pn-junction principles

pn junction open circuit

Consider a sample of Si is doped ntype and the other p-type.

Assume there is an abrupt discontinuity

between the p and n regions, called meta
llurgical junction, M.
the fixed ionized donors and free
electrons in the n-region
fixed ionized acceptors and holes in
the p region.

Depletion region
Due to the hole concentration gradient
from the p-side (p= ppo) to then n-side (p= pno)
Holes diffuse towards n-region and recombine
with the electrons in this region.
The n-side near the junction becomes depleted
of majority carriers and therefore has exposed
positive donor ions (As+) of concentration Nd.

Similarly, the electron concentration

gradient drives the electrons by diffusion t
owards the p-side, which exposes acceptor i
ons (B) of concentration Nd in the region.

Depletion Region

Depletion region

Space charge layer (SCL)

Figs (a) & (b): The regions on both sides of the
junction M consequently becomes depleted of fr
ee carriers in comparison with the bulk p and n r
egions far away from the junction.
There is therefore a space charge layer around M.
Also known as the depletion region around M.

Fig (c): the hole & electron concentration

profiles
Fig (d): the net space charge density across the
semiconductor
Fig (e): the variation of the electric field across
the pn-junction
Fig (f): taking the potential on the p-side far away
form M as zero, then V(x) increases in the depl
etion region towards the n-side.

n
A s+

Bh+

(a )
e
M

Metallurgical Junction

Eo

N eutral p-region

E (x)

N eutral n-region

W p

Wn

(e)
(b )

Eo
V(x)

M
log(n), log(p)

Wp

Wn

Vo

Space charge region

(f)

ppo

nno
(c )

ni
pno

npo

eV o
H ole P E(x)

x=0

net

PE(x)

M
eN d

Electron PE(x)
W p

x
Wn

-eN a

(d)

eV o

(g)

pn-junction band diagram

p
Ec

Eo
M

(a)
n
Ec
Ec
EFn

EFp
Ev
p

SCL

(b)

Eo E

Ev

eVo
eV

EFp
Ev

Ec
EFn

e(Vo V)

Ev
p

I
V

Energy band diagrams for a pn junction under (a) open circuit, (b) forward
bias

Eo +E
Ec

Ec

e(Vo +Vr )

EFp
Ev

Ec
EFn

(d)

Eo+E

(c)

EFp
Ev

e(Vo +Vr )

Thermal
generation

Ec
EFn

Ev
p

Vr

Ev

Vr

I = Very Small

Energy band diagrams for a pn junction under

(c) reverse bias conditions. (d) Thermal generation of electron hole
pairs in the depletion region results in a small reverse current.
1999 S.O. Kasap, Optoelectronics (Prentice Hall)

Open Circuit
Given EFp and EFn are the Fermi levels in
the p and n sides, then in equilibrium and i
n the dark, the energy band diagram for o
pen circuit is shown Fig (a):

The Fermi level must be uniform through the

two materials
Far from the metallurgical junction M, we
should still have an n-type semiconductor and (
Ec EFn) should be the same as in the isolated ntype material
Similarly, (EFn Ev) far away from M inside the ptype material should also be same as in the iso
lated p-type material
Keeping EFp and EFn the same & the band gap
Ec Ev the same.
To draw the band diagram, we have to bend
the bands E and E near the junction at M

Forward Bias
When the pn-junction is forward biased,
majority of the applied voltage drops across t
he depletion region, Fig (b) shows the effect o
f forward bias:
The applied voltage is in opposition to the built-in
potential to reduce the PE barrier from eVo to e(V
o V)
The electrons at Ec in the n-side can now readily
overcome the PE barrier and diffuse to the p-si
de
The diffusing electron from the n-side can be
replenished easily by the negative terminal of the
battery and the positive terminal of the battery ca

Reverse Bias
When a reverse bias, V= -Vr, is applied to the
pn-junction the voltage again drops across
the SCL.

