quantum structures of iii-v compound semiconductorssangho10/publications/11pn-bma.pdf · structures...
TRANSCRIPT
11
B. M. Arora received his B Tech degree in Electronics & Communication Engineering
from the Indian Institute of Technology, Kharagpur in 1965 and PhD degree in
Electrical Engineering from the University of Illinois, Urbana, in 1972. His early work
was in the area of plasma display panel. Subsequently, he studied photoluminescence
of CdSe and CdS. In 1972, he joined the Tata Institute of Fundamental Research (TIFR),
Mumbai where he worked on the growth of III-V compounds by liquid phase epitaxy,
anodic oxidation and passivation of GaAs, Schottky barrier junction properties and
defects in III-V compounds. Later at TIFR Professor Arora worked extensively on the
growth of III-V strained layer heterostructures using metal-organic vapor phase
epitaxy and their application to optoelectronic devices such as lasers and quantum well infrared
photodetectors. Professor Arora is currently with the Electrical Engineering Department of Indian Institute
of Technology, Bombay. (email: [email protected])
of size, the basic premise is very different. Reduction of size in MOS structures
is based on scaling [1, 2], while quantum structures are based on behaviour of
free electron confined to dimension smaller than the DeBroglie wavelength
(typical value for electrons in GaAs is ~50 nm), which is visualized as “particle
in a box” and referred to as quantum size effect or quantum confinement [3].
The three quantum structures: quantum wells (QWs), quantum wires (QWIs),
and quantum dots (QDs) are shown schematically in Fig.1, depending on
confinement of electrons in 1, 2 or 3 dimensions respectively. The dimension in
the confinement direction is typically 10 nm or smaller .Apart from these, two
other quantum structures in common use are i) multi-quantum wells (MQWs)
and ii) superlattices (SLs) as shown in Fig.1. As seen in the figure, the quantum
structures are composed of multiple semiconductors of different band gaps
in contact. Together they comprise a family called heterostructures.
Introduction
Quantum structures have been at the
forefront of major developments in
the science and technology of
semiconductors for nearly four
decades. Together with metal-oxide-
semiconductor (MOS) structure,
they form one of the two branches of
semiconductor science which led the
way to modern nano-science and
nano-technology. Although both
MOS and quantum structures have a
common feature, namely reduction
QUANTUM STRUCTURES OF III-V COMPOUND
SEMICONDUCTORSB. M. Arora
Sandip Ghosh
Department of Electrical Engineering,
Indian Institute of Technology-Bombay, Powai, Mumbai 400076
Department of Condensed Matter Physics and Material Science
Tata Institute of Fundamental Research, Homi Bhabha Road, Mumbai 400005
Sandip Ghosh received his B.Sc. (Hons.) degree in physics from the University of Delhi
in 1990 and the M.Sc. degree in physics from the Indian Institute of Technology, Delhi,
in 1992 and Ph.D. from the Tata Institute of Fundamental Research (TIFR), Mumbai in
1998. After holding post doctoral positions at the University of Surrey, Guildford and
the Paul Drude Institute, Berlin, he joined the faculty at TIFR in 2002. His research
interests involve the study of electronic band structure related physics of
semiconductors and optoelectroinc devices, using optical spectroscopic techniques.
(email: [email protected])
Bulletin of Indian Physics Association Vol. 41 Issue 1 January 2011
Quantum structures generally
confine electrons in the smaller band
gap semiconductor with the larger
band gap semiconductor acting as
barrier.
Historically, advances in quantum
s t r u c t u r e s h a ve h a d s t r o n g
relationship with III-V compound
semiconductors since near ideal
heterostructures can be synthesized
from them. III-V semiconductors,
with the most well known member
GaAs, have been known since the
pioneering work of Welker [4]. They
are formed by combining elements
from column III (B,Al,Ga,In) and
column V (N,P, As ,Sb) of the periodic
table. They may be binary (GaAs, InP
etc), ternary (AlGaAs, GaAsP etc), or
quaternary (AlGaInAs, InGaAsP etc)
materials [5]. Quantum structures
are formed generally from single
crystalline materials. Most of the III-
V compounds grow in cubic zinc-
blende crystal structure (Fig.2). Over
the last decade III-nitrides (AlN,
GaN and InN ) have been extensively
investigated [6]. Nitrides have
predominantly hexagonal close
packed wurzite crystal structure
(Fig.2).
An impetus for quantum structures
came after the invention of GaAs
diode laser in 1962 in order to make it
a practical device operating at low
current density. The first crucial step
was the introduction of hetero-
structure GaAs-AlGaAs which forms
the basis of many subsequent
discoveries and inventions including
quantum well lasers [7], quantum
well light modulators [8], quantum
cascade lasers [9], quantum well
infra-red detectors [10], resonant
tunnel structures [11], high electron
mobility transistors [12], and integral
12
and fractional quantum Hall effects [13]. Kroemer's early theoretical work
[14] had paved the way for subsequent developments of hetero-structures.
The next advance came from the work of i) Esaki and colleagues on electron
transport in superlattices [15], and of ii) Dingle and coworkers on optical
properties of quantum wells [16]. Apart from these composition modulated
structures, doping modulated structures nipi were also investigated [17].
