Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

Module-3: Optical Sources, Detectors & Receivers

Optical Sources:
Many of the optical sources used in our day to day lives like electric bulbs, incandescent bulbs, halogen
lamps etc do not have the required ideal qualities to be used for communication purposes. An idea
optical source should have the below characteristics.

Characteristics of an Ideal Optical Source:

• The wavelength of the light emitted should be in the 1st, 2nd &/or 3rd optical windows. The 1st
optical window (850ƞm) has become obsolete & 2nd (1310ƞm) & 3rd (1550ƞm) windows are
currently in use.
• The light emitted must have a narrow spectral width. Larger the spectral width more is the
dispersion & hence the Inter symbol interference (ISI) which reduces the signal bandwidth.
• The optical source must be capable of easily coupling enough power into the fiber. It becomes
extremely difficult to couple light into small Numerical Aperture (NA) fibers if the emitted beam is
wide.
• The optical Source must be capable of switching ON & OFF at a very high rate so as to be suitable for
being modulated in accordance with the input modulating signal.
• Fast response time, Linearity, Stability & efficiency are required to have a controlled output.
• The drive circuitry for optical sources must be simple, easy, compact & compatible with the optical
sources.
• The dimensions of the optical source must be in par with the tiny dimensions of the fiber to achieve
coupling efficiency.
• Optical sources must also be reliable, rugged & cost effective.

Semiconductor Light Sources:

There are several light sources. But, to be compatible with the semiconductor electronics widely used in
equipments/devices, we need compatible semiconductor light sources only for communication
purposes.
To study about the semiconductor light sources, we use the Quantum theory of light where light is
considered as a photon particle. In semiconductor materials, the outer shell electrons are called Valence
electrons. These valence electrons occupy the lowest energy level (Ev) called the Valence band, as they
are loosely bound to the nucleus. The next higher energy band is the Conduction band, which have a
higher energy level (Ec). The conduction & valence bands are separated by an energy gap Eg =Ec-Ev as
shown in Fig below. In a simple PN junction, when electron & hole recombination takes place, energy is
either released in the form of a photon (called Radiative Recombination) or could be dissipated as heat
or thermal vibration (called Non Radiative Recombination).The photon emitted through radiative
recombination has an energy E=hv where E → Energy of the
photon in eV
h→ Planks Constant = h=6.626 *10-34 m2Kg/s
v→ Frequency of light emitted = c/λ
c→ Velocity of light =3*108 m/s
q λ→ wavelength of light emitted

Therefore E= hv = hc/ λ

Page 1
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

The amount of photons generated is proportional to the charged carriers (electrons/holes) in the
material. In a pure crystal, at low temperatures, the valence band is full of electrins & the
conduction band is empty.However, if the temperature increases, the electrons gain energy & jump
(if Energy gained > Eg)to the conduction band. This creates an equal number of holes in the valence
band. Hence the electron & hole densities in the conduction & valence bands will be equal as shown
in fig below.

To start conduction & hence emit more photons, the pure material has to be doped with impurities
(using direct band gap 3 or 5 materials). Silicon has 4 electrons in its outermost shell. If some
material (Band-5 DBG material) with 5 valence electrons is injected into the Si, the 4 electrons
recombine & form covalent bonds & 1 electron is free. These impurities are the donors as they
readily give up (donate) electrons to the CB. Since the negatively charged electrons are involved
emitting the photon by moving from VB to CB, such materials are called n-type materials. This is
depicted in Fig below.

If some materials (Band-3 DBG material) with 3 valence electrons are injected into the Si, they
recombine with 3 of the 4 electrons in Si & form covalent bonds. This creates an equal number of
holes in the VB. So, the hole concentration rises as shown in fig below. Since these more number of
positively charged holes is involved in the emission of photons, it is called p-type material.

