Laser Diodes

an Introduction

University of Hannover, Germany

Matthias Pospiech, Sha Liu
May 2004
Contents 1

Contents
1 The physics behind laser diodes 3
1.1 Band structure of semiconductor . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Recombination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 State density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Direct band gap and indirect band gap . . . . . . . . . . . . . . . . . . . 6
1.5 Optical Feedback Mechanism . . . . . . . . . . . . . . . . . . . . . . . . 7

2 Principles of AlGaAs Laser Diodes 8

2.1 Heterostructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 Threshold current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 Oscillation Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

3 Technology of laser diodes 14

3.1 Real lasers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.1.1 Gain-guided lasers . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.1.2 Index-guided lasers . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.1.3 Gain vs. index-guided structure . . . . . . . . . . . . . . . . . . . 16
3.2 DBR and DFB Lasers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2.1 Distributed Bragg Reflector (DBR) Laser . . . . . . . . . . . . . 17
3.2.2 Distributed Feedback (DFB) . . . . . . . . . . . . . . . . . . . . 17
3.2.3 DFB vs. DBR Lasers . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3 Laser arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4 VCSEL (Vertical Cavity Surface Emitting Lasers) . . . . . . . . . . . . 19

authors:
Matthias Pospiech ([email protected]), student of technical physics
Sha Liu ([email protected]), master student of physics
This work was done for the advanced laboratory work, 2004
This skript in available online under:
http://www.matthiaspospiech.de/studium/artikel/
2 Contents

Introduction
Laser diodes have grown to a key component in modern photonics technology. This
article provides a general introduction in the physics and technology of laser diodes.
First the physical concepts behind laser diodes are explained. In the following an
example of an AlGaAs laser is given and analyzed in details. The last part introduces
technological approaches to build competitive laser diodes. The focus is thereby on
single mode laser diodes. We consider ways to introduce a waveguide in the laser diode
and concepts to make the laser diode wavelength selective. Briefly we take a look at laser
diode arrays for use in high power applications and finally we introduce the VCSEL’s
(Vertical Cavity Surface Emitting Lasers).

Comparison with other laser systems

The following features distinguish laser diodes from other lasers:

• Compact. They are build on one chip that contains everything necessary for a
laser. This enables semiconductor lasers to be easily be inserted in other instru-
ments.

• High efficiency up to 50%. This enables them to be driven by low electrical power
compared to other lasers.

• Direct excitation with small electric currents, so that conventional transistor based
circuits can supply the laser

• Possibility of direct modulation with applied current

• Small beam waist

Low costs due to mass production and high reliability made them a key component in
various applications. However they also have some disadvantages. They are highly sen-
sitive to temperature. Although this allows wide wavelength tunability it is unwanted
for most applications. Another important unwanted feature is their highly divergent
beam. [1]
 3

1 The physics behind laser diodes

1.1 Band structure of semiconductor
In a crystal, the discrete energy levels of the individual atom broaden into energy bands.
Each quantum state of the individual atom gives rise to a certain energy band. The
bonding combinations of states become the valence bands (VB) of the crystal, and
the anti-bonding combinations of these states become the conduction band (CB). The
energy difference between VB and CB is called energy gap. If the valence bands are
partly filled, this material is p-type, if the conductive bands are partly filled, this
material is n-type. Here Fermi level is used to label the occupation conditions of
electrons in the semiconductor, it is the energy level to which electrons occupy. Fermi
level (EF P ) on p-type is near the valence band and EF N on the n-type is near the
conductive band.
When two semiconductors with different band structures are combined, a heterojunc-
tion is formed, a p-n heterojunction is called a diode. Electrons and holes transfer
to the other side, because of different Fermi levels respectively. They recombine with
each other, leaving the p-side with negative charge and n-side with positive charge, this
region is called space charge layer (SCL). A built-in voltage V0 appears because of the
charge transfer and recombination. When there is no applied voltage, the Fermi level
is continuous across the diode EF P = EF N , as indicated in Figure 1(a). The built-in
voltage prevents electrons in conduction band on n-side from diffusing into conduction
band on p-side, it is similar for holes in valence band, so the majority carriers can not
flow into the space charge layer. An applied voltage V can separate EF P and EF N by
eV, and the Fermi level is not continuous across the diode. The seperate Fermi level in
each side is called quasi Fermi level, as indicated in Figure 1(b).

(a) p-n junction without bias (b) p-n junction after forward bias

Figure 1: p-n junction with/without bias [2]

Due to the applied voltage, the recombination process takes place and the diode
current establishs. When applied voltage is greater than built-in voltage , the space
charge layer is no longer depleted. Now at the junction, more electrons are injected into
4 1 The physics behind laser diodes

the conductive band at energies near Ec than electrons in valence band at energies near
Ev . This is the population inversion, and the inversion region as indicated in Figure
1(b) is called active region. In Figure 2, the carrier concentration-x is shown in two
dimensional coordinate in the space charge layer (SCL). [2].

