Phonons in Nanostructures-Cambridge University Press (2001)

This page intentionally left blank
Phonons in nanostructures
This book focuses on the theory of phonon interactions in nanoscale structures

with particular emphasis on modern electronic and optoelectronic devices.
The continuing progress in the fabrication of semiconductor nanostructures with
lower dimensional features has led to devices with enhanced functionality and
even to novel devices with new operating principles. The critical role of phonon
effects in such semiconductor devices is well known. There is therefore a
pressing need for a greater awareness and understanding of confined phonon
effects. A key goal of this book is to describe tractable models of confined
phonons and how these are applied to calculations of basic properties and
phenomena of semiconductor heterostructures.
The level of presentation is appropriate for undergraduate and graduate students
in physics and engineering with some background in quantum mechanics and
solid state physics or devices. A basic understanding of electromagnetism and
classical acoustics is assumed.
ii
D R M ICHAEL A. S TROSCIO earned a Ph.D. in physics from Yale University

and held research positions at the Los Alamos Scientific Laboratory and the
Johns Hopkins University Applied Physics Laboratory, before moving into the
management of federal research and development at a variety of US government
agencies. Dr Stroscio has served as a policy analyst for the White House Office
of Science and Technology Policy and as Vice Chairman of the White House
Panel on Scientific Communication. He has taught physics and electrical
engineering at several universities including Duke University, the North
Carolina State University and the University of California at Los Angeles. Dr
Stroscio is currently the Senior Scientist in the Office of the Director at the US
Army Research Office (ARO) as well as an Adjunct Professor at both Duke
University and the North Carolina State University. He has authored about 500
publications, presentations and patents covering a wide variety of topics in the
physical sciences and electronics. He is the author of Quantum Heterostructures:
Microelectronics and Optoelectronics and the joint editor of two World Scientific
books entitled Quantum-based Electronic Devices and Systems and Advances in
Semiconductor Lasers and Applications to Optoelectronics. He is a Fellow of
both the Institute of Electrical and Electronics Engineers (IEEE) and the
American Association for the Advancement of Science and he was the 1998
recipient of the IEEE Harry Diamond Award.
D R D UTTA earned a Ph.D. in physics from the University of Cincinnati; she was
a research associate at Purdue University and at City College, New York, as well
as a visiting scientist at Brookhaven National Laboratory before assuming a
variety of government posts in research and development. Dr Dutta was the
Director of the Physics Division at the US Army’s Electronics Technology and
Devices Laboratory as well as at the Army Research Laboratory prior to her
appointment as the Associate Director for Electronics in the Army Research
Office’s Engineering Sciences Directorate. Dr Dutta recently assumed a senior
executive position as ARO’s Director of Research and Technology Integration.
She has over 160 publications, 170 conference presentations, 10 book chapters,
and has had 24 US patents issued. She is the joint editor of two World Scientific
books entitled Quantum-Based Electronic Devices and Systems and Advances in
Semiconductor Lasers and Applications to Optoelectronics. She is an Adjunct
Professor of the Electrical and Computer Engineering and Physics departments
of North Carolina State University and has had adjunct appointments at the
Electrical Engineering departments of Rutgers University and the University of
Maryland. Dr Dutta is a Fellow of both the Institute of Electrical and Electronics
Engineers (IEEE) and the Optical Society of America, and she was the recipient
in the year 2000 of the IEEE Harry Diamond Award.
Phonons in
Nanostructures
Michael A. Stroscio and Mitra Dutta
US Army Research Office, US Army Research Laboratory
         
The Pitt Building, Trumpington Street, Cambridge, United Kingdom
  

The Edinburgh Building, Cambridge CB2 2RU, UK
40 West 20th Street, New York, NY 10011-4211, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
Ruiz de Alarcón 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South Africa
http://www.cambridge.org
© Michael A. Stroscio and Mitra Dutta 2004
First published in printed format 2001
ISBN 0-511-03187-4 eBook (Adobe Reader)

ISBN 0-521-79279-7 hardback
Mitra Dutta dedicates this book to her parents
Dhiren N. and Aruna Dutta
and
Michael Stroscio dedicates this book to

his friend and mentor Morris Moskow and
his friend and colleague Ki Wook Kim
Contents
Preface xi
Chapter 1 Phonons in nanostructures 1
1.1 Phonon effects: fundamental limits on carrier mobilities and dynamical

processes 1
1.2 Tailoring phonon interactions in devices with nanostructure components 3
Chapter 2 Phonons in bulk cubic crystals 6
2.1 Cubic structure 6

2.2 Ionic bonding – polar semiconductors 6
2.3 Linear-chain model and macroscopic models 7
2.3.1 Dispersion relations for high-frequency and low-frequency modes 8
2.3.2 Displacement patterns for phonons 10
2.3.3 Polaritons 11
2.3.4 Macroscopic theory of polar modes in cubic crystals 14
Chapter 3 Phonons in bulk würtzite crystals 16
3.1 Basic properties of phonons in würtzite structure 16

3.2 Loudon model of uniaxial crystals 18
3.3 Application of Loudon model to III-V nitrides 23
Chapter 4 Raman properties of bulk phonons 26
4.1 Measurements of dispersion relations for bulk samples 26

4.2 Raman scattering for bulk zincblende and würtzite structures 26
vii
viii Contents
4.2.1 Zincblende structures 28

4.2.2 Würtzite structures 29
4.3 Lifetimes in zincblende and würtzite crystals 30
4.4 Ternary alloys 32
4.5 Coupled plasmon–phonon modes 33
Chapter 5 Occupation number representation 35
5.1 Phonon mode amplitudes and occupation numbers 35

5.2 Polar-optical phonons: Fröhlich interaction 40
5.3 Acoustic phonons and deformation-potential interaction 43
5.4 Piezoelectric interaction 43
Chapter 6 Anharmonic coupling of phonons 45
6.1 Non-parabolic terms in the crystal potential for ionically bonded atoms 45
6.2 Klemens’ channel for the decay process LO → LA(1) + LA(2) 46
6.3 LO phonon lifetime in bulk cubic materials 47
6.4 Phonon lifetime effects in carrier relaxation 48
6.5 Anharmonic effects in würtzite structures: the Ridley channel 50
Chapter 7 Continuum models for phonons 52
7.1 Dielectric continuum model of phonons 52

7.2 Elastic continuum model of phonons 56
7.3 Optical modes in dimensionally confined structures 60
7.3.1 Dielectric continuum model for slab modes: normalization of
interface modes 61
7.3.2 Electron–phonon interaction for slab modes 66
7.3.3 Slab modes in confined würtzite structures 71
7.3.4 Transfer matrix model for multi-heterointerface structures 79
7.4 Comparison of continuum and microscopic models for phonons 90
7.5 Comparison of dielectric continuum model predictions with Raman
measurements 93
7.6 Continuum model for acoustic modes in dimensionally confined structures 97
7.6.1 Acoustic phonons in a free-standing and unconstrained layer 97
7.6.2 Acoustic phonons in double-interface heterostructures 100
7.6.3 Acoustic phonons in rectangular quantum wires 105
7.6.4 Acoustic phonons in cylindrical structures 111
7.6.5 Acoustic phonons in quantum dots 124
Contents ix
Chapter 8 Carrier–LO-phonon scattering 131
8.1 Fröhlich potential for LO phonons in bulk zincblende and würtzite

structures 131
8.1.1 Scattering rates in bulk zincblende semiconductors 131
8.1.2 Scattering rates in bulk würtzite semiconductors 136
8.2 Fröhlich potential in quantum wells 140
8.2.1 Scattering rates in zincblende quantum-well structures 141
8.2.2 Scattering rates in würtzite quantum wells 146
8.3 Scattering of carriers by LO phonons in quantum wires 146
8.3.1 Scattering rate for bulk LO phonon modes in quantum wires 146
8.3.2 Scattering rate for confined LO phonon modes in quantum wires 150
8.3.3 Scattering rate for interface-LO phonon modes 154
8.3.4 Collective effects and non-equilibrium phonons in polar quantum wires 162
8.3.5 Reduction of interface–phonon scattering rates in metal–semiconductor
structures 165
8.4 Scattering of carriers and LO phonons in quantum dots 167
Chapter 9 Carrier–acoustic-phonon scattering 172
9.1 Carrier–acoustic-phonon scattering in bulk zincblende structures 172

9.1.1 Deformation-potential scattering in bulk zincblende structures 172
9.1.2 Piezoelectric scattering in bulk semiconductor structures 173
9.2 Carrier–acoustic-phonon scattering in two-dimensional structures 174
9.3 Carrier–acoustic-phonon scattering in quantum wires 175
9.3.1 Cylindrical wires 175
9.3.2 Rectangular wires 181
Chapter 10 Recent developments 186
10.1 Phonon effects in intersubband lasers 186

10.2 Effect of confined phonons on gain of intersubband lasers 195
10.3 Phonon contribution to valley current in double-barrier structures 202
10.4 Phonon-enhanced population inversion in asymmetric double-barrier quantum-well
lasers 205
10.5 Confined-phonon effects in thin film superconductors 208
10.6 Generation of acoustic phonons in quantum-well structures 212
Chapter 11 Concluding considerations 218
11.1 Pervasive role of phonons in modern solid-state devices 218

11.2 Future trends: phonon effects in nanostructures and phonon engineering 219
x Contents
Appendices 221
Appendix A: Huang–Born theory 221
Appendix B: Wendler’s theory 222
Appendix C: Optical phonon modes in double-heterointerface structures 225
Appendix D: Optical phonon modes in single- and double-heterointerface würtzite
structures 236
Appendix E: Fermi golden rule 250
Appendix F: Screening effects in a two-dimensional electron gas 252
References 257
Index 271
Preface
This book describes a major aspect of the effort to understand nanostructures,

namely the study of phonons and phonon-mediated effects in structures with
nanoscale dimensional confinement in one or more spatial dimensions. The neces-
sity for and the timing of this book stem from the enormous advances made in the
field of nanoscience during the last few decades.
Indeed, nanoscience continues to advance at a dramatic pace and is making
revolutionary contributions in diverse fields, including electronics, optoelectronics,
quantum electronics, materials science, chemistry, and biology. The technologies
needed to fabricate nanoscale structures and devices are advancing rapidly. These
technologies have made possible the design and study of a vast array of novel
devices, structures and systems confined dimensionally on the scale of 10 nanome-
ters or less in one or more dimensions. Moreover, nanotechnology is continuing
to mature rapidly and will, no doubt, lead to further revolutionary breakthroughs
like those exemplified by quantum-dot semiconductor lasers operating at room
temperature, intersubband multiple quantum-well semiconductor lasers, quantum-
wire semiconductor lasers, double-barrier quantum-well diodes operating in the
terahertz frequency range, single-electron transistors, single-electron metal-oxide–
semiconductor memories operating at room temperature, transistors based on carbon
nanotubes, and semiconductor nanocrystals used for fluorescent biological labels,
just to name a few!
The seminal works of Esaki and Tsu (1970) and others on the semiconductor
superlattice stimulated a vast international research effort to understand the fabrica-
tion and electronic properties of superlattices, quantum wells, quantum wires, and
quantum dots. This early work led to truly revolutionary advances in nanofabrication
xi
xii Preface
technology and made it possible to realize band-engineering and atomic-level

structural tailoring not envisioned previously except through the molecular and
atomic systems found in nature. Furthermore, the continuing reduction of dimen-
sional features in electronic and optoelectronic devices coupled with revolutionary
advances in semiconductor growth and processing technologies have opened many
avenues for increasing the performance levels and functionalities of electronic and
optoelectronic devices. Likewise, the discovery of the buckyball by Kroto et al.
(1985) and the carbon nanotube by Iijima (1991) led to an intense worldwide
program to understand the properties of these nanostructures.
During the last decade there has been a steady effort to understand the optical
and acoustic phonons in nanostructures such as the semiconductor superlattice,
quantum wires, and carbon nanotubes. The central theme of this book is the
description of the optical and acoustic phonons in these nanostructures. It deals
with the properties of phonons in isotropic, cubic, and hexagonal crystal structures
and places particular emphasis on the two dominant structures underlying modern
semiconductor electronics and optoelectronics – zincblende and würtzite. In view
of the successes of continuum models in describing optical phonons (Fasol et al.,
1988) and acoustic phonons (Seyler and Wybourne, 1992) in dimensionally confined
structures, the principal theoretical descriptions presented in this book are based
on the so-called dielectric continuum model of optical phonons and the elastic
continuum model of acoustic phonons. Many of the derivations are given for the
case of optical phonons in würtzite crystals, since the less complicated case for
zincblende crystals may then be recovered by taking the dielectric constants along
the c-axis and perpendicular to the c-axis to be equal.
As a preliminary to describing the dispersion relations and mode structures for
optical and acoustic phonons in nanostructures, phonon amplitudes are quantized in
terms of the harmonic oscillator approximation, and anharmonic effects leading to
phonon decay are described in terms of the dominant phonon decay channels. These
dielectric and elastic continuum models are applied to describe the deformation-
potential, Fröhlich, and piezoelectric interactions in a variety of nanostructures
including quantum wells, quantum wires and quantum dots. Finally, this book
describes how the dimensional confinement of phonons in nanostructures leads to
modifications in the electronic, optical, acoustic, and superconducting properties of
selected devices and structures including intersubband quantum-well semiconductor
lasers, double-barrier quantum-well diodes, thin-film superconductors, and the thin-
walled cylindrical structures found in biological structures known as microtubulin.
The authors wish to acknowledge professional colleagues, friends and family
members without whose contributions and sacrifices this work would not have been
undertaken or completed. The authors are indebted to Dr C.I. (Jim) Chang, who is
both the Director of the US Army Research Office (ARO) and the Deputy Director
of the US Army Research Laboratory for Basic Science, and to Dr Robert W. Whalin
and Dr John Lyons, the current director and most recent past director of the US Army
Preface xiii
Research Laboratory; these leaders have placed a high priority on maintaining an

environment at the US Army Research Office such that it is possible for scientists
at ARO to continue to participate personally in forefront research as a way of
maintaining a broad and current knowledge of selected fields of modern science.
Michael Stroscio acknowledges the important roles that several professional
colleagues and friends played in the events leading to his contributions to this
book. These people include: Professor S. Das Sarma of the University of Maryland;
Professor M. Shur of the Rensselaer Polytechnic Institute; Professor Gerald J. Iafrate
of Notre Dame University; Professors M.A. Littlejohn, K.W. Kim, R.M. Kolbas, and
N. Masnari of the North Carolina State University (NCSU); Dr Larry Cooper of the
Office of Naval Research; Professor Vladimir Mitin of the Wayne State University;
Professors H. Craig Casey Jr, and Steven Teitsworth of Duke University; Professor
S. Bandyopadhyay of the University of Nebraska; Professors G. Belenky, Vera B.
Gorfinkel, M. Kisin, and S. Luryi of the State University of New York at Stony
Brook; Professors George I. Haddad, Pallab K. Bhattacharya, and Jasprit Singh and
Dr J.-P. Sun of the University of Michigan; Professors Karl Hess and J.-P. Leburton
at the University of Illinois; Professor L.F. Register of the University of Texas at
Austin; Professor Viatcheslav A. Kochelap of the National Academy of Sciences of
the Ukraine; Dr Larry Cooper of the Office of Naval Research; and Professor Paul
Klemens of the University of Connecticut. Former graduate students, postdoctoral
researchers, and visitors to the North Carolina State University who contributed
substantially to the understanding of phonons in nanostructures as reported in this
book include Drs Amit Bhatt, Ulvi Erdogan, Daniel Kahn, Sergei M. Komirenko,
Byong Chan Lee, Yuri M. Sirenko, and SeGi Yu. The fruitful collaboration of Dr
Rosa de la Cruz of the Universidad Carlos III de Madrid during her tenure as a
visiting professor at Duke University is acknowledged gratefully. The authors also
acknowledge gratefully the professionalism and dedication of Mrs Jayne Aldhouse
and Drs Simon Capelin and Eoin O’Sullivan, of Cambridge University Press, and
Dr Susan Parkinson.
Michael Stroscio thanks family members who have been attentive during the
periods when his contributions to the book were being written. These include:
Anthony and Norma Stroscio, Mitra Dutta, as well as Gautam, Marshall, and
Elizabeth Stroscio. Moreover, eight-year-old Gautam Stroscio is acknowledged
gratefully for his extensive assistance in searching for journal articles at the North
Carolina State University.
Mitra Dutta acknowledges the interactions, discussions and work of many
colleagues and friends who have had an impact on the work leading to this book.
These colleagues include Drs Doran Smith, K.K. Choi, and Paul Shen of the Army
Research Laboratory, Professor Athos Petrou of the State University of New York
at Buffalo, and Professors K.W. Kim, M.A. Littlejohn, R.J. Nemanich, Dr Leah
Bergman and Dimitri Alexson of the North Carolina State University, as well
as Professors Herman Cummins, City College, New York, A.K. Ramdas, Purdue
xiv Preface
University and Howard Jackson, University of Cincinnati, her mentors in various

facets of phonon physics. Mitra Dutta would also like to thank Dhiren Dutta, without
whose encouragement she would never have embarked on a career in science, as well
as Michael and Gautam Stroscio who everyday add meaning to everything.
Michael Stroscio and Mitra Dutta

Chapter 1
Phonons in nanostructures
There are no such things as applied sciences, only applications of

sciences.
Louis Pasteur, 1872
1.1 Phonon effects: fundamental limits on carrier

mobilities and dynamical processes
The importance of phonons and their interactions in bulk materials is well known to
those working in the fields of solid-state physics, solid-state electronics, optoelec-
tronics, heat transport, quantum electronics, and superconductivity.
As an example, carrier mobilities and dynamical processes in polar semiconduc-
tors, such as gallium arsenide, are in many cases determined by the interaction of
longitudinal optical (LO) phonons with charge carriers. Consider carrier transport
in gallium arsenide. For gallium arsenide crystals with low densities of impurities
and defects, steady state electron velocities in the presence of an external electric
field are determined predominantly by the rate at which the electrons emit LO
phonons. More specifically, an electron in such a polar semiconductor will accelerate
in response to the external electric field until the electron’s energy is large enough for
the electron to emit an LO phonon. When the electron’s energy reaches the threshold
for LO phonon emission – 36 meV in the case of gallium arsenide – there is a
significant probability that it will emit an LO phonon as a result of its interaction
with LO phonons. Of course, the electron will continue to gain energy from the
electric field.
In the steady state, the processes of electron energy loss by LO phonon emission
and electron energy gain from the electric field will come into balance and the
electron will propagate through the semiconductor with a velocity known as the
saturation velocity. As is well known, experimental values for this saturated drift
velocity generally fall in the range 107 cm s−1 to 108 cm s−1 . For gallium arsenide
this velocity is about 2 × 107 cm s−1 and for indium antimonide 6 × 107 cm s−1 .
1
2 1 Phonons in nanostructures
For both these polar semiconductors, the process of LO phonon emission plays a
major role in determining the value of the saturation velocity. In non-polar materials
such as Si, which has a saturation velocity of about 107 cm s−1 , the deformation-
potential interaction results in electron energy loss through the emission of phonons.
(In Chapter 5 both the interaction between polar-optical-phonons and electrons –
known as the Fröhlich interaction – and the deformation-potential interaction will
be defined mathematically.)
Clearly, in all these cases, the electron mobility will be influenced strongly by the
interaction of the electrons with phonons. The saturation velocity of the carriers in
a semiconductor provides a measure of how fast a microelectronic device fabricated
from this semiconductor will operate. Indeed, the minimum time for the carriers to
travel through the active region of the device is given approximately by the length
of the device – that is, the length of the so-called gate – divided by the saturation
velocity. Evidently, the practical switching time of such a microelectronic device
will be limited by the saturation velocity and it is clear, therefore, that phonons play
a major role in the fundamental and practical limits of such microelectronic devices.
For modern integrated circuits, a factor of two reduction in the gate length can be
achieved in many cases only through building a new fabrication facility. In some
cases, such a building project might cost a billion dollars or more. The importance
of phonons in microelectronics is clear!
A second example of the importance of carrier–phonon interactions in modern
semiconductor devices is given by the dynamics of carrier capture in the active
quantum-well region of a polar semiconductor quantum-well laser. Consider the
case where a current of electrons is injected over a barrier into the quantum-well
region of such a laser. For the laser to operate, an electron must lose enough energy
to be ‘captured’ by the quasi-bound state which it must occupy to participate in
the lasing process. For many quantum-well semiconductor lasers this means that
the electron must lose an energy of the order of a 100 meV or more. The energy
loss rate of a carrier – also known as the thermalization rate of the carrier – in
a polar-semiconductor quantum well is determined by both the rate at which the
carrier’s energy is lost by optical-phonon emission and the rate at which the carrier
gains energy from optical-phonon absorption. This latter rate can be significant
in quantum wells since the phonons emitted by energetic carriers can accumulate
in these structures. Since the phonon densities in many dimensionally confined
semiconductor devices are typically well above those of the equilibrium phonon
population, there is an appreciable probability that these non-equilibrium – or ‘hot’
– phonons will be reabsorbed. Clearly, the net loss of energy by an electron in such
a situation depends on the rates for both phonon absorption and phonon emission.
Moreover, the lifetimes of the optical phonons are also important in determining the
total energy loss rate for such carriers. Indeed, as will be discussed in Chapter 6, the
longitudinal optical (LO) phonons in GaAs and many other polar materials decay
into acoustic phonons through the Klemens’ channel. Furthermore, over a wide
1.2 Devices with nanostructure components 3
range of temperatures and phonon wavevectors, the lifetimes of longitudinal optical

phonons in GaAs vary from a few picoseconds to about 10 ps (Bhatt et al., 1994).
(Typical lifetimes for other polar semiconductors are also of this magnitude.) As
a result of the Klemens’ channel, the ‘hot’ phonons decay into acoustic phonons in
times of the order of 10 ps. The LO phonons undergoing decay into acoustic phonons
are not available for absorption by the electrons and as a result of the Klemens’
channel the electron thermalization is more rapid than it would be otherwise; this
phenomenon is referred to as the ‘hot-phonon-bottleneck effect’.
The electron thermalization time is an important parameter for semiconductor
quantum-well lasers because it determines the minimum time needed to switch the
laser from an ‘on’ state to an ‘off’ state; this occurs as a result of modulating the
electron current that leads to lasing. Since the hot-phonon population frequently
decays on a time scale roughly given by the LO phonon decay rate (Das Sarma
et al., 1992), a rough estimate of the electron thermalization time – and therefore
the minimum time needed to switch the laser from an ‘on’ state to an ‘off’ state –
is of the order of about 10 ps. In fact, typical modulation frequencies for gallium
arsenide quantum-well lasers are about 30 GHz. The modulation of the laser at
significantly higher frequencies will be limited by the carrier thermalization time
and ultimately by the lifetime of the LO phonon. The importance of the phonon in
modern optoelectronics is clear.
The importance of phonons in superconductors is well known. Indeed, the
Bardeen–Cooper–Schrieffer (BCS) theory of superconductivity is based on the
formation of bosons from pairs of electrons – known as Cooper pairs – bound
through the mediating interaction produced by phonons. Many of the theories
describing the so-called high-critical-temperature superconductors are not based on
phonon-mediated Cooper pairs, but the importance of phonons in many supercon-
ductors is of little doubt. Likewise, it is generally recognized that acoustic phonon
interactions determine the thermal properties of materials.
These examples illustrate the pervasive role of phonons in bulk materials.
Nanotechnology is providing an ever increasing number of devices and structures
having one, or more than one, dimension less than or equal to about 100 ångstroms.
The question naturally arises as to the effect of dimensional confinement on the
properties on the phonons in such nanostructures as well as the properties of the
phonon interactions in nanostructures. The central theme of this book is the descrip-
tion of the optical and acoustic phonons, and their interactions, in nanostructures.
1.2 Tailoring phonon interactions in devices with

nanostructure components
Phonon interactions are altered unavoidably by the effects of dimensional confine-
ment on the phonon modes in nanostructures. These effects exhibit some similarities
4 1 Phonons in nanostructures
to those for an electron confined in a quantum well. Consider the well-known

wavefunction of an electron in a infinitely deep quantum well, of width L z in the
z-direction. The energy eigenstates n (z) may be taken as plane-wave states in the
directions parallel to the heterointerfaces and as bound states in an infinitely deep
quantum well in the z-direction:

eik ·r 2
n (z) = √ sin k z z, (1.1)
A Lz
where r and k are the position vector and wavevector components in a plane
parallel to the interfaces, k z = nπ/L z , and n = 1, 2, 3, . . . labels the energy
eigenstates, whose energies are
h̄ 2 (k )2 h̄ 2 π 2 n 2
E n (k ) = + . (1.2)
2m 2m L 2z
A is the area of the heterointerface over which the electron wavefunction is

normalized. Clearly, a major effect of dimensional confinement in the z-direction
is that the z-component of the bulk continuum wavevector is restricted to integral
multiples of π/L z . Stated in another way, the phase space is restricted.
As will be explained in detail in Chapter 7, the dimensional confinement of
phonons results in similar restrictions in the phase space of the phonon wavevector
q. Indeed, we shall show that the wavevectors of the optical phonons in a dielectric
layer of thickness L z are given by qz = nπ/L z (Fuchs and Kliewer, 1965) in analogy
to the case of an electron in an infinitely deep quantum well. In fact, Fasol et al.
(1988) used Raman scattering techniques to show that the wavevectors qz = nπ/L z
of optical phonons confined in a ten-monolayer-thick AlAs/GaAs/AlAs quantum
well are so sensitive to changes in L z that a one-monolayer change in the thickness
of the quantum well is readily detectable as a change in qz ! These early experimental
studies of Fasol et al. (1988) demonstrated not only that phonons are confined in
nanostructures but also that the measured phonon wavevectors are well described by
relatively simple continuum models of phonon confinement.
Since dimensional confinement of phonons restricts the phase space of the
phonons, it is certain that carrier–phonon interactions in nanostructures will be
modified by phonon confinement. As we shall see in Chapter 7, the so-called di-
electric and elastic continuum models of phonons in nanostructures may be applied
to describe the deformation-potential, Fröhlich, and piezoelectric interactions in a
variety of nanostructures including quantum wells, quantum wires, and quantum
dots. These interactions play a dominant role in determining the electronic, optical
and acoustic properties of materials (Mitin et al., 1999; Dutta and Stroscio, 1998b;
Dutta and Stroscio, 2000); it is clearly desirable for models of the properties
of nanostructures to be based on an understanding of how the above-mentioned
interactions change as a result of dimensional confinement. To this end, Chapters
1.2 Devices with nanostructure components 5
8, 9 and 10 of this book describe how the dimensional confinement of phonons in

nanostructures leads to modifications in the electronic, optical, acoustic, and su-
perconducting properties of selected devices and structures, including intersubband
quantum-well semiconductor lasers, double-barrier quantum-well diodes, thin-film
superconductors, and the thin-walled cylindrical structures found in the biological
structures known as microtubulin. Chapters 8, 9, and 10 also provide analyses of the
role of collective effects and non-equilibrium phonons in determining hot-carrier
energy loss in polar quantum wires as well as the use of metal–semiconductor
structures to tailor carrier–phonon interactions in nanostructures. Moreover, Chapter
10 describes how confined phonons play a critical role in determining the properties
of electronic, optical, and superconducting devices containing nanostructures as
essential elements. Examples of such phonon effects in nanoscale devices include:
phonon effects in intersubband lasers; the effect of confined phonons on the gain
of intersubband lasers; the contribution of confined phonons to the valley current in
double-barrier quantum-well structures; phonon-enhanced population inversion in
asymmetric double-barrier quantum-well lasers; and confined phonon effects in thin
film superconductors.
Chapter 2
Phonons in bulk cubic crystals
The Creator, if He exists, has a special preference for beetles.

J.B.S. Haldane, 1951
2.1 Cubic structure

Crystals with cubic structure are of major importance in the fields of electronics and
optoelectronics. Indeed, zincblende crystals such as silicon, germanium, and gallium
arsenide may be regarded as two face-centered cubic (fcc) lattices displaced relative
to each other by a vector (a/4, a/4, a/4), where a is the size of the smallest unit of
the fcc structure. Figure 2.1 shows a lattice with the zincblende structure.
A major portion of this book will deal with phonons in cubic crystals. In
addition, we will describe the phonons in so-called isotropic media, which are
related mathematically to cubic media as explained in detail in Section 7.2. The
remaining portions of this book will deal with crystals of würtzite structure, defined
in Chapter 3. More specifically, the primary focus of this book concerns phonons
in crystalline structures that are dimensionally confined in one, two, or three
dimensions. Such one-, two-, and three-dimensional confinement is realized in
quantum wells, quantum wires, and quantum dots, respectively. As a preliminary
to considering phonons in dimensionally confined structures, the foundational case
of phonons in bulk structures will be treated. The reader desiring to supplement this
chapter with additional information on the basic properties of phonons in bulk cubic
materials will find excellent extended treatments in a number of texts including
Blakemore (1985), Ferry (1991), Hess (1999), Kittel (1976), Omar (1975), and
Singh (1993).
2.2 Ionic bonding – polar semiconductors

As is well known, the crystal structure of silicon is the zincblende structure shown
in Figure 2.1. The covalent bonding in silicon does not result in any net transfer
of charge between silicon atoms. More specifically, the atoms on the two displaced
6
face-centered cubic (fcc) lattices depicted in Figure 2.1 have no excess or deficit
of charge relative to the neutral situation. This changes dramatically for polar
semiconductors like gallium arsenide, since here the ionic bonding results in charge
transfer from the Group V arsenic atoms to the Group III gallium atoms: Since
Group V atoms have five electrons in the outer shell and Group III atoms have
three electrons in the outer shell, it is not surprising that the gallium sites acquire
a net negative charge and the arsenic sites a net positive charge. In binary polar
semiconductors, the two atoms participating in the ionic bonding carry opposite
charges, e∗ and −e∗ , respectively, as a result of the redistribution of the charge
associated with polar bonding. In polar materials such ionic bonding is characterized
by values of e∗ within an order of magnitude of unity. In the remaining sections of
this chapter, it will become clear that e∗ is related to the readily measurable or known
ionic masses, phonon optical frequencies, and high-frequency dielectric constant of
the polar semiconductor.
2.3 Linear-chain model and macroscopic models

The linear-chain model of a one-dimensional diatomic crystal is based upon a system
of two atoms with masses, m and M, placed along a one-dimensional chain as
depicted in Figure 2.2. As for a diatomic lattice, the masses are situated alternately
along the chain and their separation is a. On such a chain the displacement of
one atom from its equilibrium position will perturb the positions of its neighboring
atoms.
Figure 2.1. Zincblende

crystal. The white spheres
and black spheres lie on
different fcc lattices.
Figure 2.2. One-dimensional linear-chain representation of a diatomic lattice.

8 2 Phonons in bulk cubic crystals
In the simple linear-chain model considered in this section, it is assumed that

only nearest neighbors are coupled and that the interaction between these atoms is
described by Hooke’s law; the spring constant α is taken to be that of a harmonic
oscillator. This model describes many of the basic properties of a diatomic lattice.
However, as will become clear in Chapter 6, it is essential to supplement the so-
called ‘harmonic’ interactions with anharmonic interactions in order to describe the
important process of phonon decay.
2.3.1 Dispersion relations for high-frequency and

low-frequency modes
To model the normal modes of this system of masses, the atomic displacements
along the direction of the chain – the so-called longitudinal displacements of each
of the two types of atoms – are taken to be
u 2r = A1 ei(2rqa−ωt) (2.1)
and
u 2r +1 = A2 ei[(2r +1)qa−ωt] (2.2)
where q is the phonon wavevector and ω is its frequency. In the nearest-neighbor
approximation, these longitudinal displacements satisfy
m(d 2 u 2r /dt 2 ) = −α(u 2r − u 2r −1 ) − α(u 2r − u 2r +1 )
= α(u 2r +1 + u 2r −1 − 2u 2r ) (2.3)
and
M(d 2 u 2r +1 /dt 2 ) = −α(u 2r +1 − u 2r ) − α(u 2r +1 − u 2r +2 )
= α(u 2r +2 + u 2r − 2u 2r +1 ). (2.4)
The signs in the four terms on the right-hand sides of these equations are
determined by considering the relative displacements of neighboring atoms. For
example, if the positive displacement of u 2r is greater than that of u 2r −1 there is
a restoring force −α(u 2r +1 − u 2r ). Hence
−mω2 A1 = α A2 (eiqa + e−iqa ) − 2α A1 (2.5)
and
−Mω2 A2 = α A1 (eiqa + e−iqa ) − 2α A2 . (2.6)
Eliminating A1 and A2 ,

1 1 1 1 2 4 sin2 qa 1/2
ω =α
2
+ ±α + − . (2.7)
m M m M mM
This relationship between frequency and wavevector is commonly called a dis-
persion relation. The higher-frequency solution is known as the optical mode
since, for many semiconductors, its frequency is in the terahertz range, which
happens to coincide with the infrared portion of the electromagnetic spectrum.
The lower-frequency solution is known as the acoustic mode. More precisely, since
only longitudinal displacements have been modeled, these two solutions correspond
to the longitudinal optical (LO) and longitudinal acoustic (LA) modes of the
linear-chain lattice. Clearly, the displacements along this chain can be described
in terms of wavevectors q in the range from −π/2a to π/2a. From the solution
for ω, it is evident that over this Brillouin zone the LO modes have a maximum
frequency [2α(1/m + 1/M)]1/2 at the center of the Brillouin zone and a minimum
frequency (2α/m)1/2 at the edge of the Brillouin zone. Likewise, the LA modes have
a maximum frequency (2α/M)1/2 at the edge of the Brillouin zone and a minimum
frequency equal to zero at the center of the Brillouin zone.
In polar semiconductors, the masses m and M carry opposite charges, e∗ and
−e∗ , respectively, as a result of the redistribution of the charge associated with
polar bonding. In polar materials such ionic bonding is characterized by values of
e∗ equal to 1, to an order-of-magnitude. When there is an electric field E present
in the semiconductor, it is necessary to augment the previous force equation with
terms describing the interaction with the charge. In the long-wavelength limit of the
electric field E, the force equations then become
−mω2 u 2r = m(d 2 u 2r /dt 2 ) = α(u 2r +1 + u 2r −1 − 2u 2r ) + e∗ E

= α(ei2qa + 1)u 2r −1 − 2αu 2r + e∗ E (2.8)
and
−Mω2 u 2r +1 = M(d 2 u 2r +1 /dt 2 ) = α(u 2r +2 + u 2r − 2u 2r +1 ) − e∗ E

= α(1 + e−i2qa )u 2r +2 − 2αu 2r +1 − e∗ E. (2.9)
Regarding the phonon displacements, in the long-wavelength limit there is no

need to distinguish between the different sites for a given mass type since all atoms
of the same mass are displaced by the same amount. In this limit, q → 0. Denoting
the displacements on even-numbered sites by u 1 and those on odd-numbered sites
by u 2 , in the long-wavelength limit the force equations reduce to
−mω2 u 1 = 2α(u 2 − u 1 ) + e∗ E (2.10)
and
−Mω2 u 2 = 2α(u 1 − u 2 ) − e∗ E. (2.11)
Adding these equations demonstrates that −mω2 u 1 − Mω2 u 2 = 0 and it is clear

that mu 1 = −Mu 2 ; thus

m
−mω u 1 = 2α − u 1 − u 1 + e∗ E
2
(2.12)
M
and

m
−Mω u 2 = 2α u 1 + u 1 − e∗ E;
2
(2.13)
M
accordingly,
−(ω2 − ω02 )u 1 = e∗ E/m (2.14)
and
−(ω2 − ω02 )u 2 = −e∗ E/M (2.15)
where ω02 = 2α(1/m + 1/M) is the resonant frequency squared, in the absence of
Coulomb effects; that is, for e∗ = 0. The role of e∗ in shifting the phonon frequency
will be discussed further in the next section.
Clearly, the electric polarization P produced by such a polar diatomic lattice is
given by

N e∗ u N e∗ (u 1 − u 2 ) 1 N e∗2 1 1
P= = = + E, (2.16)
(∞) (∞) (∞) (ω02 − ω2 ) m M
where u = u 1 − u 2 , N is the number of pairs per unit volume, and e∗ is as

defined previously. This equation may be rewritten to show that it describes a driven
oscillator:

1 1
(ω02 − ω2 )u = e∗ + E. (2.17)
m M
2.3.2 Displacement patterns for phonons

As discussed in subsection 2.3.1, in the limit q → 0 the displacements, u 1 and u 2 ,
of the optical modes satisfy −mu 1 = Mu 2 and the amplitudes of the two types of
mass have opposite signs. That is, for the optical modes the atoms vibrate out of
phase, and so with their center of mass fixed. For the acoustic modes, the maximum
frequency is (2α/M)1/2 . This maximum frequency occurs at the zone edge so that,
near the center of the zone, ω is much less than (2α/M)1/2 . From subsection 2.3.1,
the ratio A2 /A1 may be expressed as
A2 2α cos qa 2α − mω2
= = , (2.18)
A1 2α − Mω 2 2α cos qa
and it is clear that the ratio of the displacement amplitudes is approximately equal
to unity for acoustic phonons near the center of the Brillouin zone. Thus, in contrast
to the optical modes, the acoustic modes are characterized by in-phase motion of
the different masses m and M. Typical mode patterns for zone-center acoustic
and optical modes are depicted in Figures 2.3(a), (b). The transverse modes are
illustrated here since the longitudinal modes are more difficult to depict graphically.
The higher-frequency optical modes involve out-of-plane oscillations of adjacent
ions, while the lower-frequency acoustic modes are characterized by motion of
adjacent ions on the same sinusoidal curve.
2.3.3 Polaritons
In the presence of a transverse electric field, transverse optical (TO) phonons of
a polar medium couple strongly to the electric field. When the wavevectors and
frequencies of the electric field are in resonance with those of the TO phonon, a
coupled phonon–photon field is necessary to describe the system. The quantum of
this coupled field is known as the polariton. The analysis of subsection 2.3.1 may
be generalized to apply to the case of transverse displacements. In particular, for a
transverse field E, the oscillator equation takes the form

N e∗2 1 1
(ωTO
2
− ω2 )P = + E, (2.19)
(∞) m M
where ω02 of subsection 2.3.1 has been designated ωTO

2 = 2α(1/m + 1/M) since
the resonant frequency in the absence of Coulomb effects, e∗2 = 0, corresponds to
Figure 2.3. Transverse displacements of heavy ions (large disks) and light ions
(small disks) for (a) transverse acoustic modes, and (b) transverse optical modes
propagating in the q-direction.
the transverse optical frequency. As will become apparent later in this section, the
LO phonon frequency squared differs from the TO phonon frequency squared by an
amount proportional to e∗2 .
According to the electromagnetic wave equation, ∂ 2 D/∂t 2 = c2 ∇ 2 E, where
D = E + 4π P, the dispersion relation describing the coupling of the field E of the
electromagnetic wave to the electric polarization P of the TO phonon is
c2 q 2 E = ω2 (E + 4π P) (2.20)
or, alternatively,
4πω2 P = (c2 q 2 − ω2 )E, (2.21)
where waves of the form ei(qr −ωt) have been assumed. The driven oscillator
equation and the electromagnetic wave equation have a joint solution when the
determinant of the coefficients of the fields E and P vanishes,

ω2 − c2 q 2 4π ω2

N e∗2 1 1 =0 (2.22)
−(ωTO − ω )
(∞) m + M
2 2
At q = 0, there are two roots: ω = 0 and

N e∗2 1 1
ω = ωTO + 4π
2 2
+ = ωLO
2
. (2.23)
(∞) m M
The dielectric function (ω) is then given by

D(ω) 4π Pe (ω) 4π P(ω)
(ω) = =1+ +
E(ω) E(ω) E(ω)

4π Pe (ω) 4π N e∗2 1 1
=1+ + 2 + , (2.24)
E(ω) (ωTO − ω2 ) (∞) m M
where the polarization due to the electronic contribution, Pe (ω), has been included
as well as the polarization associated with the ionic contribution, P(ω).
As is customary, the dielectric constant due to the electronic response is denoted
by
4π Pe (ω)
(∞) = 1 + , (2.25)
E(ω)
and it follows that

4π N e∗2 1 1
(ω) = (∞) + 2 + . (2.26)
(ωTO − ω2 ) (∞) m M
The so-called static dielectric constant (0) is then given by


4π N e∗2 1 1
(0) = (∞) + 2 + . (2.27)
ωTO (∞) m M
From these last two results it follows straightforwardly that
[(0) − (∞)]ωTO
2
(ω) = (∞) +
(ωTO
2 − ω2 )
(0) − (∞)
= (∞) + . (2.28)
1 − ω2 /ωTO
2
From electromagnetic theory it is known that the dielectric function (ω) must
vanish for any longitudinal electromagnetic disturbance to propagate. Accordingly,
the frequency of the LO phonons, ωLO , must be such that (ωLO ) = 0; from the last
equation, this condition implies that
(0) − (∞)
(ωLO ) = 0 = (∞) + (2.29)
1 − ωLO
2 /ω2
TO
or, equivalently,
1/2
(0)
ωLO = ωTO . (2.30)
(∞)
It then follows that
(0) − (∞) (ωLO /ωTO )2 (∞) − (∞)
(ω) = (∞) + = (∞) +
1 − ω2 /ωTO
2 1 − ω2 /ωTO
2

(ωLO /ωTO )2 − 1
= (∞) 1 +
1 − ω2 /ωTO
2

2 − ω2
ωTO ωLO
2 − ω2
TO
= (∞) + 2
ωTO
2 − ω2 ωTO − ω2
2 − ω2
ωLO
= (∞) , (2.31)
ωTO
2 − ω2
or alternatively
(ω) ω2 − ω2
= LO . (2.32)
(∞) ωTO
2 − ω2
In the special case where ω = 0, this relation reduces to the celebrated Lyddane–
Sachs–Teller relationship
(0) ω2
= LO . (2.33)
(∞) ωTO
2
When ω = ωLO the dielectric constant vanishes, (ωLO ) = 0; as stated above,

this condition is familiar from electromagnetics as a requirement for the propagation
of a longitudinal electromagnetic wave. That is, a longitudinal electromagnetic wave

propagates only at frequencies where the dielectric constant vanishes; accordingly,
ωLO is identified as the frequency of the LO phonon. From the relation

N e∗2 1 1
ωTO
2
+ 4π + = ωLO
2
,
(∞) m M
it follows that ωTO = ωLO for zone-center phonons in materials with e∗ = 0;

this is just as observed in non-polar materials such as silicon. In polar materials
such as GaAs there is a gap between ωTO and ωLO , associated with the Coulomb
energy density arising from e∗ . When ω = ωTO , (ωTO )−1 = 0 and the pole in (ω)
reflects the fact that electromagnetic waves with the frequency of the TO phonon are
absorbed. Throughout the interval (ωTO , ωLO ), (ω) is negative and electromagnetic
waves do not propagate.
2.3.4 Macroscopic theory of polar modes in cubic crystals

As was apparent in subsections 2.3.1 and 2.3.3, polar-optical phonon vibrations
produce electric fields and electric polarization fields that may be described in
terms of Maxwell’s equations and the driven-oscillator equations. Loudon (1964)
advocated a model of optical phonons based on these macroscopic fields that has
had great utility in describing the properties of optical phonons in so-called uniaxial
crystals such as würtzite crystals. The Loudon model for uniaxial crystals will be
developed more fully in Chapters 3 and 7. In this section, the concepts underlying
the Loudon model will be discussed in the context of cubic crystals.
From the pair of Maxwell’s equations,
1 ∂B 1 ∂D
∇ ×E+ =0 and ∇ ×B− = J, (2.34)
c ∂t c ∂t
it follows that
1 ∂(∇ × B) 1 ∂ 2D
∇ × (∇ × E) + = ∇(∇ · E) − ∇ 2 E + 2 2 = 0, (2.35)
c ∂t c ∂ t
where the source current, J, has been taken to equal zero. Then since ∇ · D =
∇ · E + 4π∇ · P = 4πρ = 0, it follows that
1 ∂ 2E 1 ∂ 2P
−4π∇(∇ · P) − ∇ 2 E + + 4π = 0. (2.36)
c2 ∂ 2 t c2 ∂ 2 t
Assuming that P and E both have spatial and time dependences of the form
ei(q·r−ωt) ,
this last result takes the form
−4π [q(q · P) − ω2 P/c2 ]

E= . (2.37)
q 2 − ω2 /c2
The condition q · P = 0 corresponds to the transverse wave; in this case,
−4πω2 P/c2
E= . (2.38)
q 2 − ω2 /c2
From Appendix A, E and P are also related through

1 [(0) − (∞)]ωTO 2
P= + [(∞) − 1] E; (2.39)
4π ωTO
2 − ω2
thus
q 2 − ω2 /c2 [(0) − (∞)]ωTO

2
= + [(∞) − 1], (2.40)
ω2 /c2 ωTO
2 − ω2
or, equivalently,
q 2 c2 2 (0) − ω2 (∞)
ωTO
= . (2.41)
ω2 ωTO
2 − ω2
For longitudinal waves, q · P = q P, so that q = (q/P)P, and it follows that
4π ω2 P/c2 4πq Pq
E= − 2
q 2 − ω2 /c2 q − ω2 /c2
4π ω2 P/c2 4πq 2 P
= 2 −
q − ω2 /c2 q 2 − ω2 /c2

4π ω2
= 2 −q P
2
q − ω2 /c2 c2
= −4π P. (2.42)
Then

1 [(0) − (∞)]ωTO 2
P= + [(∞) − 1] E
4π ωTO
2 − ω2

[(0) − (∞)]ωTO
2
=− + [(∞) − 1] P (2.43)
ωTO
2 − ω2
or, equivalently,
1/2
(0)
ω = ωTO = ωLO ,
(∞)
and the Lyddane–Sachs–Teller relation is recovered once again! In Chapter 3, we
shall return to the Loudon model to describe uniaxial crystals of the würtzite type.
Chapter 3
Phonons in bulk würtzite crystals
Next when I cast mine eyes and see that brave vibration, each
way free; O how that glittering taketh me.
Robert Herrick, 1648
3.1 Basic properties of phonons in würtzite structure
The GaAlN-based semiconductor structures are of great interest in the electronics

and optoelectronics communities because they possess large electronic bandgaps
suitable for fabricating semiconductor lasers with wavelengths in the blue and
ultraviolet as well as electronic devices designed to work at elevated operating
temperatures. These III-V nitrides occur in both zincblende and würtzite structures.
In this chapter, the würtzite structures will be considered rather than the zincblende
structures, since the treatment of the phonons in these würtzite structures is more
complicated than for the zincblendes. Throughout the remainder of this book,
phonon effects in nanostructures will be considered for both the zincblendes and
würtzites. This chapter focuses on the basic properties of phonons in bulk würtzite
structures as a foundation for subsequent discussions on phonons in würtzite
nanostructures.
The crystalline structure of a würtzite material is depicted in Figure 3.1. As in
the zincblendes, the bonding is tetrahedral. The würtzite structure may be generated
from the zincblende structure by rotating adjacent tetrahedra about their common
bonding axis by an angle of 60 degrees with respect to each other. As illustrated in
Figure 3.1, würtzite structures have four atoms per unit cell.
The total number of normal vibrational modes for a unit cell with s atoms in
the basis is 3s. As for cubic materials, in the long-wavelength limit there are three
acoustic modes, one longitudinal and two transverse. Thus, the total number of
optical modes in the long-wavelength limit is 3s − 3. These optical modes must,
of course, appear with a ratio of transverse to longitudinal optical modes of two.
16
3.1 Basic properties of phonons in würtzite structure 17
The numbers of the various long-wavelength modes are summarized in Table 3.1.
For the zincblende case, s = 2 and there are six modes: one LA, two TA, one
LO and two TO. For the würtzite case, s = 4 and there are 12 modes: one LA,
two TA, three LO and six TO. In the long-wavelength limit the acoustic modes are
simple translational modes. The optical modes for a würtzite structure are depicted
in Figure 3.2.
From Figure 3.2 it is clear that the A1 and E 1 modes will produce large electric
polarization fields when the bonding is ionic. Such large polarization fields result
in strong carrier–optical-phonon scattering. These phonon modes are known as
infrared active. As we shall see in Chapter 5, the fields associated with these infrared
modes may be derived from a potential describing the carrier–phonon interaction of
Figure 3.1. Unit cell of the

hexagonal würtzite crystal.
Table 3.1. Phonon modes associated with a unit cell

having s atoms in the basis.
Type of mode Number of modes

Longitudinal acoustic (LA) 1
Transverse acoustic (TA) 2
All acoustic modes 3
Longitudinal optical (LO) s−1
Transverse optical (TO) 2s − 2
All optical modes 3s − 3
All modes 3s
18 3 Phonons in bulk würtzite crystals
such modes. In Chapter 5, this carrier–phonon interaction potential will be identified

as the Fröhlich interaction. The dispersion relations for the 12 phonon modes of the
würtzite structure are depicted in Figure 3.3.
The low-frequency behavior of these modes near the point makes it apparent
that three of these 12 modes are acoustic modes. This behavior is, of course,
consistent with the number of acoustic modes identified in Table 3.1.
3.2 Loudon model of uniaxial crystals
As discussed in subsection 2.3.4, Loudon (1964) advanced a model for uniaxial

crystals that provides a useful description of the longitudinal optical phonons in
würtzite crystals. In Loudon’s model of uniaxial crystals such as GaN or AlN, the
angle between the c-axis and q is denoted by θ, and the isotropic dielectric constant
of the cubic case is replaced by dielectric constants for the directions parallel and
perpendicular to the c-axis, (ω) and ⊥ (ω) respectively. That is,
Figure 3.2. Optical phonons in würtzite structure. From Gorczyca et al. (1995),
American Physical Society, with permission.
 
⊥ (ω) 0 0
(ω) =  0 ⊥ (ω) 0  (3.1)
0 0 (ω)
with
ω2 − ωLO,⊥
2 ω2 − ωLO,
2
⊥ (ω) = ⊥ (∞) and (ω) = (∞) ,
ω2 − ωTO,⊥
2 ω2 − ωTO,
2
(3.2)
as required by the Lyddane–Sachs–Teller relation. The c-axis is frequently taken to

be in the z-direction and the dielectric constant is then sometimes labeled by the
z-coordinate; that is, (ω) = z (ω). Figure 3.4 depicts the two dielectric constants
for GaN as well as those for AlN.
In such a uniaxial crystal, there are two types of phonon wave: (a) ordinary waves
where for any θ both the electric field E and the polarization P are perpendicular
to the c-axis and q simultaneously, and (b) extraordinary waves, for which the
orientation of E and P with respect to q and the c-axis is more complicated. As
discussed in subsection 2.3.4, the ordinary wave has E 1 symmetry, is transverse,
and is polarized in the ⊥-plane. There are two extraordinary waves, one associated
with the ⊥-polarized vibrations and having A1 symmetry and the other associated
with -polarized vibrations and having E 1 symmetry. For θ = 0, one of these modes
is the A1 (LO) mode and the other is the E 1 (TO) mode. As θ varies between 0 and
π/2, these modes evolve to the A1 (TO) and E 1 (TO) modes respectively. For values
of θ intermediate between 0 and π/2 they are mixed and do not have purely LO or
[110] K M [100] [001] A

100
80
Energy (meV)
60
40
20
0
Reduced wavevector
Figure 3.3. Phonon dispersion curves for GaN crystal of würtzite structure. From
Nipko et al. (1998), American Institute of Physics, with permission.
TO character or A1 or E 1 symmetry (Loudon, 1964). For würtzite structures at the

point, it will be obvious in Chapter 7 that only three of the nine optical phonon
modes, the A1 (Z ) and E 1 (X, Y ) modes, produce significant carrier–optical-phonon
scattering rates. These are the so-called infrared-active modes. For the case of
würtzite structures, Loudon’s model of uniaxial crystals is based upon generalizing
Huang’s equations, equations (A.8) and (A.9) of Appendix A, and the relationship
of subsection 2.3.4, equation (2.43). Specifically, for each of these equations there
is a set of two more equations, one in terms of quantities along the c-axis and the
other in terms of quantities perpendicular to the c-axis:
1/2
V
(ωTO,⊥
2
− ω2 )u⊥ = ⊥ (0) − ⊥ (∞) ωTO,⊥ E⊥ , (3.3)
4π µN
1/2
V
(ωTO,
2
− ω2 )u = (0) − (∞) ωTO, E , (3.4)
4π µN
1/2
µN ⊥ (∞) − 1
P⊥ = ⊥ (0) − ⊥ (∞) ωTO,⊥ u⊥ + E⊥ , (3.5)
4π V 4π

µN 1/2 (∞) − 1
P = (0) − (∞) ωTO, u + E , (3.6)
4π V 4π
Figure 3.4. Dielectric constants for GaN, 1⊥ (GaN) and 1z (GaN), and for AlN,
2⊥ (AlN) and 2z (AlN). From Lee et al. (1998), American Physical Society, with
permission.
−4π [q⊥ (q · P) − ω2 P⊥ /c2 ]

E⊥ = , (3.7)
q 2 − ω2 /c2
−4π [q (q · P) − ω2 P /c2 ]
E = . (3.8)
q 2 − ω2 /c2
Eliminating u⊥ and u in the first four of these equations yields

1 [⊥ (0) − ⊥(∞)]ωTO,⊥
2
P⊥ = + [⊥ (∞) − 1] E⊥
4π ωTO,⊥
2 − ω2
1
= A⊥ E⊥ (3.9)
4π
and

1 [ (0) − (∞)]ωTO,
2
P = + [ (∞) − 1] E
4π ωTO,
2 − ω2
1
= A E , (3.10)
4π
where A⊥ and A may be written as
ωLO,⊥
2 − ω2
A⊥ = ⊥ (∞) − 1, (3.11)
ωTO,⊥
2 − ω2
ωLO,
2 − ω2
A = (∞) − 1, (3.12)
ωTO,
2 − ω2
upon using the Lyddane–Sachs–Teller relations

1/2 1/2
⊥ (0) (0)
ωTO,⊥ = ωLO,⊥ and ωTO, = ωLO, .
⊥ (∞) (∞)
(3.13)
For the ordinary wave, E = 0, P = 0, and q · P = 0, so that the derivation of

subsection 2.3.4 now gives
q 2 c2 ωTO,⊥
2 ⊥ (0) − ω2 ⊥ (∞)
= . (3.14)
ω2 ωTO,⊥
2 − ω2
For the ordinary mode it also follows that u = u⊥ = 0.