Vr adds to the built-in potential Vo so that the PE

barrier becomes e(Vo+Vr) as shown in Fig (c).
The field in the SCL at M increases to Eo+E
where E is the applied field
There is hardly any reverse current because if
an electron were to leave the n-side to travel t
he positive terminal, it cannot be replenished
from the p-side.
Virtually no electrons on the p-side

However, there is a small reverse current

arising from thermal generation of Electron-H
ole Pairs (EHP) in the SCL as shown in Fig (d).
The generated electron falls down the PE hill to the nside to be collected by the battery.

Light Emitting Diode

Introduction
One of the most popular
optoelectronics sources

Inexpensive, consumes very little power &

easily adaptable to electronics circuitry

In an LED, the semiconductor has a

high energy gap and the junction is con
structed so that radiation from the junct
ion can escape
In a normal Si diode, the radiated
wavelength is long (infrared range), the r
adiation is absorbed by the surrounding se
miconductor materials

Electrical characteristic
LED is a semiconductor diode.

Its characteristics and limitation

are similar to a normal p-n junction
diode

The breaking voltage is about

1.2 to 2V depending on the sem
iconductor material.
Dynamic resistance ranges from a
few ohms to tens of ohms.

IV characteristic of a p-n junction.

LED types

Infrared - 1.6V
Red - 1.8 to 2.1V
Orange - 2.2V
Yellow - 2.4V
Green - 2.6V
Blue - 3.0 to 3.5V (White same as blue)
UltraViolet - 3.5V

Principles of LED

Principles of LED
A LED is typically made from a direct
band gap semiconductor e.g. GaAs

in which the Electron-Hole Pairs (EHP)

recombination results in the emission of a p
hoton
The emitted photon energy is
approximately equal to the band gap energ
y, h Eg.

Energy band diagram of unbiased pn+junction device in Fig.1(1)

n side is more heavily doped than p side
The band diagram is drawn to keep the Fermi level
uniform through the device,
which is a requirement of equilibrium without bias.

The depletion region extends mainly into the pside

There is a Potential Energy (PE) barrier eVo from EC on
the n -side to EC on the p-side, Vo is the built-in
voltage
The higher concentration of conduction electrons in
the n-side encourages the diffusion from the n to the
p side.
The net electron diffusion is prevented by the

Fig.1: Principles of
LED
Electron energy
Ec
(1)

EF

n+

p
Eg

eVo

Ev

Ec
EF
eVo

n+

Eg
h Eg

(2 )

Ev

Distance into device

Electron in CB
Hole in VB
(1) The energy band diagram of p-n+ (heavily n-type doped) junction without any bias.
Built-in potential Vo prevents electrons from diffusing from n+ to p side. (2) The applied
bias reduces Vo and thereby allows electrons to diffuse, be injected, into the p-side.
Recombination around the junction and within the diffusion length of the electrons in the
p-side leads to photon emission.
1999 S.O. Kasap, Optoelectronics (Prentice Hall)

Energy band diagram of pn+-junction

device with a forward bias V in Fig.1(2)
As soon as a forward bias V is applied, this
voltage drops across the depletion region
since this is the most resistive part of the device.

The built-in potential Vo is reduced to Vo V

Allows the electrons from the n+ side to be injected
into the p-side
The hole injection component from p into n+ side is

much smaller than electron injection component from

n+ to p side

The recombination of injected electrons in the

depletion region & p-side results in the sponta
neous emission of photons

Injection
Electroluminescence
Recombination primarily occurs within the
depletion region and within a volume extending
over the diffusion length of the electron in the pside
This recombination zone is frequently called the active
region

The phenomenon of light emission from EHP

recombination as a result of minority carrier injecti
on is called injection electroluminescence
Because of the statistical nature of the
recombination process between electrons and hole
s, the emitted photons are in random direction
They result from spontaneous processes in contrast to
stimulated emission

Device structures

Fig 2: Device structures

Lightoutput

Lightoutput
p
n+

Epitaxiallayers

n+

n+

Substrate
(a)

Metalelectrode

Insulator(oxide)

p
n+

Epitaxiallayer

Substrate

(b)

AschematicillustrationoftypicalplanarsurfaceemittingLEDdevices.(a)player
grownepitaxiallyonann+ substrate.(b)Firstn+ isepitaxiallygrownandthenpregion
isformedbydopantdiffusionintotheepitaxiallayer.
1999S.O.Kasap,Optoelectronics (PrenticeHall)

GaP: Gallium Phosphide

GaAs: Gallium Arsenide

Device structures of LED

In its simplest technological form, LEDs are

typically fabricated by epitaxially growing do
ped semiconductor layers on a suitable substrat
e (e.g. GaAs or GaP) as shown in Fig.2(a).