Figure 2: The carrier concentration in SCL under forward bias [2]

1.2 Recombination
There are three kinds of transitions that are important in laser diodes, which occur
between the conduction and valence bands of the material. They are stimulated absorp-
tion, spontaneous emission and stimulated emission in Figure 3.

(a) Stimulated Absorption (b) Stimulated Absorption (c) Stimulated Emission

Figure 3: Recombination [3]

After defining R(abs) , R(spon) , R(stim) as the rate of absorption, spontaneous emission
and stimulated emission respectively, the relationship between the three processes can
be described by the following equation

R(abs) = R(spon) + R(stim)

1.3 State density 5

And the rates can be expressed by the Einstein coefficients which are defined in the
following way:
B(12) transition probability of induced absorption
A(21) transition probability of spontaneous emission
B(21) transition probability of induced emission
Here we only cite the the ratio of spontaneous emission to stimulated emission, the
deduction details can be read in website [3]

A(21)
= %(~ω) exp[~ω/kB T ] − 1
B(21)

In the stimulated emission, a photon is strongly coupled with the electron, the photon
can cause the electron to decay to a lower energy level, releasing a photon of the same
energy. The emitted photon has the same direction and phase as the incident photon.
When the stimulated emission is dominant, the light is amplified, and laser occurs. From
this equation, we can see that stimulated emission is dominant when ~ω  kb T . From
Fermi-Drac statistic law, under this condition, the probability of finding an electron in
the conduction band has to be greater than the probability of finding an electron in
the valence band, so there must be a population inversion. As mentioned before, in
a laser diode, population inversion is achieved when EF N − EF P > Eg, where Eg is
the bandgap energy and EFc and EFv are the Fermi levels of the conduction band and
valence band, respectively. These Fermi levels can only be separated by pumping energy
in the form of electrical current into the semiconductor laser. Electrons and holes are
injected into the active region from n- and p-doped semiconductor cladding layers. The
injection current required to achieve lasing is known as the threshold current, details
will be given in section 2.2.

1.3 State density

When we calculate various optical properties, such as the rate of absorption or emission
and how electrons and holes distribute themselves within a solid, we need to know the
state density. The density of state is described in the number of available states per
unit volume per unit energy. So in conductive band and valence band, the state density
of energy is
3/2 
2m∗
 
1 
%c,v (E)dE = E 1/2 dE
2π 2 ~2

m∗ → mc or mv E → E − Ec or Ev − E
according to Fermi-Dirac statistic law, electrons in the band obey the following equa-
tions
1
fc (E) =
exp ((E − Fn )/kT ) + 1

1
fv (E) =
exp ((E − Fp )/kT ) + 1
6 1 The physics behind laser diodes

so in conductive band and valence band, the electrons and holes distributions are
 ∗ 3/2
1 2me (E − Ec )1/2 dE
nc (E)dE = 2 E > Ec
2π ~2 exp ((E − Fn )/kT ) + 1

3/2
2m∗h (Ev − E)1/2 dE

1
pv (E)dE = 2 E < Ev
2π ~2 exp ((Fp − E)/kT ) + 1
The above equations are mainly cited from website [3]
Figure 4 is the schematic graph of the distributions described by the functions above.
(a) is at 0 K, and the charges first fill the lowest energy state. (b) is at a certain
temperature above 0, so some charges are excited to higher energy states.

Figure 4: Electron and hole distribution [4]

1.4 Direct band gap and indirect band gap

The recombination process mainly depends on the band structure. Generally, there
are two kinds of band structure, direct band gap and indirect bandgap. Direct band
gap means that in the E-k diagram, electrons at the minimum of the conduction band
have the same momentum as electrons at the maximum of the valence band, and for an
indirect band gap, the electrons do not have the same momentum, as indicated in Figure
5. The recombination of an electron near the bottom of the conduction band with a hole
near the top of the valence band requires the exchange of energy and momentum. For
indirect band gap recombination, the energy may be carried off by a photon, but one
or more phonons are required to conserve momentum. This multiparticle interaction is
improbable and the recombination efficiency in the indirect band gap material is lower
than in the direct band gap material. [6]
The majority part of semiconductors are indirect band gap material, compared with
them, direct bandgap materials are preferred for laser diodes. Direct bandgap structures
1.5 Optical Feedback Mechanism 7

(a) direct-band-gap struc- (b) indirect-band-gap structure

ture

Figure 5: Recombination [5]

maximize the tendency of electrons and holes to recombine by stimulated emission,

thus increasing the laser effciency. For example, the direct band gap crystal aluminium
gallium arsenide (AlGaAs) is often used for laser diodes with wavelengths between
750nm and 880nm. Alx Ga1−x As, through changing the x, the ratio of the aluminium
to gallium can be adjusted to vary the band gap and thereby control the wavelength.
[7].

