=n2i =const. At low temperature,electrons return to
donor levels; the dopant is frozen out. Next, suppose Si atom is replaced with Boron,
agin, we haye a perturbed lattice and a local E-level created.
Ec
Figure 1.17:Silicon crystal doped with boron and Energy levels of a doped silicon crystal.
Boron is missing an electron and accepts an electron from valence band, creating a hole.
Therefore doping with B increases hole concentration. We call this p-type doping, the electron
concentration n is reduced fig 1.18. Er moves closer to Ev.
Ec
Figure 1.18; Energy band structure of a boron dopedsillcon crystal.
p=Nif pi<<NA and at a reasonable temperature. At low temperatures, electrns p ≈ N A i

return to valence band, holes are frozen out,
27
2-1.2.1 Temperature dependence
Figure 1.19. Plot of electron density against temperature.
·At low temperature electrons drop to lowest E state.

·At medium temperature, donors/acceptors activated.
·At high temperature large amount of excitation over the bandgap(ni becomes large)
Calculation of temperature limits.

16
N D=1 ×1 0
Look at two cases as limits:
(1)n=N D /2−10 lower T limit. (Impurities freezing out)
(2)ni =N D /2−uupper T limit (Intrinsic concentration taking over)
Lower Limit
Define useful activation at 50% of donor levels occupied when Eg= Ep.We have typically
EC −E D=EC −E P=0.05 eV .
Therefore, n=N D /2=N c ¿

And (1.21)
15 15
5 ×1 0
EE 271 SEMICONDUCTOR DEVICES =5.4 × 10 ¿
Upper Limit-whenn1=N D /2
28
(1.22)
15 15
5 x 1 0 =3.33 x 1 0 ¿
T =675 K
Worked Examples
Useful equations:
˙
−(E C−E F )/kT −(E F −E v )/kT −( EF −E 1)/kT 2
n=N C e p=N V e Or n=ni e p=ni e
−(E1 −EF )/ k T
an np=n1
−¿ σr ¿
Charge neutrality relationship: p+ N A
+¿=n+ N A ¿
P+ N D=n+ N A ¿ionization)
An intrinsic Si sample (sample #1) at 300k, what n, p and Ep?

¿ 10 −3
Solution ⇒ n= p=ni=1. 5 1 0 c m , E F =Ei
If sample #1 is doped with 15cm3 donors.What's n,p and Er?
Solution
N D >ni
15 3 2
n=N D =6 ×1 0 c m , p=ni /n=¿
E F −ET =KT . ln ⁡(n /ni)=0.0259 ln

15
6 x10
( 10
)=0.334 eV
1.5 x 1 0
If sample#1 is also doped with 22 ×1 015 c m−3acceptors in addition to:6x1015cm-3donors,
what is n, p, and Ep?
29
Solution
15 −3 10 −3
n=N D −N A =4 ×1 0 c m ni=1.5 ×1 0 c m
15
4 x 10
E F −Ei=KT ln ⁡(n /n i)=0.0259 x ln ⁡( 10
)=0.324 eV
1.5 x 1 0
15 −3
If N D=N A =6 × 10 c m what is n,p,and Ep?
Solution
N D−N A =0 , dono nors and acceptors negate each other
SO n=p=n i , E F=Ei
The sample is compensaied and looks like intrinsic
General solutions to problems above:

Assume a sample with intrinsic concentration ni, doped with NA & Np ,we want to
calculate n,pand EF.
Solution
p+ N p =n+ N A (1)
2
np=ni (2)
from (2),we have p=n2i /n
(3)plug in(1)
we hanave n /n+ ND=n+ NA
2
1
2
⇒ n +(N A −N D). n−n1=0
2
N D −N A
⇒ n= +√ ¿ ¿
2
N A −N D
pi = + √¿ ¿
2
E F −ET =KT . ln ⁡(n /ni)
29
Temperature Dependence
If for some reason n; of sample #1 becomes 10cm3, and the sampleis still doped with
6x10'5cm-3 donors. What does the sample look like? n-type? p-type? or intrinsic?
Solution
S ∈ceND ≪ξ , , n= p=¿ , EF F=Ei so the sample is intrinsic.
31
CARRIER FLOW
SESSION 1-2: CONDUCTIVITY OF SEMICONDUCTOR AND

DRIFT CURRENT
1-2.1 Conductivity Of Semiconductor
We know that Bloch waves (which describe electron flow) move freely through a perfect
crystal.We also have the definition from classical physics:
F
Acceleration=Force/mass ie a= (2.1)
M
If the force is a constant, velocity (dv/dt) goes to infinity and current will also.
A few free electrons or holes would give infinite conductivity! What limits the
velocity? We do know that certain materials (metals, ceramics) under certain
conditions (usually low temperature) show zero resistivity, however, most materials
at reasonable temperatures show significant resistance to electron flow.
In practice, the crystal potential V(r) is not perfectly periodic:

lattice vibrates due to thermal energy ("phonons").
·ion cores are displaced from their equilibrium positions.
· lattice contains impurities (doping) and imperfections.
·all of these effects “scatter” the electron waves and limit the packet velocity.
We find that a model for electron motion is arandom walk figure 2.1 (not a straight trajectory)
32
electron
Figure 2.1:Model for electron motion
The electron is in random thermal motion (with a drift due to the e'field).Mean time between
collisions is tτ (∼ 10−13 s)(1) drift and (2) diffusion.
1-2.2 Drift Current
If an external field is present, it accelerates the electrons during time t between scattering
events figure 2.2.
Figure 2.2: Movement of electrons between scattering events
The average velocity acquired from field is:V = q E¿ τ =( qτ¿ ) E

m m
(2.2)
qτ
Where ( ¿ ) is defined as the electron mobility μn (c m V
2 −1 −1
s )and relates field E to velocity as:
m
q 2
v=μn , p E ¿< x ≥w= n
τ E ,the distance traveled at time t.
m
32
(2.3)
32
32
EE 271 SEMICONDUCTORI DEVICES
The model breaks down at high field, when energy acquired by electrons between collisions
becomes comparable to thermal energy. Then vis.limited by Veat the "scatter limited velocity"
fig2.3.
Figure 2.3:V is limited by V sat. the “scatter limited velocity”.
Many modern electronic devices work in the linear regime and it is important.to.have long t as
well as small m*, for high μ. We must therefore have clean material and a good band structure
(E-k relation).
To calculate the drift current carried by moving electrons; consider figure 2.4
Figure 2,4: An electron with an area A
In time Δt,all electrons within value vAt cross area A where v is the electron's velocity.
Asn=densi ity of electrons Δ Q=n ⋅v Δ t (q is the electron charge)
ΔQ
¿ ∈ingJ =curtentden sin ⁡ty= =qnv we get J=qn v
Δt ⋅ A
(2.4)
Therefore the Drift Currents in response to electric field is J n drif =q μ n n E
for electrons (2.5)

and J p drift=q μ p p E for holes (2.6)
And the total drift current is
32 33
→ →
J n , p drifi=(q μ p p +q μn n)E (2.7)
→
We can define a conductivity oo [Ω−1 m−1]∧J =σ E (OhmsLaw) Thus for
holes and electrons:
σ =q μ p p+q μ n n
(2.8
)
1 2 τ
forn−type σ = =cm μ n=n q ¿
ρ m
Consider figure 2.5
Figure 2.5: A piece of conductor
The total current is given by:

I =J × A=σ ⋅ A ⋅ E
V V
I =σ ⋅ A ⋅( ) E=( )
L L
Resistivity 1
we can also define the resistivity. ρ= [Ωm]
σ
Controlled by two factors:
o Carrier density
o Mobility: important also for frequency response f(frequency)
Two types of carriers:
J total=J elecrons +J holes
1
J T =q (n μn + p μ p) E and ρ=
q (n μn + p μ p )
(2.9
)
SESSION 2-2: DIFFUSION OF ELECTRONS AND HOLES

AND GENERATION/RECOMBINATION
2-2.1 Diffusion Of Electrons And Holes
The random thermal motion of electrons causes a net flow.of electrons from a region of
high concentration to region of low concentration fig 2.6.
of electrons
of electrons
Figure 2.6: Randorn thermal motion of electrons
dn
Mathematically we have J n dif =q Dn
dx
(2.10a) (2.10b)
dp
J p diff =q D p
Where Dmop is the
dx
diffusion coefficient. Consider an example of a linear dropin concentration
→
dn Δn
J n dif =q Dn =q Dn
dx L
Figure 2.7: Linear drop in concentration
36
kT
D and u are related as D= μ which is the "Einstein relation" from kinetic theory both
q
diffusion and drift depend on μ. The total current is sum of diffusion and drift in any
→ →
situation ie. J total=J drtji + J difuston
dp dn
and J Tρml =q (μ p p+ μn n) E+ q(D p + Dn ) (2.11)
dx dx
2-2.2 Generation/Recombination
2-2.2.1 Generation of electrons and holes
Generation:The movement of an electron from the valence band to the conduction band.
This results in the creation of an electron-hole pair. We have talked about the thermal
generation of holes and electrons but they can also be generated by light
(Photogeneration).
Photogeneration
Remember that E-k diagrams of real crystals (whose structure is still periodic but more
coiplex than a simple chain of atoms) are not simple parabolas.
Figure 2.8:E-k relationship for real solids.
Figure 2.8 shows the E-k relationship for real solids. (a) Silicon (which has an indirect
bandgap) and (b) gallium arsenide (which has a direct bandgap).
37
One point that arises is that the maximum of the valence band does not always occur at the
same k value as the minimum of the conduction band. We speak of a direct bandgap
semiconductor when they do and an indirect bandgap semiconductor when they do not.
Analyzing the two cases:
a) Direct gap material figure 2.9 (e.g. GaAs) -(min of conduction band at same k as max of
valence band)
Figure 2.9:Direct gap material
(Ak=0) Change of electron momentum is zero
A photon (a “particle of light") with energy E=hω can be absorbed by promoting a

valence band electron to the conduction band, creating an electron-hole pair. This is a
two body collision (electron,photon). Direct band gap materials have strong light
absorption and are modeled by an absorption parameter.
−αx
Model for light absorption fig 2.10:0 : I ( x)=I 0 e
Where a - absorption coefficient [cm]Io -incident intensity
Figure 2.10; Model for light absorption

b) Indirect gap material fig 2.11(c.g. Si) -(min of conduction band not! at same k
as max of valence band)
Figure 2.11:Indirect gap material
(Δk!=0) Change of electron momentum is non zero.