For the extraordinary wave, q⊥ = q sin θ and q = q cos θ, where θ is the angle
between q and the c-axis. Then, it follows that
q · P = (q sin θ, q cos θ) · (P⊥ , P ) = q P⊥ sin θ + q P cos θ. (3.15)
Thus,
q⊥ (q · P) = q 2 (P⊥ sin2 θ + P sin θ cos θ ),

(3.16)
q (q · P) = q 2 (P⊥ sin θ cos θ + P cos2 θ ).
In the limit where retardation effects are neglected, c → ∞ and it follows that
−4π [q⊥ (q · P) − ω2 P⊥ /c2 ]

E⊥ = → −4π(P⊥ sin2 θ + P sin θ cos θ)
q 2 − ω2 /c2
= − sin2 θ A⊥ E ⊥ − sin θ cos θ A E , (3.17)
and
−4π [q (q · P) − ω2 P /c2 ]
E = → −4π(P⊥ sin θ cos θ + P cos2 θ)
q 2 − ω2 /c2
= − sin θ cos θ A⊥ E ⊥ − cos2 θ A E . (3.18)
These equations may be written as

1 + sin2 θ A⊥ sin θ cos θ A E⊥
= 0, (3.19)
sin θ cos θ A⊥ 1 + cos2 θ A E
and it follows that the condition for non-trivial solutions to exist is

ωLO,⊥
2 − ω2
1 + sin2 θ A⊥ + cos2 θ A = ⊥ (∞) sin2 θ
ωTO,⊥
2 − ω2
ωLO,
2 − ω2
+ (∞) cos2 θ
ωTO,
2 − ω2
= ⊥ (ω) sin2 θ + (ω) cos2 θ = 0 (3.20)
or equivalently
⊥ (ω)q⊥
2
+ (ω)q2 = 0. (3.21)
Since the high-frequency electronic response of a medium should not depend

strongly on the crystalline structure, it is usually assumed (Loudon, 1964) that
⊥ (∞) ≈ (∞). Thus
ωLO,⊥
2 − ω2 ωLO,
2 − ω2
sin2 θ + cos2 θ = 0 (3.22)
ωTO,⊥
2 − ω2 ωTO,
2 − ω2
or equivalently
ω4 − (ω12 + ω22 )ω2 + ωTO,⊥

2
ωLO,
2
cos2 θ + ωLO,⊥
2
ωTO,
2
sin2 θ = 0,
(3.23)
where
ω12 = ωTO,
2
sin2 θ + ωTO,⊥
2
cos2 θ,
(3.24)
ω22 = ωLO,
2
cos2 θ + ωLO,⊥
2
sin2 θ.

When ωTO, − ωTO,⊥ is very much less than ωLO, − ωTO, and ωLO,⊥ − ωTO,⊥
this equation has roots
1
ω2 = (ω12 + ω22 ) ± [(ω12 − ω22 ) + 2ω2 (θ)] , (3.25)
2
where
(ωLO,
2 − ωLO,⊥
2 )(ωTO,
2 − ωTO,⊥
2 )
ω2 (θ) = 2 sin2 θ cos2 θ; (3.26)
ω22 − ω12
thus
ω2 = ωTO,
2
sin2 θ + ωTO,⊥
2
cos2 θ
(ωLO,
2 − ωLO,⊥
2 )(ωTO,
2 − ωTO,⊥
2 )
− sin2 θ cos2 θ
ω22 − ω12
≈ ωTO,
2
sin2 θ + ωTO,⊥
2
cos2 θ (3.27)
and
ω2 = ωLO,
2
cos2 θ + ωLO,⊥
2
sin2 θ
(ωLO,
2 − ωLO,⊥
2 )(ωTO,
2 − ωTO,⊥
2 )
+ sin2 θ cos2 θ
ω22 − ω12
≈ ωLO,
2
cos2 θ + ωLO,⊥
2
sin2 θ. (3.28)
3.3 Application of Loudon model to III-V nitrides

The conditions ωTO, − ωTO,⊥ ωLO, − ωTO, and ωTO, − ωTO,⊥
ωLO,⊥ − ωTO,⊥ are satisfied reasonably well for a number of würtzite materials
including the III-V nitrides. Indeed, for GaN, (∞) = 5.26, ωLO,⊥ = 743 cm−1 ,
ωLO, = 735 cm−1 , ωTO,⊥ = 561 cm−1 , and ωTO, = 533 cm−1 (Azuhata
et al., 1995). For AlN, (∞) = 5.26, ωLO,⊥ = 916 cm−1 , ωLO, = 893 cm−1 ,
ωTO,⊥ = 673 cm−1 , and ωTO, = 660 cm−1 (Perlin et al., 1993). For these and
other würtzite crystals (Hayes and Loudon, 1978), Table 3.2 summarizes the various
frequency differences appearing in the previously stated frequency conditions.
As is clear from Table 3.2, the inequalities assumed in Section 3.2 are reasonably
well satisfied for both GaN and AlN as well as for the other materials listed. The
infrared-active modes in these III-V nitrides are the A1 (LO), A1 (TO), E 1 (LO), and
E 1 (TO) modes and the frequencies associated with these modes, ω A1 (LO) , ω A1 (TO) ,
ω E 1 (LO) , and ω E 1 (TO) are given by ωLO, , ωTO, , ωLO,⊥ , and ωTO,⊥ , respectively.
Let us consider the case of GaN in more detail. From the results of Section 3.2, it
follows immediately that
E⊥ sin θ cos θ A sin θ cos θ A sin θ
=− = = , (3.29)
E 1 + sin2 θ A⊥ cos2 θ A cos θ
and
1/2
u⊥ ωTO,
2 − ω2 ⊥ (0) − ⊥ (∞) ωTO,⊥ E⊥
= 2
u ωTO,⊥ − ω2 (0) − (∞) ωTO, E
1/2
ωTO,
2 − ω2 ⊥ (0) − ⊥ (∞) ωTO,⊥ sin θ
= . (3.30)
ωTO,⊥
2 − ω2 (0) − (∞) ωTO, cos θ
Since q = (q⊥ , q ) = (q sin θ, q cos θ), the first of these relations illustrates the
fact that E q, as expected from q 2 E = −4πq(q · P); this last equality follows from
∇ ·(E+4πP) = 0. The ratio u ⊥ /u may be estimated for GaN for the transverse-like
modes, with ω2 = ωTO, 2 sin2 θ + ωTO,⊥
2 cos2 θ, as

u⊥ ⊥ (0) − ⊥ (∞) 1/2 ωTO,⊥ cos θ cos θ
=− ≈ −0.95 , (3.31)
u (0) − (∞) ωTO, sin θ sin θ
and for the longitudinal-like modes, with ω2 = ωLO,
2 cos2 θ + ωLO,⊥
2 sin2 θ, as

u⊥ ωTO,
2 − ω2 ⊥ (0) − ⊥ (∞) 1/2 ωTO,⊥ sin θ
= 2
u ωTO,⊥ − ω2 (0) − (∞) ωTO, cos θ

ωTO,
2 − ωLO2
⊥ (0) − ⊥ (∞) 1/2 ωTO,⊥ sin θ
= 2
ωTO,⊥ − ωLO 2 (0) − (∞) ωTO, cos θ
sin θ
= 1.07 , (3.32)
cos θ
Table 3.2. Difference frequencies in cm−1 for GaN and AlN as well
as for other würtzite crystals.

Würtzite ωTO, − ωTO,⊥ ωLO, − ωTO, ωLO,⊥ − ωTO,⊥
GaN 27 211 186
AlN 59 279 243
AgI 0 18 18
BeO 44 403 375
CdS 9 71 64
ZnO 33 199 178
ZnS 0 76 76
where ωLO
2 is taken to be equal to both ω2
LO, and ωLO,⊥ since ωLO, ≈ ωLO,⊥ .
2 2 2
The properties of uniaxial crystals derived in this section and in Section 3.2 will
be used extensively in Chapter 7 to determine the Fröhlich potentials in würtzite
nanostructures.
Chapter 4
Raman properties of bulk phonons
When you measure what you are speaking about and express it
in numbers, you know something about it; but when you
cannot measure it, when you cannot express it in numbers, your
knowledge is of a meagre and unsatisfactory kind; it may be the
beginning of knowledge but you have scarcely in your thoughts
advanced to the stage of science, whatever the matter may be.
Lord Kelvin, 1889
4.1 Measurements of dispersion relations for bulk

samples
This chapter deals with the application of Raman scattering techniques to measure
basic properties of phonons in dimensionally confined systems. It is, however,
appropriate at this point to emphasize that non-Raman techniques such as neutron
scattering (Waugh and Dolling, 1963) have been used for many years to determine
the phonon dispersion relations for bulk semiconductors. Indeed, for thermal
neutrons the de Broglie wavelengths are comparable to the phonon wavelengths.
For bulk samples, neutron scattering cross sections are large enough to facilitate
the measurement of phonon dispersion relations. This is generally not the case for
quantum wells, quantum wires, and quantum dots, where Raman and micro-Raman
techniques are needed to make accurate measurements of dispersion relations
in structures of such small volume. Further comparisons of neutron and Raman
scattering measurements of phonon dispersion relations are found in Section 7.5.
4.2 Raman scattering for bulk zincblende and würtzite

structures
Raman scattering has been a very effective experimental technique for observing
phonons; it involves measuring the frequency shift between the incident and
26
scattered photons. It is a three-step process: the incident photon of frequency ωi

is absorbed; the intermediate electronic state which is thus formed interacts with
phonons or other elementary excitations of energy via several mechanisms, creating
or annihilating them; finally, the scattered photon, of different energy ωs , is emitted.
Energy and momentum are conserved and are given by the following equations:
h̄ωi = h̄ωs ± h̄, (4.1)

ki = ks ± q. (4.2)
Since the momenta of the incident and scattered photons are small compared with
the reciprocal lattice vectors, only excitations with q 0 take part in the Raman
process illustrated in Figure 4.1. In the case of Raman scattering in semiconductors,
the absorption of photons gives rise to electron–hole pairs; hence the intensity of the
Raman scattering and the resonances reflect the underlying electronic structure of
the material. The Raman intensity, I (ωi ), is given by
1
I (ωi ) ∝ ωs4 |εε s Tεε i |2 (4.3)
α,β
(E α − h̄ωi )(E β − h̄ωs )
where the ωi and ωs are the frequencies of the incoming photon and of the
scattered photon respectively, E α and E β are the energies of the intermediate states,
T the Raman tensor, and ε i and ε s are the incident and scattered polarization
vectors. The summation is over all possible intermediate states. In general, for
semiconductors there may be the following real intermediate states: Bloch states,
which form the conduction or valence bands, exciton states and in-gap impurity
states. In equation (4.3), the second factor gives the Raman selection rules, which
come about from symmetry considerations of the interactions involved in a Raman
process. The selection rules are conveniently summarized in the form of Raman
tensors. These selection rules are essential tools for determining crystal orientation
and quality.
Details of the theoretical description of Raman scattering and these effects in
the vicinity of the critical points of the semiconductor are given in excellent books
and reviews elsewhere (Loudon, 1964; Hayes and Loudon, 1978; Cardona, 1975;
Cardona and Güntherodt, 1982a, b, 1984, 1989, 1991) and will not be repeated here.
Figure 4.1. Diagrammatic

representation of the Raman
process. The broken line
represents the phonon, the
wavy lines represent the
photons, and the dotted line
represents the electronic
state.
28 4 Raman properties of bulk phonons
Instead we will summarize key results in zincblende and würtzite crystals both for
the bulk case and, in Chapter 7, for quantum wells and superlattices. While first-rate
articles and book chapters exist for the results of the zincblende structures, the work
on the nitrides, with their würtzite structure, is more recent and hence in this book
we will cover the latter results in more detail.
4.2.1 Zincblende structures

The features that can be observed in a Raman experiment for particular values of
incident and scattered polarization can be determined from the symmetry properties
of the second-order susceptibility for the excitation concerned as well as from the
spatial symmetry of the scattering medium. The cubic zincblende structure has
a space-group symmetry Td2 , and there is one three-fold Raman active mode of
the T2 representation. The optic mode is polar so that the macroscopic field lifts
the degeneracy, producing a non-degenerate longitudinal mode that is at a higher
frequency than the two transverse modes. The allowed light-scattering symmetries,
as indicated by the second-order susceptibilities for the zincblende structure are
given below by appropriate matrices for the tensor T in the T2 representation:
 
0 0 0
 0 0 d  R(x) mode,
0 d 0
 
0 0 d
 0 0 0  R(y) mode, (4.4)
d 0 0
 
0 d 0
 d 0 0  R(z) mode.
0 0 0
Raman scattering has been used now for several decades as a characterization
tool in understanding, for example, crystal structure and quality, impurity content,
strain, interface disorder, and the effects of alloying and sample preparation. Much
work has been done in this class of cubic zincblende crystals since the first laser
measurements of Hobden and Russell (1964) in zincblende GaP. The prototypical
system that has been studied extensively is GaAs, and comprehensive reviews
are available (Loudon and Hayes, 1978; Cardona, 1975; Cardona and Güntherodt,
1982a, b, 1984, 1989, 1991). Frequencies of the LO and TO modes, ωLO and ωTO
respectively, for some of these systems are listed in Table 4.1.
4.2.2 Würtzite structures

In the last several years Raman scattering has also contributed a great deal to the
advances in understanding of the III-V nitride materials. The wealth of experiments
and information collected over the past 25 years on the GaAs-based material systems
is now starting to be duplicated in the nitride system, albeit somewhat slowly, as the
growth techniques and material systems continue to improve.
GaN-, AlN- and InN-based materials are highly stable in the hexagonal würtzite
structure although they can be grown in the zincblende phase and unintentional
phase separation and coexistence may occur. The würtzite crystal structure belongs
to the space group C6v 4 and group theory predicts zone-center optical modes are
A1 , 2B1 , E 1 and 2E 2 . The A1 and E 1 modes and the two E 2 modes are Raman
active while the B modes are silent. The A and E modes are polar, resulting in
a splitting of the LO and the TO modes (Hayes and Loudon, 1978). The Raman
tensors for the würtzite structure are as follows:
 
a 0 0
 0 a 0  A1 (z) mode
0 0 b
 
0 0 c
 0 0 0  E 1 (x) mode
c 0 0
  (4.5)
0 0 0
 0 0 c  E 1 (y) mode
0 c 0
  
f 0 0 0 −f 0
 0 −f 0   −f 0 0  E 2 mode.
0 0 0 0 0 0
The vibrational modes in würtzite structures are given in Figure 3.3. Details of the
frequencies are given in Table 4.2.
Table 4.1. Frequencies in cm−1 of the

LO and TO modes for zincblende crystals.
ωLO (cm−1 ) ωTO (cm−1 )

AlN 902 655
GaAs 292 269
GaN 740 554
GaP 403 367
InP 345 304
ZnS 352 271
Following some early work (Manchon et al., 1970; Lemos et al., 1972; Burns
et al., 1973) there has been a number of more recent experiments (Murugkar et al.,
1995; Cingolani et al., 1986; Azuhata et al., 1995) identifying the Raman modes
in these nitride materials. The early work was mainly on crystals in the form of
needles and platelets and the more recent work has been on epitaxial layers grown
on sapphire, on 6H-SiC, and on ZnO as well as some more unusual substrates. Table
4.2 gives the Raman modes as well as the scattering geometry in which they were
observed in the experiments of Azuhata et al. (1995). Experiments on AlN and InN
crystallites and films, particularly for the latter material, are more scarce, reflecting
the difficulties in achieving good growth qualities for these materials. In uniaxial
materials, when the long-range electrostatic field interactions of the polar phonons
dominate the short-range field of the vibrational force constants, phonons of mixed
symmetry can be observed (Loudon, 1964) under specific conditions of propagation
direction and polarization. They have been seen in the case of AlN (Bergman et al.,
1999).
4.3 Lifetimes in zincblende and würtzite crystals

Phonon–carrier interactions have an impact on semiconductor device performance
and, hence, a knowledge of the phonon lifetimes is important. Phonon lifetimes
demonstrate the effects of anharmonic interactions as well as scattering via point
defects and impurities. Anharmonic interactions (Klemens, 1958; Klemens, 1966;
Borer et al., 1971; Debernardi, 1998; Menéndez and Cardona, 1984; Ridley, 1996)
include the decay of phonons into other normal modes with the conservation of
energy and momentum. For a three-phonon decay process, a phonon of frequency
ω1 and wavevector q1 decays into two phonons of frequencies ω2 and ω3 , with
wavevectors q2 and q3 respectively, such that ω1 = ω2 + ω3 and q1 = q2 + q3 .
The investigation of the dynamical behavior of the vibrational modes provides a
direct measure of the electron–phonon interaction. The measurement of the decay
Table 4.2. Frequencies in cm−1 of the vibrational

modes in some würtzite structures.
ωAlN ωCdS ωGaN ωInN ωZnO

E 21 252 44 144 101
E 22 660 252 569 495 437
A1 (TO) 614 228 533 380
A1 (LO) 893 305 735 596 574
E 1 (TO) 673 235 561 407
E 1 (LO) 916 305 743 583
4.3 Lifetimes in zincblende and würtzite crystals 31
of the optical modes, which involves the anharmonic effects mentioned previously,
will be discussed here. Other processes that give experimental information on the
electron–phonon interaction include the generation of optical phonons by high-
energy carriers, intervalley scattering between different minima in the conduction
band, and carrier–carrier scattering; these are reported by Kash and Tsang (1991)
for the prototypical system of GaAs.
Measurements of phonon linewidths for Raman and infrared measurements in
GaAs, ZnSe, and GaP give phonon lifetimes of 2–10 ps (von der Linde, 1980;
Menéndez and Cardona, 1984). For systems that are not far from equilibrium, the
lifetimes of the phonons can be described by anharmonic processes. The decay of
an optical phonon is frequently via pairs of acoustic phonons or via one acoustic
phonon and one optical phonon of appropriate energies and momenta (Cowley,
1963; Klemens, 1966). The first measurements with continuous-wave pumping
(Shah et al., 1970) of highly non-equilibrium LO phonons in GaAs yielded estimates
of LO-phonon lifetimes of approximately 5 ps at room temperature. This was
consistent with values obtained from linewidth studies. von der Linde (1980) used
time-resolved Raman scattering to obtain directly the time decay of non-equilibrium
LO phonons. They obtained a value of 7 ps for GaAs LO phonons at 77 K.
Subsequent experiments by Kash et al. (1985) led to the conclusion that the LO
phonon lifetime in GaAs was limited by its anharmonic decay into two acoustic
phonons.
Kash et al. (1987, 1988) and Tsen and Morkoç (1988a, b) used time-resolved
Raman scattering for the alloy system AlGaAs. The results for the lifetimes are
similar to those for pure GaAs; here, though, the phonon linewidths are broadened
owing to the disorder of the alloys and these inhomogeneous broadening effects need
to be considered. Secondly, although the dispersion relations of AlAs are different
from those of GaAs there is a similarity in decay times that is interesting and
unexpected. Tsen (1992) and Tsen et al. (1989) reported on the use of time-resolved
Raman studies of non-equilibrium LO phonons in GaAs-based structures.
Tsen et al. (1996, 1997, 1998) have studied the electron–phonon interactions in
GaN of würtzite structure via picosecond and sub-picosecond Raman spectroscopy.
Results on undoped GaN with an electron density of n = 5 × 1016 cm−3 showed
that the relaxation mechanism of the hot electrons is via the emission of LO phonons
and that the Fröhlich interaction is much stronger than the deformation-potential
interaction in that material. The measured lifetime was found to be 3 ps at 300 K and
5 ps at 5–25 K (Tsen et al., 1996, 1997, 1998). The electron–LO-phonon scattering
rate was seen to be an order of magnitude larger than that for GaAs and was
attributed to the much larger ionicity in GaN. These experiments also indicated that
the longitudinal phonons decay into a TO and an LO phonon or two TO phonons.
Raman investigations of phonon lifetimes have been reported by Bergman et al.
(1999) in GaN, AlN, and ZnO würtzite crystals. These lifetimes were obtained
from measured Raman linewidths using the uncertainty relation, after correcting
for instrument broadening (Di Bartolo, 1969). These results demonstrate that the E 21
mode has a lifetime of 10 ps, an order of magnitude greater than that of the E 22 ,
E 1 (TO), A1 (TO) and A1 (LO) modes. This result was found to be true for samples
of high-quality GaN, AlN, ZnO as well as for AlN with a high level of impurities.
An explanation of the relative long lifetime of the E 22 phonons was given in terms
of factors including energy conservation constraints, density of final states, and
anharmonic interaction coefficients. The E 22 mode lies at the lowest energy of the
optical phonon modes in the würtzite dispersion curves (Nipko et al., 1998; Nipko
and Loong, 1998; Hewat, 1970) and only the acoustic phonons provide channels of
decay. At the zone edges, the acoustic phonons are equal to or larger than those of
the E 22 mode. Thus, for energy conservation to hold, the E 22 phonons have to decay
to acoustic phonons at the zone center, where their density is low.
4.4 Ternary alloys

The phonons of the ternary alloys ABx C1−x formed from the binaries AB and AC
crystals in the III-V as well as the II-VI semiconductors have been studied for some
time (Chang and Mitra, 1968). The III-nitrides have been studied more recently and
the alloys of the würtzite materials show some interesting features (Hayashi et al.,
1991; Behr et al., 1997; Cros et al., 1997; Demangeot et al., 1998; Wisniewski
et al., 1998). The ABx C1−x mixed crystals of the zincblende materials fall into
two main groups when classified according to the characteristics of the phonons.
These two classes are generally referred to as one-mode or two-mode behavior,
where ‘one-mode’ refers to the situation where the frequency of the AB phonons
gradually approaches the frequency of the AC phonons as the x-value of the alloy
increases. In the two-mode situation, the phonon frequencies are distinct and in
the limit of x = 0 (1) the AC (AB) phonon frequency is a local mode in the
AB (AC) crystal. Intermediate behavior has also been observed for certain crystals
(Lucovski and Chen, 1970). While there is no general agreement, several criteria
for phonon-mode behavior based on the mass differences of the atoms have been
proposed (Chang and Mitra, 1968). Typically, when the frequencies of the phonons
in the AB and the AC binary crystals are very different a two-mode behavior is
expected; otherwise, a one-mode behavior is seen. There is more uncertainty as well
as a smaller number of reports in the case of the würtzite nitrides. Hayashi et al.
(1991) reported studies on AlGaN würtzite films in the range 0 < x < 0.15. The
E 2 , E 1 (TO), E 1 (LO) and A1 (TO) modes were investigated and, in the composition
range studied, one-mode behavior was observed. Similar results were obtained by
Behr et al. (1997) in a narrow composition range. The E 2 mode was seen to be
unaffected by a change in composition. Cros et al. (1997) studied the AlGaN alloys
over the whole concentration range. They concluded that the E 2 mode exhibits
two-mode character, while the A1 (LO) mode is one-mode; the results for the
4.5 Coupled plasmon–phonon modes 33
A2 (TO) mode were inconclusive. Studies by Demangeot et al. (1998) concluded that
the A1 (LO), A1 (TO), and E 1 (TO) modes all exhibit one-mode behavior. However,
infrared reflectance experiments indicate that the E 1 (TO) mode displays two-mode
behavior (Wisniewski et al., 1998).
Yu et al. (1998) extended the modified random-element isodisplacement model
developed for zincblende structures by including the additional phonon modes and
the anisotropy of the würtzite structure. According to this model, the E 1 and the
A1 phonon modes should show one-mode behavior. In zincblende AlGaN crystals,
the results of Harima et al. (1999) indicate that the LO-phonon shows one-mode
behavior while the TO mode shows two-mode behavior.
Raman experiments using two ultraviolet wavelengths were performed by Alex-
son et al. (2000) on InGaN in the range 0 < x < 0.5. They investigated the A1 and
the E 2 phonons. These studies show a one-mode behavior of the A1 (LO) phonon
while the E 2 phonon demonstrates a two-mode characteristic.
The fact that the E 2 mode behaves differently from the E 1 and A1 modes is
not surprising, when one considers the specific atoms giving rise to the vibrations.
This in fact emerges from experiments on GaN würtzite films from natural GaN as
well as from GaN containing the isotope 15 N reported by Zhang et al. (1997). All
the A1 and the E 1 modes observed in the 15 N isotope were seen to shift to lower
frequencies. Niether the E 21 nor E 22 mode showed a similar shift; the E 21 mode was
essentially unaffected by the different isotopic mass. The authors thus concluded
that the E 21 vibration is due to the motion of the Ga atoms, which are heavy, alone
and, thus, that there is no frequency response to an isotropic change in the nitrogen
mass, which is considerably lighter.
4.5 Coupled plasmon–phonon modes

A coupling of the LO phonons to the plasma oscillations of the free carriers –
these oscillations are known as plasmons – occurs when an appreciable free-carrier
concentration is present in a polar semiconductor. These coupled phonon–plasmon
modes may be observed by Raman scattering, so providing information about the
free-carrier density of a given sample.
The free electrons scatter light weakly, although in solids the effect is enhanced
by band structure effects. Detailed accounts are provided by Platzman and Wolff
(1973), Yafet (1966), and Klein (1975). Mooradian and Wright (1966) made the
first observation of plasmons in n-GaAs using a Nd:YAG laser. These results
demonstrated the coupling between the LO phonon and the plasmon via the
interaction of the longitudinal electric fields produced by each of these excitations.
Other systems for which plasmon–phonon scattering has been observed include
zincblende structures such as InSb, GaP, and InAs (Hon and Faust, 1973; Patel and
Slusher, 1968). Various würtzite crystals have been investigated, for example CdS
(Scott et al., 1969), SiC (Klein et al., 1972) and, more recently, the nitrides discussed
in the next paragraph.
Here, we summarize the experimental Raman scattering results of phonon–
plasmon coupled modes in GaN, which as grown tends to be an n-type material
(Edgar, 1994); a higher carrier concentration is achieved via intentionally doping
the material. Due to this fairly high carrier concentration, the plasmons in GaN
are considered to be overdamped, similarly to those in SiC (Klein et al., 1972).
Experiments by Kozawa et al. (1998) in GaN films with a relatively low concentra-
tion of carriers show a broadening and a weakening of the intensity of the Raman
features as well as a shift to higher frequencies. The carrier concentrations obtained
by fitting the Raman lineshape as in Klein et al. (1972), Hon and Faust (1973), and
Irmer et al. (1983) give values similar to those obtained from Hall measurements
(Kozawa et al., 1998). Other experiments on GaN to study these effects at lower
free-carrier concentrations have been carried out by Wetzel et al. (1996) and Ponce
et al. (1996) and show similar results. Kirillov et al. (1996) studied the effect on
the Raman modes in the high-carrier-concentration limit. This study found that with
these higher carrier concentrations the Raman features corresponding to the upper
branch of the phonon–plasmon coupled modes are too broad to extract meaningful
information. Demangeot et al. (1997) carried out a study of the lower branch of
the coupled phonon–plasmon mode for GaN for higher carrier concentrations. The
broad Raman peak observed could be fitted by the models referenced above (Klein
et al., 1972; Hon and Faust, 1973).
Chapter 5
Occupation number representation
Oh mighty-mouthed inventor of harmonies.
Alfred, Lord Tennyson, 1863
5.1 Phonon mode amplitudes and occupation numbers
In the study of carrier–phonon interactions in nanostructures it is convenient to

use the so-called phonon-number-occupation basis. In this basis the phonon system
is modeled by the Hamiltonian for a simple harmonic oscillator. Specifically, the
familiar conjugate variables of position and momentum are replaced by creation
and annihilation operators. These creation and annihilation operators act on states
each having a given number of phonons. In particular, the creation operator acting
on a state of n q phonons of wavevector q increases the number of phonons to n q + 1
and the phonon annihilation operator acting on a state of n q phonons of wavevector
q decreases the number of phonons to n q − 1.
In some applications – such as those involving the ground state of a Bose–
Einstein condensation – the creation and annihilation operators are essential to
describing the physical properties of the system. However, in the applications
considered in this book, these operators are used merely as a convenient way of
keeping track of the number of phonons before and after a carrier–phonon scattering
event; they do not introduce new physics. Nevertheless, it is important to understand
these operators since they are used widely by the semiconductor community in
modeling the electronic and optical properties of bulk semiconductors as well as
nanostructures.
The Hamiltonian describing the harmonic oscillator associated with a phonon
mode of wavevector q is
35
36 5 Occupation number representation
pq2 1
Hq = + mωq2 u q2 , (5.1)
2m 2
where m is the mass of the oscillator, ωq is the frequency of the phonon, u q is the
displacement associated with it, and pq is its momentum. Introducing the operators,
aq and aq† ,

mωq 1
aq = uq + i pq (5.2)
2h̄ 2h̄mωq
and

mωq 1
aq† = uq − i pq , (5.3)
2h̄ 2h̄mωq
it is straightforward to show that

mωq 1 mωq 1
aq aq =
†
uq − i pq uq + i pq
2h̄ 2h̄mωq 2h̄ 2h̄mωq
1 mωq2 2 1 1 2 i
= u + p + [u q , pq ]. (5.4)
2 h̄ωq q 2m h̄ωq q 2h̄
Here the commutator [u q , pq ] ≡ u q pq − pq u q = i h̄, from the properties of the
quantum mechanical operators u q and pq . Thus

pq21 1
+ mωq u q = h̄ωq aq aq +
2 2 †
. (5.5)
2m 2 2
Since the energy of a quantum-mechanical harmonic oscillator is h̄ωq (n q + 12 ),

where n q is the number of phonons having wavevector q, it is clear that Nq =
aq† aq operating on an eigenstate of Nq phonons |Nq has eigenvalue Nq . That
is, Nq |Nq = n q |Nq . Moreover, by calculating aq aq† in the same manner used
to derive an expression for aq† aq , it follows that aq aq† − aq† aq = [aq , aq† ] = 1,
[aq , Nq ] = aq aq† aq − aq† aq aq = [aq , aq† ]aq = aq and [aq† , Nq ] = aq† aq† aq −
aq† aq aq† = aq† [aq† , aq ] = −aq† .
Accordingly,
Nq (aq† |Nq ) = (aq† Nq + aq† )|Nq = (n q + 1)aq† |Nq (5.6)
and
Nq (aq |Nq ) = (aq Nq − aq )|Nq = (n q − 1)aq |Nq . (5.7)
Thus, aq† acting on |Nq gives a new eigenstate with eigenvalue increased by 1
and aq acting on |Nq gives a new eigenstate with eigenvalue decreased by 1; that
is,

aq† |Nq = n q + 1|Nq + 1 (5.8)
and
√
aq |Nq = n q |Nq − 1, (5.9)
where the eigenvalues are consistent with the relation Nq |Nq = n q |Nq with Nq =
aq† aq . The eigenstates |Nq are orthonormal, so that
Nq ||Nq ≡ Nq |Nq = δ Nq ,Nq (5.10)
where δ Nq ,Nq is the Kronecker delta function.

Thus, the only non-zero matrix elements of aq† are those that couple states Nq |
and |Nq for which Nq = Nq + 1:

Nq |aq† |Nq = n q + 1 δ Nq ,Nq +1 . (5.11)
Likewise,
√
Nq |aq |Nq = n q δ Nq ,Nq −1 . (5.12)
The phonon occupation number n q may be determined at a temperature T from

the relation = h̄ωq (n q + 12 ), where the average energy is calculated for the
case where the eigenenergies are those of a harmonic oscillator, E n = h̄ωq (n + 12 ).
Then taking the weighting factor to be the Boltzmann factor f (E n ) = e−E n /k B T ,
we find that

= E n f (E n ) f (E n ), (5.13)
where the sums are over all n from 0 to ∞. Then

h̄ωq
= h̄ωq ne−h̄ωq n/k B T e−h̄ωq n/k B T +
2

h̄ω q
= h̄ωq nx n xn + , (5.14)
2
where x = e−h̄ωq /k B T . Then

d n h̄ωq
= h̄ωq x x xn +
dx 2
x/(1 − x)2 h̄ωq
= h̄ωq +
1/(1 − x) 2
h̄ωq x h̄ωq
= +
1−x 2
h̄ωq h̄ωq
= −h̄ω /k T + , (5.15)
e q B −1 2
and it follows that

1
nq = , (5.16)
e−h̄ωq /k B T −1
which is known as the Bose–Einstein distribution. Equations (5.10)–(5.12) and
(5.16) are used frequently in calculating carrier–phonon scattering rates in nano-
structures. Such calculations will be the subject of following chapters. In these
calculations the carrier wavefunctions and the phonon eigenstates are written as
products to express the total wavefunctions for the system of carriers and phonons.
When matrix elements are evaluated between the final and initial states of the
system, the phonon eigenstates and the phonon operators are grouped together since
they commute with the carrier wavefunctions and operations. Thus, carrier and
phonon matrix elements always appear as products and are evaluated separately.
This procedure will be illustrated in Chapters 8, 9, and 10. In Sections 5.2–5.4
we will relate the phonon displacement amplitudes to the interaction Hamiltonians
describing the dominant carrier–phonon interaction processes. In these calculations
and those of Chapters 8, 9, and 10 it will be convenient to express the normal-mode
phonon displacement u q in terms of the phonon creation and annihilation operators,
aq† and aq respectively. By adding equations (5.11) and (5.12) it will be seen readily
that

h̄
uq = (aq + aq† ). (5.17)
2mωq
In calculations of carrier–phonon scattering probabilities, the normal-mode

phonon displacement u(r) will appear in various linear forms in the Hamiltonians
entering the matrix elements being evaluated. In evaluating these matrix elements,
it will be important to keep track of not only the phonon occupation numbers n q
emerging from the phonon matrix elements but also the necessary conservation of
momentum and energy for each scattering process. In these matrix elements either
phonon creation or phonon absorption will take place but not both; therefore, as is
manifest from equations (5.11) and (5.12), only one of the two terms appearing in
equation (5.17) will contribute to the process under consideration.
u(r) is, of course, a Fourier series over the modes u q . In phonon absorption pro-
cesses the phonon appears as an incoming wave and the factor ei(q·r−ωt) multiplies
the amplitudes associated with the phonon fields. Likewise, in phonon emission
processes the phonon appears as an outgoing wave and the factor ei(−q·r−ωt) mul-
tiplies the amplitudes associated with the phonon fields. These factors are essential
in ensuring proper conservation of momentum and energy and it is convenient to
include them along with the associated creation or annihilation operator. Indeed,
they appear naturally in the Fourier decomposition of u(r).
Moreover, each incoming or outgoing phonon will be associated with a unit
polarization vector; these unit polarization vectors will be denoted by êq, j for
incoming waves and by ê∗q, j for outgoing waves. The factor e−iωt is common to
both incoming and outgoing phonons and it is generally included – along with
factors of E/h̄ associated with the carrier phases – in the integral over time that
appears in the Fermi golden rule. As a means of including the phase factors to
ensure proper accounting of energy and momentum as well as the appropriate unit
polarization vectors, (5.17) will now be written as a sum over all wavevectors q; the
appropriate non-temporal phase factors appear as multipliers of the corresponding
phonon operators:

1 h̄ iq·r
u(r) = √ aq e êq, j + aq† e−iq·r ê∗q, j
N q j=1,2,3 2mω q

1 h̄ †

=√ êq, j (aq + a−q ) eiq·r ≡ u(q) eiq·r .
N q j=1,2,3 2mωq q
(5.18)
where q is summed over all wavevectors in the Brillouin zone and N is the
number of unit cells in the sample. Chapter 7 will treat the role of dimensional
confinement in modifying the optical and acoustic phonon modes in nanostructures.
Two major modifications result: the phase space is restricted and the plane-wave
nature of the phonon is modified. Both of these effects will be treated by appropriate
modifications of equation (5.18): the changes in the phase space due to dimensional
confinement will be addressed by modifying the sum over q and the dimensional
confinement of the plane wave will be described by introducing suitable envelope
functions. Equations (5.8)–(5.18) will find many applications here and in Chapters
8, 9 and 10.

The factor h̄/2mωq in the final expression for u(r) ensures that the desired
Hamiltonian is consistent with the phonon mode amplitude. It is convenient in
practical calculations of u(r) to cast the amplitude constraints implied by [u q , pq ] =
i h̄ in the form of an integral:
h̄
∗ 1
u (r)u(r)d r =
3
√ êq · êq e−i(q −q)·r d 3 r
q q 2N m ω ω
q q
1 h̄
= , (5.19)
nm q 2ωq
where n = N /V and the integral is performed through use of the identity

−i(q −q)·r 3
e d r = V δq,q . For a single mode q, the so-called phonon normalization
condition then becomes

√ √ h̄
[ nmu∗ (r)] · [ nmu(r)]d 3 r = . (5.20)
2ωq
In terms of u(q), the normalization condition is then
√ √ h̄ 1
[ nmu∗ (q)] · [ nmu(q)] = . (5.21)
2ωq V
In the literature, quantized fields – such as the displacement field u(r) – are
expressed in terms of aq and aq† in a number of different ways. Indeed, it is possible
to perform canonical transformations that essentially exchange the roles of u q and
pq . Specifically, from the definitions of aq and aq† in terms of u q and pq , it follows
that

h̄ †
pq h̄
uq = (aq + aq ) and = −i (a − aq† ),
2mωq mωq 2mωq q
(5.22)
where the primes have been added in preparation for the following substitutions:
aq → −iaq and aq† → +iaq† . With these substitutions it follows that
pq pq
u q → and → −u q . (5.23)
mωq mωq
Thus, the canonical transformation aq → −iaq has the result that the roles of
u q and pq /mωq are interchanged. Clearly, this canonical transformation leaves
the harmonic-oscillator Hamiltonian unchanged. This is, of course, true for this
Hamiltonian whether expressed in terms of u q and pq or in terms of aq and aq† .
As a result of this canonical transformation, the quantized fields – such as the
displacement field u(r) – may be expressed in terms of aq and aq† in a number
of different ways. Indeed, the literature testifies to the fact that all the different
forms are used widely. Evidently, factors of ±i as multipliers of aq and aq† do
not change the matrix element of any such field times its complex conjugate, since
(±i)(±i)∗ = 1; thus, quantum-mechanical transition rates are not affected by this
transformation, as should clearly be the case. There is one other invariance that
is used frequently to rewrite fields expressed in terms of aq and aq† in the most
convenient form for a particular application. Namely, in expressions where a sum
or integral over positive and negative values of q is present, substitutions of the
type aq eiq·r → a−q e−iq·r leave the expression unchanged if all other multiplicative
factors contain even powers of q. Again, the quantum-mechanical transition rates
are not affected by such a change of variable.
5.2 Polar-optical phonons: Fröhlich interaction

One of the most important carrier–phonon scattering mechanisms in semiconductors
occurs when charge carriers interact with the electric polarization, P(r), produced
by the relative displacement of positive and negative ions. In low-defect polar semi-
conductors such as GaAs, InP, and GaN, carrier scattering in polar semiconductors
at room temperature is dominated by this polar–optical-phonon (POP) scattering
mechanism. The POP–carrier interaction is referred to as the Fröhlich interaction,
5.2 Polar-optical phonons: Fröhlich interaction 41
after H. Fröhlich, who formulated the first qualitatively correct formal description.
In this book, the potential energy associated with the Fröhlich interaction will be
denoted by φFr (r). Clearly the polarization P associated with polar-optical phonons
and the potential energy associated with the Fröhlich interaction, φFr (r), are related
by
∇ 2 φFr (r) = 4π e∇ · P(r). (5.24)
In terms of the phonon creation and annihilation operators of Section 5.1, P(r)
may be written as
d 3 q iq·r
P(r) = ζ aq e eq, j + aq† e−iq·r e∗q, j (5.25)
j=1,2,3
(2π ) 3
where eq, j represents the polarization vector associated with P(r) and q is the
phonon wavevector; then, it follows that
d 3 q iq·r
4π∇ · P(r) = 4πiζζ aq e q · eq, j − aq† e−iq·r q · e∗q, j .
j=1,2,3
(2π ) 3
(5.26)
Consider the case of a polar crystal with two atoms per unit cell, such as GaAs.
Clearly, the dominant contribution to P(r) results from the phonon modes in which
the normal distance between the planes of positive and negative charge varies. Such
modes are obviously the LO modes since in the case of LO modes eq, j is parallel
to q. However, TO phonons produce displacements of the planes of charge such
that they remain at fixed distances from each other; that is, the charge planes ‘slide’
by each other but the normal distance between planes of opposite charge does not
change. So, TO modes make negligible contributions to P(r). For TO phonons,
eq, j · q = 0. Accordingly,

d 3 q iq·r
4π ∇ · P(r) = 4πiζζ aq e q − aq† e−iq·r q , (5.27)
(2π ) 3
and the potential energy associated with the Fröhlich interaction, φFr (r), is given by

d 3 q 1 iq·r
HFr = φFr (r) = −4πieζζ aq e − aq† e−iq·r , (5.28)
(2π ) 3 q
where φFr (r), has been denoted by HFr , the Fröhlich interaction Hamiltonian since
φFr (r) is the only term contributing to it.
The dependence of φFr (r) on q −1 is familiar from the Coulomb interaction; the
coupling constant ζ remains to be determined.
From subsection 2.3.3, the electric polarization P(r) may be written as
N e∗
P(r) = uq (r)
(∞)

N e∗ 1 h̄
= √
(∞) N q j=1,2,3 mM
2 ωLO
m+M

× aq eiq·r eq, j + aq† e−iq·r e∗q, j ), (5.29)
where the division by (∞) accounts for screening, and the normal-mode expression
(5.18) has been used for u(r). By noticing that

1 d 3q
√ ↔ (5.30)
V q (2π )3
and by comparing expressions (5.25) and (5.29), it follows that

N e∗ 1 h̄
ζ = √ . (5.31)
(∞) N mM
2 ωLO
m+M
However, from (2.26) evaluated for ω = ω0 , it follows that

N e∗2 1 1 ωLO
2 − ω2
TO
+ =
(∞)2 m M 4π (∞)
2
ω2 1 1 ωTO
= LO −
4π (∞) (∞) ωLO 2

ω2 1 1
= LO − , (5.32)
4π (∞) (0)
so that

2
N e∗ 1
h̄ h̄ ωLO 1 1
ζ = √ = − ,
(∞) N mM 2ωLO 4π (∞) (0)
2 ωLO
m+M
(5.33)
and

d 3 q 1 iq·r
HFr = φFr (r) = −4πieζζ aq e − aq† e−iq·r
(2π ) q
3

2π e2 h̄ωLO 1 1 1 iq·r
= −i − aq e − aq† e−iq·r . (5.34)
V (∞) (0) q q
5.3 Acoustic phonons and deformation-potential interaction 43
5.3 Acoustic phonons and deformation-potential

interaction
The deformation-potential interaction arises from local changes in the crystal’s en-
ergy bands arising from the lattice distortion created by a phonon. The deformation-
potential interaction, introduced by Bardeen and Shockley, is one of the most
important interactions in modern semiconductor devices and it has its origin in
the displacements caused by phonons. Indeed, the displacements associated with
a phonon set up a strain field in the crystal. In the simple case of a one-dimensional
lattice, the energy of the conduction band, E c , or the energy of the valence band,
E v , will change by an amount
E c,v = E c,v (a) − E c,v (a + u), (5.35)
where a is the lattice constant and u is the displacement produced by the phonon
mode. Since a u, it follows that
E c,v (a) = (d E c,v (a)/da)u. (5.36)
Thus the phonon displacement field u produces a local change in the band energy;
the energy associated with the change is known as the deformation potential and
it represents one of the major scattering mechanisms in non-polar semiconductors.
Indeed, the deformation-potential interaction is a dominant source of electron energy
loss in silicon-based electronic devices. The three-dimensional generalization of
E c,v is
E c,v (a) = (d E c,v (a)/d V )V, (5.37)
where V is a volume element and V is the change in the volume element due to
the phonon field. For an isotropic medium V /V = ∇ · u and the last expression
becomes,
E c,v (a) = V (d E c,v (a)/d V )∇ · u, (5.38)
which is usually written as
c,v
Hdef = E c,v (a) = E 1c,v ∇ · u. (5.39)
c,v
The superscripts on Hdef and E 1c,vare necessary since the deformation potential
for electrons is different from that for holes. Chapter 9 provides a discussion of the
case where the medium is not assumed to be isotropic.
5.4 Piezoelectric interaction

The piezoelectric interaction occurs in all polar crystals lacking an inversion
symmetry. In the general case, the application of an external strain to a piezoelectric
crystal will produce a macroscopic polarization as a result of the displacements

of ions. Thus an acoustic phonon mode will drive a macroscopic polarization in
a piezoelectric crystal. In rectangular coordinates, the polarization created by the
piezoelectric interaction in cubic crystals, including zincblende crystals, may be
written as

P = 12 ex4 (∂w/∂ y + ∂v/∂z), 12 ex4 (∂u/∂z + ∂w/∂ x), 12 ex4 (∂u/∂ y + ∂v/∂ x) ,
(5.40)
where ex4 is the piezoelectric coupling constant and, as will be described in

Section 7.2, the factors multiplying ex4 are the components of the strain tensor
that contribute to the piezoelectric polarization in a zincblende crystal. As with
the Fröhlich and deformation-potential interactions, phonons play an essential role
in producing piezoelectric interactions. Piezoelectric interactions will be discussed
further in Chapter 9.
Chapter 6
Anharmonic coupling of phonons

With a name like yours, you might be any shape almost.
Lewis Carroll, Through the Looking Glass, 1872
6.1 Non-parabolic terms in the crystal potential for

ionically bonded atoms
The crystal potential may be expanded in powers of the displacements of the ions
from their equilibrium positions to yield a sum over quadratic and higher-order
terms. The quadratic terms, of course, represent the harmonic modes considered
at length in Chapter 5. The cubic and higher-order terms, containing products of
three or more displacements, are generally known as the anharmonic terms. These
anharmonic terms lead to modifications of the harmonic modes in the quadratic
approximation of the harmonic oscillator. Indeed, while the harmonic approximation
may be used to describe the phonon dispersion relations it cannot describe the decay
of phonon modes: the anharmonic interaction is necessary to describe the decay of
a phonon into other phonons. The leading term in the anharmonic interaction is the
cubic term and it may be written as
1
Hq, j;q , j ;q , j = √ P(q, j; q , j ; q , j )uq, j uq , j uq , j , (6.1)
N
where q, q , and q are the phonon wavevectors and j, j , and j designate the
polarizations of the phonon modes participating in the anharmonic interaction. N
is the number of unit cells in the crystal, and P(q, j; q , j ; q , j ) describes the
anharmonic coupling. The normal process where q − q − q = 0 is taken into
account in this analysis but the umklapp process where q − q − q = b = 0
is ignored since the normal process is expected to dominate for small phonon
wavevectors. From Section 5.1, the displacement is given by
45
46 6 Anharmonic coupling of phonons

1 h̄ iq·r
u(r) = √ aq e êq, j + aq† e−iq·r ê∗q, j
N q j=1,2,3 2mωq, j
1
=√ uq, j (6.2)
N q j=1,2,3
and, accordingly, the normal-mode coordinates uq, j are

h̄ iq·r
uq, j = aq e êq, j + aq† e−iq·r ê∗q, j , (6.3)
2mωq, j
where, as defined in Chapter 5, the unit polarization vectors are denoted by êq, j
for incoming waves and by ê∗q, j for outgoing waves, and the phonon creation and
annihilation operators are given by aq† and aq , respectively.
6.2 Klemens’ channel for the decay process LO → LA(1)

+ LA(2)
Klemens (1966) considered the anharmonic decay of an optical phonon into two
longitudinal acoustic phonons, LA(1) and LA(2). The Klemens channel, LO →
LA(1) + LA(2), is now recognized as the dominant LO phonon decay channel in
a number of cubic crystals including GaAs (Tsen, 1992). For the Klemens channel,

qLO = qLA(1) + qLA(2) . (6.4)
Writing q = qLO , q = qLA(1)

, and q = qLA(2) , the anharmonic interaction is
1/2
1 3 h̄ 1
Hq, j;q , j ;q , j = √ P(q, q , q )aq aq† aq† ei(q−q −q ) .
N 2m ω ω
q q qω
(6.5)
The Klemens process is depicted in Figure 6.1 for bulk GaAs.
Figure 6.1. Decay of a

near-zone-center LO phonon
into two acoustic phonons
for idealized dispersion
curves (thick lines). The
arrows begin at the energy
and wavevector of the initial
LO phonon and terminate at
the final state LA phonon
energies and wavevectors.
Momentum and energy are
both conserved in this
process.
6.3 LO phonon lifetime in bulk cubic materials 47
To determine P(q, q , q ) it is necessary to characterize the anharmonic cou-

pling. Direct measurement of the anharmonic potential is difficult and few ab initio
calculations have been undertaken (Tua and Mahan, 1982; Tua, 1981). Keating
(1966) used the theory of elasticity and exploited the relationship between the
third-order elastic coefficients and the anharmonic terms of the crystal potential to
determine P(q, q , q ).
The theory of elasticity will be discussed further in Chapter 7 as the basis
for a continuum model for acoustic phonons. Bhatt et al. (1994) used Keating’s
approach to estimate P(q, q , q ) as the strain energy density Ua associated with
the anharmonic third-order potential. They gave expressions for Ua in terms of the
third-order elastic constants and the linear components of the strain variables. Ua
is related to a Grüneisen constant and is treated frequently as a parameter. Herein,
P(q, q , q ) will be denoted as Ua .
6.3 LO phonon lifetime in bulk cubic materials

For the Klemens channel, LO → LA(1) + LA(2), the Fermi golden rule predicts a
transition rate
2π
= |M|2 δ(h̄ωq − h̄ωq − h̄ωq ), (6.6)
h̄ q ,q
where the matrix element satisfies

h̄ 3 Ua3 1
|M| =
2
n q (n q + 1)(n q + 1)δq,q +q (6.7)
8N m 3 ωq ωq ωq
and, as derived in Section 5.1, the Bose–Einstein occupation number is given by
1
nq = . (6.8)
e−h̄ωq /k B T − 1
To calculate the net rate of phonon decay through the Klemens channel it is
necessary to consider not only the decay rate but also the generation rate for the
reverse process, G. The total loss rate for the LO phonons is then given by − + G.
Writing as n q (n q + 1)(n q + 1) it follows that G is given by (n q + 1)n q n q
and the inverse lifetime takes the form (Bhatt et al., 1994; Klemens, 1966; Ferry,
1991)
1 π h̄ 2 U 3 1

a
= (1 + n q + n q )
τ q ,q
4N m 3 ωq ωq ωq
× δq,q +q δ(h̄ωq − h̄ωq − h̄ωq ). (6.9)
In the general case the sum over final states must include a sum over j and j
as well. Bhatt et al. (1994) applied this result to calculate the lifetime of the bulk
LO phonon in GaAs as a function of temperature and as a function of phonon
wavevector along the 100 and 111 directions at selected temperatures; these
results are shown in Figures 6.2 and 6.3.
6.4 Phonon lifetime effects in carrier relaxation

The energy loss rate of a carrier in a polar semiconductor is determined by both
the rate at which the carrier’s energy is lost by phonon emission and the rate at
which the carrier gains energy from phonon absorption. This latter rate can be
significant in dimensionally confined structures – such as quantum-well lasers –
since the phonons emitted by energetic carriers can accumulate in these structures.
Indeed, the phonon densities in dimensionally confined semiconductor devices may
well be above those of the equilibrium phonon population and there is a large
probability that these non-equilibrium phonons will be reabsorbed. The net loss
of energy by a carrier – also known as the carrier relaxation rate – depends on
Figure 6.2. LO phonon lifetime in GaAs as a function of temperature. From Bhatt

et al. (1994), American Institute of Physics, with permission.
6.4 Phonon lifetime effects in carrier relaxation 49
the rates for both phonon absorption and emission. Clearly, the lifetimes of the
optical phonons are important in determining the total energy loss rate for such
carriers. As discussed in the last section, the LO phonons in GaAs and many
other polar materials decay into acoustic phonons through the Klemens channel.
Over a wide range of temperatures and phonon wavevectors, the lifetimes of LO
phonons in GaAs vary from a few picoseconds to about 10 ps (Bhatt et al., 1994), as
illustrated by Figures 6.2 and 6.3. Longitudinal-optical phonon lifetimes for other
polar semiconductors are also of this magnitude. As a result of the Klemens channel,
the ‘hot’ LO phonons decay into acoustic phonons in times of the order of 10 ps.
These LO phonons undergoing decay into acoustic phonons are not available for
absorption by the carriers and, as a result of the Klemens channel, the carrier energy
relaxation rate is influenced strongly by the LO phonon lifetime. Indeed, it follows
that the time required to modulate a semiconductor quantum-well laser through
the switching of the electronic current being used to pump the laser is limited by
the time it takes the carrier to thermalize in the active quantum-well region of the
laser.
Figure 6.3. Wavevector dependence of the lifetime of bulk LO phonons in GaAs for
wavevectors along the 100 (squares) and 111 (circles) directions at 77 K (upper
set of points) and 300 K (lower set of points). From Bhatt et al. (1994), American
Institute of Physics, with permission.
Moreover, this thermalization time of carriers in a polar semiconductor quantum

well such as in GaAs is determined by the LO phonon emission rate minus the
LO phonon absorption rate. Since there is generally a large probability that such
non-equilibium LO phonons in the quantum well will be reabsorbed, the net rate of
carrier energy loss is reduced considerably from the thermalization rate that could
be achieved if the carrier lost energy through the sequential emission of LO phonons
without reabsorption. This effect is referred to frequently as the phonon bottleneck.
In fact, the thermalization times of semiconductor quantum-well lasers are generally
tens of picoseconds – in accord with the LO phonon lifetime – and not the fractions
of picoseconds characteristic of the carrier–LO-phonon emission times for many
polar semiconductors over a wide range of carrier energies. To reduce the phonon-
bottleneck effect in quantum-well lasers, Zhang et al. (1996) conceived the tunneling
injection laser, in which the carrier tunnels into the active region of the quantum-well
laser with energies only one or two times the LO-phonon energy. As a result of
this ingenious design, the carriers can thermalize by emitting only one or two LO
phonons and the probability of significant LO phonon absorption is reduced greatly.
These tunneling injection lasers have intrinsic switching speeds close to 100 GHz,
which is several times faster than for the semiconductor quantum-well lasers, where
the carriers have to be captured and thermalize in quantum wells having well depths
several times the LO phonon energy. The phonon-bottleneck effect will be discussed
further in subsection 8.3.4 in connection with carrier energy loss in quantum wires
and in Section 10.1 on intersubband lasers.
6.5 Anharmonic effects in würtzite structures: the Ridley

channel
Longitudinal-optical phonon decay in GaN cannot proceed by the Klemens’ channel

since the LO phonon energy is more than twice the energy of any of the available
acoustic phonons. Ridley (1996) observed that a four-phonon decay process would
be much slower than the rates typical of polar semiconductors and proposed that a
three-phonon decay channel where the LO phonon decays into a TO phonon and an
LA phonon is possible. The Ridley channel, LO → TO + LA, competes with the
channel LO → TO + TA but it is expected that the Ridley channel will dominate,
owing to symmetry considerations (Ridley, 1996). Through an analysis similar to
that presented in Sections 6.1–6.3, Ridley showed that the net rate of annihilation of
LO phonons via the Ridley channel in GaN is given by
1 2
= h̄ωTO (ωLO − ωTO )3
τGaN 2πρLO ωLO vLA
3
× [n(ωTO ) + n(ωLO − ωTO ) + 1], (6.10)