1.5 Optical Feedback Mechanism

In a high efficiency laser, a resonator must be formed which has the ability not only to
amplify the electromagnetic wave, but also to feedback it. A laser resonator generally
consists of two parallel mirrors perpendicular to the optical axis. Space between the
mirrors is partially occupied by the amplifying material. This structure, called a Fabry-
Perot Resonator, is obtained in a laser diode by cleaving the ends of the crystal. Because
the refractive index has a jump at the interface of the crystal and other material, the
mirror facet functions as a reflective surface. In some cases, special coatings are used to
enhance either the reflectivity r, or transmissivity t, of the facet. When the resonator is
brought to a state of population inversion, photons produced by spontaneous emission
are amplified and repeatedly reflected by the front and rear facets. In homojunction
LD, there is no optical confinement in the direction perpendicular to the optical axis,
so the elelctromagetic waves in any direction not parallel to the optical axis of the
resonator will pass through the sides of the resonator. In heterostructure LD, the
waveguide will confirm the wave in the active region (see section 2). The component
of the spontaneously emitted photons, which travels parallel to the optical axis, will be
repeatedly reflected by the mirror facets. As the electromagetic wave travels through
the semiconductor material, it is amplified by stimulated emission. At each reflection,
the wave is partially transmitted through the reflective facets. Laser oscillation begins
8 2 Principles of AlGaAs Laser Diodes

when the amount of amplification becomes equal to the total amount lost through the
sides of the resonator, through the mirror facets and through absorption by the crystal.
The details of Fabry-Perot can be found in the ‘Protokoll laser diodes, theory part’. [8]

Figure 6: Light feedback [3]

2 Principles of AlGaAs Laser Diodes

In the double heterostructure, stimulated emission occurs only within a thin active
layer of GaAs, which is sandwiched between p- and n- doped AlGaAs layers that have
a wider band gap. Laser diodes use heterojunctions to achieve simultaneous carrier and
photon confinement in the active region. A high laser efficiency demands that the light
and injected charge carriers be confined as closely as possible to the same volume.

2.1 Heterostructure

Figure 7: Double heterojunction laser diode [9]

2.1 Heterostructure 9

As illustrated in Figure 7, the AlGaAs Laser Diode consists of a double hetero-

junction formed by an undoped (or lightly p-doped) active region surrounded by higher
bandgap p and n Alx Ga1−x As cladding layers. The surrounding cladding layers pro-
vide an energy barrier to confine carriers to the active region. The actual operation
wavelengths may range from 750 – 880 nm due to the effects of dopants, the size of
the active region, and the compositions of the active and cladding layers. When a
certain parameter is fixed, the wavelength can vary in several nanometers due to other
variables. For example, when the active layer has an energy gap Eg = 1.424 eV, the
nominal emission wavelength is λ = hc/Eg = 871 nm. When a bias voltage is applied
in the forward direction, electrons and holes are injected into the active layer. Since the
band gap energy is larger in the cladding layers than in the active layer, the injected
electrons and holes are prevented from diffusing across the junction by the potential
barriers formed between the active layer and cladding layers (Figure 7). The electrons
and holes confined to the active layer create a state of population inversion, allowing
the amplification of light by stimulated emission. [9]
The cladding layers serve two functions. First, inject charge carriers. Second, light
confinement. Since the active region has a smaller bandgap than the cladding layers,
its refractive index will be slightly larger than that of the surrounding layers. The
GaAs refractive index at these wavelengths is n = 3.5 while the refractive index of
the Alx Ga1−x As cladding layers is slightly smaller. The Figure 8 indicates the elec-
tromagnetic field distribution due to the heterostructure. For the confinement in the

(a) Refractive index (b) Field distribution

Figure 8: Refractive index of the waveguide and the field distribution [4]

horizontal (lateral) direction, in real laser structures, index or gain guiding is always
used, as mentioned in section 3.1.
The characteristics of a three-layer slab waveguide are conveniently described in terms
of the normalized waveguide thickness D, defined as
  p
2π
D= d ηa2 − ηc2
λ
where ηa and ηc are the refractive indices of the active and cladding layers respectively
10 2 Principles of AlGaAs Laser Diodes

and d is the active layer thickness.

The confinement factor Γ, defined as the fraction of the electromagnetic energy of the
guided mode that exists within the active layer, is an important parameter representing
the extent to the active layer. Γ for a fundamental mode is approximately given by [1]

D2
Γ≈
2 + D2

For a GaAs/Al0.3 Ga0.7 As waveguide with d = 0.1 µm , Γ ≈ 0.27.

