Chapter 4 Stimulated Emission Devices

Optical Amplifiers and LASERS

CHAPTER 4 -
STIMULATED
EMISSION DEVICES
OPTICAL AMPLIFIERS
AND LASERS
STIMULATED EMISSION DEVICES
OPTICAL AMPLIFIERS AND LASERS

Zhores Alferov (on the right) with Valery Kuzmin (technician) in 1971 at the Ioffe Physical
Technical Institute, discussing their experiments on heterostructures. Zhores Alferov carried
out some of the early pioneering work on heterostructure semiconductor devices that lead to
the development of a number of important optoelectronic devices, including the heterostructure
laser. Zhores Alferov and Herbert Kroemer shared the Nobel Prize in Physics (2000) with Jack
Kilby. Their Nobel citation is "for developing semiconductor heterostructures used in high-
speed- and opto-electronics" (Courtesy of Zhores Alferov, Ioffe Physical Technical Institute)
 THE LASER PATENT WARS
Arthur L. Schawlow is adjusting a ruby optical maser during
an experiment at Bell Labs, while C.G.B. Garrett prepares to
photograph the maser flash. In 1981, Arthur Schawlow
shared the Nobel Prize in Physics for his "contribution to the
development of laser spectroscopy"
(Courtesy of Bell Labs, Alcatel-Lucent)
Gordon Gould (19202005) obtained his BSc in Physics (1941) from Union
College in Schenectady, and MSc from Yale University. Gould came up with
the idea of an optically pumped laser during his PhD work at Columbia
University around 1957. He is now recognized for the invention of optical
pumping as a means of exciting masers and lasers. He has been also credited
for collisional pumping as in gas lasers, and a variety of application-related
laser patents. After nearly three decades of legal disputes, in 1987, he
eventually won rights to the invention of the laser. Gould's laboratory
logbook even had an entry with at he heading "Some rough calculations on
the feasibility of a LASER: Light Amplification by Stimulated Emission of
Radiation,", which is the first time that this acronym appears. Union College
awarded Gould an honorary Doctor of Sciences in 1978 and the Eliphalet
Nott Medal in 1995.
4.1 STIMULATED
EMISSION,
PHOTON
AMPLIFICATION,
AND LASERS
4
LIGHT INTERACTION WITH MATTER
 The rules of quantum mechanics, limit atoms and molecules to having
discrete amounts of stored energy that depend on the nature of the atom
or molecule.
 The lowest energy level for an individual atom occurs when its electrons
are all in the nearest possible orbits to its nucleus.
 This condition is called the ground state.
 When one or more of an atom’s electrons have absorbed energy, they can
move to outer orbits, and the atom is then referred to as being “excited.”
STIMULATED EMISSION DEVICES
OPTICAL AMPLIFIERS AND LASERS

 Einstein recognized that this emission could be produced in two

ways.
 Usually, discrete packets of light known as photons are emitted
spontaneously, without outside intervention.
 Alternatively, a passing photon could stimulate an atom or molecule to
emit light—if the passing photon’s energy exactly matched the energy
that an electron would release spontaneously when dropping to a lower-
energy configuration.
 BASIC CONCEPT (LASER)
Spontaneous Emission

Photons emitted in the spontaneous

emission process do not flow exactly in the
same direction of incident photons.
SPONTANEOUS EMISSION

 An electron in an excited state may decay to a lower energy

state which is not occupied, according to a particular time
constant characterizing that transition.
 When such an electron decays without external influence,
emitting a photon, that is called "spontaneous emission".
 The phase and direction associated with the photon that is
emitted is random.
 A material with many atoms in such an excited state may thus
result in radiation which has a narrow spectrum (centered
around one wavelength of light), but the individual photons
would have no common phase relationship and would also
emanate in random directions.
 This is the mechanism of fluorescence and thermal emission.
 BASIC CONCEPT (LASER)
Stimulated Emission
• an incoming photon of energy stimulates the whole emission
process by inducing the electron at E2 to transit down to E1.
• The emitted photon has the same energy, since ,
• The liberated energy transfers to the electromagnetic field,
creating a new photon with a frequency, polarization, and
direction of travel that are all identical to the photons of the
incident wave.
STIMULATED EMISSION
 An external electromagnetic field at a frequency associated with
a transition can affect the quantum mechanical state of the atom
without being absorbed.
 A transition from the higher to a lower energy state produces an
additional photon with the same phase and direction as the
incident photon; this is the process of stimulated emission.
LIGHT AND MATTER
INTERACTION SUMMARY

11
PHOTON AMPLIFICATION
 Stimulated emission is the basis for obtaining photon
amplification since one incoming photon results in two outgoing
photons which are in phase.
 Which process dominates depends on the ratio of lower-energy
to higher-energy configurations.
 Ordinarily, lower-energy configurations predominate.
 This means that a spontaneously emitted photon is more likely to be
absorbed and raise an electron from a lower-energy configuration to a
higher-energy configuration than to stimulate a higher-energy
configuration to drop to a lower-energy configuration by emitting a second
photon.
 As long as lower-energy states are more common, stimulated emission will
die out.
POPULATION INVERSION
 In physics, specifically statistical mechanics, a population
inversion occurs while a system (such as a group of atoms or
molecules) exists in a state in which more members of the
system are in higher, excited states than in lower, unexcited
energy states.
 It is called an "inversion" because in many familiar and
commonly encountered physical systems, this is not possible.
 This concept is of fundamental importance in laser science
because the production of a population inversion is a necessary
step in the workings of a standard laser.
POPULATION INVERSION
PUMPING
TWO LEVEL LASER SYSTEM
 To understand the concept of a population inversion, and how
much pumping is required to generate LASER consider a very
simple assembly of atoms forming a laser medium.
 Assume there is a group of N atoms, each of which is capable of
being in one of two energy states: either
 The ground state, with energy E1 with N1 atoms; or
 The excited state, with energy E2, with N2 atoms such that E2 > E1 and
N=N1+N2

16
TWO LEVEL LASER SYSTEM
 The energy difference between the two states, given by

 determines the characteristic frequency of light which will

interact with the atoms;
 This is given by the relation
 being Planck's constant.