Fig.1 : Semiconductor quantum structures. Electrons can tunnel across the layers in a
superlattice. In a MQW structure, they are confined to the smaller band gap. Also
shown are the density-of-states (DOS) for the electrons.
It was realized early that experiments on quantum wells and superlattices
required very high degree of crystal perfection of multiple thin crystalline
layers (known as epitaxial layers) as well as perfection of interfaces between
the epitaxial layers. All early work on heterostructures was done by growing
layers using liquid phase epitaxy (LPE) [18]. New techniques i) molecular
beam epitaxy (MBE) [19, 20] and ii) metal-organic vapour phase epitaxy
(MOVPE) [21] were invented for synthesis of very thin layers with near perfect
/ abrupt hetero-interfaces. Most of the above advances have come by using
arsenides and phosphides of Al, Ga and In. For example, communication
diode lasers use InGaAs for the light emitting QW layer and InGaAsP barrier
layers to form the waveguide, all grown on InP substrates.
More recently, great advances have been made in the use of nitrides of Al, Ga
and In . Apart from giving us sources and detectors in the UV, blue and green
regions, nitrides have provided us a vision of solid state lighting [22], very
high frequency compact electronics [23], and high temperature electronics
[24]. In this brief review, we discuss salient properties of heterostructures and
quantum structures, methods of synthesis, electronic structure and optical
properties, and conclude with a few applications.
A combination of two semiconductors of different band gaps forms a hetero-
junction. The band diagram of a hetero-junction has discontinuities at the
junction as shown in Fig.3. The discontinuity between the conduction band
Hetero-structures and Quantum structures of III-V compounds
13
edges is called the conduction band offset and corresponding discontinuity
between the valence band edges is the valence band offset. Three different
types of discontinuities are observable in different combinations of
semiconductors as shown in Fig.3 [25]. Although there are examples of all the
three types of heterostructures, in this review we dwell mainly on the type-I
heterostructures, which are the ones most widely used. Fig.4 shows band
diagram of a type I QW. As seen, transition from the small band gap to the
larger band gap is abrupt, making it a square potential QW. The synthesis of
hetero-structures has been refined such that the transition region can be
designed to create potential well of a desired shape such as parabolic or
triangular well.
E (k) = k /2m * + E (y,z) ....(2)
E = E (x,y,z) ....(3)
e x e em,n
e el,m,n
2 2
In a QD, the electron energy is
discrete in all three dimensions and
given as
In these equations, the localization
energy of the electrons is E (z) in a
QW, E (y,z) in a QWI, and E
(x,y,z) in a QD; l,m,n have integral
values. Similar relations hold for
localization energy of holes in the
valence band of quantum structures.
Localization energy is obtained by
solving the Schrodinger equation of
motion of electron/hole in the finite
potential well. [26].
Growth of high quality single crystal
multiple layers of different materials,
with different chemical compositions
and doping, different thicknesses of
layers with sub-nanometer control
and near perfect and abrupt
interfaces, imposes very stringent
requirement on the choice of
materials and method of synthesis
[27]. Molecular beam epitaxy (MBE)
and organometallic vapour phase
epitaxy (MOVPE) are widely used for
the synthesis of quantum structures.
Schematic diagrams of these
apparatus are shown in the Fig.5.
The MBE apparatus [20] is an ultra
high vacuum (UHV 10 Torr)
chamber with ovens containing
elemental Al, Ga, In, As (for growing
arsenides), Si (n type dopant) and Be (
p type dopant). Atomic/molecular
beams evaporate from the ovens kept
at elevated temperatures. The beams
reach the heated substrate where they
react and form the compounds which
grow epitaxially. The ovens are fitted
with shutters to prevent unwanted
beam from reaching the substrate.
en
em,n el,m,n
��10
�
Choice of Materials and
S y n t h e s i s o f Q u a n t u m
Structures
Fig.2 Zinc Blende and Wurzite crystal structures
Fig.3 The three types of heterostructure band lineups, Type-I, Type-II and Type-III.
In a QW, electrons are confined in one
dimension (along ), while they are
free in the other 2 dimensions ( and
). Such a system of electrons is called
2 dimensional electron gas (2DEG).
The energy of electrons in a QW of
width L is given by
E (k) = k /2m * + k /2m * + E (z)
.....(1a)
where
E (z) = n /2L m * .....
In a QWI electrons are free to move in
1 direction (say ) and confined in the
other two dimensions ( and ). For a
QWI, the electron energy is given as
z
x
y
x
y z
w
e x e y e en
en w e
2 2 2 2
22 2 2� (1b)
� �
�
Fig.4 A type-I GaAs/Al Ga As Quantum
Well (well width = L ) potential profile.
x 1-x
w
15
the area of optoelectronic devices such
as quantum dot lasers and quantum
dot infrared photodetectors [37].
materials) because of disorder in the random placement of components in the
sublattices. Ternary and quaternary materials are alloys and even though
crystalline, they have alloy disorder. For example, in Al Ga As, the number of
Al/Ga atoms will fluctuate from one cubic cell to another, even though globally
the mean composition is fixed. The alloy disorder effect is particularly
prominent in determining carrier transport properties at low temperatures in
quantum well structures. For example, electron mobility in GaAs QW can rise
to 10 cm /(V-s) or even higher at liquid He temperature [28]. On the other
hand, the corresponding electron mobility in an InGaAs QW does not exceed
10 cm /V-s because of alloy disorder in the In/Ga sublattice [30].