2.2 Threshold current

When a sufficient number of electrons and holes are accumulated to form an inverted
population, the active region exhibits optical gain and can amplify electromagnetic
waves passing through it, since stimulated emission overcomes interband absorption.
The wave makes a full round trip in the cavity without attenuation, which means that
the optical gain should equal the losses both inside the cavity and through the partially
reflecting end facets. Thus, the gain coefficient at threshold gth is given by the relation

1 1
Γgth = Γαa + (1 − Γ)αc + αs + ln
| {z } L R
αi

Here, αa and αc denote the losses in the active and cladding layers respectively, due
to free-carrier absorption. αs accounts for scattering loss due to heterointerfacial im-
perfections between the active and cladding layer. The first three loss terms on the
right-hand side combined are termed internal loss αi and add up to 10 to 20 cm−1 . The
reflection loss L1 ln R1 ≈ 40 cm−1 for L ≈ 300 mum, R ≈ 0.3) due to output coupling is
normally the largest among the loss terms.
There is a phenomenological linear relationship between the maximum gain g and the
∂g
injected carrier density n, supposing ∂n and nt are constant to a good approximation.

∂g
g(n) = (n − nt )
∂n

Here, ∂g/∂n is termed differential gain, and nt denotes the carrier density required
to achieve transparency where stimulated emission balances against interband absorp-
tion corresponding to the beginning of population inversion. Taking GaAs lasers
as an example, ∂g/∂n ≈ 3.5 × 10−6 cm2 and nt ≈ 1.5 × 1018 cm−3 , and remember
Γ=0.27, α=10 cm−1 and L−1 ln R−1 =40 cm−1 , we get a threshold carrier density n
≈ 2 × 1018 cm−3 . The threshold current density Jth is expressed as

edn
Jth =
τs

where τs is the carrier lifetime due to spontaneous emission. Assuming that τs =3ns and
d=0.1µm, we obtain a threshold current density Jth ≈ 1KAcm−2 . [1]
2.3 Oscillation Modes 11

2.3 Oscillation Modes

In laser diodes, length determines longitudinal modes where width and height of the
cavity determines transverse or lateral modes.

Figure 9: AlGaAs LD Spectrum [9]

Longitudinal Mode The longitudinal modes, or optical resonances of the Fabry-Perot

formed by the cleaved facet end mirrors, are determined by the length L of the resonator
and the refractive index n of the semiconductor. For electromagnetic wave of wavelength
λ, the half-wavelength in the medium is λ/2n, and for a standing wave, qλ/2n = L, q is
the integral multiple. Variation of the integer q by 1, causes a wavelength variation, ∆λ
of 0.35nm, and the laser resonator may simultaneously support several standing waves,
or longitudinal modes, of slightly different wavelength. In a laser diode, the oscillation
arises at the wavelength corresponding to the band gap energy of the semiconductor,
the intensity decreases as the wavelength goes far away from the central wavelength,
as indicated in Figure 9. Since the band gap energy varies with temperature, the
wavelength with maximum intensity also varies with temperature. For example, the
AlGaAs laser diode, the wavelength increases by approximately 0.23nm for an increase
in temperature of 1 degree. The free spectral range ff sr is defined as
 
1 1 c∆λ c
ffsr = f2 − f1 = c − = 2 =
λ2 λ1 λ 2nL
The peak separation is defined as
c
∆λ = λ2 − λ1 = λ − c
λ + ff sr

Figure 9 is an example with the center wavelength λ = 817.5 nm and the peak
separation ∆λ = 0.45 nm, the free spectral range ffsr = c∆λ
λ2
= 200GHz. [9]
12 2 Principles of AlGaAs Laser Diodes

Figure 10: Index guided laser diode [9]

Transverse Modes The transverse mode represents the state of the electromagnetic
standing wave in the direction perpendicular to the optical axis of the laser resonator.
The transverse mode has two components, one parallel and the other perpendicular to
the active layer of the laser. As stated above, there exist steps in the refractive index
on each side of the active layer, which serve to confine the light to the active layer. The
laser beam displays a diverging field due to the diffraction at the ends of the cavity. Fig
10 shows the construction of a typical index guided laser diode with cladding layers,
electrodes, and GaAs active region. The laser cavity mirrors are the end facets of the
semiconductor crystal, which has been cleaved.