 If the group of atoms is in thermal equilibrium, it can be shown

from Maxwell–Boltzmann statistics that the ratio of the number
of atoms in each state is given by the ratio of two Boltzmann
distributions, the Boltzmann factor:

 where is the thermodynamic temperature of the group of atoms,

and is Boltzmann's constant.
17
TWO LEVEL LASER SYSTEM -
ANALYSIS
 We may calculate the ratio of the populations of the two states at room
temperature (T ≈ 300 K) for an energy difference ΔE that corresponds to light
of a frequency corresponding to visible light (ν ≈ 5×1014 Hz).
 In this case ΔE = E2 - E1 ≈ 2.07 eV, and kT ≈ 0.026 eV.
 Since E2 - E1 ≫ kT, it follows that the argument of the exponential in the
equation above is a large negative number, and as such N2/N1 is vanishingly
small; i.e., there are almost no atoms in the excited state.
 When in thermal equilibrium, then, it is seen that the lower energy state is
more populated than the higher energy state, and this is the normal state of
the system.
 As T increases, the number of electrons in the high-energy state (N2)
increases, but N2 never exceeds N1 for a system at thermal equilibrium;
rather, at infinite temperature, the populations N2 and N1 become equal.
 In other words, a population inversion (N2/N1 > 1) can never exist for a
system at thermal equilibrium.
 To achieve population inversion therefore requires pushing the system into a
non-equilibrated state. 18
TWO LEVEL LASER SYSTEM -
ANALYSIS
 It is clear, that in the two-level atom the pump is, in a way, the
laser itself!
 Such a two-level laser would work only in jolts. That is to say,
once the population inversion is achieved the laser would lase.
 But immediately it would end up with more atoms in the lower
level.
 Such two-level lasers involve a more complicated process.
 We will see, in later material, examples of these in the context of
excimer lasers, which are pulsed lasers.
 For a continuous laser action we need to consider other
possibilities, such as a three-level atom.

19
THREE LEVEL
SYSTEM

20
THREE LEVEL SYSTEM
 To achieve lasting non-equilibrium conditions, an indirect method
of populating the excited state must be used.
 Consider a group of N atoms, this time with each atom able to
exist in any of three energy states, levels 1, 2 and 3, with energies
E1, E2, and E3, and populations N1, N2, and N3, respectively.
 We assume that E1 < E2 < E3; that is, the energy of level 2 lies
between that of the ground state and level 3.

21
THREE LEVEL SYSTEM
 Initially, the system of atoms is at thermal equilibrium, and the
majority of the atoms will be in the ground state, i.e., N1 ≈ N,
N2 ≈ N3 ≈ 0.
 If we now subject the atoms to light of a frequency

 the process of optical absorption will excite electrons from the

ground state to level 3.
 The level 3 is sometimes referred to as the pump level or pump
band, and the energy transition E1 → E3 as the pump transition,
which is shown as the arrow marked P in the diagram

22
THREE LEVEL SYSTEM
 Upon pumping the medium, an appreciable number of atoms will transition
to level 3, such that N3 > 0.
 To have a medium suitable for laser operation, it is necessary that these
excited atoms quickly decay to level 2.
 The energy released in this transition may be emitted as a photon (spontaneous
emission), however in practice the 3→2 transition (labeled R in the diagram) is usually
radiationless, with the energy being transferred to vibrational motion (heat) of the host
material surrounding the atoms, without the generation of a photon.
 An electron in level 2 may decay by spontaneous emission to the ground
state, releasing a photon of frequency such that
 , which is shown as the Laser transition L,
 If the lifetime of this transition, is much longer than the lifetime of the
radiationless 3 → 2 transition (if , known as a favorable lifetime ratio), the
population of the E3 will be essentially zero (N3 ≈ 0)
 If over half the N atoms can be accumulated in this state, this will exceed the
population of the ground state N1.
 Thus population inversion (N2 > N1 ) can be achieved between level 1/ 2, 23
and optical amplification at the frequency can be obtained.
RATE EQUATION
 Consider a process in a two-level atomic system
 The rate equation for any given level is:

 where is the probability of a transition and is the population of

the level.
 Convension is the probability of transition from level to
 We may also relate this
probability to the Einstein B
coefficient, which can be
expressed as:

 Where ρ is the energy density.

24
THREE LEVEL SYSTEM – RATE
EQUATION

25
THREE LEVEL SYSTEM – RATE
EQUATION
 Analysis of a three-level laser is progressing with an examination of
energy flow into and out of each level in the system.
 Pump Level
 The rate equation for the pump level can be expressed as the change of population of
atoms in the pump level as follows:

 Where is the probability of an atom making the transition from level 1 to level 3 and
is the decay lifetime of the pump level.
 Flow of energy into the pump level
 Described by first half of the equationi n which (N1 - N3) represents the number of
atoms available at the ground state to be pumped to level 3 by W13
 Flow of energy out of pump level
 Described by the second half of the equation (N3/τ3) in units of number of atoms
per second.
 The decay lifetime τ3 is a total lifetime representing both the decay from level 3 to
level 2 and the decay from level 3 to level 1. If each decay path 3→2 and 3 →1 has 26
time constants τ32 and τ31, respectively, 1/τ3 = 1/τ32+ 1/τ31.
THREE LEVEL SYSTEM – RATE
EQUATION
 E2 Level

 where τ32 is the decay lifetime from the pump level to the E2 and τ21
is the lifetime of the E2.
 Energy flows into the E2 from above via decay from the pump level,
the number of atoms per second being the population of the pump
level (N3) divided by the decay lifetime from the pump level.
 We use the specific decay rate 1/τ32 to describe the flow of energy from the
pump level only to the E2.
 The other path that may be possible is from the pump level directly to
ground (τ31), but that is undesirable and in any case, does not affect
the rate equation for the E2 (except that the process may serve to
depopulate the pump level, which would otherwise decay to the E2).
 Energy flows out of the E2 via the transition to ground (which serves
as the lower-lasing level in a three-level system such as this), which is
modeled by the second part of the equation. 27
THREE LEVEL SYSTEM – RATE
EQUATION
 E2 Level

 where τ32 is the decay lifetime from the pump level to the E2 and τ21
is the lifetime of the E2.
 Energy flows into the E2 from above via decay from the pump level,
the number of atoms per second being the population of the pump
level (N3) divided by the decay lifetime from the pump level.
 We use the specific decay rate 1/τ32 to describe the flow of energy from the
pump level only to the E2.
 The other path that may be possible is from the pump level directly to
ground (τ31), but that is undesirable and in any case, does not affect
the rate equation for the E2 (except that the process may serve to
depopulate the pump level, which would otherwise decay to the E2).
 Energy flows out of the E2 via the transition to ground (which serves
as the lower-lasing level in a three-level system such as this), which is
modeled by the second part of the equation. 28
STEADY STATE CONDITION
 Under a steady-state condition these rates must both be equal to zero;
that is, the population of each level remains constant, so the flow of
energy into the level is the same as the flow of energy out of the level.

 Since τ3 is a small quantity i.e., 1/τ3 is much greater than W13 so that
the latter term can essentially be ignored, giving us an expression for N3:

 The population of the pump level depends on

 the rate of pumping itself and
 the total lifetime of that level 29
STEADY STATE CONDITION
 Similarly,

 Assuming that τ32 =τ3 (i.e., assuming that there is no leakage

from the pump level to ground and that all atoms in the pump
level decay solely to E2, we get

 To ensure population inversion i.e., a minimum pumping rate

equal to (where is proportional to pumping rate) is needed just
to get half of the population at ground state to the upper lasing 30
level
THREE LEVEL LASER -
ANALYSIS

31
Energy-level populations for a three-level system.
 3-LEVEL LASERS: THE RUBY LASER

(a) A more realistic energy diagram for the Cr 3+ ion in the ruby crystal (Al2O3), showing the optical pumping
levels and the stimulated emission. (b) The laser action needs an optical cavity to reflect the stimulated
radiation back and forth to build-up the total radiation within the cavity, which encourages further stimulated
emissions. (c) A typical construction for a ruby laser, which uses an elliptical reflector, and has the ruby
crystal at one focus and the pump light at the other focus.
4 LEVEL LASER SYSTEM

A four energy level laser system

Highly simplified representation of Nd3+:YAG laser
 FOUR LEVEL LASER SYSTEM
 The change of population of atoms in the pump level as follows:

• A population inversion results when any pump energy is supplied,

• Unlike the three-level system, where inversion occurs only after a threshold pump rate is
reached, inversion occurs immediately in a four-level system. Most
35
FOUR LEVEL LASER SYSTEM
• A population inversion results when any pump energy is supplied
• Unlike the three-level system, where inversion occurs only after a
threshold pump rate is reached, inversion occurs immediately in a four-
level system.
• Most common laser systems, such as the He Ne, argon, CO2, and YAG
lasers, are four level systems.

Energy-level populations for

a four-level system

36
4.2 STIMULATED
EMISSION RATE
AND EMISSION
CROSS-SECTION

37
EINSTEIN COEFFICIENTS
 In atomic, molecular, and optical physics, the Einstein
coefficients are quantities describing the probability of
absorption or emission of a photon by an atom or molecule.
 The Einstein A coefficients are related to the rate of spontaneous
emission of light, and the Einstein B coefficients are related to
the absorption and stimulated emission of light.
 "light" refers to any electromagnetic radiation, not necessarily in
the visible spectrum.
 These coefficients are named after Albert Einstein, who
proposed them in 1916.

38
EINSTEIN COEFFICIENTS
 A useful laser medium must have a higher efficiency of
stimulated emission compared with the efficiencies of
spontaneous emission and absorption.
 We need to determine the controlling factors for the rates of
stimulated emission, spontaneous emission, and absorption.
 The rate equations discussed before can be given in terms of
Einstein coefficients as well.

39
 EINSTEIN COEFFICIENTS

R12 = B12N1r(u) R21 = A21N2 + B21N2r(u)

-dN1 /dt Absorption -dN2 /dt Spontaneous Stimulated

emission emission
, are Einstein coefficients
𝑖𝑠 𝑝ℎ𝑜𝑡𝑜𝑛 𝑒𝑛𝑟𝑔𝑦 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 𝑝𝑒𝑟 𝑢𝑛𝑖𝑡 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦

We need A21, B12 and B21

 EINSTEIN COEFFICIENTS

Consider equilibrium R12 = R21

N2 / N1 = exp[-(E2 - E1)/kBT]
Boltzmann statistics
E1 and E2 have the same degeneracy

8h 3
Planck’s black body  eq ( ) 
radiation law   h  
c exp
3
  1
  k BT  
EINSTEIN COEFFICIENTS

B12 = B21
A21/B21 = 8phu /c
3 3

R21 (stim ) B21N 2  ( ) B21 ( ) c3

    ( )
R21 (spon ) A21N 2 A21 8h 3

R21 (stim ) N2

R12 (absorp ) N 1
LASER REQUIREMENTS

R21 (stim ) N2
 Population inversion
R12 (absorp ) N 1
R21 (stim )
  ( ) Optical cavity
R21 (spon )
CONCLUSIONS
 For stimulated photon emission to exceed photon absorption, we
need to achieve population inversion, that is, N2 > N1.
 For stimulated emission to far exceed spontaneous emission, by
Eq. (4.2.9) we must have a large photon concentration, which is
achieved by building an optical cavity to contain the photons.