A photon carries little momentum, for à transition we need.Δk (a momentum change) and this can not
be provided by the photon itself.
Promotion of valence band electron to conduction band requires momentum transfer from crystal
lattice (interaction with a “phonon”); this is a relatively unlikely event as it is a three body collision
(electron, photon, phonon). “phohonon =quan lattice vibration ("particle" of heat) quantized
:. light absorption is relatively weak.

Note: all semiconductors are effectively - transparent for light withhv < E Gpossible transition).
absorb for hbforhv > EG(transitions allowed)
2-2.2.2 Recombination of electrons and holes
Recombination: The movement of an electron from the conduction band to the valence band. This
results in the destruction of and electron-hole pair.
Once created holes and electronscan recombine and be eliminated, in doing so, the energy released in
the recombination is given up as light (radiative recombination)or heat (non radiative recombination).
38
39
b) Indirect gap material fig 2.11(e.g. Si) -(min of conduction band not! at same k as max
of valence band)
Figure 2.11:Indirect gap material
(Δk !=0) Change of electron momentum is non zero.

A photon carries little momentum, for a transition we need Δk (a momentum change) and this
can not be provided by the photon itself.
Promotion of valence band electron to conduction band requires momentum transfer from
crystal lattice (interaction with a “phonon”); this is a relatively unlikely event as it is a three
body collision (electron, photon, phonon). “phonor uantized lattice vibration (“particle" of
heat)
:. light absorption is relatively weak.
Note: all semiconductors are effectively - transparent for light with hv < E G(no possible
transition).
absorbforhv > EG (transitions allowed)
2-2.2.2 Recombination of electrons and holes
Recombination: The movement of an electron from the conduction band to the valence band.
This results in the destruction of and electron-hole pair.
Once created holes and electrons can recombine and be eliminated, in doing so, the energy
released in the recombination is given up as light (radiative recombination)or heat (non radiative
recombination).
40
Figure 2.12: Direct radiative recombination
a) Direct radiative recombination (e.g. GaAs)

The conduction band electron fills the valence band hole, releasing a photon with the energy
Ephoton~Ec.(figure 2.12)
This is the basis of semiconductor LEDs and lasers, and is a very unlikely process in indirect
gap materials (need a phonon for momentum). Therefore can not make Si LEDs or lasers
easily, without modifying material properties.
b) Recombination through Midgap Energy levels (traps) (two stage processes).
Defect centers or traps are energy levels (ET) in the forbidden gap which are associated with
defect states caused by the presence of. impurities or Iattice imperfections.
Consider figure 2.13
1) trap captures conduction band
electron
40
2) trap transfers electron to valence band (capture hole).
·In silicon, midgap defect centers generally associated with transition inetal contamination: Cu, Fe, etc
like dopant.

40
·Traps are generally undesirable as we want long carrier lifetimes (elapsed time before recombination) in
most devices.
·Traps shorten lifetime, produce multiple frequencies and energy pathways (contribute to non-
radiative recombination).
Thermal Generation (trap aided)

Consider fig 2.14;
·heat (phonon) gives energy to electron.
·usually enabled by midgap energy levels (smaller energy).
B。
process repeats
1)trap captures valence band
electron(create hole)
2) trap emits electron to conduction band
Figure 2.14: Thermal Generation (trap aided)
40
displaced relative to those of the n-type side. This produces the Energy Band
Diagram for a p-n Junction shown in figure 3.4.
Figure 3.4:The same Fermi energy throughout the system at equilibrium
Majority carriers transfer to the other side because of difference in Fermi level
·A voltage difference appears because of the charge transfer
The voltage builds up until the currents are equilibrated
- At equilibrium the currents of minority carriers must balance those of majority
carriers.
1-3.2.1 After equilibrium
Bands are continuous through the semiconductor

Fermi.level is constant
The P side is charged negatively
The N side is charged positively
Equilibrium is based on balance of the majority and minority currents

The device is inherently asymmetric
45
Figure 3.5: Charge Density, Field,Potential V(x), carrier Density and Electron Energy
Diagrams
1-3.2.2 Built-in-potential
Can be obtained directly from basic definitions without resorting to Poisson's equation
and abrupt or non-abrupt case give same result. It is simply given as:
Barrier=Ei (P−side)−E i (N−side) (3.1)
Now consider semiconductor away from junction Ei ¿

Left: p ≈ N A ;
(3.2)
Right: n ≈ N D ;
Ei ¿ (3.3)
The built-in voltage is simply given by:

1
V bl = ¿ (3.4)
q
Vbr is called built-in potential
· OR Vb can be obtained simply from using the bulk potentials we defined earlier:
45 45
ND
q V n=E F−E i=kT ln ⁡ (3.5)
ni
ni
Hence q V p =Ei−E F =kT ln ⁡ (3.6)
NA NpN A
q V bi =q V n−q V p=kT ln ⁡ 2 (3.7)
kT ND NA ni
And V bi = ln ⁡ 2 (3.8)
q ni
Potential barrier acts like a dam
Majority carriers confined to one side of junction

Number of majority carriers able to overcome barrier equals number of
minority carriers on other side
We find that Vbi depends on doping and Temperatre
Carriers and Potential Barrier
A useful formula is derived from the usual carrier density relation:
(3.9)
1-3.2.3 Fields
·The depletion region gives rise to an electric field
·Fields goes from N to P region

Force opposes motion of majority carriers
Poisson's Equation
We use the basic electrostatic equation (3.10)to determine the depletion
width,built-in potential and the electric field.
ρ
∇ ⋅ E= (3.10)
ϵs ϵ0
ρ(x )=q [ N A (x )−N D (x )+ p (x)−n(x)]

· In 1 dimension:
(3.11)
-Constant charge density: ρ(x )=cons tan ⁡t , E (x)∝ x
- Zero charge density: zero field

Quantitative Relations
46
45
x n + x p=
Electric field (depleton extends from-x x p (0+ x n ):
−q N A
El (x )= (x p + x)
10
ϵs ϵ0
−q N D
En ght (x )=
' (x n−x)
ϵs ϵ0
N A x p=N D x n
NB:Field is zero outside depletion region.
Potential:
−q N A
V tef (x )= ¿
2ϵs ϵ0
V r¿ ¿
Bands Lineup
Define Electrostatic Potential V
1
V (x )= (E F −Ei (x))
q
Define the field as the gradient of the potential:

−d V bi (x)
E(x )=
dx
Define the charge density in regionsofE ≠ 0
Depletion Width
Depletion area width can be computed to yield:
x p=
√ 2ϵs ϵo N D
q N A (N A + N D )
V bl
W =x n+ x p =
√ 2ϵs ϵ0 N A N D
q (N A + N D )
V bi ×
48
(3.17)
(3.12) (3.18)
(3:13)
(3.14)
(3.19)
(3.15) (3.20)
(3.16) (3.21)
48
ϵ s =11.8, ϵ 0=8.85× 1 0−14 F /cm
W ≈ .2−20 μm
the other) abrupt junctions (abrupt transitions in doping from n to:p) and.the previous
equation reduces to:
√
(3.22) 2ϵs ϵo ˙
W= V bi , ,Where N B =N A N D > N A ∨N B =N D if N A > N D
q NB
Even for zero bias, there are electron and hole flows across the junction which exactly
balance, and
J n (dijf )+ J n (dijfusion)=0∧J n (drji)+J n (dijfusion)=0 foxV =0
Carriers and Potential Barrier
· A useful formula is derived from the usual carrier density relation:

n p =n j exp [(E F− Ei ( p−side))/ kT ]
n n=ni exp[(E F −Ei (n−side))/kT ]