Figure 11: Beam profile from an index guided laser diode [9]

The dimensions of the crystal determine the pattern of the emitted beam (the trans-
verse mode pattern) and also the possible laser emission frequencies (the longitudinal
2.3 Oscillation Modes 13

mode pattern). The output pattern is dominated by diffraction because the width
W ≈ 10µm and height H ≈ 2µm of typical LDs are comparable to the emission wave-
length. The divergence angle of the emission along these two directions is inversely
proportional to the dimensions as shown in Fig 11. The angular width ϑ of the emis-
sion pattern from a slit or rectangular opening of width d is

ϑ = 2 arcsin(λ/d)

For example, a laser wavelength of 850 nm and strip width W = 10µm has a divergence
angle ϑw ≈ 10 deg, ϑH ≈ 45 deg as shown in Fig 11. The dimensions W and H of the
active region of a laser diode can be determined by measuring the output emission cone
angles. The smaller the aperture the greater the diffraction, with a sufficiently small
W and H, only the lowest transverse mode TEM00 exits. [9]
14 3 Technology of laser diodes

3 Technology of laser diodes

In the further proceedings we are going to take a closer look at different techniques of
constructing a laser diode. The focus thereby is on single mode laser diodes. Single
mode waves are preferred in most cases. The realm of multimode lasers are high power
lasers, where single mode operating is not of importance.

3.1 Real lasers

Most modern semiconductor lasers adopt a structure, where the current is injected only
within a narrow region beneath a stripe contact several µm wide, in order to keep the
threshold current low and to control the optical field distribution in the lateral direction.
Compared with broad-area lasers, where the entire laser chip is excited, the threshold
current of lasers with stripe geometry is reduced roughly proportional to the area of
contact.
We differ mainly between two different types of structures. In case that the current
injection is restricted to a small region along the junction plane these are termed gain-
guided. Devices incorporating a built-in refractive index variation in the lateral direction
are termed index-guided lasers. [1]

3.1.1 Gain-guided lasers

With these lasers the current injection is re-
stricted to a narrow region beneath a stripe. The
active region is planar and continuous. Lasing
however occurs only in a limited region of the ac-
tive layer beneath the stripe contact where high
density of current flows. This horizontal confine-
ment of the em–wave propagating through the ac-
tive region is thereby accomplished by the small
refractive index variation produced by the cur-
rent generated population inversion. If the em–
wave spreads in the horizontal plane outside of Figure 12: Gain-guided laser [10]
the horizontal dimensions of the stripe, it will be
absorbed by the unexcited region of the active layer. In the vertical directions the lower
refractive indices of the surrounding layers reflects the em–wave back into the active
region. [11, 10]
The current restriction serves several purposes: [11]

• it allows CW operation with reasonable low threshold

currents (10-100 mA)

• it can allow fundamental-mode operation along the junc-

tion plane, which is necessary for applications where the
em–wave is coupled into a single mode optical fiber

• the requirements for heat sinking are low.

Figure 13: Cross section
[11]
3.1 Real lasers 15

Such lasers are determined gain-guided lasers because the optical intensity distribu-
tion in the lateral direction is determined by the gain profile produced by carrier density
distribution.[1]

3.1.2 Index-guided lasers

The transversal mode control in laser diodes can
be achieved using index guiding along the junction
plane. The mode control is necessary for improving
the em–wave current linearity and the modulation
response of lasers. The active region is thereby sur-
rounded by materials with lower refractive indices in
both the vertical (y) and lateral (x) transverse direc-
tions – the active region is buried in lower refractive
indices layers (e.g. InP) on all sides. For this reason,
these lasers are called buried-heterostructure lasers.
Figure 14: Index-guided laser [9]
[9, 1]
The lateral index step along the junction plane is
about two magnitudes larger than the carrier induced effects. As a result, the lasing
characteristics of buried-heterostructure lasers are primarily determined by the rect-
angular waveguide that confines the mode inside the buried region. The transverse
dimensions of the active region and the index discontinuities are chosen so that only
the lowest order transverse modes can propagate in the waveguide. [9, 11]
These index guided devices produce beams with much higher beam quality, but are
according to [10] typically limited in power to only a few hundred milliwatts. Another
important feature of this laser is the confinement of the injected carriers to the active
region. [9]

Buried Heterostructure Lasers Figure 15 shows the schematic cross sections of different
types of buried-heterostructure lasers. They are (i) the buried heterostructure that is
also called etched-mesa buried-heterostructure (EMBH) to distinguish it from other BH
lasers. And (ii) the double-channel planar buried heterostructure (DCPHB) laser. [11]

Figure 15: Schematic cross sections of buried-heterostructure lasers [11]

The fabrication of the EMBH structure is technologically complex and two epitaxial
16 3 Technology of laser diodes

steps are necessary. Less critical is the DCPHB laser although this necessitate two
epitaxial steps as well. There are other structures which are easier in fabrication, but
these are not discussed here.