44
SPONTANEOUS DECAY TIME

R12 = -dN1/dt and R21 = -dN2/dt

R21 = rate at which N2 is decreasing by spontaneous and stimulated

emission

Consider N2 changes by spontaneous emission

dN2/dt = -A21N2 = - N2/tsp,

tsp = 1/A21 = spontaneous decay time; or the lifetime of level E2.

EXAMPLE: Minimum pumping power for three level laser systems
Consider the 3-level system Figure 4.2(a). Assuming that the transitions
from E3 to E2 are fast, and the spontaneous decay time from E2 to E1 is tsp,
show that the minimum pumping power Ppmin that must be absorbed by the
laser medium per unit volume for population inversion (N2 > N1) is
Ppmin/V = (N0/2)hu13/tsp Minimum pumping for population inversion for 3-level laser (4.2.12)
where V is the volume, N0 is the concentration of ions in the medium and
hence at E0 before pumping. Consider a ruby laser in which the
concentration of Cr3+ ions is 1019 cm-3, the ruby crystal rod is 10 cm long and
1 cm in diameter. The lifetime of Cr3+ at E2 is 3 ms. Assume the pump takes
the Cr3+ ions to the E3-band in Figure 4.3 (a), which is about 2.2 eV above
E0. Estimate the minimum power that must be provided to this ruby laser to
achieve population inversion.

Solution
Consider the 3-level system in Figure 4.2 (a). To achieve population inversion we
need to get half the ions at E1 to level E2 so that N2 = N1 = N0 /2 since N0 is the total
concentration of Cr3+ ions all initially at E1. We will need [(N0 / 2)hu13 × volume]
amount of energy to pump to the E3-band.
EXAMPLE: Minimum pumping power for three level laser systems

Solution (continued)

The ions decay quickly from E3 to E2. We must provide this pump energy before
the ions decay from E2 to E1, that is, before tsp Thus, the minimum power the ruby
needs to absorb is
Ppmin = V(N0 / 2)hu13/tsp

which is Eq. (4.2.12). For the ruby laser

Ppmin = [p(0.5 cm)2(10 cm)][(1019 cm-3)/2](2.2 eV)(1.6×10-19 J/eV)]/(0.003 s)

= 4.6 kW

The total pump energy that must be provided in less than 3 ms is 13.8 J.
 ABSORPTION CROSS SECTION

Optical power absorbed by an ion

= Light intensity × Absorption cross section of ion

= Isab
I
   ab N1  
Ix
EMISSION CROSS SECTION

Stimulated optical power emitted by an ion

= Light intensity × Emission cross section of ion

= Isem
I
  em N 2
Ix
OPTICAL GAIN COEFFICIENT

Definition

 I  g = semN2 - sabN1
g 
 I x  net g(u) = sem(u)N2 - sab(u)N1
Optical gain is G
G = exp(gL)
51
 He-Ne LASER

Ali Javan and his associates William Bennett Jr. and Donald Herriott at Bell Labs wee
first to successfully demonstrate a continuous wave (CW) helium-neon laser operation
(1960-1962). (Reprinted with permission from Alcatel-Lucent USA Inc)
He-Ne LASER: PRINCIPLES

He* + Ne ® He + Ne*
 He-Ne LASER
What are other lasing emissions?
 What are other lasing emissions?

Wavelength (nm) 543.5 594.1 612 632.8 1523

Color Green Yellow Orange Red Infrared
EXAMPLE: Efficiency of the He-Ne laser
A typical low-power 5mW He-Ne laser tube operates at a dc voltage of 2000 V
and carries a current of 7 mA. What is the efficiency of the laser ?

Solution
From the definition of efficiency,

Output Light Power 5  10 3 W

Efficiency  
Input Elec trical Power (7  10 3 A)( 2000 V)
 0.036%
Typically He-Ne efficiencies are less than 0.1%. What is important is the highly
coherent output radiation. Note that 5 mW over a beam diameter of 1 mm is 6.4 kW m -2.
LASER DIODE
PRINCIPLE

57
LASER DIODE PRINCIPLE
 Consider a p-n junction
 In order to design a laser diode, the p-n junction must be heavily
doped.
 In other word, the p and n materials must be degenerately doped
 A degenerately doped semiconductor is one that is so heavily
doped that it starts acting like a metal.
 By degenerated doping, the Fermi level of the n-side will lies in
the conduction band whereas the Fermi level in the p-region will
lie in the valance band.
The highest energy level that an electron can occupy at the absolute zero
temperature is known as the Fermi Level. The Fermi level lies between the
valence band and conduction band because at absolute zero temperature,
the electrons are all in the lowest energy state.
 DIODE LASER OPERATION
p+ Junction n+
E
c
p+ n+
E
g eV E E
o c Fn
Inversion E
E region c
E
Ev Holes in VB E g eV
Fp Electrons Electrons in CB E Fn
c
E
Fp

E
(a) v (b)

•P-n junction must be degenerately doped. V

•Fermi level in valance band (p) and •Forward bias, eV> Eg
conduction band (n). •Built in potential diminished to zero
•No bias, built n potential; eVo barrier to •Electrons and holes can diffuse to the space
stop electron and holes movement charge layer
Application of Forward Bias