6.5 Anharmonic effects in würtzite structures: the Ridley channel 51
where ρLO is the reduced-mass density, vLA is the group velocity, and qLA is the LA
wavevector associated with the LA frequency ωLO −ωTO ; see also Ridley and Gupta
(1991). By taking ≈ 108 cm−1 as for GaAs, Ridley obtained a zero-temperature
rate of 5.0 × 1011 s−1 . A full understanding of phonon decay in the wide-bandgap
III-V nitride materials is yet to be obtained. However, it is likely that the optical-
phonon lifetimes in the III-V nitrides will play an important role in the design of
electronic and optoelectronic devices fabricated from them. The Raman analysis of
anharmonic phonon decay rates given in Chapter 4 provides essential information
for the systematic design of such devices.
Chapter 7
Continuum models for phonons

There is not a single thing, however small, in the world that
does not depend on something that is higher . . . for everything
is interdependent.
Zohar, Kabbalah, Book 1, 156b
7.1 Dielectric continuum model of phonons

The dielectric continuum model of optical phonons in polar materials is based on
the concept that the associated lattice vibrations produce an electric polarization
P(r) that is describable in terms of the equations of electrostatics for a medium
of dielectric constant (ω) (Fuchs and Kliewer, 1965, 1966a, b, c; Engelman and
Ruppin, 1968a, b, c; Ruppin and Engelman, 1970; Licari and Evrard, 1977; Wendler,
1985; Mori and Ando, 1989). The volume of the structure is assumed to be L 3
(−L/2 ≤ x, y, z ≤ +L/2) with periodic boundary conditions. The potential (r)
associated with P(r) is given by (Kim and Stroscio, 1990)
∇ 2 (r) = 4π∇ · P(r) (7.1)
and the electric field E(r) is given by
E(r) = −∇(r). (7.2)
Moreover, E(r) and P(r), in medium n are related through the dielectric suscepti-
bility, n (ω):
P(r) = n (ω)E(r) (7.3)
where
n (ω) = [n (ω) − 1]/4π. (7.4)
Using the results of Chapter 5, the Lyddane–Sachs–Teller relation for medium n

may be written as
52
ω2 − ωLO,n
2
n (ω) = n (∞) , (7.5)
ω2 − ωTO,n
2
for a binary polar semiconductor AB and as
ω2 − ωLO,n,a
2 ω2 − ωLO,n,b
2
n (ω) = n (∞) , (7.6)
ω2 − ωTO,n,a
2 ω2 − ωTO,n,b
2
for a ternary polar material A y B1−y C, where the subscript a denotes frequencies
associated with the dipole pairs AC and the subscript b denotes frequencies
associated with the dipole pairs BC. As in subsection 2.3.1, the displacement field is
related to the fields E(r) and P(r) through the driven oscillator equation and through
the effective charge, en∗ : for a binary medium n,
−µn ω2 un (r) = −µn ω0n

2
un (r) + en∗ Elocal (r),
(7.7)
P(r) = nn en∗ un (r) + nn αn Elocal (r),
where nn is the number of unit cells in region n, µn = m n Mn /(m n + Mn ) is the

reduced mass and αn is the electronic polarizability per unit cell and where, by the
Lorentz relation,
4π
Elocal (r) = E(r) + P(r). (7.8)
3
Within the virtual-crystal model, for the dipole pairs AC (BC) in a ternary medium
m, we have
∗
−µm,a(b) ω2 um,a(b) (r) = −µm ω0m,a(b)
2
um,a(b) (r) + em,a(b) Elocal (r),
(7.9)
∗ ∗
P(r) = n m [yem,a um,a (r) + (1 − y)em,b um,b (r)] + n m αm Elocal (r).
An alternative and useful form of these equations for the case of a binary material
results straightforwardly from the relations of Appendix A. Indeed, it is shown in
Appendix A, equations (A.8) and (A.9), for a diatomic polar material that
1/2
V
ü = −ωTO
2
u+ (0) − (∞) ωTO E,
4π µN
(7.10)
µN 1/2 (∞) − 1
P= (0) − (∞) ωTO u + E.
4π V 4π
In the first equation it has been assumed that ü has a general form for the
time dependence and may not be simply sinusoidal in ω. As will become evident,
this pair of equations is well suited as the basis for an alternative method for
performing the calculations of subsection 7.3.1. Moreover, it provides a convenient
starting point for the derivation of the macroscopic equations describing optical
phonons in polar uniaxial materials (Loudon, 1964). Uniaxial materials such as
54 7 Continuum models for phonons
the hexagonal würtzite structures GaN, AlN, and Gax Al1−x N have relatively wide
bandgaps and are suited for high-temperature electronics and short-wavelength
optoelectronic devices. Loudon (1964) introduced a useful model for describing
the macroscopic equations of a uniaxial polar crystal by introducing one dielectric
constant associated with the direction parallel to the c-axis, , and another dielectric
constant associated with the direction perpendicular to the c-axis, ⊥ . In Loudon’s
model a separate set of Huang–Born equations is necessary for the phonon mode
displacements parallel to the c-axis, u , and perpendicular to it, u⊥ . For a medium
denoted by n it then follows straightforwardly that

V
ü⊥,n = −ωTO,⊥,n u⊥,n +
2
(0)⊥,n − (∞)⊥,n ωTO,⊥,n E⊥,n ,
4π µn N

µn N (∞)⊥,n − 1
P⊥,n = (0)⊥,n − (∞)⊥,n ωTO,⊥,n u⊥,n + E⊥,n ,
4π V 4π
2
ω − ωLO,⊥,n
2
⊥,n (ω) = ⊥,n (∞) , (7.11)
ω2 − ωTO,⊥,n
2
and

V
ü,n = −ωTO,,n u,n +
2
(0),n − (∞),n ωTO,,n E,n ,
4π µn N

µn N (∞),n − 1
P,n = (0),n − (∞),n ωTO,,n u,n + E,n ,
4π V 4π
2
ω − ωLO,,n
2
,n (ω) = ,n (∞) . (7.12)
ω2 − ωTO,,n
2
Of course, in Loudon’s model these six equations must be supplemented by the

following three equations of electrostatics for the case where there is no free charge:
E(r) = −∇φ(r),
D(r) = E(r) + 4π P(r)
(7.13)
ρ + (ω)E (r)ẑ,
= ⊥ (ω)E ⊥ (r)ρ̂
∇ · D(r) = 0,
ρ are the unit vectors in the and ⊥ directions respectively.

where ẑ and ρ̂
In the first and third of this set of nine equations, it has been assumed that ü⊥,n
and ü,n have a general form of the dependence and may not be simply sinusoidal
in ω; the assumption of sinusoidal time dependence made in Appendix A was not
necessary and simply by replacing −ω2 u by ü it is straightforward to rederive the
results of Appendix A without assuming a sinusoidal time dependence.
The above set of nine equations provides a convenient basis for describing
carrier–optical-phonon scattering in würtzite crystals. Indeed, using the relations for
the displacement perpendicular (parallel) to the c-axis,

1 h̄ †
u(r)⊥() =√ êq, j,⊥() (aq + a−q ) eiq·r , (7.14)
N q j=1,2,3 2mωq
which follow from the results of Section 5.1, and defining

φ(r)⊥() = φ(q)⊥() eiq·r (7.15)
q

E(r)⊥() = −∇φ(r)⊥() = −iq φ(q)⊥() eiq·r , (7.16)
q
it follows, by taking m = µn , ω2 = ωq2 and assuming a sinusoidal dependence for

u(r, t), that (Lee et al., 1997)

h̄ †
(ωTO,⊥(),n − ωq )
2 2
êq, j,⊥() (aq + a−q )
2µn N ωq

V
= (0)⊥(),n − (∞)⊥(),n ωTO,⊥(),n (−i)q⊥() φ(q),
4π µn N
(7.17)
with q⊥ = q sin θ, where θ is the angle between q and the c-axis, which is taken as
the z-axis. Moreover, q = q cos θ and ê2q, j,⊥ + ê2q, j, = 1. Hence

2π h̄ †
êq, j,⊥() (aq + a−q )
V ωq

(0)⊥(),n − (∞)⊥(),n ωTO,⊥(),n
= (−i)q⊥() φ(q), (7.18)
ωTO,⊥(),n
2 − ωq2
so that

2π h̄ † 2
(aq + a−q )
V ωq

[(0)⊥ − (∞)⊥ ] ωTO,⊥ 2
= −q 2
sin θ φ 2 (q),
2
(ω⊥,TO
2 − ωq2 )2

[(0) − (∞) ] ωTO,
2
−q 2
cos θ φ 2 (q),
2
(7.19)
(ωTO,
2 − ωq2 )2
and

2π h̄ †
φ(q) = −i (aq + a−q )(ωTO,⊥
2
− ωq2 )(ωTO,
2
− ωq2 )
V q 2 ωq

× [(0)⊥ − (∞)⊥ ]ωTO,⊥
2
(ωTO,
2
− ωq2 )2 sin2 θ
−1/2
+ [(0) − (∞) ]ωTO,
2
(ωTO,⊥
2
− ωq2 )2 cos2 θ . (7.20)
Thus
†
H= (−e)φ(q) eiq·r (aq + a−q )
q

2π e2 h̄ 1 iq·r †
=i e (aq + a−q )(ω⊥,TO
2
− ωq2 )(ωTO,
2
− ωq2 )
q V ωq q

× [(0)⊥ − (∞)⊥ ]ωTO,⊥
2
(ωTO,
2
− ωq2 )2 sin2 θ
−1/2
+ [(0) − (∞) ]ωTO,
2
(ω⊥,TO
2
− ωq2 )2 cos2 θ
1/2
4π e2 h̄V −1 1 iq·r †
=i e (aq + a−q ).
q (∂/∂ω)[(ω)⊥ sin θ + (ω) cos θ]
2 2 q
(7.21)
In the isotropic case, (ω)⊥ = (ω) and this result must reduce to the
expression, (5.34), obtained in Section 5.2 for the interaction Hamiltonian de-
scribing the carrier–LO-phonon interactions. Indeed, since the general form of the
Lyddane–Sachs–Teller relation implies that

2 − ω2
ωTO LO 1 1 1/2
√ = −ωLO − , (7.22)
ωTO (0) − (∞) (∞) (0)
the Hamiltonian for the uniaxial case reduces to (5.34) upon taking (ω)⊥ = (ω)
and ωq = ωLO .
7.2 Elastic continuum model of phonons

As will become clear, the elastic continuum model of acoustic phonons provides
an adequate description of acoustic phonons for nanostructures having confined
dimensions of about two atomic monolayers. A simple and illustrative application
of the elastic continuum model is found in the case of a longitudinal acoustic mode
propagating in a quasi-one-dimensional structure. Consider an element d x located
along this structure between x and x +d x. Let u(x, t) be the elastic displacement at x
along the axis of the one-dimensional structure; that is, u(x, t) describes the uniform
longitudinal displacement of the element d x. In the elastic continuum model the
dynamics of the mass-containing element, d x, are described in terms of Newton’s
laws. Indeed, defining the strain as e = du/d x and the stress, T (x), as the force per
unit area in the quasi-one-dimensional structure of area A, it follows from Hooke’s
law that
T = Y e, (7.23)
where Y is a proportionality constant known as Young’s modulus. The force

equation describing the dynamics of the element d x of density ρ(x) is given by
Newton’s second law:

∂ 2 u(x, t)
ρ(x)A d x = [T (x + d x) − T (x)]A, (7.24)
∂t 2
where ρ A d x is the mass associated with the element d x and ∂ 2 u/∂t 2 . By Hooke’s
law
2
∂T ∂e ∂ u
T (x + d x) − T (x) = dx = Y d x = Y 2 d x, (7.25)
∂x ∂x ∂x
and it follows that

∂ 2u ρ(x) ∂ 2 u
= . (7.26)
∂x2 Y ∂t 2
Seeking solutions of this one-dimensional wave equation of the form u(x) =
ξ ei(q x−ωt) ,where q = 2π/λ and ω is the angular frequency of the wave, it follows
that the dispersion relation for the longitudinal acoustic (LA) mode is ρω2 = Y q 2
√
or ω = vl q, where vl = Y /ρ. The longitudinal sound speed, vl , has typical values
(3–5) × 105 cm s−1 and for ρ = 4 g cm−3 it follows that Y must have an order of
magnitude of 1012 g cm s−2 .
The three-dimensional generalization of these results may be accomplished
through the replacements (Auld, 1973) u(x) → u(x, y, z) = (u, v, w) and T =
Y e → T = c : S with Ti = ci j Sj . In this generalization, Young’s modulus is
replaced by a 6 × 6 matrix of elastic constants ci j ; T is replaced by a six-component
object Ti ; e is replaced by a six-component object Sj . For the cubic, zincblende, and
würtzite crystals the most general form of the stress–strain relation, Ti j = ci jkl Skl ,
where i, j, l, k run over x, y, z, may be represented by Ti = ci j Sj . In this last result,
i and j run over the integers 1–6; 1 ≡ x x, 2 ≡ yy, 3 ≡ zz, 4 ≡ yz or zy, 5 ≡ x z or
zx, and 6 ≡ x y or yx. The resulting forms for Sj are
∂u ∂v ∂w
S1 = Sx x = , S2 = S yy = , S3 = Szz = ,
∂x ∂y ∂z

1 ∂w ∂v
S4 = S yz = Szy = + ,
2 ∂y ∂z
(7.27)
1 ∂u ∂w
S5 = Sx z = Szx = + ,
2 ∂z ∂x

1 ∂u ∂v
S6 = Sx y = S yx = + .
2 ∂y ∂x
For Ti the forms are
T1 = Tx x , T2 = Tyy , T3 = Tzz ,
(7.28)
T4 = Tyz = Tzy , T5 = Tx z = Tzx , T6 = Tx y = Tyx.
For the elastic energy to be single valued ci j = c ji , and it follows that only
21 distinct elements are necessary to define the 6 × 6 matrix ci j . Nanostructures
of widespread interest in modern electronics and optoelectronics are generally

fabricated from zincblende and würtzite crystals. For cubic crystals, including
zincblende crystals, the matrix ci j is of the form
 
c11 c12 c12 0 0 0
 c 0 
 12 c11 c12 0 0 
 
 c12 c12 c11 0 0 0 
  (7.29)
 0 0 0 c44 0 0 
 
 0 0 0 0 c44 0 
0 0 0 0 0 c44
and for würtzite crystals ci j is of the form
 
c11 c12 c13 0 0 0
 c 
 12 c11 c13 0 0 0 
 
 c13 c13 c33 0 0 0 
 . (7.30)
 0 0 0 c44 0 0 
 
 0 0 0 0 c44 0 
0 0 0 0 0 (c11 − c12 )/2
For a cubic medium such as the zincblende crystal, only three independent elastic
constants, c11 , c12 , and c44 , are needed to specify all the ci j . For an isotropic cubic
medium c12 = c11 − 2c44 , and only two constants λ and µ are necessary to define
the ci j :
λ = c12 = c13 = c21 = c23 = c31 = c32 ,
µ = c44 = c55 = c66 = 12 (c11 − c12 ), (7.31)

λ + 2µ = c11 = c22 = c33 .
The constants λ and µ are known as Lamé’s constants. Thus, in the cubic case three
independent constants replace Y : c11 , which relates the compressive stress to the
strain along the same direction, [100]; c44 , which relates the shear stress and the
strain in the same direction; and c12 , which relates the compressive stress in one
direction and the strain in another direction.
For the isotropic case, it follows that
Tx x = λ(Sx x + S yy + Szz ) + 2µSx x = λ + 2µSx x ,

Tyy = λ(Sx x + S yy + Szz ) + 2µS yy = λ + 2µS yy ,
(7.32)
Tzz = λ(Sx x + S yy + Szz ) + 2µSzz = λ + 2µSzz ,
Tyz = µS yz , Tzx = µSzx , Tx y = µSx y ,
where = ∂u/∂ x + ∂v/∂ y + ∂w/∂z represents the dilatation of the medium.
Then, it is straightforward to show that the three-dimensional generalization of
ρ(x)∂ 2 u/∂t 2 = Y ∂ 2 u/∂ x 2 = ∂ T /∂ x is given by the equations
∂ 2u ∂ Tx x ∂ Tyx ∂ Tzx ∂
ρ = + + = (λ + µ) + µ∇ 2 u,
∂t 2 ∂x ∂y ∂z ∂x
∂ 2v ∂ Tx y ∂ Tyy ∂ Tzy ∂
ρ 2 = + + = (λ + µ) + µ∇ 2 v, (7.33)
∂t ∂x ∂y ∂z ∂y
∂ 2w ∂ Tx z ∂ Tyz ∂ Tzz ∂
ρ 2 = + + = (λ + µ) + µ∇ 2 w,
∂t ∂x ∂y ∂z ∂z
where, as usual,
∂2 ∂2 ∂2
∇2 ≡ + + . (7.34)
∂x2 ∂ y2 ∂z 2
Two alternative forms of the three-dimensional force equations are encountered
frequently in the literature. The first of these is derived by writing the components
of u(x, y, z) = (u 1 , u 2 , u 3 ) as u α , α = 1, 2, 3; it then follows that the three force
equations may be rewritten as
∂ 2uα ∂ Tαβ
ρ = , (7.35)
∂t 2 ∂rβ
where
Tαβ = λSαα δαβ + 2µSαβ . (7.36)
In these equations, the subscripts α and β run over 1, 2, 3 (corresponding to
x, y, z). A repeated index in a term implies summation. δαβ is the Kronecker
delta function. In a second alternative form the three force equations are written
straightforwardly as the single vector equation
∂ 2u
= ct2 ∇ 2 u + (cl2 − ct2 ) grad , (7.37)
∂t 2
where it is now clear that
= ∇ · u = div u. (7.38)
Here, ct and cl are the transverse and longitudinal sound speeds and we have
λ λ + 2µ
ct2 = and cl2 = . (7.39)
ρ ρ
In physical acoustics the solutions for the displacement fields are frequently
specified in terms of two potential functions, a scalar potential φ and a vector
potential = (x , y , z ), through
∂φ ∂x ∂ y
u= + − ,
∂x ∂y ∂z
∂φ ∂x ∂z
v= + − , (7.40)
∂y ∂z ∂x
∂φ ∂ y ∂x
w= + − ,
∂z ∂x ∂y
where φ and i , i = x, y, z, satisfy
1 ∂ 2φ λ + 2µ
∇ 2φ = , cl2 = ,
cl2 ∂t 2 ρ
(7.41)
1 ∂ 2 i λ
∇ i = 2
2
, i = x, y, z, ct2 = .
ct ∂t 2 ρ
The scalar potential φ corresponds to the ‘irrotational’ part of the solution and
the vector potential corresponds to any remaining ‘rotational’ fields. In the literature
the irrotational solutions are also referred to as the longitudinal, compressional, or
dilatational solutions. Moreover, seismologists frequently refer to these solutions as
P waves. Likewise, the rotational vector-potential solutions based on i are iden-
tified as transverse, shear, distortional, or equivoluminal solutions. In seismology
these solutions are commonly identified as S waves.
Herein, the irrotational solutions will generally be referred to as longitudinal
modes and the corresponding sound speed will be denoted by cl . Likewise, the
rotational fields will be denoted as transverse modes and the associated sound speed
will be denoted by ct . The principal interest in this book is on using the longitudinal
and transverse solutions of the elastic continuum model to describe the longitudinal
acoustic (LA) and transverse acoustic (TA) phonons in nanostructures. Hence,
the notation adopted herein is that corresponding most naturally to the solid-state
community’s descriptions of phonons as longitudinal and transverse.
Every year, experimental observations of acoustic modes in nanostructures are
being reported in the literature in increasing numbers. Examples of such experi-
mental studies include the works of Nabity and Wybourne (1990a, b), Seyler and
Wybourne (1992), and Sun et al. (1999). Wybourne and his coworkers report
measurements of confined acoustic phonons in very thin metallic foils and wires
on dielectric substrates. Sun et al. (1999) presented data on folded acoustic modes
in semiconductor superlattices. In view of such observations, Section 7.6 presents
extensions of the elastic continuum theory of this section to the case of acoustic
modes in dimensionally confined structures and Chapter 9 focuses on carrier–
acoustic-phonon scattering rates in both bulk materials and dimensionally confined
nanostructures.
7.3 Optical modes in dimensionally confined structures

The dielectric continuum model has been applied to describe the properties of
dimensionally confined optical phonons in many electronic and optoelectronics
devices fabricated from semiconductor nanostructures (Dutta and Stroscio, 1998,
2000; Mitin et al., 1999). These include quantum wells, superlattices, quantum
wires, and quantum dots. To illustrate the basic features of the dielectric continuum
model of optical phonons, the case of confinement in just one dimension – as

in a quantum well or superlattice – is considered first. In addition, the dielectric
continuum model will be compared with other continuum models, including the
hydrodynamic model and the reformulated dielectric continuum model. These
models predict different sets of confined optical phonon modes but each model
predicts the same carrier–phonon scattering rate as long as it includes a com-
plete set of orthogonal phonon modes (Nash, 1992). Following a comparison
of these models, a microscopic treatment of confined phonon modes will be
discussed.
7.3.1 Dielectric continuum model for slab modes:

normalization of interface modes
The dielectric continuum model predicts a set of confined optical phonon modes
commonly referred to as the slab modes. These slab modes may be determined by
applying the dielectric continuum model and by imposing electrostatic boundary
conditions at each heterointerface. The normal-mode frequencies and orthogonal
confined phonon modes are obtained through the simultaneous solution of the
equations arising from the dielectric continuum model, subject to the boundary con-
ditions that the potential, (r) and the normal component of D(r) are continuous at
each heterointerface. Taking the heterointerfaces to be normal to the z-direction, the
electrostatic potential i (r) in the region Ri = (z i , z i+1 ) and its two-dimensional
Fourier transform i (q, z) are related by

i (r) = i (q, z) e−iq·ρρ , (7.42)
q
where ρ ≡ (x, y) and q is the two-dimensional wavevector in the x y-plane, that is,
q = qx x̂ + q y ŷ, where x̂ and ŷ are unit vectors. Then

Ei (r) = −∇i (r) = Ei (q, z) e−iq·ρρ ,
q
(7.43)
Pi (r) = i (ω)Ei (r) = Pi (q, z) e−iq·ρρ .
q
Following the concepts of Section 5.1, the mode normalization condition requires
that the energy of a phonon of mode q is h̄ωq ; for the case of a single interface at
z = 0 separating two layers n and m, this condition is
∞ ∗
L2 (µn nn )un (q, z) · (µn nn )un (q, z) dz
0
0 ∗
+y L2 (µm nm )um,a (q, z) · (µm nm )um,a (q, z) dz
−∞
0 ∗
+ (1 − y) L2 (µm nm )um,b (q, z) · (µm nm )um,b (q, z) dz
−∞
h̄
= . (7.44)
2ωq
To illustrate the normalization procedure, let us consider one of the classes of

optical phonon modes existing for this one-heterointerface structure. The wave
equations for the fields of relevance here admit both oscillating and exponential
solutions. In particular, let us consider the solutions having an exponential character.
For these modes, known as the interface (IF) modes, we take

(r) = (q, z) e−iq·ρρ = ce−q|z| e−iq·ρρ (7.45)
q q
so that
!
c(iqe−qz q̂ − qe−qz ẑ) z ≥ 0,
Ei (q, z) =
c(iqeqz q̂ + qeqz ẑ) z ≤ 0,
! (7.46)
i c(iqe−qz q̂ − qe−qz ẑ) z ≥ 0,
Pi (q, z) =
i c(iqeqz q̂ + qeqz ẑ) z ≤ 0.
region 1 region 2
Figure 7.1. A ternary–binary structure. From Kim and Stroscio (1990), American
where q̂ is the unit vector specifying the direction of q ≡ (qx , q y ). Let material n
be a binary layer filling the space z ≥ 0, region 1, and material m be a ternary layer
filling the space z ≤ 0, region 2, as illustrated in Figure 7.1. Then, for the right-hand
medium, material 1, using the notation of Appendix B, it follows that
−µ1 ω2 u1 (q, z) = −µ1 ω01

2
u1 (q, z) + e1∗ Elocal (r),
µ1 (ω01
2 − ω2 )
P1 (q, z) = n1 en∗ u1 (q, z) + n1 α1 u1 (q, z)
e1∗
(7.47)
2 − ω2 )
α1 µ1 (ω01
= n1 e1∗ 1 + u1 (q, z)
e1∗2
= 1 c(iqe−qz q̂ − qe−qz z);
thus
1 c(iqe−qz q̂ − qeqz ẑ)
u1 (q, z) = . (7.48)
n1 e1∗ [1 + α1 µ1 (ω01
2 − ω2 )/e∗2 ]
1
For material 2, there are two driven-oscillator equations, one for the AC pair,
denoted by a, and one for the BC pair, denoted by b:
−µ22,a(b) ω2 u2,a(b) (q, z) = −µ2,a(b) ω02,a(b)

2
u2,a(b) (q, z)
∗
+ e2,a(b) Elocal (q, z), (7.49)
and the electric polarization in the virtual-crystal approximation is

∗ ∗
P2 (q, z) = n2 [(1 − y)e2b u2b (q, z) + ye2a u2a (q, z)] + n2 α2 Elocal (q, z)
∗2 ∗2
ye2a (1 − y)e2b
= n2 + + α2 Elocal (q, z)
µ2,a (ω02,a
2 − ω2 ) µ2,b (ω02,b
2 − ω2 )
= 2 c(iqeqz q̂ + qeqz ẑ). (7.50)
Then
2 c(iqeqz q̂ + qeqz ẑ)
Elocal (q, z) = ∗2 ∗2 ,
ye2a (1 − y)e2b
n2 + + α2
µ2,a (ω02,a
2 − ω2 ) µ2,b (ω02,b2 − ω2 )
∗
e2,a(b)
u2,a(b) (q, z) =
µ2,a(b) (ω02,a(b)
2 − ω2 )
2 c(iqeqz q̂ + qeqz ẑ)
× ∗2 ∗2 .
ye2a (1 − y)e2b
n2 + + α2
µ2,a (ω02,a
2 − ω2 ) µ2,b (ω02,b2 − ω2 )
(7.51)
With these expressions for u1 (q, z) and u2,a(b) (q, z), the normalization condition,
with n = 1 and m = 2, yields

h̄ 1 µ1 n1 χ12 q
=
2ω L 2 c2 {n1 e1∗ [1 + α1 µ1 e1∗−2 (ω01
2 − ω2 )]}2
∗ 2
e
+ µ2a n2 y 2a (ω02,a 2
− ω2 ) q
µ2,a
! 2 ∗2 ∗2 2 "
2 ye2a (1 − y)e2b
× + + α2
n22 µ2,a (ω02,a
2 − ω2 ) µ2,b (ω02,b
2 − ω2 )
∗ 2
e
+ µ2b n2 (1 − y) 2b (ω02,b 2
− ω2 ) q
µ2,b
! 2 ∗2 ∗2 2 "
2 ye2a (1 − y)e2b
× + + α2 ,
n22 µ2,a (ω02,a
2 − ω2 ) µ2,b (ω02,b
2 − ω2 )
(7.52)
where the integral in the first term has been performed using
∞ ∞
dz(iqe−qz q̂ − qe−qz z)∗ · (iqe−qz q̂ − qe−qz ẑ) = dz 2q 2 e−2qz = q,
0 0
(7.53)
and the second and third integrals have been performed using
0 ∞
dz(iqeqz q̂ + qeqz ẑ)∗ · (iqeqz q̂ + qeqz ẑ) = dz 2q 2 e−2qz = q. (7.54)
−∞ 0
Thus the normalization constant c is determined. It is convenient to rewrite this

expression using conditions derived by Wendler (1985). These conditions are
discussed in Appendix B. The conditions that are useful at this point in our derivation
are
n n αn
n (∞) = 1 + 4π ,
1 − 43 π nn αn
1
ωLO,n
2
= ω0,n
2
+ 23 ωplasma,n
2
,
1 + 3 π n n αn
8
1
ωTO,n
2
= ω0,n
2
+ 13 ωplasma,n
2
, (7.55)
1 − 3 π nn αn
4
1 1
ωLO,n
2
− ωTO,n
2
= 23 ωplasma,n
2
+ 13 ωplasma,n
2
1 + 3 π n n αn
8
1 − 3 π n n αn
4
ωplasma,n
2
= .
(1 + 83 π nn αn )(1 − 43 π nn αn )
Here the subscript n represents either material 1 or material 2. In these relations, the
plasma frequency squared, ωplasma,n
2 , is given by
∗2
ωplasma,n,a(b)
2
= 4π nn en,a(b) /µn,a(b) . (7.56)
As may be verified algebraically, the Lyddane–Sachs–Teller relations of subsec-

tion 2.3.3 are satisfied by these frequencies (Wendler, 1985).
With these results of Wendler, a straightforward but lengthy derivation yields

h̄ 8π ω 1/2 ∂ E 1 (ω) ∂ E 2 (ω) −1/2
c=− +
2ω L 2 q ∂ω ∂ω
1/2 −1/2
4π h̄ ∂ E 1 (ω) ∂ E 2 (ω)
=− + . (7.57)
2
L q ∂ω ∂ω
Thus,

(r) = (q, z) e−iq·ρρ = ce−q|z| e−iq·ρρ
q q
4π h̄ 1/2 ∂ E 1 (ω) ∂ E 2 (ω)
−1/2
=− + eq|z| e−iq·ρρ , (7.58)
q L 2q ∂ω ∂ω
where, as discussed previously,

2
ω − ωLO,1
2
1 (ω) = 1 (∞) ,
ω2 − ωTO,1
2
2 2 (7.59)
ω − ωLO,2,a
2 ω − ωLO,2,b
2
2 (ω) = 2 (∞) .
ω2 − ωTO,2,a
2 ω2 − ωTO,2,b
2
Finally, multiplying (r) by −e and introducing aq and aq† , according to the

procedure described in Section 5.1, the interaction Hamiltonian for the interface
(IF) optical phonon mode may be written as (Kim and Stroscio, 1990)
4π e2 h̄ 1/2 ∂ E 1 (ω) ∂ E 2 (ω)

−1/2
HIF = + e−q|z| eiq·ρρ (aq + a−q
†
),
q L 2q ∂ω ∂ω
(7.60)
where êq, j and ê∗q, j of equation (6.2) have been taken as unit vectors in the
longitudinal direction, since the IF phonon modes considered here are longitudinal
optical (LO) phonons. The dispersion relation for this optical phonon mode is
given from the requirement that the normal components
of the electric
displacement
field be continuous at z = 0, that is, 2 (ω)E 2,z z=0 = 1 (ω)E 1,z z=0 . From this
condition, it follows immediately that the frequencies of the IF optical phonons must
satisfy 1 (ω) + 2 (ω) = 0.
This result is similar to that for a bulk semiconductor, where the optical phonon
frequencies must satisfy (ω) = 0. Moreover, since this is the condition necessary
for the propagation of any longitudinal electromagnetic disturbance, it was expected
that the frequencies of longitudinal optical phonons should satisfy this dispersion
relation.
In the case of the two-region, single-heterointerface structure, the IF longitudinal
optical phonon frequencies depend on both 1 (ω) and 2 (ω). Therefore, the IF
optical phonon mode is a joint mode of both materials. Indeed, the electrostatic
phonon potential for this mode falls off exponentially with distance from the
heterointerface in both materials and, of course, the IF phonon electrostatic potential
has only one common value at the interface. Hence, such an IF optical phonon mode
must have one common electrostatic phonon potential throughout the two-region
heterostructure. It is therefore not surprising that the frequency of such an interface
LO phonon mode depends on the dielectric constants of both materials. Clearly,
the IF optical phonon mode described by HIF is a joint mode of both materials. In
general, IF optical phonon modes are joint modes of all the materials in a given
heterostructure. This property is manifest throughout this book.
Clearly, the IF optical phonon modes do not form a complete set of optical
phonon modes for the case of two semi-infinite regions joined at a single hetero-
interface. Indeed, these IF modes vanish exponentially as |z| → ∞ and it is
clear that bulk-like optical phonons must exist in regions significantly removed
from the heterointerface. These additional modes are known as half-space modes.
For a structure with a single heterointerface these half-space modes have been
given by Mori and Ando (1989) for the case where the two semi-infinite regions
are composed of binary semiconductors. Mori and Ando also gave the full set
of optical phonon modes for double-heterointerface structures, where one type of
binary semiconductor layer with interface planes situated at −a/2 and +a/2 is
bounded by two semi-infinite regions of a different binary semiconductor. Appendix
C provides a summary of the phonon modes of the double-heterointerface structure
for three models, the slab modes of the dielectric continuum model with electrostatic
boundary conditions, as in Mori and Ando (1989), and two other models discussed
widely in the literature. Appendix C also discusses Raman measurements useful
in understanding the behavior of these modes (Sood et al., 1985). The modes
arising from the second and third models are known as the guided modes and the
reformulated (or Huang–Zhu) modes. As discussed in Appendix C, all these sets
of phonon modes predict the same intrasubband and intersubband scattering rates
provided that each set is composed of a complete, orthogonal set of phonon modes.
7.3.2 Electron–phonon interaction for slab modes

Here, it is instructive to consider an earlier – and intuitively very appealing –
theory of electron–phonon interactions in a dielectric slab given by Licari and
Evrard (1977). In this theory, a single dielectric slab of infinite extent in the x- and
y-directions is situated with its faces at −a and +a and with its surface bounded by
a vacuum in the regions with |z| ≥ a. Within this dielectric slab ∇ · D = 0, where,
as usual, D(r) = (ω)E(r) = E(r) + 4πP(r); (ω) is the dielectric constant of

the slab and P(r) is the electric polarization associated with the optical phonons
in the slab. Defining a scalar potential through E(r) = −∇φ(r) and, since the
system is translationally invariant in the x y-plane, taking φ(r) to be of the form
φ(r) = φ(z) eiq ·ρρ , where ρ = (x, y) and q = (qx , q y ), it follows that

∂2
(ω) − q φ(z) = 0
2
(7.61)
∂z 2
where q2 = qx2 + q y2 . This equation is satisfied when (ω) = 0 or when (∂ 2 /∂ 2 z −

q2 )φ(z) = 0. As shown previously, from the general form of (ω) – as given for
example in Appendix A – and the Lydanne–Sachs–Teller relation, the condition
(ω) = 0 is satisfied for a single-material system when ω = ωLO . In this case an
arbitrary function, φ(z), is a solution of the wave equation; Licari and Evrard took
this solution (as did Fuchs and Kliewer, 1965, and Kliewer and Fuchs, 1966a, b) to
be of the form

φ(z) = (φ1 sin qz z + φ2 cos qz z), (7.62)
qz
inside the slab i.e., in the range (−a, +a). #Outside the slab, where = 1, the
solutions have the form φ(z) = φ± exp(± qx2 + q y2 z), where the positive sign
applies for z ≤ −a and the negative sign applies for z ≥ +a. The constants φ1 ,
φ2 , φ+ , and φ− are determined by the usual boundary conditions that the tangential
component of E and the normal component of D are continuous at z = ±a. From
these conditions it is seen that φ± = 0 and it is thus clear that for this mode
φ(z), E(r), and D(r) are zero in the regions surrounding the slab; in particular φ(z)
vanishes at the surfaces of the layer, where z = ±a. For z in the range (−a, +a),
the boundary conditions may be satisfied by taking either φ1 = 0 or φ2 = 0, so that
there are two solutions corresponding to the two polarization vectors:
φ2 iq ·ρρ $ mπ mπ mπ %
+ (r) =
Pm e i iq a cos z − ẑ sin z
4πa 2a 2 2a
m = 1, 3, 5, . . . ,
φ $ mπ mπ mπ %
(7.63)
1 iq ·ρρ
Pm
− (r) = e i iq a sin z + ẑ cos z
4πa 2a 2 2a
m = 2, 4, 6, . . . ,
where ẑ is the unit vector in the z-direction. Of course, ∇ · D(r) = 0 implies that
∇ 2 φ(r) = −∇ · E(r) = +4π ∇ · P(r). These standing modes are now widely known
as the confined optical phonon modes in a slab. They exist for m running from 1 to
some maximum number N2a ; the values of m must terminate at N2a , the number of
unit cells in thickness 2a, since the continuum model adopted here must fail when
the number of half-wavelengths in 2a becomes equal to or greater than the number
of unit cells in the same thickness.
The remaining solution corresponds to the case where (ω) = 0 inside the slab
# #
φ(x) = φ1 exp + qx2 + q y2 z + φ2 exp − qx2 + q y2 z .
#
As before, when |z| ≥ a, the solution is φ(z) = φ± exp(± qx2 + q y2 z). The
boundary conditions then restrict the modes to the forms
# 1 − iq ·ρρ
P0+ (r) = −φ1 qx2 + q y2 e i
$ # 4π # %
× i q̂ cosh qx2 + q y2 z + ẑ sinh qx2 + q y2 z ,
(7.64)
# 1 − iq ·ρρ
P0− (r) = −φ1 qx2 + q y2 e i
$ # 4π # %
× i q̂ sinh qx2 + q y2 z + ẑ cosh qx2 + q y2 z ,
where q̂ is the unit two-dimensional wavevector. These last two modes describe the
so-called IF optical phonon modes in the polar semiconductor slab of thickness 2a.
The boundary conditions imply that the frequencies for these modes are solutions of
1 + (ω) #
= ± exp −2 qx2 + q y2 a , (7.65)
1 − (ω)
which, using (ω) = (∞) + [(0) − (∞)]/(1 − ω2 /ωTO

2 ), may be written as
#
[(0) + 1] ∓ [(0) − 1] exp − qx2 + q y2 a
ω± = ωTO # ,
2 2
(7.66)
[(∞) + 1] ∓ [(∞) − 1] exp − qx2 + q y2 a
where the plus sign corresponds to the even mode, the minus sign to the odd mode.
As pointed out by Licari and Evrard (1977), this continuum model is capable
of predicting both the confined LO phonons and the interface IF optical phonons
because for both of these modes there exists a polarization charge density. In
particular, both ρ = −∇ · P, the volume charge density, and σ = −P · n̂, the
surface charge density, contribute to the confined LO modes; here, n̂ is the unit
vector normal to the surface and pointing into the vacuum. For the IF modes, only
σ makes a contribution. Clearly, in this model the polarization charge acts as the
source of the fields associated with these phonon modes. Transverse modes are not
predicted by this continuum approach since for such modes ∇ · P = 0 and P · n̂ = 0.
Licari and Evrard (1977) used this model to study the effects of electronic
polarizability on the phonon modes and they derive conditions for the slab which
are equivalent to Wendler’s conditions for the two-layer system described in Ap-
pendix B. Licari and Evrard also used their model to construct the normalized polar-
ization eigenvectors and frequencies for the phonon modes of the dielectric slab.
Moreover, they constructed the Hamiltonian for the electron–polar-optical-phonon
interaction and showed that the correct harmonic oscillator energy is recovered when
the eigenvectors of the slab are used to evaluate the Hamiltonian; in particular, it
can be shown that the normal modes are consistent with the harmonic oscillator
energy of Section 5.1. Finally, Licari and Evrard presented a very enlightening
physical derivation of the electron–phonon interaction Hamiltonian for a slab by
applying boundary conditions to the electron–phonon interaction Hamiltonian for
a bulk semiconductor. Specifically, starting with expression for the bulk Fröhlich
interaction, which we take as the expression (5.34) derived in Section 5.2,
! "1/2
2π e2 h̄ωLO 1 1 1
HFr = −i − †
(aq + a−q ) e−iq·r
V (∞) (0) q q
! 2π e2 h̄ωLO 1 1
"1/2
= −i − †
(aq + a−q ) e−iq·r
q V q 2 (∞) (0)

= −i Vq (aq + a−q †
) e−iq·r , (7.67)
q
Licari and Evrard took q = (q , qz ) and split the sum over q into a sum over q and
a sum over qz > 0:

HFr = Vq e−iq ·r e−iqz z (aq ,qz + a−q
†
,−qz
)
q ,qz >0
†
+ eiqz z (aq ,−qz + a−q ,qz
). (7.68)
Then, using eiθ = cos θ + i sin θ to write e±iqz z in terms of sines and cosines,
√
HFr = 2 Vq e−iq ·r
q ,qz >0
† †
× {cos qz z[a+ (q ) + a+ (−q )] + sin qz z [a− (q ) + a− (−q )]},
(7.69)
where
1 −i
a+ (q ) = √ (aq ,qz + aq ,−qz ), a− (q ) = √ (aq ,qz − aq ,−qz ).
2 2
(7.70)
† †
The operators a+ (−q ) and a− (−q ) are given by taking the adjoints. These oper-
ators describe phonons which propagate as plane waves in the x- and y-directions
but as standing modes in the z-direction. Indeed, since qz = mπ/2a the Fröhlich
Hamiltonian for the two-dimensional slab takes the form
1/2
4π e2 h̄ωLO 1 1
HFr = − e−iq ·r
V (∞) (0) q
!
cos(mπ/2a)z †
× [a (q ) + am+
2 + (mπ/2a)2 ]1/2 m+
(−q )]
m=1,3,5...
[q 2
x + q y
"
sin(mπ/2a)z †
+ [am− (q ) + am− (−q )] .
m=2,4,6...
[qx2 + q y2 + (mπ/2a)2 ]1/2
(7.71)
This Hamiltonian vanishes for z = ±a, as it must since the Fröhlich interaction
Hamiltonian is given by −eφ, as explained in Section 5.2, and since φ(±a) = 0 for
the potential describing the fields associated with phonon modes in the dielectric
slab. This heuristic derivation makes manifest the fact that the confined phonon
modes in the slab located between −a and +a are standing modes with an integer
number of half-wavelengths confined within the slab. This Hamiltonian does not
contain the contributions of the IF optical phonons in the slab since it satisfies only
the boundary conditions for the confined optical phonon modes at z = ±a, namely
HFr (a) = −eφ(±a) = 0. As shown by Licari and Evrard (1977), the Fröhlich
interaction Hamiltonian for the IF optical phonon modes in the dielectric slab is
! "1/2
2π e2 h̄ωTO
HFr =− [(0) − (∞)]
L2
 # 1/2
#
sinh 2 qx2 + q y2 a
 e− qx +q y a
2 2
× e−iq ·r  #
q qx2 + q y2
& #
× G + qx2 + q y2 , z [a0+ (q ) + a0+ †
(−q )]
# '
+ G − qx2 + q y2 , z [a0− (q ) + a0−
†
(−q )] , (7.72)
where
# ( #
# cosh qx2 + q y2 z cosh qx2 + q y2 a
G + qx2 + q y2 , z = #
−2 qx2 +q y2 a
[(∞) + 1] − [(∞) − 1] e
 # 1/4
 [(∞) + 1] − [(∞) − 1] e−2 qx2 +q y2 a 
× # ,
 −2 qx2 +q y2 a 
[(0) + 1] − [(0) − 1] e
(7.73)
and
# ( #
# sinh qx2 + q y2 z sinh qx2 + q y2 a
G − qx2 + q y2 , z = #
−2 qx2 +q y2 a
[(∞) + 1] + [(∞) − 1] e
 # 1/4
 [(∞) + 1] + [(∞) − 1] e−2 qx2 +q y2 a 
× # .
 −2 qx2 +q y2 a 
[(0) + 1] + [(0) − 1] e
(7.74)
In the result (7.72) the phonon creation and annihilation operators are not
summed over m; since there are just two IF optical phonon modes, the subscript
m on the creation and annihilation operators for the confined optical phonon modes
is replaced by 0 and the plus sign in the subscript corresponds to the even mode
while the minus sign corresponds to the odd mode. The modes discussed by Licari
and Evrard are recognized as the optical modes of a double-interface heterostructure
in the special case where = 1 outside the region bounded by the two heteroint-
erfaces. As mentioned previously, Appendix C compares the three frequently used
complete sets of optical phonon modes for a double-interface heterostructure for
the case where all the material layers are polar semiconductors. As may be seen
straightforwardly, the modes considered by Licari and Evrard correspond to the slab
modes derived with electrostatic boundary condition for the special cases where the
quantum well is bounded by a vacuum, so that = 1.
7.3.3 Slab modes in confined würtzite structures

The so-called slab modes for dimensionally confined würtzite semiconductor
structures have been derived using an extension of Loudon’s model of uniaxial
semiconductors, which was introduced in Chapter 3. The normalization condition
of Chapter 5 must be modified to take into account the fact that for uniaxial
semiconductor crystals there are separate equations governing u⊥,n and u,n , where,
as in subsection 7.3.2, the subscripts denote the components normal and parallel
to the c-axis of a uniaxial semiconductor with materials properties associated with
those of a medium n.
For such a medium, the normalization condition of Section 5.1 for a single mode
q (with ωq = ω) becomes

√
nn µn u⊥,n (r)2 + √nn µn u,n (r)2 dr = h̄ , (7.75)
2ω
where µn denotes the reduced mass and, as in Section 5.1, nn = Nn /V . Then,
defining

u⊥, (r) = u⊥, (q) eiq·r , (7.76)
q
it follows that
√
nn µn u⊥,n (q)2 + √nn µn u,n (q)2 = h̄ 1 , (7.77)
2ω V
provides the necessary generalization to the case of a uniaxial crystal. From
Section 7.1, the equations governing u⊥,n and u,n may be written as

1 (0)⊥(),n − (∞)⊥(),n
u⊥(),n = √ ωTO,⊥(),n E⊥,n , (7.78)
4π µn nn ωTO,⊥(),n
2 − ω2
where the time dependence of each displacement has been assumed to be of the form
eiωt . Then, it follows that
√
nn µn u⊥,n (r)2 + √nn µn u,n (r)2
1 [(0)⊥,n − (∞)⊥,n ]ωTO,⊥,n 2

2
= E⊥,n
4π (ωTO,⊥,n − ω )
2 2 2
1 ((0),n − (∞),n )ωTO,,n 2

2
+ E,n . (7.79)
4π (ωTO,,n − ω )
2 2 2
However, from the generalized Lyddane–Sachs–Teller relation, subsection 2.3.3, it

is known that
2
ωLO,⊥(),n − ω2
(ω)⊥(),n = (∞)⊥(),n , (7.80)
ωTO,⊥(),n
2 − ω2
and it follows that

1 ∂(ω)⊥(),n [(0)⊥(),n − (∞)⊥(),n ] 2
= ωTO,⊥(),n , (7.81)
2ω ∂ω (ωTO,⊥(),n
2 − ω 2 )2
so that
√
nn µn u⊥,n (r)2 + √nn µn u,n (r)2
1 1 ∂(ω)⊥,n 2 1 1 ∂(ω),n 2
= E⊥,n + E,n . (7.82)
4π 2ω ∂ω 4π 2ω ∂ω
Using this identity, the normalization condition becomes

1 1 ∂(ω)⊥,n 2
1 1 ∂(ω),n 2
h̄
E⊥,n + E,n dr = .
4π 2ω ∂ω 4π 2ω ∂ω 2ω
(7.83)
As discussed in Section 7.3, the normalization condition for the case where there is
dimensional confinement in only the z-direction is then

1 1 ∂(ω)⊥,n 2
1 1 ∂(ω),n 2
h̄
L 2
E⊥,n + E,n dz = ,
4π 2ω ∂ω 4π 2ω ∂ω 2ω
(7.84)
where ω = ωq and L 2 = L x L y . It should be noted that this integral is of the same

form as (7.44). As for the zincblende case, the electron–optical-phonon Hamiltonian
is then given by

HIF = −e (q, z) eiq·ρρ (aq + a−q
†
), (7.85)
q
where the potential, (q, z), is associated with E⊥,n and E,n . The form of
normalization (7.84) is particularly convenient for optical modes in the dielectric
continuum model since the phonons may be described in terms of the associated
electric fields and potentials (Kim and Stroscio, 1990; Lee et al., 1998; Komirenko
et al., 2000a).
In this subsection, the quantization condition for a uniaxial crystal will be
applied to determine the electron–optical-phonon interaction Hamiltonian for in-
terface phonons in a single-heterointerface structure. Appendix D summarizes the
electron–optical-phonon interaction Hamiltonians for all the optical phonon modes
in single- and double-heterointerface uniaxial crystals for the case where the c-axis
is perpendicular to the heterointerface(s).
As an illustration, consider the interface optical mode in a würtzite structure
composed of two semi-infinite regions separated by a single heterointerface situated
at z = 0. The c-axis is taken to be normal to the heterointerface. In the region
z < 0 the dielectric functions are (ω)⊥(),2 and in the region z > 0 they
are (ω)⊥(),1 . Each of the four functions, (ω)⊥(),1(2) obeys the generalized
Lyddane–Sachs–Teller relation; in particular, each is less than zero for frequencies
selected from one range of interest from among the four ranges ωTO,⊥(),1(2) <
ω < ωLO,⊥(),1(2) . In addition, for ωLO,⊥(),1(2) < ω < ωTO,⊥(),1(2) , it is clear
that (ω)⊥(),1(2) > 0 for the range corresponding to the one dielectric function of
interest. For each of these four ranges, the positivity or negativity properties are as
for zincblende crystals.
However, for uniaxial crystals the products (ω)⊥,1 (ω),1 and (ω)⊥,2 (ω),2
may be either positive or negative since, for a material n, (ω)⊥,n and (ω),n may
have different signs depending on the overlap of the two regions ωTO,⊥,n < ω <
ωLO,⊥,n and ωTO,,n < ω < ωLO,,n .
For binary zincblende heterostructures such as GaAs/AlAs, the IF modes exist
for the frequencies in the two ranges ωTO,AlAs < ω < ωLO,AlAs and ωTO,GaAs <
ω < ωLO,GaAs . Since these two ranges do not overlap, the frequency condition
for the existence of IF modes in such zincblende heterostructures is typified by
GaAs (ω)AlAs (ω) < 0 for all allowed IF mode frequencies ω. Such a simple
characterization is not possible for uniaxial crystals.
As will become obvious, this situation leads to significant differences in the
optical phonon modes in würtzite and zincblende structures. For the zincblende case
!
ρ Aeκ2 z z<0
φ(r ) = e iq·ρ
× −κ (7.86)
Be 1 z z>0
and in the absence of free charge ∇ · D = 0 so that
(ω),1 κ12 − (ω)⊥,1 q 2 = 0 (ω),1 (ω)⊥,1 > 0,

(7.87)
(ω),2 κ22 − (ω)⊥,2 q = 0
2
(ω),2 (ω)⊥,2 > 0.
From the continuity of the tangential component of the electric field at z = 0, it then
follows that A = B. From the continuity of the normal component of the electric
displacement at z = 0, ,2 κ2 A = −,1 κ1 B; thus ,1 κ1 + ,2 κ2 = 0 must be

satisfied in the allowed range of IF modes, which corresponds to the frequencies
satisfying ,1 (ω),2 (ω) < 0. However,
(ω)⊥,1 2
κ12 = q (ω),1 (ω)⊥,1 > 0,
(ω),1
(7.88)
(ω)⊥,2 2
κ22 = q (ω),2 (ω)⊥,2 > 0,
(ω),2
so that

(ω)⊥,1 (ω)⊥,2
,1 κ1 + ,2 κ2 = (ω),1 q + ,2 (ω) = 0. (7.89)
(ω),1 (ω),2
Accordingly,

− (ω)⊥,1 (ω),1 + (ω)⊥,2 (ω),2 = 0 (ω),1 < 0 and (ω),2 > 0,

+ (ω)⊥,1 (ω),1 − (ω)⊥,2 (ω),2 = 0 (ω),1 > 0 and (ω),2 < 0,
(7.90)

and it follows that (ω)⊥,1 (ω),1 = (ω)⊥,2 (ω),2 . Thus
!
Aeκ2 z z<0
φ(r ) = φ0 eiq·ρρ ×
Be−κ1 z z>0
(7.91)
ρ
exp( (ω)⊥,2 /(ω),2 qz) z<0
= φ0 e iq·ρ
×
exp(− (ω)⊥,1 /(ω),1 qz) z > 0.
Now E⊥ and E are given by the appropriate gradients of φ0 , and the integrals
2 2
needed to calculate the normalization condition are related to E⊥,n and E,n
through
0 0
2 q 2 1 (ω)⊥,2
E⊥,2 dz = φ q2 2
e 2 dz =
2κ z
φ0 =
2
qφ02 ,
0
−∞ −∞ 2κ 2 2 (ω),2
∞ ∞
q2 2 1 (ω)⊥,1
E⊥,1 2 dz = φ 2 q 2 e −2κ1 z
dz = φ0 = qφ02 ,
0
0 0 2κ1 2 (ω),1
0 0
2 κ 1 (ω)⊥,2
E,2 dz = φ 2 κ 2 2
e2κ2 z dz = φ02 = qφ02 ,
0 2
−∞ −∞ 2 2 (ω),2
∞ ∞
2 κ 1 (ω)⊥,1
E,1 dz = φ 2 κ 2 e−2κ1 z dz = φ02 =
1
qφ02 .
0 1
0 0 2 2 (ω),1
(7.92)
Then we have
1 h̄
L 22ω
1 1 ∂(ω)⊥,n 2
1 1 ∂(ω),n 2

= E⊥,n + E,n dz
4π 2ω ∂ω 4π 2ω ∂ω
0 ∞
1 1 ∂(ω)⊥,2 2 1 1 ∂(ω)⊥,1 2
= E⊥,2 dz + E⊥,1 dz
4π 2ω −∞ ∂ω 4π 2ω 0 ∂ω
0 ∞
1 1 ∂(ω),2 1 1 ∂(ω),1 2
E,2 dz + E,1 dz
2
+
4π 2ω −∞ ∂ω 4π 2ω 0 ∂ω

1 1 q ∂⊥,1 q ∂,1 κ1 ∂⊥,2 q ∂,2 κ2
= + + + φ02 ,
4π 2ω 2 ∂ω κ1 ∂ω q ∂ω κ2 ∂ω q
(7.93)
and it follows that
−1
4π h̄ 2 ∂⊥,1 q ∂,1 κ1 ∂⊥,2 q ∂,2 κ2
φ02 = 2 + + +
L q ∂ω κ1 ∂ω q ∂ω κ2 ∂ω q
(7.94)
and

HIF = −(e(q, z)) eiq·ρρ (aq + a−q
†
)
q

4π e2 h̄
= (2q −1 )1/2
q L2
−1/2
∂⊥,1 q ∂,1 κ1 ∂⊥,2 q ∂,2 κ2
× + + +
∂ω κ1 ∂ω q ∂ω κ2 ∂ω q
× eiq·ρρ (aq + a−q
†
)
√
− (ω)⊥,1 /(ω),1 qz
e√ z > 0,
× (7.95)
e (ω)⊥,2 /(ω),2 qz z < 0,
which may be written in an alternative form by use of the relation

%
∂ $

2 (ω)⊥,1 (ω),1 − (ω)⊥,2 (ω),2
∂ω

∂⊥,1 q ∂,1 κ1 ∂⊥,2 q ∂,2 κ2
= + + + . (7.96)
∂ω κ1 ∂ω q ∂ω κ2 ∂ω q
The dispersion relation for the modes described by this Hamiltonian is given by the

condition, (ω)⊥,1 (ω),1 = (ω)⊥,2 (ω),2 resulting from the requirements
that the tangential component of the electric field as well as the normal component
of the displacement field be continuous at the heterointerface. A similar analysis for
the case of a structure with two heterointerfaces is given in Appendix D (Komirenko
et al., 1999), where the dispersion of polar optical phonons in würtzite quantum
wells is considered at length. Gleize et al. (1999) have extended Komirenko’s results
to the case of würtzite superlattices. For the specific case of a GaN/AlN superlattice
with the c-axis normal to the heterointerfaces and with a superlattice period, d, as
shown in Figure 7.2, the requirement of periodicity and the boundary conditions
imposed on the fields for the quantum-well case lead to dispersion relations of the
form
a1 (ω) cosh[γ1 (ω)d/2] + a2 (ω) sinh[γ1 (ω)d/2] = 0, (7.97)
the antisymmetric modes and
a1 (ω) sinh[γ1 (ω)d/2] + a2 (ω) cosh[γ1 (ω)d/2] = 0, (7.98)
for the symmetric modes. In these dispersion relations we have
# (
γ1 (ω) = q⊥ (ω)⊥,1 (ω),1 ,

a1 (ω) = sign[(ω),1 ] (ω)⊥,1 (ω),1 , (7.99)

a2 (ω) = sign[(ω),2 ] (ω)⊥,2 (ω),2 .
The dispersion relations (7.97), (7.98) are depicted in Figures 7.3 and 7.4 for the
AlN(5 nm)/GaN(5 nm) superlattice along with quasi-confined modes. In the limit

q⊥ d → ∞, these dispersion relations reduce to the condition (ω)⊥,1 (ω),1 =

(ω)⊥,2 (ω),2 , as they must since in the short-wavelength limit the frequencies
of these modes cannot depend on d and, in fact, should be given by the dispersion
relation for a single heterointerface between GaN and AlN.
Let us consider in more detail the Hamiltonian for the single-heterointerface
structure. For the case when (ω)⊥,1 = (ω),1 and (ω)⊥,2 = (ω),2 , q 2 =
κ12 = κ22 and the Hamiltonian reduces to

4πe2 h̄ L 2 1 −q|z| iq·ρρ †
HIF = e e (aq + a−q ),
q ∂ 1 (ω)/∂ω + ∂ 2 (ω)/∂ω q
(7.100)
d2 d1
GaN AlN GaN AlN GaN
z
z=0
Figure 7.2. Würtzite superlattice considered by Gleize et al. (1999). From Gleize
et al. (1999). American Physical Society, with permission.
which can be recognized as the Hamiltonian for the IF optical phonon modes in a
zincblende single-heterointerface system.
In this subsection the interface optical phonon modes in a single-heterointerface
würtzite structure have been normalized to construct the Fröhlich-like electron–
Figure 7.3. Dispersion of interface and quasi-confined modes for an infinite and
unstrained AlN(5 nm)/GaN(5 nm) superlattice in the transverse optical (TO)
frequency range. The shaded areas depict the bands for all values of Q z ,
−π/d < Q z < π/d, lying in the first Brillouin zone of the superlattice: - — -,
Q z = 0; ——, Q z = π/d. The S ( j) and the AS ( j) are the symmetric and
antisymmetric modes with respect to the middle plane of any layer. j = 1 for GaN
and j = 2 for AlN. The quasi-confined modes are identified by their order m, an
integer following a comma. From Gleize et al. (1999), American Physical Society,
with permission.
optical-phonon interaction Hamiltonian HIF . Appendix D provides a summary of all

the optical phonon modes in single- and double-heterointerface würtzite structures
based on the macroscopic dielectric continuum model and Loudon’s model for
uniaxial crystals.
Figure 7.4. Dispersion of interface and quasi-confined modes for an infinite and
unstrained AlN(5 nm)/GaN(5 nm) superlattice in the longitudinal optical (LO)
frequency range. The shaded areas depict the bands for all values of Q z ,
−π/d < Q z < π/d, lying in the first Brillouin zone of the superlattice: - — -,
Q z = 0; ——, Q z = π/d. The S ( j) and the AS ( j) are the symmetric and
antisymmetric modes with respect to the middle plane of any layer. j = 1 for GaN
and j = 2 for AlN. The quasi-confined modes are identified by their order m, an
integer following a comma. From Gleize et al. (1999), American Physical Society,
with permission.
7.3.4 Transfer matrix model for multi-heterointerface

structures
Yu et al. (1997) derived a very useful set of normalization conditions for het-
erostructures containing multiple parallel heterointerfaces separating different semi-
conductor layers. These normalization conditions are essential for examining the
optical phonon bandstructure in superlattices and they provide the basis for rela-
tively straightforward calculations of the normalization factors for heterostructures
containing just a few heterointerfaces. Since translational invariance holds in the
two-dimensional planes parallel to the heterointerfaces, the electrostatic potential
describing the carrier–optical-phonon interaction in each region Ri = (z i , z i+1 ) is
denoted by i (r) and is taken to be of the form

i (r) = e−iq·ρρ i (q, z) (7.101)
q
with
i (q, z) = ci− e−qz + ci+ e+qz ≡ ci− φi− + ci+ φi+ , (7.102)
where the z-axis is taken to be normal to the heterointerfaces and where, as usual,
ρ = (x, y) and q denote the position and wavevector in two dimensions. ci− and ci+
are the relative amplitudes of the exponentially decaying and growing potentials,
respectively, in layer i; as will become clear, these relative amplitudes are related
through a transfer matrix. Figure 7.5 depicts a generic potential i (z) for regions
R0 , R1 , . . . , Rn .
According to the electrostatic boundary conditions the electrostatic poten-
tial i (q, z) and the normal component of the electric displacement, i Ei =
Figure 7.5. A possible generic potential for regions R0 , R1 , . . . , Rn . From Yu et al.