3.1.3 Gain vs. index-guided structure

Historically, gain-guided devices based on the stripe geometry were developed first in
view of their ease of fabrication. However gain-guided lasers exhibit higher threshold
currents than index-guided lasers and have other undesirable characteristics that be-
come worse as the laser wavelength increases. [11]
In contrast, the fabrication of strongly index-guided lasers requires either a single
epitaxial growth over non-planar surfaces or two epitaxial growths and, in addition,
careful attention to processing. In spite of these difficulties in fabrication, their superior
performance characteristics – low threshold current, stable fundamental mode operation,
and good high speed modulation characteristics – make them a prime candidate for
high-performance applications.[11]

3.2 DBR and DFB Lasers

The standard Fabry Perot Laser are not wavelength selective. This lead to lasing of
higher modes and allows for mode jumps. So the question is, how wavelength selection
can be achieved. Shorter optical cavities are not practical since it is difficult to handle
very small chips. Another possible method is to insert an optical feedback in the device
to eliminate other frequencies.
Periodic gratings incorporated within the lasers waveguide can be utilized as a means
of optical feedback. Devices incorporating the grating in the pumped region are termed
Distributed Feedback (DFB) lasers, while those incorporating the grating in the passive
region are termed Distributed Bragg Reflector (DBR) Laser.
DFB and DBR lasers oscillate in a single-longitudinal mode even under high-speed
modulation, in contrast to Fabry-Perot lasers, which exhibit multiple-longitudinal mode
oscillation when pulsed rapidly. [1]

Figure 16: Schematic Illustration of DFB and DBR Lasers [11]

3.2 DBR and DFB Lasers 17

3.2.1 Distributed Bragg Reflector (DBR) Laser

DBR mirrors The gratings or distributed Bragg reflec-
tors (DBRs) are used for one or both cavity mirrors.
The grating thereby consist of corrugations with a pe-
riodic structure. They are used because of their fre-
quency selectivity of single axial mode operation. The
period of grating is chosen as half of the average optical
wavelength, which leads to a constructive interference
between the reflected beams. Significant reflections can
also occur in harmonics frequencies of the medium. The
corrugations are typically etched on the surface of the
waveguide, and these are refilled with a different index
material during a second growth. [4] This is illustrated
in figure 17. Figure 17: Schematic of a DBR
The concept of the grating is that many reflections mirror [4]
can add up to a large net reflection. At the Bragg
frequency the reflections from each discontinuity add up exactly in phase. As the
frequency is deviated from the Bragg condition, the reflections from discontinuities
further into the grating return with progressively larger phase mismatch. [4]

DBR Lasers A DBR Laser can be formed

by replacing one or both of the discrete
laser mirrors with a passive grating reflec-
tor. Figure 18 shows a schematic of such
a laser with one grating mirror. Besides
the single frequency property provided by
the frequency-selective grating mirrors, this
laser can include wide tunability. Since the
refractive index depends on the carrier den-
sity this can be exploited to vary the refrac-
tive index electro optically on the sections
by separate electrodes. Figure 18: Schematic of a DBR Laser [4]
The potential tunability of DBR Lasers
is one of the main reasons why they are of great importance. As indicated in Figure
18, there are usually three sections, one active, one passive, and the passive grating.
The first provides the gain, the second allows independent mode phase control and the
grating is a mode selective filter. By applying a current or voltage to the sections the
refractive index changes, shifting the axial modes of the cavity. [4, 1]

3.2.2 Distributed Feedback (DFB)

A distributed feedback laser (DFB) also uses grating mirrors, but the grating is included
in the gain region. Reflections from the ends are suppressed by antireflection coatings.
Thus, it is possible to make a laser from a single grating, although it is desirable to
18 3 Technology of laser diodes

have at least a fraction of a wavelength shift near the center to facilitate lasing at the
Bragg frequency. The idea behind this concept is, apart from the wavelength selectivity,
to improve the quality of the laser, as the active length is a quarter-wavelength long.
This applies for no shift in the gratings, where the cavity can be taken to be anywhere
within the DFB, since all periods look the same.
For the standard DFB grating, we can see that the
laser is antiresonant at the Bragg frequency. The modes
of this laser are placed symmetrically around the Bragg
frequency. However only the modes with lowest losses
will lase. With symmetrical gain profile around the
Bragg frequency, this means that two modes are reso-
nant.
To suppress one mode we need to apply additional
perturbation reflections, such as from uncoated cleaves
at the end. An alternative method is to introducing an
extra quarter wavelength element in the grating. This
quarter wave shifted grating is shown in figure 19. This
however is more difficult to fabricate. [12, 4]
Figure 19: Standard and quarter-
wave shifted DFB Lasers [4]
3.2.3 DFB vs. DBR Lasers
DFB Lasers are easier to fabricate and show less losses
and therefore have a lower threshold current. The DBR
is widely tunable, but relatively complex since a lot of structure must be created along
the surface of the wafer. for this reason DBR Lasers are only formed when their prop-
erties are required. Both lasers however work in single mode. [4]

3.3 Laser arrays

Semiconductor laser structures described so far have
been developed for low power applications, such as
providing a source in em–wave communication system.
Their limitation in the output power mainly arises form
the leakage current, which increases with an increase in
the applied current. [11]
The major drawbacks of semiconductor lasers, the
relatively low output powers, and the high beam di-
Figure 20: laser diode array [1] vergence can be lowered by integrating multiple laser
stripes into an array as shown in Figure 20. The stripes
are closely spaced so that the radiation from neighboring stripes is coupled to form a
coherent mode of the entire array. These laser diodes operate in multimode, in contrary
to those described so far.
The array modes, often referred to as supermodes, are phaselocked combinations of
the individual stripe modes and characterized by the phase relationship between the
optical fields supported by adjacent stripes. [1]
3.4 VCSEL (Vertical Cavity Surface Emitting Lasers) 19

Figure 21: Laserarray with heat sink (by Laser Diode Array Inc.)