Suppose that the degenerately doped p-n

junction is forward biased by a voltage
greater than the band gap; eV > Eg
The separation between EFn and EFp is now
the applied potential energy
The applied voltage diminished the built-in
potential barrier, eVo to almost zero.
Electrons can now flow to the p-side
Holes can now flow to the n-side
 POPULATION INVERSION IN
DIODE LASER
More electrons in
the conduction
band near EC
CB
EFn
Electrons in CB

eV Eg
Than electrons in
Holes in VB the valance band
EFp near EV
VB

EFn-EfP = eV
There is therefore a population inversion between
eV > Eg
energies near EC and near EV around the junction.
eV = forward bias voltage
This only achieved when degenerately doped p-n
Fwd Diode current pumping  junction is forward bias with energy > Egap
injection pumping
THE LASING ACTION
 The population inversion region is a layer along the junction 
also call inversion layer or active region
 Now consider a photon with E = Eg
 Obviously this photon can not excite electrons from EV since
there is NO electrons there
 However the photon CAN STIMULATE electron to fall down
from CB to VB.
 Therefore, the incoming photon stimulates emission than
absorption
 The active region is then said to have ‘optical gain’ since the
incoming photon has the ability to cause emission rather than
being absorbed.
PUMPING MECHANISM IN LASER
DIODE
 It is obvious that the population inversion between energies near
EC and those near EV occurs by injection of large charge carrier
across the junction by forward biasing the junction.
 Therefore the pumping mechanism is FORWARD DIODE
CURRENT  Injection pumping
 For Successful Lasing Action:
1. Optical Gain (not absorb)
Achieved by population inversion
2. Optical Feedback
Achieved by device configuration
Needed to increase the total optical amplification by
making photons pass through the gain region multiple
times
Insert 2 mirrors at each end of laser
This is term an oscillator cavity or Fabry Perot cavity
Mirrors are partly transmitted and party reflected
LASER DIODE
CHARACTERISTI
CS

67
OUTPUT SPECTRUM
 The output spectrum from a laser diode depends on two factors:
 the nature of the optical resonator used to build the laser oscillations and
 the optical gain curve (lineshape) of the active medium.

 The optical resonator is essentially a Fabry–Perot cavity with a

length L, width W, and height H.

68
FABRY–PEROT CAVITY

69
 OPTICAL OUTPUT VS. DRIVE CURRENT
• The LD output characteristics are
temperature sensitive.
• Figure shows the changes in the optical
output power vs. diode current
characteristics with the case
temperature.
• As the temperature increases, the
threshold current increases steeply,
typically as the exponential of the
• absolute temperature, that is,

• where A and To are constants that have the units of current (A) and
temperature (K), respectively;
• is sometimes called the characteristic temperature for the Ith vs. T
dependence.
 LASER DIODE EFFICIENCIES
Slope Efficiency

 Po  Po
slope    
 I  abovethreshold I  I th

Increase in optical output power

slope 
Increase in input current above threshold
Slope efficiency determines the optical power Po in the lasing emission in terms
of the diode current I above the threshold current Ith.
 LASER DIODE EFFICIENCIES

External Quantum Efficiency

Number of output photons from the diode per unit second

EQE 
Number of injected electrons into the diode per unit second

Po / h ePo
EQE  
I /e Eg I
 LASER DIODE EFFICIENCIES

External Differential Quantum Efficiency

Increase in number of output photons from diode per unit second

 EDQE 
Increase in number of injected electrons into diode per unit second

Po / h e  e  Po
 EDQE   slope 
I / e h  E g  I  I th
 LASER DIODE EFFICIENCIES

Internal Quantum Efficiency

Rate of radiative combination

IDQE 
Rate of all recombination process

1/ r
IQE 
1 /  r  1 /  nr

tnr = Nonradiative recombination time

tr = Radiative recombination time

 LASER DIODE EFFICIENCIES

Extraction efficiency
=
(Loss from the exit cavity end) / (Total loss)

hEE = (1/2L)ln(1/R1) / at
Suppose that we are interested in the output from one of the mirrors,
say R1. The loss at R1 is the transmitted output. If there were no
losses at the other mirror (R2 = 1) and no internal losses (), then the
radiation in the cavity would eventually be coupled out from R1 and
the EE efficiency would be unity. Thus, what is important is the
relative ratio of the loss due to R1 to all losses in the cavity.
The extraction efficiency is defined as the ratio of loss coefficient due
to R1 to that due to total losses,
LASER DIODE EFFICIENCIES

Power Conversion Efficiency

Optical output power Po  Eg 
 PCE     EQE  
Electrical input power IV  eV 

Po  Eg 
PCE   EQE  
IV  eV 
 LASER DIODE EFFICIENCIES
Typical characteristics for a few selected red and violet commercial laser diodes. All LDs are
MQW structures and have FP cavities. Violet lasers are based on InGaN/GaN MQW, and red
LEDs use mainly AlGaInP/GaInP MQW.

LD Po l Ith I V q q// hslope hPCE

(mW) (nm) (mA) (mA) (V) (mW/mA) %
Red 500 670 400 700 2.4 21 10 1.0 30
Red 100 660 75 180 2.5 18 9 1.0 22
Red 50 660 60 115 2.3 17 10 0.90 19
Red 10 639 30 40 2.3 21 8 1.0 11
Violet 400 405 160 390 5.0 45 15 1.7 21
Violet 120 405 45 120 5.0 17 8 1.6 20
Violet 10 405 26 35 4.8 19 8.5 1.1 6.0
 LASER DIODE EFFICIENCIES

Typical values for the threshold current Ith, slope efficiency (hslope ) and power conversion efficiency
(hPCE) for 36 commercial red LDs with different optical output powers from 3 mW - 500 mW.
EXAMPLE: Laser diode efficiencies for a sky blue LD
Consider a 60 mW blue LD (Nichia SkyBlue NDS4113), emitting at a peak wavelength of 488 nm. The
threshold current is 30 mA. At a forward current of 100 mA and a voltage of 5.6 V, the output power is
60 mW. Find the slope efficiency, PCE, EQE and EDQE.