(1997), American Institute Physics, with permission.
−i ∂i (q, z)/∂z, must be continuous at each heterointerface; thus, at the hetero-
interface located at z i ,
∂i (q, z i ) ∂i−1 (q, z i )
i (q, z i ) = i−1 (q, z i ) and i = i−1
∂z ∂z
(7.103)
are the boundary conditions at the heterointerface separating regions Ri−1 =

(z i−1 , z i ) and Ri = (z i , z i+1 ). Yu et al. (1997) wrote these results in matrix form by
defining

 
ci− φi− (z) φi+ (z)
Ci = and Mi (z) =  ,
ci+ (z) φ (z)
i φi− i i+
(7.104)
so that
Mi (z i )Ci = Mi−1 (z i )Ci−1 . (7.105)
Then with the matrix C0 for region R0 , the column vector Ci and therefore
the electrostatic potential i (q, z i ) can be determined in any region through the
sequence
Ci = Qi (z i )Ci−1 = Qi (z i )Qi−1 (z i−1 ) · · · Q1 (z 1 )C0 , (7.106)
where the transfer matrix relating Ri−1 and Ri is given by
Qi (z i ) = Mi (z i )−1 Mi−1 (z i ).
Clearly, the electrostatic potential for an n-interface heterostructure, (q, z i ), is

given by joining the solutions i (q, z) for each region:
0 (q, z), z ∈ R0 ; · · · i (q, z), z ∈ Ri ; · · · n (q, z), z ∈ Rn .
Moreover, in each region (ω) and (∞) are related through the generalized
Lyddane–Sachs–Teller relations, subsection 2.3.3 and Section 7.1. For interface
optical phonons, the potentials must decrease exponentially as z → ±∞ so that,
for an n-region heterostructure, cn+ = 0 and c0− = 0. Thus, the dispersion relation
for this interface mode is obtained by setting the (2, 2) component of the transfer
matrix equal to zero; that is,
[Qn (z n , q, ω)Qn−1 (z n−1 , q, ω) · · · Q1 (z 1 , q, ω)]2,2 = 0. (7.107)
The number of interface optical phonon modes may be determined by examining

the dispersion relation obtained in this way; indeed, since a given Qi is proportional
to i (ω)i−1 (ω), and since for a heterostructure with n interfaces Cn is the product
of n Q-matrices and C0 , it follows that for a system with binary layers only the
dispersion relation goes as (ω2 )2n . Thus, for such an n-interface structure with only
binary layers, there are 2n interface optical phonons. By extending this argument,
it follows that such a heterostructure with alternating layers of binary and ternary
semiconductors has 3n interface optical phonon modes, since each ternary layer has
two binary-like optical phonon modes.
The normalization condition for these modes is a straightforward generalization
of the normalization condition for optical phonon modes in simple heterostructures,
namely,
h̄
L2 dz |µi ni ui (q, z)|2 = . (7.108)
i Ri 2ω
To normalize the optical phonon modes of such multiple heterointerface struc-

tures, it is convenient to take
∂i (q, z)
Ei (q, z) = −∇i (q, z) = −iqi (q, z)q̂ − ẑ,
∂z
(7.109)
∂i (q, z)
Pi (q, z) = −χi (ω) iqi (q, z)q̂ + ẑ ,
∂z
where q̂ and ẑ are the unit vectors for q and the z-direction respectively. Then, for
Ri it follows from our previous expression relating the displacement to the electric
polarization that
Pi (q, z)
ui (q, z) =
1 + αi µi ei∗−2 (ω0i
ni ei∗ 2 − ω2 )

∂i (q, z)
−χi (ω) iqi (q, z)q̂ + ẑ
∂z
= . (7.110)
ni ei∗ [1 + αi µi ei∗−2 (ω0i
2 − ω2 )]
Then, it follows straightforwardly that

/ 0
∂i (q, z) 2
χi2 (ω) q 2 |i (q, z)|2+

∂z
|µi ni ui (q, z)|2 = µi ni & '2 . (7.111)
ni ei∗ [1 + αi µi ei∗−2 (ω0i
2 − ω2 )]
Generalizing our previous expression for Pi (q, z) to the corresponding result for
layer i,
Pi (q, z) = ni ei∗ ui (q, z) + ni αi Elocal,i (q, z)

/ 0
∗ ei∗
= ni ei + ni αi Elocal,i (q, z)
µi (ω0i
2 − ω2 )
/ 0
∗ ei∗ 4π
= ni ei + n α
i i E i (q, z) + Pi (q, z) . (7.112)
µi (ω0i
2 − ω2 ) 3
Thus
Pi (q, z) = χi (ω)Ei (q, z)
ei∗
ni ei∗ + ni αi
µi (ω0i
2 − ω2 )
= / 0 Ei (q, z), (7.113)
ei∗
1− 3π
4
ni ei∗ + ni αi
µi (ω0i
2 − ω2 )
and it follows that

|µi ni ui (q, z)|2
q 2 |i (q, z)|2 + |∂i (q, z)/∂z|2
/ 0−2
∗ αi µi (ω0i2 − ω2 )
= µi ni ni ei 1 +
ei∗2
/ 02
∗ ei∗
ni ei + ni αi
µi (ω0i
2 − ω2 )
× / 02
e ∗
1 − 43 π ni ei∗ i
+ ni αi
µi (ω0i
2 − ω2 )
/ 0−2
ei∗ ei∗
2 2
4
= ni 2 −1 + π nn + αi
(ω0i − ω2 )2 3 µi (ω0i
2 − ω2 )
/ 0−2
1 ω2pi 3 ω pi
1 2
= −(ω0i − ω ) +
2 2
, (7.114)
4π (1 − 4π 3 nn αn )
2 1 − 4π 3 nn α n
where as defined previously ω2pi = 4π ni ei∗ /µi is the plasma frequency squared.
2
Finally, using the expressions for ωLO,i

2 , ωTO,i
2 , and i (∞) given in Appendix B,
equation (B.15), we have
|µi ni ui (q, z)|2 1 ωLO,i
2 − ωTO,i
2
2
= 4π i (∞) 2
∂i (q, z) (ω − ωTO,i
2 )2
q 2 |i (q, z)|2 +

∂z
1 1 ∂i (ω)
= (7.115)
4π 2ω ∂ω
and

1 1 ∂i (ω) ∂i (q, z) 2

dz q |i (q, z)| +
2 2 = h̄ .
4π 2ω ∂ω ∂z 2ωL 2
i Ri
(7.116)
Thus, for a phonon potential of the form

i (q, z) = A(ci− e−qz + ci+

e+qz ) = Ai (q, z), (7.117)
the normalization constant A is given by

1/2
h̄ 1 1 ∂i (ω)
A=
2ωL 2 i
4π 2ω ∂ω

∂i (q, z) 2 −1/2
× dz q |i (q, z)| +
2 2
,
Ri ∂z
(7.118)
and so the Hamiltonian is

HIF = ei (r) = e e−iq·ρρ i (q, z)(a−q
†
+ aq )
q

=e e−iq·ρρ Ai (q, z)(a−q
†
+ aq ). (7.119)
q
The utility of this formulation is illustrated by its application for heterostructures

with known phonon modes. Consider the interface optical phonons for the case
of two semi-infinite semiconductor regions joined at a single interface at z = 0.
Clearly, 0 (q, z) = c0− e+qz = e+qz for z ≤ 0 and 1 (q, z) = c1− e−qz = e−qz for
z ≥ 0. Then

4π h̄ 1/2 1 ∂0 (ω) 0
A= dz 2q 2 e+2qz
2ωL 2 2ω ∂ω −∞
−1/2
1 ∂1 (ω) ∞
+ dz 2q 2 e−2qz
2ω ∂ω 0
1/2 ! "
4π h̄ 1 ∂0 (ω) ∂1 (ω) −1/2
= q + , (7.120)
2ωL 2 2ω ∂ω ∂ω
so that

4πe2 h̄ L −2 1 −iq·ρρ −q|z| †
HIF = e e (a−q + aq ),
q ∂0 (ω)/∂ω + ∂1 (ω)/∂ω q
(7.121)
which is identical to the result obtained previously.
A second illustrative example is given by the case of a layer of one material
situated in the region from z = −d/2 to z = +d/2 and bounded by two semi-infinite
regions of another material; for example, we might consider a GaAs quantum well
of thickness d embedded in AlAs barriers; the center of the quantum well is at
z = 0. For this case, the phonon potential must decrease exponentially for z → ±∞
and the phonon potential in the quantum well must be a combination of increasing
and decreasing exponentials. Consider the case where the phonon potential in the
quantum well is even. Let the dielectric constant in the quantum well be 1 (ω) and
that of the barriers be 0 (ω). Since the barriers are taken to be the same material,
2 (ω) = 0 (ω). Then, it is clear that 0 (q, z) = e+q(z+d/2) for z ≤ −d/2,
1 (q, z) = (cosh qz)/(cosh qd/2) for |z| < d/2, and 2 (q, z) = e−q(z−d/2) for
z ≥ d/2 define an admissible envelope for the phonon potential. It follows that

−d/2
∂0 (q, z) 2
dz q |0 (q, z)| +
2 2 = dz 2q 2 e2qz eqd = q,
∂z
R0 −∞

∂1 (q, z) 2
dz q |1 (q, z)| +
2 2

R1 ∂z
+d/2
cosh2 qz + sinh2 qz
= dz q 2 = 2q tanh qd/2,
−d/2 cosh2 qd/2
∞
∂2 (q, z) 2
dz q |2 (q, z)| +
2 2 =
dz 2q 2 e−2qz eqd = q,
R2 ∂z d/2
(7.122)
and, accordingly,

4π h̄ 1/2 1 ∂0 (ω) 1 ∂1 (ω) qd −1/2
A= 2q + 2q tanh (7.123)
2ωL 2 2ω ∂ω 2ω ∂ω 2
so that for the symmetric case

4π e2 h̄ L −2 1 −iq·ρρ
HIF,S = e †
f S (q, z)(a−q + aq ),
∂0 (ω) ∂1 (ω) 2q
q + tanh qd/2
∂ω ∂ω
(7.124)
where f S (q, z) = i (q, z). The dispersion relation for this optical phonon mode is
given from the requirement that the normal components of the electric displacement

field be continuous at the heterointerfaces. At z = −d/2, 0 (ω)E 0,z z=−d/2 =

1 (ω)E 1,z z=−d/2 .
From this condition it follows immediately that the frequencies of the IF optical
phonons must satisfy 0 (ω) + 1 (ω) tanh qd/2 = 0. This same dispersion relation is
obtained from the continuity of the normal component of the electric displacement
field at z = d/2. Recall that in a bulk semiconductor the optical phonon frequencies
must satisfy (ω) = 0; indeed, since this is the condition necessary for the
propagation of any longitudinal electromagnetic disturbance, it is expected that
the frequencies of longitudinal optical phonons should be given by this dispersion
relation.
In the case of a two-material, double-heterointerface structure, the IF longitudinal
optical phonon frequencies depend on both 0 (ω) and 1 (ω). The expression for
HIF,S is identical to that of Kim and Stroscio (1990) and can be rewritten to be in
the form given by Mori and Ando (1989). The mode described by this Hamiltonian
is the symmetric IF optical phonon for the quantum well being considered. As a
final example, if 0 (q, z) = −e+q(z+d/2) for z ≤ −d/2, 1 (q, z) =

(sinh qz)/(sinh qd/2) for |z| < d/2, and 2 (q, z) = e−q(z−d/2) for z ≥ d/2, it
follows that for the antisymmetric case

4πe2 h̄ L −2 1 −iq·ρρ
HIF,A = e †
f A (q, z)(a−q + aq ),
∂0 (ω) ∂1 (ω) qd 2q
q + coth
∂ω ∂ω 2
(7.125)
where f A (q, z) = i (q, z). As before, the dispersion relation for this mode follows
from the requirement that the normal component of the electric displacement field
be continuous at the heterointerface. In this case, 0 (ω) + 1 (ω) coth qd/2 = 0.
This result reproduces the Hamiltonian derived by Kim and Stroscio (1990) for the
antisymmetric IF optical phonon of the quantum-well system in question.
The transfer-matrix approach of Yu et al. (1997) may be used to gain insights
into the nature of phonons in superlattices. Indeed, application of the transfer
matrix method to a multiple-barrier AlAs/GaAs structure, Figure 7.6(a), leads to
the dispersion relations such as those depicted in Figure 7.7 for various AlAs/GaAs
heterostructures.
As will become evident in Chapter 10, the five-interface heterostructure of Figure
7.6(b) is of importance in narrow-well semiconductor lasers. The transfer-matrix
method of Yu et al. (1997) may be applied to determine the IF phonon dispersion
relations and the associated IF phonon potentials. For the case where the two
barriers in Figure 7.6(b) are Al0.6 Ga0.4 As and the shallow barrier to the far left is
Al0.25 Ga0.75 As, the dispersion relations are as in Figure 7.8 and the five AlAs-like
interface modes are as shown in Figure 7.9. As discussed previously, there are in
total 15 IF modes in such a five-interface binary–ternary heterostructure. Indeed, as
Figure 7.6. Potential profiles for (a) a multiple-barrier AlAs/GaAs heterostructure

and for (b) a five-interface asymmetric heterostructure. From Yu et al. (1997),
American Institute of Physics, with permission.
Figure 7.7. Longitudinal optical IF phonon dispersion relations for (a) one-barrier,
(b) two-barrier, (c) three-barrier, (d) four-barrier structures of the type shown in
Figure 7.6(a). From Yu et al. (1997), American Institute of Physics, with permission.
Figure 7.8. Dispersion relations for the 15 interface modes in a five-interface

AlAs/GaAs heterostructure. From Yu et al. (1997), American Institute of Physics,
with permission.
deduced previously, such an n-interface structure with alternating layers of binary

and ternary semiconductors has 3n IF optical phonon modes.
The transfer-matrix approach of Yu et al. (1997) is also useful in understanding
the IF phonons in the double-barrier heterostructures discussed widely in connection
with new types of electronic diodes and transistors. As will be discussed in Chapter
10, the so-called valley current in such double-barrier structures is due in large
measure to phonon-assisted tunneling. Consider Figure 7.10, which depicts the
potential profile for a typical double-barrier heterostructure. The Fröhlich interaction
Hamiltonian and the dispersion for a heterostructure of the generic type shown in
Figure 7.10 was derived by Mori et al. (1992) and by Kim et al. (1992) without the
benefit of the transfer-matrix method of Yu et al. These results may be summarized
as follows:

HIF,S = (4πe2 h̄ L −2 )1/2
q
−1/2
∂1 (ω) qd1 ∂2 (ω) 2
× tanh + a2 + (b + c2 )(1 − e−2qd2 )
∂ω 2 ∂ω

1 −iq·ρρ †
× e (a−q + aq )
2q


 ae−q(z−d1 /2−d2 ) d1 /2 + d2 ≤ z,

 −q(z−d1 /2) + ceq(z−d1 /2−d2 )

 be d1 /2 ≤ z ≤ 2 + d2 ,
× (cosh qz)/(cosh qd1 /2) |z| ≤ d1 /2,

 /2) −q(z+d /2+d )
be

q(z+d1 + ce 1 2 −d1 /2 − d2 ≤ z ≤ −d1 /2,

 q(z+d1 /2+d2 )
ae z ≤ −d1 /2 − d2
(7.126)
is the Fröhlich interaction Hamiltonian for the symmetric LO phonon interface
modes, and

HIF,A = (4πe2 h̄ L −2 )1/2
q
−1/2
∂1 (ω) qd1 ∂2 (ω) 2
× coth + a2 + (b + c2 )(1 − e−2qd2 )
∂ω 2 ∂ω

1 −iq·ρρ †
× e (a−q + aq )
2q


 ae−q(z−d1 /2−d2 ) d1 /2 + d2 ≤ z,

be−q(z−d1 /2) + ceq(z−d1 /2−d2 )

 d1 /2 ≤ z ≤ 2 + d2 ,
× (sinh qz)/(sinh qd1 /2) |z| ≤ d1 /2,

 q(z+d1 /2) − ce−q(z+d1 /2+d2 )

 −be −d 1 /2 − d2 ≤ z ≤ −d1 /2,


−ae q(z+d1 /2+d 2 ) z ≤ −d1 /2 − d2
(7.127)
is the Fröhlich interaction Hamiltonian for the antisymmetric modes.
Figure 7.9. The phonon potentials for the five AlAs-like interface phonon modes of
Figure 7.8. The line codes used in Figure 7.8 are employed here also to indicate
which phonon potentials correspond to which. From Yu et al. (1997), American
Figure 7.10. Potential profile for a double-barrier heterostructure with a quantum

well of thickness d1 and dielectric constant 1 , and barriers of thickness d2 and
dielectric constant 2 . The growth direction is taken to be the z-direction. From Kim
For the symmetric modes,

1 qd1
a = cosh qd2 + tanh sinh qd2 ,
2 2

1 1 qd1
b= 1 − tanh , (7.128)
2 2 2

1 1 qd1
c = eqd2 1 + tanh ,
2 2 2
and the dispersion relation is given by
& '1/2
0 = 1 ± 1 − [(2 tanh qd2 )/(1 + tanh qd1 /2)] tanh qd1 /2
2
× 1 (1 + tanh qd1 /2)/(2 tanh qd2 ) + 2 . (7.129)
The antisymmetric modes a, b, c and the dispersion relation are obtained from these
results by substituting coth qd1 /2 for tanh qd1 /2.
The dispersion relations determined by Kim et al. (1992) are displayed in
Figures 7.11 and 7.12 for the case of a 60-ångstrom-wide GaAs quantum well
with 60-ångstrom-wide AlAs barriers. There are four symmetric (S) and four
antisymmetric (A) IF optical phonon modes for this heterostructure. The AlAs-like
modes are denoted by the subscript 2 and the GaAs-like modes by the subscript 1.
Figure 7.11. Dispersion relation for the four symmetric (S) LO phonon interface
modes for the heterostructure shown in Figure 7.10 for the case of a
60-ångstrom-wide GaAs quantum well with 60-ångstrom-wide AlAs barriers. From
Kim et al. (1992), American Institute of Physics, with permission.
The ± subscripts are used to distinguish the two different roots of the dispersion
relation for each of the modes.
As will become clear in Chapter 10 the interface LO phonon modes contribute
significantly to the valley current in such a double-barrier quantum-well structure.
There it will be explained how the mode labeled by ω S2− in Figure 7.11 makes
a major contribution to the valley current in certain double-barrier quantum-well
structures, through phonon-assisted tunneling of carriers into the quantum well.
7.4 Comparison of continuum and microscopic models

for phonons
The dielectric continuum model of optical phonons has been compared with the
microscopic theories by many authors, including Rücker et al. (1992), Molinari
et al. (1992), Molinari et al. (1993), Bhatt et al. (1993a), and Lee et al. (1995).
In view of the variety of continuum models of dimensionally confined phonons, as
discussed in Appendix C, there has been a clear motivation to perform microscopic
calculations in order to understand the properties of phonons in nanostructures.
These microscopic models have included both ab initio models (Molinari et al.,
1992) and simplified microscopic models based on empirical lattice force constants
Figure 7.12. Dispersion for the four antisymmetric (A) LO phonon interface modes
for the heterostructure shown in Figure 7.10 for the case of a 60-ångstrom-wide
GaAs quantum well with 60-ångstrom-wide AlAs barriers. From Kim et al. (1992),
7.4 Comparison of continuum and microscopic models for phonons 91
(Bhatt et al., 1995). Molinari et al. (1993) applied a microscopic model to calculate
the atomic displacement amplitudes and the Fröhlich potentials for quantum wells
and quantum wires. These calculations show that none of the macroscopic models –
including the so-called ‘slab’ model – gives a completely accurate representation of
the microscopic situation. However, as discussed in Appendix C the intersubband
and intrasubband scattering rates computed with the phonon modes of all these
models are in good agreement as long as the macroscopic model selected is
based on a complete set of orthogonal modes. Further discussions concerning the
comparisons of these macroscopic models are given in Appendix C. In this section,
the intersubband and intrasubband scattering rates calculated with slab modes will
be compared with the results of the microscopic models (Bhatt et al., 1993a). Figure
7.13 presents a comparison of the scattering rates for intersubband and intrasubband
electron–phonon scattering for a GaAs quantum well embedded in AlAs barriers.
For well widths in the 2 to 10 nanometer range and for a temperature of 300 K,
the scattering rates calculated from the microscopic model (Bhatt et al., 1993a) and
from the slab modes are in excellent agreement for an electron energy of 50 meV.
Figure 7.13. Intrasubband (1 → 1) and intersubband (2→1) scattering rates as a

function of quantum-well width at a temperature of 300 K and for an electron energy
of 50 meV: solid line, microscopic theory; dots, macroscopic theory. From Bhatt
et al. (1993a), American Physical Society, with permission.
For the specific case of a 20-monolayer-wide GaAs quantum well embedded

in AlAs, Figure 7.14 presents a comparison of the scattering rates at 300 K for
(a) intrasubband (1 → 1) and (b) intersubband (2 → 1) electron–optical-phonon
Figure 7.14. Comparison of

the scattering rates at 300 K
for (a) intrasubband (1 → 1)
and (b) intersubband
(2 → 1)
electron–optical-phonon
transitions in a
20-monolayer-wide GaAs
quantum well embedded in
AlAs. The result from three
different phonon models are
given: solid lines,
macroscopic models with
slab modes; broken and
dotted lines, ab initio model
of Molinari et al. (1992);
broken lines, microscopic
model with empirical force
constants (Bhatt et al.,
1993a). From Bhatt et al.
(1993a), American Physical
Society, with permission.
7.5 Comparison of dielectric continuum model predictions 93
transitions for three different types of phonon model: macroscopic models with slab
modes (solid lines), the ab initio model (broken and dotted lines) of Molinari et al.
(1992), and a microscopic model (Bhatt et al., 1993a) with empirical force constants
(broken lines). These results indicated that the slab model and the simplified
microscopic model provide good approximations to the scattering rates predicted
by the fully microscopic model.
7.5 Comparison of dielectric continuum model

predictions with results of Raman measurements
Chapters 8 and 10 will survey numerous examples on the applications of dimen-
sionally confined phonon modes of the dielectric continuum model. In the present
section, selected Raman measurements of optical phonon modes in dimensionally
confined polar semiconductors are discussed, since these measurements illuminate
the properties of such modes.
Figure 7.15 depicts Raman spectra taken at 15 K for GaAs N /AlAs N quantum-
well heterostructures with thicknesses of N = 10±1 monolayers, for three different
laser energies: (a) E L = 1.933 eV, (b) E L = 1.973 eV, and (c) E L = 2.410 eV
(Fasol et al., 1988). In cases (a) and (b) the measurements were made in the z(x x)z̄
polarization configuration and for case (c) the polarization configuration was z(x y)z̄.
The peaks on the various spectra are denoted by 10 N , 8 N , 6 N , 6 N −1 , 5 N , 5 N −1 , 4 N ,
4 N −1 , 3 N , 2 N , and 1 N . This notation is used to identify the LO4 mode – confined
mode with n = 4 – in a quantum well with N monolayers as the 4 N peak. Likewise,
the peak associated with the LO4 mode in a quantum well with N − 1 monolayers is
identified by 4 N −1 . A remarkable feature of Figure 7.15 is that the confined-phonon
wavevectors, qz = nπ/L z , for n = 4, 5, and 6 are sensitive to even a one-monolayer
fluctuation in the thickness of the GaAs quantum well. Moreover, Fasol et al. (1988)
showed that the change in qz corresponds to changing L z from 10 monolayers to
9 monolayers. These observations indicate that the phonons in AlAs/GaAs/AlAs
quantum-well heterostructures are confined very strongly at the heterointerfaces.
This result is expected on the basis of the dielectric continuum model, since the LO
phonon energies differ substantially for the two materials; the LO phonon modes in
AlAs at zone center have energies of about 50 meV while those of GaAs are about
36 meV.
One of the major predictions of the dielectric continuum model of optical phonon
modes is the existence of IF optical phonons which are joint modes of the two or
more heterostructure materials and which have frequencies characteristic of their
individual optical phonon modes. A recent example of the Raman analysis of such
an interface mode – and the first known for a würtzite heterostructure – is provided
by Dutta et al. (2000). In these measurements, the IF modes were observed for
GaN/AlN würtzite superlattices, as shown in Figure 7.16 for a 20-period GaN(9
nm)/AlN(8.5 nm) superlattice (lower curve) and a 40-period GaN(3 nm)/AlN(3

nm) superlattice (upper curve). In both cases, the wavelength of the incident laser
radiation is 244 nm.
The second case, with the narrower wells, is characterized by a broader Raman
peak as expected (Komirenko et al., 1999; Gleize et al., 1999). Moreover, the
energies and asymmetries of the IF modes are as expected from the dielectric
continuum model, as is clear by comparison with Figure D.2 of Appendix D.
The superperiodicity of the multilayered structures lowers the crystal symmetry
and increases the size of the unit cell so that it includes more atoms per cell. The
effect on the Raman properties is to allow more modes to exist and perhaps to
allow optically inactive modes to become active. For instance, in the zincblende
structure, for heterostructures grown in the [001] direction, the crystal symmetry
is lowered from Td to D2d and the A1 mode, previously inactive, becomes active.
Similar behavior has been predicted by group-theory analysis (Kitaev et al., 1998)
in würtzite (GaN)m (AlN)n superlattices, in cases where the Raman-active modes
Figure 7.15. Raman spectra

taken at 15 K for
GaAs N /AlAs N
quantum-well
heterostructures with
thicknesses of N = 10 ± 1
monolayers for three
different laser energies: (a)
E L = 1.933 eV, (b)
E L = 1.973 eV, and (c)
E L = 2.410 eV. The peaks
in the spectra are denoted by
n N and n N −1 , where n
represents the nth confined
mode for the quantum well.
In this case, there are only
two quantum-well
thicknesses, N = 10
monolayers and N − 1 = 9
monolayers. From Fasol
et al. (1988), American
Physical Society, with
permission.
7.5 Comparison of dielectric continuum model predictions 95
vary non-monotonically with the increase in supercell size. The material quality in
these structures, however, is not good enough yet for these effects to be observed
experimentally.
Resonant Raman studies on (AlGa)N/GaN quantum wells have allowed the
observation of the A1 (LO) phonons in the quantum wells (Behr et al., 1997; Gleize
et al., 2000), but the observation of a series of confined phonons similar to those
observed in the zincblende structures has yet to be made in the würtzite nitride
system.
The first really significant observation of the impact of heterostructures on light
scattering was the observation of doublets in the acoustic phonons in superlattices
by Colvard et al. (1980). A number of other studies on different materials have since
been done. These have been reviewed extensively by Jusserand and Cardona (1991).
As yet no similar observations have been reported for the würtzite nitride system,
although recently Göppert et al. (1998) have reported that confined optical and
folded acoustic phonons have been observed in the würtzite CdSe/CdS superlattices.
As a final example of where Raman measurements provide insights into the prop-
erties of confined optical phonon modes in polar semiconductor heterostructures,
Figure 7.16. Raman spectra for a 20-period GaN(9 nm)/AlN(8.5 nm) würtzite
superlattice (lower line) and a 40-period GaN(3 nm)/AlN(3 nm) würtzite superlattice
(upper line). The incident laser wavelength was 244 nm. From Dutta et al. (2000), to
be published.
Figure 7.17 (Fasol et al., 1988) illustrates that the confined phonon energies and
wavevectors for superlattices and quantum wells fall on the dispersion curve (solid
line) measured by neutron scattering at 10 K for bulk GaAs (Richter and Strauch,
1987). The neutron scattering data are in agreement with the earlier results of Waugh
and Dolling (1963).
The Raman data in Figure 7.17 are from works as follows: circles, Klein (1986);
diagonal crosses, Worlock (1985); squares, Castro and Cardona (1987); diamonds,
Colvard et al. (1980); and upright crosses, Sood et al. (1985), Jusserand and Paguet
(1986), and Sood et al. (1986).
The agreement between the Raman and neutron scattering measurements is
extremely enlightening. These results tell us – at least for the GaAs/AlAs system –
Figure 7.17. Comparison of LO phonon energies and wavevectors determined by

Raman scattering with those of neutron scattering (solid line). See text for references
to the works summarized in this figure. From Fasol et al. (1988), American Physical
that phonon confinement effects serve to restrict the phase space of the phonons but
not to alter substantially the basic energy–wavevector relationship for the phonons.
This may be understood by considering the dominant role of the nearest-neighbor
coupling of adjacent ions in a polar semiconductor. Indeed, the presence of a hetero-
interface will not influence the local LO phonon frequency in the region between two
ions located about two or more monolayers away from the heterointerface. Thus, the
basic energy–wavevector relationship – the dispersion relationship – is expected to
be approximately that of the bulk phonons. The phase space, however, is altered
dramatically by the effects of dimensional confinement: only wavevectors corre-
sponding to multiples of half-wavelengths, as discussed previously, are allowed.
7.6 Continuum model for acoustic modes in

dimensionally confined structures
In this section, the elastic continuum model of Section 7.2 is applied to the analysis
of acoustic phonon modes in a variety of nanostructures. The acoustic phonon
amplitudes and dispersion relations are determined for free-standing layers as well
as for double-interface structures. In addition, the acoustic phonon modes are
analyzed for both rectangular and cylindrical wires. For the case of hollow wires,
the acoustic phonon modes are applied to the analysis of acoustic disturbances in
the microtubulin structures found in many biological systems. Finally, the elastic
continuum model is applied to describe acoustic phonon modes in quantum wires.
7.6.1 Acoustic phonons in a free-standing and

unconstrained layer
The acoustic phonons in a free-standing and unconstrained layer illustrate key fea-
tures of the confined modes in dimensionally confined structures. Such free-standing
layers are referred to occasionally as free-standing quantum wells in the case when
the layer is thin enough for quantum confinement to modify the properties of the
electron de Broglie waves in the layer. From Section 7.2 the three force equations
(7.33) describing the amplitudes of the phonon displacements may be written as the
vector equation
∂ 2 u/∂t 2 = ct2 ∇ 2 u + (cl2 − ct2 ) grad , (7.130)
where
= ∇ · u = div u, cl2 = (λ + 2µ)/ρ, and ct2 = µ/ρ.

(7.131)
Let us consider a slab of width a and free standing in space with unconstrained
surfaces at z = ±a/2. At such surfaces the material displacements are unrestricted
but the normal components of the stress tensor – the traction force – must vanish;
with the expressions for stress given in Section 7.2 it follows that at z = ±a/2
(Bannov et al., 1994a, b, 1995),

∂u x ∂u z
Tx,z = µ + = 0,
∂z ∂x

∂u y ∂u z
Ty,z =µ + = 0, (7.132)
∂z ∂y
∂u z
Tz,z = λ + 2µ = 0.
∂z
Taking the displacement eigenmodes to be of the form
2 ∞
1
u(r, t) = d 2 q eiq ·r un (q , z), (7.133)
2π n −∞
the force equations reduce to an eigenvalue problem
Dun (q , z) = −ωn2 un (q , z), (7.134)
where
 2 
2 d d
 t dz 2 − cl qx (cl2 − ct2 )iqx
2 2
c 0
 dz 

 d2 
D=
 0 ct2 − ct2 qx2 0 ,
 (7.135)
 dz 2 
 2 
d d
(cl2 − ct2 )iqx 0 cl 2 − ct qx
2 2 2
dz dz
with boundary conditions corresponding to zero traction force at z = ±a/2
du x du y du z c2 − 2c2
= −iqx u z , = 0, and = −iqx l 2 t u x .
dz dz dz cl
(7.136)
As discussed in Appendix A, it is convenient to consider the quantity w =

√ √
ρu; for acoustic modes the quantity ρ replaces µN /V , which applies for
optical modes, µ being the reduced mass. Accordingly, let us take the orthogonal
eigenvectors for these modes to satisfy

w†n (q , z) · wm (q , z)dz = 0, n = m. (7.137)
From classical acoustics (Auld, 1973) it is known that the problem at hand admits
to three types of solution: shear waves, dilatational waves, and flexural waves. First,
consider the shear waves. For these modes, the only component of the displacement
is parallel to the surfaces z = ±a/2; taking this non-zero component to be in the
y-direction, we have un (q , z) = (0, u y , 0) with
!
cos qz,n z for n = 0, 2, 4, . . .
uy = (7.138)
sin qz,n z for n = 1, 3, 5, . . .
and qz,n = nπ/a. These transverse modes are designated as rotational modes. The
frequency–wavevector relation for these shear waves is
#
ωn = ct qz,n
2 + q 2.
x (7.139)
Clearly, these shear modes have wavelengths such that an integral number n of
half wavelengths fits into the confinement region of length a. In nanoscale crystalline
layers the number of half-wavelengths n is limited to the number of unit cells in the
thickness a. Thus, for an elemental semiconductor layer with Nm monolayers in the
thickness a, n takes on integer values from 0 to Nm .
The second class of solutions is associated with so-called dilatational modes.
These dilatational modes are irrotational modes and they are associated with
compressional distortions of the medium. The compressional character of these
modes leads to local changes in the volume of the medium. They have two non-zero
components: un (q , z) = (u x , 0, u z ) with
$ qa ql a %
t
u x = iqx (qx2 − qt2 ) sin cos ql z + 2ql qt sin cos qt z ,
2 2
2 q a q a 3 (7.140)
t l
u z = ql −(qx2 − qt2 ) sin sin ql z + 2qx2 sin sin qt z ,
2 2
where ql and qt are solutions of
tan qt a/2 4qx2 ql qt

=− ,
tan ql a/2 (qx2 − qt2 )2 (7.141)
cl2 (qx2 + ql2 ) = ct2 (qx2 + qt2 ).
For each value of qx this pair of equations has either pure imaginary or real
solutions, denoted by ql,n (qx ) and qt,n (qx ); here, the label n is used to denote the
different branches of the solutions ql,n (qx ) and qt,n (qx ). The dilatational modes have
frequencies ωn satisfying
# #
ωn = cl ql,n
2 + q2 = c
x t qt,n + q x .
2 2 (7.142)
Numerical solutions of these dispersion relations were given by Bannov et al. (1995)
for a 100-ångstrom-wide GaAs slab under the assumption that cl = 5.7×105 cm s−1
and ct = 3.35 × 105 cm s−1 .
Finally, the third class of solutions is referred to as the flexural modes. These
flexural modes are of the form un (q , z) = (u x , 0, u z ) with
2 qt a ql a 3
u x = iqx (qx2 − qt2 ) cos sin ql z + 2ql qt cos sin qt z ,
2 2
2 q a q a 3 (7.143)
t l
u z = ql (qx2 − qt2 ) cos cos ql z − 2qx2 cos cos qt z ,
2 2
ql and qt being determined as solutions of the pair of equations
tan ql a/2 4qx2 ql qt

=− ,
tan qt a/2 (qx2 − qt2 )2 (7.144)
cl2 (qx2 + ql2 ) = ct2 (qx2 + qt2 ).
Just as for the dilatational modes, this pair of equations for the flexural modes admits
solutions of the form ql,n (qx ) and qt,n (qx ), where n labels the different branches of
the solutions. These modes are normalized, according to the procedures of Section
5.1, in terms of wn (q , z) instead of un (q , z) since, as mentioned above, the
√
considerations of Appendix A make it clear that it is convenient to use wn = ρun :

h̄
u(r) = †
(an,q + an,−q )wn (q , z) eiq ·r . (7.145)
q ,n 2L 2 ρωn (q )
This last result is, of course, consistent with the normalization condition of subsec-
tion 7.3.1,

√ √ h̄
L 2 dz{ µnu(q, z)}∗ · { µnu(q, z)} = ; (7.146)
2ω(q)
this equivalence follows straightforwardly by noting that q = q and by taking

ρ = µn, as is appropriate since the mass density ρ is clearly the appropriate quantity
for acoustic modes.
7.6.2 Acoustic phonons in double-interface

heterostructures
Wendler and Grigoryan (1988) considered the acoustic phonon modes supported by
an isotropic slab of density ρ1 and width a bounded at z = ±a/2 by semi-infinite
embedding materials of density ρ2 . The slab of width a is taken to have transverse
and longitudinal sound speeds ct1 and cl1 respectively. The sound speeds of the two
identical embedding materials are taken to be ct2 and cl2 respectively. According to
the cardinal boundary conditions of classical acoustics, the material displacements
u and the normal components of the stress tensor Ti3 , also known as the traction
force, must be continuous at z = ±a/2. The Young’s modulus, E α , and the Poisson
ratio, να , for medium α may be expressed in terms of the Lamé constants as
µα (3λα + 2µα ) λα
Eα = and να = (7.147)
λα + µα 2(λα + µα )
where α = 1 for the slab and α = 2 for the embedding materials. Then, the inverse
relations are
E α να Eα
λα = and µα = , (7.148)
(1 − 2να )(1 + να ) 2(1 + να )
and it follows that
Eα λα
λ α + µα = and να = . (7.149)
2(1 − 2να )(1 + να ) 2(λα + µα )
Writing the displacement field u as the sum of its longitudinal part, ul , satisfying
∇ × ul = 0, and its transverse part, ut , satisfying ∇ · ut = 0, it follows from the
wave equation (7.130) of subsection 7.6.1 that
∂2 ∂2
ul − clα
2
ul = 0 and ut − ctα
2
ut = 0, (7.150)
∂t 2 ∂t 2
where
E α (1 − να ) Eα
2
clα = and 2
ctα = . (7.151)
ρα (1 + να )(1 − 2να ) 2ρα (1 + να )
Following Wendler and Grigoryan (1988), the media are assumed to be isotropic
and, without loss of generality, we may consider acoustic modes propagating in
the x-direction with wavevector q . Wendler and Grigoryan (1988) classified the
acoustic modes for such an embedded quantum well as symmetric shear vertical
waves, antisymmetric shear vertical waves, symmetric shear horizontal waves, and
antisymmetric shear horizontal waves. Defining u(z) through the relationship

u(x, y, z) = u(z) · exp i(q x − ωt) , (7.152)
it is possible to consider all the acoustic phonons as belonging to two classes of

waves: shear vertical (SV) modes with two non-zero components,

u(z) = u 1 (z), 0, u 3 (z) , (7.153)
and shear horizontal (SH) modes with

u(z) = 0, u 2 (z), 0 . (7.154)
As in subsection 7.6.1, the symmetric modes satisfy
u 1 (z) = u 1 (−z), u 2 (z) = u 2 (−z), u 3 (z) = −u 3 (−z), (7.155)

and the antisymmetric modes satisfy
u 1 (z) = −u 1 (−z), u 2 (z) = −u 2 (−z), u 3 (z) = u 3 (−z). (7.156)
The localized modes for the embedded slab under consideration must satisfy the
boundary conditions
u(z)|z=±∞ = 0. (7.157)
Then from the wave equations for the displacements, the symmetric shear vertical
(SSV) modes must have the form


 A S exp (−ηl2 z) + B2S exp (−ηt2 z) z > a/2,
 2
u 1S (z) = A1S cosh ηl1 z + B1S cosh ηt1 z a/2 > z > −a/2,


 A S exp η z + B S exp η z z < −a/2,
2 l2 2 t2

 ηl2 S q S

 i A exp (−ηl2 z) + B exp (−ηt2 z) z > a/2,

 q 2 ηt2 2


 ηl1 S q S
u 3 (z) =
S
i A sinh ηl1 z − B sinh ηt1 z a/2 > z > −a/2,

 q 1 ηt1 1



 ηl2 S q S

 i A exp (ηl2 z) − B exp (ηt2 z) z < −a/2.
q 2 ηt2 2
(7.158)
and the antisymmetric shear vertical (ASV) modes have the form


 A A exp (−ηl2 z) + B2A exp (−ηt2 z) z > a/2,
 2
u 1 (z) =
A
A1 sinh ηl1 z + B1 sinh ηt1 z
A A
a/2 > z > −a/2,


 −A A exp η z − B A exp η z z < −a/2,
2 l2 2 t2

 ηl2 A q A

 i A exp (−η z) + B exp (−η z) z > a/2,

 q 2
l2
ηt2 2
t2


 ηl1 A q A
u 3 (z) =
A
i − A cosh ηl1 z − B cosh ηt1 z a/2 > z > −a/2,

 q 1 ηt1 1



 ηl2 A q A

 i A exp ηl2 z − B exp ηt2 z z < −a/2.
q 2 ηt2 2
(7.159)
The functions ηlα and ηtα are defined by
ηlα = (q2 − ω2 /clα )

2 1/2
and ηtα = (q2 − ω2 /ctα ) ,
2 1/2
(7.160)
where α = 1 for the slab and α = 2 for the embedding materials. The conditions
ηlα = 0 and ηtα = 0 are again recognized as the bulk dispersion relations for
medium α for the longitudinal and transverse acoustic modes, respectively. From
the definitions of the stress, Ti j , and strain, Si j , of Section 7.2, it follows that for
medium α
%
(α) Eα (α) να $ (α) (α) (α)
Ti j = S + S + S22 + S33 δi j . (7.161)
1 + να i j 1 − 2να 11
From the expressions (7.158) and (7.159) for u 1S (z), u 3S (z), u 1A (z), and u 3A (z) it then
follows that
(α) (α)
T13 = 2ρa ctα
2
S13 ,
(α)
T23 = 0, (7.162)
(α) (α) 2 (α)
T33 = ρa (clα
2
− ctα
2
)S11 + ρa clα S33 .
(α)
Requiring that Ti3 and u be continuous at z ± a/2 yields four equations. The
determinant of these equations, det dmn , then yields the dispersion relations for the
SSV modes. In dmn , m and n take on the the values 1, 2, 3, and 4, and the dmn are
given by

ηl2 a ηt2 a
d11 = exp − , d12 = exp − ,
2 2
ηl1 a ηt1 a
d13 = − cosh , d14 = − cosh ,
2 2

ηl2 a q2 ηt2 a
d21 = −ηl2 exp − , d22 = − exp − ,
2 ηt2 2
ηl1 a q2 ηt1 a
d23 = −ηl1 sinh , d24 = − sinh ,
2 ηt1 2

ηl2 a η2
2 t2
+ q2 ηt2 a
d31 = −2ρ2 ct2 ηl2 exp −
2
, d32 = −ρ2 ct2 exp − ,
2 ηt2 2
ηl1 a η2
2 t1
+ q2 ηl1 a
d33 = −2ρ1 ct1
2
ηl1 sinh , d34 = −ρ1 ct1 sinh,
2 ηt1 2
$ %
ηl2 a ηt2 a
d41 = ρ2 ct2
2
ηt2
2
+ q2 exp − , d42 = 2ρ2 ct2
2 2
q exp − ,
2 2
$ %
ηl1 a ηt1 a
d43 = −ρ1 ct1 ηt1 + q cosh
2 2 2
, d44 = −2ρ1 ct1 q cosh
2 2
.
2 2
(7.163)
For the ASV modes, the corresponding elements dmn are given in terms of these
results by making the replacements cosh sinh, d14 → −d14 , d24 → −d24 , d31 →
−d31 , d32 → −d32 , d33 → −d33 , and d44 → −d44 . As in Wendler and Grigoryan
(1988), A1S,A , A2S,A , B1S,A , and B2S,A are then given in terms of the new elements
dmn by

−d14 d12 d13
1
A1S,A = −d24 d22 d23 ,
det dmn
−d34 d32 d33

d11 d12 −d14
1
A2S,A = d21 d22 −d24 , (7.164)
det dmn
d31 d32 −d34

d11 −d14 d13
1
B1S,A = d21 −d24 d23 ,
det dmn
d31 −d d
34 33
and
B2S,A = 1, (7.165)
where in det dmn m and n take on the values 1, 2, and 3.