Laser arrays can be found in applications, where high beam powers and high beam
densities are demanded. It therefore suggests itself for pumping of solid state lasers,
because the wavelength range from 700 to 900 nm can be covered with appropiate
semiconductor materials. [13]
The one-dimensional arrays are also referred to as laserbars. Through stacking of
these bars one can build up two-dimensional arrays. With such devices powers up to a
few kW have already been achieved. [10]

3.4 VCSEL (Vertical Cavity Surface Emitting Lasers)

A vertical cavity surface emitting laser (VCSEL1 ) is a special-
ized laser diode with improved efficiency and increased data
speed.
Older laser diodes, called edge-emitting diodes, emit coher-
ent em–waves parallel to the boundaries between the semi-
conductor layers. The VCSEL emits its coherent em–waves
perpendicular to the boundaries between the layers.
The VCSEL has several advantages over edge-emitting
diodes. The VCSEL is cheaper to manufacture in quantity, is
Figure 22: VCSEL laser
easier to test, and is more efficient. In addition, the VCSEL
diode Simulation [14]
requires less electrical current to produce a given coherent
energy output. The VCSEL emits a narrow, more nearly cir-
cular beam than traditional edge emitters. And because of its short cavity length it
emits only one mode. [15]

1
The acronym VCSEL is pronounced ‘vixel’
20 3 Technology of laser diodes

There are many designs of VCSEL structure however,

they all have certain aspects in common. The cavity length
of VCSELs is very short typically 1-3 wavelengths of the
emitted em–wave. As a result, in a single pass of the cav-
ity, a photon has a small chance of triggering a stimulated
emission event at low carrier densities. Therefore, VC-
SELs require highly reflective mirrors to be efficient. In
edge-emitting lasers, the reflectivity of the facets is about
30%. For VCSELs, the reflectivity required for low thresh-
Figure 23: Layers of DBR
old currents is greater than 99.9%. Such a high reflectivity mirrors
can not be achieved by the use of metallic mirrors. VC-
SELs make use of Distributed Bragg Reflectors. (DBRs) as described on page 17. These
are formed by laying down alternating layers of semiconductor or dielectric materials
with a difference in refractive index. Semiconductor materials used for DBRs have a
small difference in refractive index therefore many periods are required. Since the DBR
layers also carry the current in the device, more layers increase the resistance of the
device therefore dissipation of heat may become a problem. [3]

(a) mesa etched (b) ion implanted

Figure 24: Different designs of VCSELs [11]

As well as reducing the dimensions of the cavity one can reduce the threshold current
of a VCSEL device by limiting the cross-sectional area in which gain occurs. There
are a number of different types of lasers, which generally differ in the way in which
high-reflectivity mirrors and current confinement is achieved.
One simple method is to etch a pillar down to the active layer. These are termed
mesa2 etched structure. Etched mesas are typically a few micrometer in diameter
which allows fabrication of a large number of lasers on a single substrate. The large
difference in refractive index between the air and device material also act to guide the
em–wave emitted. However, for practical use a suitable bonding scheme is required. A
polyamide can be used for filling up the region around the etched mesas, to allow a
practical bonding scheme. Another problem with this type of structure are the loss of
carriers due to surface recombination at the sidewalls and poor dissipation of heat from
the laser cavity.
2
Webster Definition for ‘mesa’: isolated hill having steeply sloping sides and a level top
3.4 VCSEL (Vertical Cavity Surface Emitting Lasers) 21

Another technique for current confinement is ion implantation.

By selectively implanting ions into a semiconductor a region of
high resistivity around a 10 µm diameter opening is produced.
This scheme reduces current spreading, producing a region of
high gain at the center of the opening where laser action takes
place.
Since the substrate is absorbing in these cases, em–waves are
emitted from the top. It can also be emitted from the bottom by
using a structure in which the substrate near the emitting region
Figure 25: Etched
has been etched away. [11, 3]
Mesa

(a) Etched Well VCSEL (b) Air Post VCSEL (c) Burried Regrowth VCSEL

Figure 26: Other possible designs of VCSELs [3]

22 3 Technology of laser diodes

Appendix
Materials, wavelenght and typical applications

GaN AlGaAs/GaAs InGaAsP/InP AlGaAsSb/GaSb

420 - 520 nm 780 - 870 nm 1300 - 1550 nm 1700 - 5000 nm

blue - green near infrared near infrared middle/far in-

frared

optical datastor- optical transims- optical transmis- research

age sion for short sion for large
ranges ranges

Abbrevitations
This table provides a list of frequently used abbrevitations in the subject of laser diodes.