Solution
From the definition in Eq. (4.12.2),
hslope = Po / (I - Ith)
= (60 mW) / (100 - 30 mA) = 0.86 mW/mA-1
From Eq. (4.12.8), PCE is
hPCE = Po / IV
= (60 mW) / [(100 mA)(5.6 V)] = 0.11 or 11%
We can find the EQE from Eq. (4.12.3) but we need hu, which is hc/l. In eV,
hu (eV) = 1.24 / l (mm)
= 1.24 /0.488 = 2.54 eV
EQE is given by Eq. (4.12.3)
hEQE = (Po /hu) / (I /e )
= [(60×10-3)/(2.54×1.6×10-19)]/[(100×10-3) /(1.6×10-19)]
= 0.24 or 24%
EXAMPLE: Laser diode efficiencies for a sky blue LD
Solution (continued)

Similarly, hEDQE is given by Eq. (4.12.4b) above threshold,

hEDQE = (DPo /hu) / (DI /e )  (Po /hu) / [( I - Ith )/e )]

= [(60×10-3)/(2.54×1.6×10-19)] / [(100×10-3-30×10-3)/1.6×10-19)]

= 0.34 or 34%

The EDQE is higher than the EQE because most injected electrons above Ith are
used in stimulated recombinations. EQE gauges the total conversion efficiency from
all the injected electrons brought by the current to coherent output photons. But, a
portion of the current is used in pumping the gain medium.
EXAMPLE: Laser diode efficiencies
Consider an InGaAs FP semiconductor laser diode that emits CW radiation at 1310 nm. The cavity
length (L) is 200 mm. The internal loss coefficient as = 20 cm-1, R1 = R3  0.33 (cleaved ends). Assume
that internal differential quantum efficiency, IDQE, is close to 1. The threshold current is 5 mA. What is
the output power Po at I = 20 mA? The forward voltage is about 1.3 V. What is the EDQE and
conversion efficiency?
Solution
From the definition of IDQE in Eq. (4.12.6), the number of internal coherent
photons generated per second above threshold is hIDQE(I - Ith)/e. Thus,

Internal optical power generated = hu ×hIDQE(I - Ith)/e

The extraction efficiency hEE then couples a portion of this optical power into the
output radiation. The output power Po is then hEE × hu ×hIDQE(I - Ith)/e. Thus,

Po = hEEhIDQEhu(I - Ith)/e Output power vs current (4.12.9)

The slope efficiency from Eq.(4.12.2) is

hslope = DPo / DI = hEEhIDQE(hu/e ) Slope efficiency (4.12.10)
EXAMPLE .3: Laser diode efficiencies
Solution (continued)

Further, from the definition of EDQE and Eq.(4.12.9) is

hEDQE = (DPo / hu) / (DI /e) = (Po / hu) / [(I - Ith)/e] = hEEhIDQE

External differential quantum efficiency (4.12.11)

We can now calculate the quantities needed. The total loss coefficient is

at = as + (1/2L) ln (1 / R1R2)

=2000 + (2×200×10-6)-1 ln(0.33×0.33)-1 = 7543 m-1

The extraction efficiency is

hEE = (1/2L)ln (1/R1) / at = (2×200×10-6)-1 ln(1/0.33) / (7543)

= 0.37 or 37%
EXAMPLE: Laser diode efficiencies
Solution (continued)

Thus, using I = 20 mA in Eq. (4.12.9),

Po = (0.37)(1)[(6.62×10-34)(3×108)/(1310×10-9)][(0.02 - 0.005) / (1.6×10-

19
)] = 5.2 mW

The slope efficiency from Eq. (4.12.10) is

hslope = DPo / DI = (5.2 mW - 0) / (20 mA - 5 mA) = 0.35 mW mA-1

The EDQE from Eq. (4.12.11) is

hEDQE = hEEhIDQE = 0.37 or 37%

The power conversion efficiency hPCE = Po / IV

= 5.2 mW / (20 mA × 1.3 V)
= 0.20 or 20%
LASER
DIODE
EQUATION

A highly simplified and idealized description of a

semiconductor laser diode for deriving the LD
equation. (a) The heterostructure laser diode structure.
(b) The current I injects electrons in the conduction
band, and these electrons recombine radiatively with
the holes in the active region. (c) The coherent
radiation intensity across the device; only a fraction G
is within the active region where there is optical gain.
(d) Injected electron concentration n and coherent
radiation output power Po vs. diode current I. The
current represents the pump rate.
 LASER DIODE EQUATION

I n
  CnNph
eLWd τ r
Radiative lifetime
Rate of electron injection by current I
= Rate of spontaneous emissions
+ Rate of stimulated emissions

N ph
 CnN ph Rate of coherent photon loss in the cavity
 ph = Rate of stimulated emissions

Photon cavity lifetime

 LASER DIODE EQUATION

I th nth
 Threshold

eLWd τ r
1
nth  Threshold

C ph

Substitute back into steady state rate equation

I

nth
 CnthN ph
 ph
eLWd τr N ph  ( I  I th )
eLWd
LASER DIODE EQUATION
 ph
N ph  ( I  I th )
eLWd
( 12 N ph )(Cavity Volume)(Ph oton energy)
Po  (1  R )
t

 hc 2 ph (1  R ) 
Po    ( I  I th )
 2enL 

 hc 2 ph (1  R ) 
Io    ( J  J th )
 2end 
Light intensity = Po / A
THRESHOLD GAIN