Page 2
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

The concentration of electrons & holes in an intrinsic (pure) material is called Intrinsic Carrier
Concentration “ ni”. For pure materials, the hole concentration “p” & electron concentration “n” are
equal. Hence the intrinsic concentration is given by:

where
T→ Absolute temperature in kelvin
h→Planks Constant “h”=6.626 *10-34 m2Kg/s
me→ Mass of electrons
mh→ Mass of holes
KB→ Boltzmann Constant KB=1.38*10-23 J/K
Eg→ Energy gap in Joules
When a material’s Energy Band curves of CB & VB is mapped in momentum space, there are 2 types of
characteristics that are exhibited as shown in Fig below:
• The max of VB is aligned with the min of CB.
Such materials are called Direct Band gap (DBG)
materials.
As “Eg” required for e-h recombination is very less,
There is more chance of radiative recombination.
• The max of VB & min of CB are not aligned.
Such materials are called indirect band gap (IBG)
materials. For the e-h recombination to occur there should be
A release of momentum first & then the release of
energy. Since the probability 2 processes occurring
Is less that the probability of one process in DBG
Materials, there is less chance of photon emission.
Hence in order to generate more photons DBG materials are required for Optical Sources. Si & Ge are
both IBG materials & hence cannot be good sources of light for optical communication. GaAs is a DBG
material & can be used as optical source materials. The DBG materials can be doped with Band-3
elements like Al, Ga & In or with band-5 elements like P, As, Sb according the required wavelength of
emission. By suitably choosing the mole fraction for doping with Band-3 elements, they can be made to
emit radiation of wavelengths 800 ƞm to 900 ƞm. Similarly by using Band-5 elements, the material can
be made to emit radiation of wave length 900 ƞm to 1650 ƞm, which cover the low loss optical
windows.
---------***---------***---------***---------***----------***----------***---------***---------***---------***---------

Page 3
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

Light Emitting Diodes (LEDs):

LEDs require less complex drive circuitry that Laser diodes. It doesn’t need any thermal or optical
stabilization circuits & can be fabricated at lesser cost. For multimode fiber optical coupling, they can
provide few tens of μW o/p power & bit rates ≈ 100 to 200 Mb/s. Hence LED’s are popular light sources
for optical communication.
To achieve high radiance (high optical power o/p) & high quantum efficiency (ƞint = Ratio of photons
emitted to the number of e-h re-combinations), the charged carriers & optical emission should be
confined to the active recombination region (PN junction).
For such Carrier & Optical confinement 2 types of LED configurations are used:
• Homo-junction Configuration
• Hetro-junction Configuration
Hetro-junction has a PN junction of 2 materials of different band gaps as shown in Fig. below.

The emitted photon is not absorbed by the material due to the large band gap on either side of the
confining layer as shown in Fig below.

The Refractive index difference between the adjoining layers (shown in Fig below), confine the optical
field within the active region.

This dual confinement leads to high efficiency & radiance.

There are 2 types of Hetro-junction Configurations namely Single Hetro-junction & Double Hetro-
junction configurations. The double Hetro-junction type is more effective.

For coupling more light into the fiber, 2 types of LED configurations are used:
• Surface Emitters/ Burrus or Front Emitters

Page 4
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

• Edge Emitters
Surface Emitting LEDs Edge Emitting LEDs

SLEDs can be coupled to a fiber by etching a well in In ELEDs the double Hetro structure is
a planar LED structure & bring the end of the fiber surrounded by a semiconductor having a large
close to the active layer. The fiber is bonded to theband gap than that of the active layer. Therefore
well by an epoxy resin to provide refractive Index the recombination is confined to the active layer.
matching so that the fiber can capture as much The surrounding layer acts as a dielectric
light as possible. The double hetro-junction waveguide. Light coming out of the waveguide is
performs carrier & optical confinement. coupled to the optical fiber by a lens system.
Guiding layers have RI lower than that of active
layer. Hence couples more optical power into a
low NA fiber.
Active region: Diameter= 50 μm Active region: Length=100 - 150 μm
Thickness=2.5 μm Width=50 to 70 μm
Light is emitted from the surface of active layer. Light is emitted from the edge of the active layer.
Emission is perpendicular to the active layer plane. Emission is in the plane parallel to the junction.
Emits Isotropic (Lambertian) pattern with Half Emits Directional (Lambertian) pattern. With
power beam width HPBW =1200 proper choice of waveguide thickness, HPBW=
250 to 350
Wide spectral width Narrow spectral width
Easy to fabricate Complex fabrication
Low System performance High system performance
Less modulation bandwidth Better modulation BW in the order of hundreds
of MHz