And hence:
(3.23)
−q V kv / xτ
n p (−x p )=nn (+ x n) e (¿ n p 0) (3.24)
This relation shows. very directly it is the potential barrier that "holds back the majority carriers to
completely spill out on the other side.
Carriers and Voltage
Excess carriers are obtained from the barrier height:V b 1 →V b 1−V

qv
Leading to is the edge of the depletion layer on the p-side, and xnon the n-side):n (−x )=n e KT
p p p0
qV
n'p (−x p )=n p (−x p)−n p 0=n p 0 (e KT −1)(3.25)
(3.26)
nV
pn ( x n )=nn 0 e KT (3.27)
48
qv
p'n ( x n )= pn (x n )− pn 0= p k0 (e RT −1) (3.28)
1-3.3 Biasing
Formulae remain valid. Simply:

V bi →V bi −V applicd
· Depletion width:W ( V )=W (0)

√
V kl −V
V bi
=0.2−20 μm
(3.29)
qND qNA 3 5
Electric field: Emax =E (0)=¿ x n∨¿∨ x ∨¿ 10 −1 0 V /cm
ϵ y ϵc ϵz ϵe p
(3.30)
Dcfinitions:
- Forward Bias: lower barrier, lower field, narrower depletion layer
- Reverse Bias: higher barrier, higher field, wider depletion layer
1-3.3.1 Reverse bias
If we apply a bias VV <0 fi<0 fig 3.6, a very small current will flow.
Reverse bius
Figure 3.6: Reverse bias
Connecting the positive terminal of a d.c. source to the n side withdraws electrons and
holes from the depletion area which becomes wider and the potential barrier grows
higher figure 3.7.
50
Figure 3.7:Depletion area becomes wider and the potential barrier grows higher.
We no longer have an equilibrium situation. The barrier is now so high that few electrons can cross
from n-type to p type region reducing the diffusion current.However,electrons are still generated
(thermally) in p-type region (not many it's p-type). And holes are still generated in n-type region
(not many it's n-type).
There is also the generation of hole/electron pairs in depletion region and these cause small drift
current:Therefore we have a small leakage current as these electrons and holes are swept across.the
Dp Dn
depletion region (drift current). J o =q( L P n+ L n p)
p n
(3.31)
length,pn is the hole concentration.n the n-side and np is the electron concentration on the p-side. In
order to keep this current small, the minority carriers (pn and np)have to be kept at low levels. This
can be accomplished by selecting semiconductors having a large energy gap and by high doping.
The drift current is relatiyely insensitive to the height of the potential barrier since all of the
minority carriers generated may diffuse to the depletion region and be swept across it,whatever the
size of the e there.
1-3.3.2 Forward bias

Consider now the case when the n-side is connected to the negative terminal of the d.c. source
figure 3,8, the depletion region narrows.
51
Forward bias
Figure 3.8: Forward bias
This has the effect of lowering the height of the potential barrier to (Φ-V) figure
3.9.Consequently,majority carriers are able to surmount the potential barrier much more
easily than in the equilibrium case so that the diffusion current becomes much larger than
the drift current.
Flgure 3:9:Forward bias lower the height of the potential barrier (Φ-V)The diffusion current
J ∫ ∨¿∝ e
nv/kT
¿and the total current density is:
(3.32)
η/ kT
J=J o (e −1)As the equation 3.32.indicates, this
current is very strong temperature and. bias dependent. We can also see that for forward
bias (positive V) the net current increases exponentially with. voltage. For reverse bias
(negative V) the current is essentially constant and equal·to-J
The reverse bias generation current is small compared with the forward bias current.:A
typical I-V curve for a diode is shown in figure 3.10:
52
Figure 3.10:A typica I-V curve for a diode
Current-Voltage characteristic of a p-n junction diode. The reverse saturation current io is

equal J。 multiplied by the junction cross-sectional area(A).
53
SESSION 2-3: DIODE CURRENT, DC MODEL,

METAL-SEMICONDUCTOR JUNCTION AND
"OHMIC” CONTACTS TO SEMICONDUCTORS
2-3.1 Diode Current

Problem is easily solved by using the diffusion equation:
2 ' '
δ n p δn p
Dn 2
− =0 (3.33)
δx δ τn
The solution,for the current density,Jr:
(3.34)
2 2
Dn ni D pnl
J T =J n (−x p )+ J p (+ x n)=[q +q ]¿ Ln= √ Dn τ n∧. L p= √ D p τ p are the
Ln N A L p N p Where
diffusion lengths of n and p respectively
Or: (3:35)
VA
η
Where A is the cross sectional area I =I (e KT
−1)
T snt
And
(3.36)
·Forward current increases sharply with bias voltage
Reverse current remains constant until breakdown
Electrical Breakdown
At large negative voltages the current increases sharply

Cause is the generation process in the depletion region
Minority carriers are accelerated to very large energies as they “slide down”the
barrier
- One electron (1e) loses its energy by creating an e-h pair(2e+1h)
54
-Each electron creates another e-h pair(4e+3h)
- May become a runaway process: avalanching
Current eventually limited by series resistor
Breakdown may be a problem or a feature
Breakdown as a voltage regulator

Sharp current increase may be.used to provide a voltage reference
Variations in load resistance or voltage supply end up in small output changes Zener effect:
special case for high-doping
Figure 3.11: Diode in breakdown region
R_G currents quite important in Silicon diodes
- Carriers unable to overcome barrier may recombine in the depletion layers

-Recombination process via traps
Forward current effect
High-level injection:
Series resistance of neutral regions and of contacts becomes relevant at high currents:
V Dladt=V bt + Rs I D → V bl =V D−R s I D
(3.37)
Conclusions
Reviewed main aspects of diode response
Forward case:
-Series resistance
- High-level injection
- R-G currents
55
Reverse case
54
55
- Breakdown: avalanche,zener
Worked Examples
1. If the Energy Band Diagram of a Si sample is as follows Ee
a) Do equilibrium conditions prevail?
b) Sketch V(x)
c) Sketch ε(x)
c) Sketch n(x) and p(x)
Solution
a) Yes the Er has the same energy value as a function of position
b) If we assume the potentialatx=0 isV (0)=0 , then
x) has the sare functional form as the "upside down" of

−1
Ec (x) V (x)= (Ec ( x )−Eref ) reweselect Eref =Ec (0)
8
−dv (x)
c) ϵ (x )=
dx
.
56
d) n=ni exp ⁡[(EF −E i)/kT ] P=n1 exp ⁡[(E i−E F )/kT ]
17 −3 14 −3
2. A Si step junction operated at 300K with N A =1 0 c m andND=1 0 c m :Calculate the Fermi
level positions in the p and n regions.
a) Draw the band diagram for the junction and determine the contact
potentialV o ∨V bt
b) Calculate using equation.(5-8)

c) Calculate x no , X po∧ϵ o
d) Sketch )and charge density.
Solution
a)
pp 10
17
Eip −E F =KT . ln ⁡ =0.0259 . ln ⁡ 10
=0.407 eV .
ni 1.5 ×1 0
nn 10
14
b) F ∈
E −E =kT . ln ⁡
ni
=0.0259 . ln ⁡
1.5 x 1 0
10
=0.228 eV
57
KT NA ND 10
31
V bi = ln ⁡ 2
=0.0259 ln ⁡ 20
=0.635 V
q ni 2.25 x 1 0
2-3.2 DC Model
Applicable for very low frequencies where capacitance can be neglected
Piecewise linear model,VT=7V for Si,Slope determined by forward conductance

Application:rectifier
57
57
Figure 3.12: DC model
Figure 3.13:Circuit and Loadline
2-3.2.1 Small signal analysis

·The diode frequency response is controlled by Go-slope at Q-point
For a simple circuit with a vs AC signal over imposed on the operating supply:
Z D (ω)
v D =V D + v (3.38)
Z D (ω)+ R L s
Notice that the response may be highly asymmetric
Figure 3,14: DC and AC loadlines
57
2-3.2.2 Small-signal model
High-frequency model:
- Capacitance due to the space charge Cp
-Capacitance due to injected carriers: C1
-Conductance:Gp
All are voltage dependent, i.e., depend on the Q-point
Figure 3.15:Small signal model
Small Signal Analysis

Consider a diode biased at a Q-point (load-line analysis)
A small signàl is superimposed on the bias condition(Q-point): v D (t)=V D (Q)+v d (t)
(3.39)
In general the system will exhibit a Q-point dependent admittance:Y =G+ jωC
(3.40)
In reverse, G= and only the reactive part is important
Reverse Admittance
Small AC.signal va(t) will modulate the width of the depletion layer,leading to a small-signal
capacitance
ϵ jϵo
C j= A (3.41)
W
For an asymmetric jun ction p+¿−¿¿ − p
W=
√2ϵs ϵo
q NB
(V bl −V A )
(3.42)
After some manipulation we have a simple expression

C j0 for the capacitance of the junction:
C J=
√ 1−V A /V Di
(3:43)
60
·The capacitance is thus a direct function of applied voltage. This can be used in a circuit
varactors
The frequency dependence is also very important(microwave and fast logic)
2-3.3 Metal-Semiconductor Junction
2-3.3.1 Schottkey barriers (metal-semiconductor junctions)

These are devices based on a metal-semiconductor junction.They are the oldest semiconductor
devices used as rectifiers in “crystal” radios but were replaced by pn-junctions.
Band Structures
At surface of any solid there is an energy barrier holding in the electrons (work function).
2-3.3.1.1 metal band structure (halffilled band).

electron energy
Figure 3.161Metal band structure
om is the energy to remove the electron from the Fermi level to vacuum level (ionization
energy). The work function om is strongly dependent on surface preparation and is not a
precisely defined quantity (varies depending on surface).
Semiconductor Band Structure (conduction and valance band, n-type):
61
work function.Φ, depends on doping and it is.

more useful to deal with the electron affinity X
(energy to remove electron from conduction band)
X5~4JeV for Si
Figure 3,17: Semiconductor Band Structure
Both materials not touching (фm.>
Figure 3.18: Metal and semiconductor in close proximity
After the metal:and semiconductor have been brought into contact, electrons start to flow
from the semiconductor “down" into the metal until the Fermi energies of both solids are
equal, figure 3.19 (equilibrium condition).
·In semiconductor, a depletion region of width W is formed (uncompensated donors)

·In metal,electron current forms a negative surface charge layer
·An electric field is created
·An electric potential and thus a bending of the bands results
62
Figure 3,19: Metal and Semiconductor in contact (equilibrium condition).