It is formed by the epitaxial growth of first the n-layer

and then the p-layer

The substrate is essentially a mechanical

support for the pn-junction device and can be of
different material
The p-side is on the surface
from which light is emitted
it is also made narrow (a few microns) so that photons
is allowed to escape without reabsorbed

Low pressure chemical vapor deposition

(LPCVD) reactor for Epitaxially Growing Doped
Not only a film is deposited, but single crystal
growth must also be maintained

Fig 2: Device structures

Lightoutput

Lightoutput
p
n+

Epitaxiallayers

n+

n+

Substrate
(a)

Metalelectrode

Insulator(oxide)

p
n+

Epitaxiallayer

Substrate

(b)

AschematicillustrationoftypicalplanarsurfaceemittingLEDdevices.(a)player
grownepitaxiallyonann+ substrate.(b)Firstn+ isepitaxiallygrownandthenpregion
isformedbydopantdiffusionintotheepitaxiallayer.
1999S.O.Kasap,Optoelectronics (PrenticeHall)

Device structures, cont

To ensure that most of the recombination
takes place in the p-side, the n-side is heavil
y doped
Those photons that are emitted towards the nside become either absorbed or reflected back.
In Fig 2(a), the use of a segmented back
electrode will encourage reflections from the se
miconductor-air interface.

It is also possible to form the p-side by

diffusing dopants into the epitaxial n+-layer,
which is diffused junction planar LED as sho
wn in Fig 2(b).

Defect of Device structures

If the epitaxial layer and the substrate crystals have
different crystal lattice parameters
Mismatch between the two crystal structures will exist
This causes lattice strain in the layer and leads to crystal
defects.
Such defects encourage radiationless EHP
recombination acting as recombination center

Such defects are reduced by lattice matching the

LED epitaxial layer to the substrate crystal
AlGaAs alloys is a direct bandgap semiconductor in the
red emission region.
It can be grown on GaAs substrate with excellent lattice
match to produce high efficiency LED devices.

Fig. 3: Optical design

(a)

Light output

(b)
Light

p
n+

Domed
semiconductor

(c)

Plastic dome

pn Junction
n+

Substrate
Electrodes

Electrodes

(a) Some light suffers total internal reflection and cannot escape. (b) Internal reflections
can be reduced and hence more light can be collected by shaping the semiconductor into a
dome so that the angles of incidence at the semiconductor-air surface are smaller than the
critical angle. (c) An economic method of allowing more light to escape from the LED is
to encapsulate it in a transparent plastic dome.
1999 S.O. Kasap,Optoelectronics (Prentice Hall)

Optical Design of LED

Figs. 2(a) & 2(b) show the planar pn-junction
based simple LED structures.

Not all light rays reaching the semiconductor-air

interface can escape because of Total Internal R
eflection (TIR)
Those rays with incidence angle > critical
angle (c) will be reflected back as shown in Fig.3(
a)
For GaAs-air interface, c is only 16

It is possible to shape the surface of the

semiconductor into a dome/hemisphere

So that light strikes the surface at angles less than

Encapsulation of LED
The main drawback is

the additional difficult process in fabricating

such domed LEDs and
the associated increase in expense

An inexpensive and common procedure that

reduces TIR is the encapsulation of the sem
iconductor junction within a transparent pl
astic medium (epoxy)
has higher reflective index than air
has a domed surface on one side of pn-junction
as shown in Fig.3(c)

LED materials
There are various direct bandgap
semiconductor materials that can be re
adily doped to make commercial pn-jun
ction LEDs that emit radiation in the re
d & infrared range of wavelength

In visible spectrum is III-V ternary alloys

based on alloying GaAs & GaP donated as
GaAs1yPy.
In this compound, As & P atoms (Group V)
are distributed randomly at normal As sites
in the GaAs crystal structure.