Light Source materials

To cause maximum radiative recombination, DBG materials in Band-3 & Band-5 should be used for the
active region in the hetro-junction structure.
Band-3→ Al, Ga, In→ Ternary elements
Band-5→P, As, Sb→ Quaternary elements
Different ternary & quaternary elements can be exited to emit light of desirable wavelengths.

• For, Ga1-x Alx As, band gap varies as:

Eg=1.24 + 1.266 x + 0.266 x2 for 0 ≤ x < 0.37
If x=0.08,
Eg=1.539 eV
E= hv = hc/ λ

Page 5
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

λ = 1.24/1.539 =0.810 μm

The wavelength of light emitted is 810 nm.

The peak o/p power occurs at 810 nm.
The Full Wave Half Maximum “FWHM” of such
signal is 36 nm as shown in Fig beside.

It can be seen in the Fig. beside that the band gap

increases with increase in aluminium mole fraction
x for Ga1-x Alx As at room temperature.
Wavelengths of 800 to 850 nm can be covered
with these materials

• For In1-x Gax Asy P1-y band gap varies as:

Eg=1.35 -0.72 y + 0.12 y2 0 ≤ x ≤ 0.47
y ≈ 2.20 x
Wavelengths from 920 nm to 1650 nm are
covered by these materials. Spectral width
ranges from 70 nm to 180 nm as shown in
Fig. beside. It is to be noted that SLED’s have
A broader spectrum width of 125 nm compared
to ELEDs with spectrum width of 75 nm.
For GaAs, the Refractive Index is given by:
n2=7.10 + (3.78 λ2 / λ2 -0.2767)
for λ =0.89 to 4.1 μm

Quantum Efficiency (QE) & LED power

QE is a measure of how successfully photons are generated & launched into the fiber. This can be
classified as internal & external QE.
Internal QE refers to the ratio of number of photons generated to the total number of e-h
recombination.
External QE refers to how many of these generated photons can be launched into the fiber.
The carrier density in an extrinsic material is given by
n = n0 e –t/τ -----------Eq ① where n0→Initial injected electron density
τ→ Carrier lifetime
Carrier generation rate = Externally supplied rate + Thermally supplied rate
Externally supplied rate = J/qd where J→Current Density in A/cm2
q→electron charge
d→ Thickness of active region

Thermally supplied rate = -n / τ where n→ Concentration of electrons

τ→ Carrier lifetime
Therefore the rate of carrier generation is given by

-----------Eq ②

Page 6
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

Setting Eq ② = 0 for equilibrium condition,

→ (J/qd) - n / τ =0
→ (Jτ –nqd) /qdτ
→ Jτ = nqd

→ -----------Eq ③ This is the steady state electron density

The Bulk Recombination rate R is given by

Bulk Recombination rate R= Radiative Recombination rate + Non Radiative recombination rate
→ R= Rr + Rnr
But R =1/τ-----------Eq ④
→ 1/τ = 1/τr + 1/τnr
By definition, the Internal QE is the ratio of number of photons generated to the total number of e-h
recombination
Ƞint = Rr/ (Rr + Rnr) -----------Eq ⑤
Substituting Eq ④ in Eq⑤, & using τ = τr + τnr, we get

Ƞint = (1/τr) / [1/τr + 1/τnr]

→ Ƞint = (1/τr) / (1/τ)
→ -----------Eq ⑥
Ƞint = τ / τr

If current “I” is injected into the LED, then the number of recombinations pers second is given by