We find that
(3.44)
N
√2ϵs V 0
˙
kT
V 0=(ϕ m −ϕ s )/q= ln ⁡( c ),∧. W = which mainly depend on doping
q ND qND
concentration (ND) and temperature. The equilibrium is reached when there are two equal and
opposite electron flows over the barrier aandJ =0 atV =0.
2-3.3.1.2 reverse bias (V
If the metal is connected to the negative terminal of a battery, the metal is charged even more
negatively than without bias. Thus, the electrons in the semiconductor are repelled even more, figure
3.20.
·depletion region becomes wider
· potenitial barrier is increased.
The electron flows are unbalanced since the Jsm→0.
63
Figure 3.20: Reverse bias
A current flows as a result of a few electrons in the metal acquiring enough thermal cnergy to
overcome barrier. This is a small current as barrier (фm - X) not changed,and JM-S is as before.
However, at high reverse bias, electrons can tunnel through barrier “field emission" and cause a
charge current to flow.
2-3.3.1.3 forward bias

Now consider the case when the polarity of the battery is reversed (metal connected to the positive
terminal of the battery).The depletion region is narrowed,figure 3.21.The potential barrier in the
semiconductor is reduced and electrons "spill" from semiconductor into metal. We get a large current
exponentially related to V flowing in the device. To compute the electron flow from semiconductor to
metal we use thermionic emission theory.
Figure 3.21:Forward bias

4 ¿ 2 2
q π k T −q (ϕ − x)/ kT
The current that flows from the metal into the semiconductor is J M →S = 3
e m
h
(3.45)
Defining:
¿ 2
q4 π n k −q(ϕ − χ)/ KT
C= 3 is Richardson's Constant, we have j M → S=C e
m
The
h
current flowing from the semiconductor into the metal is
−q (φ m−ϕ s− v)/ kT
J S −M =C e The net current is given by:
nV / kT
J nei=J s [e −1] (3.46)
Just like diode,where J S =C T i e−q(φ −ϕ )/kT m s

(3.47)
As typically there is very little minority carrier injection from semiconductor into metal;
Schottkey diodes are said to be majority carrier devices.
63
63
2-3.4 "Ohmic" Contacts To Semiconductors

In Integrated Circuits the external connections are always metallic: A metal-problematic.
Need to make a low resistance contact with linear V-I characteristics.
Figure 3.22:A metal-semiconductor junction
Consider now the case when Φm<Φs:
dnce
Figure 3.23: Bands of the metal and semiconductor placed together
Electrons flow from the metal into the semiconductor figure 3.23.
·In.metal,electron current forms a positive surface charge layer

·An electric field is created (in the opposite direction as before)
An electric potential and thus a bending of the bands results, figure 3.24
65
Figure 3.24: Bending of the bands
The bands of the semiconductor bend downward and no barrier exists for the.flow of
electrons in either direction. The current increases linearly with increasing voltage and is
symmetric about the origin.
Schottkey barrier what we need for an ohmic contact
Figure 3.25: I-V characteristics of a Schottkey barrier and an ohimic contact
Other applications:
Since only majority carriers are involved, no mutual annihilation (recombination) of

electrons and holes can occur. This results in faster devices.
The metal base provides better heat removal and is helpful in high power devices.
Used for detection of long wavelength IR (low energy photons).
This is useful n temperature sensors.
65
TRANSISTORS AND PHOTODETECTORS
SESSION 1-4: TRANSISTORS
Transistors are devices which control a large current with a smaller current or an (on/off
switching).There are two basic types: bipolar structures which use a back to.back diode structure and
FET (field effect transistor) which is typically a surface controlled device such as JFETs, and MOSFETs.
1-4.1 The Junction FET
First FET device was the Junction FET (JFET) whose structure is shown in figure 5.1. Two junctions are
formed on a Si substrate Reverse bias causes the depletion areas to expand and reduces the size of
conduction region.
Figure 4.1:The Junction FET (JFET)
66
67
In the linear region there is a continuous path of charge (majority carriers) from drain to source: this
is called the channel. As Va is increased the width of the channel decreases because the depletion
regions widen. We need to take into account the 2D character of the device. For ¿ V Ds> 0 ,,the
potential difference near the drain is given by Va - Vps. Increasing Vps increases the local reverse
bias and thus narrows the channel further near the drain; source side is not affected. At some critical
voltage Vp. the two depletion layers will merge and the channel disappear;further increases in Vps
will not affect much the situationV Dsat =V G−V P .Vp is the gate voltage for pinchoff whenV DS=0.
1-4.1.1 Linear equations

Equations describing the I-Vcharacteristic are complicated. Compute the local cross-section and
integrate: ¿
dV
I D ( y)=q μn N D S( y ) E=q μn N D Za(1−W ( y )/a) and the final expression looks
dy
like this: I D =I D (V GS−V DS)
(4.1)
This equation has the shape of an inverted parabola and reaches its maximum value at:
V Dsar =V G−V p . (4.2)
1-4.1.2 Saturation case
Saturation Voltage: V Dsat =V G−V P .

Saturation Current: Simplified saturation current (Square law)
I Dsat =I 0 ¿ (4.3)
The JFET is a non-linear resistor where R changes with voltage:
Two current sources in parallel
o. A conductance of value:
2 qZ μ n N d a
G= (4.4)
L
A.voltage-dependent source, with a relation:
68
(4.5)
I =−k ¿ The first corresponds to the resistance of the
channel, the second stemnsfrom the areas that are lost to conduction because of depletion.
1-4.1.3 AC response
˙
Small signal analysis: I
D (V DS , V DS )=I D (Q)+i d (v gs , v d 5 ).
(4.6)
˙ δ ID δ ID
I D (V DS , V DS)=I D (Ω)+ v 85+ v
δ V GS δ V DS ds
The small signal coeficients are thus: gd =¿
Drain conductance (4.7)
gm =¿ -Transconductance (4.8)
Finally we get for the ac component: i D =gm v gs + g d v ds (figure 4.2)
(4.9)
Figure 4.2:AC model
1-4.2 Bipolar Transistors
The structure is a back to back diode (ie npn of pnp). Two depletion regionis are bending
between the regions.There are three contacts an emitter,base, and collector figure 4:3. It is
typically the base that is the controlling contact.
68
Figure 4.3: Basic npn structure
11 Pn
Figure 4.4: NPN band structure in equilibrium with no applied biases.
At equilibrium (no applied bias) the fermi levels of all three regions must line up ;figure 4.4.
There is no net current through either junction. The junction widths and built in voltages are
simply calculated using the diode expressions. Across each junction there are two exactly equal
flows of both electrons and holes. As with the diodethe built in voltages across each junction are
precisely that needed to produce equal and opposite flows of both electrons and carriers and
guarantee that no cuirent is flowing.
There are many ways of biasing a BJT but the primary use is as an amplifier when biased in
forward active: We will use this as an example of the current transport in a BJT.In forward
active the Base-emitter junction is forward biased by Veb and the base collector is reverse
biased by Vob figure 4.5. This results in the lowering of the barrier height between the emitter
and the base and an increase in the barrier height between the base and collector.
70
EF
Figure 4-5:npn biased in forward active mode
The lowering of the base emitter junction barrier height wil unbalance the electron (electrons) into
the p type base. This injection of electrons is exponentially related to the applied base-emitter
voltage.
At the collector junction, the reversc bias provides a means to collect electrons that are present in the
base near the junction. Any electron that is in the base near the collector junction will fall over the
barrier and lower its energy.This results in very low electron (minority) concentrations in the base at
the collector junction figure 4.6.
Electron Transport Current in a npn transistor
Figure 4,6: Electron transport current in a npn transistor
In forward active the basic mode of operation is therefore.
71
1. Electrons are injected into the base at the.emitter junction. Causing a high concentration of electrons
in the base at that junction,n ≈ e q v / λT
kl
2.Electrons are collected at the collector junction as they fall over into the collector region. n ≈ 0
3. A diffusional transport current is present from the emitter junction through the base to the collector
current. If the base is thin then electron distribution is linear and
The net result of this is a transport current that is controlled by the base emitter current. The equation of
q v /kT
this transport current is: I C ≈ I E ≈ I 0 e
m
A closer analysis of the base emitter junction shows that there is a base current flowing that can
I 0 q v / kτ
be given by: I B ≈ e 1z
and we have: I C =β I B
β
A small base current can therefore control the collector current. β can be made large by
choosing a good clean material and making the base width small (large diffusion current). This
formula is only valid for forward active, but other modes of operation (cutoff, saturation, etc)
can be understood in a similar manner.
1-4.3 MOSFET Transistors