0.5 0.5
7 2
5 2
Fig.6 Effect of mismatch between lattice parameters of the substrate and the epi-
layer
The lattice matching requirement for near perfect growth is relaxed in the case
of very thin layers such as QWs. The difference in lattice mismatch is
accommodated by the biaxial compression (or tension) in the plane of the QW,
with concomitant extension (or shrinkage) of lattice parameter normal to the
plane of the QW [31] as shown in Fig.6. If the thickness of the epi layer exceeds
a critical limit, the layer relaxes with production of dislocations. The ability to
grow mismatched defect free layers has greatly enhanced the choice of
materials that can be used in devices. As is to be expected, built-in strain in
these layers modifies the electronic band structure in the layer. This turns out
to be highly beneficial in many cases and plays very important role in modern
devices which employ strain engineering [32]. Strained QW laser diodes have
lower lasing threshold current [33]. Strained Si-Ge layers are used to enhance
the electron mobility [34]. The role of strain in the synthesis of quantum dots is
even more extraordinary. Large lattice mismatch (~5%) between the substrate
and the layer (a > a ), reduces the critical thickness of defect free planar layer
to one or two atomic monolayers ( also called wetting layer), such that further
growth results in self assembled growth of nanometer dimension islands ,
which are essentially defect free QDs [35, 36]. It is even possible to build
multiple layers of QDs in this manner. Substantial advance has been made in
L S
Fig.7 Schematic of the electronic band
structure of bulk GaAs in terms of
e l e c t ro n e n e rg y ve rs u s c r ysta l
momentum ( ) around the band gap Eg.k
Spectroscopy of Quantum
Structures
As discussed above, the reduction in
the physical size of a semiconductor
crystal and the formation of
heterostructures, lead to significant
changes in the electronic band
structure (EBS). In order to tailor the
EBS to ones' needs for designing
better electronic devices, it is
important to be able to measure the
changes in the EBS. Optical
spectroscopy is one of the most
powerful tools for studying the EBS
and in this section we will describe a
few of the optical techniques used in
the study of semiconductor quantum
structures. First we note that the
o u t e r m o s t e l e c t r o n i c s h e l l
configuration of group III and group
V elements consist of and type
atomic orbitals (eg. in GaAs, Ga:4s ,
4p and As: 4s , 4p ). When the atoms
come close together to form a
compound III-V semiconductor
crystal, these orbitals undergo
s p2
1 2 3
16
MOVPE on a GaAs substrate, which
has several GaAs/Al Ga As QWs
of different thicknesses. We see that
although the material of the well
(GaAs) is the same throughout, the
change in the thickness has led to
different effective band gaps for the
wells, as identified by the different
peak PL emission energies. The GaAs
related emission is from the thick
GaAs cap layer on top of the sample.
Light emission through the edges of a
QW sample (perpendicular to )
shows addit ional interest ing
features.
Fig.8 b) shows that such electro-
luminescence emission from a
GaInP/AlGaInP QW is strongly
polarized with the transverse electric
(TE || ) component peaking at a
longer wavelength when compared
to the transverse magnetic (TM || )
component, the latter being weaker in
strength [38]. This can be understood
as follows. The confinement energies
are different for the hh and the lh
bands due to the different effective
masses of the holes in these bands.
This lifts the hh-lh degeneracy and
the lowest energy hh band now has
more holes resulting in a stronger e1-
hh1 emission at longer wavelengths
(lower energy). Also transition from
an like CB state to a like VB state
involves polarized light and so on.
The hh state around k , k = 0 is purely
composed of and orbitals and
therefore this emission is polarized.
The lh state has a small component
and therefore the e1-lh1 transition is
seen in the TM polarized emission at
shorter wavelengths (higher energy).
While PL spectroscopy is useful for
estimating the effective band gap
change, it does not reveal any
information about confined states of
the quantum structure lying at
0.27 0.73
z
x
z
x
x
s p
p p
p
x
x y
z
x y
hybridization and result in the formation of a single conduction (CB) and
three closely spaced valence bands (VB). Fig.7 shows a schematic of the EBS of
bulk GaAs in terms of electron energy versus crystal momentum ( ). Here one
is concerned with only the CB and VBs, that too near =0 (Brillouin zone
center) while neglecting filled bands at lower energies and empty ones at
higher energies because they typically do not directly affect device
performance. The curvature of the bands in space defines the effective mass
and accordingly the top two valence bands, which are degenerate (in cubic
crystals) at =0, are labeled as the heavy hole (hh), light hole (lh) and the third
as the spin orbit split off (so) hole band. The character of the cell periodic parts
of the wavefunction in these bands, which are like for CB and , or like
for the VBs, decide how these materials respond to polarized light. This topic
will be discussed later. Fig.4 showed the potential energy diagram of a GaAs
quantum well where one had plotted the energy of the band extrema at =0 as
a function of distance along , the film growth direction. We see that the lowest
energy state that the electron can occupy in the QW is not the CB bottom but an
energy higher by E (z). Similarly the hh occupies an energy higher by E (z)
(hole energies increase downwards). Thus the effective band gap of this
structure becomes Eg = Eg + E (z) + E (z). Referring to equation (1b) we
see that effective band gap increases as the width of well L reduces.
k
k
k
k
k
s p p px y z
z
e1 hh1
eff GaAs e1 hh1
w
Fig.8 a) Photo-luminescence(PL) spectrum of a sample which has several
GaAs/Al Ga As QWs of different well widths Lw. The QW emissions are all e1hh1
transitions. b) Polarization resolved edge emission spectrum from a GaInP/AlGaInP QW.