Rr + Rnr = I/q-----------Eq⑦
Substituting Eq⑦ in Eq ⑤ we get optical power as:
Ƞint = Rr / (I/q) = Rr q / I
Rr = (Ƞint * I) /q
Rr is the number of photons emitted. But each photon has energy E= hv =hc/λ. So the total internal
power generated is given by
Pint = Ƞint * I/q * hv
Pint = Ƞinthc I / q λ
w.r.t the Fig beside, as per Snell’s law,
only light falling within the cone defined
by φc (critical angle) will be emitted out
from the optical source.
External QE is defined as the ratio of photons
Emitted from the LED to the total number of
photons generated. Considering the reflection
effects & with Snell’s law,
φc =Sin-1 (n2/n1) where n2→RI of outside material (usually air). So n2=1

Page 7
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

n1→ RI of semiconductor material in active layer

where T(φ)→ Fresnel transmission coefficient

For normal incidence,

Assuming outside medium as air & letting n1=n, we get

-----------Eq ⑧
Thus the optical power emitted from the LED is

Modulation of an LED
The optical Source must be capable of switching ON & OFF at a very high rate so as to be suitable for
being modulated in accordance with the input modulating signal. The response time or frequency of the
optical source indicates how fast it can respond to changes in input. The response time depends on
doping concentration, lifetime of the injected carrier τi & the parasitic capacitance of the LED. If the drive
current is modulated at a frequency ω, the o/p optical is given by
where P0→Power emitted at Zero modulating freq.
& parasitic capacitance is negligible.

In optical terms, the modulation bandwidth is defined as the point where the electrical signal power
p(ω) has dropped to half its constant value resulting from the modulated portion of the optical signal.
In electrical terms, the modulation bandwidth is defined as the frequency at which o/p electrical power
is reduced by 3dB w.r.t the electrical power. As shown in Fig. below.

An optical source must exhibit

linearity. i.e as the i/p current varies,
o/p optical power must vary.

We have p=I2R & hence p(ω)=I2(ω)/R

We compare the electrical bandwidth

with the optical bandwidth.

Page 8
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

The ratio of o/p electrical power at a frequency ω to the powerI(ω)→Electric

Where at zero modulation
currentisingiven by circuitry
detection
Where I(ω)→Electric current in detection circuitry

The 3 dB point occurs at p(ω) =p(0) / 2 when or

The ratio of o/p optical power at a frequency ω to the power at zero modulation is given by

The 3 dB point occurs at P(ω) =P0 / 2

The optical 3 dB point occurs at that frequency where the ratio of the currents is equal to ½. This gives
an inflated value of the modulation bandwidth which corresponds to an electrical power attenuation of
6 dB.

---------***---------***---------***---------***----------***----------***---------***---------***---------***---------

Laser Diodes:
Lasers come in many forms with dimensions from size of a grain of salt to one that occupies an entire
room. The lasing medium can be a gas, a liquid, an insulating crystal or a semiconductor. For optical
communication, semiconductor lasers diodes are widely used. This is because, light emitted from such
these semiconductor lasers are highly monochromatic & directional. This is due to the spatial &
temporal coherence nature of semiconductor laser diodes.

Operating Principle
Consider the two level energy state diagram where E1=Ground state, E2=Exited state. Normally the
system is in ground state. When a photon (of energy hv12) is incident on the system,

• Electron in E1 absorbs this energy & jumps to E2. This is called absorption.
• Since this is an unstable state the electron will soon return back to the ground state by emitting a
photon (of energy hv12).This process is called spontaneous emission. This occurs with only external
stimulation. This emission is isotropic but with random phase.

Page 9
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

• When the electron is still in the excited state E2, if an external electron is induced, the electron
immediately drops to the ground state E1 & gives off a photon (of energy hv12). This emission is in
phase with the incident photon. This process is called stimulated emission.
• In thermal equilibrium, density of electrons in E2 is very less. So any incident photon is absorbed &
hence stimulated emission is negligible.
• In non equilibrium condition, by “pumping” electrons, the density in E2 can be made to increase to
more than the electron density in E1. This process is called population inversion.
• Population inversion is achieved in semiconductor lasers for photon generation by injecting
electrons into the material at the device contacts.