The MOSFET (Metal Oxide Silicon Field Effect Transistor) is a device that controls a current
between two contacts (Source ànd Drain) using a voltage contact (Gate).The device uses a
surface effect to create an n-type region in a p-type substrate figure 4.7(or the converse).
Applied Voltage>Vth
and a n region forms on

surface
Flow of electrons
Figure 4.7: Flow of electrons
oxide layer and a metal gate. If we apply a positive potential to the gate (the
72
substrate is grounded) electrons will be attracted to the gate and will pile up at the surface underneath the
gate.
At some voltage Vin called the threshold voltage the region under the gate will have The oxide is very
important as it stops the current flow towards the gate and forces
The basic MOSFET structure uses the capacitor structure figure 4.8 with n type regions placed at either
edge, known as the source and drain.
p substrate
Figure 4.8: The NPN MOSFET
and form an n-The

basic operation of the device is to bias the gate with V G >V TH type region between the source and the
drain. This provides a simple n-type path between the n-type source and drain regions for electrons to
flow figure 4.9. This region is called a channel. Note that with out forming the channel there are two
back to back diodes formed which will not allow appreciable current to flow.
Flow of electrons to the poaitive gate pobential.
Figure 4.9: Flow of electrons to the positive gate potential
73
The formation of this channel provides a simple resistive path between the.source and the
drain. The thickness of the channel is function of the difference between the gate potential and
the potential in the substrate near the surface. We can place a voltage between the source and
the drain and cause a current to flow.Typically,we ground the source and bias the drain with
Vas.The ability to change the thickness of the channel using the gate potentiál provides a means
of controlling the current from the source to the drain. We basically can form a voltage
controlled resistor. One thing to note is that the application of a drain voltage raises the
potential of the region of the substrate at the surface near the drain. This results in a thinner
channel at that end (as the gate to substrate potential is reduced).
Vdg<Vdh p substrate
Figure 4.10: No inversion layer
A new condition arises if we increase the drain voltage substantially ¿and VDs, sat is called the
saturation voltage). The drain voltage becomes.large enough that the gate to substrate potential
at the drain is smaller than threshold.Therefore,the channel thickness at this end goes to zero
figure 4.10. We call this pinch off.Electrically,the effect of pinch off is that the channel no
longer acts like a simple resistor.The current becomes fixed (saturated) at the value just prior to
pinch.off.
1-4.3.1 The MOSFET current-voltage characteristics
The MOS is normally considered with source shorted to the bulk and grounded and drain at
supply voltage. The behavior of the transistor is controlled by the voltages between gate and
source (bulk) VGs and between drain and source Vps.The oxide,isolation. The drain current 1
I DS =I DS ¿VDs)is a controlled voltage
74
transistor.The MOSFET is therefore much more of a programmable resistor than the BJT.
The operation is very similar to JFET,three operational areas; namely Cutoff: Ios=0,Linear
region:I: I =I (V DS)and Saturation: where Ips=constant.
At cutoff: The device is in accumulation and no conducting path exists. The current is the reverse
current of the two junctions biased in reverse.
Linear region: In this region.V CS > V TH H but the two dimensional nature of the device complicates
the analysis. At any point under the gate, the surface potential is the sum of two contributions
contribution from the gate voltage and thatfrom ohmic drop along channel from current flowing.
Thus surface potential varies from point to point under gate, the amount of charge and width is
changing along the channel;and current is proportional to the amount of charge under the gate which
is also proportional to VGs-VTH.That is, larger gate voltages produce more charge and hence more
current but Vps counteracts VTH-
In Saturation: When the local voltage near the drain becomes less than VTH-the inversion disappears
and the channel is pinched-off. The curren continues to flow where the depletion region acts as a high
resistance section of the channel. Further increases in Vps simply increase the width of the Pinched-
off area. If Lpinch off<<the current saturates. of ≪ L
1-4.3.2 Computation of MOS currents
Total current (not density) is continuous, depends on how much charge is available
to conduct.We need to decompose in x (across channel) and y (along channel). The

x-field determines amount of charge and y -field determines current density.
The current density is given as in equation:
(4.10)
δϕ
J n ≈ q μn E=−q μn n The total current is therefore given by equation where W is the width
δy ¿ δϕ
of the device I D =−W μn Qn
δy
(4.11)
But Ip is independent of y (charge conservation) and hence,
75
L vp
∫ ❑ I D dy=I D L=−μ n W ∫❑ Qn dϕ (5.12)

0 0
And
vp
−W μ n
L ∫
I D= ❑ Qn dϕ (4.13)
D
But for simplicity we approximate the total charge as that under the gate which becomes:
Qn ( y )=−C ox (V ss −V m −Φ)
(4.14)
ϵ ox ϵ o
Where Cox is the oxide capacitance per unit area, given by: C ox = ∧t ox=¿oxide thickness.
t ox
By integration of Ip we obtain:
W μ n C ox ˙
2
I D= [(V GS −V TH )V DS−V DS /2] (4.15)
L
And the threshold voltageVTH is given as:
ϵ t
V TH =2Φ B + s ox
ϵo
4qNA
√ Φ
ϵs ϵo B
Example fo¿ N A =1 0 , φ B=0.288 ,V th =0.978V
15
(4.16)
From equation (4.155), we note the following:

·Two terms in the current expression: one increases. withps,the other decrease.
Maximum is reached at
W μn ⋅COX
Then I D S = ¿
SAT
L
Example: Cox=34.4 ×1 0−9 V DSS Δ t =2.03V , I DSSat (V GS =3V )=7.0 × 10−5 A
Therefore the equations for the MOSFETare:
1. The triode regicon → V GS >V c h1 V DS <V D Si sat
2
I DS =I 0 [(V GS−V ¿ )V DS −V vs /2] (4.17)
75
The V2os/2 term takes the narrowing of the channel at the gate region into account as Vps
approaches Vps,sat
2. The saturation region η−V GS >V t / 3 V DS >V DS ,sat
V DS ,Sat =V GS−V th , ∧I DS , Sat =I 0 ¿ ¿
It can be shown that (4.18)
In saturation, the MOSFET acts like a nonlinear voltage controlled current.
1-4.3.3 Small signal analysis
capacitor thus it is an open circuit at low frequency. The output is:

i D (v DS , v GS)=I D (V DS , V GS)+¿ (4.19)
(4.20)
1
i d (v DS , v GS )=g d v dS + gm v gs = v + g v Where ga and gm are the drain conductance and
r d ds m gs
transconductance respectively;and ra is the drain resistance shown in figure 4.11.
Figure 4.11: Small signal model of MOSFET
W μ n C ox
Below pinchoff: gd = (V gs−V th −V ds )
L
(4.21)
W μn C ox
gm = V ds (4.22)
L
Above pinchoff:
75
gd =0 (4.
W μ n C nx W μn C ρx
8 m= (V gs−V Ih )= V ds(sn) (4.2
L L
Transit Time:
In digital logic the transit time may be more interesting than the cut off frequer
2
L L L
t t= = = (4.2
v drf μn E μ n V ds(sωt )
78
Worked Examples
Example 1
Find the maximum width of the depletion region for an ideal MOS capacitor on p-type Si
16
withh N a =10 cm
Solution
The relative dielectric constant of Si is1611.8. We.get Φr from

kT N a 10
Thus Φ F = ∈ =0.0259∈ 10
=0.347 V
q nf 1.5 ×1 0
W m =2
√ ϵ s Φr
qNa
=2¿
−5
¿ 3.01 ×1 0 cm=0.301 μm
Example 2
Using the conditions of example 1 and a 100-Å thick SiO2 layer, we can calculate major points on
the C-V curye. The relative dielectric constant of SiO2 is 3.9.
Solution
ϵ 1 (3.9)(8.85 × 10−14)
C 1= = −6
=3.45 ×1 0−7 F /c m2
d 10
−19 16 −4
Qd =−q N a W m=−(1.6 ×1 0 )(10 )(0.301 ×1 0 )
−8 2
¿−4.82 ×1 0 C /c m
−Qd 4.82× 1 0
−8 ˙
V T= + 2 ϕF = −8
+ 2(0.347)=0.83 4 V
Ci 34.5 × 10
At maximum depletion
79
ϵ s (11.8)(8.85 ×1 0−14 )
C d= = −4
=3.47 ×1 0−8 F /c m2
Wm 0.301 ×1 0
C i C i 34.5 ×34.7 −8 −8 2
C min= = ( 10 )=3.15 ×1 0 F /c m
C i+ Cd 34.5+3.47
Example 3
CalculateVfor the MOS structure described in example 1 and 2, including the effects of flat band
−3
voltage.If nn+polysilicon is used for the gate. Φ ms=−0.95 Vfo N a =10 c m .Assuming an interface
16
10 2
charge oof 5 ×1 0 q(C /c m ).
Solution;
1
V T =Φ ms+2 ϕ F− (Q −Qd )
C1 i
10 −19 −8
(5 ×1 0 ×1.6 ×1 0 )−4.82× 10
¿−0.95+0.694−
34.5× 1 0−8
¿−0.14 V
Example 4
16 −3 10
Consider an n+polysilicon SiC -p-channel device with N d =1 0 c m ∧¿ Q i=5 ×1 0 ㎡.
Calculate Vrfor a gate oxide thickness of 0.01 jum and repeat for.a field oxide thickness of 0.5 μm.
Solution
Values of ΦF, Oi and Qu can be obtained from examples 2 and 3 if we use.appropriate
signs. The value. of Ci for the thin oxide case is same as in example 2.Φ ms=−0.25 V .
−9 −8
8× 1 0 + 4.82×1 0
V τ =−0.25−0.694− −8
=−1.1V
34.5 × 10
In the field region wher ℜ⋅d =0.5 μm,