Band gaps of some common semiconductors

relative to the optical spectrum.

Direct & Indirect bandgap

semiconductor
When y <0.45, the alloy GaAs1yPy is a direct
bandgap semiconductor and hence the EHP re
combination process is direct.

The rate of recombination the product of electron

and hole concentration
The emitted wavelength range from 630nm (red)
for y=0.45 (GaAs0.55P0.45) to 870nm (Infrared) for y = 0 (
GaAs)

When y >0.45, the alloy GaAs1yPy is a indirect

bandgap semiconductor.
The EHP recombination processes occur through
recombination centers

Isoelectronic Impurites
If isoelectronic impurites, N atoms (Group
V), is added into semiconductor crystal,
some of N atoms substitute for P atoms to fo
rm the same number of bonds.
The positive nucleus of N is less shielded
by electrons compared with that of the P
atom.

A conduction electron in the neighborhood will be

attracted and trapped at this site
Therefore N atoms introduce localized energy
level (electron traps), EN near the conduction ban

Fig.4: LED materials

Ec

EN

Ec

Eg
Ev

(a) GaAs1-yPy
y < 0.45

Ev

(b) N doped GaP

Ea

(c) Al doped SiC

(a) Photon emission in a direct bandgap semiconductor. (b). GaP is an

indirect bandgap semiconductor. When doped with nitrogen there is an
electron trap at EN. Direct recombination between a trapped electron at EN
and a hole emits a photon. (c) In Al doped SiC, EHP recombination is
through an acceptor level like Ea.
1999 S.O. Kasap, Optoelectronics (Prentice Hall)

Isoelectronic Impurites, cont

When the electron is captured at EN, it can
attract a hole in its vicinity by Coulombic att
raction
Eventually recombine with it directly and emit a
photon.
The emitted photon energy is slightly less than
Eg.

The recombination process depends on N

doping, it is not as efficient as direct recomb
ination

Thus the efficiency of LEDs from N doped indirect

band gap GaAs1yPy semiconductors is less than
those from direct band gap semiconductor

GaN: Gallium Nitride

Two types of blue LED materials

SiC: Silicon Carbide

1.

2.

GaN is a direct bandgap semiconductor with Eg

of 3.4eV
The blue GaN LEDs actually use the GaN alloy
InGaAs has a bandgap of about 2.7eV, which corresponds
to blue emission.

The less efficient type is the Al doped SiC, which

is an indirect bandgap semiconductor

The acceptor type localized energy level captures a

hole from the valence band
A conduction electron then recombines with this hole to
emit a photon
As the recombination process is not direct and therefore
not as efficient, the brightness of blue SiC LEDs is limi
ted.

Commercially important direct

bandgap semiconductor materials

Ternary (3 elements) and Quarternary (4 elements)

alloys based on III & V elements (III-V alloys)
Ternary alloy, Al1xGaxAs, in which x <0.43 are direct
bandgap semiconductors

The composition can be varied to adjust the bandgap and

hence the emitted radiation from 640nm-870nm (deep red
-infrared light)

InGaAlP is a quarternary III-V alloy that has a

direct bandgap variation with composition over the
visible range
It can be lattice-matched to GaAs substrate when in the
composition range In0.49Al0.17Ga0.34P to In0.49Al0.058Ga0.452P

This LED material is likely to dominate high

intensity visible range.

GaSb

In0.7Ga0.3As0.66P0.34

In0.57Ga0.43As0.95P0.05

InP

GaAs

InGaN
SiC(Al)

Indirect
bandgap

In0.14Ga0.86As

GaP(N)
GaAs0.55P0.45

Fig.5: Spectrum range

GaAs1-yPy
In1-xGaxAs1-yPy

x = 0.43

AlxGa1-xAs
In0.49AlxGa0.51-xP

0.6

Red

0.5

Blue
Green
Yellow
Orange

Violet

0.4

0.7 0.8
Infrared

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

Free space wavelength coverage by different LED materials from the visible spectrum to the
infrared including wavelengths used in optical communications. Hatched region and dashed
lines are indirect Eg materials.
1999 S.O. Kasap, Optoelectronics (Prentice Hall)

External efficiency
External efficiency ext of an LED quantifies the
efficiency of conversion of electrical energy into an
emitted external optical energy

ext

Pout optical

100%
IV

The input of electrical power into an LED is simply the

diode current and diode voltage product (IV).
Pout is the optical power emitted by the device.