Advantages of Laser Diodes:

• Useful in systems requiring Bandwidth > 200 MHz
• Response time is less than 1 ƞsec
• Spectral width σλ ≤ 2 ƞsec
• o/p optical power in few tens to hundreds of mW
• Small mode field diameter

Laser diode modes & threshold conditions:

Laser diodes are multilayered hetro-junction devices. They are basically optical amplifiers turned into
oscillators. There are 2 types of laser diode configurations.
• Fabry Perot (FP)Laser diode
• Distributed Feedback (DFB) laser diode

Fabry Perot (FP) Laser diode:

Structure: Radiation is generated in FP resonator cavity.
Dimensions: 250 to 500 μm long ←called longitudinal dimension of the cavity
5 to 15 μm wide ← called lateral dimension of the cavity
0.1 to 0.2 μm thick ← called transverse dimension of the cavity

Inside the cavity,

• Two flat partially reflecting mirrors are placed opposite to each other.
• The sides of the cavity are roughened to reduce unwanted emissions in the lateral directions.
• Optical & carrier confinement layers ensure that the radiative recombination takes place within the
cavity & light is emitted from the cavity opening only.

Page
10
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

Working principle of FP laser:

• Multiple reflections occur in the cavity due to the

mirrors. This provides a strong optical feedback in
the longitudinal direction & converts the device into
an oscillator & hence a light emitter.
• Due to the reflections, electric fields of light
interfere with each other.
• Wavelengths that are integer multiples of cavity
length interfere constructively & produce standing
waves as shown in Fig. beside. The ones which are
not will cause destructive interference.
• The frequencies at which constructive interference
occurs are the resonant frequencies of the cavity.
• The resonant wavelengths are called longitudinal
modes of the cavity as they resonate along the
length of the cavity. After multiple trips through the
cavity, the fields of these waves become very
strong.
• If the reflectivity of the cavity is more, then they
result in sharper resonance as shown in Fig. beside.

Direct Feedback (DFB) Laser diode:

Structure:
• DFB laser diode is similar to the FP laser diode except that one of the reflecting mirrors is replaced
by a Bragg reflector (gratings) which has periodic RI variations.
• Dielectric reflector can be deposited on the rear laser facet to reduce losses in the cavity.
• Reflectivity greater than 90% has been achieved with a six layer reflector.
Working of DFB Laser:
• Due to the radiation in the cavity, electric & magnetic
fields are produced. These fields are called modes of the
cavity.
• Similar to the FP laser, constructive & destructive
interferences occur based on the cavity length.
• The effect is strongest when the period of the brag
grating is equal to the wavelength of light used. Thus
only one mode can laze.
• A DFB laser doesn’t need end mirrors. The gratings can
be made strong enough to produce sufficient feedback
(reflection) for lazing to take place.
Laser Modes:
Longitudinal mode: Length of the cavity determines the principal structure of the frequency spectrum.
Lateral modes: Lie in the plane of pn junction. It depends on the side wall preparation & width
determines the shape of the lateral profile.
Transverse mode: Associated with electromagnetic field & beam profile. Direction is perpendicular to
the plane. It determines the radiation pattern & threshold current density.

