Chapter 5

Optical Sources
General requirements for a light source for use
in optical communications
1. The emission wavelength compatible with the loss spectrum of glass fibers,
820nm, 1300nm & 1550nm.
2. The sources should be capable of modulation at rates in excess of 1GHz for
high data rate transmission.
3. The spectral width of the sources should be narrow in order to minimize the
bandwidth limiting pulse dispersion in the fibers.
4. The average emitted power of the source that is needed is typically few milliwatts,
although higher power values are needed for very long continuous fiber links or
if high loss fibers are used.
5. The radiance of the source should be as high as possible for effective coupling into
the low-loss fiber with small NA ( ~0.2). This means that the beam spread of
the sources must be minimized.
6. The sources must have long lifetime and it must be possible to operate the device
continuously at room temperature.
7. The sources must be highly reliable.
8. The sources should be reasonably low cost.

The principal light sources used for fiber optic communications applications are
heterojunction-structured semiconductor laser diodes or injection laser diodes
(ILDs) and light-emitting diodes (LEDs).
 TYPES OF OPTICAL SOURCES

• Wideband continuous spectra sources

(Incandescent Lamps)
• Monochromatic incoherent sources
(Light Emitting Diodes - LED)
• Monochromatic coherent sources
(Light Amplification by Stimulated
Emission of Radiation - LASER)
SPECTRAL LINEWIDTHS
 Absorption and emission of radiation

Quantum theory suggests that atoms exist only in certain discrete energy states. The
interaction of light with matter takes place in discrete packets of energy, called
photons.

Two level atomic system

Absorption

Spontaneous
emission

Stimulated
emission

The frequency of the absorbed or emitted radiation f is related to the difference in

energy E between the higher energy state E2 and the lower energy state E1 by the
expression:
 Spontaneous & Stimulated
Emission

Spontaneous emission - the atoms returns to the lower

energy state in an entirely random manner, gives
incoherent radiation.
Stimulated emission
„ The proton produced is generally of an identical

energy to the one which caused it and hence the light

associated with them is of the same frequency, f.
„ Light associated with the simulating and simulated
photon is in phase and has the same polarization.
"Coherence" is the term used to describe the
in-phase property of light waves within a
beam.
 The Einstein Relations

At thermal equilibrium, the population of the two energy levels of such

system are described by Boltzmann statistics which give:

where N1 and N2 represent the density of atoms in energy levels E1 and

E2, respectively,with g1 and g2 being the corresponding degeneracies of
the levels.
The absorption transition rate
Where is the spectral density of the radiation energy at the transition
frequency f and B12 is the Einstein coefficient of absorption.
The total transition rates from level 2 to level 1 (emission rate), R21, is
the sum of thespontaneous and stimulated contributions.
where A21 is the Einstein coefficient of spontaneous emission, is
equal to the reciprocal of spontaneous lifetime. The second term
which is the rate of stimulated downward transition is similar to that
of stimulated upward transition.
For a system in thermal equilibrium, R12=R21 and gives

When g12=g21, then the probabilities of absorption and stimulated

emission rate are equal, and the stimulated emission rate to the
spontaneous emission rate is given by

In thermal equilibrium, spontaneous emission is the dominant

mechanism.
For stimulated emission to dominate over absorption and
spontaneous emission in a two level system, both the radiation
density and the population density of the upper energy level N2 must
be increased in relation to the population density of the lower energy
level N1.
Population inversion

Boltzmann distribution for a system in

thermal equilibrium, the lower energy
level E1 of the two level system contains
more atoms than upper energy level E2.

To achieve optical amplification it is

necessary to create a nonequilibrium
distribution of atoms such that the
population of the upper energy level is
greater than that of the lower energy
level (N2 > N1).

This condition is known as population

inversion.

The excitation of atoms into the upper

energy level is referred to as “pumping”.
 Optical feedback

The basic laser structure incorporating plane mirror. The optical signal is fed
back many times whilst receiving amplification (stimulated emission) as it
passes through the medium. One mirror is made partially transmitting to
allow useful radiation escape from the cavity.
The structure acts as a Fabry-Perot resonator.
A stable output is obtained at saturation when the optical gain is exactly
matched by the losses incurred in the amplifying medium.
The major losses results from
„ Absorption and scattering in the amplifying medium
„ Absorption, scattering and diffraction at the mirrors
„ Nonuseful transmission through the mirrors.
 Laser oscillation

Oscillations occur in the laser cavity over a small range of frequencies where the
cavity gain is sufficient to overcome the various losses. Hence the device is not
a perfectly monochromatic source but emits over a narrow spectral band.
The central frequency of this spectral
is determined by the mean energy level
difference of the stimulated emission
transition.
Other oscillation frequencies within the
spectral band result from the frequency
variations due to thermal motion of
atoms within the amplifying medium,
known as Doppler broadening
(inhomogeneous broadening
mechanism) and by atomic collisions
(homogeneous broadening).
 Longitudinal or axial modes
Since the structure forms a resonant cavity, when sufficient population inversion
exists in the amplifying medium the radiation builds up and becomes established
as standing waves between the mirrors.