0.27 0.73
Photo-luminescence (PL) spectroscopy is a technique commonly used to
estimate such effective band gap change. Here one excites the sample with an
intense pump laser beam with photon energy much higher than the band gap
to be measured. The absorbed pump beam excites electrons to the CB leaving
behind a hole in the VB. Within a short time (typically pico-second) these
excited electrons (holes) loose energy and trickle down to the CB bottom (VB
top) and thereafter recombine radiatively. The photon energy of the resultant
emission will then correspond to the effective band gap. Thus by measuring
the emission spectrum, typically using a reflection-grating based
monchromator to disperse the emitted light and then detect it with an
appropriate photo-detector, one can estimate the effective band gap change.
Fig.8 a) shows the PL spectrum of a sample grown in our laboratory using
17
higher energies. One way to obtain such information is through photo-
luminescence excitation (PLE) spectroscopy, where in one detects the PL
emission at a particular wavelength while varying the wavelength of the
pump excitation. The working principle is that more the absorption, more
electron-hole pairs would be generated and hence more will be the
luminiscence output. Fig.9 shows the PLE spectrum of a 4.6nm thick
GaAs/Al Ga As single QW, which essentially mimics its absorption
spectrum. It consists of some sharp peaks riding on a staircase like
background. The latter is the signature of the step-function 2D density of
states previously shown in Fig.1. Each step indicates the onset of transition
between a CB and VB confined energy state in the QW. The sharp peaks just at
the edge of the steps are due to resonant absorption of photons that lead to
exciton formation. Excitons can be visualized as electron-hole pairs bound
together by coulomb attraction, just like the electron-proton in a hydrogen
atom. In 3D, the exciton binding energy for GaAs is about 4 meV and is
therefore easily ionized at room temperature (k T=26meV at 300K). However
an interesting aspect of 2D systems is that when the electron-hole pair is
constrained to move in 2D then the solution of the Schrodinger equation
shows that the binding energy becomes four times that in case of 3D motion.
In practice one gets a value close to 10 meV for exciton binding energy in a 5
nm GaAs QW. This allows the excitons to exist up to much higher
temperatures and they show up prominently in the absorption spectrum of a
QW. We also notice two closely placed transitions close to the band edge.
These arise because the hh-lh degeneracy is lifted in a QW as explained before.
0.36 0.64
B
second intense pump laser beam
incident on the sample. The pump
beam has photon energy higher than
the band gap and it generates
electron-hole pairs which modify the
built-in electric fields at the interfaces
of the QW. This leads to an electric
field modulation of the dielectric
function and consequently the
refractive indices change. In ER the
electric field is applied directly using
appropriate electrode arrangements
[39]. The dielectric function depends
on the electronic joint-density-of-
states. Since the joint-density-of-state
changes abruptly at a critical point
(van-Hove singularity) in the EBS,
such as at the onset of a new inter-
band transition, a perturbation like
the electric field affects the optical
response more around a critical
point. Thus the change in reflectance
shows sharp spectral features at
energies where a new inter-band
transition arises such as between the
higher confined states in a QW. The
difference nature of modulation
spectroscopy helps it pick up small
changes in reflectance over a large
background and from the resulting
spectral features one can determine
the critical point energies. Fig.10
shows the PR spectrum of another
GaAs/Al Ga As QW where one
finds sharp spectral features which
one can identify as arising due to the
GaAs (cap), Al Ga As barrier and
four transitions in the QW. The e1-
hh3 feature is weak because it is a
normally forbidden transition and
the e2-hh2 transition is weak because
the e2 level is close to the conduction
band edge and is therefore weakly
confined.
An absorption spectrum directly
gives information on higher energy
states. However, direct measurement
0.36 0.64
0.36 0.64
Energy (eV)1.6 1.7 1.8 1.9 2.0 2.1
Sig
na
l (a
rb.
un
it)
0
100
200
300
400
GaAs/Al 0.36Ga0.64As SQW
Lw~46A, 8K
Al0.36Ga0.64Ase1hh1
e1lh1
e1hh3
e2hh2
�D
PLE
PL
Fig.9 Photo-luminescence excitation (PLE) spectrum of a single GaAs/Al Ga As
QW of width Lw ~4.6 nm measured at 8K. The PL spectrum is also shown. The PLE
detection was at 1.62eV ( =765nm).
0.36 0.64
D�
PLE has a drawback in that it typically requires sample cooling to low
temperatures for the PL signal to be detected. However it is also possible to
identify higher energy transitions in quantum structures like single QWs
using the techniques of modulation spectroscopy such as Photo-reflectance
(PR) and Electro-reflectance (ER). In PR one measures the change in the
reflectance of a probe beam of variable wavelength due to influence of a
18
plane of the QW. The strain mixes the
VBs and modifies the relative
contributions of and to the hh
and lh bands giving rise to this
anisotropy. The anisotropic strain in
this case was specially engineered,
such structures are of interest for
polarization sensitive detection
applications.