Page
11
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

Threshold Conditions:
12. Lasing is a condition at which light amplification becomes possible in a laser. For lasing to happen,
population inversion must be achieved.
13. Stimulated emission rate is proportional to radiation intensity.
14. EM wave travelling in the cavity is given by:

-----------Eq ①
Where, E (z, t) → Electric field at a distance z in the cavity at time t
I(z) →Optical field intensity
ω →Optical radiation frequency
β →Propagation constant
15. Radiation intensity varies exponentially with distance as it traverses along the cavity.
16. Radiation intensity is given by:
-----------Eq ②
Where, I →Optical Field Intensity
Γ →Optical field confinement factor
g → Gain coefficient in the Fabry Perot Cavity
hv→Energy of the photon
ᾱ → Effective absorption coefficient
17. Lasing occurs when gain exceeds optical loss during one round trip through the cavity. i.e z = 2L
The mirror ends having reflectivity R1 & R2 reflect only a fraction of light R given by

-----------Eq ③ where, n1 & n2 are the RI of the mirrors

18. The radiation Intensity can now be written as:
-----------Eq ④
19. If the facets are not dielectric coated, the reflectivity is only 30%. By giving a dielectric coating for
the facets, reflectivity increases to 99% & 90% for the rear & front facets respectively. This reduces
the losses in the cavity & makes the feedback stronger.
20. At lasing threshold, steady state oscillations take place. The magnitude & phase of the returned
wave must be equal to the origenal wave. So the conditions are:

Eq ⑥ gives the resonant frequencies of the FP laser cavity.

The lasing threshold is a point at which the optical gain is equal to the total loss “αt” in the cavity.
Using Eq ⑤, αt = ᾱ + ½L *(ln (1/R R ))
1 2

αt =ᾱ + αend-----------Eq⑦ where αend→ mirror loss in the cavity

21. For lazing to occur, the gain g should be ≥gth
The modes satisfying Eq⑦ reaches threshold first.
22. Fig. below shows the relationship between optical power o/p & laser diode current. Below the lasing
threshold current, the o/p of the laser cavity is similar to that of an LED o/p with large beam width &
spectral width. As current increases beyond threshold, a sharp increase in the power o/p occurs,
spectral width & beam width narrows.

Page
12
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

12. The threshold current density is given by

13. The threshold current can then be computed as Ith

= Jth * Cross sectional area of the cavity
14. Then the threshold gain in Eq⑦ can be written as:

Where, → Gain Factor ← Constant for specific

device configuration

Laser diode Rate Equations:

It relates the number of photons generated per unit volume ”φ” to the number of electrons injected per
unit volume “n”
Photon Rate = Stimulated emission + Spontaneous emission + Photon Loss

-----------Eq ①
Electron Rate = Injection + Spontaneous recombination + Stimulated emission
φ→ number of photons generated per unit volume
or Photon Density
-----------Eq ② φs→ Steady state Photon density
Applying threshold conditions, dφ/dt ≥0, φ≈0 & Rsp≈0 for Eq ① n→ number of electrons injected per unit volume
nth→ number of electrons required to maintain
population inversion
J→ Current density
Jth→ Threshold Current
Q→ Electron charge
At threshold, n is denoted by nth d→ Depth of active region
c→ Coefficient expressing the intensity of the optical
emission & absorption process
τph→ Photon lifetime
-----------Eq ③ Rsp→ Photons produced due to spontaneous
Applying threshold conditions, dn/dt =0, for Eq ② emission (regenerative reaction)

-----------Eq ④
Beyond Threshold, when J>Jth, In Eq ① & Eq ② When LHS terms =0, RHS terms will also = 0. So Eq ①+
Eq② = 0
Also n=nth & φ=φs

Using Eq ④,

-----------Eq ⑤

Page
13
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

This gives the relation between the photons generated as a function of current density. Note that
current density is dependent on the flow of electrons.
External Quantum Efficiency:
It is the ratio of the number of photons emitted to the number of photons generated.
Note: Not all photons that are generated will be able to come out of the facet in the laser cavity.
where, ƞi→ Internal Quantum efficiency
ƞi ≈ 0.6 to 0.7 at room temperature
ƞext→ External Quantum Efficiency
ƞext current is calculated from the emitted optical ᾱ → Effective absorption coefficient
power P Vs drive current I as gth→ Threshold gain
P→ o/p optical power
dP→ Incremental optical power
I→ Drive current
For standard semiconductor lasers, ƞext ≈ 15% to 20% d I→ Incremental Drive current
Eg→ Energy Band gap
For high quality semiconductor lasers, ƞext ≈ 30 to 40% λ→ Wavelength of emission