The SSV modes are of special interest because they possess a non-zero defor-
mation potential. They will be considered further in Chapter 10 in connection with
the radiation of coherent acoustic phonons from an electron current in an embedded
layer. The shear horizontal (SH) modes may be determined by a derivation analo-
gous to that for the SV modes (Wendler and Grigoryan, 1988) and take the form


 D S exp (−ηt2 z) z > a/2,
 2
u 2 (z) =
S
D1S cos θt1 z a/2 > z > −a/2, (7.166)


 D S exp η z z < −a/2
2 t2
for the SSH modes and



 D A exp (−ηt2 z) z > a/2,
 2
u 2A (z) = D1A sin θt1 z a/2 > z > −a/2, (7.167)


 −D A exp η z z < −a/2,
2 t2
for the ASH modes, where

$ %
θt1 = ω2 /ct1
2
− q2 = iηt1. (7.168)
By procedures analogous to those for the SSV modes, it follows that the
dispersion relation for the SSH modes is given by
2 η
ρ2 ct2 t2 θt1 a
− tan = 0. (7.169)
ρ1 ct1
2 θ
t1 2
The dispersion relation for the ASH modes is given by

2 θ
ρ1 ct1 t1 θt1 a
+ tan = 0, (7.170)
ρ2 ct1
2 η
t2 2
and the amplitude coefficients for the displacement fields are given by
ηt2 a θt1 a ηt2 a θt1 a
D1S = D1A = 1, D2S = exp cos , D2A = exp sin .
2 2 2 2
(7.171)
The SH modes are seen to exist only in the region ω2 /ct2 2 < q 2 < ω2 /c2 .
t1
Wendler and Grigoryan (1988), see also Mitin et al. (1999), classified the
localized acoustic modes in a symmetrical embedded layer in terms of the regions
in the ωq -plane. They found that localized acoustic modes exist in this structure
provided that ct1 < ct2 , cl1 , cl2 . These localized modes propagate along the layer
and decay outside the layer. For such a symmetrical embedded layer, Wendler
and Grigoryan (1988) give many numerical results for the dispersion relations
and mode amplitudes for the localized symmetric shear vertical waves (SSVWs),
antisymmetric shear vertical waves (ASVWs), symmetric shear horizontal waves
(SSHWs), and antisymmetric shear horizontal waves (ASHWs). As mentioned
previously, the SV modes will be considered further in Chapter 10 in connection
with the radiation of coherent acoustic phonons from an electron current in an
embedded layer.
7.6.3 Acoustic phonons in rectangular quantum wires

The classical compressional acoustic modes in free-standing rods with rectangular
cross sections have been examined experimentally (Morse, 1948) and theoretically
(Morse, 1949, 1950). The solutions obtained by Morse are based on the elastic
continuum model as well as on the approximation method of separation of variables.
As illustrated previously, these classical elastic continuum solutions provide the
basis for describing the compressional – that is, the longitudinal – phonon modes
in a nanoscale quantum wire with a rectangular cross section. For cross-sectional
dimensions with aspect ratios of approximately two or greater, Morse (1948,
1949, 1950) found that these solutions provide simple and accurate analytical
expressions in agreement with the experimentally observed modes over a wide
range of conditions. Consider a free-standing rectangular rod of infinite length in
the z-direction with an x-directed height 2a and a width 2d in the y-direction, as
shown in Figure 7.18.
Taking the origin of the coordinates in the geometric center of the of the
x y-plane, the acoustic mode displacements determined by Morse are given by
u(x, y, z) = (u 1 , v1 , w1 ), where
u 1 = u(x, y) eiγ (z−ct) , v1 = v(x, y) eiγ (z−ct) , w1 = w(x, y) eiγ (z−ct) .
(7.172)
Here c is the phase velocity and γ is the wavevector in the z-direction. Morse
considered the wave equations

∂2 ∂2 ∂2 ∂2
+ 2 + ql φ = 0,
2
+ 2 + ql ψi = 0,
2
(7.173)
∂x2 ∂y ∂x2 ∂y
by writing solutions in the form of a product of trigonometric functions:

! "! " ! "! "
sin q1 x sin h 1 y sin q2 x sin h 2 y
φ∝ , ψ∝ (7.174)
cos q1 x cos h 1 y cos q2 x cos h 2 y
where
ql2 = q12 + h 21 and qt2 = q22 + h 22 . (7.175)
The boundary conditions at x = ±a and y = ±d are taken as
Tx x = 0, Tyx = 0, Tzx = 0 at x = ±a,

(7.176)
Tyy = 0, Tx y = 0, Tzy = 0 at y = ±d.
These boundary conditions cannot be satisfied completely. Morse adopted a simple

approach to satisfying them. Specifically, he noted that the simplest way to attempt
to meet such conditions is to make the stress components factor into products of
functions of x and y. This is possible if either q1 = q2 or h 1 = h 2 .
The approximate separation-of-variables solution given by Morse for the com-
pressional modes in the case where h = h 1 = h 2 is
Figure 7.18. Rectangular

quantum wire of width 2d in
the y-direction and height 2a
in the x-direction. From Yu
Physical Society.
u = (A sin q1 x + B sin q2 x) cos hy,

h
v= A cos q1 x + C cos q2 x sin hy,
q1 (7.177)

γ 1
w = i − A cos q1 x + (q2 B + hC) cos q2 x cos hy,
q1 γ
where
q12 + h 21 = γ 2 [(c/cl )2 − 1], q22 + h 22 = γ 2 [(c/ct )2 − 1], (7.178)
ρ is the density of the elastic medium, and the longitudinal and transverse sound
speeds are given by

cl = (λ + 2µ)/ρ and ct = µ/ρ. (7.179)
The modes associated with this case are known as the ‘thickness modes’, as
designated by Morse, who showed that h = h 1 = h 2 leads to an adequate
description of the experimentally determined modes when d ≥ 2a (Morse, 1948).
Using the expressions (7.177) for u, v, and w to evaluate Tx x = Tyx = Tzx = 0 at
x = ±a, it follows that
 
2h sin q1 a h sin q2 a q2 sin q2 a  
  A
 −(γ 2 + h 2 − q 2 ) cos q1 a 2q1 q2 cos q2 a 0  B 
 2 
2(h 2 + γ 2 ) sin q1 a (γ 2 + h2 − q22 ) sin q2 a 0 C
= 0. (7.180)
The dispersion relation for q2 = 0 is given by the expression resulting from the
condition that the determinant of the coefficients vanishes, that is,
tan q2 a 4q1 q2 (h 2 + γ 2 )
=− , (7.181)
tan q1 a (h 2 + γ 2 − q22 )2
which is similar to the dispersion relation for a free-standing layer discussed in

subsection 7.6.1. For calculating the acoustic phonon frequencies as functions of
the wavevector, γ , it is convenient to rewrite this dispersion relation as

tan π χ 2 − ψ 2 4ψ 2 χ 2 − ψ 2 χ 2 − ψ 2
=− 2 , (7.182)
tan π χ 2 − ψ 2 2ψ 2 − χ 2

whereχ 2 and ψ 2 are related to q1 and q2 through q1 = (π/a) χ 2 − ψ 2 , q2 =
(π/a) χ 2 − ψ 2 and in accordance with ωγ2 = cl (γ 2 + h 2 + q12 ) and ωγ2 = ct (γ 2 +
h 2 +q22 ), where = (ct /cl )2 = (1−2σ )/(2−2σ ). Defining s = aγ /π and recalling

that ωγ = cγ , it follows that

c 2 $ a %2 2 c 2 $ a %2 ω 2
χ 2 = s2 = γ = ,
ct π ct π ct
2 $ %
ah a 2 2
ψ =s +
2 2
= (γ + h 2 ), (7.183)
π π
$ π %2 $π % 2
r̃
γ2 = ψ 2 − h2 = ψ2 −
a a 2
with r̃ = a/d.
Solving (7.180) for B and C in terms of A, it follows that
u 1 = A(sin q1 x + α sin q2 x) cos hy eiγ (z−ct) ,

h
v1 = A cos q1 x + β cos q2 x sin hy eiγ (z−ct) ,
q1 (7.184)

γ 1
w1 = i A − cos q1 x + (q2 α + hβ) cos q2 x cos hy eiγ (z−ct) ,
q1 γ
where
sin q1 a 2(h 2 + γ 2 )
B = αA = − A,
sin q2 a (γ 2 + h 2 − q22 )
(7.185)
q2 h sin q1 a 2q2 h
C =− B= A = β A.
h2 + γ 2 sin q2 a (γ 2 + h 2 − q22 )
The remaining constant A is determined by quantizing the phonon amplitude

according to
a d
1 h̄
dx dy (u ∗ u + v ∗ v + w ∗ w) = , (7.186)
4ad −a −d 2Mωγ
where ωγ is the angular frequency of the mode with wavevector γ . Performing the
indicated integrations, it follows that
! 2 3
A2
f 1 (h, d) f 2 (q1 , a) + 2αg1 (q1 , q2 , a) + α 2 f 2 (q2 , a)
4ad
/ 0
h2 2βh
− f 1 (h, d) 2 f 1 (q1 , a) + g2 (q1 , q2 , a) + β 2 f 1 (q2 , a)
q1 q1
2
γ 2
+ f 1 (h, d) 2 f 1 (q1 , a) − (q2 α + hβ)g2 (q1 , q2 , a)
q1 q 1

q1 α + hβ 2
+ f 1 (q2 , a)
γ2
/ 0"
h2 2βh
+ 2d f 1 (q1 , a) + g2 (q1 , q2 , a) + β f 1 (q2 , a)
2
q12 q1
h̄
= , (7.187)
2Mωγ
with

sin 2hd
f 1 (h, d) = d 1 + , f 2 (h, d) = 2d − f 1 (h, d),
2hd
sin(q1 − q2 )a sin(q1 + q2 )a
g1 (q1 , q2 , a) = − , (7.188)
q1 − q2 q1 + q2
sin(q1 − q2 )a sin(q1 + q2 )a
g2 (q1 , q2 , a) = + .
q1 − q2 q1 + q2
It is convenient to define a new normalization constant Bγ , through
2h̄
A2 = ≡ A2γ . (7.189)
Mωγ Bγ
As discussed by Morse, h must be chosen to satisfy the boundary condition on

the stress components at y = ±d, that is, Tyy = Tx y = Tzy = 0. This can be
accomplished for d 2a since in this case Tx y and Tzy become negligible; with
Tyy = 0 this implies that

hd = n + 12 π, n = 0, 1, 2, . . . . (7.190)
The principal propagation mode has no nodal surfaces parallel to the length of
the quantum wire; this corresponds to the case n = 0. Morse found close agreement
between theory and experiment for a/d = 1/8 and as expected less agreement
for a/d = 1/2. In addition to the ‘thickness modes’ another set of modes was
observed experimentally by Morse (1948, 1950). These modes are known as ‘width
modes’ and are determined by a procedure used to analyze the ‘thickness modes’.
Specifically, Morse took q1 = q2 = q and obtained a set of equations similar to
those for u 1 , v1 , and w1 but with x and y interchanged. By imposing the boundary
conditions at y = ±d, the ‘width modes’ were found to have a dispersion relation
identical in form to that for the ‘thickness modes.’ The dispersion curves for selected
acoustic modes are shown in Figure 7.19 for a 28.3 Å × 56.6 Å GaAs quantum wire
and in Figure 7.20 for a 50 Å × 200 Å GaAs quantum wire.
For carriers at the non-degenerate point in band α, E α (k), the deformation-
potential interaction Hamiltonian Hdef α is given in terms of the displacement operator
û(r) by
α
Hdef = E 1α ∇ · û(r). (7.191)
At such a symmetry point, only the irrotational – that is, the longitudinal – compo-
α . Accordingly, only the the potential φ contributes
nents of u(r) contribute to Hdef
α . Since there are multiple modes for a given value of n, another index, m,
to Hdef
is needed to describe the phonon spectrum at each value of γ . For the case of a
quantum wire, the quantization of the acoustic phonons may be performed by taking
1
û(r) = √ [u(γ , x, y)aγ + c.c.] eiγ z , (7.192)
L n,m,γ
where the components of u(γ , x, y) = (u, v, w) were normalized previously over
the area 4ad. The deformation potential is then given by

α E 1α ∂u ∂v
Hdef = √ [an,m (γ ) + an,m (−γ )] + + iγ w eiγ z ,
L n,m,γ ∂x ∂y
(7.193)
Figure 7.19. The six lowest-order (m = 1, 2, . . . , 6) width modes (solid lines) and
thickness modes (broken lines) for a 28.3 Å × 56.6 Å GaAs quantum wire. From Yu
et al. (1994), American Physical Society, with permission.
and, upon applying the Fermi golden rule, these combinations lead to conditions
enforcing the conservation of energy.
The Hamiltonian is independent of time. In Chapters 8 and 9, such time-
independent carrier–phonon Hamiltonians will be used in applying the Fermi golden
rule to calculate carrier–phonon scattering rates. The carrier–phonon interaction also
has a time dependence of the form eiωγ t , where ωγ is the phonon frequency. As will
become obvious in Chapters 8 and 9, such time-dependent factors are combined
with the time-independent factors of carrier wavefunctions. Since
∂u
= A(q1 cos q1 x + αq2 cos q2 x) cos hy,
∂x

∂v h2
=A cos q1 x + β cos q2 x cos hy, (7.194)
∂y q1
/ 0
γ2
−iγ w = A − cos q1 x + (q2 α + hβ) cos q2 x cos hy,
q1
it follows that

∂u ∂v h2 γ2
+ + iγ w = A q1 + + cos q1 x cos hy
∂x ∂y q1 q1
1 $ 2 %
=A q1 + h 2 + γ 2 cos q1 x cos hy
q1
ωγ
= A 2 cos q1 x cos hy (7.195)
cl q1
and, accordingly,
α Eα ωγ
Hdef = √1 A 2 cos q1 x cos hy[an,m (γ ) + an,m (−γ )] eiγ z .
L n,m,γ cl q1
(7.196)
In Chapter 8 this expression for the deformation-potential interaction will be used to

calculate electron–acoustic-phonon scattering rates for a rectangular quantum wire.
7.6.4 Acoustic phonons in cylindrical structures

The acoustic phonons in a cylindrical waveguide and in a cylindrical shell il-
lustrate key features of the confined modes in dimensionally confined structures.
The cylindrical waveguide is of obvious practical importance. Furthermore, the
cylindrical shell is of interest because it approximates a single-walled buckytube
and also because it resembles the microtubuline structure found in many parts
of the human body. As discussed previously in this section and in Section 7.3,
the elastic continuum model provides an approximate description of the acoustic
phonon modes in such dimensionally confined nanostructures. The force equations
for a cylindrical elastic medium may be written as (Auld, 1973; Sirenko et al., 1995)
∂ 2 ur ∂ Trr 1 ∂ Tr ϕ ∂ Tr z Trr − Tϕϕ

ρ = + + + ,
∂t 2 ∂r r ∂ϕ ∂z r
∂ 2uϕ ∂ Tr ϕ 1 ∂ Tϕϕ ∂ Tϕz 2Tr ϕ

ρ = + + + , (7.197)
∂t 2 ∂r r ∂ϕ ∂z r
∂ 2 u zϕ ∂ Tr z 1 ∂ Tϕz ∂ Tzz Tr z
ρ = + + + ,
∂t 2 ∂r r ∂y ∂z r
where the axis of the cylinder is oriented along the z-direction, ϕ is the azimuthal
angle, and r is the radial coordinate of the cylindrical structure. As before, the stress
tensor T is related to the strain tensor S through the Hooke’s law relationship
Tµv = λSµµ δµν + 2µSµν ; (7.198)
in this stress–strain relation, λ and µ are the Lamé constants. Alternatively, these
force equations are frequently written in the form
Figure 7.20. The six lowest-order (m = 1, 2, . . . , 6) width modes (solid lines) and
thickness modes (broken lines) for a 50 Å × 200 Å GaAs quantum wire. From Yu

∂ ∂u k (r, t)
ρ(r)ü i (r, t) = λi jkl (r) , (7.199)
∂ xj ∂ xl
where the elastic stiffness tensor for a particular isotropic medium is expressed as
λi jkl = λδi j δkl + 2µδik δ jl . (7.200)
These equations are more complicated than their counterparts in rectilinear co-
ordinates, Section 7.2. Indeed, the additional complexity of the force equations
in cylindrical coordinates is a direct consequence of the fact that in curvilinear
coordinates the basis vectors are coordinate dependent.
Consider the acoustic phonon modes in a cylindrical waveguide of radius a
embedded in an elastic medium. Both of these media are taken to be isotropic. From
the normalization procedures of Section 5.1, the modes are normalized in terms
of w instead of u since the considerations of Appendix A make it clear that it is
√
convenient to use w = ρu; the displacement operator û(r) is then given by

h̄
û(r) = [wmn,q (r )amn,q + w∗mn,−q (r )amn,−q
†
] eimϕ+iqz/a .
q,mn 2Lρωmn (q)
(7.201)
The quantum number n labels modes with the same m and q in the set wmn,q (r ),
where q represents the z-component of the wavevector qz . In determining the
normalization constants for the normal modes wmn,q (r ), it is convenient to write
1
wmn,q (r) = wmn,q (r ) eimϕ+iqz/a = √ u(r ) eimϕ+iqz/a , (7.202)
πa 2 N
where u(r ) is the classical displacement given by the elastic continuum model and
the normalization constant N is then determined by the normalization condition

d 2r ρ(r)w∗n,m,q (r) · wn ,m ,q (r) = δn,m,q;n ,m ,q , (7.203)
and q ≡ aqz .
Let the density and Lamé constants of the cylindrical waveguide be ρ1 , λ1
and µ1 respectively, and those of the surrounding material be ρ2 , λ2 and µ2 . The
general solution of the classical elastic continuum equations for such a cylindrical
structure may be written (Beltzer, 1988; Stroscio et al., 1996) in terms of three scalar
potentials φ, ψ, and χ as
u = ∇φ + ∇ × êz ψ + a∇ × ∇ × êz χ, (7.204)
where êz is a unit vector along the z-direction. The second and third terms in this
last result correspond to the usual irrotational contribution to u, expressed as a sum
of two mutually normal vectors. The potentials φ, ψ, and χ satisfy scalar wave
equations with longitudinal and transverse sound speeds given by

clξ = (λξ + 2µξ )/ρξ and ctξ = µξ /ρξ ; (7.205)
the subscript ξ takes on the value 1 to designate the material parameters of the
cylinder and the subscript 2 to designate those of the surrounding material. Solutions
of the classical elastic continuum equations are sought with vibration frequency ω,
wavevector qz = q/a, and azimuthal quantum number m. Seeking acoustic modes
confined near the cylindrical waveguide, the scalar potentials for r < a are taken to
be
   
φ ibl1 Jm (kl r/a)
1
 ψ  =  Bt1 Jm (kt r/a)  eimϕ+iqz/a−iωt . (7.206)
a
χ bt1 Jm (kt r/a)
Outside the cylindrical waveguide, where r > a, the solutions are taken to be
   
φ ibl2 K m (κl r/a)
1
 ψ  =  Bt2 K m (κt r/a)  eimϕ+iqz/a−iωt , (7.207)
a
χ bt2 K m (κt r/a)
where kl,t and κl are defined by
2
kl,t = q 2 − ω2 a 2 /c(l,t)1
2
and κl,t
2
= ω2 a 2 /c(l,t)2
2
− q2 (7.208)
and the bl1 etc. are normalization constants, to be determined. In the expressions for
φ, ψ, and χ, it is assumed that kl,t 2 > 0 and κ 2 > 0, since confined acoustic modes
l,t
are desired. Substituting these potentials into the general expression for u it follows
that

kl r a kt r kt r
−iu r (r ) = bl1 kl Jm + Bt1 m Jm + bt1 qkt Jm ,
a r a a

a kl r kt r a kt r
−u ϕ (r ) = bl1 m Jm + Bt1 kt Jm + bt1 mq Jm ,
r a a r a

kl r kt r
−u z (r ) = bl1 q Jm − bt1 kt Jm
2
a a
(7.209)
for r < a and
a
−iu r (r ) = bl2 kl K m (κl r/a) + Bt2 m K m (κt r/a) + bt2 qκt K m (κt r/a),
r

a κl r κt r a κt r
−u ϕ (r ) = bl2 m K m + Bt1 κt K m + bt1 mq K m ,
r a a r a
(7.210)

κl r κt r
−u z (r ) = bl2 q J K m − bt1 κt2 K m
a a
for r > a. By applying the boundary conditions of continuity of displacement and
continuity of the normal components of the stress tensor at r = a, it follows that

U1 −U2 B1
= 0, (7.211)
µ1 F1 −µ2 F2 B2
where Bξ = [blξ , Btξ , btξ ]T and where the ‘displacement’ matrices Uξ are evaluated
at r = a and are given by
 
Gξ myξ qYξ
 −mgξ −Yξ −mqyξ ; (7.212)
−qgξ 0 ktξ yξ
the matrices Fξ are given by

 
−2qG ξ −mqyξ (ktξ
2 − q 2 )Y
ξ
 
 2m(gξ − G ξ ) (ktξ − 2m 2 )yξ + 2Yξ
2 2mq(yξ − Yξ ) ,
(2m 2 + q 2 − klξ
2 )g − 2G
ξ ξ 2mq(yξ − Yξ ) 2q[(m − ktξ )yξ − Yξ ]
2 2
(7.213)

=1 = k(l,t) , k(l,t)ξ =2 = −κ(l,t) , g1 = Jm (kl ), G 1 = kl Jm (kl ), y1 =
2
where k(l,t)ξ 2 2 2
Jm (kt ), Y1 = kt Jm (κt ), g2 = K m (kl ), G 2 = κl K m (κl ), y2 = K m (κt ), and

Y2 = κt K m (κt ). From the components of u(r ), u r (r ), u ϕ (r ), and u z (r ), it is
then straightforward to construct wmn,q (r) and û(r). The case where m = 0 is
of practical interest and leads to relatively simple results. Specifically, for m =
0, submatrices of the 6 × 6 matrix decouple into matrices representing distinct
axisymmetric torsional and radial–axial modes. For the axisymmetric torsional
modes the dispersion relation is given by
J2 (kt ) K 2 (κt )
µ1 k t = µ2 κt (7.214)
J1 (kt ) K 1 (κt )
and the components of the normal modes are given by
!
1 K 1 (κt )J1 (kt r/a) r < a,
wr = 0, wϕ = √ wz = 0,
πa 2 N ϕ J1 (kt )K 1 (κt r/a) r > a,
(7.215)
where the normalization condition gives
Nϕ = ρ1 K 12 (κt )[J12 (kt ) − J0 (kt )J2 (kt )]

+ ρ2 J12 (kt )[(K 0 (κt )K 2 (κt ) − K 12 (κt )]. (7.216)
For the axisymmetric radial–axial modes, the dispersion relation is defined by

 
b1l

Q11 Q12  
 b1t  = 0,
 (7.217)
Q21 Q22 b2l 
b2t
with A = q 2 − kl2 , B = q 2 + κl2 , and

−kl J1 (kl ) −qkt J1 (kt )

Q11 = ,
−q J0 (kl ) kt2 J0 (kt )

κ1 K 1 (κl ) qκt K 1 (κt )

Q12 = ,
q K 0 (κl ) κt2 K 0 (κt )

2µ1 qkl J1 (kl ) µ2 kt (q 2 − kt2 )J1 (kt )

Q21 = ,
µ1 [A J0 (kl ) + 2kl J1 (kl )] 2µ1 qkt [J1 (kt ) − kt J0 (kt )]

−2µ2 qκl K 1 (κl ) −2µ2 κt (κt2 + q 2 )K 1 (κl )

Q22 = .
−µ2 [B K 0 (kl ) + 2κl K 1 (kl )] −µ2 qκt [J (kt ) + κt J0 (kt )]
(7.218)
The components of the normal modes are given by

1 bl1 kl J1 (kl r/a) + bt1 qkt J1 (kt r/a) r < a,
iwr = √
πa 2 N r z bl2 κl K 1 (κl r/a) + bt2 qκt K 1 (κt r/a) r > a,
wϕ = 0,

1 bl1 q J0 (kl r/a) − bt1 kt2 J0 (kt r/a) r < a,
−wz = √
πa 2 N r z bl2 q K 0 (κl r/a) + bt2 κt2 K 0 (κt r/a) r > a.
(7.219)
For the axisymmetric radial–axial modes, the normalization condition requires
that
2 2 2
Nr z = ρ1 b1t {q [J0 (kl ) + J12 (kl )] + kl2 [J12 (kl ) − J0 (kl )J1 (kl )]}
+ b1t kt {kt [J0 (kt ) + J12 (kt )] + q 2 [J12 (kt ) − J0 (kt )J2 (kt )]}
2 2 2 2

− 4b1l b1t qkt J0 (kl )J1 (kt )
2 2 2
+ ρ2 b2l {q [K 1 (κl ) − K 02 (κl )] + κl2 [K 0 (κl )K 2 (κl ) − K 12 (κl )]}
+ b2t κt {κt [K 12 (κt ) − K 02 (κt )] + q 2 [K 0 (κt )K 2 (κt ) − K 12 (κt )]}
2 2 2

− 4b2l b2t qκt K 0 (κt )K 1 (κt ) . (7.220)
As for the case of rectangular quantum wires, subsection 7.6.3, for carriers at
the non-degenerate point in band α, E α (k), the deformation-potential interaction
α is given in terms of the displacement operator û(r) by
Hamiltonian Hdef
α
Hdef = E 1α ∇ · û(r). (7.221)
Again, at such a symmetry point only the irrotational – that is, the longitudinal
– components of u(r) contribute to Hdef α . Accordingly, only the the potential φ,
α
(7.204), contributes to Hdef . Indeed, from the normalized components wmn,q (r)
(7.202), û(r) is obtained readily and by using the relation ∇ 2 φ = −(ω/cl )2 φ it
follows that

ωmn (q) 2 h̄
α α
Hdef = −E 1
q,mn acl 2π LρN ωmn (q)

× mn,q (r )amn,q + ∗mn,−q (r )amn,−q
†
eimϕ+iqz/a , (7.222)
the potential mn,q (r ) being given by

!
ibl1 Jm (kl r/a) r < a,
mn,q (r ) = (7.223)
ibl2 K m (κl r/a) r > a.
Let us consider the case of a thin cylindrical shell. For a cylindrical shell, the
boundary conditions on the inner and outer surfaces are
Pµ = Tµν n ν , (7.224)
where P represents an external pressure that would be present, for example, in the
case where the cylindrical shell is in contact with a liquid, Tµν is the stress tensor,
and n ν is the normal to the surface of the shell. In particular, P = ±P in êr and
out
n = ∓êr where êr is the unit vector in the r -direction. The subscripts ‘in’ and ‘out’
are alternatives. For a cylindrical shell of infinite length in the z-direction and of
thickness h and radius R such that h R, the boundary conditions are

Tr ϕ R∓h/2 = Tr z | R∓h/2 = 0, Trr | R∓h/2 = P in . (7.225)
out
Assuming that all quantities except Trr are nearly constant with respect to r over
the interval from R − h/2 to R + h/2, it is possible to show that
Tϕϕ Pin −P out

− + = ρ ü r ,
R h
1 ∂ Tϕϕ ∂ Tϕz
+ = ρ ü ϕ , (7.226)
R ∂ϕ ∂z
1 ∂ Tϕz ∂ Tzz
+ = ρ ü z .
R ∂ϕ ∂z
These results follow straightforwardly by integrating the right- and left-hand sides
of each force equation over the interval from R − h/2 to R + h/2, invoking the
boundary conditions in the radial force equation, and cancelling factors of h. From
the stress–strain relation (7.198), it follows that
Tϕϕ = (λ + 2µ)Sϕϕ + λ(Szz + Srr ),

Tϕz = 2µSϕz ,
Tzz = (λ + 2µ)Szz + λ(Sϕϕ + Srr ), (7.227)
Trr = (λ + 2µ)Szz + λ(Sϕϕ + Szz )
= 0.
These stress–strain relations then imply that

λ
Srr = − (Sϕϕ + Szz ),
λ + 2µ
4µ(λ + µ) 2λµ
Tϕϕ = Sϕϕ + Szz , (7.228)
λ + 2µ λ + 2µ
4µ(λ + µ) 2λµ
Tzz = Szz + Sϕϕ .
λ + 2µ λ + 2µ
The Young’s modulus E and the Poisson ratio ν may be expressed in terms of the
Lamé constants as
µ(3λ + 2µ) λ
E= and ν= . (7.229)
λ+µ 2(λ + µ)
Then the inverse relations are
Eν E
λ= and µ= (7.230)
(1 − 2ν)(1 + ν) 2(1 + ν)
and it follows that
E λ
λ+µ= and ν= . (7.231)
2(1 − 2ν)(1 + ν) 2(λ + µ)
Using these stress–strain relations and eliminating the Lamé constants in favor of
Young’s modulus and the Poisson ratio, the force equations may be written as
E Pin −P out
− (Sϕϕ + ν Szz ) + = ρ ü r ,
R(1 − ν 2 ) h
E ∂ E ∂
(Sϕϕ + ν Szz ) + Sϕz = ρ ü ϕ , (7.232)
R(1 − ν ) ∂ϕ
2 1 + ν ∂ϕ
E ∂ E ∂
Sϕz + (Szz + ν Sϕφ ) = ρ ü z .
R(1 + ν) ∂ϕ 1 − ν 2 ∂z
Then using the relations (Auld, 1973)

1 ∂u ϕ ∂u z 1 ∂u z ∂u ϕ
Sϕϕ = + ur , Szz = , and Sϕz = + ,
r ∂ϕ ∂z r ∂ϕ ∂z
(7.233)
the force equations may be written as

u r + u ϕ,ϕ νu z,z p ü r
− − + = 2,
R2 R R c
u r,ϕ + u ϕ,ϕϕ 1−ν 1+ν ü ϕ
+ u ϕ,zz + u z,ϕz = 2 , (7.234)
R2 2R 2R c
νu r,z 1+ν 1−ν ü z
+ u ϕ,ϕz + 2
u z,ϕϕ + u z,zz = 2 ,
R 2R 2R c
the thin-plate longitudinal sound speed squared, c2 , and the dimensionless pressure
p are defined by

E R 1 − ν2
c =
2
and p= (Pin −P out ). (7.235)
ρ(1 − ν 2 ) h E
These equations have been used (Sirenko et al., 1995) to describe the free
vibrations of a thin shell and the vibrations of such a shell immersed in liquid. For
the case of a free shell, p = 0 and solutions are taken to have the form
   
ur −icr $ %
 u ϕ  =  cϕ  exp imϕ + i q z − iωt , (7.236)
R
uz cz
where the dimensionless wavevector in the z-direction, q, and the dimensionless

frequency are defined by
ωR
q = qz R and = . (7.237)
c
Then the force equations yield
icr − imcϕ − iνqcz = i2 cr ,
mcr − m 2 cϕ − ν− q 2 cϕ − ν+ mqcz = −2 cϕ , (7.238)
νqcr − ν+ mqcϕ − ν− m 2 cz = −2 cz

or
  2 
cr −1 m νq
 cϕ   m 2 − m 2 − ν− q 2 −ν+ mq  = 0.
cz νq −ν+ mq − ν− m − νq
2 2 2
(7.239)
The dispersion relation is then given by

2
−1 m νq

det D = m 2 − m 2 − ν− q 2 −ν+ mq = 0.

νq −ν+ mq 2 − ν− m 2 − νq 2
(7.240)
Hence
6 − [(ν− + 1)(m 2 + q 2 ) + 1]4

+ (m 2 + ν− q 2 )(ν− m 2 + q 2 ) + (1 + ν− )(m 2 + q 2 )
− ν 2 q 2 − m 2 − ν+ m q 2 − ν− (1 − ν 2 )q 4 = 0
2 2 2
(7.241)
or
Aq 4 − Bq 2 + C = 0, (7.242)
where
A = ν− (1 + ν 2 ),
B = ν− (2 − 1)(2 − ν− m 2 ) + (2 − 1)(2 − m 2 )

(7.243)
+ 2νν+ m 2 + ν 2 (2 − m 2 ) + ν+ m (2 − 1) − ν− m 4 ,
2 2
C = (2 − m 2 − 1)(2 − ν− m 2 )2 .

Let us examine these solutions for three several special cases: symmetric az-
imuthal modes with m = 0, small q, and large q.
For the modes with m = 0,
  2 
cr −1 0 νq
 cϕ   0 2 − ν− q 2 0  = 0. (7.244)
cz νq 0 −q
2 2
For cr = cz = 0, the pure torsional mode has frequency

√
torsional ≡ II
m=0 = ν− q. (7.245a)
In addition, for cϕ = 0 it follows that two coupled radial–longitudinal modes have
frequencies
$ %2 2 # 3
2rad−long ≡ I,III
m=0 = 1
2 q 2
+ 1 ± (q 2 − 1)2 + 4ν 2 . (7.245b)
For small q, the q → 0 limit yields

  2 
cr −1 m 0
 cϕ   m − m2
2 0  = 0. (7.246)
cz 0 0 − ν− m
2 2
= 0, the two torsional–radial modes have frequencies torsional−radial,a =

For cz √
Im = m 2 + 1 and torsional−radial,b ≡ III
m = 0. In addition, for cr = cz = 0,
√
it follows that the longitudinal mode has frequency longitudinal ≡ II m = ν− m.
Since m ≈ 0 was obtained in the lowest order in q it is necessary to find the first
III
2
non-vanishing term. For m = 0, by making the assumption that III m=0 = αq 2
and collecting terms up to order q 4 in the dispersion equation, it follows that
−α 2 q 4 + αq 2 (1 + ν− − ν 2 )q 2 − ν− (1 − ν 2 )q 4 = 0. (7.247)
One of the roots of this equation, α = ν− , corresponds

√ to the mode IIm=0 , and
another root, α = 1 − ν , implies that m=0 ≈ 1 − ν 2 q. For the case where
2 III
2
m = 0, by making the assumption III m = αq 4 and collecting terms up to q 4 in
the dispersion relation, it follows that
α[ν− m 4 + (1 + ν− )m 2 − m 2 ]q 4 − ν− (1 − ν 2 )q 4 = 0, (7.248)

so that α = (1 − ν 2 )/(m 4 + m 2 ) and IIIm ≈ (1 − ν 2 )/(m 2 + 1)(q 2 /m). In the
limit of large q the leading terms of the dispersion relation imply that
6 − (1 + ν− )q 2 4 + ν− q 4 2 − ν− (1 − ν 2 )q 4 = 0 (7.249)
or making the guess 2 → αq 2 , it follows that
α 3 − (1 + ν− )α 2 + ν− α = 0, (7.250)
so that α1 = 1, α2 = ν− , and α3 = 0. The result α3 = 0 is inconsistent with

the initial guess, so solutions are of the form 2 → constant. Then from the last
two terms of the dispersion relation, with leading fourth power of q, it follows that
ν− q 4 2 − ν− (1 − ν 2 )q 4 ≈ 0, and 2 → 1 − ν 2 . Convenient interpolation formulae
between the small-q and large-q solutions are given by
#
Im (q) q 2 + m 2 + 1,
#
II
m (q) ν− (q 2 + m 2 ),
q (7.251)
III
m=0 (q) 1 − ν2 ,
1+q
q2
III
m (q) 1 − ν 2 √ .
m m2 + 1 + q2
In the axisymmetric case with m = 0 the I, II, and III modes correspond to pure
radial, torsional, and longitudinal modes respectively. When m = 0, the radial
and torsional modes are coupled. In the limit where q m + 1 the asymptotic
√
√ depend on m; indeed, in this limit m (q) q, m (q) ν− q,
expressions do not I II
m (q)
and III 1 − ν 2 . Analysis of the coefficients cr , cϕ , and cz (Sirenko et al.,
1996a, b) reveals that in the limit of large q the I, II, and III modes correspond to
pure longitudinal, torsional, and radial vibrations, respectively.
The case of a cylindrical shell immersed in fluid was considered by M̌arkus

(1988) and solved numerically by Sirenko et al. (1996b). In this case, p = 0 and the
dispersion relation takes the form
2
(1 + Wm Q ) − 1 m νq

det D = m − m − ν− q
2 2 2 −ν+ mq = 0,

νq −ν+ mq − ν− m − νq
2 2 2
(7.252)
where the term describing the coupling between the shell and the fluid is given by

α Im (Q) K m (Q)
Wm Q = − . (7.253)
Q Im (Q) K m (Q)
Here

2
ω2 c
Q =R
2 2
qz2 − 2 ≡ q2 − 2 , (7.254)
sf sf
s f being the sound speed in the fluid and
Im (−i|Q|) = i −m Jm (|Q|). (7.255)
In the case of a thin cylindrical shell immersed in fluid, interface modes in the fluid
are localized near the cylindrical surface and correspond to the region Q 2 > 0. In
the case where ω > s f qz the acoustic disturbances are radiated from the cylindrical
shell. Indeed, the ωqz -plane is divided into two regions by the curve ω > s f qz . As
just indicated, the region defined by Q 2 > 0, or ω < s f qz , is the region where
interface modes are localized on the scale of R/Q from the cylindrical shell. When
ω > s f qz , Q 2 < 0 and the relation
π (1)
K m (−i|Q|) = i m+1 H (|Q|),
2 m
makes manifest the radiation of cylindrical waves from the cylindrical shell into the
(1)
region surrounding it, since the Hankel function of the first kind, Hm , represents
outgoing cylindrical waves.
The case of a cylindrical shell immersed in fluid was solved numerically by
Sirenko et al. (1996b) in order to model the vibrational behavior of microtubules
(MTs) immersed in water. These microtubules are of great interest in biology as
they are present in many biological structures including cytoskeleton and eukariotic
cells. In these numerical calculations, the inner and outer radii of the MTs are taken
to be 11.5 nm and 14.2 nm respectively, the length of the MTs is taken to be 8 nm,
their mass is taken to be 1.83 × 10−19 g, the Poisson ratio, ν, is taken to be 0.3, and
the Young’s modulus is taken to be 0.5±0.1 GPa. These parameters give a thin-plate
longitudinal sound speed c of 610 m s−1 and a density of 1.47 g cm−3 . The sound
speed and density of water are taken to be 1.50 km s−1 and 1 g cm−3 . The calculated
dispersion relations, im (q), for m = 1, 2, and 3 are shown in Figures 7.21, 7.22,
and 7.23, respectively.
From Figures 7.21–7.23, it is apparent that for qz m/R the mode frequencies
of the immersed MTs tend to those of the free-standing MTs and do not depend on
m. These modes are seen to have maximum frequencies of the order of tens of GHz.
Moreover, the sound speeds of the axisymmetric acoustic modes are in the range
200–600 m s−1 .
Sirenko et al. (1996b) also considered the dynamical behavior of cytoskeletal
filaments, by using the elastic continuum model to determine the mode structure
for the vibrations of a solid cylinder. Particular attention was given to (a) the
axisymmetric torsional mode, (b) the axisymmetric radial–longitudinal mode, and
(c) the flexural mode, as depicted in Figure 7.24.
Figure 7.21. Dispersion relations im for m = 1 modes of the MT sample discussed
in the text. I, pure longitudinal; II, torsional; III, radial. The frequency is in units of
= ω R/c and q = qz R. = 1 yields ω = 7.6 GHz for the parameters given in the
text. The dotted line separating the regions of interface and radiated waves
corresponds to the condition ω = s f qz , as discussed in the text. The solid and broken
lines correspond to MTs immersed in water and free standing respectively. From
Sirenko et al. (1996a), American Physical Society, with permission.
7.6.5 Acoustic phonons in quantum dots

In quantum dots, phonons and carriers alike are modified as a result of abrupt
changes in the material properties at the interface between the quantum dot and
the surrounding material. Indeed, carrier wavefunctions are modified as a result of
the variations in the electron and hole band energies near the boundaries of the
quantum dot. In the case of acoustic phonons, the changes in elastic properties near
the quantum-dot boundaries lead to modifications in the displacement amplitudes.
The acoustic phonon modes for spherical quantum dots and for quantum dots
with rectangular-face confinement have been considered previously (Stroscio et al.,
1994). For the case of a free-standing spherical quantum dot, the quantization of the
acoustic phonons may be performed by taking
1
u(r) = √ [u(q, r)aq + c.c.], (7.256)
N q
and normalizing the acoustic phonon Fourier amplitude, u(q, r, ϕ, z), according to
in the text. I, pure longitudinal; II, torsional; III, radial. The frequency is in units
of ω R/c and q = qz R. = 1 yields ω = 7.6 GHz for the parameters given in the
lines correspond to MTs immersed in water and free standing respectively. From
2π π a
3 h̄
dϕ dθ sin θ dr r 2 u(q, r, θ, ϕ) · u∗ (q, r, θ, ϕ) = .
4πa 3 0 0 0 2Mωq
(7.257)
Here, a is the radius of the quantum dot, N is the number of unit cells in the
normalization volume V , aq is the phonon annihilation operator, q is the phonon
wavevector, ωq is the angular frequency of the phonon mode, M is the mass of the
ions in the unit cell, and r, θ, φ are the usual spherical coordinates. For a quantum
dot with rectangular faces the normalization condition for the acoustic phonon mode
amplitude is given by
a/2 b/2 c/2
1 h̄
dx dy dz u(q, x, y, z) · u∗ (q, x, y, z) = .
abc −a/2 −b/2 −c/2 2Mωq
(7.258)
in the text. I, pure longitudinal; II, torsional; III, radial. The frequency is in units of
= ω R/c and q = qz R. = 1 yields ω = 7.6 GHz for the parameters given in the
lines correspond to MT immersed in water and free standing, respectively. From
The classical acoustic modes in an isotropic elastic medium have been analyzed
previously, and many of the most useful known results summarized, by Auld (1973).
The lowest-order pure-compressional mode is referred to frequently as the breathing
mode. The displacement field associated with this lowest order compressional mode
of a sphere of radius a is given by

ωq r
u(q, r ) = r̂γ j1 e−iωq t , (7.259)
cl
where γ is the normalization constant, r̂ is the unit vector in the radial direction, j1
is the spherical Bessel function of order unity, j1 (x) = sin x/x 2 − cos x/x, ωq is the
√
mode frequency, and the longitudinal sound speed cl is equal to (λ + 2µ)/ρ. The
frequency for a free-standing sphere is determined by the condition that the normal
component of the traction force at the surface of the sphere vanishes; that is, Trr = 0
at r = a:

d2 ωq r 2λ d ωq r
(λ + 2µ) 2 j0 + j0 = 0, (7.260)
dr cl r dr cl r =a
where j0 (x) = sin x/x is the spherical Bessel function of order zero. This last result
implies that

ωq r 4µωq a/c
tan = . (7.261)
cl 4µ − (λ + 2µ)(ωq a/c)2
The normalization condition for this lowest-order breathing mode is
Figure 7.24. Schematic

illustrations of the
displacement fields of (a) the
axisymmetric torsional
mode, (b) the axisymmetric
radial–longitudinal mode,
and (c) the flexural mode for
a solid elastic cylinder. From
Sirenko et al. (1996b),
American Physical Society,
with permission.
ω̃l
3γ 2 h̄
dr r 2 j12 (r ) = , (7.262)
ω̃l3 0 2Mωq
with ω̃l representing the quantity ωq a/cl . This integral may be performed analyti-
cally and it follows that

1 h̄
γ =# , (7.263)
j 2 (ω̃ ) − j (ω̃ ) j (ω̃ ) 3Mωq
1 l 0 l 2 l
where the second-order spherical Bessel function, j2 (x), is defined by j2 (x) =

[(3/x 3 ) − (1/x)] sin x − (3/x 2 ) cos x. The lowest-order breathing and torsional
modes for a spherical quantum dot are shown in Figure 7.25.
The lowest-order torsional mode – a pure shear mode – of an free-standing
isotropic spherical quantum dot may be determined from the known elastic-
continuum solution for the lowest-order classical pure shear mode of an isotropic
elastic medium (Auld, 1973):
Figure 7.25. Lowest-order

(a) breathing mode and (b)
torsional mode in a spherical
quantum dot. From Stroscio
Institute of Physics, with
permission.

ωq r
ϕ τ cos θ j1
u(q, r, θ) = ϕ̂ e−iωq t , (7.264)
ct
ϕ is the unit vector in the ϕ -direction, τ is the normalization constant to
where ϕ̂
be determined from the phonon normalization condition, and the transverse sound
√
speed of the shear wave is given by ct = µ/ρ. This mode is depicted in Figure
7.25(b). The normalization condition for this mode is

τ 2 ω̃t h̄
dr r 2 j12 (r ) = , (7.265)
ω̃t 0
3 2Mω q
with ω̃t = ωq a/ct . Thus, the normalization constant τ may be evaluated in terms of
the same integral used to calculate γ ; indeed,

1 h̄
γ =# . (7.266)
j 2 (ω̃ ) − j (ω̃ ) j (ω̃ ) Mω q
1 t 0 t 2 t
Following the same procedure as for the breathing mode, it follows that the
dispersion relation for the lowest-order torsional mode is

ωq a 3ωq a/ct
tan = . (7.267)
ct 3 − (ωq a/ct )2
Krauss and Wise (1997) have recently observed the coherent acoustic phonons in
spherical quantum dots. The damping of the lowest-order acoustic phonon modes
observed by Krauss and Wise has been described in terms of the elastic continuum
model by Stroscio and Dutta (1999).
McSkimin (1944) gave approximate classical flexural thickness modes for a
structure with rectangular faces. The structure considered in this section has faces
joining each other at right angles, and the faces in the x y-, yz-, and x z-planes are
rectangles such that the width of the structure – in the x-direction – is a, the height
of the structure – in the y-direction – is b, and the length of the structure – in the
z-direction – is c. The approximate flexural thickness modes given by McSkimin are
u(x, y, z) = A sin mx (sin l1 y + α sin l2 y + β sin l3 y) cos nz,

−l1 αl2 β(m 2 + n 2 )
v(x, y, z) = A cos mx cos l1 y − cos l2 y + cos l3 y cos nz,
m m l3 m
/ 0
n α(m 2 + l22 ) βn
w(x, y, z) = A cos mx sin l1 y − sin l2 y + sin l3 y sin nz,
m nm m
(7.268)
where α and β are determined by applying desired boundary conditions on two

sets of rectangular faces. The discrete mode indices for the x- and z-dimensions
are labeled by m and n respectively. For the y-dimension, the mode index for the
irrotational, or longitudinal, part of the mode is l1 , and l2 and l3 correspond to the

indices for the rotational parts of the mode. As will become clear, the deformation
potential depends only on l1 , as expected. Indeed, it follows that
∂u ∂v ∂w
(x, y, z) = + +
∂x ∂y ∂z

l12 n2
= A cos mx m sin l1 y + sin l1 y + sin l1 y cos nz. (7.269)
m m
The flexural acoustic modes of McSkimin are used in this section to illustrate the
quantization of modes for such a structure. Obviously, these modes do not form a
complete set. Indeed, in addition to compressional modes there are flexural modes
corresponding to the width and length modes. The normalization condition for
McSkimin’s flexural thickness modes,

1 a/2 b/2 c/2 h̄
dx dy dz uu ∗ + vv ∗ + ww ∗ = , (7.270)
abc −a/2 −b/2 −c/2 2Mω
reduces to

1
f 2 (m, a/2) f 1 (n, c/2)[ f 2 (l1 , b/2) + 2αg1 (l1 , l2 , b/2)
abc
+ 2βg1 (l1 , l3 , b/2) + α 2 f 2 (l2 , b/2) + 2αβg1 (l2 , l3 , b/2) + β 2 f 2 (l3 , b/2)]
/
l2 2αl1l2
+ f 1 (m, a/2) f 1 (n, c/2) 12 f 1 (l1 , b/2) + g2 (l1 , l2 , b/2)
m m2
2βl1 (m 2 + n 2 ) α 2l22 2αβl2 (m 2 + n 2 )
− g2 (l 1 , l 3 , b/2) + f 1 (l 2 , b/2) −
l3 m 2 m2 l3 m 2
0
2β (m + n )
2 2 2
× g2 (l2 , l3 , b/2) + f 1 (l3 , b/2)
l32 m 2
/
n2 2αn(m 2 + l22 )
+ f 1 (m, a/2) f 2 (n, c/2) f 2 (l 1 , b/2) − g1 (l1 , l2 , b/2)
m2 nm 2
2βn 2 α 2 (m 2 + l22 )
+ g1 (l 1 , l 3 , b/2) + f 2 (l2 , b/2)
m2 n2m 2 0
2αβn(m 2 + l22 ) β 2n2
− g1 (l2 , l3 , b/2) + f 2 (l3 , b/2)
nm 2 m2
h̄
= , (7.271)
2Mω
where
+a/2
a sin ha
f 1 (h, a/2) = dy cos hy =
2
1+ ,
−a/2 2 ha
+a/2
f 2 (h, a/2) = dy sin2 hy = a − f 1 (h, a/2),
−a/2
+b/2
g1 (li , l j , b/2) = dy sin li y sin l j y
−b/2
(7.272)
sin(li − l j )b/2 sin(li + l j )b/2
= − ,
li − l j li + l j
+b/2
g2 (li , l j , b/2) = dy cos li y cos l j y
−b/2
sin(li − l j )b/2 sin(li + l j )b/2

= + .
li − l j li + l j
The deformation potential for these flexural thickness modes has an especially
simple form at a non-degenerate -point, namely,
Hdef = E 1 (x, y, z)

E1 A 2
=√ (m + l1 + n ) (cos mx sin l1 y cos nz) aq + c.c. .
2 2
N q m
(7.273)
In fact, the spatially dependent terms in Hdef do not depend on α and β.

Chapter 8
Carrier–LO-phonon scattering
Beauty is Nature’s brag, and must be shown in courts, at feasts,

and high solemnities where most may wonder at the
workmanship.
John Milton, 1637
8.1 Fröhlich potential for LO phonons in bulk

zincblende and würtzite structures
In this section, the Fermi golden rule and the dielectric continuum model of optical
phonons are applied to determine the electron–LO-phonon scattering rates in bulk
zincblende and bulk würtzite structures.
8.1.1 Scattering rates in bulk zincblende semiconductors

A transition rate for electron–LO-phonon scattering may be estimated by calculating
the transition rate predicted by the Fermi golden rule for the perturbation Hamilto-
nian HFr of Section 5.2. According to the Fermi golden rule, the transition rate for
emission or absorption,
S {e,a}
4 state |k and an initial
4 (k, k ), from an initial electron

phonon state Nq + 2 ± 2 – denoted by k, Nq + 12 ± 12 – to a final electron state
1 1
5 6 6

k and a final phonon state Nq + 12 ± 12 – denoted by k , Nq + 12 ± 12 – per unit
time per unit volume is given by
2π {e,a} 2
S {e,a} (k, k ) = M (q) δ(E(k ) − E(k) ± h̄ω), (8.1)
h̄ q
where
6 4

M {e,a} (q) = k , Nq + 1
2 ± 12 HFr k, Nq + 1
2 ± 1
2 . (8.2)
In these expressions, the positive and negative signs are chosen to select the de-
sired initial and final phonon states. In the case of phonon emission, the final phonon
131
132 8 Carrier–LO-phonon scattering
state contains one phonon more than the initial phonon

6 state; accordingly,
the initial
7 1

and final states are selected to be k, Nq and k , Nq + 2 ± 2 respectively. For
1
phonon absorption processes the initial and final states are chosen in a similar
manner. In the energy-conserving delta function in S {e,a} (k, k ), the upper sign
corresponds to phonon emission and the lower sign to phonon absorption. In this
expression,
7 the electron and phonon states represent different spaces so the ‘product’
7
|k Nq ≡ k, Nq ; that is, calculating the matrix elements of electron states involves
only integration over coordinate space and calculating matrix elements of phonon
states involves only taking matrix elements of the phonon creation and annihilation
operators as in Section 5.1. Clearly these two types of matrix element may 7 be
|k Nq ≡
calculated
7
independently of each other. For this reason, the ‘product’
k, Nq should not cause any confusion. The phonon states are as defined in Section
5.1 and the Fröhlich Hamiltonian is that of Section 5.2,

2π e2 h̄ωLO 1 1 1 iq·r
HFr = −i − aq e − aq† e−iq·r
V (∞) (0) q q
C
= aq eiq·r − aq† e−iq·r , (8.3)
q q
where

2πe2 h̄ωLO 1 1
C = −i − . (8.4)
V (∞) (0)
The electron states are plane-wave states normalized in the volume V ,

√ 5 √
|k = eik·r / V and k = e−ik ·r / V . (8.5)
The matrix element, M {e,a} (q), is then given by

e−ik ·r+ik·r C 6

7
M {e,a} (q) = d 3r Nq + 1
2 + 12 (aq eiq·r − aq† e−iq·r ) Nq
V q

e−ik ·r+ik·r∓iq·r C
1 1/2
= d 3r (∓) n q + 1
2 ± 2 , (8.6)
V q
where the phonon matrix elements have been evaluated using the expressions of
Section 5.1; in particular,
5 7 √ 5 7
Nq − 1 aq Nq = n q and Nq + 1 aq† Nq = n q + 1. (8.7)
In (8.6) the upper signs correspond to phonon emission and the lower signs to

phonon absorption. The integral over the factor e−ik ·r+ik·r∓iq·r /V is, of course,
8.1 Fröhlich potential for LO phonons in bulk zincblende and würtzite structures 133
dimensionless and contributes a delta function that depends on the quantity

k − k ± q. Hence, it follows that the final state electron wavevector, k , is related to
the initial state electron wavevector k and the phonon wavevector q by k = k ∓ q.
The case where k = k + q corresponds to an electron of wavevector k in the
initial state and an electron and phonon in the final state with wavevectors k and q
respectively. That is, this case represents phonon emission or phonon creation.
For the remaining case, k + q = k and it is clear that the final state represents
an electron of wavevector k and that the initial state is composed of an electron and
phonon with wavevectors k and q respectively. Clearly, this case represents phonon
absorption or phonon annihilation. The two conditions k = k + q and k + q = k
are equivalent to momentum conservation, since the product of each wavevector
times h̄ is the momentum of the particle in question. The integral over the factor
exp(−ik · r + ik · r ∓ iq · r)/V may be written as a Kronecker delta function
requiring k − k ± q = 0 or it may be expressed as V −1 δ(k − k ± q). Then it
follows that
2 d 3 q |C|2

M {e,a} (q) = n q + 12 ± 12 δ(k − k ± q)
q (2π ) q
3 2
|C|2 1
= nq + 1
± 1
, (8.8)
4π h̄ 2 |k − k |2 2 2
where

2π 2 1 1
|C| =
2
e h̄ω − . (8.9)
V (∞) (0)
In evaluating the integral over q it has been assumed that the frequency of the
phonon mode, ωq , is independent of q and is equal to the zone-center phonon
frequency, ωq . This is a good approximation for many materials. For example,
in GaAs the LO phonon frequency varies by about 10% over the entire Brillouin
zone. Moreover, since the largest contributions to the integral occur near the zone
center – as a result of the q −1 dependence in the integrand – the approximation that
ωq = ωLO causes little error in the value of the integral in question.
Accordingly, the transition rate S {e,a} (k, k ) is given by
|C|2 1
S {e,a} (k, k ) = nq + 1
± 1
δ(E(k ) − E(k) ± h̄ω)
4π h̄ 2 |k − k |2 2 2
|C|2 1
= nq + 1
2 ± 1
2 δ(E(k ) − E(k) ± h̄ω), (8.10)
4π h̄ 2 q 2
where q 2 = |q|2 and the delta function expresses the conservation of energy, E(k )−
E(k) = ∓h̄ω, the upper sign representing phonon emission and the lower sign
representing phonon absorption. The |q|−2 dependence in S {e,a} (k, k ) is the same
as that appearing in the Coulomb interaction. Accordingly, it is anticipated that finite
scattering rates will be obtained only if screening is treated properly. Therefore,

S {e,a} (k, k ) is rewritten as
|C|2 q2
S {e,a} (k, k ) = nq + 1
2 ± 1
2 δ(E(k ) − E(k) ± h̄ω),
4π h̄ (q + q0 )
2 2 2 2
(8.11)
where q0 is a cut-off parameter introduced to represent the effect of screening of

the Coulomb-like interaction. The final scattering rate will be calculated in the limit
q0 → 0.
{e,a}
The total, three-dimensional, scattering rate, 1/τ3D (k), associated with
{e,a} {e,a}
S3D (k, k ) results from integrating S3D (k, k ) over q and multiplying by the
volume of the heterostructure, V :
1 |C|2 V
{e,a}
= 2
n q + 12 ± 12
τ3D (k) 4π h̄
2π π
q2
× dϕ dθ sin θ dq q 2 δ(E(k ) − E(k) ± h̄ω).
0 0 (q 2 + q02 )2
(8.12)
The integral over ϕ contributes a factor of 2π since there is no ϕ-dependence in the

problem at hand. In performing the integration over θ it is necessary to consider the
fact that the argument of the delta function depends on θ. By assuming parabolic
carrier bands and considering the case where the initial carrier wavevector is k and
the carrier mass is m, the argument of the delta function may be rewritten as follows:
E(k ) − E(k) ± h̄ω = E(k ∓ q) − E(k) ± h̄ω

h̄ 2 (k ∓ q)2 h̄ 2 (k)2
= − ± h̄ω
2m 2m
h̄ 2 2
= (k ) − (k)2 ± h̄ω
2m
h̄ 2 2
= (q ∓ 2qk cos θ ) ± h̄ω. (8.13)
2m
The integral over θ is now performed by use of the relation

g(θ )
g(θ)δ[ f (θ) − a]dy = , (8.14)
d f /dθ θ =θ0
where θ0 is determined by f (θ0 ) = a. Then, taking
h̄ 2
f (θ) = ∓ (2qk cos θ), (8.15)
2m
it follows by letting u = cos θ that
2π h̄ 2 2
dθ sin θ δ (q ∓ 2qk cos θ ) ± h̄ω
0 2m
1
h̄ 2 h̄ 2 2
= du δ (∓2qku) + q ± h̄ω
−1 2m 2m
m
=∓ 2 , (8.16)
h̄ qk
so that
{e,a}
|C|2 V
1 2π m qmax q3
{e,a}
= nq + 1
2 ± 1
2 dq ,
τ3D (k) 4π 2 h̄ h̄ 2 k qmin
{e,a} (q 2 + q02 )2
{e,a}
|C|2 V m qmax q3
= nq + 1
2 ± 1
2 dq . (8.17)
2π h̄ 3 k qmin
{e,a} (q 2 + q02 )2
{e} {e}
We consider the ranges of q associated with emission and absorption, {qmin , qmax }
{a} {a}
and {qmin , qmax } respectively. The range of q is determined by
2mω
q 2 − 2qk cos θ ± = 0, (8.18)
h̄
where the upper sign corresponds to emission and the lower sign to absorption. The
roots of this equation are

{e} 2mω
q = k cos θ ± k cos θ −
2 2 , (8.19)
h̄
and

{a} 2mω
q = −k cos θ ± k 2 cos2 θ + . (8.20)
h̄
Hence, by taking cos θ = ±1, the minimum and maximum values are found:

{e} {e} 2mω 2mω
{qmin , qmax } = k − k 2 − , k + k2 − , (8.21)
h̄ h̄

{a} {a} 2mω 2mω
{qmin , qmax } = k2 + − k, k2 + +k . (8.22)
h̄ h̄
Then the total scattering rate may be written as

qmax{e}
1 1 2mω q3
{e,a}
= αω (n q + 1) dq
τ3D (k) k h̄ {e}
qmin (q 2 + q02 )2
qmax
{a}
1 2mω q3
+ αω nq dq , (8.23)
k h̄ {a}
qmin (q 2 + q02 )2
where the relation

|C|2 V m 1 2mω
= αω (8.24)
2π h̄ 3 k k h̄
with

1 e2 1 1
α= √ − (8.25)
2h̄ω h̄/2mω (∞) (0)
{e,a}
has been used to replace |C|2 by α. The scattering rate, 1/τ3D (k), may be evaluated
analytically through use of the identity
 
{e,a} {e,a}2 {e,a}2 {e,a}2
qmax q3 1  qmax + q02 q02 qmin − qmax
dq = ln {e,a}2 + $ %$ % .
{e,a}
qmin (q 2 + q02 )2 2 qmin + q02 2 q {e,a}2 + q 2 q {e,a}2 + q 2
max 0 min 0
(8.26)
Ignoring the terms of O(q02 ) it follows that
$ %2
1 αω 1 2mω k + k 2 − 2mω/h̄ + q02
{e,a}
= (n q + 1) ln $ %2
τ3D (k) 2 k h̄
k − k 2 − 2mω/h̄ + q02
$ %2
k 2 + 2mω/h̄ + k + q02
αω 1 2mω
+ n q ln $ %2 . (8.27)
2 k h̄
k 2 − 2mω/h̄ − k + q02
Finally, in the limit as q0 → 0