1  1 
g th   t   s  ln  
2 L  R1R2 

G= Fraction of the coherent optical

radiation within the active region

The gain g works on the radiation within the

cavity, which means that we must multiply g
with G to account for less than perfect optical
confinement
 OPTICAL GAIN CURVE
(a)
(b)

(a) The density of states and energy distribution of electrons and holes in the conduction and
valence bands respectively at T > 0 in the SCL under forward bias such that EFn - EFp > Eg.
Holes in the VB are empty states. (b) Gain vs. photon energy (hu).
 OPTICAL GAIN CURVE

Optical gain g vs. photon energy for an InGaAsP active layer (in a 1500 nm LD) as a function of injected
carrier concentration n from 1×1018 to 3×1018 cm-3. (The model described in Leuthold et al, J. Appl. Phys.,
87, 618, 2000 was used to find the gain spectra at different carrier concentrations.) (Data combined from J.
Singh, Electronic and Optoelectronic Properties of Semiconductor Structures, Cambridge University Press,
203, p390; N.K. Dutta, J. Appl. Phys., 51, 6095, 1980; J. Leuthold et al, J. Appl. Phys., 87, 618, 2000.)
OPTICAL GAIN CURVE

The dependence of the peak gain coefficient (maximum g) on the injected carrier concentration n for
GaAs (860 nm), In0.72Ga0.28As0.6P0.4 (1300 nm), and In0.60Ga0.40As0.85P0.15 (1500 nm) active layers. (Data
combined from J. Singh, Electronic and Optoelectronic Properties of Semiconductor Structures,
Cambridge University Press, 203, p390; N.K. Dutta, J. Appl. Phys., 51, 6095, 1980; J. Leuthold et al, J.
Appl. Phys., 87, 618, 2000.)
EXAMPLE: Threshold current and optical output power from a Fabry-Perot
(FP) heterostructure laser diode
Consider GaAs DH laser diode that lases at 860 nm. It has an active layer (cavity) length L of 250 mm.
The active layer thickness d is 0.15 mm and the width W is 5 mm. The refractive index is 3.6, and the
attenuation coefficient as inside the cavity is 103 m-1. The required threshold gain gth corresponds to a
threshold carrier concentration nth  2×1018 cm-3. The radiative lifetime tr in the active region can be
found (at least approximately) by using tr = 1/Bnth, where B is the direct recombination coefficient, and
assuming strong injection as will be the case for laser diodes [see Eq. (3.8.7) in Chapter 3]. For GaAs, B
 2×10-16 m3 s-1. What is the threshold current density and threshold current? Find the output optical
power at I = 1.5Ith, and the external slope efficiency hslope. How would G = 0.5 affect the calculations?

Solution
The reflectances at the each end are the same (we assume no other thin film
coating on the ends of the cavity) so that R = (n-1)2/ (n+1)2 = 0.32. The total
attenuation coefficient at and hence the threshold gain gth, assuming G = 1 in Eq.
(4.13.9), is
1  1 
g th   t  (10 cm ) 
1
4
ln   = 55.6 cm-1
( 2  250  10 cm )  (0.32)( 0.32) 
From Figure 4.48(b), at this gain of 56 cm-1, nth  2×1018 cm-3. This is the threshold
carrier concentration that gives the right gain under ideal optical confinement,
with G = 1.
EXAMPLE: Threshold current and optical output power from a Fabry-Perot (FP)
heterostructure laser diode
Solution (continued)

The radiative lifetime tr = 1/Bnth = 1/[2×10-16 m3 s-1)( (2×1024 m-3)] = 2.5 ns

Since J = I/WL, the threshold current density from Eq. (4.13.4) is

nth ed (2  10 24 m 3 )(1.6  10 19 C)(0.15 10 6 m)
J th 
τr = (2.5 10 9 s)

= 1.9×107 A m-2 or 1.9 kA cm-2 or 19 A mm-2.

The threshold current itself is,
Ith = (WL)Jth = (5×10-6 m) (250×10-6 m)(1.9×107 A m-2)
= 0.024 A or 24 mA
The photon cavity lifetime depends on at, and is given by
tph = n/(cat) = 3.6 / [(3×108 m s-1)(5.56×103 m-1)]
= 2.16 ps
The laser diode output power is
 hc  ph (1  R ) 
2 34 12
(6.626  10 )( 3  10 ) ( 2.16  10 )(1  0.32)
8 2
Po   ( I  I th )  ( I  I th )
 2 e n  L  2 (1 . 6  10 19
)( 3 . 6 )( 860  10 9
)( 250  10 6
)
EXAMPLE: Threshold current and optical output power from a Fabry-Perot (FP)
heterostructure laser diode
Solution (continued)

That is Po = (0.35 W A-1)(I – Ith ) = (0.35 mW mA-1)(I – 24 mA)

When I = 1.5Ith = 36 mA,
Po = (0.35 mW mA-1)(36 mA-24 mA) = 4.2 mW
The slope efficiency is the slope of the Po vs. I characteristic above Ith,
Po  hc  ph (1  R ) 
2

slope   
I  2enL  = 0.35 mW mA-1

We can now repeat the problem say for G = 0.5, which would give Ggth = at, so that
gth = 55.6 cm-1 / 0.5 = 111 cm-1. From Figure 4.48 (b), at this gain of 111 cm-1, nth 
2.5×1018 cm-3. The new radiative lifetime,
tr = 1/Bnth = 1/[2.0×10-16 m3 s-1)(2.5×1024 m-3)] = 2.0 ns
The corresponding threshold current density is
Jth = nthed/tr = (2.5×1024 m-3)(1.6×10-19 C)(0.15×10-6 m)/(2.0×10-9 s)
= 30 A mm-2
and the corresponding threshold current Ith is 37.5 mA
EXAMPLE: Threshold current and optical output power from a Fabry-Perot (FP)
heterostructure laser diode
Solution (continued)

There are several important notes to this problem

• First, the threshold concentration nth  2×1018 cm-3 was obtained graphically from Figure
4.48 (b) by using the gth value we need.