As the expertise of synthesizing
good quality III-nitrides is
growing, there is continuous
refinement of the measurements of
their basic properties. Among the
heterostructure properties, the most
significant are the conduction band
discontinuity and the valence band
discontinuity at the heterojunction. It
turns out that determination of these
band discontinuities has presented
difficulties of reliability even for
the most well known heterojunction
AlGaAs-GaAs [41]. We chose to
measure the band discontinuities
b e t w e e n I n N a n d G a N
heterojunction by internal photo-
emission technique [43] using
junction of heavily doped n type InN
and lightly doped n type GaN. We
measure short circuit photo current
of this heterostructure, while shining
monochromatic light (through GaN
side) over a broad range of photon
energies 1.5 to 3.5 eV. Results of these
measurements are shown in
p px y
of the absorption coefficient through optical transmission experiments is
often difficult in case of quantum structures because the small dimensions
(for example the width of a single QW) result in barely measurable light
attenuation. To estimate the absorption spectrum, indirect methods are
adopted such as PLE discussed above as also Photo-Voltage (PV) and Photo-
Current (PC) spectroscopy. In PV measurements the quantum structure
typically has a built-in field across it, such as by virtue of being placed in the
i-region of a p-i-n diode structure. In this case, light absorbed by the
Fig.10 Photo-reflectance (PR) spectrum of a GaAs/Al Ga As single QW with well
width Lw ~4.6 nm at room temperature.
0.36 0.64
structure creates electron-hole pairs which get separated by the field and in
the process modifies the built in potential which is detected as a voltage
signal, much like in the case of a solar-cell. Fig.11 shows our measured PV
spectrum of MBE grown InAs QDs. These pancake shaped QDs with average
height ~9 nm and diameter ~40 nm are in a GaAs matrix. The PV spectrum
mimics the absorption spectrum of the QDs. Instead of very narrow atom like
transitions expected for a 0D system, we find that the spectrum can be
resolved into broadened Gaussians which actually represent inhomogenous
broadening of the spectrum due to a distribution in the size of the QDs (about
4% of the mean here) [40]. However we do see the dramatic effect of 3D
confinement in terms of the shift of the band gap from 0.35 eV for bulk InAs to
about 1.03 eV for these InAs QDs. It is also possible to estimate the absorption
spectrum by applying a bias to a quantum structure and then shining light on
it to measure the photo-current as a function of incident photon-energy. Fig.12
shows the lateral PC spectra of GaAs/Al Ga As QWs where one has ensured
that the conduction happens along the plane of the QW. The two spectra are
for two orthogonal polarizations of the light incident normally on the QW
plane (x-y plane). We again see spectral features similar to those seen in the
PLE spectrum, however one also notices that the hh and lh related exciton
transitions have different strength for the two polarizations. This is usually
not expected in a QW since both hh and lh bands have identical and
contributions. In this case our analysis showed [41] that the in-plane
polarization-anisotropy arises due to the presence of anisotropic strain in the
x 1-x
p px y
Energy (eV)1.0 1.1 1.2 1.3 1.4
PV
sig
nal (a
rb.
un
i ts)
0
1
2
3
4
5
6
p=1p=2
p=3
p=4
WL
x1/20
x1/300
GaAs
QD
Fig.11 Photo-voltage spectrum of InAs
quantum dots embedded in GaAs,
measured at room temperature.
19
phase epitaxy [44]. Fig.14 b) shows
crossectional micrograph of a typical
device fabricated by LPE [45].
Devices incorporating QWs and
other quantum structures are
fabricated by using MBE or MOVPE.
Diode lasers used for optical
communication are single mode
devices, integrated and tunable [46,
47]. On the other hand, diode lasers
have also been combined to make
arrays for delivering kilowatts of
power for industrial applications
[48]. Fig.15 shows characteristics of a
high power 0.98 micron laser device
fabricated at TIFR.
Inter-subband transitions provide a
mechanism for generating and
d e t e c t i n g i n f r a r e d r a d i a t i o n
i n vo l v i n g o n l y e l e c t r o n s i n
conduction band or holes in the
valence band. The emission or
absorption wavelength of radiation
i s d e t e r m i n e d b y q u a n t u m
confinement and is independent of
the fundamental band gap. The light
emitting device called quantum
cascade laser may have 20 to 100
periods of coupled quantum wells
and is perhaps the most intricate
quantum device [9]. Detectors based
on quantum wells are relatively
s i m p l e r a n d a r e e s s e n t i a l l y
photoconductors as shown in the
Fig.16. Inter subband transitions in
quantum wells require, apart from
e n e r g y a n d m o m e n t u m
conservation, that the polarization of
the radiation should be in direction
normal to the plane of the QWs.
Coupling of light to electrons in the
QW is achieved in a single detector
device by using the angle lapped
configuration as shown in Fig.16 [49].
In imaging devices, it is desirable that
2. Inter-subband transition
devices
Fig.13 Internal-photo emission spectrum of an InN/GaN hetero-junction. The
adjacent schematic shows the inferred band alignment at the heterojunction.