1 αω 1 2mω k + k 2 − 2mω/h̄
{e,a}
= (n q + 1) ln
τ3D (k) 2 k h̄ k − k 2 − 2mω/h̄

αω 1 2mω k 2 + 2mω/h̄ + k
+ n q ln . (8.28)
2 k h̄ k 2 − 2mω/h̄ − k
8.1.2 Scattering rates in bulk würtzite semiconductors

The Fröhlich interaction Hamiltonian given in Section 7.1 for a polar uniaxial (UA)
crystal may be written as
†
UA
HFr = (−e)φ(q) eiq·r (aq + a−q ) (8.29)
q

2π e2 h̄ 1 iq·r †
=i e (aq + a−q )(ω⊥,TO
2
− ωq2 )(ωTO,
2
− ωq2 )
q V ωq q

× [(0)⊥ − (∞)⊥ ]ω⊥,TO
2
(ω,TO
2
− ωq2 )2 sin2 θ
−1/2
+ [(0) − (∞) ]ω,TO
2
(ω⊥,TO
2
− ωq2 )2 cos2 θ
1/2
4π e2 h̄V −1 1 iq·r †
=i e (aq + a−q ),
q (∂/∂ω)[(ω)⊥ sin θ + (ω) cos θ]
2 2 q
(8.30)
where
ω2 − ω⊥,LO
2 ω2 − ω,LO
2
⊥ (ω) = ⊥ (∞) , (ω) = (∞) , (8.31)
ω2 − ω⊥,TO
2 ω2 − ω,TO
2
and θ is the angle between the phonon wavevector q and the c-axis. Moreover, from
Section 3.2, equation (3.21),
⊥ (ω) sin2 θ + (ω) cos2 θ = 0. (8.32)
As was discussed in Section 3.2, the high-frequency electronic response of a

medium should not depend strongly on the crystalline structures; it is usually
assumed (Loudon, 1964) that ⊥ (∞) ≈ (∞). Thus
ωLO,⊥
2 − ω2 ωLO,
2 − ω2
sin θ +
2
cos2 θ = 0 (8.33)
ωTO,⊥
2 − ω2 ωTO,
2 − ω2
or, equivalently,
ω4 − (ω12 + ω22 )ω2 + ωTO,⊥

2
ωLO,
2
cos2 θ + ωLO,⊥
2
ωTO,
2
sin2 θ = 0
(8.34)
where
ω12 = ωTO,
2
sin2 θ + ωTO,⊥
2
cos2 θ, ω22 = ωLO,
2
cos2 θ + ωLO,⊥
2
sin2 θ.
(8.35)

In Section 3.2 it was shown that when ωTO, − ωTO,⊥ ωLO, − ωTO, and
ωLO,⊥ − ωTO,⊥ this equation has roots

ω2 = 1
2 (ω12 + ω22 ) ± [(ω12 − ω22 ) + 2ω2 (θ)] , (8.36)
where
(ωLO,
2 − ωLO,⊥
2 )(ωTO,
2 − ωTO,⊥
2 )
ω2 (θ) = 2 sin2 θ cos2 θ; (8.37)
ω22 − ω12
thus, one root is
2TO = ω2 = ωTO,
2
sin2 θ + ωTO,⊥
2
cos2 θ
(ωLO,
2 − ωLO,⊥
2 )(ωTO,
2 − ωTO,⊥
2 )
− sin2 θ cos2 θ
ω22 − ω12
≈ ωTO,
2
sin2 θ + ωTO,⊥
2
cos2 θ (8.38)
and the other root is
2LO = ω2 = ωLO,
2
cos2 θ + ωLO,⊥
2
sin2 θ
(ωLO,
2 − ωLO,⊥
2 )(ωTO,
2 − ωTO,⊥
2 )
+ sin2 θ cos2 θ
ω22 − ω12
≈ ωLO,
2
cos2 θ + ωLO,⊥
2
sin2 θ. (8.39)
The inequalities assumed for the derivation of 2TO and 2LO are satisfied for the
parameters of both GaN and AlN as may be verified from the numerical values of
the parameters given in Section 3.3. For the case of GaN, the dependence for the
phonon frequencies on θ is shown in Figure 8.1 for these infrared-active phonons. θ
is the angle between the phonon wavevector q and the c-axis.
As in subsection 8.1.1, the transition rate for electron–optical-phonon scattering
may be estimated by calculating the transition rate predicted by the Fermi golden
rule for the perturbation Hamiltonian, HFr UA . According to the Fermi golden rule, the
{e,a}
transition rate, SUA (k, k ), from an initial electron state |k and an initial phonon
Figure 8.1. Angular

dependence of the phonon
frequencies ω for LO-like
and TO-like infrared-active
phonons in GaN. θ is the
angle between the phonon
wavevector q and the c-axis.
From Lee et al. (1997),
with permission.
4 4 5

state Nq + 1
±– denoted by k, Nq + 12 ± 12 – to a final electron state k and
1
2
6 2
6

a final phonon state Nq + 12 ± 12 – denoted by k , Nq + 12 ± 12 – per unit time
per unit volume is given by
{e,a} 2π {e,a} 2
SUA (k, k ) = MUA (q) δ(E(k ) − E(k) ± h̄ω), (8.40)
h̄ q
where
6 4
{e,a} UA
MUA (q) = k , Nq + 1
2 ± 12 HFr k, Nq + 1
2 ± 1
2 . (8.41)
As in subsection 8.1.1, the electron states are plane-wave states normalized in

volume V ,
√ 5 √
|k = eik·r / V and k = e−ik ·r / V . (8.42)
2
{e,a},TO
For ω2 = 2TO , the matrix element, MUA (q) , is then given by (Lee et al.,
1997)

{e,a},TO 2
MUA (q)
2π e2 h̄ (ω⊥,TO
2 − ω,TO
2 )2 sin2 θ cos2 θ
=
V q 2 TO [⊥ (0) − ⊥ (∞)] ω⊥,TO
2 cos2 θ + (0) − (∞) ω,TO
2 sin2 θ

× n q + 12 ± 12 . (8.43)
This matrix element does not vanish, in general, since the TO-like mode is in reality
not a pure TO mode in a uniaxial material. In the limit of an isotropic material,
ω⊥,TO = ω,TO and the transverse matrix element vanishes. Likewise, for ω2 =

{e,a},LO 2
2LO , the matrix element MUA (q) is given by
/
2π e2 h̄ sin2 θ
{e,a},LO 2
MUA (q) =
V q 2 LO [1/⊥ (∞) − 1/⊥ (0)] ω⊥,LO
2
0−1
cos2 θ
+
1/ (∞) − 1/ (0) ω,LO
2

× n q + 12 ± 12 . (8.44)
In the isotropic limit, ω⊥,LO = ω,LO and the longitudinal matrix element reduces
to
2π e2 h̄ωLO
{e,a},LO 2
MUA (q) = [1/⊥ (∞) − 1/⊥ (0)] n q + 12 ± 12 . (8.45)
V q2

{a} 2

Numerical evaluation of Mq = MUA (q) indicates that the LO-like and TO-
like contributions to the absorption matrix element for bulk GaN are as shown in
Figure 8.2.
The scattering rates for LO-like and TO-like phonons have been evaluated
numerically by Lee et al. (1997) for two cases: as a function of the incident angle of
the electron with respect to the c-axis, θk , for an initial electron energy of 0.3 eV; and
for θk = 0 and θk = π/2, for a range of electron energies. These results are depicted
in Figures 8.3 and 8.4. In Figure 8.3 the scattering rates for emission and absorption
of LO-like and TO-like phonons are presented as a function of the incident angle of
the electron with respect to the c-axis, θk , for an initial electron energy of 0.3 eV. In
Figure 8.4 the scattering rates for LO-like and TO-like absorption are presented for
a range of incident electron energies for θk = 0 and θk = π/2.
8.2 Fröhlich potential in quantum wells
The Fröhlich interaction Hamiltonian of equation (5.34) in Section 5.2 describes

the carrier–polar-optical-phonon interaction in the bulk. As discussed in subsec-
tion 7.3.2, the Fröhlich Hamiltonian for the two-dimensional slab takes the form
(7.71). Moreover, the Fröhlich interaction Hamiltonian for the IF optical phonon
modes in the dielectric slab is given by equation (7.72). The interaction Hamiltoni-
ans of subsection 7.3.1 describe the Fröhlich potential for a two-dimensional slab
bounded by regions with = 1, as discussed by Licari and Evrard (1977).
In this section, the Hamiltonians describing carrier–polar-optical-phonon in-
teractions will be used to calculate carrier–LO-phonon scattering rates in the
approximation of the Fermi golden rule.
Figure 8.2. Matrix element

|Mq | for optical phonon
absorption as a function of θ
for GaN. θ is the angle
between the phonon
wavevector q and the c-axis.
From Lee et al. (1997),
with permission.
8.2.1 Scattering rates in zincblende quantum-well

structures
A particularly simple treatment of carrier–LO-phonon interactions in quantum-well
structures is possible when the carrier confinement is considered in the limit of an
infinitely deep confining potential for electrons – known as the so-called extreme
quantum limit – and the phonons are taken to be bulk phonons. In this case the
calculation of subsection 8.1.1 must be modified to account for the confinement
of the charge carriers. As in subsection 8.1.1, the transition rate for electron–LO-
phonon scattering may be estimated by calculating the transition rate predicted by
the Fermi golden rule for the perturbation Hamiltonian HFr of Section 5.2. However,
the carrier states are now those of an infinitely deep quantum well. According to the
{e,a}
Fermi golden rule, the transition rate, S2D (k, 4 k ), from an initial electron state
4

|k2D and an initial phonon state Nq + 12 ± 12 – denoted by k2D , Nq + 12 + 12 –
5 6

to a final electron state k2D and a final phonon state Nq + 12 ± 12 – denoted by
6

k2D , Nq + 12 ± 12 – per unit time per unit heterostructure area A, is given by
Figure 8.3. Scattering rates for emission and absorption of LO-like and TO-like
phonons as a function of the incident angle of the electron with respect to the c-axis,
θk , for an initial electron energy of 0.3 eV. From Lee et al. (1997), American
Physical Society, with permission.
{e,a} 2π {e,a} 2

S2D (k, k ) = M2D (q) δ(E(k ) − E(k) ± h̄ω), (8.46)
h̄ q
where
6 4
{e,a}
M2D (q) = k2D , Nq + 1
2 ± 12 HFr k2D , Nq + 1
2 ± 1
2 . (8.47)
In these expressions, the positive and negative signs are chosen to select the
desired initial and final phonon states. In the case of phonon emission, the final
phonon state contains one phonon more than the initial phonon6state; accordingly,
7
the initial and final states are selected to be k2D , Nq and k2D , Nq + 12 ± 12
respectively. For phonon absorption processes, the initial and final states are chosen
in a comparable manner. In the energy-conserving delta function in S {e,a} (k, k ), the
upper sign corresponds to phonon emission and the lower sign to phonon absorption.
As in subsection 8.1.1, the
electron and phonon states represent different spaces, so
7 7

we have |k2D Nq ≡ k2D , Nq . The phonon states are as defined in Section 5.1
and the Fröhlich Hamiltonian is given by equation (5.34),
C
HFr = aq eiq·r − aq† e−iq·r , (8.48)
q q
Figure 8.4. Scattering rates for LO-like and TO-like absorption are presented for a
range of incident electron energies for θk = 0 (solid line) and θk = π/2 (broken
line). From Lee et al. (1997), American Physical Society, with permission.
where

2π e2 h̄ωLO 1 1
C = −i − . (8.49)
V (∞) (0)
The electron states are taken as plane waves in the directions parallel to the
heterointerfaces and as the ground state of an electron in an infinitely deep quantum
well in the z-direction (Leburton, 1984),

eik ·r 2 πz 5 e−ik ·r 2 πz
|k2D = √ sin and k2D = √ sin ,
A Lz Lz A Lz Lz
(8.50)
where A is the area of the heterointerface over which the electron wavefunction is
normalized.
{e,a}
Following the procedures of subsection 8.1.1, the matrix element M2D (q) is
then given by

{e,a} e−ik ·r+ik ·r 2 2 πz C
M2D (q) = d r $√ %2 3
sin
Lz Lz q
A
6 4

× Nq + 12 ± 12 (aq eiq·r − aq† e−iq·r ) Nq + 1
2 ± 1
2
C
= (∓)(n q + 1
2 ± 12 )1/2 δk −k ±q F(qz L z ), (8.51)
q
where F(qz L z ) is defined as

2 Lz πz
F(qz L z ) = dz e−iqz z sin2
0 Lz Lz
π 2 sin qz L z /2 −iqz L z /2
= 2 e . (8.52)
π − (qz L z /2)2 qz L z /2
Then, following the procedure of subsection 8.1.1, it follows that
{e,a} L z |C|2
S2D (k , k ) = (n q + 12 ± 12 )I2D (q , L z )
4π h̄ 2
× δk −k ±q δ(E(k ∓ q ) − E(k ) ± h̄ω), (8.53)
where

2π 2 1 1
|C| =
2
e h̄ω − , (8.54)
V (∞) (0)
and
+∞ |F(qz L z )|2
I2D (q , L z ) = dqz
−∞ q2 + qz2
/
0
π 1 1 S2 1
= Lz [1 − G(S)] + + 2− 2 G(S)
2 S2 2 S + π2 S2 + π 2
(8.55)
with
G(S) = [1 − e−2S ]/2S and S = q L z /2. (8.56)
As in subsection 8.1.1, it has been assumed that the frequency of the phonon mode,
ωq , is independent of q and is equal to the zone-center phonon frequency, ωq .
{e,a} {e,a}
The total scattering rate, 1/τ2D (k ), associated with S2D (k , k ) results from
{e,a}
integrating S2D (k , k ) over q and multiplying by the area of the heterostructure,
V2D :
1 |C|2 L z V2D
{e,a}
= n q + 12 ± 12
τ2D (k ) 4π 2 h̄

× dφ dq I2D (q , L z )δ(E(k ∓ q ) − E(k ) ± h̄ω).
(8.57)
To evaluate this integral, it is necessary to consider the argument of the delta

function. By assuming parabolic carrier bands and considering the case where a
carrier of wavevector k and mass m emits a phonon of wavevector q , so that the
final carrier wavevector is k = k − q , it follows that
E(k ∓ q ) − E(k ) ± h̄ω

h̄ 2 (k − q )2 h̄ 2 π 2 h̄ 2 (k )2 h̄ 2 π 2
= + − − ± h̄ω
2m 2m L 2z 2m 2m L 2z
h̄ 2
= [(k )2 − (k )2 ] ± h̄ω
2m
h̄ 2
= [(q )2 − 2q k cos φ] ± h̄ω, (8.58)
2m
where φ is the angle between k and q , the ground state energy for the carrier
is h̄ 2 π 2 /(2m L 2z ), and q and k represent the absolute magnitudes of q and k
respectively. The argument of the delta function vanishes when
2mω
q 2 − 2q k cos φ ± = 0, (8.59)
h̄
where the upper sign corresponds to emission and the lower sign to absorption. The
roots of this equation are

{e} 2mω
q± = k cos φ ± k2 cos2 φ − , (8.60)
h̄
and

{a} 2mω
q± = −k cos φ ± k2 cos2 φ + . (8.61)
h̄
{e}
Both roots of q± are allowed for φ when it is in the range 0 < φ < φmax , where
√ {a}
φmax = arccos h̄ω/E with E = h̄ 2 k2 /2m but with q− forbidden. For the case of
emission,
h̄ 2 {e,} {e,}
E(k ∓ q ) − E(k ) ± h̄ω = q − q+ q − q−
2m

h̄ 2 2mω
= q − 2q k cos φ +
2
. (8.62)
2m h̄
The integral over q is now performed by use of the relation

g(y)
dy g(y)δ( f (y) − a) = , (8.63)
d f /dy y=y0
where y0 is determined by f (y0 ) = a. Then, taking
h̄ 2 $ 2 %
f (q ) = q − 2q k cos φ , (8.64)
2m
it follows that
d f (q ) h̄ 2
= 2q − 2k cos φ
dq 2m

h̄ 2 2mω
=± k cos φ −
2 2 . (8.65)
m h̄
With these results, it follows that
1
{e}
τ2D (E)
φmax {e} {e} {e} {e}
I2D (q+ (E, φ))q+ (E, φ) + I2D (q− (E, φ))q− (E, φ)
αω
= (n q + 1) dφ ,
π 0 (E/h̄ω) cos2 φ − 1
(8.66)
and by an analogous derivation (Leburton, 1984),
π {a} {a}
I2D (q+ (E, φ))q+ (E, φ)
1 αω
{a}
= nq dφ , (8.67)
τ2D (E) π 0 (E/h̄ω) cos2 φ + 1
where

1 e2 1 1
α= √ − . (8.68)
2h̄ω h̄/2mω (∞) (0)
8.2.2 Scattering rates in würtzite quantum wells

Komirenko et al. (2000a) have calculated energy-dependent electron scattering
rates for electron–LO-phonon scattering in würtzite quantum wells. The Fröhlich
potential used to calculate these scattering rates (Komirenko et al., 1999; Lee et al.,
1998) is based on the Hamiltonians given in Appendix D and on the Loudon model
of uniaxial crystals. In these calculations the Fermi golden rule is used to calculate
the scattering rate for GaN free-standing quantum wells and for GaN quantum wells
embedded in AlN. In all cases the c-axis is normal to the heterointerfaces and the
GaN quantum wells are 50 ångstroms wide. The interacting electrons are restricted
to energies close to the point and only the lowest subband is considered. Figures
8.5 and 8.6 present the scattering rates calculated for bulk GaN, for a free-standing
GaN quantum well, and for an AlN/GaN/AlN quantum well. Figure 8.5 depicts the
total electron–optical-phonon scattering rates in bulk GaN and a free-standing GaN
quantum well as functions of the electron energy, E k , over a range of 0 to about
500 meV; this figure also presents the angular dependence of the scattering rate
for bulk GaN crystals. Figure 8.6 depicts the total scattering rate as a function of
the electron energy E k over a range of 0 to 400 meV for a free-standing quantum
well and a GaN quantum well embedded in AlN; this figure also indicates the
contributions of the confined modes and the interface modes.
These results illustrate clearly the importance of including the effects of dimen-
sional confinement in calculating the electron–optical-phonon scattering rates in
dimensionally confined würtzite semiconductor structures.
8.3 Scattering of carriers by LO phonons in quantum

wires
In this section, scattering rates for the scattering of charge carriers by LO phonons
in quantum wires are calculated in the approximation of the Fermi golden rule.
8.3.1 Scattering rate for bulk LO phonon modes in

quantum wires
A particularly simple treatment of carrier–LO-phonon interactions in quantum-wire
structures is possible when the phonons are taken to be bulk phonons and the carrier
confinement is considered in the limit of an infinitely deep confining potential
for electrons, the so-called extreme quantum limit. In this case the calculation of
subsection 8.1.1 must be modified to account for the confinement of the charge
carriers. As in Sections 8.1.1 and 8.3.1, the transition rate for electron–LO-phonon
scattering may be estimated by calculating the transition rate predicted by the Fermi
golden rule for the perturbation Hamiltonian, HFr , of Section 5.2. However, the
carrier states are now those of an infinitely deep quantum well. According to the
{e,a}
Fermi golden rule, the transition rate, S1D (k, 4 k ), from an initial electron state
4

|k1D and an initial phonon state Nq + 12 ± 12 – denoted by k1D , Nq + 12 ± 12 –
5 6
to a final electron state k1D and a final phonon state Nq + 1 + 1 – denoted by
6 2 2
, N + 1 + 1 – per unit time per unit heterostructure area A is given by
k1D q 2 2
{e,a} 2π {e,a} 2

S1D (k, k ) = M1D (q) δ(E(k ) − E(k) ± h̄ω), (8.69)
h̄ q
where
6 4
{e,a}
M1D (q) = k1D , Nq + 1
2 + 12 HFr k1D , Nq + 1
2 ± 1
2 . (8.70)
Figure 8.5. Main figure: total electron–optical-phonon scattering rates in bulk GaN
(thin lines) and in a free-standing GaN quantum well (thick solid line) as functions
of the electron energy E k over a range of 0 to about 500 meV. Scattering rates at
three particular angles for bulk GaN crystals are shown as follows: thin solid lines,
θk = π/2; thin dotted lines, θk = π/4; thin broken lines, θk = 0. The broken and
dotted line indicates the scattering rate calculated for confined electrons and bulk
phonons.
Inset: the LO-like and TO-like emission and absorption rates in bulk GaN for
three different angles: solid lines, θk = π/2; dotted lines, θk = π/4; broken lines,
θk = 0. From Komirenko et al. (2000a), American Physical Society, with
permission.
In these expressions, the positive and negative signs are chosen to select the desired
initial and final phonon states. In the case of phonon emission, the final phonon
state contains one phonon more than the initial phonon4 state;6 accordingly, the
, N + 1 ± 1
initial and final states are selected to be k1D , Nq + 12 ± 12 and k1D q 2 2
respectively. For phonon absorption processes the initial and final states are chosen
in similar manner. In the energy-conserving delta function in S {e,a} (k, k ), the upper
sign corresponds to phonon emission and the lower sign corresponds to phonon
absorption. As in subsection 8.1.1, the electron
and phonon states represent different
7 7

spaces, so that we have |k1D Nq ≡ k1D , Nq . Again, the phonon states are as
defined in Section 5.1 and the Fröhlich Hamiltonian is given by equation (5.34) of
Section 5.2,
C
HFr = aq eiq·r − aq† e−iq·r , (8.71)
q q
where
Figure 8.6. Total electron–optical-phonon scattering rates calculated for a

free-standing GaN quantum well (broken line) and an AlN/GaN/AlN quantum well
(solid line) as functions of the electron energy E k over a range of 0 to about
400 meV.
Inset: the scattering rates for confined modes (line with circles) and for interface
modes for the AlN/GaN/AlN structure (solid line) as well as for the free-standing
GaN quantum well (broken line). From Komirenko et al. (2000a), American

2π e2 h̄ωLO 1 1
C = −i − . (8.72)
V (∞) (0)
The electron states are taken as plane-wave states along the quantum-wire axis,
taken here as the x-axis, and as the ground states of an electron in an infinitely deep
quantum wire in the y- and z-directions (Leburton, 1984),

eik x x 2 πy 2 πz
|k1D = √ sin sin , (8.73)
Lx Ly Lz Lz Lz
and
−i
A− (qz )± = √ [a± (qx , q y ) − a± (qx , −q y )], (8.74)
2
where L x is the length in the x-direction over which the free-electron wavefunction
is normalized and L y and L z are the height and the width of the quantum wire; it is
assumed that L x L y , L z .
{e,a}
Following the procedures of subsection 8.1.1, the matrix element, M1D (q), is
then given by

{e,a} e−ik x x+ik x x 2 πy 2 πz
M1D (q) = d 3 r √ 2 sin2 sin2
Lx Ly L y Lz Lz
6 7
C
× Nq + 12 + 12 (aq eiq·r − aq† e−iq·r ) Nq
q
C
= (∓)(n q + 12 ± 12 )1/2 δk x −k x ±qx F(q y L y )F(qz L z ), (8.75)
q
where F(q y L y ) and F(qz L z ) are defined by
Ly
2 πy
F(q y L y ) = dy e−iq y y sin2
Ly 0 Ly
π2 sin q y L y /2 −iq y L y /2
= e ,
π − (q y L y /2) (q y L y /2)
2 2
Lz (8.76)
2 πz
F(qz L z ) = dz e−iqz z sin2
Lz 0 Lz
π 2 sin qz L z /2 −iqz L z /2
= 2 e .
π − (qz L z /2)2 (qz L z /2)
Then, following the procedure of subsection 8.1.1, it follows that
{e,a} L z |C|2
S1D (k x , k x ) = nq + 1
2 ± 1
2 I1D (qx , L y , L z )
4π h̄ 2
× δk x −k x ±qx δ(E(k x ∓ qx ) − E(k x ) ± h̄ω), (8.77)
where

2π 2 1 1
|C|2 = e h̄ω − (8.78)
V (∞) (0)
and

+∞ +∞ F(q y L y )2 |F(qz L z )|2
I1D (qx , L y , L z ) = dq y dqz , (8.79)
−∞ −∞ qx2 + q y2 + qz2
which has been evaluated numerically (Leburton, 1984).

As in subsection 8.1.1, it has been assumed that the frequency of the phonon
mode is independent of q and equal to the zone-center phonon frequency, ωq .
8.3.2 Scattering rate for confined LO phonon modes in

quantum wires
The bulk Fröhlich interaction Hamiltonian, HFr 3D , of Section 5.2 may be used to
derive straightforwardly the Fröhlich interaction Hamiltonian for a one-dimensional

quantum wire, HFr 1D , by an analysis similar to that of subsection 7.3.2, where
the Fröhlich interaction Hamiltonian for a quantum well, HFr 2D , was obtained

from HFr by imposing the boundary condition that HFr (x, y, z)z=L /2 = 0. In
3D 3D
z
particular, the analysis of subsection 7.3.2 may be extended by applying not just
the boundary condition HFr 3D (x, y, z) = 0 but also the boundary condition
z=L z /2
H (x, y, z)
3D = 0. Starting with the canonical form of the bulk Fröhlich
Fr y=L y /2
interaction Hamiltonian,

2πe2 h̄ωLO 1 1 1
3D
HFr =− − †
(aq + a−q ) e−iq·r , (8.80)
V (∞) (0) q q
we write the sum over q as a sum over positive values of qz in equation (8.84):

HFr = Vq e−iq ·r e−iqz z aq ,qz + a−q
†
,−qz
q ,qz >0
†
+ eiqz z aq ,−qz + a−q ,qz
, (8.81)
where

2πe2 h̄ωLO 1 1
Vq = − − . (8.82)
V (∞) (0)
Then, defining
1 −i
a+ (q ) = √ (aq ,qz + aq ,−qz ) a− (q ) = √ (aq ,qz − a−q ,qz ),
2 2
(8.83)
with
1
A+ (qz )± = √ [a± (qx , q y ) + a± (qx , −q y )],
2
(8.84)
−i
A− (qz )± = √ [a± (qx , q y ) − a± (qx , −q y )],
2
expanding e±iq y y and e±iqz z , and taking q y = ±mπ/L y and qz = ±mπ/L z to
enforce the electromagnetic boundary conditions that the confined phonon modes
vanish at y = ±L y /2 and z = ±L z /2, leads to the result (Stroscio, 1989)
1D
HFr

mπ y nπ z
= 2α e−iqx x λmn cos cos [A+ (qx )+ + A†+ (−qx )+ ]
qx m=1,3,5,... n=1,3,5,...
Ly Lz
mπ y nπ z
+ λmn cos sin [A+ (qx )− + A†+ (−qx )− ]
m=1,3,5,... n=2,4,6,...
Ly Lz
mπ y nπ z
+ λmn sin cos [A− (qx )+ + A†− (−qx )+ ]
m=2,4,6,... n=1,3,5,...
Ly Lz

mπ y nπ z †
+ λmn sin sin [A− (qx )− + A− (−qx )− ] ,
m=2,4,6,... n=2,4,6,...
Ly Lz
(8.85)
where
/ 2 2 0−1/2
mπ nπ
λmn = qx2 + + (8.86)
Ly Lz
and
! "1/2
2π e2 1 1
α = h̄ω − . (8.87)
V (∞) (0)
Equation (8.85) may be written more compactly as (Campos et al., 1992),

∞
∞
1D
HFr = 2α e−iqx x ξm (y)ξn (y)λmn βmn , (8.88)
qx m=1 n=1
where
βmn = Amn (qx ) + A†mn (−qx ), (8.89)
Amn (qx ) = 12 [(−1)m+1 a(qx , q y , qz ) + a(qx , q y , −qz )

+ (−1)n+m a(qx , −q y , qz ) + (−1)n+1 a(qx , −q y , −qz )], (8.90)

nπ t π
ξn (t) = sin + δn (8.91)
Lt 2
with δn = 1 for odd n and δn = 0 for even n.
The Hamiltonian (8.88) includes contributions only from the confined modes of
the dielectric continuum model with electrostatic boundary conditions – also known
as the slab modes. The interface and half-space modes are not included and must be
considered separately. For calculations based on the guided modes, the contributions
to the Hamiltonian in the interior of the quantum wire may be written in terms of
the Hamiltonian for the slab modes by taking

nπ t π
ξn (t) = cos + δn . (8.92)
Lt 2
Of course, it is necessary to augment this expression with terms containing
the evanescent contributions of the guided modes in the regions surrounding the
quantum wire. In the present account, scattering rates will be calculated within the
dielectric continuum model with electrostatic boundary conditions. As discussed in
Appendix C, intrasubband and intersubband transitions rates are the same in either
the slab or guided models as long as complete sets of orthogonal phonon modes are
used in performing the calculation.
{e,a}
As in the previous subsection, the transition rate, S1D (k, k ), from
4 an ini-

tial electron state |k1D and an initial phonon state Nq + 2 ± 2 – denoted
1 1
4 5
and a final phonon state
by k1D , Nq + 12 ± 12 – to a final electron state k1D
6 6
, N + 1 + 1 – per unit time per unit length, L,
Nq + 12 + 12 – denoted by k1D q 2 2
is given by
{e,a} 2π {e,a} 2
S1D (k, k ) = M1D (q) δ(E(k ) − E(k) ± h̄ω), (8.93)
h̄ q
where
6 4
{e,a} 1D
M1D (q) = k1D , Nq + 1
2 + 12 HFr k1D , Nq + 1
2 ± 1
2 . (8.94)
The electron wavefunction is taken to be

eik x x 2 πy 2 πz
|k1D = √ cos cos , (8.95)
Lx Ly Lz Lz Lz
within the finitely long quantum wire with cross-sectional area defined by −L y <
y < L y /2, −L z < z < L z /2 and is taken to vanish outside this region. In this
so-called extreme quantum limit, the electron energy for the ground state is

h̄ 2 k x2 h̄ 2 π 2 1 1
E 1D = + + 2 . (8.96)
2m ∗ 2m ∗ L 2y Lz
Then the procedures of the last subsection lead straightforwardly to

{e,a} 1/2
M1D,m,n (q) = ∓2α δk x −k x ∓qx λmn 2Pmn n q + 12 ± 12 , (8.97)
where λmn is given in (8.96) and where (Stroscio, 1989)

L y /2 L z /2
dy dz πy πz
Pmn = cos2 cos2
L
−L y /2 y /2 L
−L z /2 z /2 L y Lz
 mπ y nπ z

 cos cos

 L Lz


y

 mπ y nπ z


 cos L sin L
y z
× (8.98)


mπ y nπ z

 sin cos

 Ly Lz



 mπ y nπ z
 sin sin
Ly Lz
The four expressions for Pmn correspond to the four terms of (8.85). The numerical
values for the dominant values of Pmn are: P11 = (8/3π)2 , P13 = P31 =
(1/5)P11 , P15 = P51 = −(1/35)(8/3π )2 , P33 = (1/25)(8/3π)2 , P35 = P53 =
−(1/175)(8/3π)2 , and P55 = (1/1225)(8/3π )2 . The transition rate is then given
by
1
{e,a}
τ1D (k x )
+∞
|α |2 V
= nq + 1
2 ± 1
2 dqx I1D (qx , L y , L z )δ(E(k x ± qx ) − E(k x ) ± h̄ω),
4π 2 h̄ −∞
(8.99)
where
(2π )2
I1D (qx , L y , L z ) = 16 |Pmn |2 λmn . (8.100)
L y Lz m=1,3,5,... n=1,3,5,...
Figure 8.7 provides numerical values for I1D (qx , L y , L z ) for wires with selected
cross-sectional dimensions.
Rewriting the argument of the delta function δ(E(k x ±qx )− E(k x )± h̄ω) through
the use of
h̄ 2 {e,a} {e,a}
E(k x ± qx ) − E(k x ) ± h̄ω = ∗
q x − q+ q x − q− , (8.101)
2m
where
1/2
{e,a} 2m ∗ ω
q± = k x ± k x2 − (8.102)
h̄
with = 1 for emission and = −1 for absorption, it follows that

{e,a} {e,a}
1 αω 1 + I1D q+ + I1D q−
{e,a}
= nq + √ , (8.103)
τ1D (k x ) 2π 2 E/h̄ω −
with

1 e2 1 1
α= √ − . (8.104)
2h̄ω h̄/2mω (∞) (0)
8.3.3 Scattering rate for interface-LO phonon modes

The Fröhlich interaction Hamiltonian, HIF 1D , associated with the interface phonons in
rectangular polar-semiconductor quantum wires has been derived analytically using

an approximation method (Stroscio et al., 1990; Kim et al., 1991) involving the
separation of variables (Marcatili, 1969) and has been treated exactly by the use
of numerical techniques (Knipp and Reinecke, 1992). As a result, it is known that
the approximate separation-of-variables approach fails to predict so-called ‘corner’
modes considered by Knipp and Reinecke – the well-known ‘corner’ problem.
Unlike the confined optical phonon modes of the last subsection, the interface modes
do not vanish on the interfaces of the quantum wire. Therefore, it is clear that
difficulties will be encountered in the vicinity of the corners when attempting to have
consistent boundary conditions on adjacent sides of the quantum wire. Nevertheless,
Figure 8.7. Numerical

values for I1D (qx , L y , L z )
for wires with selected
cross-sectional dimensions
defined by the four pairs
(L y /L 0 , L z /L 0 ) =
(1, 1), (2, 2), (3, 3), (4, 4),
where L 0 = [h̄/(2m ∗ ω)]1/2 .
The solid and broken lines
represent results obtained for
bulk phonons and for the
lowest-order confined modes
respectively.
the approximate solution based on separation of variables will be presented since

it provides physical insight into the nature of the interface modes. Subsequent to
deriving the IF modes in this approximation, the exact treatment of Knipp and
Reinecke (1992) will be presented and the corner modes discovered by Knipp and
Reinecke will be described.
Taking the quantum wire to have infinite extent in the x-direction, the electrostatic
potential, (r), may be written as a one-dimensional Fourier transform:

(r) = (qx , y, z) e−iqx x . (8.105)
qx
Then

E(r) = −∇(r) = E(qx , y, z) e−iqx x ,
qx
(8.106)
P(r) = (ω)E(r) = P(qx , y, z) e−iqx x .
qx
The normalization condition is given in terms of u i (qx , y, z) by

+L y /2 +L z /2 2
h̄
dy dz ni µi Lu i (qx , y, z) = , (8.107)
−L y /2 −L z /2 2ω
where the subscript i refers to a general material region i. From the results of
subsection 7.3.4 for a quantum well, it is clear that
|µi ni u i (qx , y, z)|2 1 ωLO,i

2 − ωTO,i
2
= i (∞)
∂i (qx , y, z) 2 4π (ω2 − ωTO,i
2 )2
qx2 |i (qx , y, z)|2 +

∂z
1 1 ∂i (ω)
= , (8.108)
4π 2ω ∂ω
holds for the case of a quantum wire and it follows that

1 1 ∂i (ω) ∂i (qx , y, z) 2
dz qx |i (qx , y, z)| +
2 2 = h̄ ,
4π 2ω ∂ω ∂z 2ωL
i Ri
(8.109)
where the sum over i is over the two material regions in the problem at hand: i = 1
inside and i = 2 outside the quantum wire. The Fröhlich interaction Hamiltonian,
1D , is then given by
HIF

1D
HIF = −e (qx , y, z) e−iqx x [AIF (qx ) + A†IF (−qx )], (8.110)
qx
where AIF (qx ) and A†IF (−qx ) are the annihilation and creation operators for the
appropriate IF-phonon modes of the quantum wire and where in the separation-of-
variables approximation (qx , y, z) has the form


 cosh αy cosh βy

 C |y| ≤ L y /2, |z| ≤ L z /2,

 cosh αL y /2 cosh β L z /2



 C cosh αy eβ L z /2 e−β|z|

|y| ≤ L y /2, |z| ≥ L z /2,
s (qx , y, z) = cosh αL y /2

 cosh βy



 CeαL y /2 e−α|y| |y| ≥ L y /2, |z| ≤ L z /2,

 cosh β L z /2

 αL y /2 −α|y| β L z /2 −β|z|
Ce e e e |y| ≥ L y /2, |z| ≥ L z /2,
(8.111)
for the symmetric IF mode and


 sinh αy sinh βy

 +C |y| ≤ L y /2, |z| ≤ L z /2,

 sinh αL y /2 sinh β L z /2



 sinh(αy) β L z /2 −β|z|
 ∓C e e |y| ≤ L y /2, |z| ≥ L z /2,
s (qx , y, z) = sinh(αL y /2)



 sinh(βy)

 ∓CeαL y /2 e−α|y| |y| ≥ L y /2, |z| ≤ L z /2,

 sinh(β L z /2)


±CeαL y /2 e−α|y| eβ L z /2 e−β|z| |y| ≥ L y /2, |z| ≥ L z /2,
(8.112)
for the antisymmetric IF modes; in these potentials, α and β satisfy the relation
α 2 + β 2 − qx2 = 0, (8.113)
and in (8.112) the plus sign is taken when yz > 0 and the minus sign when
yz < 0. As usual, the dispersion relations are obtained by imposing the boundary
conditions. For the symmetric solution this dispersion relation is
1 (ω) tanh αL y /2 + 2 (ω) = 0, (8.114)
and for the antisymmetric mode the dispersion relation is
1 (ω) coth αL y /2 + 2 (ω) = 0, (8.115)
where
αL y = β L z (8.116)
must also be satisfied for both modes. These dispersion relations are of the same
forms as those for a quantum well, as is expected since the separation-of-variables
approximation has been assumed thus far in this treatment of the optical phonons
in quantum wires. The numerical solutions for these dispersion relations are
depicted in Figure 8.8 for the case where L y = L z = d. In this figure the
dispersion relations are given for a GaAs quantum wire in two cases: a free-
standing quantum wire having frequencies ω S, f and ω A, f for the symmetric and
antisymmetric interface modes respectively, and a GaAs quantum wire embedded
in AlAs having frequencies for the symmetric and antisymmetric interface modes
of ω S,± and ω A,± respectively. The subscripts ± designate the high-frequency
(+) and low-frequency (−) modes for both the symmetric and antisymmetric
cases.
Figure 8.8. Frequency versus wavevector for the IF optical phonons of a

free-standing GaAs quantum wire having frequencies ω S, f and ω A, f for the
symmetric and antisymmetric interface modes respectively, and a GaAs quantum
wire embedded in AlAs having frequencies ω S,± and ω A,± for the symmetric and
antisymmetric interface modes respectively. The subscripts ± designate the
high-frequency (+) and low-frequency (−) modes for both the symmetric and
antisymmetric cases. From Kim et al. (1991), American Institute of Physics, with
permission.
The normalization condition leads to the following equation in terms of the

normalization constant, C, for the symmetric mode:

ωLO,1 − ωTO,1 β L z −2
2 2
h̄ −1 −2 αL y
4π L C = 1 (∞) cosh cosh
2ω ω2 − ωTO,1
2 2 2
/
α2 L z sinh αL y sinh β L z
× sinh αL y +
2 α αβ
0
β2 L y sinh αL y sinh β L z
+ sinh β L z +
2 β αβ

ωLO,2 − ωTO,2 αL y −2
2 2
+ 2 (∞) cosh
ω2 − ωTO,22 2

sinh a L y sinh αL y Ly
× α2 + β2 +
αβ αβ β
−2
ωLO,2 − ωTO,2
2 2
β Lz
+ 2 (∞) cosh
ω − ωTO,2
2 2 2

2 sinh β L z Lz 2 sinh αL y
× α + +β
αβ α αβ
/ 2 0
ωLO,2 − ωTO,2 2 α + β 2
2 2
+ 2 (∞) . (8.117)
ω2 − ωTO,22 αβ
A similar expression may be derived straightforwardly for the antisymmetric mode.

{e,a}
As in the previous subsection, the transition rate, S1D,IF (k, k ), from an ini-
4

tial electron state |k1D and an initial phonon state Nq + 12 ± 12 – denoted
4 5
and a final phonon state
by k1D , Nq + 12 ± 12 – to a final electron state k1D
6 6
, N + 1 + 1 – per unit time per unit heterostruc-
Nq + 12 + 12 – denoted by k1D q 2 2
ture area A is given by
{e,a} 2π {e,a} 2

S1D,IF (k, k ) = M1D,IF (qx ) δ(E(k ) − E(k) ± h̄ω), (8.118)
h̄ q
6 4
{e,a} 1D
M1D,IF (qx ) = k1D , Nq + 12 + 12 HIF k1D , Nq + 12 ± 12 . (8.119)
The electron wavefunction and the electron energy for the ground state are as
given in the last subsection. Then the procedures of the last subsection lead
straightforwardly to the equation
{e,a} {e,a}
1 α ω 1 + I1D,IF q+ + I1D,IF q−
{e,a}
= nq + √ , (8.120)
τ1D,IF (k x ) 2π 2 E/h̄ω −
with = 1 for emission and = −1 for absorption,

1/2
{e,a} 2m ∗ ω
q± = k x ± k x2 − (8.121)
h̄
and
1 e2
α = √ . (8.122)
2h̄ω h̄/2mω
Furthermore,

(2π )2 L
I1D,IF = C 2 Ps2 , (8.123)
ω2 4π
where
1
Ps =
cosh αL y /2 cosh β L z /2
L y /2 L z /2
dy dz πy πz
× cos2 cos2
−L y /2 L y /2 −L z /2 L z /2 Ly Lz
× cosh αy cosh βz
1
=
cosh αL y /2 cosh β L z /2
4 sinh(αL y /2) sinh (β L z /2)
× . (8.124)
αβ L y L z (αL y /2π )2 + 1 (β L z /2π)2 + 1
Only the symmetric mode contributes to the scattering rate for the ground state
electronic wavefunction in the extreme quantum limit.
{e,a} {e,a}
The scattering rate 1/τ1D,IF (k x ) has the same form as 1/τ1D (k x ), (8.103), of

the last subsection. Indeed, α and α differ only by the factor −1 (∞) − −1 (0) ,
and I1D is replaced by I1D,IF . Otherwise the expressions are identical. Therefore, it
follows that the scattering rate due to both confined optical phonons and interface
optical phonons is given by
{e,a} {e,a}
1 α ω 1 + Iwire q+ + Iwire q−
{e,a}
= nq + √ , (8.125)
τwire (k x ) 2π 2 E/h̄ω −
where

1 1
Iwire = − I1D + I1D,IF . (8.126)
(∞) (0)
The electron–optical-phonon scattering rates for emission and absorption for se-
lected phonon modes are displayed in Figure 8.9 for a 40 Å × 40 Å GaAs quantum
wire embedded in AlAs. The interface phonons are labeled as SO (surface-optical)
modes and the confined phonons are labeled simply as LO modes.
The one-dimensional density-of-states peaks are prominent features of the curves
representing the phonon emission rates. The electrons are in the ground states in the
y- and z-directions. The maximum scattering rates for SO-phonon emission occur
at the energies of the symmetric high-frequency interface mode (about 50 meV) and
the symmetric low-frequency interface mode (about 33 meV).
Knipp and Reinecke (1992) examined the accuracy of using the separation-
of-variables approximation to the dielectric continuum model for calculating the
Fröhlich potential of the interface optical phonons in quantum wires. They consid-
Figure 8.9. The electron–optical-phonon scattering rates for emission and

absorption for selected phonon modes for a 40 Å × 40 Å GaAs quantum wire
embedded in AlAs. The interface phonons are labeled as SO (surface-optical) modes
and the confined phonons are labeled simply as LO modes. From Kim et al. (1991),
ered a variety of wire geometries, as illustrated by the cross-sectional areas of Figure

8.10.
As indicated in Figure 8.10, the material inside the quantum wire has dielectric
constant 1 (ω) and the surrounding material has dielectric constant 2 (ω). For the
cases of quantum wires with circular and elliptical cross sections these workers
obtained elegant analytical results. For quantum wires with nearly rectangular cross
sections having rounded corners they used numerical techniques to determine the
Fröhlich potentials.
An interesting feature of the calculations of Knipp and Reinecke is the localiza-
tion of interface modes near the corners of the nearly rectangular quantum wires in
the vicinity of the rounded corners, as illustrated in Figure 8.11. In Figure 8.11 the
Fröhlich potential in the plane of the quantum wire – denoted by φ – is depicted for
the mode of highest azimuthal symmetry (m = 0) for a quantum wire with R/r = 2
and for a corner curvature a = r/10. The Fröhlich potential shown in Figure 8.11
represents the case where qr = 2.
The Fröhlich potential in the plane of an elliptical quantum wire with R/r = 2
is shown in Figure 8.12. The potential shown in Figure 8.12 is again the Fröhlich
potential for the mode of highest azimuthal symmetry (m = 0), with qr = 2. The
potential is enhanced in the regions near the vertices of the rectangle into which the
ellipse fits.
Figure 8.10. Cross sections considered by Knipp and Reinecke (1992). The
long-broken line represents an elliptical cross section. The shorter broken line
designates a cross section that is basically elliptical but has flattened sides. The solid
and dotted lines (visible at the corners) represent ‘rectangles’ with rounded corners.
The geometrical parameters R and r are as indicated in this figure. From Knipp and
Reinecke (1992), American Physical Society, with permission.
8.3.4 Collective effects and non-equilibrium phonons in

polar quantum wires
Campos et al. (1992) and Das Sarma et al. (1992) considered the influence of
collective carrier screening on carrier–optical-phonon scattering in quantum wires.
Since the Fröhlich Hamiltonian represents the interaction between a carrier and the
fields produced by polar optical phonons, it is clear that the screening of such a
Coulomb-like interaction by carriers is possible.
To estimate the magnitude of such a collective screening of the Fröhlich inter-
action, Campos et al. and Das Sarma et al. calculated the power loss per carrier
caused by carrier–optical-phonon interactions in a GaAs quantum wire for the case
where dynamical screening, phonon confinement, and the presence of the so-called
‘hot-phonon’ effect are all taken into account. The energy loss rate of a carrier in
such a polar-semiconductor quantum wire is determined by both the rate at which
the carrier’s energy is lost by phonon emission and the rate at which the carrier gains
Figure 8.11. Interface modes in a nearly rectangular quantum wire. The Fröhlich
potential φ in the plane of the quantum wire is depicted for the mode of highest
azimuthal symmetry (m = 0) for a quantum wire with R/r = 2 and for a corner
curvature a = r/10; qr = 2. From Knipp and Reinecke (1992), American Physical
energy from phonon absorption. This latter rate can be significant in quantum wires
and quantum wells, since the phonons emitted by energetic carriers can accumulate
in these structures. Indeed, the phonon densities in many dimensionally confined
semiconductor devices are typically well above those of the equilibrium phonon
population and there is an appreciable probability that these non-equilibrium – or
‘hot’ – phonons will be reabsorbed. The net loss of energy by a carrier in such a
situation depends on the rates for both phonon absorption and emission.
The lifetimes of the optical phonons are also important in determining the total
energy loss rate for such carriers. As discussed in Chapter 6, the longitudinal optical
(LO) phonons in GaAs and in many other polar materials decay into acoustic
phonons through the Klemens’ channel. Over a wide range of temperatures and
phonon wavevectors, the lifetimes of these longitudinal optical phonons in GaAs
vary from a few picoseconds to about 10 ps (Bhatt et al., 1994). Typical lifetimes for
other polar semiconductors are also of this magnitude. The LO phonons undergoing
decay into acoustic phonons are not available for absorption by the carriers and,
Figure 8.12. The Fröhlich potential in the plane of an elliptical quantum wire with
R/r = 2. This potential was derived analytically by Knipp and Reinecke (1992) for
the mode of highest azimuthal symmetry (m = 0), with qr = 2. From Knipp and
Reinecke (1992), American Physical Society, with permission.
as a result of the Klemens’ channel, the carrier energy loss is more rapid than it
would be otherwise; this phenomenon is referred to as the ‘hot-phonon-bottleneck’
effect. In fact, the numerical calculations of Campos et al. and Das Sarma et al. for
a 50 Å × 50 Å GaAs quantum wire indicate, over a temperature range from 50 K to
300 K, for wire dimensions from 50 Å × 50 Å to 500 Å × 500 Å, and for a linear
density range of 104 to 106 cm−1 , that the hot-phonon-bottleneck effect changes the
net power loss by an order of magnitude or more while the effect of carrier screening
results in a change of only, roughly, a factor of two.
Figure 8.13 illustrates these findings for the case of a linear charge density of
106 cm−1 (Campos et al., 1992). The power loss P per carrier is given as a function
of the inverse electron temperature for both the slab modes and the guided modes.
The curve of highest power loss (long-broken line) is calculated for the slab modes
with no screening effects. The thinner solid line represents the power loss with
screening effects included for a phonon decay time τphonon of 0 ps; the thicker solid
line represents the power loss with screening effects included for a phonon decay
time τphonon of 7 ps. For the case of the guided modes, the unscreened power loss
function is represented by the broken-and-dotted line, and the screened power loss
functions for τphonon = 0 ps and τphonon = 7 ps are represented by the thinner and
thicker short-broken lines respectively.
Figure 8.13. Results of Campos et al. for a 50 Å × 50 Å GaAs quantum wire for the
case of a linear charge density of 106 cm−1 . The power loss per carrier is given as a
function of the inverse electron temperature. The curve of highest power loss
(long-broken line) is calculated for the slab modes with no screening effects. For the
guided modes, the unscreened power loss function is represented by the
broken-and-dotted line and the screened power loss functions for τphonon = 0 ps and
τphonon = 7 ps are represented by the thinner and thicker short-broken lines
respectively. From Campos et al. (1992), American Physical Society, with
permission.
Thus while dynamical screening effects can be noticeable the effects of the
hot-photon bottleneck can lead to a change of an order of magnitude or more in the
carrier energy loss rates associated with carrier–LO-photon scattering in quantum
wires. The screening calculations of Campos et al. (1992) and Das Sarma et al.
(1992) involve non-trivial numerical calculations. For additional information on
dynamical screening and hot-phonon effects the reader is referred to these papers.
8.3.5 Reduction of interface–phonon scattering rates in

metal–semiconductor structures
The first calculations of carrier–LO-phonon scattering rates in quantum wires made
of polar semiconductors (Stroscio et al., 1990; Kim et al., 1991) revealed that
the scattering of carriers with interface optical phonon modes dominates over
the scattering associated with confined and half-space phonons. As a means of
reducing the unwanted inelastic scattering caused by the interface LO phonons in
semiconductor quantum wires, it was suggested (Stroscio et al., 1992) that the large
carrier–LO-phonon scattering caused by interface phonons could be eliminated by
encapsulation of the quantum wire in a metal. This proposal is of little use for most
metal–semiconductor systems, but such a metal–semiconductor interface would lead
to a depletion of the semiconductor when the cross-sectional dimensions – L y and
L z – of the quantum wire are small enough (i.e., less than about q −1 = 1/0.02 Å)
for the scattering rate due to interface phonons to be appreciable.
However, in some metal–semiconductor systems, such as the InAs–Al system,
the Fermi level is pinned so that there is no depletion of carriers in the semiconductor
but an accumulation of carriers near the metal–semiconductor interface. In cases
where the semiconductor is not depleted, quantum wires with L y , L z q −1 =
1/0.02 Å will exhibit large carrier–LO-interface phonon scattering rates detrimental
to high-mobility transport.
To reduce these effects, the interface optical phonons may indeed be modified or
practically eliminated through the use of metal boundaries at the semiconductor
interfaces. The highly mobile carriers in a metal generally screen electric fields
and these fields penetrate only very short distances, of order δ, into the metal; δ
is known as the Thomas–Fermi screening length and is generally of the order of an
ångstrom or so. The significance of this strong screening of electric fields near the
metal–semiconductor interface is that the phonon potential φ must vanish near the
interface. In such a situation, it is clear that both the symmetric and antisymmetric
interface optical phonons will have very small amplitudes in quantum-wire and in
quantum-well structures.
To estimate the magnitude of this reduction, Bhatt et al. (1993a) consid-
ered a metal–semiconductor–semiconductor (MSS) double-heterointerface structure
with the metal–semiconductor interface at z = −d and the semiconductor–
semiconductor interface at z = d. For this heterostructure, the interface phonon

potential is of the form

 Ae
z/δ z < −d,
ρ
φ(r ) = φ0 e iq·ρ
× Beqz + Ce−qz |z| < d, (8.127)
 −q(z−d)
e z > d,
where the constants A, B, and C as well as the dispersion relation for the mode
are determined by the conditions that φi and i ∂φi /∂z are continuous at the
two heterointerfaces. It is straightforward to show by imposing these boundary
conditions that
−1
1 m
A = eqd 1+ ,
1 − 2 qδ1

1 −qd 2
B= e 1− , (8.128)
2 1

1 qd 2
C= e 1+
2 1
and that the dispersion relation is

1 − 2 −2qd qδ1 − m
e − 1 = 0. (8.129)
1 + 2 qδ1 + m
In these results, the dielectric function of the metal is given by
ω2p τ 2
m = 1 − , (8.130)
1 − ω2 τ 2
where ω p is the plasma frequency for the carriers in the metal and τ is the dielectric
relaxation time for the carriers in the metal.
Since the ratio |A/B| is a measure of the amplitude of the IF optical phonon
mode in the metal and since

A 2
=
B 1 + /qδ 1, (8.131)
m 1
for typical metal–semiconductor structures, it follows that the fields associated

with the IF optical phonon mode are damped dramatically by electron screening
as it enters the metal. Thus, to a very good approximation, φ(r ) = 0 at a
metal–semiconductor interface. Indeed, numerical calculations (Bhatt et al., 1993a)
show that |A/B| is in the range 4 × 10−3 –10−8 when δ is in the range 1–10
Å and q = 0.01 Å−1 . Since the IF optical phonon modes, in general, have
non-zero amplitudes at heterointerfaces, it follows that these modes do not satisfy
the boundary condition φ(r ) = 0. Accordingly, these results imply that the IF
modes suffer dramatic reductions in amplitude when semiconductor–semiconductor
heterostructures are replaced with metal–semiconductor structures. Indeed, for the

metal–semiconductor–metal (MSM) structure of Figure 8.14, it follows trivially that
the optical phonon potentials should vanish at both z = 0 and z = d.
Clearly, neither the symmetric nor the antisymmetric optical phonon modes
can satisfy these conditions. The MSM structure of Figure 8.14 does, however,
support confined LO phonons. It is straightforward to extend these conclusions
to the case of a semiconductor quantum wire embedded in a metal. In closing
we note that Constantinou (1993) also considered the reduction of IF optical
phonon mode amplitudes in metal–semiconductor systems. The concept of using
a metal–semiconductor interface to reduce electron–interface–phonon interactions
in quantum wires was considered further by Dutta et al. (1993).
8.4 Scattering of carriers and LO phonons in quantum

dots
As discussed in Sections 5.2 and 7.1, the Fröhlich Hamiltonian for a bulk polar
semiconductor may be written as
(3D)
C $ †
%
HFr = aq − a−q eiq·r , (8.132)
q |q|
where