−8
5.62× 10
V T =−0.944− −9
=−9.1 V
6.9× 1 0
Example 5
For the p-channel transistor of example 4, calculate the boron ion dose FB (B+ions/cm2)
required to reduce Vr from - 1.1V to - 0.5V. Assume that the implanted acceptors form a
sheet of negative charge just below the Si surface.
Solution
q Fg
−0.5=−1.1+
Cf
−7
3.45 ×1 0 12 −2
F B= −19
(0.6)=1.3× 1 0 c m
1.6 ×1 0
For a beam current of 10μA scanned over a 650 c㎡ target area,
−5
10 (C /s) 12 2 −19
2
t (s)=1.3 ×1 0 (ioms /c m )× 1.6× 1 0 (C /ion )
650 c m
The implanted time ist=13.5 s .
80
81
Practice Exercises
1. a) Give two or three example of an n-type and a p-type dopant for Si n−type=¿
p−type=¿
b) Explain (briefly) why carrier mobility usually
i) decreases with increasing doping concentration
ii) decreases with increasing temperature
c) Briefly explain the difference between a shallow impurity and a deep level impurity in terms of how readily they
ionize in Si
16 −3
d). A bar of Si is doped with a donor concentration N D=1 0 cm 3.Describe (and justify your answer) the hole
and electron concentration at:
i) T =0
∘
ii) T =30 0 K
∘
iii) T =70 0 K
2 Consider a pn junction with
14 −3 15 −3
N A =2 ×1 0 c m nd N D =1 0 c m
a) Given the space charge distribution shown for the n-side,
sketch the distribution for the.p-side.
b) Sketch the electric field distribution,E(x).Write the

expression relating charge density, P(x),to the electric field
distribution at xatx=0 , E(0). I .Determine the value of
E(0).Use identical graph as in (a)
(c)Write the expression relating the built-in voltage,V to the
electric field distribution, E(x).Sketch the electrostatic
potential, VV (x )ComputeVo1, Use identical graph as in (a)
(d) Sketch the energy band diagram and label Ec,EF, Ev,Ei.
81
81
3. Imagine that at time t=0 we sandwich a 2 μm thick undoped Si slab between an n-doped
−6
τ n=1 0
What is the hole current density (magnitude and direction)across the undoped region at t=07
State any (reasonable) simlifying assumptions.
82
83
SESSION 2-4:PHOTODETECTORS
When a p-n junction is formed in a semiconductor materíal, a region depleted of mobile

charge carriers is formed with a high internal electric field across it known as the depletion
region. If an electron-hole pair is generated by photon absorption within this region, the
internal field will cause the electron and hole to separate as shown in figure 5.1.
J ph=−enφ
Figure 5.1: Separation of an electron and hole
Photons per second:φ

photocarriersgenerated
Quantum efficiency: ¿=
photonsabsorbed
We may detect this charge separation in two ways. If the device is left on open circuit, an
externally measurable potential will appear between the p and n regions.This is known as the
photovoltaic mode of operation.
On the other hand, we may short circuit the device externally (usually operated under reverse
bias) in which case an external current flows between the p and n regions figure 5.2. This is
known as the photoconductive mode of operation.
84
Figure 5.2:External current flows between the p and n regions
The current-voltage characteristics of a p-n. junction under various levels of illumination is

shown in figure 5.3. The dark characteristic (absence of light) is that of an ordinary p-n junction
diode. Under increasing levels of illumination the curve is progressively shifted downwards.
Figure 5.3: The current-voltage characteristics of a p-n junction under various levels of
illumination
sensitivity of the device.
The main drawback is the presence of a dark current which limits the ultimate
85
2-4.1 Structure Of pn Photodiode
Simple pn photodiode
Figure 5,4: Simple pn photodiode structure
For high speed operation the depletion region must be kept thin to reduce the transit.time of
photocarriers. On the other hand, to increase the quantum efficiency,the depletion region must be
sufficiently thick to allow a large fraction of the incident light to be absorbed. Thus there is a trade-
off between the speed of response. and quantum efficiency figure 5.4. The depletion width
W=
'
√ 2 ϵ s (V bi −V )
qNA
is therefore a strong determinant of efficiency and speed of the device:
By introducing an intrinsic layer of material in between the junction a p-i-n diode is created figure
5.5.The depletion region thickness (the intrinsic layer) can be tailored to optimize the quantum
efficiency and frequency response.
Figure 5.5:A p-i-n diode strncture
We have a nominally intrinsic"1” layer, but in practice very lightly doped fully.depleted
region is formed. Photogenerated carriers are swept out of depletion region by high e field
and ideally,carriers move at~Vsat.
2-4.2 Avalanche Detectors.

Useful internal amplificationis achieved in the avalanche photodiode. In this device,a basic p-n structure is
operated under very high (close·to breakdown ~100V)enable other carriers to be excited across the energy
gap by impact excitation. This results in an 'avalanche' of carrier multiplication,
Thus, one photon produces many electrons/holes and lots of current (high sensitivity).
Figure 5.6: The principle of operation of an avalanche photodiode
Figure 5.6 illustrates the principle of operation of an avalanche photodiode. An electron having reached the point A
has sufficient energy above the conduction band-bottom to enable it to excite an electron from the valence band into
the conduction band (CD). In doing so, it falls from A.to B.
(vii) Band shuefre

2-4.3 Photodetector Materials
NA > NA
Silicon:
band gap enerenergy=1.1 eV , ,sensitive for 00 bn
indirect gap
cheap,mature technology
x p > x nGermanium:
band gap eneergy=0.7 eV , works in telecommunications range
· direct gap forλ< 1550

difficult material on
NH
NH + ND √ v
1− H
v lh
x nB=x n √ 1−v ∥ bi
16 ⋅61 √ 1−410 ⋅8498 86
−11
¿ 39.68 ×1 0 cm.
x PB =x PO √ 1−vb /vhi
¿ 41.625 × 10
−8
√1−−
E,
ANS:0.0892cmΩ 2.
c. For the doped sample calculate the resistance if thelength of the sample is 0.2cm and the cross
sectional area is 6c㎡.
3.
ρl 0.0892 ×0.2
R= R=
A 6
ANS:2.97mΩ
d. Calculate the current through the sample if 10mV is applied across its terminals.
4
ANS:3.36A
QUESTION AND SOLUTIONS

Q1. Boron is implanted into one part of an n-type Si sample with doping concentration of 1016cm-3
forming an abrupt junction of square cross section with 1 016 cm area=2× 10−3 c m−¿ ¿cm3. The acceptor
concentration in the p-region is 4x1018cm-3.
1. Calculate Fermi level on the p-side and then on the n-side.
2. Calculate the built-in voltage.
3. Calculate the maximum electric field
4. Calculate the depletion width
5. Calculate the depletion capacitance.
SOLUTION
16 −3 18 −3
N D=1 0 c m N A =4 ×1 0 c m
ND 10
16
E −E
1, ¿ 1 =KTII ( )=0.0259∈( 10
)=0.347 eV In
n1 1.5 ×1 0
64
NA 4 × 10
18
E y −E Fy=KT ln ⁡( )=0.0259∈( 10
)=0.502cV
nj 1.5 ×1 0
le is 2.
its
NA −6
X n= W =33.25 ×1 0 cm
N A+N D
3.
−19 15
−q N D X n −1.6 ×1 0 ×1 0 × X n
ϵ= = −11
=−50.98 KV /cm
ϵ0 ϵs 8.854 ×1 0 ×11.8
4. W =
√
2ϵeϵr 1
q
( +
1
NA ND k
)V =0.33 μm
ϵ e ϵ r , A 8.854 × 10−4 ×11.8× 2× 1 0−3
1
5. C=(
W
)=
W
=62.66 pF
Q2. Calculate the conductivity of an elemental semiconductor silicon, Si at 300K for which the Fermi
level is at 0.36eV below the conduction band edge, assuming the effective density of states in the
conduction band is 2.82 ×1 019 c m−3-3 and the electron and hole mobility are 1350c ㎡ /Vs and 480c
㎡/Vs respectively.
h SOLUTION
Er −E j=0.36 2
μc =1350 c m /V s
2
μ p=480 c m /V s
−(K 1−F 2) −0.36

13 −3
n=N C e KI n=2.82× 1 019 × e 0.0259 n=2.59 ×1 0 c m
2 10 −3
np=ni ni =1.5× 1 0 c m p=¿ ¿
σ =q (μe n+ μ p p)
−19 13 6 −3 −1 −1
σ =1.602 ×1 0 (1350 ×2.59 ×1 0 +480 × 8.67 ×1 0 )=5.60 ×1 0 Ω m
Q3.What type of band gap material is:

1.Silicon
2pTuasr 目七〕2.
3.Which of the two semiconductor materials has better performance? Why?