Fig.13.Three prominent features can be seen in this figure marked Region I,
Region II, and Region III. Region III is sharp peak arising from excitations of
electrons across the GaN band gap. Region II is caused by excitation of
electrons from the valence band maximum of InN to conduction band edge of
GaN. Region I is caused by excitation of electrons from the conduction band
(level marked E ) of InN to the bottom of conduction band of GaN. From these
threshold energies, we obtain the valence band discontinuity of InN/GaN
heterojunction to be ~0.85 eV and the corresponding conduction band
discontinuity to be ~1.82 eV as shown in the adjacent schematic band diagram.
Quantum structures form the basis of a very large number of optoelectronic
and electronic devices. Perhaps the most widely used device is the
semiconductor laser diode. They are small in size, have high gain and are very
reliable. Fig.14 shows the progressive reduction in the threshold current
density for lasing by using quantum structures for the active layer, because of
the higher gain. Higher gain of diode lasers is a result of modification of the
density of states.All laser structures till 1970 were synthesized by using liquid
F
Selected Applications
1. Diode Lasers
Fig.12 Polarized lateral photo-current (PC) spectrum of a GaAs/Al Ga As QW based
device structure (inset) where the QW is under anisotropic strain in the x-y plane.
x 1-x
Fig.17 Quantum Hall Effect measurement results. The figure shows the sample structure
with the InGaAs/InP QW, the potential profile, a photograph of the sample and results of
Hall effect measurements showing the QHE plateaus at various temperatures [52].
20
light is imaged on the plane of the
QWs. To achieve coupling to
electrons, the surface is formed into a
grating. Infrared detectors are also
made utilizing inter subband
transitions in quantum dots (QDIPs).
Coupling to electrons in QDs does
n o t h a v e t h e p o l a r i z a t i o n
dependence required by the QWs.
Infrared imaging cameras based on
QWIPs are used in astronomy (extra
terrestrial bodies emitting in
i n f r a r e d ) , m e d i c i n e ( c a n c e r
detection), defence (night vision),
and for industrial applications where
different parts of objects may be at
different temperatures.
A large concentration of high
mobility electrons can be generated
i n a q u a n t u m we l l t h r o u g h
modulation doping, where the
doped layer in the larger band gap
semiconductor is separated from the
QW and donates electrons to the QW.
A 2DEG structure of this type using
undoped InGaAs QW and n doped
InP as the donor layer is shown in
Fig .17 . The most prominent
electronic device based on 2DEG is
the high electron mobility transistor
(HEMT). Transistors fabricated from
structures similar to that in Fig.17
show excellent characteristics with
high gain and low noise upto
100GHz in frequency[50]. Further
improvements have been achieved
by using metamorphic HEMT, where
the substrate is GaAs, the QW is
InGaAs and the donor layer is
I n A l A s [ 5 1 ] . M o r e r e c e n t
developments of HEMTs based on
III- Nitrides [23] give advantages of
reduced size and higher power at
high frequencies because of electron
densities approaching 10 /cm .13 2
3. Two Dimensional Electron
Gas (2DEG) Devices
Fig.14 a) Progressive reduction in the threshold current density for lasing over the
years. b) Crossectional micrograph of an LPE grown QW laser
Fig.15 Structure and characteristics of a high power QW laser device fabricated at TIFR.
Fig.16 Structure and characteristics of a QW infrared photodetector fabricated at TIFR [49]
21
References[1] R. Noyce, ScientificAmerican ,
63 (1977)
[2] D. L. Critchlow, Proc. IEEE , 6 5 9
(1999)
[3] G. Bastard, Wave Mechanics
a p p l i e d t o S e m i c o n d u c t o r
H e t e r o s t r u c t u r e s , W i l e y -
interscience (1991)
[4] H. Welker, Z Naturforsch , 7 4 4
( 1 9 5 2 ) ; H . We l k e r a n d H .
Weiss, Solid State Physics, Vol.
3,Academic Press (1956)
[5] O. Madelung, Physics of III-V
Compounds, Wiley (1964)
[6] Gallium Nitride Vol I and Vol II
i n S e m i c o n d u c t o r s a n d
S e m i m e t a l s , ( E d s . T . D .
Moustakas and J. I. Pankove)
Academic Press (1998) and
(1999)
[7] C. H. Henry in Quantum Well
L a s e r s ( E d . P. S . Z o r y )
Academic Press (1993)
[8] C. Weisbuch and B. Vinter,
Q u a n t u m S e m i c o n d u c t o r
Structures, Academic Press
(1991)
[9] J. Faist, F. Capasso, D. L. Sivco,
C. Sirtori,A. L. Hutchison and A. Y.
C h o , S c i e n c e , 5 5 3
(1994)
[10] H. Schneider and H. C. Liu,
Q u a n t u m w e l l I n f r a r e d
P h o t o d e t e c t o r s , S p r i n g e r
(2006)
[11] B. R. Nag, Physics of Quantum
W e l l D e v i c e s , K l u w e r
Academic (2000)
[12] W. Liu, Fundamentals of III-V
Devices, Wiley (1999)
[13] R. E. Prange, S M Girvin, Eds., T h e
Q u a n t u m H a l l E f f e c t ,
Springer-Verlag (1987)
[14] H. Kroemer, Proc IRE 45, 1535
(1957)
[15] L. Esaki and R. Tsu, IBM J. Res.