2π e2 h̄ωLO 1 1
C = −i − . (8.133)
V (∞) (0)
Consider a quantum dot with the geometry of a free-standing cube each of whose
sides is of length L (de la Cruz et al., 1993). Then by writing the sum over q as a
Figure 8.14. A metal–semiconductor–metal heterostructure in which the

metal–semiconductor interfaces are at z = 0 and z = d. Interface LO phonon modes
are not present in such a structure since the LO phonon potential vanishes at each
metal–semiconductor interface. From Bhatt et al. (1993a), American Institute of
Physics, with permission.
sum over positive values of qx , q y , and qz , expanding the exponentials in a manner

similar to that for a free-standing quantum wire (Stroscio, 1989), and requiring that
qx = ±mπ/L, q y = ±nπ/L, and qz = ± pπ/L, so that the Fröhlich potential
vanishes on the faces of the box, it follows that the Fröhlich Hamiltonian for the
quantum box must have the form
!
(0D)
HFr = 23/2 C β(m, n, p)
m=1,3... n=1,3... p=1,3...
mπ x nπ y pπ z 2 3
× cos cos cos A++ (q)+ + A†++ (−q)+
L L L
mπ x nπ y pπ z
+ β(m, n, p) sin cos cos
m=2,4... n=1,3... p=1,3...
L L L
2 3
× A+− (q)+ + A†+− (−q)+
mπ x nπ y pπ z
+ β(m, n, p) cos sin cos
m=1,3... n=2,4... p=1,3...
L L L
2 3
× A−+ (q)+ + A†−+ (−q)+
mπ x nπ y pπ z
+ β(m, n, p) sin sin cos
m=2,4... n=2,4... p=1,3...
L L L
2 3
× A−− (q)+ + A†−− (−q)+
mπ x nπ y pπ z
+ β(m, n, p) cos cos sin
m=1,3... n=1,3... p=2,4...
L L L
2 3
× A++ (q)− + A†++ (−q)−
mπ x nπ y pπ z
+ β(m, n, p) sin cos sin
m=2,4... n=1,3... p=2,4...
L L L
2 3
× A+− (q)− + A†+− (−q)−
mπ x nπ y pπ z
+ β(m, n, p) cos sin sin
m=1,3... n=2,4... p=2,4...
L L L
2 3
× A−+ (q)− + A†−+ (−q)−
mπ x nπ y pπ z
+ β(m, n, p) sin sin sin
m=2,4... n=2,4... p=2,4...
L L L
2 3"
× A−− (q)− + A†−− (−q)− ,
(8.134)
where
$ 1/2
mπ %2 $ nπ %2 $ pπ %2
β(m, n, p) = + + , (8.135)
L L L
and where the phonon annihilation operators for the zero-dimensional confined
modes are
1
A++ (q)± = √ A+ (qx )± + A+ (−qx )± ,
2
−i
A+− (q)± = √ A+ (qx )± − A+ (−qx )± ,
2
(8.136)
1
A−+ (q)± = √ A− (qx )± + A− (−qx )± ,
2
−i
A−− (q)−− = √ A− (qx )± − A− (−qx )± .
2
The operators A± (qx )± can be written in terms of the annihilation operator for a
three-dimensional semiconductor, in a manner analogous to that employed for the
quantum wire (Stroscio, 1989), through use of the relations
1
A+ (qx )± = √ a± (qx , q y ) + a± (qx , −q y ) ,
2
(8.137)
−i
A− (qx )± = √ a± (qx , q y ) − a± (qx , −q y ) ,
2
where a± (qx , q y ) are defined by
1
a+ (qx , q y ) = √ a(qx , q y , qz ) + a(qx , q y , −qz ) ,
2
(8.138)
−i
a− (qx , q y ) = √ a(qx , q y , qz ) − a(qx , q y , −qz ) .
2
For the case of a GaAs quantum dot in vacuum, the electron energy levels are
generally separated by more than the typical longitudinal optical phonon energy,
h̄ωLO , unless the dimensions of the dot are such that L is greater than several
hundred ångstroms. Phonon confinement effects are generally not significant for
such large confinement lengths. Specifically, for such a GaAs quantum ‘box’ the
confined phonons and the interface phonons have energies within a few meV of the
bulk phonon energy, 36 meV. In the case where the quantum dot is embedded, in, say,
AlAs, there would be interface phonons with energies close to 50 meV but even these
energies are small compared to typical separations between electron energy states
unless the length L is several hundred ångstroms or more. For holes, the de Broglie
wavelength is considerably shorter and energy level separations are smaller than
for the case of electrons. Accordingly, for transitions involving holes the quantum
‘box’ may be small enough for phonon confinement effects to be significant. It will
be instructive to summarize key features of the calculation of de la Cruz (1993)
because it illustrates the general procedure within the effective-mass approximation
and the Fermi golden rule approximation (Appendix E).
For a ‘box’ in vacuum with infinitely high potential barriers at its boundaries,
defined by −L/2 ≤ x, y, z ≤ L/2, the carrier wavefunctions are represented by

3 2 iπ x jπ y lπ z
|i, j, l = cos cos cos , for odd i, j, l,
L L L L
(8.139)
3 2 iπ x jπ y lπ z
|i, j, l = sin sin sin , for even i, j, l.
L L L L
The carrier eigenenergies for this system in the effective-mass approximation, with
effective mass m ∗ , are given by
/ 2 2 2 0
h̄ 2 iπ jπ lπ
E i, j,l = + + . (8.140)
2m ∗ L L L
The transition probability given by the Fermi golden rule is then
2π {e,a} 2
W {e,a} = M δ(E i , j ,l − E i, j,l ± h̄ωLO ), (8.141)
h̄
with
5 7
(0D)
M {e,a} = i , j , l ; Nq + 1
2 ± 1
2 HFr i, j, l; Nq + 1
2 ± 1
2 (8.142)
where, as discussed several times previously, the signs are chosen to describe
the desired phonon process. The energy-conservation constraint required by the
δ-function may be expressed as
$ % $ % $ % 2m ∗ L 2
i 2 − i 2 + j 2 − j 2 + l 2 − l 2 ± ωLO = 0. (8.143)
h̄π 2
For the system under consideration, non-zero matrix elements correspond to tran-
sitions between states of the same parity. Accordingly, the transitions corresponding
to the smallest quantum boxes are those between the |1, 1, 1 states and states where
the carrier quantum numbers, i , j , l and i, j, l, are selected from 1, 3, and 5.
For these latter states, all the degenerate states for a given set of carrier quantum
numbers are denoted by {|i, j, l}. The transition probabilities for a selection of
such transitions were calculated by de la Cruz (1993) for GaAs at room temperature.
These probabilities are summarized in Table 8.1, where, in the case of emission, the
carrier transition occurs between the indicated set of degenerate states {|i, j, l} and
the |1, 1, 1 state. Likewise, for the case of absorption, the initial state is the |1, 1, 1
state and the final state is indicated by the specific set {|i, j, l} corresponding to the
transition.
Table 8.1. Transition probabilities W for LO phonon emission and absorption

resulting from the Fröhlich interaction in a GaAs quantum box of side L at room
temperature. The |1, 1, 1 state is the final state in emission and the initial state in
absorption.
L (nm) Wemission Initial states Wabsorption Final states

(× 1013 s−1 ) (×1012 s−1 )
352 2.51 {|3, 1, 1} 5.77 {|3, 1, 1}
498 1.71 {|3, 3, 1} 3.93 {|3, 3, 1}
610 1.02 {|3, 3, 3} 2.33 {|3, 3, 3}
610 1.69 {|5, 1, 1} 3.88 {|5, 1, 1}
704 1.20 {|1, 3, 5} 2.76 {|1, 3, 5}
787 0.81 {|5, 3, 3} 1.86 {|5, 3, 3}
863 0.91 {|5, 5, 1} 2.09 {|5, 5, 1}
1056 0.86 {|5, 5, 5} 1.56 {|5, 5, 5}
Chapter 9
Carrier–acoustic-phonon scattering
O praise ye the Lord, all things that give sound, each jubilant
chord re-echo around; Loud organs, his glory tell in deep tone
and, sweet harp, the story of what he hath done.
Henry Williams Baker, 1875
9.1 Carrier–acoustic-phonon scattering in bulk

zincblende structures
In this section the strain tensor introduced in Section 7.2 will be related to both the
deformation-potential interaction and the piezoelectric interaction.
9.1.1 Deformation-potential scattering in bulk zincblende

structures
The deformation-potential interaction of Section 5.3 applies to the case of an
isotropic medium. In the general case of a carrier band with energy E α (k), the
deformation potential depends on the strain tensor and the deformation-potential
α , must also be represented as a second-rank tensor; accordingly,
constant, E µν1
unpert α
Hα,def = E αstrained (k) − E α (k) = E µν1 Sµν , (9.1)
where Sµν is the strain tensor defined in Section 7.2, (7.27), and the deformation
potential is the difference in the energy of band α when strain is present, E αstrained (k),
unpert
and the energy of band α when there is no strain, E α (k). Since the phonon field is
always present, the case where there is no strain is unphysical; it corresponds to the
situation where there is no deformation potential. For cubic crystals and for carriers
experiencing a spherically symmetric energy surface, symmetry considerations
make it evident that the crystalline potential should be proportional to ∇V /V , as
in Section 5.3. This conclusion holds for carriers at a non-degenerate point but
172
9.1 Carrier–acoustic-phonon scattering in bulk zincblende structures 173
not for the cases of other symmetry points such as the X or L points. Accordingly,
for the case of a cubic crystal and for carriers at the point,
Hα,def = E 1α ∇V /V = E 1α ∇ · u, (9.2)
or, in the notation of Section 7.2,
Hα,def = E 1α , (9.3)
since here = ∇ ·u. Consider an acoustic mode of wavevector q and a displacement

field which varies as e−iq·r . Then (r) =∇ · u = −iq · u and it follows that
longitudinal acoustic (LA) but not transverse acoustic (TA) modes lead to non-zero
values of Hα,def . However, for cubic crystals with carriers at energy minima at other
symmetry points – such as X or L points – E µν1 α does not reduce to a scalar but
instead takes the form
α α α
E µν1 = E 1l δµν + E 1t iµiν , (9.4)
where the minimum is on an axis along the unit vector î. For this more general case,
α term contributes to the deformation potential and, in general, the TA and
the E 1t
optical phonons produce deformation potentials leading to intravalley processes.
9.1.2 Piezoelectric scattering in bulk semiconductor

structures
As described in Section 5.4, piezoelectric scattering results from the interaction of
carriers with the macroscopic piezoelectric polarization produced by strain fields, in
polar crystals lacking an inversion symmetry. These strain fields may be generated
either externally as, for example, by a mismatch between the lattice constants in
pseudomorphic systems, or internally, as by an acoustic phonon. In the notation of
Section 7.2, the piezoelectric polarization of the cubic crystal, considered in Section
5.4 takes the general form
piezo
Pλ = eλ,µν Sµν , (9.5)
where Sµν is the strain tensor and eλ,µν is the generalized tensor describing the
piezo
piezoelectric coupling of the polarization, Pλ , to the strain field Sµν . In practice,
the components of eλ,µν may be determined through knowledge of the strain and
electromagnetic fields and the static dielectric constant, λµ
0 , via the relation
1 0
eλ,µν Sµν = Pλ − ( − δλµ )E µ . (9.6)
4π λµ
As described previously (Vogl, 1980), eλ,µν is manifestly a bulk quantity. In
general, Pλ and E λ depend on the geometry of the crystal structure but the
174 9 Carrier–acoustic-phonon scattering
components of eλ,µν are bulk-like quantities determined by the local electronic

structure of the crystal.
The piezoelectric interaction may be modeled in terms of the interaction of
carriers with the macroscopic electric potential produced by the piezoelectric field.
To calculate this potential for a bulk semiconductor, the phonon displacement is
taken to vary as e−iq·r , and it follows that ∇ · u = iq · u. In a region without free
charges ∇ · D = 0. Hence, applying qλ to
Dλ = 4πeλ,µν Sµν + λµ
0
Eµ (9.7)
it follows that
qλ λµ
0
E µ = −4πqλ eλ,µν Sµν . (9.8)
For the phonon displacement field under consideration, Sµν = 12 i(qµ u ν + u ν qµ )

and, by defining the potential associated with the piezoelectric potential through
−iqV piezo (q) = E, it follows that
qλ eλ,µν qµ
V piezo (q) = 4π uν . (9.9)
qλ λµ
0 q
µ
For isotropic, cubic, trigonal, tetragonal, and hexagonal structures qλ λµ

0 q reduces
µ
to relatively simple expressions since in these cases λµ may be represented as a
0
3 × 3 matrix with all non-diagonal elements equal to zero. For cubic structures,
qλ λµ
0 q reduces to 0 q · q since the three diagonal elements of 0 are all equal to
µ λµ
0.
For non-cubic crystals it is convenient to use the tensor notation of Section 7.2 to
piezo piezo
determine Pλ from Pλ = eλ,µν Sµν . As in Section 7.2, Sµν may be represented
piezo
as a six-element vector, Pλ is a three-element vector, and eλ, µν is a 3 × 6 matrix;
for zincblende structures (Auld, 1973), the 3 × 6 matrix is
 
0 0 0 ex4 0 0
 0 0 0 0 ex4 0  (9.10)
0 0 0 0 0 ex4
and for würtzite crystals this matrix takes the form
 
0 0 0 0 ex5 0
 0 0 0 ex5 0 0 . (9.11)
ex1 ex2 ex3 0 0 0
9.2 Carrier–acoustic-phonon scattering in

two-dimensional structures
For a single band and for a non-degenerate point the results of subsection 9.1.1
reduce to
Hdef = E 1 ∇V /V = E 1 ∇ · u = E 1 , (9.12)
and with the expression for the quantized displacement derived in (7.145), Section
7.6 (Bannov, 1995), it follows that

h̄ 2 3
Hdef = E 1 †
(an,q + an,−q )∇ · wn (q , z) eiq ·r
q ,n 2L ρωn (q )
2

= eiq ·r (q , n, z)(an,q + an,−q
†
) (9.13)
q ,n
where

E 12 h̄ ∂wnz (q , z)
(q , n, z) = iq · wn (q , z) + . (9.14)
2L 2 ρωn (q ) ∂z
9.3 Carrier–acoustic-phonon scattering in quantum

wires
In this section electron–acoustic-phonon scattering rates will be considered for two
different quantum-wire geometries, cylindrical and rectangular. In both cases, the
scattering rates will be calculated by applying the Fermi golden rule and by treating
the acoustic phonon modes as modes of an elastic continuum with the normalization
condition of Chapter 7.
9.3.1 Cylindrical wires

For the case of a cylindrical quantum wire, the quantization of the acoustic phonons
may be performed by taking
1
u(r) = √ [u(q, r)aq + c.c.] (9.15)
N q
and by normalizing the acoustic phonon Fourier amplitude, u(q, r, ϕ, z), according
to
2π a
1 h̄
dϕ dr r u(q, r, ϕ, z) · u∗ (q, r, ϕ, z) = . (9.16)
πa 2 0 0 2Mω q
Here a is the radius of the quantum wire, N is the number of unit cells in the
normalization volume V , aq is the phonon annihilation operator, q is the phonon
wavevector, ωq is the angular frequency of the phonon mode, M is the mass of the
ions in the unit cell, and r, ϕ, z are the usual cylindrical coordinates. As discussed
in Chapter 7, the elastic continuum model provides an adequate description of such
a quantum wire as long as the diameter of the quantum wire is several times larger
than the linear dimension of the unit cell. Within the elastic continuum model, the
equation describing the acoustic phonon modes is (Yu et al., 1996)
∂ 2u
ρ = c44 ∇ 2 u + (c12 + c44 )∇(∇ · u)
∂t 2
∗ ∂ 2u1 ∂ 2u2 ∂ 2u3

+ c î1 2 + î2 2 + î3 2 (9.17)
∂ x1 ∂ x2 ∂ x3
where c∗ = c11 − c12 − 2c44 . For the isotropic case of Section 7.6, c∗ = c11 − c12 −
2c44 = 0, and
∂ 2u
ρ = c44 ∇ 2 u + (c44 − c11 )∇(∇ · u)
∂t 2
= ρct2 ∇ 2 u + ρ(cl2 − ct2 )∇(∇ · u), (9.18)
or, equivalently,
∂ 2u
= cl2 ∇(∇ · u) − ct2 ∇ × ∇ × u, (9.19)
∂t 2
√
where cl = c11 /ρ is the sound speed for longitudinal acoustic waves and ct =
√
c44 /ρ is the sound speed for transverse acoustic waves. In considering the acoustic
phonon modes in a cylinder it is convenient to introduce two equivalent quantities,
the Young’s modulus velocity v0 and the Poisson ratio σ , through

2 (cl /ct ) − 1
1
(3cl2 − ct2 ) 2
c0 = cl and σ = . (9.20)
cl2 − ct2 (cl /ct )2 − 1
In this subsection, the electron–acoustic-phonon scattering rate is evaluated for

an isotropic, cylindrical quantum wire. From subsection 9.1.1, the deformation-
potential interaction for non-degenerate carriers at a point may be written as
Hdef = E 1 ∇ · u. In this case, only longitudinal acoustic modes contribute to the
scattering rate. As discussed by Auld (1973), the classical longitudinal modes for an
isotropic cylindrical structure have no azimuthal components and are of the form
! "
d
u r (r, z) = [B J0 (ql r ) + A J0 (qt r )] ei(qz−ωt) ,
dr
! " (9.21)
qt2
u z (r, z) = i q B J0 (ql r ) − A J0 (qt r ) e i(qz−ωt)
,
q
where q is the z-component of the wavevector, ω is the angular frequency, and J0 is
the lowest-order Bessel function of the first kind, and where ql and qt are given by
ω2
2
ql,t = − q 2. (9.22)
vl,t
2
We define dimensionless velocities, wavevectors, and frequencies by dividing by c0 ,

q0 = π/a, and ω0 = q0 c0 respectively. That is, we take c̃l,t = cl,t /c0 , q̃ = q/q0 =
q(a/π), and ω̃ = ω/ω0 = ω(a/πc0 ). Upon evaluating
2π a
1 h̄
dϕ dr r [u r (r, z)u r∗ (r, z) + u z (r, z)u ∗z (r, z)] = ,
πa 02
0 2Mωq
(9.23)
expressing u(r) in terms of u(q, r), and assuming Hdef = E 1 ∇ · u it follows that

h̄
Hdef = i E 1
n q 2N Mω
a
× √ β(q 2 + ql2 )J0 (ql r )(aq eiqz − aq† e−iqz ), (9.24)
2sn
where

q̃t4 π 2 2
sn = J (q̃t π ) + J02 (q̃t π )
q̃ 2 2 1
π
− 2q̃t2 β 2 [q̃l J1 (q̃l π )J0 (q̃t π ) − q̃t J1 (q̃t π)J0 (q̃l π)]
q̃l − q̃t2
π2 2 π2
+ q̃ 2 β 2 [J1 (q̃l π ) + J02 (q̃l π )] + q̃t2 [J12 (q̃t π) − J0 (q̃t π)J2 (q̃t π)]
2 2
π
+ 2q̃t q̃l β 2 [q̃t J0 (q̃t π )J1 (q̃l π ) − q̃l J0 (q̃l π)J1 (q̃t π)]
q̃l − q̃t2

2 2π
2
+ q̃l β [J (q̃l π ) − J0 (q̃l π )J2 (q̃l π )] ;
2
(9.25)
2 1
here β = B/A.
In classical acoustics (Auld, 1973), there are two particularly simple types
of boundary condition. They are referred to frequently as the cardinal boundary
conditions and are the free-surface boundary condition (FSBC) and the clamped-
surface boundary condition (CSBC). In the free-surface boundary condition, at the
boundary between the cylindrical wire and the vacuum the normal components of
the stress tensor are zero and the displacement is unrestricted. In the clamped-surface
boundary condition, at the boundary between the cylindrical wire and the vacuum
the normal components of the stress tensor are unrestricted and the displacement is
zero. The dispersion relation for the free-surface boundary condition is then given
by
(q̃ 2 − q̃t2 )T1 (q̃l π ) − 2q̃l2 (q̃ 2 + q̃t2 ) + 4q̃ 2 q̃l2 T1 (q̃t π) = 0, (9.26)
and the dispersion relation for the clamped-surface boundary condition is given by
q̃l2 T1 (q̃t π) + q̃ 2 T1 (q̃l π ) = 0, (9.27)

where
T1 (x) = x J0 (x)/J1 (x). (9.28)
For the free-surface boundary condition β is given by
q̃t (q̃ 2 − q̃t2 ) J1 (q̃t π )

β=− , (9.29)
2q̃ 2 q̃l J1 (q̃l π )
and for the clamped-surface boundary condition β is given by
q̃t J1 (q̃t π )
β=− . (9.30)
q̃l J1 (q̃l π )
The electron–acoustic-phonon scattering rates may be evaluated straightfor-

wardly, in the limit where electrons are assumed to be confined in the cylinder by an
infinite potential at its surface. In this case (Wang and Lei, 1994)
eikz
|k = √ ψ(r, θ), (9.31)
Lz
where

1 X 10
ψ(r, θ ) = √ J0 r , (9.32)
π Y10 a a
X 10 being the position of the first zero of J0 (x) and Y10 being equal to J1 (X 10 ); L z ,
the length of the cylinder, is assumed to be much larger than a. The energy of the
ground state electron energy is [(h̄k)2 + (X 10 /a)2 ]/2m ∗ , where m ∗ is the electron’s
effective mass. Then it follows that the matrix element of the deformation potential,
Hdef , takes the form
6

Mn{e,a} = k , Nn + 12 + 12 Hdef |k, Nn

h̄ a
= E1 √ βn (q 2 + ql2 )
2N Mωn 2sn
#
× F(ql a) Nn + 12 + 12 δ−k +k−q , (9.33)
where Nn is the phonon occupation number of Section 5.1, = 1, −1 in the cases

of phonon emission and absorption respectively and F is given by
1 2

Fn (t) = d x x J0 (X 10 x) J0 (t x). (9.34)
0
Then, assuming the Fermi golden rule,

1 2π {e,a} 2
= Mn δ(E(k ) − E(k) + h̄ω)
τ {e,a} (k) h̄
π m ∗ |E 1 |2
= 0 4 2
n Y1 h̄ ρ
1 1 {e,a} {e,a}
× 4 2
I1cn q̃+ + I1cn q̃− , (9.35)
c0 c̃l a
with
|βn (q̃)|2 $ %
I1cn (q̃) = |ω̃n |3 |Fn (q̃l π )|2 Nn + 12 + 12
|sn (q̃)|
1
× . (9.36)
∗
(q̃ + k̃) − (m c0 a/π h̄)(d ω̃/d q̃)
{e,a}
From energy- and momentum-conservation conditions it follows that q̃± is the
solution of (h̄ 2 /2m ∗ )(q̃ 2 ± 2q̃ k̃) − h̄ω = 0; accordingly, the plus sign is taken for
forward scattering and the negative sign for backward scattering. As usual, n labels
the quantized acoustic phonon modes.
For the piezoelectric scattering of carriers in a cylindrical quantum wire, the
normalization procedure is of course the same as that for scattering by a deformation
potential but the interaction Hamiltonian is based on the interaction potential of
subsection 9.1.2. For cubic materials this piezoelectric interaction potential takes
the form
qλ eλ,µν qµ
V piezo (q) = 4π uν . (9.37)
0q · q
In cylindrical coordinates the piezoelectric tensor of subsection 9.1.2 takes the form
(Auld, 1973)
 
0 0 0 ex4 A(ϕ) 2ex4 B(ϕ) 0
e= 0 0 0 −2ex4 B(ϕ) ex4 A(ϕ) 0 
2ex4 B(ϕ) −2ex4 B(ϕ) 0 0 0 ex4 A(ϕ)
(9.38)
where A(ϕ) = cos2 ϕ − sin2 ϕ and B(ϕ) = cos ϕ sin ϕ. Moreover, the polarization
vector P and strain components Sν are given by
 
P1
P =  P2  (9.39)
P3
and
∂u r ur 1 ∂u ϕ
S1 = Srr = , S2 = Sϕϕ = + ,
∂r r r ∂ϕ
∂u z ∂u ϕ 1 ∂u z
S3 = Szz = , S4 = 2Szϕ = 2Sϕz = + ,
∂z ∂z r ∂ϕ
(9.40)
∂u r ∂u z
S5 = 2Sr z = 2Szr = + ,
∂z ∂r
1 ∂u r ∂u ϕ uϕ
S6 = 2Sr ϕ = 2Sϕr = + − .
r ∂ϕ ∂r r
Accordingly, the relation P = e S implies that (Stroscio and Kim, 1993)
P1 = 2ex4 (cos2 ϕ − sin2 ϕ)Szϕ + 4ex4 cos ϕ sin ϕ Sr z ,
P2 = −4ex4 cos ϕ sin ϕ Szϕ + 2ex4 (cos2 ϕ − sin2 ϕ)Sr z , (9.41)
P3 = 2ex4 cos ϕ sin ϕ (Srr − Sϕϕ ) + 2ex4 (cos2 ϕ − sin2 ϕ)Sr ϕ .

From classical acoustic theory, the lowest-order azimuthally symmetric torsional
mode of a cylindrical structure of radius a has the form (Auld, 1973)
u wire
ϕ = γ r e−iωt/vt , (9.42)
where vt = (c44 /ρ)1/2 and γ is a normalization factor to be determined. The

solutions for a free-standing cylindrical structure satisfy the boundary condition of
vanishing stress components on the surface of the cylinder. That is, the vibrational
modes are such that the cylinder undergoes distortions until there is no remaining
force on its surface to cause additional distortion. For such a case the displacement
field on the surface of the cylinder is unrestricted. For the case of a free-standing
structure, the displacement field of lowest-order azimuthally symmetric torsional
mode is characterized by vanishing u r and u z and q = qz = ω/vt . For a wire of
radius a and length L the classical solution for this mode is found by requiring the
zϕ-component of the stress to be zero at the ends of the cylinder located at z = 0
and z = L, so that the classical mode is given by
νπ
u finite
ϕ
wire
= γ r cos z, (9.43)
L
where νπ/L = ω/vt and γ is the normalization constant. The normalization
condition,
L 2π a
1 h̄
dz dϕ dr r u(q, r, ϕ, z) · u∗ (q, r, ϕ, z) = ,
πa L 0
2
0 0 2Mωq
(9.44)
yields

h̄ 2 νπ
u finite
ϕ
wire
= r cos z. (9.45)
2Mωq a L
Strictly speaking, this case of a quantum wire of finite length also corresponds to
a quantum dot since dimensional confinement is considered in all three dimensions.
For a quantum wire of infinite length the normalization introduced at the beginning
of this section requires the condition
1/2
1 h̄
γ = , (9.46)
a Mωq
so that

h̄ 1 −iωt/vt
u finite
ϕ
wire
= re . (9.47)
Mωq a
These normalized displacement fields determine the components of S and, therefore,

P; and the piezoelectric interaction potential V piezo (q) is thus determined for the
lowest-order azimuthally symmetric torsional modes of quantum wires and cylindri-
cal quantum wires of finite length. As illustrated for the deformation potential, the
carrier–acoustic-phonon scattering rate may be estimated by perturbation theory. In
particular, the Fermi golden rule of Appendix E provides a convenient formalism
for such calculations.
9.3.2 Rectangular wires

The calculation of the carrier–acoustic-phonon scattering rate in a rectangular quan-
tum wire proceeds along the lines of the analogous calculation of subsection 9.3.1
for a cylindrical quantum wire. The deformation-potential interaction for acoustic
phonons interacting with carriers in a rectangular quantum wire was derived in
subsection 7.6.3, (7.196), where it was shown that
α Eα ωγ
Hdef = √1 Aγ 2 cos q1 x cos hy [an,m (γ ) + an,m (−γ )] eiγ z ,
L n,m,γ cl q1
(9.48)
with
2h̄
A2γ = ,
Mωγ Bγ (9.49)
hd = (n + 12 )π, n = 0, 1, 2, . . . .
In the extreme quantum limit where the carriers are assumed to be confined within
the rectangular boundaries of the quantum wire by an infinitely high potential, the
carrier wavefunction is
1 πx π y ikz
ψk (x, y, z) = |k = √ cos cos e , (9.50)
ad 2a 2d
and the ground-state carrier eigenenergy is

2
h̄ 2 π π2
E= + + k 2
. (9.51)
2m (2a)2 (2d)2
To estimate the carrier–acoustic-phonon scattering rate, it is necessary to evaluate
the matrix element of the deformation-potential interaction. This matrix element is
given by
6 4

Mn{e,a} = k , Nn + 12 + 12 Hdef k, Nn
Eα ωγ2
= √1 A 2 k | cos q1 x cos hy eiγ z |k
L n,m,γ cl q1
6 4

× N + 12 + 12 [an,m (γ ) + an,m (−γ )]N
Eα ωγ2 1
= √1 A 2 δk−k +γ
L n,m,γ cl q1 ad
a d
πx πy
× d x cos2 cos q1 x dy cos2 cos (hy)
−a 2a −d 2d
6 4

× N + 12 + 12 [an,m (γ ) + an,m (−γ )]N
Eα ωγ2 π 2 sin q1 a 1
= √1 A 2 2a2) 2 2 3
L n,m,γ c q
l 1 1 q a(π 2 − q 1 π n + 2 1 − n + 12
1
6 4

× N + 12 + 12 [an,m (γ ) + an,m (−γ )]N
Eα ωγ2 π 2 sin (q1 a)

= √1 A 2
L n,m,γ cl q1 q1 a(π 2 − q12 a 2 )
1 1/2
× 2 3 n + 12 ± 12 δk−k ±γ . (9.52)
2
π n + 12 1 − n + 12
We have used

1 a πx π 2 sin q1 a
d x cos2 cos q1 x = (9.53)
a −a 2a q1 a(π 2 − q12 a 2 )
and

1 d πy π 2 sin hd 1
d
dy cos2
2d
cos hy =
hd(π 2 − h 2 d 2 )
= 2 2 3 .
−d π n + 2 1 − n + 12
1
(9.54)
As usual the plus signs in the phonon-occupancy and momentum-conservation terms
correspond to phonon absorption and the negative signs to phonon emission. Then
the scattering rate predicted by the Fermi golden rule is given by

1 L +∞2π {e,a} 2
= dγMn δ(E(k ) − E(k) ± h̄ωγ )
τ 2π −∞ h̄

2 / 02
L +∞ 2π E 1α ωγ2 π 2 sin q1 a
= dγ √ Aγ 2
n,m 2π −∞ h̄ L cl q1 q1 a(π 2 − q12 a 2 )
 2
 1 
× 2 3 n + 12 ± 12
π n + 1 1 − n + 1 2 
2 2

h̄ 2 2
×δ (γ ∓ 2kγ ) ± h̄ωγ , (9.55)
2m
where the sum over γ has been converted to an integral, n is the Bose–Einstein
occupation number for the acoustic phonons, and L is the normalization length.
Yu et al. (1994) evaluated this deformation-potential scattering rate numerically for
GaAs quantum wires at 77 K for three different cross sections, 28.3 Å × 56.6 Å,
100 Å × 200 Å, and 50 Å × 200 Å. The results of these numerical calculations
Figure 9.1. Deformation-potential emission and absorption scattering rates for bulk
and confined acoustic phonons in a 28.3 Å × 56.6 Å GaAs quantum wire at 77 K.
Confined modes: –•—•—•–, emission; – – – – – –, absorption. Bulk modes:
–◦—◦—◦–, emission; – – – – – –, absorption. Energy thresholds for the width
modes are at 0.03, 2.36, 2.55, 4.90, 7.30, and 7.40 meV. Energy thresholds for
thickness modes are at 2.06, 4.44, 5.90, 9.87, 14.5, and 15.1 meV. One-dimensional
density-of-states peaks are evident in the emission rate for the quantum wire. From
Yu et al. (1994), American Physical Society, with permission.
are shown in Figures 9.1, 9.2, and 9.3 respectively. In each of these cases the
acoustic phonon emission and absorption rates are shown for the GaAs quantum
wire and for bulk GaAs. The distinct one-dimensional density-of-states peaks are
evident for the case of phonon emission in each of the quantum-wire structures.
These peaks occur at the threshold electron energy for the emission of one of the
various width or thickness modes. The peak magnitudes of the scattering rates
in these quantum wires suggest that carrier–acoustic-phonon scattering via the
deformation potential is significant in quantum-wire elements of nanoscale devices.
The calculations of Yu et al. (1994) are based on the assumption that the wires
have perfectly uniform cross sections. In envisioned practical situations it is unlikely
that the dimensional tolerances associated with the cross-sectional areas will be
small enough to realize subatomic dimensional control. It is thus unlikely that the
measured one-dimensional density-of-states peaks will be as pronounced as they are
in Figures 9.1, 9.2, and 9.3.
and confined acoustic phonons in a 100 Å × 200 Å GaAs quantum wire at 77 K.
–◦—◦—◦–, emission; – – – – – –, absorption. Energy thresholds for the confined
modes are at 0.03, 0.65, 0.75, 1.39, 2.06, and 2.12 meV. Energy thresholds for the
bulk modes are at 0.59, 1.26, 1.68, 2.80, 4.11, and 4.28 meV. One-dimensional
and confined acoustic phonons in a 50 Å × 200 Å GaAs quantum wire at 77 K.
–◦—◦—◦–, emission; – – – – – –, absorption. Energy thresholds for the confined
modes are at 0.03, 0.65, 0.75, 1.39, 2.06, and 2.12 meV. Energy thresholds for the
bulk modes are at 0.61, 2.58, 3.05, 5.56, 8.24, and 8.43 meV. One-dimensional
Chapter 10
Recent developments
Nothing is too wonderful to be true, if it is consistent with the
laws of nature, and in such things as these, experiment is the
best test of such consistency.
Michael Faraday, 1849
10.1 Phonon effects in intersubband lasers
The effects of dimensional confinement on optical phonons, phonon-assisted elec-

tron intersubband transition rates (Teng et al., 1998), and gain (Kisin et al.,
1997) have been evaluated in a series of studies on semiconductor lasers.
Many of the novel semiconductor lasers – such as the tunneling injection laser
and the quantum cascade laser – contain quantum wells with confinement di-
mensions of about 50 Å or less. An example of such a semiconductor laser
structure is given in Figure 10.1. In this laser, the conduction band is engi-
neered so that the upper and lower energy levels are E 3 and E 2 respectively,
with a third level, E 1 , such that phonon-assisted tunneling from level E 2 to
level E 1 is promoted. The IF optical phonons are of special importance in
such heterostructures. Two properties of the IF optical phonons account for
their special significance (Stroscio, 1996) in narrow-well semiconductor lasers:
(a) in narrow wells the interface phonons have appreciable interaction poten-
tials throughout the quantum well since, as was demonstrated in Chapter 7,
the interface optical phonons have potentials near the heterointerfaces of the
form ce−q|z| e−iq·ρρ , where, for example, q has values of very roughly 0.02
Å−1 for a typical intrasubband transition in GaAs; and (b) the energies asso-
ciated with phonon-assisted processes in these heterostructures can be substan-
tially different from those in semiconductor lasers without significant dimen-
sional confinement, since the interface phonons may have frequencies ωq which
are significantly different from those of the other phonons in the quantum
well.
186
As an example, optical phonon (LO) transitions play an especially important

role in novel intersubband lasers operating at infrared wavelengths, since selected
intersubband transition rates are critical to establishing and maintaining a popu-
lation inversion. In many specific cases, such lasers (Faist et al., 1996a, b; Zhang
et al., 1996) can be designed only if quantum-well regions have thicknesses as
small as 25–50 Å. In such polar-semiconductor structures, the dominant electron
energy relaxation processes are due to the electron–LO-phonon interaction. In this
subsection these processes will be considered and their role in the operation of novel
semiconductor lasers will be examined.
Teng et al. (1998) determined the optical phonon (LO) modes and the electron–
LO-phonon intersubband transition rates due to the electron–LO-phonon interac-
tions in two generic heterostructures similar to the regions where phonon-induced
transitions occur in a variety of dimensionally confined semiconductor lasers. These
structures are depicted in Figure 10.2, where materials 1, 2, and 3 are GaAs,
Al0.25 Ga0.75 As and Al0.4 Ga0.6 As respectively.
The heterostructure of Figure 10.2(a) has three interfaces and an asymmetrical
potential profile. Figure 10.2(b) is a four-interface structure with a symmetrical
potential. The dielectric constants of regions 1, 2, and 3 are taken as 1 (ω), 2 (ω),
and 3 (ω) respectively. The dielectric continuum model of optical phonons, Sections
7.1 and 7.3 and Appendix C, leads immediately to the following LO phonon
Figure 10.1. Step quantum-well laser with narrow quantum wells.

188 10 Recent developments
potentials: in polar material 1, where i = 2, the confined LO modes are given by

 mπ

 m = 1, 3, 5, . . .
cos a z,
φ(z) ∝ (10.1)


sin mπ z, m = 2, 4, 6, . . .
a
|z| < .
a 2
The confined LO modes in polar material 2, where i = 1, are given by

 2mπ b+a

 cos z+ , m = 1, 3, 5, . . .
 b−a 4
φ(z) ∝

 b+a

sin
2mπ
z+ , m = 2, 4, 6, . . .
b−a 4
b a
− <z<− . (10.2)
2 2
Figure 10.2. Two generic heterostructures common to dimensionally confined

semiconductor lasers. Materials 1, 2, and 3 are GaAs, Al0.25 Ga0.75 As and
Al0.4 Ga0.6 As respectively. Structure (a) has an asymmetrical potential profile and
structure (b) a symmetrical potential profile. From Teng et al. (1998), American
For the half-space LO modes in polar material 3, where i = 0, 3,




2mπ b b

sin L z+ , m = 1, 2, 3, . . . z≤−
2 2
φ(z) ∝ (10.3)

 $ %

sin 2mπ a a
z− , m = 1, 2, 3, . . . z≥ .
L 2 2
Moreover, the interface LO phonon modes are as given in subsection 7.3.4 and have
the form

φi (q, z) ∝ ci− e−qz + ci+

e+qz , (10.4)
where the values of
ci± are determined by the boundary conditions:

φ1 (q, z = −L/2) = 0 )⇒ c0− = 0,

φ3 (q, z = L/2) = 0 )⇒ c3+ = 0,
(10.5)
φi (q, z i ) = φi−1 (q, z i ),
∂φi (q, z i ) ∂φi−1 (q, z i )
i = i−1 ,
∂z ∂z
where L is taken to be large compared with other dimension scales in the system, and
z i designates the z-coordinate at the location of the heterointerface between regions i
and i − 1. The conditions on the continuity of the potential and the continuity of the
normal component of the electric displacement may be used to relate recursively
the non-zero coefficients ci± to the dielectric constants of the various materials
composing the heterostructure. The boundary condition at z = L/2 requires that

c3+ = 0 and the dispersion relation for the interface modes for the heterostructure
of Figure 10.2(a) is
1 2
0= (3 − 1 )(1 + 2 )(2 − 3 ) e−q(a+b)
1 2 3
+ (3 − 1 )(1 − 2 )(2 + 3 ) e−2qa
+ (3 + 1 )(1 − 2 )(2 − 3 ) e−q(b−a)
+ (3 + 1 )(1 + 2 )(2 + 3 )], (10.6)
where the generalized Lyddane–Sachs–Teller relationship is used to write n (ω) in
terms of n (∞), equations (7.5), (7.6); for binary materials,
ω2 − ωLO,n
2
n (ω) = n (∞) , (10.7)
ω2 − ωTO,n
2
and for ternary materials of the form A y B1−y C

2 2
ω − ωLO,n,a
2 ω − ωLO,n,b
2
n (ω) = n (∞) , (10.8)
ω2 − ωTO,n,a
2 ω2 − ωTO,n,b
2
where, as discussed in Section 7.1, the subscript a denotes frequencies associated

with the dipole pairs AC and the subscript b denotes frequencies associated with
the dipole pairs BC. The condition that c3+ (ω) = 0 is met for 10 frequencies
and, accordingly, there are 10 interface modes: six of these correspond to GaAs-like
modes with energies in the range of 32 to 37 meV and four are AlAs-like modes with
energies close to 46 meV. The dispersion relations for these 10 modes are shown in
Figure 10.3.
As in Teng et al. (1998), the dielectric constant is taken as
(∞) = 10.89 − 2.73x, (10.9)
where x represents the content of the Al in Alx Ga1−x As. The phonon energies used
in determining these dispersion relations are given in Table 10.1.
For the structure of Figure 10.2(b), the boundary condition determines that

c4+ /c0+ = 0 and the dispersion relations for the interface modes – as determined
by the previously described iteration procedure – are given by
Figure 10.3. Dispersion relations for the interface LO phonon modes of the
structure depicted in Figure 10.2(a). From Teng et al. (1998), American Institute of
1 2
0= −(3 − 2 )2 (1 + 2 )2 e2qb − 2(32 − 22 )(22 − 12 ) eq(a+b)
1 22 3
+ (32 − 22 )2 (2 − 1 )2 e2q(b−a) + 2(32 − 22 )(22 − 12 ) eq(b−a)
3
− (3 + 2 )2 (2 − 1 )2 e2qa +(3 + 2 )2 (1 + 2 )2 .
(10.10)
The 14 solutions of this equation correspond to six GaAs-like interface modes and
eight AlAs-like interface modes. The potentials for the six GaAs-like modes are
plotted in Figure 10.4 for qa = 0.5.
These LO phonon modes are either symmetric or antisymmetric, as is to be
expected from the symmetry of the heterostructure. They must still be normalized.
The normalization condition of subsection 7.3.4, equation (7.116),
2
1 1 ∂i (ω) ∂ (q, z)
dz q 2 |i (q, z)|2 +
i = h̄ ,
4π 2ω ∂ω ∂z 2ωL 2
i Ri
(10.11)
provides the necessary condition to determine the normalization constant for each
mode. Applying this condition leads to the following normalized LO phonon
potentials: in polar material 1, where i = 2, the confined LO modes are given by
1/2 1/2
4π h̄ 1 1
φ(z) =
L 2 ∂i (ω)/∂ω q 2 + (mπ /a)2
 mπ
1/2  
 cos z, m = 1, 3, 5, . . .
2 a a
× |z| < . (10.12)
a 
 mπ 2
 sin z, m = 2, 4, 6, . . .
a
In polar material 2 where i = 1, the confined LO modes are given by
Table 10.1. Values of phonon energies adopted for these calculations.
Phonon energy GaAs AlAs Alx Ga1−x As

(meV)
h̄ωLO,a 36.25 36.25 − 6.55x + 1.79x 2
h̄ωLO,b 50.09 44.63 + 8.78x − 3.32x 2
h̄ωTO,a 33.29 33.29 − 0.64x − 1.16x 2
h̄ωTO,b 44.88 44.63 + 0.55x − 0.30x 2
1/2 ! "1/2
4π h̄ 1 1
φ(z) =
L 2 ∂i (ω)/∂ω q 2 + [2mπ/(b − a)]2


 2mπ b+a
1/2  cos z + m = 1, 3, 5, . . .
4 b−a 4
×
b−a 
 2mπ b+a

sin z+ m = 2, 4, 6, . . .
b−a 4
b a
− <z<− . (10.13)
2 2
For the half-space LO modes in polar material 3, where i = 0, 3,
1/2 1/2
4π h̄ 1 1
φ(z) =
L 2 ∂i (ω)/∂ω q 2 + (2mπ /L)2

 2mπ b b
1/2   sin z+ m = 1, 2, 3, . . . z≤−
4 L 2 2
×
L 
 $ %
sin 2mπ z − a m = 1, 2, 3, . . . z≥ .
a
L 2 2
(10.14)
Figure 10.4. Potential profiles for the six GaAs-like interface LO phonon modes for
the structure of Figure 10.2(b). The mode frequencies are (1) 35.86 meV, (2) 34.91
meV, (3) 34.26 meV, (4) 34.01 meV, (5) 33.17 meV, and (6) 33.07 meV. From Teng
As in subsection 7.3.4, the interface LO phonon mode potentials are of the form,
equations (7.117), (7.118),

i (q, z) = A(ci− e−qz + ci+

e+qz ) = Ai (q, z), (10.15)
where the normalization constant A is given by

1/2
h̄ 1 1 ∂i (ω)
A= dz
2ωL 2 i
4π 2ω ∂ω Ri

−1/2
∂i (q, z) 2
× q |i (q, z)| +
2 2
. (10.16)
∂z
Finally, from subsection 7.3.4, equation (7.119),

HIF = ei (r) = e e−iq·ρρ i (q, z)(a−q
†
+ aq )
q

ρ
−iq·ρ †
=e e Ai (q, z)(a−q + aq ). (10.17)
q
For a particular phonon mode j the Fröhlich interaction Hamiltonian is given by

Hj = eφ j (r) = e e−iq·ρρ φ j (r)(a−q
†
+ aq ) (10.18)
q
and, from Section 8.1,
{e,a} 2π {e,a} 2

Sn,n (k, k ) = Mn,n (q) δ(E(k ) + E n − E(k) − E n ± h̄ω),
h̄ q
(10.19)
with
6 4
{e,a}
Mn,n (q) = n , k , Nq + 1
2 ± 12 Hj n, k, Nq + 1
2 ± 1
2 , (10.20)
where the initial and final energies each have been written as a sum of the in-plane
and subband energies, E(k) + E n and E(k ) + E n respectively; E(k) = h̄ 2 k 2 /2m.
Since the in-plane energy associated with the two-dimensional wavevector k has
been separated from the subband energy, the subscript 2D of Section 8.1 has been
omitted. Phonon-assisted transitions between the initial and final carrier states are
depicted in Figure 10.5.
Teng et al. (1998) solved the Schrödinger equation numerically to determine
the electron wavefunctions, for selected phonon modes, needed to evaluate the
matrix elements for phonon-assisted transitions. These have focused on the emission
process to gain insights into phonon-assisted processes of importance in narrow-well
semiconductor lasers. As expected (Stroscio, 1996), their results show that the
half-space modes and confined LO modes yield phonon-assisted rates that are small
relative to those associated with the interface LO phonon modes. Moreover, it is

found that certain interface modes make contributions about an order of magnitude
larger than most of the remaining interface LO phonon modes. Figure 10.6 shows the
transition rate for the dominant phonon-assisted transition in the symmetric structure
of Figure 10.2(b). This maximum transition is associated with the emission of the
antisymmetric interface longitudinal optical phonon mode having an energy close to
Figure 10.5. Phonon-

assisted transitions between
states |n, k and |n , k . (a)
An emission event, between
the state |n, k (upper plane)
and the state |n, k = k − q
(lower plane). (b) An
absorption event between
|n, k (lower plane) and
|n, k = k + q (upper
plane). The potential
function is parabolic in the
wavevector. From Teng et al.
(1998), American Institute of
50 meV. As is apparent from Figure 10.6, the rate is substantially larger than the rate
associated with the emission of a bulk phonon. Clearly, in the design of narrow-well
semiconductor lasers it is essential that the interface modes be considered. The usual
approximation of treating the phonons as bulk phonons is inadequate.
10.2 Effect of confined phonons on gain of intersubband

lasers
In the previous section, it was demonstrated that it is essential to take into account
the spectrum of dimensionally confined phonon modes in nanostructure-based
intersubband semiconductor lasers. In the present section, it will be shown that the
gain of an intersubband narrow-well semiconductor laser can be modeled accurately
only if the realistic heterostructure phonon spectrum is taken into account (Kisin
et al., 1997). The system considered in calculating the gain of an intersubband
quantum-well laser is an Alx Ga1−x As/GaAs/Alx Ga1−x As double-heterostructure
of width a. The electron wavefunctions and dispersion relations for this system
Figure 10.6. For the symmetric structure of Figure 10.2(b), the maximum transition
rates for the antisymmetric interface mode and the bulk mode are depicted as
functions of the difference E 2 − E 1 between the final and initial energies.
are modeled (Kisin et al., 1997) with a four-band Kane model. The electronic
wavefunctions (Gorfinkel et al., 1996) for the states in subband n are taken to be
(n) 1
k (ρρ , z) = √ eik·ρρ ψk(n) (z), (10.21)
Sa
where, as defined previously, ρ = (x, y) and S is the area of the heterostructure. For
intersubband lasers – such as the quantum-cascade laser – involving only transitions
(n)
in the conduction band, the four-component Kane wavefunction envelopes, ψk ,
may be described simply in terms of scalar wavefunctions. For the lowest subband,
n = 1,

√ cos k z1 z |z| < a/2,
ψ1 (z) = 2C1 (10.22)
−λ
cos(k z1 a/2) e 1 (|z|−a/2) |z| > a/2,
with
k z1 tan k z1 a/2 = λ1 E w1 /E b1 . (10.23)
For the second subband, n = 2,

√ sin k z2 z |z| < a/2,
ψ2 (z) = 2C2 (10.24)
sign z sin k z2 a/2 e−λ2 (|z|−a/2) |z| > a/2,
with
k z2 tan k z2 a/2 = −λ2 E w2 /E b2 . (10.25)
Here E wn and E bn are the energies for the nth subband of the well and the barrier
measured with respect to the valence band in each material. For these approximate
wavefunctions, it is clear that phonon-assisted transitions from the second to the low-
est subband will involve only the antisymmetric interface phonons, the odd-parity
confined phonon modes, and barrier modes. However, intrasubband phonon-assisted
transitions will include contributions from the symmetric interface phonons, the
even-parity confined phonon modes, and barrier modes. As will become apparent,
such phonon-assisted processes have line-broadening effects that exert a major
influence on the gain of the laser. Moreover, it will become apparent that the
energy dependence of the line broadening resulting from the energy spectrum of
the phonons also plays a significant role in determining the properties of the gain of
the laser.
Subband energy dispersion curves for 60-ångstrom-wide and 100-ångstrom-wide
quantum wells are depicted in Figure 10.7 along with examples of intersubband and
intrasubband phonon-assisted processes; in Figure 10.7, x = 0.3 and the conduction
band and valence band off-sets are c = 300 meV and v = 150 meV respectively.
The phonon spectrum used in the calculation of the intersubband optical gain
in the quantum-well system under consideration includes the symmetric interface
modes, the antisymmetric interface modes, the two lowest-order confined modes,
and the half-space modes, which are referred to as the barrier modes in this
section. The Fröhlich interaction Hamiltonians used here are those of the dielectric
continuum model with electrostatic boundary conditions, as discussed in Chapter 7
and in Appendix C.
These modes may be obtained straightforwardly from the modes for the double-
heterojunction, uniaxial structure of Appendix D by taking the limit where both
(ω)⊥ and (ω) are equal to (ω). Figures 10.8(a), (b) and 10.9(a), (b) depict
the contributing scattering rates (Kisin et al., 1997) for the scalar wavefunctions
presented previously and for the phonon modes of the dielectric continuum model
with electrostatic boundary conditions.
In calculating these results, the parameters have been selected as follows: x =
0.4, c = 300 meV, v = 200 meV, and a = 60 Å. The energy spectrum of Figure
10.8(a) illustrates clearly that there are several distinct thresholds.
Figure 10.7. Energy dispersion curves for 60-ångstrom-wide (solid lines) and
100-ångstrom-wide (broken lines) Al0.3 Ga0.7 As/GaAs/Al0.3 Ga0.7 As quantum
wells. Typical intersubband and intrasubband transitions are shown for both quantum
wells. The energy gap for the GaAs well, E g (GaAs), is taken as 1.4 eV and the ratio
of the effective mass to the electron mass is taken as 0.067 for GaAs. From Kisin
Figure 10.8. (a) Intrasubband scattering rates in the second subband as a function of
the electron energy in the second subband. The high- and low-frequency symmetric
interface LO phonon modes, the confined modes, and the barrier modes are labeled
by is +, is −, c and b respectively. (b) Intersubband scattering rates from the second
to the lowest subband as a function of the electron energy in the second subband.
The high- and low-frequency antisymmetric interface LO phonon modes, the
confined modes, and the barrier modes are labeled by ia+, ia−, c and b respectively.
From Kisin et al. (1997), American Institute of Physics, with permission.
Figure 10.9. (a) Intrasubband scattering rates in the lowest subband as a function of
well thickness. The initial electron energy in the first subband is 60 meV. (b)
Intersubband scattering rates for transitions from the second to the first subband as a
function of well thickness. The initial electron energy in the second subband is
10 meV. From Kisin et al. (1997), American Institute of Physics, with permission.
The optical gain spectrum g() for intersubband transitions between the second
and the lowest subband (Gelmont et al., 1996; Gorfinkel et al., 1996) is given by