VB Valance band
CB Conduction band
SCL space charge layer
BH Buried Heterostructure
CW Contiuous Wave
DBR Distributed Bragg Reflector
DFB Distributed Feedback Bragg
DHS Double Heterostructure
LD Laserdiode
MQW Multiple Quantum Well
QW Quantum Well
QC Quantum Cascade
QD Quantum Dot
VCSEL Vertical Cavity Surface Emitting Laser
List of Figures 23

List of Figures
1 p-n junction with/without bias [2] . . . . . . . . . . . . . . . . . . . . . 3
2 The carrier concentration in SCL under forward bias [2] . . . . . . . . . 4
3 Recombination [3] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4 Electron and hole distribution [4] . . . . . . . . . . . . . . . . . . . . . . 6
5 Recombination [5] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
6 Light feedback [3] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
7 Double heterojunction laser diode [9] . . . . . . . . . . . . . . . . . . . . 8
8 Refractive index of the waveguide and the field distribution [4] . . . . . 9
9 AlGaAs LD Spectrum [9] . . . . . . . . . . . . . . . . . . . . . . . . . . 11
10 Index guided laser diode [9] . . . . . . . . . . . . . . . . . . . . . . . . . 12
11 Beam profile from an index guided laser diode [9] . . . . . . . . . . . . . 12
12 Gain-guided laser [10] . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
13 Cross section [11] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
14 Index-guided laser [9] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
15 Schematic cross sections of buried-heterostructure lasers [11] . . . . . . . 15
16 Schematic Illustration of DFB and DBR Lasers [11] . . . . . . . . . . . 16
17 Schematic of a DBR mirror [4] . . . . . . . . . . . . . . . . . . . . . . . 17
18 Schematic of a DBR Laser [4] . . . . . . . . . . . . . . . . . . . . . . . . 17
19 Standard and quarter-wave shifted DFB Lasers [4] . . . . . . . . . . . . 18
20 laser diode array [1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
21 Laserarray with heat sink (by Laser Diode Array Inc.) . . . . . . . . . . 19
22 VCSEL laser diode Simulation [14] . . . . . . . . . . . . . . . . . . . . . 19
23 Layers of DBR mirrors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
24 Different designs of VCSELs [11] . . . . . . . . . . . . . . . . . . . . . . 20
25 Etched Mesa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
26 Other possible designs of VCSELs [3] . . . . . . . . . . . . . . . . . . . . 21
24 References

References
[1] Nagaatsu Ogasawara. Lasers, semiconductor. Technical report, Nagaatsu Oga-
sawara University of Electro-Communications, www.pro-physik.de/Phy/pdfs/
OE042_1.pdf.

[2] S.O.Kasap. Optoelectronics. Prentice Hall, 1 edition, 1999.

[3] Carl Hepburn. Britneys guide to semiconductor physics. http://www.

britneyspears.ac/lasers.htm.

[4] Scott W. Corzine Larry A. Coldren. Diode Lasers and Photonic Integrated Circuits.
John Wiley & Sons, 3 edition, 1995.

[5] Mark Fox. Optical Properties of solids. Oxford University Press, 2001.

[6] M. C. Teich B. E. A. Saleh. Fundamentals of Photonics. Wiley, New York, 1

edition, 1991.

[7] Shane Eaton. Distributed Feedback Semiconductor Lasers.

[8] DEPARTMENT OF PHYSICS and APPLIED PHYSICS. 3rd/4th year laboratory.

[9] Ammon Yariv. Quantum Electronics. John Wiley & Sons, 3 edition, 1989.

[10] Richard Scheps. Laser Diode-Pumped Solid State Lasers. SPIE Press, 2002.

[11] Niloy K. Dutta Govind P. Agrawal. Semiconductor Lasers. Van Nostrad Reinhold,
2 edition, 1993.

[12] K. J. Ebeling. Integrierte Optoelektronik. Springer Verlag, 1989.

[13] Dirk Jansen. Optoelektronik. Vieweg Verlag, 1993.

[14] University of Stuttgart. http://www.physik.uni-stuttgart.de/ExPhys/4.

Phys.Inst./Forschung/VCSEL/vcsel.html.

[15] http://whatis.techtarget.com/. vertical cavity surface emitting laser. http://

whatis.techtarget.com/definition/0,,sid9_gci803517,00.html.