Develop. ,61 (1970); L. Esaki
and L. L. Chang Phys. Rev. Lett.
, 495 (1974)
[16] R. Dingle, W. Wiegemann, and
C. H. Henry, Phys. Rev. Lett. ,
827 (1974)
2 3
87
7a
2 6 4
14
33
33
Application of a suitable potential to
the gate electrodes essentially isolates
the region contained by them,
mimicking a QD structure. This has
been widely used in single electron
transistors to investigate coulomb
blockade and tunnel current transport
[55]. With the advent of the subject
of quantum information processing
and quantum computing, there has
been increasing expectation that
semiconductor quantum structures
will provide the scalable basic logic
units called qubits. Connected to these
developments is the use of two level
electron spin system in QD as a qubit.
Electrical and optical techniques have
been used to prepare and read out the
individual spin polarized state as well
as to control interaction between
neighbouring spin states [56, 57]. There
are also proposals for using the spin
resolved edge states of the Quantum
Hall effect in the QH insulating regime
for qubits [58]. Before closing, it may be
mentioned that quantum wires have
been relatively less investigated
because of synthesis difficulties.
However, much progress has occurred
in the synthesis of nanowires [59] and
further progress may be expected.
The authors are grateful to many
colleagues who have contributed to
the development of the work at TIFR.
We would like to acknowledge K. L.
Narasimhan, A. Bhattacharya, S. S.
Chandvankar, T. K. Sharma (RRCAT
Indore), S. Ganguly (RRCAT Indore), S
Sen (Calcutta Univ.), G. Dohler
(Erlangen Univ.), S. Malzer (Erlangen
Univ.), K. S. Chandrasekharan, M. K.
Sanyal (SINP, Kolkata), S. Datta, B.
Karmakar, A. Arora, J. Bhattacharya,
M. R. Gokhale, A. P. Shah, S.
Subramanian, A. K. Srivastava, A.
Majumdar and Z. H. Mahmood. BMA
was with TIFR till 2008.
Acknowledgements
2DEG electrons are also the source of
integral and fractional Quantum Hall
Effects (QHE). A measurement of the
integral QHE based on the sample of
Fig.17 is shown with several plateau to
plateau (PP) transitions. The mobility
of 2DEG in this sample is limited to ~
10 cm /V-s because the QW is an alloy.
The temperature dependence of PP
transitions has been used to
investigate their scaling behaviour
[52]. The most well known use of QHE
is the resistance standard as the
resistance in the each plateau region is
constant and assumes value h/ne ,
depending only on fundamental
quantities: h the Planck constant and e
the electron charge, with n an integer.
[53].
It is apparent that quantum hetero-
structures have become an essential
part of semiconductor science and
technology. Most of the developments
are based on quantum wells as
outlined above, because of the relative
ease of synthesis of these structures.
Investigations and applications of
quantum dots have picked up over the
years. Researchers have used two
routes to produce QD structures. One
of these is the self assembly route and
has been used for optoelectronic
applications as mentioned above. Self
assembled QDs have dispersion of
sizes which broadens the energy level
structure. In addition their placement
is by and large random. While these
do not adverse ly af fec t the
optoelectronic devices using ensemble
of dots [54], single or coupled dot
structures require tighter control over
these properties. For these reasons, an
a l t e r n a t i v e r o u t e i s u s e d
predominantly in single dot and
coupled dot related work by
surrounding a small region of a
QWwith gate electrodes defined by
e l e c t r o n b e a m l i t h o g r a p h y.
5 2
2
Prospects
22
[17] G. H. Dohler in Properties of III-V
Q u a n t u m W e l l s a n d
S u p e r l a t t i c e s ( E d . P. K .
Bhattacharya) Inspec (1995)
[18] M. G. Astles, Liquid Phase
E p i t a x i a l G r o w t h , A d a m
Hilger (1990)
[19] A. Y. Cho, Appl. Phys. Lett. ,
467 (1971)
[20] E. H. C. Parker (Ed) , The
Technology and Physics of
M o l e c u l a r B e a m E p i t a x y,
Plenum Press (1985)
[21] G . B . S t r i n g f e l l o w ,
Organometallic Vapor Phase
Epitaxy,Academic Press (1989)
[22] C. J. Humphreys, MRS Bulletin
, 459 (2008)
[23] U. K. Mishra, L. Shen, T. E.
Kazior, Y. F. Wu, Proc. IEEE ,
287 (2007)
[24] P. G. Neudeck, R. S. Okojie and
L. Y. Chen, Proc. IEEE , 1065
(2002)
[25] H e t e r o j u n c t i o n B a n d
D i s c o n t i n u i t i e s , ( E d s . F .
Capasso and G. Margaritondo)
North Holland (1984)
[26] P. Harrison, Quantum Wells,
Wires and Dots, Wiley (2000)
[27] Light Wave Communications
Technology: Materials Growth
Technologies: Semiconductors
and Semimetals Vol. 22A (Ed.
W. T. Tsang) Academic press
(1985)
[28] K. J. Friedland, R. Hey, H.
Kostial, R. Klam and K. Ploog,
Phys. Rev. Lett. , 4616 (1996)
[29] M . H e n i n i , D i l u t e N i t r i d e
S e m i c o n d u c t o r s , E l s e v i e r
(2004)
[30] T. Kitada, T. Aoki, I. Watanabe,
S . S h i m o m u r a a n d S .