For indirect bandgap semiconductor ext are

generally less than 1%, but for direct band gap se
miconductor with the right device structure, ext ca
n be larger.

Heterojunction high intensity

LEDs

A junction between 2 differently doped

semiconductor that are of the same materi
al (same band gap) is called a homojuncti
on.
A junction between 2 different band gap
semiconductors is called a heterojunctio
n
A semiconductor device structure that has
junctions between different band gap mate
rials is called a heterostructure device (
HD)

Refractive index and band

gap
The refractive index of a semiconductor
materials depends on its band gap.
Wider band gap semiconductor has a lower
refractive index

This means that by constructing LEDs

from heterostructures, we can engineer a
dielectric waveguide within the device
Thereby channel photons out from the
recombination region.

The homojunction LED has two

drawbacks.
1.

2.

The p-region must be narrow to allow the

photons to escape without much re-absorption

When p side is narrow, some of the injected electrons in

the p side reach the surface by diffusion and recombine t
hrough crystal defects near the surface
The radiationless recombination process decreases the
light output

If the recombination occurs over a relatively large

volume (distance), due to long electron diffusion
length, then the chances of re-absorption of emitt
ed photons becomes higher.
The amount of re-absorption increases with the
material volume

Double Heterostructure
LED constructions for increasing the intensity of the
output light make use of the Double
Heterostructure (DH) structure.
Two junctions between different semiconductor materials
with different band gaps

In Fig 6(a), DH consists of AlGaAs with Eg2eV and

GaAs with Eg1.4eV.

DH has an n+p heterojunction between n+-AlGaAs and p-GaAs

Another heterojucntion between p-GaAs and p-AlGaAs.
The p-GaAs region is a thin layer (a fraction of micron) and is
lightly doped

Fig. 6: Double Heterostructure

n+

(a)

AlGaAs

GaAs

AlGaAs

~ 0.2 m

Electrons in CB
EF
Ec

(b)

Ec

Ec
eVo

1.4 eV

No bias

2 eV

EF
Ev

2 eV

Holes in VB

Ev

(a) A double
heterostructure diode has
two junctions which are
between two different
bandgap semiconductors
(GaAs and AlGaAs)

(b) A simplified energy

band diagram with
exaggerated features. EF
must be uniform.

Fig. 6: Double Heterostructure

With
forward
bias

(c)

n+

(d)
AlGaAs

GaAs

AlGaAs

(c) Forward biased

simplified energy band
diagram.
(d) Forward biased LED.
Schematic illustration of
photons escaping
reabsorption in the
AlGaAs layer and being
emitted from the device.

Band Diagram of Double

Heterostructure
The simplifies energy band diagram for the
whole device in the absence of an applied
voltage is shown in Fig.6(b)

The Fermi level EF is continuous through the

whole structure
There is potential energy barrier eVo for electrons
in the CB of n+-AlGaAs against diffusion into p-GaA
s.
There is a band gap change at the junction
between p-GaAs and p-AlGaAs that results in a ste
p change, Ec.
Ec is effectively a potential energy barrier that
prevents any electrons in the CB in p-GaAs passin
g to the CB of p-AlGaAs

When a forward bias is applied,

majority of this voltage drops between
n+-AlGaAs and p-GaAs
Reduces the potential barrier eVo
This allows electrons in the CB of n+-AlGaAs to be
injected into p-GaAs as shown in Fig.6(c)
These electrons are confined to the CB of p-GaAs
since there is a barrier between p-GaAs and p-AlGaAs
.
The wide bandgap AlGaAs layers act as confining
layers that restrict injection electron to the p-GaAs
layer
The recombination of injected electrons and the
holes already present in this p-GaAs layer results in
spontaneous photon emission.
Since the bandgap Eg of AlGaAs is greater than GaAs,
the emitted photon do not get reabsorbed as they
escape the active region and can reach the surface o

Defect of Double Heterostructure

Since light is also not absorbed in p-AlGaAs, it
can be reflected to increase the light output
There is no lattice mismatch between the
two crystal structure in AlGaAs/GaAs heteroju
nction.
Negligible strain induced interfacial defects
(e.g. dislocation) in the device compared with
the defects at the surface of the
semiconductor in conventional homojunction
LED structure.
The DH LED is much more efficient than
the homojunction LED.