4.Explain the principle behind lasers and light-emitting diodes in relation to band SaTmoms
NoIuT0s
Silion is an indirect band gap material.
Gallium Arsenide is a direct band gap material.
Gallium Arsenide is a better semiconductor material to silicon.
For Gallium Arsenide, the energy absorbed is only used to overcome the energy
gzp hence less energy is required by electrons in the valence band to be able to
move to the conduction band.
SOLVED QUESTIONS AND ANSWERS

1.A silicon sample is doped to 1017 Arsenic atoms/cm3. Calculate the
equilibrium hole concentration at 300K.
2. A silicon sample is doped to 1017 Arsenic atoms/cm3 calculate the Fermi-level E

relative to
3.Calculate the resistance of a silicon sample at 300K, 0.1cm long and cross sectional
area of 4 c㎡ doped with 1016 Boron atoms/cm3.
μ 2
Assumm e =480 c m / V S
n
4If carriers were allowed to defuse into a sample doped with 10 cm3phosphorus
estimate the diffusion co-efficient of the carriers. Given that
2 μ 2
μn=700 c m /V S an d p=300 c m /V S .
16 −3 17 −3
5.A Si p-n junction has on one side an N d =1 0 c m on the other N a =10 c m
side. At 300K, find the Fermi-level on the n-side.
t is the built-in potential for the device in question 5?
67
7. For the device in question 6, if the depletion layer of the p-side is 0.34km,what will be the
depletion layer on the n-side.
S. A p-type semiconductor with μ p=400 c m2 /V s and doping concentration of

1015carriers,carry a current of 1A/c㎡.What is the electric field across this sample?
9. Find the collector current of an n-p-n tranistor operation at I B=21 μA and ithα =0.99
2
10.An nMOS device with L=5 μm, W =100 μm , μ n=1000 c m /Vs and gate oxide capacitance
−8 2 16 −3
of 3.8 ×1 0 F /c m is made on a p-type substrate doped to 1 0 c m The saturation
transconductance measured at V gs=3V produced 1.50mS. Estimate the threshold voltage of
this device.
II.For the device in question 11, ifV gs=V ds=5 V is applied, what would be the source-
drain current if the device is operated in the linear region?
Solution
17 −3
n=N D =1 0 c m
2
np=ni
2
ni
z= =¿ ¿
n
−(E C−E F )/ KT
n=N C e
−(Ec − EF )/ KT
¿e
n −(EC −E F )
= ene
NF KT
n
F e e −E F ¿=KTen
NC
5
¿ [14 × 14 ÷ 14]1 ÷12 ÷ 14=12 11−1=12702 ÷ 6=¿
11−9=18
3
2.5+2.5=1.2
a b
y=−3−m a 1=a2 +b 3 12−2=185.5−2.5=1.5 =
b a
2.5−2.5=1.2
11−9=18 +1=18+1=18
1112151215181111111111111111111111111155511555552.5+2.5=1.2
.
202 ÷3=12012 ÷3=18
∘ ∘ ∘ ∘ ∘ ∘ ∘
9 0 −∠ BAC =9 0 −18 0 −90 −18 0 −18 0 −9 0
。 2. B . D . EF .
3=9 8 7 −0.5=18
¿ [18 ×25 × 25]×25AB÷ 25BC

¿ [1820 ×25 ]× 20 ÷25
1 1
1 1 =
.
11−9=18
÷ b= BC BC (2) + ¿
¿ [18 ×5 ]× 10=1818 ÷ 10=18 2 2
a a
. .
a b b b b a
= ⋅ ⋅ ⋅ ⋅
b a a a a b
12−0.5=¿
x 2=−2 x 2=2
1202 ÷3=1202 ÷3
¿ [14 × 14 ×12 ×2]11÷ 2
G
9
−202 ÷ 3=12
70
−g m × L
V th = +V gs
W μ n C ox
−(1.5 ×5)
V hh= +3 V m =−2.15+3
100 ×1000 ×3.48 ×1 0−8
V th =0.845 V
11.
2
W μ n C ox V ds
I D= [(V gs −V Ih)V ds − ]
L 2
−8 2
100 × 1000× 3.48 ×1 0 5
I D= [(5−0.845)5− ]
5 2
−4
I D =(6.96× 1 0 )(8.275)
−3
¿ 5.7594 × 10 A
15
12.A silicon step-junction diode is doped with N D=1 0 on the n-side and
20 −3
N A =4 ×1 0 c m on the p-side. At room temperature,calculate the
a. The built-in potential and
b.KTThe N
depletion
A ND
layer width and maximum electric field at the junction zero bias
V bi = en 2
q ni
20 15
( 4 × 10 )(1 0 )
V bi =0.0259 en
¿¿
(b) Depletion width W =

√ 2 ϵ r ϵ 0 (N A + N D )
qNA ND
V bi
−4 −6
W =1.089× 1 0 cm=1.089× 10 m
q N D χn
Emax (0)=¿ ∨¿
ϵr ϵ 0
70
70
70
−4 −6
χ n =1.090× 1 0 cm=1.090 × 10 m
−19 15 −4
(1.602 ×1 0 )(1 0 )(1.090 ×1 0 )
Emax (0)=¿ ∨¿
(11.8)(8.854 × 10−14)
4 6
Emax (0)=1.671 ×1 0 V /cm=1.671× 10 V /m
13.Consider a p-n,junction fabricated step diffusion of boron into a substrate

−3
wit h =2 ×1 0 c m .The surface concentration ofof 1 018 c m−3 and the junction depth is
ND 15
5 μm that the built-in potential is 0.8V, obtain capacitance at a reverse bias of 5V.
W=
√ 2 ϵ r ϵ o (N A + N D )
qNA ND
(V bi +V R )
W=
√ 2(11.8 )(8.854 ×1 0−14)(1 018 +2 ×1 015)(0.8+ 5)
(1.602 ×1 0
−19 18
)(1 0 )(2 ×1 0 )
15
−4 −6
W =1.947 ×1 0 cm=1.947 ×1 0 m
−14
11.8 ×(8.854 × 1 0 )
C j= −16
¿
1.947 ×1 0
14.Find the resistivity of intrinsic germanium and silicon at 200K. If a shallow donor impurity is
added to the extent of 1 atom per108.
1
ρ=
q (μn n+ μ p p)
But for intrinsic semiconductorr , n= p=ni
¿
1
ρ=
q n j (μ n+ μ p)
72
13
ForGe=ni=2.4 × 10 , μn=3900 , μ p=1900
1 1 1
ρ= = =4.484 ×1 0 Ω cm
q n (μn + μ p ) (1.602 ×1 0−19)(2.4 ×1 013)(3900+1900)
10
F ¿ Silcon=ni=1.5 ×1 0 , μ n=1500 , μ p=450
72
71
1 1 5
ρ= = −19
=2.134 ×1 0 Ω cm
q 1 p1 (μ n+ μ p) (1.602 ×1 0 )(1.5 ×1 0 )(1500+ 450)
10
15.A MOSFET with gate length of 0.33μm and gate oxide thickness of 60 is operated on a 2.5V rail
with a transconductance of 320mS/mm. If the threshold is 1/4of the supply voltage, calculate the
mobility of the carrier in the channel.
−6
L=0.33 × 10 m
∘
t ox =60 A
V g=2.5 V
V th =1/4 ×2.5=0.625
W μn C ox
gm = (V gs −V th )
L
gm −3
=320 ×1 0
W
C ox =oξde capacitance, t ox =oxidethickness
ϵ ox ϵ o
C ox = Where ntofSi O 2
t ox
ϵ ox =dielectric
−14
3.6 × 8.854 ×1 0
C ox = −8
60 ×1 0
−7
C ox =5.312× 10 F
gm L −3
320 ×1 0 × 0.33 ×1 0
−6
μn = = =1.006 ×
W C α (V gs−V lm ) 5.312 ×1 0−7 ( 2.5−0.625)
16 −3
16. Find the maximum width of the depletion region of an ideal MOS capacitor on p-type Si with N a =10 c m
KT N A
16
10
ϕ F= en =0.025 φn 10
=0.34 πr
q n1 1.5 × 10
73
W m =2
√ ϵ s ϕF
q Na
72
73
−5
H m =3.01× 10 cm
H r =0.301 μm
17.From question 16, calculate the total capacitance for a 1004° thick SiOz layer.
Cox C D ϵ ox ϵ o
Cf = C ox = constant of
C ex +C D t ox
ϵ¿
Wher e =dielectric
50:
−14
3.6 × 8.854 ×1 0 −7
C ox = −8
=3.45 × 10 F
100 ×1 0
ϵs ϵr ϵo
C D= =
H V max W max
−14
(11.8)(8.854 ×1 0 ) −6
C 0= −6
=3.47 ×1 0 F
0.301 ×1 0
−7 −6
(3.45 ×1 0 ) ×(3.47 ×1 0 )
C 7= −7 −6
=3.138 ×1 0−7 F
(3.45× 1 0 )+(3.47 ×1 0 )
I8.A germanium p-n junction diode at 300K gives a current of 1.0mA when forward
Hiased by 0.2V. What is the saturation current for the diode K=1.38 × 10−23 .
q V A / KT
I T =I SAT ( e −1)
−3
I =1.0 ×1 0 A
−23
K=1.38 × 10
T =300 K
V =0.2V
−19 −20
9 V =(1.602 ×1 0 )×(0.2)=3.204 × 10
−23 −21
KT =(1.38 ×1 0 )×300=4.14 ×1 0
74
−20
qV 3.204 ×1 0
= =7.739
KT 4.14 ×1 0−21
I −7
I SAT = 7.739
=4.35 ×1 0 A
e −1
19. The following parameters of a Schottky diode are given