Hiyamizu, Appl. Phys. Lett. ,
4043 (2004).
[31] Strained Layer Superlattices:
S e m i c o n d u c t o r s a n d
Semimetals Vols. 32 and 33 (Ed.
T. P. Pearsall) Academic Press
(1990, 1991)
[32] J . S i n g h , P h y s i c s o f
S e m i c o n d u c t o r s a n d t h e i r
19
33
90
90
77
85
heterostructures, McGraw Hill
(1993)
[33] E. P. O'Reily, andA. R.Adams, IEEE
J. Quantum Electron. , 366 (1994)
[34] P. K. Basu and S. K. Paul, J.
Appl. Phys. , 3617 (1994)
[35] D. Bimberg, M. Grundmann a n d
N. N. Ledentsov, Quantum D o t
H e t e r o s t r u c t u r e s , W i l e y
(1999)
[36] M . S . S k o l n i c k a n d M .
G r u n d m a n n , A n n u l . R e v.
Mater. Res. , 181 (2004)
[37] P. Bhattacharya , J. Singh , H.
Yoon, X. Zhang, A. Gutierrez-
Aitken, and Y. Lam, IEEE J
Q u a n t u m E l e c t r o n ,
1620(1996); B. Kochman,A. D. Stiff
R o b e r t s , S . C h a k r a b a r t i ,
J.D.Phillips, S. Krishna, J. Singh
and P. Bhattacharya , IEEE J
Q u a n t u m E l e c t r o n 4 5 9
(2003)
[38] S. Ghosh, S. Constant and T. J. C .
Hosea, J. Appl. Phys. , 1432
(2000)
[39] S. Datta, S. Ghosh and B. M.
Arora, Rev. Sci. Instrum. , 177
(2001)
[40] J. Bhattacharyya, S. Ghosh, S.
Malzer, G. H. Dohler and B. M.
Arora, Appl. Phys. Lett. ,
212101 (2005)
[41] A. Arora, S. Ghosh, B. M. Arora,
S. Malzer and G. H. Dohler,
Appl. Phys. Lett. , 081902
(2010)
[42] R. C. Miller, D.A. Kleinmann a n d
A. C. Gossard, Phys. Rev. B
3 7 4 0 ( 1 9 8 4 ) , a n d 7 0 8 5
(1984)
[43] Z. H. Mahmood, A. P. Shah, A.
Kadir, M. R. Gokhale, S. Ghosh,
A. Bhattacharya and B. M.
Arora, Appl. Phys. Lett. ,
152108 (2007)
[44] Heterostructure Lasers, H. C.
C a s e y a n d M . B . Pa n i s h ,
Academic Press (1978)
[45] S. S. Chandvankar,A. P. Shah, A .
B h a t t a c h a r y a , K . S .
C h a n d r a s e k a r a n a n d B . M .
Arora, J. Cryst. Growth 348
(2004).
30
71
34
3 2
3 9
88
72
87
97
2 9 ,
91
260,
[46] Semiconductor Lasers (Ed. E.
Kapon)Academic Press (1999)
[47] P. Vasilev, Ultra fast Diode
Lasers,Artech House (1995)
[48] Diode Laser Arrays (Eds. D.
B o t e z a n d D . R . S c r i f e s )
Cambridge Univ. Press (1994)
[49] A. Majumdar, A.P. Shah, M.R.
Gokhale, S. Sen, S. Ghosh, B.M.
Arora and D.C. Tsui, IEEE J.
Quant. Electron, , 872 (2005)
[50] R. Lai, H. Wang, Y. C. Chen, T.
Block, P. H. Lie, D. C. Streit, D.
Tran, P. Siegel, M. Barsky, W.
Jones and T. Gaiver, Proc.
Indium Phosphide and Related
Materials (IPRM) p241 (1997)
[51] K. Van der Zander, M. Behet a n d
G. Borghs, Proc. Int. Conf.
G a l l i u m A r s e n i d e
Manufacturing Technology, p201
(1999)
[52] B . K a r m a k a r , M a g n e t o -
t r a n s p o r t s t u d y o f t w o -
dimensional electron gas in
InP/InGaAs/InP quantum well
structures (Ph.D. Thesis TIFR,
Mumbai, 2005, unpublished)
[53] B. Jeckelmann and B. Jeanneret,
Rep. Prog. Phys. , 1603 (2001)
[54] T. J. Bukowski and J. H. Simmons,
Critical Reviews of Solid State and
Materials Science , 119 (2002)
[55] M. A. Kastner, Ann. Phys. , 885
(2000)
[56] L. P. Kouwenhoven, J. M. Elzeman,
R. Hanson, L. H. Willems van
Beveren and L. M. K. Vandersypen,
Physica Stat. Solidi (b) , 3682
(2006)
[57] J. Berezovsky, M. H. Mikkelsen, O.
Gywat, W. G. Stoltz, L. A. Coldren
and D. D. Awschalom, Science ,
1916 (2006)
[ 5 8 ] B . K a r m a k a r ( p r i v a t e
communication)
[59] Z. Zhong, C. Yang and C. M. Leiber
in Nanosilicon, Ed. V. Kumar,
Elsevier (2007)
41
64
27
9
243
314