LED characteristics
The energy of an emitted photon from LED
is not simply equal to band gap energy Eg
because electrons in the CB are distributed in
energy and so are holes in the valence band.

Fig.7(a) and 7(b) illustrate the energy band

diagram and the energy distribution of
electrons and holes in the CB and VB respe
ctively
The energy concentration per unit energy in
CB is given by g(E)f(E)
g(E) is the density of states
f(E) is the Fermi Dirac function

LED characteristics, 1
In Fig 7(b), the electron concentration
in the CB as a function of energy is as
ymmetrical
has a peak at kBT above Ec.
The energy spread of these electrons is
typically about 2kBT from Ec.

Similarly, hole concentration spread from

Ev in the VB

Fig.7: LED characteristics

(a)

(b)
Electrons in CB

CB

2kBT
Ec
1

1/ k T
2B

Eg
3

Ev

VB

Holes in VB
Carrier concentration
per unit energy

(a) Energy band

diagram with possible
recombination paths.
(b) Energy
distribution of
electrons in the CB
and holes in the VB.
The highest electron
concentration is
(1/2)kBT above Ec

Fig.7: LED characteristics

(c)

(d)
Relative intensity

Relative intensity
Eg + kBT
1

(2.5-3)kBT
h

Eg

(c) The relative light intensity as a function of photon energy

based on (b). (d) Relative intensity as a function of wavelength in
the output spectrum based on (b) and (c).

The rate of direct recombination is

proportional to both electron and hole con
centrations
1. The transition, which is identified as 1 in Fig
7(a), has the relative small intensity of light
with photon energy hv1.

The carrier concentrations near the band edges

are very small and hence does not occur frequen
tly

2. The relative intensity of light corresponding

to transition hv2 is maximum
The transitions that involve the largest electron
and hole concentration occur most frequently.

3. The light intensity at the relative high

photon energies hv3 occurred through transit
ion 3 is small.

Output Spectrum

The relative light intensity vs photon energy

characteristic of the output spectrum is
shown in Fig 7(c).

It represents an important LED characteristic

Given the spectrum in Fig 7(c), we can also

obtain the relative light intensity vs wavelen
gth characteristic as shown in Fig (d) becaus
e =c/

The linewidth of the output spectrum, or ,

is defined as width between half-intensity points.

Fig.7: LED characteristics

Relative
intensity

(f )

(e)

(g)
V

Relative light intensity

655nm

1.0
0.5
0
600

24 nm

0
650

700

I (mA)
0

20

40

I (mA)
20

40

(e) A typical output spectrum (relative intensity vs wavelength) from a red GaAsP LED.
(f ) Typical output light power vs. forward current. (g) Typical I-V characteristics of a
red LED. The turn-on voltage is around 1.5V.
1999 S.O. Kasap, Optoelectronics(Prentice Hall)

Output Spectrum, 2

The wavelength for the peak intensity and

the linewidth of the spectrum are obviously
related to

Energy distributions of the electrons and holes

in the CB and VB
Density of states in these bands (individual
semiconductor properties)

The photon energy for the peak emission is

roughly Eg+kBT

It corresponds to peak-to-peak transitions in the

energy distributions of the electrons and holes
The linewidth (h) is typically between 2.5kBT to
3kBT as shown in Fig.7(c)

Output spectrum, 3
The output spectrum (relative
intensity vs wavelength
characteristics) from an LED depends

The semiconductor material

The structure of the pn-junction diode
including the dopant concentration levels

The spectrum in Fig 7(d) represents an

idealized spectrum without including
the effects of heavy doping on the ene
rgy bands

Red LED characteristics

Typical characteristics of a red LED
(655nm) are shown in Fig 7(e) to 7(g).