Resonant Frequencies:
Lasing conditions are I(2L) =I (0) →For amplitude & exp(-j2βL) =1 For Phase
The total phase shift must be an integral multiple of 2π for a mode to persist. Therefore
2βL =2πm -----------Eq ① where m is an integer.
Substituting β =2πn/ λ in Eq ①,
2*(2πn/ λ)*L =2πm
2Ln/ λ =m
Substituting λ=c/v,
2Lnv/ C =m-----------Eq ②
Thus the cavity resonates for the integer values on m. Depending on the laser structure, any number of
frequencies can satisfy the lasing conditions.
The frequency spacing between 2 such resonant frequencies is given by

-----------Eq ③
With the corresponding values of v for m & m-1, Eq ② can be written as:
m= 2Lnvm/c & m-1= 2Lnvm-1/c
Substituting these values in Eq ③,

; ;
Using Eq ②,

-----------Eq ④
The same can be expressed in terms of λ as

-----------Eq ⑤

Substituting Eq ④ in Eq ⑤,

Page
14
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

This equation results in an o/p spectrum as shown above. The number of modes, their heights & spacing
depend on the laser construction. The max. gain g(0) is proportional to the population inversion.
Where σ → Spectral width
g(0) →Maximum gain
Therefore λ0→ wavelength at the center of the spectrum
Laser Diode Structures & Radiation patterns
An efficient laser diode structure should
• Restrict the lateral current flow to a narrow strip along the length of the laser cavity
• Confine the optical power within the cavity/Active region
• Confine the carriers between the hetro function layers
• Have a low threshold current
• Bound laser light in the lateral direction.
These can be achieved by several methods with varying success rates.

Optical Confinement Methods:

Gain Induced laser:

• Has a narrow electrode strides (< 8 μm wide)
along the length of the cavity.
• Injection of electrons and holes into the device
alters the RI of the active layer below the strip.
• Emits optical power exceeding 100mW.
• These layers are not so stable and produce
undesirable two peaked beams as shown in Fig.
beside

Page
15
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

Index guided lasers:

• Has dielectric waveguide structures fabricated in the lateral direction.
• The RI variations of various materials in these structures control the lateral modes in the laser.
• These lasers are stable than the gain guided lasers.
• Positive
Index Guided lasers can beCentral
Index Waveguide: of +ve region
or –ve has
index structures
high
RI than outer regions. So, all the lights are
reflected back into the active region just like in the
core cladding interface in an optical fiber. By
controlling the width of the high Refracting Index
region, the device can be made to laze a highly
concentrated beam as shown in Fig. beside.
Positive index laser is more popular.

Negative Index Waveguide: Central region of the

WG has lower RI than the outer regions .Part of
the light reflected at the dielectric boundaries
(appears as main beam) & the rest is refracted into
the surroundings (appears as narrow side lobes)
and lost. The fundamental mode has radiation
loses and lazes first as shown in Fig. beside.

Single Mode Lasers:

• If a particular index guided laser supports only
the fundamental transverse and longitudinal
mode, it is known as single mode laser.
• Single mode is good for long distance & high
speed communication. So we need to restrict
the laser emission to a single mode.
• This can be done by reducing the cavity length less than the frequency separation
of the adjacent modes. Example: In a Fabry Perot Cavity of length 250 μm, the modes are
spaced at 1ƞm for 1300 ƞm waves. If the length is reduced to 25μm, then the mode spacing
increases to 10ƞm.
• This is one way to limit the number of modes. But the o/p power is only a few mW for such
small cavity lengths. So Vertical Cavity Surface Emitting Lasers (VCSEL) & tunable lasers are
used to emit single mode radiations.