These standing waves exist only at frequencies for which the distance between
the mirrors is an integral number of half wavelengths. Thus when the optical
spacing between the mirrors is L the resonance condition along the axis of the
cavity is given by
where λ is the emission wavelength, n is the refractive index of the amplifying
medium and q is an integer.
The discrete emission frequencies is where c is the velocity of light.
The different frequencies of oscillation within the laser cavity are determined by
the various integer values of q and each constitutes a resonance or mode. These
modes are separated by a frequency interval or in term of free space
wavelength, assuming
LONGITUDINAL MODES
 Transverse modes

Laser oscillation may also occur in a direction which is transverse to the

axis of the cavity. This gives rise to resonant modes which are transverse to
the direction of propagation. These transverse electron magnetic modes are
designated by TEMlm where the integers l and m indicate the number of
transverse modes

Unlike the longitudinal modes which contribute only a single spot of light to
the laser output, transverse modes may give rise to a pattern of spots at the
output. The greatest degree of coherence, together with the highest level of
spectral purity is only obtained from a laser which operates in the TEM00 (the
lowest) mode as all parts of the propagating wave front are in phase.
Higher order transverse modes only occur when the width of the cavity is
sufficient for them oscillate. Consequently, they may be eliminated by
suitable narrowing of the laser cavity.
 Threshold condition for laser oscillation

Population inversion between the energy levels providing the laser transition
is necessary for oscillation to be established but it is not alone sufficient for
lasing to occur. In addition a minimum or threshold gain within the
amplifying medium must be attained such that laser oscillations are initiated
and sustained. This threshold gain may be determined by considering the
change in energy of a light beam as it passes through the amplifying
medium.
Consider an amplifying medium occupies a length L completely filling the
region between the two mirrors which have reflectivities r1 and r2. On each
round trip the beam passes through the medium twice and the fractional
loss incurred by the light beam is
Where is the loss coefficient per unit length (cm-1) for all the losses
except those due to transmission through the mirrors.
The increase in beam intensity resulting from stimulated emission is
exponential and the fractional round trip gain is given by
where is the gain coefficient per unit length produced by stimulated
emission.
At steady state conditions the laser oscillation are achieved when the gain
in the amplifying medium exactly balances the total losses. Hence

The threshold gain pre unit length can be written as

the second term represents the transmission loss through the mirrors.
 Optical emission from semiconductors

Direct bandgap semiconductors -

• electrons and holes on either side
of the energy gap have the same
value of crystal momentum, k.
• when e-h recombination occurs k
of the electron remains virtually
constant.
• the energy released, hf ~ Eg , the
bandgap energy, may be emitted
as light.
Indirect bandgap semiconductors-
• the maximum and minimum
energies occur at different values
of crystal momentum.
• the e-h recombination only
possible with the aid of a third
particle, a phonon.
• the recombination in indirect
bandgap semiconductor is
relatively slow.
 Optical emission from p-n diodes
Under forward bias carriers are
injected into the empty electron
states in the conduction band of the
p-type material and the empty hole
states in the valence band of the n-
type material.
• E-h recombination across the
bandgap released energy as a
photon