19 3 16 3
ϕ m =4.7 V , χ=4.0 V , N C =10 c m , N D =1 0 c m , ϵ r=12
Assume the density of the interface states is negligible. At 300K, calculate barrier
height, built-in potential and depletion width at zero bias.
Barrier height, q ϕ B=q( ϕm − χ )
ϕ B=(ϕ m− χ )=4.7−4.0=0.7
q ϕ B=q(0.7)
−19 −19
q ϕ B=(1.602× 1 0 )( 0.7)=1.1214 ×1 0 V
q V bi =q ϕ B −q V n
NC 10
19
But V n=E C −E F =KT e n =0.0259 en( 16 )=0.179 V
ND 10
V bi =ϕ B−V n
V bi =0.7 eV −0.179=0.521 eV
W=
√ 2ϵr ϵo
V
q N D bi
W=
√ 2(12)(8.854 × 10−14 )(0.521)
(1.602 ×1 0
−19
)×1 0
16
=2.629 ×1 0 cm
−7
75
20. A silicon Schottky barrier has constant area of 0.01c㎡ and the donor
concentration in the semiconductor is 1 016 c m−3 let V bi=0.7∧V R =10.3 V .
Calculate
(a)The thickness of the depletion.
(b)The barrier capacitance.
(c)The field strength at the surface.
75
W=
√ 2 ϵ r ϵ 0 (V bi +V R)
qND
W=
√ 2(11.8 )(8.854 ×1 0−14)(0.7 +10.3)
(1.602 ×1 0
−19
)(10 )
16
−4 −6
W =1.197 ×1 0 cm=1.197 ×1 0 m
ϵr ϵo
C= A
W
−14 4
11.8×8.854 ×1 0 ×0.001 ×1 0 −13
C= −6
=8.73× 1 0 F
1.197 ×1 0
˙
2(V bi +V R ) 2(0.7+ 10.3) 6
Emax = = −6
=1.83 ×1 0
W 1.197 ×1 0
17 16 V bi
21.An Abrupt junction has N A =1 0 , N D=1 0 on the n-side at 300K. Find the Fermi-levels andd
ND
E F −Ei=KTen
ni
ni
Ei −EF =KTen
NA
E F −Ei=0.347
Ei −EF =0.407
ni
en
NA
p=KTe
10
1.5 ×1 0
q V p =0.0259 en( 17
)=−0.407 eV
10
76
q V n=0.0259 en ¿
16
10
n( 10
)=0.347 eV
1.5 ×1 0
q V bi =q V n−q V p
V bl =0.347+0.407=0.754 V
76
22. Find intrinsic carrier concentration
25 −3
N=3× 10 m , Eg =1.1 eV , μ e=0.14 μ p=0.05
Find the conductivity.

(E g/ 2 KT )
ni =N e
25 −1.1 /2 (0.0259)
ni =(3 ×1 0 )e
16 −3
ni =1.797 ×1 0 m
σ =q (μn n+ μ p p)
σ =q n i (0.14+ 0.05)
−19 16
σ =(1.602 ×1 0 )(1.797 ×1 0 )(0.19)
−4 −1 −1
σ =5.47 ×1 0 Ω m
23. Calculate the donor concentration in N-type Germanium, having resistivity of

100Ωm.
1
ρ=
σ
1 1
σ= = =0.01 σ =q μn n
ρ 100
σ 0.01 17 −3
n= = =1.734 ×1 0 m 24.A piece of p-type silicon, with a
q μ n (1.602× 1 0−19)(0.36)
cross section of 1mmx1mm and a length of 10mm, has a resistivity of 10-3Ωm at
300K.Ohmic contacts are fabricated on the faces at either end of the longest dimension of
the bar and an AC voltage of 10V is applied. Calculate the magnitude of the current,
μ 2 −1 −1
assume thata t =0.03 m v s Evaluate the current at 400K.
p
L=10 mm=0.01 m
−6
A=1× 10 m
I =J × A
I =σ × E× A
V
I =σ × ×A
L
1 10 −6
I= × × 1× 10
10
−3
0.01
I =1 A Let P400 and300 be the hole concentration at 400K and 300K respectively.
−(E −E )/ KT
p=N V e F V
−(E F −EV )/400 K

p400 =N V e
−(EF − EV )/ 300 K
p300 =N V e
p 400 −(E −EV )/ 400K −(EF − EV )/ 300 K

=e F
+e
p 300
p 400 (E −EV )/1200 K

=e F
p 300
1 1
P3300 = = −3
ρq μ p 1 0 ×¿ ¿
−(EF − EV )/ KT
p300 =N V e
−23 25 −(EF −EV )/ 0.0259

2.0827 ×1 0 =10 e
−23
2.0827 ×1 0
−(EF −E V )=0.0259 en( 25
)
10
E F −EV =0.1 eV
p 400 0.1 eV /1200 K
=e
p 300
77
0.1 eV /1200 K
p400 = p300 × e
23
P400 =2.0827 × 10 ×2.63107
23 −3
p400 =5.4793 ×1 0 m

Semiconductor Lecture Notes.

Uploaded by

Copyright:

Available Formats

Semiconductor Lecture Notes.

Uploaded by

Document Information

Original Description:

Original Title

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

Semiconductor Lecture Notes.

Uploaded by

Copyright:

Available Formats

PRESBYTERIAN UNIVERSITY COLLEGE GHANA

1-1.1 Introduction: Crystal Structure And Energy Bands

1-1.2 Electronic Structure

1-1.2.1 Number of states in a band

Figure 1.1:“free” Bloch electrons.

Figure 1.2: Region of k-space

Figure 1.3:Statesin Bands

1-1.2.2 Insulators, metals and semiconductors

Figure 1.4:Energy band of carbon

Figure 1.6:Energy band of an alkali metal

Figure 1.8: E-k curve

Consider the following E-k curve:

Suppose there are

1-1.2.4 Electrons and holes

Figure 1.10: Band structure of a semiconductor

At OK in a pure material, conduction band is completely empty, valence band

In a semi-conductor, wetherefore have two charge carriers

- Electrons +ive mass, -ive charge (conduction band)

- Holes +ive mass, +ive charge (valence band)

1-1.2.5 Occupancy of available energy levels and filling the bands

Fermi-Dirac distribution function(1.7)

Figure 1.11:Probability (f(E)) of occupation of the energy level at 1 E=E F .

sity of ele (1.8)

where:f(E)-probability of occupancy at E and g(E) -density of states in energy at E

1-1.2.7 Conduction band density of states

(Previously we calculated total number of states, not distribution.) Consider a

∫ ❑E=0 f (E) g( E)dE

The solution of this integral can be shown to be

SESSION 2-1: INTRINSIC SEMICONDUCTORS AND

2-1.1 Intrinsic Semiconductors

Figure 1.13:Energy band diagram

called the mass action law.

concentrations and suppress: the upper limits in semiconductor operating temperature. We

2-1.2 Extrinsic Semiconductors

To engineer semiconductors we modify n and p by adding impurities. Suppose an Si atom is

Figure 1.15:Energy levels of a doped silicon crystal.

Figure 1.18; Energy band structure of a boron dopedsillcon crystal.

p=Nif pi<<NA and at a reasonable temperature. At low temperatures, electrns p ≈ N A i

2-1.2.1 Temperature dependence

Figure 1.19. Plot of electron density against temperature.

·At low temperature electrons drop to lowest E state.

Calculation of temperature limits.

(1)n=N D /2−10 lower T limit. (Impurities freezing out)

(2)ni =N D /2−uupper T limit (Intrinsic concentration taking over)

Therefore, n=N D /2=N c ¿

EE 271 SEMICONDUCTOR DEVICES

An intrinsic Si sample (sample #1) at 300k, what n, p and Ep?

If sample #1 is doped with 15cm3 donors.What's n,p and Er?

E F −ET =KT . ln ⁡(n /ni)=0.0259 ln

EE 271 SEMICONDUCTOR DEVICES

N D−N A =0 , dono nors and acceptors negate each other

The sample is compensaied and looks like intrinsic

General solutions to problems above:

from (2),we have p=n2i /n

E F −ET =KT . ln ⁡(n /ni)

EE 271 SEMICONDUCTOR DEVICES

S ∈ceND ≪ξ , , n= p=¿ , EF F=Ei so the sample is intrinsic.

EE 271 SEMICONDUCTOR DEVICES