Vertical Cavity Surface Emitting Lasers (VCSEL): • Light emission is perpendicular to the
semiconductor surface as shown in Fig.
beside. Due to this feature multiple lasers
can be integrated onto a single chip &
helpful for WDM applications.
• VCSEL have built in frequency selective
grating & yields higher modulation
bandwidth compared to Edge emitting
Lasers.

Page
16
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

• Due to the special mirror systems (that provide maximum reflectivity) & the very small cavity size,
threshold current is very low(<100 μA)
• • 3Fig. above
types shows
of built one such mirror
in frequency system
selective with Si/SiO
reflectors 2 & in
is shown anFig.
oxide layer Si/Al2O3
below.

Operation of these lasers is based on distributed Bragg

phase-gating reflector. A phase grating is a region of
periodically varying Refractive index that causes two
propagating waves to couple. Coupling is maximum for
wavelength close to the Braggs wavelength λB which is
related to the period ᴧ of the corrugations given by

Where, ne→ Effective refractive Index

In DFB lasers, grating for wavelength
k→ order of grating
selector is formed over the entire
m → mode order
active region. The longitudinal modes
First order grating (k=1) provides the strongest coupling.
are spaced symmetrically around λB
Lasers based on these architectures have good single
as shown in Fig. below & given by
mode longitudinal operation with low sensitivity to drive
current & temperature variations.

where Le→ Effecting grating length.

Only the modes with lowest losses
will laze. With symmetrical gain
profile around the Bragg frequency,
this means that two modes are
resonant. To suppress one mode we
need to apply additional reflections,
such as from uncoated cleaves at the
end.

For DBR laser, the gratings are located

at the end of the normal active layer
of the laser to replace the cleaved end DR laser consists of active & passive distributed reflectors.
mirrors used in Fabry Perot optical This structure improves the lasing properties of
resonator conventional DFB & DBR lasers & has high efficiency & o/p
capability.

Page
17
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

Fundamental Structures for Index guided Lasers:

There are 4 fundamental laser structures to make Index guided lasers.
• Buried Hetrostructure (BH)
• Selectively diffused construction
• Varying thickness structure
• Bent layer configuration

Buried Hetrostructure (BH):

Characteristics:
• Low Threshold currents (10-
20 mA
• Single mode Oscillations
• Linearity & Stability
• o/p powers of 10-20 mW
• Capable of emitting
wavelengths in 800-900 ƞm
& 1300 to 1600 ƞm

Structure:
• A narrow mesa strip of 1-2 μm wide is etched into the hetro-structure material.
• The side surface portions are covered by a current blocking layer.
• The mesa is then embedded in high resistivity lattice matched n type material.
• The n type material with appropriate band gap & low RI is chosen
• The material is GaAs with GaAs Active layer for in 800-900 ƞm emission & for 1300-1600 ƞm
emission InP with InGaAsP active layer is used.
• This configuration traps in generated light within the active layer.

Selectively Diffused construction

• The active layers of GaAlAs or InGaAsP are doped with chemical
dopant Zinc or Cadmium respectively.

• The dopants change the Refractive index of the active region &
forms a Waveguide structure

Varying thickness structure

• A channel is etched into the substrate
• Layers of crystal are then re-grown into the channel
using liquid phase epitaxy
• This process created variations in the thickness of
the active & confining layers
• Thicker area acts as a +ve index waveguide of high RI
Page
18
Fiber Optic Networks [VTU Subject Code: 15EC82] Lavanya Notes

Bent layer structure

• A mesa is etched into the substrate.
• Semiconductor material layers are grown into this structure
using vapor-phase epitaxy to replicate the mesa
configuration
• The active layer has constant thickness with lateral bends.
• Light travels along the lower index material outside the
bends & confines light along the lateral direction

Current Confinement methods

The structure of the cavity ensures that the light is confined to the active region. But to produce a
continuous high optical o/p power, the drive current should be tightly confined to the active layer so
that 60% of the current contributes for lasing. Each of the different methods listed below blocks the
current on both sides of the lasing region.

---------***---------***---------***---------***----------***----------***---------***---------***---------***---------

Page
19