• Second is that, at best, the calculations represent rough values since we also need to
know how the mode spreads into the cladding where there is no gain but absorption and, in
addition, what fraction of the current is lost to nonradiative recombination processes. We
can increase as to account for absorption in the cladding, which would result in a higher
gth, larger nth and greater Ith. If tnr is the nonradiative lifetime, we can replace tr by an
 1   r1   nr
1

effective recombination time t such that , which means that the threshold
current will again be larger. We would also need to reduce the optical output power since
some of the injected electrons are now used in nonradiative transitions.

• Third, is the low slope efficiency compared with commercial LDs. hslope depends on tph,
the photon cavity lifetime, which can be greatly improved by using better reflectors at the
cavity ends, e.g. ,by using thin film coating on the crystal facets to increase R.
 Distributed Bragg Reflector (DBR) LDs

(a) The basic principle of the Distributed Bragg Reflection (DBR) laser. (b)
Partially reflected waves at the corrugations can only constitute a reflected
wave when the wavelength satisfies the Bragg condition. Reflected waves A
and B interfere constructively when q(lB/n) = 2L. (c) Typical output
spectrum. SMSR is the side mode suppression ratio.
 Distributed Feedback (DFB) LDs

LEFT: Distributed feedback (DFB) laser structure. The mode field diameter is
normally larger than the active layer thickness and the radiation spreads into the
guiding layer.

RIGHT: There are left and right propagating waves, partial reflections from the
corrugation, and optical amplification within the cavity, which has both the active
layer and the guiding layer.
 Distributed Feedback (DFB) LDs

2B
m   B  ( m  1)
2nL

LEFT: Ideal lasing emission output has two primary peaks above and
below lB. RIGHT: Typical output spectrum from a DFB laser has a single
narrow peak with a dl typically very narrow, and much less than 0.1 nm
 Distributed Bragg Reflector (DBR) and Distributed
Feedback (DFB) LDs

Selected properties of DBR, DFB and external cavity (EC) laser diodes
Note: fm is 10-15 s; du and dl are spectral widths (FWHM). SMSR is the side
mode suppression ratio, TEC is a thermoelectric cooler

LD l0 du, dl SMSR Po I hslope Comment

(nm) dB mW mA mA

DBRa 1063 2 MHz, 45 80 200 0.8 GaAs DBR LD for spectroscopy and metrology, includes
8 fm monitor current, TEC and thermistor.
DFBb 1063 2 MHz, 45 80 190 0.2 GaAs DFB LD for spectroscopy and metrology, includes
8 fm monitor current, TEC and thermistor
DFBc 1550 10 MHz, 45 40 300 0.3 Pigtailed to a fiber, includes monitor current, TEC and
0.08 pm thermistor. CW output for external modulation. For use in
long haul DWDM.
DFBd 1653 0.1 nm 35 5 30 0.23 Pigtailed to a single mode fiber, includes monitor current,
TEC and thermistor. Mainly for fiber optic sensing.

ECe 1550 50 kHz; 45 40 300 0.2 Pigtailed. Tunable over Du = 3 GHz. Mainly for
0.4 fm communications

a
Eagleyard, EYP-DBR-1080-00080-2000-TOC03-0000; bEagleyard, EYP-DFB-1083-00080-1500-TOC03-0000;
c
Furukawa-Fitel, FOL15DCWD; dInphenix, IPDFD1602; eCovega SFL1550S, marketed by Thorlabs.
Example: DFB LD wavelength
Consider a DFB laser that has a corrugation period L of 0.22 mm and a grating length of 400 mm. Suppose
that the effective refractive index of the medium is 3.5. Assuming a first order grating, calculate the Bragg
wavelength, the mode wavelengths and their separation.

Solution
The Bragg wavelength is

2n 2(0.22 μm)(3.5) = 1.5400 mm.

B  
q 1
and the symmetric mode wavelengths about lB are
B2 (1.5400 μm) 2
m  B  (m  1)  1.5400  (0  1)
2n L 2(3.5)(400 μm)
so that the m = 0 mode wavelengths are
l0 = 1.53915 or 1.54085 mm.
The two are separated by 0.0017 mm, or 1.7 nm. Due to a design
asymmetry, only one mode will appear in the output and for most practical purposes
the mode wavelength can be taken as lB. Note: The wavelength calculation was
kept to five decimal places because lm is very close to lB..
 External Cavity Laser Diodes (ECLD)

A simplified diagram of an external cavity diode laser (ECDL), which uses an angled
interference filter (IF) to select the wavelength lo (depends on the angle of the IF), and the
optical cavity has a GRIN lens with one end coated for full reflection back to the LD. The output
is taken from the left facet of the LD.
 External Cavity Laser Diodes (ECLD)

LEFT: A commercial external cavity diode laser, based on the principle shown on the right. (US Patent
6,556,599, Bookham Technology). The output is a single mode at 785 nm ( 1.5 pm) with a linewidth less
than 200 kHz, and coupled into a fiber. The output power is 35 mW, and the SMSR is 50 dB. (ECDL,
SWL-7513-P. Courtesy of Newport, USA)
105