The output spectrum exhibits less

asymmetry than the idealized spectrum
The width of the spectrum is about 24 nm,
which corresponds to a width of about
2.7kBT in the energy distribution of the em
itted photons

LED current
As the LED current increases, so does the
injected minority carrier concentration,

thus the rate of recombination and hence the

output light intensity
However, the increase in the output power is not
linear with the LED current -> Fig.7(f)

At high current levels, strong injection of

minority carriers leads to the recombination t
ime depending on the injected carrier concen
tration
Hence on the current itself; this leads to a nonlinear recombination rate with current

Current-Voltage Characteristics
Typical current-voltage characteristics
are shown in Fig 7(g)

It can be seen that the turn-on or cut-in voltage is

about 1.5V (current increases sharply with voltage
)

The turn-on voltage depends on the

semiconductor and generally increases with
energy bandgap Eg
For example, typically it is about 3.5-4.5V for a
blue LED
It is about 2V for a yellow LED
It is around 1V for a GaAs infrared LED

Example LED output

spectrum
Given that the width of the relative
light intensity vs photon energy
spectrum of an LED is typically
about ~3kBT, what is the linewidth
in the output spectrum in term
s of wavelength ?
Given = 870 nm, 1300 nm, 1550
nm
T = 300 K

Solution
We note that the emitted wavelength is related to the photon energy E ph by

c hc

E ph

If we differentiate with respect to photon energy E ph we get

d
hc
2
dE ph
E ph
We can represent small changes or intervals (or ) by differentiate
e.g. /E ph d / dE ph then

hc
E ph
2
E ph

We are given the energy width of the output spectrum, E ph h 3k BT .

Then, using the latter and substituting for E ph in terms of we find,
3k BT
hc
Thus, at 870nm, 47nm
1300nm, 105nm
1550nm, 149nm
These linewidths are typical values and the exact values depend on the LED structure
2

Example: LED output wavelength

variations
Consider a GaAs LED. The band
gap of GaAs at 300K is 1.42eV, which
changes (decreases) with temperat
ure as dEg/dT= 4.5104 eVK1. What i
s the change in the emitted wavele
ngth if the temperature change is 1
0C?

Solution
Neglecting the k BT term and taking hc / E g we have,

d
hc dEg
6.626 10 34 3 108
4
19
2

4
.
5

10

1
.
6

10
2
dT
E g dT
1.42 1.6 1019
d
So that,
2.77 1010 mK 1 or 0.277nmK 1
dT
The change in the wavelength for T 10C is

d
1
T 0.277nmK 10 K 2.8nm
dT
Since Eg decreases with temperature, the wavelength increases with temperature.
This calculated change is within 10% of typical values for GaAs LEDs quoted in
data books.

Example: InGaAs on InP substrate

The ternary alloy In1xGaxAsyP1y grown on an InP
crystal substrate is a suitable commercial se
miconductor material for infrared wavelength
LED and laser diode applications. The device r
equires that the InGaAsP layer is lattice matc
hed to the InP crystal substrate to avoid cryst
al defects in the InGaAsP layer. This in turn re
quires that y 2.2x. The bandgap Eg of the tern
ary alloy in eV is then given by the empirical r
elationship,
Eg 1.35 0.72y + 0.12y2; 0x0.47
Calculate the compositions of InGaAsP ternary alloys for
peak emission at a wavelength of 1.3 m

Solution
We first note that we need the required bandgap E g at the wavelength of interest.
The photon energy at peak emission is hc / E g k BT . Then in electron volts,
hc k BT

e
e
and at 1.3 10 6 m, taking T 300K,
Eg

3 10 6.626 10 0.0259eV 0.928eV

1.6 10 1.3 10
34

Eg

19

The InGaAsP then must have y satisfying,

0.928 1.35 0.72 y 0.12 y 2
Solving this quadratic equation on a calculator gives y 0.66. Then x 0.66 / 2.2 0.3.
The quarternary alloy is In 0.7 Ga 0.3 As0.66 P0.34