The optical wavelength is

This spontaneous emission of light

from within the diode structure is
known as electroluminescence.
The light is emitted primarily close to
the junction, recombination may take
place throughout diode structure as
carriers diffuse away from the
junction region.
Stimulated Emission in Semiconductor Diode
Population inversion may be obtained at a p-n junction by heavy doping of
both the p and n type material.
Heavy doping causes the Fermi level enter the conduction band of the n-
region and lowering the Fermi level into the valence band in p region.
When a forward bias ~ bandgap voltage is
applied, an active region exists near the
depletion layer that contains simultaneously
degenerate populations of electron and
holes, where population inversion is occur.
Any electromagnetic radiation of frequency
which is confined to the active region will be
amplified.
The degenerative doping distinguishes a p-n
junction which provides stimulated emission
from one which gives only spontaneous
emission as in the LED.
High impurity concentration within a
semiconductor causes differences in the
energy bands in comparison with an intrinsic
semiconductor
 Semiconductor Injection Laser
Electrons are injected into the device from the n-
type side - a current-controlled device.
Diode laser commonly takes the form of a
rectangular parallel piped >100mm to 1mm.
The junction is a plane within the structure.
Two of the sides perpendicular to the junction
are purposely roughened so as to reduce their
reflectivity.
The other two sides are made optically flat and
parallel, by either cleaving or polishing.
These two surfaces form the mirrors for the
laser cavity. The reflectivity of the
airsemiconductor interface is high enough that
no other mirrors are needed,
One of the reflecting surfaces may be coated to
increase the reflectivity and to enhance laser
operation.
The thickness of the junction region is small,
typically around one micrometer.
Light traveling in the plane of the junction is
amplified more than light perpendicular to it and
the laser emission is parallel to the plane of the
junction.
 Threshold Current
The rate equations for electron n, and photon density f, in the active layer of
the semiconductor laser are

where J is the injected current density (A/m2), e is the charge on an electron,

d is the thickness of the recombination region, is the spontaneous
emission lifetime, C is a coefficient which incorporates the B coefficients, δ is
the fraction of photons produced by spontaneous emission which combine to
the energy in the lasing mode and is the photon lifetime.
The fields in the optical cavity, Φ, is build up from small initial values with
when Φ is small. By setting δ = 0, we have
The threshold value for the electron density is
The threshold current density, Jth, required to maintain n = nth in the steady
state when Φ = 0, is and the steady state photon density is
 Lasing in an Injection Laser

Laser diode gives little light output in the region below the
threshold current which corresponds to spontaneous
emission .
Above the threshold light output is almost entirely due to
stimulated emission, where the laser is acting as an amplifier
of light.
The stimulated emission minority carrier lifetime is much
shorter ( typically 10-11s) than that due to spontaneous.
The coherent emission has a linewidth of a nanometer or
less, whereas incoherent spontaneous emission has a
linewidth of tens of nanometers.

The photon density is proportional to the amount by which J

exceeds its threshold value.
For strongly confined structures, the threshold gain coefficient
where the gain factor is a constant appropriate to specific
devices.
The threshold current density can be written as
 Operational Efficiency of Injection Laser

Differential external quantum efficiency, is the ratio of the increase in

photon output rate for a given increase in the number of injected electrons
(the slope quantum efficiency)

where Pe is the optical power emitted from the device, I is the current, e is
the charge on an electron, hf is the photon energy and Eg is the bandgap
(eV). For a continuous wave (CW) operation laser, usually has values in the
range 40 to 60%.

is related to by the expression

(it has values in the range 50 to 100%)
 Total efficiency

total number of output photons

(external quantum efficiency) ηT =
total number of injected electrons
Pe / hf P
= ≅ e
1/e IE g

As the power emitted Pe changes linearly when the injection current I > Ith, then

For high injection current (e.g. I = 5Ith), then , whereas for lower
currents (I =2Ith) the total efficiency is lower and around 15 to 25%.
External power efficiency (or device efficiency),

where P = IV is the d.c. electrical input power.

 Beam Profile from a Semiconductor Laser
The beam emitted from a semiconductor laser typically has an elliptical
spatial profile.
In the direction perpendicular to the junction, the beam is confined by the
narrow junction, ~ 1mm and is spread by diffraction to an angle as large as
several tens of degrees.
In the direction parallel to the junction, the beam is not confined so
stringently and spreads less to around ten degrees.

The emission from a gallium

arsenide laser tends to be an
elliptical beam with a full
angle divergence around 20°
in the direction perpendicular
to the junction and around 5°
in the direction parallel to the
junction. These angles may
vary considerably with
individual lasers.
 Semiconductor Laser Materials
Most semiconductor laser materials are
fabricated from 3-5 compounds –
structures formed of alloys that contain
three or four elements from columns 3
and 5 of the periodic table.
The three most important types of ternary
or quaternary compound semiconductor
laser materials.
Material Wavelength Range (nm)
Al1–xGaxAs 780-880
In1–xGaxAs1–yPy 1150-1650
In1–xGaxAs1–yPy 630-680
The properties of the material vary
continuously as x and y vary (0 →1).
The most common of the semiconductor lasers is the Al1–xGaxAs laser and
are used for optical disks, optical-fiber telecommunications, and laser
printers.
The quaternary alloy In1–xGaxAs1–yPy, are commonly used for optical-fiber
telecommunications.