4πn 2 e2 |z 12 |2 ∞ −ε/k B Te f1
g() = √ dε τε () e 1− , (10.26)
h̄ac (∞)k B Te 0 f2
where the line-shape function is defined by
W (ε)
τε () = , (10.27)
[ − ε ]2 + [W (ε)]2
the transition energy is related to the energy difference h̄0 between the subbands
at k = 0 by
h̄ε = E 2 (k) − E 1 (k) = h̄0 + ε2 (k) − ε1 (k), (10.28)
and the ratio of distribution functions, f 1 / f 2 , is given by

f1 n1 m 1 ε m1
= exp 1− , (10.29)
f2 n2 m 2 k B Te m2
for the simplified case of parabolic bands with effective masses m 1 and m 2 ; n 1 and
n 2 are the areal electron concentrations of the two subbands. The broadening term
in the denominator of τε (), [W (ε)]2 , is related to the phonon-assisted scattering
rate, 1/τ (ε), as well as to other scattering rates such as electron–plasmon scattering.
The rate associated with phonon-assisted scattering processes generally dominates
over the rates for other scattering processes (Kisin et al., 1997) and is the only rate
considered herein. Room-temperature gain spectra for a second subband electron
concentration of 1011 cm−2 are shown in Figures 10.10 and 10.11.
It is clear from these gain curves that the effects of dimensional confinement
of the phonons play a role in determining the optical gain of a narrow-well
intersubband laser. This is especially evident in Figure 10.10(b), where gain curves
based on the full spectrum of dimensionally confined phonon modes and on bulk
phonons only are dramatically different and, in fact, have opposite signs over an
appreciable range of photon energies. To understand this result, it is useful to
consider the steady state condition expressing the conservation of particle flow:
n2 n1
= W12 n 2 = . (10.30)
τ12 τout
This shows that the ratio of the subband populations in equilibrium depends directly
on the ratio of the intersubband transition rate and the escape rate from the lower
subband. Indeed, if the escape rate is decreased from the value consistent with this
equilibrium condition then the population inversion is not expected to be maintained
and the gain will become negative. For the parameters used in calculating the results
of Figure 10.10(b), τ12 = 0.56 ps and it is now clear why the gain curves for τout =
0.55 ps are so sensitive to the energy spectrum of the phonon modes.
Figure 10.10. Optical gain in cm−1 for different rates of electron escape, τout , from
the lowest subband: (a) τout is 0.4 ps and 0.6 ps; (b) τout is 0.55 ps. The broken lines
are calculated assuming that only bulk GaAs LO phonons contribute; the
broken-and-dotted lines assuming that only bulk GaAlAs phonons contribute. The
solid lines are calculated from the full spectrum of dimensionally confined phonon
modes. From Kisin et al. (1997), American Institute of Physics, with permission.
10.3 Phonon contribution to valley current in

double-barrier structures
Turley and Teitworth (1991a, b, 1992) and Turley et al. (1993) performed consis-
tent experimental and theoretical investigations of the effect of phonon-assisted
tunneling in GaAs/AlAs double-barrier structures. The experiments of Turley
and Teitsworth provided observations of tunneling currents and magnetotunneling
spectra indicating that a major part of the valley current in certain GaAs/AlAs
double-barrier quantum wells is due to phonon-assisted tunneling. These experi-
ments also present clear evidence that the symmetric interface phonon mode makes
a dominant contribution to the phonon-assisted tunneling current in these devices.
The dominance of the interface mode is consistent with early calculations suggesting
that the symmetric interface modes would play an important role in carrier transport
in short-period GaAs/AlAs superlattices (Stroscio et al., 1991a) and it is also
consistent with the results of the last two sections on narrow-well intersubband semi-
conductor lasers. The results of Turley and Teitsworth provided early systematic
studies demonstrating that the technologically important valley current in certain
Figure 10.11. Optical gain in cm−1 for two different sets of phonon modes: broken
lines, all phonon modes are included; solid lines, all modes but the barrier mode are
included. Results are given for three different values of the electron escape rate, τout :
(a) 0.4 ps; (b) 0.5 ps; (c) 0.6 ps. From Kisin et al. (1997), American Institute of
10.3 Phonon contribution to valley current in double-barrier structures 203
double-barrier resonant tunneling devices (DBRTDs) is, in fact, due partially to

phonon-assisted tunneling. This is an important conclusion since DBRTDs have
been the subject of intense study for almost two decades because of their perceived
potential for playing a major role in new generations of high-performance electronic
devices. Moreover, the careful and systematic results of Turley and Teitsworth
demonstrate convincingly the utility of the dielectric continuum model of confined
phonons.
Current–voltage measurements at 4.2 K by Turley et al. (1993) are shown in
Figure 10.12. The double-barrier structure in this case was an 80-ångstrom-wide
GaAs quantum well with a 33-ångstrom-wide AlAs emitter barrier and a 45-
ångstrom-wide AlAs collector barrier. All these layers are nominally undoped. In
Figure 10.12 the principal resonant tunneling peak is at about 0.54 volts and a
distinct phonon-assisted tunneling peak is seen at about 0.81 volts.
The magnetotransport experiments of Turley et al. provided even more revealing
data. Figures 10.13(a), (b) depict measured magnetic field versus applied voltage
diagrams. The two ‘Landau fans’ converge at 0.67 volts and at 0.81 volts. Moreover,
Figure 10.14 features two straight vertical lines to which other lines converge at
−0.45 volts and −0.52 volts. By calibrating their experimental voltage scale, Turley
et al. (1993) related the electronic energy scale E z and the applied voltage V
through the relation V = αE z , where α = 5.49 ± 0.15 V/eV, and concluded
that the Landau fans converging at 0.67 volts in Figures 10.13(a), (b) correspond to
confined LO phonon modes with energies in the range 36.0 ± 0.9 meV. Likewise,
the Landau fans converging at 0.81 volts in Figures 10.13(a), (b) as well as that
Figure 10.12. Measured current–voltage curve at 4.2 K for a double-barrier

structure with an 80-ångstrom-wide GaAs quantum well having a 33-ångstrom-wide
AlAs emitter barrier and a 45-ångstrom-wide AlAs collector barrier. The principal
tunneling peak is at about 0.54 volts and a strong phonon-assisted tunneling peak
occurs at 0.81 volts. From Turley et al. (1993), American Physical Society, with
permission.
converging at −0.45 volts in Figure 10.14 correspond to symmetric interface modes

with energies in the range 50.3 ± 1.3 meV. In both cases the expected phonon
energies fall well within the error bars of Turley et al. (1993). The remaining Landau
fan, converging at −0.52 in Figure 10.14, corresponds to emission of phonons with
a changed Landau-level index.
Phonon emission processes have been observed in the measured currents in other
double-barrier-based heterostructures (Choi et al., 1990; Choi et al., 1996) albeit at
scales where dimensional confinement is not expected to lead to modifications in the
phonon modes. In particular, Choi et al. (1990) determined the current transfer ratio
and its derivative r ,
α = IC /I E , (10.31)
r = dα/d VE , (10.32)
at 4.2 K. Here IC is the collector current, I E is the emitter current, and VE is the
bias voltage for the heterostructure shown in Figure 10.15.
The structure of Figure 10.15 consists of a 1000-ångstrom-wide GaAs emit-
ter with n-type doping of 2 × 1017 cm−3 , an undoped 120-ångstrom-wide
Al0.35 Ga0.65 As injection barrier, a 1520-ångstrom-wide GaAs base doped at 2 ×
1017 cm−3 , an undoped 165-ångstrom-wide Al0.25 Ga0.75 As analyzer, and a 1000-
ångstrom-wide gallium arsenide collector doped at 2 × 1017 cm−3 . Figure 10.16
presents the experimental and theoretical values of the transfer ratio and its deriva-
tive as functions of VE .
The dominant peaks are at V1 and V2 and the peaks associated with phonon
emission are the subsidiary peaks at V P1 and V P2 . These subsidiary peaks are
separated by the LO phonon energy in GaAs – h̄ω0 = 36 meV – as expected.
In separate experiments, Choi et al. (1996) determined the hot-electron distribution
ρ(E), in a regime where both phonon emission and plasmon emission are important.
The heterostructure used in these experiments consisted of a GaAs emitter layer, an
AlGaAs injection barrier, a 1500-ångstrom-wide GaAs base doped at 2×1017 cm−3 ,
a double-barrier filter with a bandpass at 280 meV, and a GaAs collector. This
structure and the theoretical and experimental electron energy distributions ρ(E)
are shown in Figure 10.17. The initial injected electron distribution is taken to be
Gaussian and to have a width of 22 meV. The injection energy is taken to be e |Ve |.
The experimental photocurrent transfer ratio α p , as a function of emitter voltage
Ve , for the heterostructure shown in Figure 10.17 is presented in Figure 10.18 as a
solid line. The theoretical contributions for plasmon emission, phonon emission, and
both phonon and plasmon emission are designated by triangles, crosses, and circles
respectively. The difference between the theoretical and experimental contributions
at high Ve is due to the non-linear relation between the injection energy and Ve .
Based on these experimental and theoretical results, Choi et al. (1996) concluded
that phonon emission dominates when the base doping density is less than 3.3 ×
1017 cm−3 . For higher doping densities, plasmon emission becomes more important.
10.4 Population inversion in asymmetric double-barrier quantum-well lasers 205
10.4 Phonon-enhanced population inversion in

asymmetric double-barrier quantum-well lasers
Phonon-assisted interwell electronic transitions have been explored as a means of
rapidly depopulating the lower laser level in intersubband semiconductor lasers
Figure 10.13. Magnetic field versus applied voltage: (a) full-field range; (b)
low-field range. The heterostructure is a double-barrier structure with an
80-ångstrom-wide GaAs quantum well having a 33-ångstrom-wide AlAs emitter
barrier and a 45-ångstrom-wide AlAs collector barrier. From Turley et al. (1993),
American Physical Society, with permission.
(Faist et al., 1996a, b; Gmachl et al., 1998; Stroscio, 1996; Stroscio et al., 1999).
This fast depopulation promotes the lasing in intersubband lasers. Phonon-assisted
depopulation rates can be enhanced in asymmetric double-quantum-well het-
erostructures if the subband being depopulated is degenerate with a second subband
in the wider quantum well, as shown in Figure 10.19 (Stroscio et al., 1999).
The double-well active region of the AlAs/GaAs heterostructure depicted in this
figure is designed so that the lower carrier state A1 of well A is depopulated
by phonon-assisted transitions to state B1 of well B. As will be demonstrated,
these phonon-assisted transitions are enhanced greatly when levels A1 and B1 are
degenerate.
The interwell transition rate between states A1 and B1 is enhanced when A1
and B2 are nearly degenerate, as is illustrated by Figures 10.20 and 10.21. This
effect is associated primarily with an enhancement in the overlap between the
final and initial state wavefunctions involved in the phonon-assisted transition. In
Figure 10.20 the interwell transition rate is shown as a function of a1 for three
different models of the optical phonon modes participating in the phonon-assisted
transitions. The broken curve (a) is calculated with the Fermi golden rule, assuming
that only the bulk GaAs phonon with an energy of 36 meV participates in the
phonon-assisted transitions between states A1 and B1. Likewise, the broken curve
(b) is calculated with the Fermi golden rule, assuming that only the bulk AlAs
Figure 10.14. Magnetic field versus applied voltage: reverse bias diagram. The
heterostructure is a double-barrier structure with an 80-ångstrom-wide GaAs
quantum well having a 33-ångstrom-wide AlAs emitter barrier and a
45-ångstrom-wide AlAs collector barrier. From Turley et al. (1993), American
10.4 Population inversion in asymmetric double-barrier quantum-well lasers 207
phonon with an energy of 51 meV participates in the phonon-assisted transitions

between states A1 and B1. The remaining curve (solid line) is calculated with the
Fermi golden rule, using the full spectrum of optical phonon modes as discussed in
Chapter 7 and in Sections 10.1 and 10.2. The features on the solid curve between
the thresholds for curves (a) and (b) are due to interface modes in the nominal
reststrahlen band.
The peak values of the interwell phonon-assisted transition rates for conditions
of near degeneracy of energy levels A1 and B2 are shown in Figure 10.21 for a
series of values of the width a3 of the wider well. For a3 = 18 nm it is clear that the
interwell transition rate is enhanced by about an order of magnitude as a result of the
increased overlap of the final and initial electronic wavefunctions near conditions of
degeneracy between levels A1 and B2.
In summary, the phonon-assisted transition rate τout for the intersubband laser
structure of Figure 10.19 is enhanced significantly when the heterostructure is
engineered so that there is a near degeneracy between states A1 and B2. In this
case, the combination of band structure engineering and phonon engineering leads
Figure 10.15. Band structure of the device studied by Choi et al. (1990) under a
bias V E . The energy distribution of the current, the so-called hot-electron
distribution ρ(E), is shown in the barrier region B. The current emerging from the
emitter E is I E and the current at the collector C is indicated by IC . The states E n
(n = 1, 2, 3, 4) are resonant states created in the barrier. E F is the Fermi energy and
E P is the energy of the emitted phonon. U and L are the effective ‘upper’ and
‘lower’ contributions to the half-width of the energy distribution after phonon
emission. From Choi et al. (1990), American Physical Society, with permission.
to a significant favorable enhancement in the expected population inversion for the

double-well intersubband laser.
10.5 Confined-phonon effects in thin film

superconductors
It has been recognized for a number of years that the dimensional confinement of
carriers in thin film supeconductors leads to modification of the superconducting
energy gap and the critical temperature Tc (Blatt and Thompson, 1963; Thompson
and Blatt, 1963; Yu et al., 1976). In particular, it has been understood that size-
Figure 10.16. Transfer ratio α (smooth curve) and its derivative (peaked curve) as
functions of V E : (a) experimental values and (b) theoretical values. The subsidiary
peaks at V P1 and V P2 are due to phonon emission. The arrows show the current
directions. From Choi et al. (1990), American Physical Society, with permission.
quantization effects are manifest, as a function of the film thickness d, each time
one of the two-dimensional subband energy levels E n (d) passes through the Fermi
level as the film thickness is varied. Moreover, the Bardeen–Cooper–Schrieffer
(BCS) theory of superconductivity is based on the formation of Cooper pairs whose
binding energy is determined by the phonon-mediated pairing of electrons. Thus,
it is reasonable to expect that the effects of phonon confinement play a role in
determining the characteristics of thin film superconductors. Hwang et al. (2000)
have recently investigated this possibility within the context of the BCS theory for
thin films, by replacing the bulk phonon by the spectrum of confined phonons for the
thin film. These results of Hwang et al. are expected to reveal selective qualitative
features associated with phonon confinement in thin film superconductors but
are not complete enough to give rigorous qualitative predictions. Only selected
qualitative features are expected, since the calculations assume infinitely high
electronic barriers at the interfaces of the thin film and since the interface phonon
modes are not included in the spectrum of phonon modes. Figure 10.22 shows the
Figure 10.17. (a) Theoretical and (b) experimental electron energy distribution
ρ(E) for the heterostructure described previously and shown in the inset. The arrow
in the inset shows the current direction. The theoretical electron energy distributions
ρ(E) are shown at 0 Å, 600 Å, and 1500 Å from the point of injection. The
experimental curve for ρ(E) was obtained at Ve = −0.3 V. From Choi et al. (1996),
superconducting transition temperature Tc for a thin film with an electron density

n = 2 × 1022 cm−3 as a function of the film thickness, d. The Debye energy,
h̄ω D , is taken to be 100 K. In this figure the broken line represents the results of
Blatt and Thompson (1963) and Thompson and Blatt (1963); these results were
obtained for the case where the electrons are treated as dimensionally confined
but only bulk phonons are considered. The solid line represents the results of
Hwang et al. (2000) where the BCS prediction for the superconducting transition
temperature Tc is modified to take into account both electron confinement and
phonon confinement.
The results of Hwang et al. (2000) demonstrate that phonon confinement effects
play a role in the superconducting properties of thin film superconductors. The
conclusion was expected since thin film superconductors may be as thin as several
tens of ångstroms and since the binding energy of the phonon-mediated Cooper pairs
is determined by the electron–phonon interaction. For films with thicknesses in the
range shown in Figure 10.22 it is reasonable to expect that the interface phonons
will also play an important role in determining selective properties of thin film
superconductors.
Figure 10.18. Experimental results for the photocurrent transfer ratio α are shown
as a solid curve. The theoretical results for plasmon emission, phonon emission, and
both phonon and plasmon emission are designated by triangles, crosses, and circles
respectively. The difference between the theoretical and experimental contributions
at high Ve is due to the non-linear relation between the injection energy and Ve .
From Choi et al. (1996), American Institute of Physics, with permission.
Figure 10.19. (a) Double-well active region of an AlAs/GaAs heterostructure

designed so that the lower quasi-bound carrier state A1 of well A is depopulated by
phonon-assisted transitions to state B1 of well B. The lasing transition is highlighted
by the outlined arrow. Diagram (b) depicts the energies in meV, E3, E2, and E1 of
the A1, B2, and B1 subbands respectively as functions of the width in nm, a1 , of the
narrow well. The barrier thickness a2 and the width of the wider well a3 are 2 nm
and 10 nm respectively. The lifetimes τ11 , τ22 , τ21 , and τout are respectively the
intrasubband transition rates for the first and second subbands in well A, the
intersubband transition rate for the second subband in well A, and the escape rate
from state A1 to state B1. The lifetime τout should be short compared with τ21 in
order to maintain a population inversion between A2 and A1. From Stroscio et al.
(1999), American Institute of Physics, with permission.
10.6 Generation of acoustic phonons in quantum-well

structures
Komirenko et al. (2000b) have calculated the rate of generation of high-frequency

confined acoustic phonons via the deformation-potential interaction between drift-
ing electrons and the confined acoustic phonon modes in a quantum well. These
authors predict that the drifting electrons cause strong Cerenkov-like generation
of confined acoustic phonons and that they determine the gain coefficient for
this process as a function of the phonon frequency and parameters describing the
quantum-well structure. A gain coefficient of several hundred cm−1 is predicted for
a p-doped, 10-nm-wide Si/SiGe/Si quantum well.
To determine the rate of confined acoustic phonon generation, it is necessary to
calculate the electron–confined-acoustic-phonon scattering rates in the Si/SiGe/Si
quantum well. The carriers interact with acoustic phonons via the deformation-
c,v
potential interaction, Hdef , of Section 5.3, equation (5.39):
Figure 10.20. Interwell transition rates as a function of a1 for phonon-assisted

transitions between A1 and B1, as calculated by the Fermi golden rule for three
different models of the optical phonons participating in the phonon-assisted process:
the broken curve (a) assumes that only bulk GaAs phonons participate; the broken
curve (b) assumes that only bulk AlAs phonons participate; and the solid curve
includes the effects of the full spectrum of optical phonon modes for the
heterostructure. As in Figure 10.19, a2 = 2 nm and a3 = 10 nm. From Stroscio et al.
(1999), American Institute of Physics, with permission.
Figure 10.21. Peak values of the interwell phonon-assisted transition rates for
conditions of near-degeneracy of energy levels A1 and B2 for a series of values of
the width a3 of the wider well. The curves corresponding to a3 = 10 nm show the
relative contributions of the confined phonons (broken curve) and interface phonons
(dotted curve). The bold broken line with a minimum at about 8.5 nm is the rate of
non-radiative intrawell intersubband transitions, 1/τ12 , for the case where
a3 = 18 nm. From Stroscio et al. (1999), American Institute of Physics, with
permission.
Figure 10.22. Superconducting

critical temperature Tc for a
thin film with an electron
density n = 2 × 1022 cm−3 ,
as a function of the film
thickness d. The solid line
denotes the result for
confined slab phonons and
the broken line represents
the corresponding result for
bulk phonons. The Debye
temperature, h̄ω D , is taken
to be 100 K. From Hwang
permission.
c,v
Hdef = E c,v (a) = E 1c,v ∇ · u, (10.33)
where, as discussed in subsection 7.6.1,

h̄
u(r) = †
(an,q + an,−q )wn (q , z) eiq ·r . (10.34)
q ,n 2L 2 ρωn (q )
These modes are normalized according to the procedures of Section 5.1 but in terms
of wn (q , z) instead of un (q , z), since the considerations of Appendix A make it
√
clear that it is convenient to use wn = ρun ; un (q , z) ≡ un,q is obtained from

the mode amplitudes of subsection 7.6.2 by dividing by exp i(q · r − ωq t) ,
since this factor is included separately in the above equation for u(r) and in
the energy-conserving delta function in the Fermi golden rule. Clearly, transverse
modes do not contribute to the deformation potential. Two of the types of mode
in Section 7.2 contain longitudinal components: the dilatational modes and the
flexural modes. The dilatational modes are irrotational and they are associated
with compressional distortions of the medium; the compressional character of
these modes leads to local changes in the volume of the medium. As discussed
in subsection 7.6.2, Wendler and Grigoryan (1988) derived the localized acoustic
modes for an embedded quantum well. For a symmetric quantum well, the electrons
couple via the deformation potential to the symmetric shear vertical (SSV) confined
acoustic modes. The quantum-well heterostructure and the lowest-order SSV mode
are depicted in Figures 10.23(a), (b).
Figure 10.23. (a) Quantum-well heterostructure and (b) distribution of

displacement components u 1 and u 3 for the SSV mode, for conditions corresponding
to those of maximum amplification, as defined in Figure 10.24. v is the drift velocity
of the electron. From Komirenko et al. (2000b), American Institute of Physics, with
permission.
The electronic wavefunction for the lowest two-dimensional subband of the layer
is given by
1
k (ρ, z) = √ exp(ik · ρ )χ(z), (10.35)
S
where k is the two-dimensional wavevector and, as defined previously, ρ ≡ (x, y).
Only this lowest electronic subband is taken to be occupied. Then, the probability
of transitions between electronic states k and k due to the emission or absorption
of a confined phonon of band n and wavevector q is given by
2π $ %
P (±) (k , k n, q = |M(q)|2 n n,q + 12 ± 12 δk x ∓q,k x δk y ,k y
h̄
$ %
× δ E(k ) − E(k ) ∓ h̄ωn,q
2 3
× F(k ) 1 − F(k ) , (10.36)
where
∞
Ec
M(q) = el 1 ∇ · un,q χ 2 (z)dz, (10.37)
κ (q) −∞

n n,q is the phonon occupation number of mode n, q , κ el (q) is the electron
permittivity described in Appendix F (Bastard, 1988; Komirenko et al., 2000b),
F(k ) = F(k x , k y ) is the electron distribution function, and q = q . Appendix
F provides a derivation of κ el (q) based on the Lindhart method as applied to a
two-dimensional electron gas (Bastard, 1988).
To determine the net rate of generation of localized acoustic phonons, Komirenko
et al. (2000b) considered
dn n,q (+) (−)
= γn,q
(1 + n n,q ) − γn,q n
n,q
− βn,q n n,q , (10.38)
dt
(±)
where the γn,q are determined by calculating the total rates of absorption and

emission of phonons for mode n, q and βn,q represents phonon losses such
as non-electronic phonon absorption or anharmonic phonon decay. Defining the
phonon increment by
(+) (−)
γn,q ≡ γn,q
− γn,q
, (10.39)
it follows that
m∗ 2 3
γn,q = |M(q)|2 I (+) (q) − I (−) (q) , (10.40)
π h̄ 3 q
where m ∗ is the effective mass and
∞
(±)
m ∗ ωn,q 1
I (q) = dk y F (sign q) ± q, k y . (10.41)
−∞ h̄|q| 2
For the case where an external electric field causes the electrons to drift with
a velocity Vdr , taking the electron distribution function to be a drifted Fermi
distribution FF , we have

m∗
F(k x , k y ) = FF k x − Vdr , k y . (10.42)
h̄
It then follows that αn,q = γn,q /Vg > 0 if the electron drift velocity exceeds the
confined phonon phase velocity, Vdr > ωn,q /|q|, where the phonon group velocity
Vg = dωn,q /dq. This condition is the same as the criterion for Cerenkov emission.
Figure 10.24. Dimensionless phonon amplification coefficient α/αmax versus

phonon frequency ω/ω0 for the two lowest SSV phonon branches. T = 50, 100,
150, and 200 K; the larger the value of T , the lower the value of α.
αmax = 290 cm−1 , ω0 = 110 GHz. The values of ω/ω0 and qd for the maxima of α
are depicted in the inset for T = 50 K. From Komirenko et al. (2000b), American
Komirenko et al. (2000b) evaluated αn,q for a p-doped, Si/SiGe/Si quantum well 10
nm thick, for the two lowest SSV phonon branches. The amplification coefficients
for these phonon branches are shown in Figure 10.24 for temperatures T of 50, 100,
150, and 200 K. In these numerical results, the hole density is taken as 1012 cm−2 ,
the drift velocity is given by Vdr = 2.5 ct1 and ct1 = 3.4 × 105 cm s−1 for the SiGe
layer.
The amplification coefficient for this p-doped Si/SiGe/Si quantum well is seen
to be of the order of tens to hundreds of cm−1 for confined modes in the sub-THz
frequency range.
Chapter 11
Concluding considerations
Now there is one outstandingly important fact regarding
Spaceship Earth, and that is that no instruction book came with
it.
R. Buckminster Fuller, Operating Manual for Spaceship Earth, 1969
11.1 Pervasive role of phonons in modern solid-state

devices
As illustrated throughout this book, phonon effects are pervasive in modern solid-
state devices. As is illustrated by the many examples for Chapters 7–10, the
importance of these effects is usually at least as great for dimensionally confined
structures as for bulk structures. Indeed, in Chapter 7 the effects of dimensional
confinement were seen to be important even for biological structures! In this case,
a cylindrical shell immersed in a fluid (Sirenko et al., 1996b) was used to model
the vibrational behavior of microtubules (MTs) immersed in water. In addition,
the examples of Chapters 7 and 9 illustrate that the elastic continuum model
provides an accurate description of acoustic phonons in dimensionally confined
structures of many geometries including thin films, nanowires with rectangular and
circular cross sections, and a variety of dot-like structures. These structures will
inevitably be pervasive as elements of nanoscale structures mimicking the well
known and larger microelectromechanical structures. Indeed, Cleland and Roukes
(1996) reported a technique for fabricating nanometer-scale mechanical structures
from bulk, single-crystal Si substrates. As another example of acoustic phonon
effects in dimensionally confined structures, it was recently predicted theoretically
that Cerenkov-like effects lead to the generation of high-frequency confined acoustic
phonons in quantum wells (Komirenko et al., 2000b); see Section 10.6.
In Chapter 8, values of carrier–optical-phonon scattering rates calculated for a
variety of dimensionally confined semiconductor structures were found to exceed
1013 s−1 . Clearly, such rates are among the largest encountered in such structures
218
11.2 Future trends: phonon effects in nanostructures and phonon engineering 219
and in many cases dominate over all other scattering mechanisms. It is well
known that these relatively large carrier–phonon interactions play a major role
in determining carrier mobilities (Ferry, 1991; Mitin et al., 1999). In Chapter
10, phonons in dimensionally confined structures were seen to make a dominant
contribution to the valley current in specific double-barrier quantum-well structures
(Turley et al., 1993) and the properties of thin film superconductors were shown
to depend on the spectrum of confined phonons for sufficiently thin films (Hwang
et al., 2000).
Moreover, it was shown in Chapter 10 that dimensional confinement of phonons
in intersubband semiconductor lasers changes the laser gain (Kisin et al., 1997,
1998a, b) and leads to enhanced population inversions in some asymmetric double-
barrier quantum-well lasers (Stroscio et al., 1999). Indeed, Educato et al. (1993),
Julien et al. (1995), Wang et al. (1996a, b), and Gauthier-Lafaye et al. (1997)
have examined optically pumped intersubband scattering in coupled quantum-well
lasers. These same authors have shown that interface–phonon-assisted transitions
are important in such structures. Such results illustrate the importance of phonon
confinement effects in intersubband lasers. As discussed in Chapter 10, the proper
treatment of optical phonon confinement in optical systems such as intersubband
semiconductor lasers depends critically on the detailed energy spectrum of the
phonons. This is also true for a number of novel current-injection semiconductor
intersubband lasers (Sun et al., 1993; Zhang et al., 1996; Sung et al., 1996; Faist
et al., 1996a, b), as is illustrated by the calculations of Stroscio (1996) and Kisin
et al. (1997, 1998a, b).
11.2 Future trends: phonon effects in nanostructures and

phonon engineering
In this section, some speculative observations concerning phonons in nanostructures
are made as a guide to potentially fruitful avenues of research on dimensionally
confined phonons. In predicting future developments in the fields of phonon effects
in nanostructures and phonon engineering, it is instructive to consider emerging in-
ternational efforts for both nanostructures and bulk structures. Indeed, novel phonon
effects in bulk materials are likely to have counterparts in nanostructures. Progress
in femtosecond lasers and ultrafast spectroscopy and the continued development of
novel techniques for fabricating nanostructures such as quantum dots (Empedocles,
1996) have been the basis for experimental observations of coherent oscillations
of acoustic phonons in superlattices (Sun et al., 1999), damped spherical acoustic
breathing modes in quantum dots (Krauss and Wise, 1997, see the supporting
analysis of Stroscio and Dutta, 1999), optical phonons near the surface of bulk GaAs
(Cho et al., 1990), optical phonons in Ge (Pfeifer et al., 1992), and the excitation of
coherent phonons in Sb, Bi, Te, and Ti2 O3 (Cheng et al., 1991). Coherent phonon
220 11 Concluding considerations
effects are likely to represent one of the future trends in in nanostructures as well
as in phonon engineering. Pötz and Shroeder (1999) pointed to the potential for
achieving coherent control in atoms, molecules, and semiconductors.
Phonon engineering in nanostructures is likely to be a future avenue for
device-related research and development. Indeed, the ability to model the phonon
modes in dimensionally confined structures has been the basis for efforts to design
nanostructures such that the resulting carrier and phonon states are tailored to yield
dissipative and scattering mechanisms different from those of the corresponding
bulk structures. Sakaki (1989) analyzed theoretically the electronic structure of
quantum-wire superlattices and coupled quantum-box arrays with the purpose of
engineering the electron energy subbands in such a way that carrier–optical-phonon
scattering is suppressed. As another example, the use of metal–semiconductor
heterointerfaces to reduce carrier–interface–phonon scattering rates (Stroscio et al.,
1992) represented one of the first approaches to phonon engineering in nanos-
tructures. Indeed, the first calculations of carrier–LO-phonon scattering rates in
polar-semiconductor quantum wires (Stroscio et al., 1990; Kim et al., 1991) revealed
that the scattering of carriers with interface optical phonon modes dominates over
the scattering associated with confined and half-space phonons. As a means of
reducing the unwanted inelastic scattering caused by the interface LO phonons
in semiconductor quantum wires, it was suggested (Stroscio et al., 1992) that the
large carrier–LO-phonon scattering caused by interface phonons could be eliminated
by the encapsulation of the quantum wire in a metal. In a related example of
phonon engineering, Leburton (1997) modeled dissipation- and scattering-time
engineering in heterostructure-based quantum devices. Indeed, Leburton and his
collaborators – Educato et al. (1993), Julien et al. (1995), Wang et al. (1996a, b),
and Gauthier-Lafaye et al. (1997) – have exploited such effects in optically pumped
intersubband scattering in coupled quantum-well lasers. Clearly, efforts in phonon
engineering will be one of the future research trends based on the theories of
confined phonons presented in this book.
Appendices
Appendix A: Huang–Born theory
Huang (1951) and Born and Huang (1954) took the most general form of the
microscopic theory of diatomic polar crystals to be described by a pair of equations
relating w, E, and P,
ẅ = aw + bE and P = dw + cE. (A.1)
Born and Huang showed that d equals b as a result of energy conservation. In (A.1),
√
w = µN /V u; u, µ, N , and V are the relative displacement of the two ions, the
reduced mass of the ion pair, the number of unit cells in the crystal, and the volume
of the crystal respectively. Assuming a time dependence e−iωt and eliminating w
from these equations yields the relation

b2
P= c+ E. (A.2)
−a − ω2
Since D = E + 4π P = (ω)E, (A.2) may be rewritten as
4π b2
(ω) = 1 + 4π c + . (A.3)
−a − ω2
From the generalized Lyddane–Sachs–Teller relation given in subsection 2.3.3,

(2.32), it is clear that
221
222 Appendices
ωLO
2 − ω2 2 − ω2 ) + (ω2 − ω2 )]
[(ωTO LO TO
(ω) = (∞) = (∞)
ωTO
2 − ω2 ωTO
2 − ω2
ωLO
2 /ω2 − 1
TO
= (∞) + (∞)
1 − ω2 /ωTO
2
(0) − (∞)
= (∞) + (∞) . (A.4)
1 − ω2 /ωTO
2
Comparing the relation (A.3) of Born and Huang with that based in the Lyddane–
Sachs–Teller equation, it is evident that
1/2
(∞) − 1 (0) − (∞)
c= , a = −ωTO
2
, and b= ωTO .
4π 4π
(A.5)
Thus, the pair of equations in (A.1) and (A.2) put forth by Huang and Born may
be written as
1/2
(0) − (∞)
(ωTO
2
− ω )w =
2
ωTO E (A.6)
4π
and
1/2
(0) − (∞) (∞) − 1
P= ωTO w + E. (A.7)
4π 4π
Re-introducing the relative displacement u these equations become

1/2
V
(ωTO
2
− ω )u =
2
(0) − (∞)ωTO E, (A.8)
4π µN
and
1/2
µN (∞) − 1
P= (0) − (∞)ωTO u + E. (A.9)
4π V 4π
Appendix B: Wendler’s theory

Wendler (1985) formulated a theory for the carrier–optical-phonon interactions in
dielectric bilayer systems that includes the effect of the electronic polarizability on
the phonon eigenmodes and their frequencies. Wendler’s model predicts several
relations that are useful in calculations of carrier–optical-phonon interactions in
dimensionally confined systems. As in subsection 2.3.1, the displacement field in
Wendler’s model is related to the fields E(r) and P(r) through the driven oscillator
Appendices 223
equation and through the effective charge, en∗ : for a binary medium labeled by
subscript n,
−µn ω2 un (r) = −µn ω0n

2
un (r) + en∗ Elocal (r), (B.1)
P(r) = nn en∗ un (r) + nn αn Elocal (r), (B.2)
where µn = m n Mn /(m n + Mn ) is the reduced mass, αn is the electronic polariz-

ability per unit cell and where by the Lorentz relation
4π
Elocal (r) = E(r) + P(r). (B.3)
3
Here, Elocal (r) is the local electric field associated with the optical modes acting
on the electron shells of the ions. From Poisson’s equation,
(r, t) = −4π (total (r, t), (B.4)
where (total (r, t) includes both bulk and surface polarization charge density; using
the Green’s function approach Wendler took a solution to Poisson’s equation of the
form
∂ 1
(r, t) = − d 3 r P (r , t).
| β
(B.5)
β
∂r β |r − r
Then, with E α (r, t) = −∂(r, t)/∂rα and with the Lorentz equation (B.3), it
follows that
4π
E αlocal (r, t) = (λn − λ0n )Pα (r, t)
3

+ 4π d 3 r αβ (r − r )Pβ (r , t), (B.6)
β
where the Green’s tensor is given by
1 ∂2 1
αβ (r − r ) = . (B.7)
4π ∂rα ∂rβ |r − r |
By assuming that P(r, t) depends sinusoidally on time, Wendler showed that

$ %
(λn − λ0n ) 1 − 43 π nn αn + 43 π
Pα (r) = −4π d 3 r αβ (r − r )Pβ (r ), (B.8)
1 − nn αn (λn − λ0n ) β
where
λn = 4πω2 /ωplasma,n
2
, λ0n = 4π ω0n
2
/ωplasma,n
2
, (B.9)
and the plasma frequency squared, ωplasma,n

2 , is given by
224 Appendices
∗2
ωplasma,n,a(b)
2
= 4π nn en,a(b) /µn,a(b) . (B.10)
By defining a two-dimensional Fourier transform for the electrostatic fields of the
form
∞
1
P(r) = d 2 q eiq ·r P(q , r3 ), (B.11)
2π −∞
with r in the plane of the heterointerface of the bilayer system, Wendler showed
that
Elocal (q , r3 ) = n P(q , r3 ), (B.12)
where
λn − λ0n
n = (B.13)
1 − nn αn (λn − λ0n )
and
1/2
θn
u(q , r3 ) = P(q , r3 ), (B.14)
nn en∗
with
1/2 1
θn = . (B.15)
1 − nn αn (λn − λ0n )
Moreover, the Green’s function approach of Wendler yields several useful
conditions:
nn αn
n (∞) = 1 + 4π ,
1 − 43 π nn αn
1
ωLO,n
2
= ω0,n
2
+ 23 ωplasma,n
2
,
1+ 3 π nn αn
8
1
ωTO,n
2
= ω0,n
2
+ 13 ωplasma,n
2
,
1− 3 π nn αn
4
1 1
ωLO,n
2
− ωTO,n
2
= 23 ωplasma,n
2
+ 13 ωplasma,n
2
1+ 3 π nn αn
8
1− 3 π nn αn
4
ωplasma,n
2
= .
(1 + 83 π nn αn )(1 − 43 π nn αn )
(B.16)
where n represents either material 1 or material 2. As may be verified alge-
braically, the Lyddane–Sachs–Teller relations of subsection 2.3.3 are satisfied by
the frequencies in (B.16) (Wendler, 1985). Wendler’s treatment of the effect of
dielectric polarizability provides the basis for many works on optical phonons in
dimensionally confined systems. The utility of Wendler’s solution is illustrated in
Chapter 7.
Appendices 225
Appendix C: Optical phonon modes in

double-heterointerface structures
Nash (1992) made an extremely important contribution to the field of phonons in
nanostructures by showing that intrasubband and intersubband electron–phonon
scattering rates are independent of the basis set used to describe the confined
polar-optical phonons in a semiconductor slab. Comas et al. (1997) also formulated
a model for treating both the mechanical and electrostatic fields associated with
phonons in heterogeneous semiconductor structures.
In particular, Nash showed that the so-called slab modes derived with electrostatic
boundary conditions, the guided modes derived with mechanical boundary condi-
tions, and the reformulated slab modes of Huang and Zhu (1988) all predict the same
scattering rates as long as each set of modes is orthogonal and complete. Moreover,
Nash’s analysis resolved the long-standing controversy over whether the normal
modes satisfy mechanical or electromagnetic boundary conditions; specifically,
√
Nash shows that the mechanical boundary conditions apply to w = ρu (where
u is the relative displacement of the ion pairs and ρ is the reduced mass per unit
volume as discussed in Appendix A) and its derivatives, and the electromagnetic
boundary conditions apply to and D. Furthermore, Nash showed that both types of
boundary condition are necessary to obtain the normal modes of the heterostructure.
Nash considered a Lagrangian density, for a polar material, of the form (Born and
Huang, 1954),
1 ∂w ∂w 1 2
L= · − ωTO w · w
2 ∂t ∂t 2
1 1 ∂wk ∂wl
+ (∞)∇ · ∇ − γ w · ∇ + Z i jkl , (C.1)
8π 2 i jkl ∂r j ∂ri
where
1/2
1
γ = (∞)(ωLO − ωTO )
2 2
4π
! "1/2
1
= [(0) − (∞)] ωTO
4π
! "1/2
(∞) 1 1
= ωLO 4π − (C.2)
4π (∞) (0)
√
and where the displacement is described in terms of w = ρu for the same reasons
as in Appendix A. As will become apparent, it is convenient to define
! "1/2
1 4π 1 1
= γ = ωLO 4π − . (C.3)
β (∞) (∞) (0)
In the Lagrangian density (C.1), the first term is the kinetic energy, the second
term is the potential energy of the lattice due to short-range forces, the third term
226 Appendices
is due to the potential energy of the macroscopic electric field in the absence of
ionic motion, the fourth term is the potential energy associated with the coupling
of the lattice to the macroscopic electric field, and the fifth term represents the
quadratic dispersion associated with the short-range forces between ions, for the
case of isotropic dispersion, Z i jkl = Aδi j δkl + Bδik δ jl + Cδil δ jk .
The Euler–Lagrange equations derived by Nash from the variations of w and
to describe the optical phonon modes in a slab situated between z = 0 and z = d
are
∂ 2w
+ ωTO
2
w + γ ∇ + A∇ 2 w + (B + C)∇(∇ · w) = 0 (C.4)
∂t 2
and
∇ · [(∞)∇ − 4π γ f (z)w] = 0 (C.5)
respectively. In these equations, γ is as given in (C.2) and f (z) = )(z) − )(d − z),
) being the Heaviside step function. Following Nash, w is written as the gradient
of a mechanical potential χ̄, w = −∇ χ̄ . The second of these equations is Poisson’s
equation,
∇ · [(∞)∇] = 4π γ ∇[ f (z)w] = −4π γ ∇ 2 χ̄ − 4πγ ∇ f (z)∇ χ̄ , (C.6)
or
1
− ∇ · [(∞)∇]
(∞)

4π γ ∂ χ̄ ∂ χ̄
= ∇ χ̄ −
2
δ(z − d + η) + δ(z − η) , (C.7)
(∞) ∂z z=d ∂z z=0
where the last two terms on the right-hand side of (C.7) arise from the surface
polarization charge due to the discontinuity of the z-component of w. In these
results, w is taken to be normalized as described previously. That is, these modes
are normalized (according to the procedures of Section 5.1) in terms of wm (q , z)
instead of um (q , z), since the considerations of Appendix A make it clear that it is
√
convenient to use wm = ρum :

h̄
um (r) = (a
2 ρω (q ) m,q
†
+ am,−q )wm (q , z) eiq ·r . (C.8)
q ,m 2L m
Here the subscript m is retained to remind us that this result holds for a general
medium m. This last result is, of course, consistent with the normalization condition
of subsection 7.3.1, (7.44),

√ √ h̄
L 2 [ µn um (q, z)]∗ · [ µn um (q, z)] dz = ; (C.9)
2ωm (q)
Appendices 227
this equivalence follows straightforwardly by noting that q = q. As discussed in

Chapters 5 and 7, the interaction potential for these normalized modes is

m (r) = m (q, z) e−iq·ρρ . (C.10)
q
Solutions of the form w = −∇ χ̄ satisfying these Euler–Lagrange equations are
(ωLO
2
− ω2 )∇ 2 χ̄ + µ∇ 2 ∇ 2 χ̄ = 0 (C.11)
and
1 2
= (ω − ω2 )χ̄ + µ∇ 2 χ̄ − B0 , (C.12)
γ TO
for 0 < z < d, with µ = A + B + C and B0 a constant. For solutions outside the
slab, satisfies Poisson’s equation. As discussed previously, the fields for a slab
exhibit translational invariance in the direction normal to the surface of the slab;
accordingly, χ̄(r) = χ(z) eiq·ρρ and (r) = φ(z) eiq·ρρ , where q and ρ are as defined
previously. Thus, for q = 0 it follows that B0 = 0. Then the last two equations may
be written as

d2 2 − ω2
ωLO d2
µ + − q 2
− q 2
χ(z) = 0 (C.13)
dz 2 µ dz 2
and
1 2 d 2 χ(z)
φ(z) = (ωTO − ω2 − µq 2 )χ(z) + µ , (C.14)
γ dz 2
respectively. Nash has applied the methods used in analyzing the Sturm–Liouville
equation to show that the normal-mode solutions for the slab must satisfy the
following boundary conditions:
= 0 at z = ±∞ (C.15)
and
wx = w y = wz = 0 or wx wx + w y wy + wz = 0

(C.16)
at z = 0 and d.
Again taking w = −∇ χ̄ , the equations (C.16) reduce to
χ = χ = 0 and χ = 0 (C.17)
respectively.
228 Appendices
At this point it is instructive to consider each of the commonly used sets of

normal modes and to compare the boundary conditions they satisfy with those
obtained within the context of the Sturm–Liouville equation. Nash took knq 2 =
(ωLO − ω )/(µ − q ) in the fourth-order equation for χ(z) to define the fourth-order
2 2 2
eigenvalue equation
2 2
d d
+ k 2
nq − q 2
χnq (z) = 0, (C.18)
dz 2 dz 2
where knq2 is the eigenvalue of χ (z) and where the subscripts nq identify different
nq
members of the basis set. As usual, q is the wavevector in the x y-plane and n is
here an integer used to distinguish different modes having the same value of q. It
is this equation that Nash uses to generate complete sets of orthonormal functions
representing the phonon modes. As discussed in Chapter 7, the orthogonality
relation for the normal modes is

d 3 r wα (r) · wβ (r) = 0 for α = β, (C.19)
where α and β each denote the set of quantum numbers nq. Recalling that
w = −∇ χ̄ and that χ̄(r) = χ(z) eiq·ρρ this orthogonality relation may be written
as
d ∗
∗ 1 ∗
dχnq dχn q
χnq · χn q = dz q 2 χnq (z) · χn q (z) + . (C.20)
2d 0 dz dz
The various sets of confined and interface optical phonon modes, which have
been discussed in the literature include the so-called slab modes that satisfy the
electrostatic boundary conditions, the Huang–Zhu modes, which are based on a
reformulation of the slab modes, and the so-called guided modes, which satisfy
mechanical boundary conditions. For the slab modes,
(n + 1)π z
χnq (z) = sin ; (C.21)
d
the boundary conditions on χnq (z) at z = 0 and d are χnq = χnq = 0 and it is
necessary to include the interface modes to obtain a complete set of modes. The
electrostatic potentials for these modes, nq , are given by
4π γ
nq = − χnq (z). (C.22)
(∞)
The normal modes satisfying = 0 at z = ±∞ and χ = χ = 0 at z = 0 and d
are the slab modes.
For the modified Huang–Zhu modes given by Nash,

cos µnq π z /d + Dnq cosh qz n even,
χnq (z) = (C.23)

sin µnq π z /d + Dnq sinh qz n odd;
Appendices 229
the boundary conditions on χnq (z) at z = 0 and d are χnq = χnq = 0 and it is
necessary to include the interface modes to obtain a complete set of modes. The
Huang–Zhu (H–Z) modes have been defined to satisfy these boundary conditions
at the interfaces since the available microscopic calculations exhibit this behavior.
In practice, the boundary conditions χnq = χnq = 0 are satisfied by selecting the
necessary values of Dnq and µnq . The H–Z modes are referred to frequently as
the reformulated slab modes and also as the reformulated modes of the dielectric
continuum model (with electromagnetic boundary conditions). Huang and Zhu
(1988) showed that the modes of the dispersionless microscopic theory at small q
are approximated well by these reformulated slab modes. The electrostatic potentials
for these modes, nq , are given by
4πγ
nq = − χnq (z). (C.24)
(∞)
For the guided modes,
nπ z
χnq (z) = cos ; (C.25)
d
the boundary conditions on χnq (z) at z = 0 and d are again χnq = χ = 0 and, as
nq
pointed out by Nash, it is not necessary to include the interface modes to obtain a
complete set of modes. The electrostatic potentials for these modes, nq , are given
by

4πγ cos nπ z/d − e−qd/2 cosh q(z − d/2) n even,
nq = −
(∞) cos nπ z/d + e−qd/2 sinh q(z − d/2) n odd.
(C.26)
While these guided modes satisfy χnq = 0 at the interfaces, in agreement with the
q = 0 modes observed in Raman backscatter experiments (Sood et al., 1985), they

do not satisfy the conditions that χnq (z) = 0 at z = 0 and d. Unlike the slab and H–Z
modes, the guided modes have non-zero exponentially decreasing values outside the
slab: nq (z) = nq (0) eqz for z < 0 and nq (z) = nq (0) eq(d−z) for z > d.
Finally, the interface modes that satisfy ∇ 2 χ̄ = 0 are given by
sinh q(z − d/2) cosh q(z − d/2)
χIF,1q (z) = and χIF,2q (z) = .
sinh qd/2 cosh qd/2
(C.27)
These expressions are consistent with those derived in Chapter 7; as argued there,
the IF optical phonon mode potentials decrease exponentially with distance from the
slab; that is, as eqz for z < 0 and as eq(d−z) for z > d.
Characterizing the interaction potential for the region m, m (q, z), in terms of the
quantum numbers n and q for that region it follows for each of the sets of complete,
normalized modes that the interaction Hamiltonian is

HFr = −e(r) = −e nq (q, z) e−iq·ρρ . (C.28)
nq
230 Appendices
Clearly, of the three commonly used sets of phonon modes, only the slab modes
satisfy the conditions that = 0 at z = ±∞ and χnq = χnq = 0 at z = 0 and

d. The alternative set of boundary conditions, = 0 at z = ±∞ and χnq = 0 at
z = 0 and d, is not satisfied by any of the commonly used sets of phonon modes. As
argued by Nash (1992), the Euler–Lagrange equations, (C.4) and (C.5) are satisfied
in each material layer and provide the basis for determining the so-called connection
rules at each heterointerface. In the dispersionless limit the Lagrangian density does
not contain terms with spatial derivatives of w and, accordingly, the significance of
mechanical boundary conditions is unclear in this case.
For this reason, Nash examined the Euler–Lagrange equations at the heteroin-
terfaces for the case where the modes have non-zero dispersion, as described by
the Lagrangian density of (C.1). In this case, the connection rules may be seen
to be that w/ρ 1/2 , , and the z-components of the electric displacement vectors,
Dz = (1/4π)(∞) − γ wz , are continuous at the heterointerfaces. Thus in
the dispersive continuum model there is no contradiction in applying mechanical
boundary conditions to w and w and at the same time applying electromagnetic
boundary conditions to and D. Indeed, both types of boundary condition must be
invoked to derive the full set of normal modes.
As an illustration leading to an appreciation of the importance of Nash’s con-
tribution to the understanding of the boundary conditions for the system at hand,
it is instructive to oversimplify the current analysis artificially by considering the
equation
(∞)∇ − 4π γ w = 0. (C.29)
The left-hand side of this equation appears in the Euler–Lagrange equation obtained
from variation of in the Lagrangian density: ∇ · [(∞)∇ − 4πγ w] = 0. Clearly,
this simplified equation is not sufficient to describe the system fully but it is in
fact the equation assumed in the mechanical models to derive longitudinal modes
for the construction of normal modes. Let us consider (C.29) and show that it in
fact gives the correct Fröhlich interaction Hamiltonian for electron–polar-optical-
phonon interactions, as derived in Section 5.1. Substituting the expression for γ in
terms of ωLO , (0), and (∞), it follows immediately that
! "1/2
1 1
∇ = ωLO 4π − w. (C.30)
(∞) (0)
√
Then using the expression of Appendix A, w = µN /V u, and the normalized
displacement of Section 5.1 for the case where ωq = ωLO and only the longitudinal
polarization contributes, it follows immediately that
Appendices 231
! "1/2
1 1
∇ = ωLO 4π −
(∞) (0)

N 1 h̄ iq·r
× aq e + aq† e−iq·r
V N q 2ωLO
! "1/2
2π h̄ωLO 1 1
= ωLO 4π −
V (∞) (0)

† −iq·r
× aq e iq·r
+ aq e . (C.31)
q
Then, using the result that ∇ = −iq and multiplying by −e to obtain the
Hamiltonian, it follows that

2πe2 h̄ωLO 1 1 1 iq·r
HFr = −i − aq e − aq† e−iq·r ,
V (∞) (0) q q
(C.32)
which is precisely the result obtained in Section 5.2. Thus, the simplified result,
(∞)∇ − 4πγ w = 0, gives the correct three-dimensional Fröhlich interac-
tion Hamiltonian. However, this successful derivation does not justify the use of
(∞)∇ − 4πγ w = 0 instead of ∇ · [(∞)∇ − 4πγ w] = 0 in analyzing the
boundary conditions for the phonon modes in a slab. Indeed, as argued by Nash
(1992) the longitudinal waves obtained from this simplified equation do not provide
sufficient degrees of freedom for the mechanical and electromagnetic boundary
conditions to be satisfied simultaneously.
The understanding gained through Nash’s paper (1992) has been critical in
sorting out the correct approaches for calculating carrier–optical-phonon scattering
rates in dimensionally confined structures. Of course, the key feature of such cal-
culations is selecting a complete, orthogonal set of phonon modes. Notwithstanding
the situation that only slab modes satisfy the desired boundary conditions, Nash
demonstrated by explicit calculations that any of the three sets of complete and
orthogonal modes – slab modes plus IF modes, H–Z modes plus IF modes, and
guided modes – may be used as a basis set for determining the intrasubband and
intersubband electron–phonon scattering rates. Nash showed for a quantum well of
width d that
5
72 1
e 2
f nq (z) i ∝ ωnq d 2 q n 1 (q) n ∗1 (q) f n (q),
2
(C.33)
q 2q
where the form factor, f n (q), is given by

2
q ωLO
2
f n (q) = dz n 0 (z)βnq (z) ∗ (z) χ (z) ω2
, (C.34)
dχnq nq nq
and where
232 Appendices
|i = ψ0 (z, i)ψ1 (ρ, i), | f = ψ0 (z, f )ψ1 (ρ, f ),

ρ ),
n 1 (q) = d 2 ρ exp(−iq · ρ )n 1 (ρ (C.35)
ρ ) = ψ1∗ (ρ, f )ψ1 (ρ, i),

n 1 (ρ n 0 (z) = ψ0∗ (z, f )ψ0 (z, i).
Thus, for a given in-plane momentum transfer q the form factor f n (q) describes
the probability that an electron will scatter from the initial subband |i to the
final subband | f as a result of interaction with phonons of branch, n. The
explicit numerical calculations of Nash (1992) showed that the form factors f n (q)
corresponding to the three sets of complete and orthogonal modes – slab modes plus
IF modes, H–Z modes plus IF modes, and guided modes – are identical. This result
is evident from Figure C.1 for the case of intrasubband transitions and Figure C.2
for the case of intersubband transitions.
Figure C.1. Form factors

f n (q) for intrasubband
scattering for optical modes
in a quantum well as
functions of qd for three
different cases: (a) slab
modes, (b) reformulated slab
modes, and (c) guided
modes. The labels 1, 2, and
∞ designate the
lowest-order mode, the
second-order mode, and the
infinite sum over all modes.
(d) depicts the total form
factor for bulk modes, f B ,
as well as that for the
interface modes, f I ;
f tot = f B + f I . From Nash
(1992), American Physical
Appendices 233
These results indicate that any of the three sets of complete and orthogonal
modes for a semiconductor layer mentioned above may be used to calculate
electron–phonon scattering for the nearly degenerate longitudinal optical phonon
modes in the layer. The key point is that the modes of each of the three sets must be
complete and orthonormal. This result explains why the relatively simple modes of
the so-called slab model may be used to perform such a scattering-rate calculation
even though these modes are not the normal modes of a heterolayer.
As further proof of the validity of this approach, Nash showed that it is possible
to perform unitary transformations between the slab modes plus IF modes, the H–Z
modes plus IF modes, and the guided modes. These calculations demonstrate that
it is essential to include the IF optical phonon modes in the set of modes for the
Figure C.2. Form factors

f n (q) for intersubband
scattering for optical modes
in a quantum well as
functions of qd for three
different cases: (a) slab
modes, (b) reformulated slab
modes, and (c) guided
modes. For a description of
the labels, see the caption to
the previous figure. From
Nash (1992), American
permission.
234 Appendices
slab model – also known as the dielectric continuum model with electromagnetic
boundary conditions – and the Huang–Zhu models. Moreover, it is essential that the
IF modes are not included in the set of guided modes; this set of guided modes
is complete without the IF modes. Indeed, the guided modes explicitly include
IF-like exponentially decaying components outside the slab, unlike the slab and H–Z
modes. The dielectric continuum model with electromagnetic boundary conditions
leads to a convenient set of modes – the slab modes plus the interface modes –
for making relatively simple calculations of electron–phonon scattering rates in the
case where it is not necessary to account for the dispersion of the confined modes,
also referred to as the bulk-like modes in a slab. For GaAs and AlAs the total
dispersion over the entire Brillouin zone is only a few meV and it follows that
in typical carrier-transport calculations – including those of solid-state electronic
devices with dimensionally confined structures – it is not necessary to consider this
dispersion. As is evident in Chapter 10, it is necessary to consider energy differences
Figure C.3. The z-component of the atomic displacement, u z , and the

corresponding potential V for each of the modes reported by Rücker et al. (1991,
1992). The optical modes are presented for a 56 Å (001)-oriented GaAs quantum
well surrounded by two AlAs layers. The confined modes of highest frequency and
the two interface modes are displayed from top to bottom in order of decreasing
phonon frequency for each type of mode. The heterointerfaces are designated by the
vertical bars and the phonon wavevectors are qz = 0 for the modes displayed in the
upper half of the figure and q = 0.15 Å−1 for the modes displayed in the lower half
of the figure. From Rücker et al. (1991, 1992), American Physical Society, with
permission.
Appendices 235
as small as a few meV when modeling optoelectronic devices such as semiconductor

quantum-well lasers.
In closing this appendix, it is extremely enlightening to consider a graphical
comparison of the various macroscopic and microscopic optical phonon modes in
two-dimensional heterostructures. Rücker et al. (1992) made direct comparisons
of the slab, guided, and H–Z modes with the modes calculated with an ab initio
microscopic model. Figure C.3 presents the z-component of the atomic displace-
ment, u z , as well as the corresponding potential, V , for each of these modes as
reported by Rücker et al. (1991, 1992). The optical modes are presented for a
56 Å (001)-oriented GaAs quantum well surrounded by two AlAs layers. The
heterointerfaces are designated by the vertical bars and the phonon wavevectors
Figure C.4. Detailed comparison of potentials for the high-frequency symmetric

(IF1) and antisymmetric (IF2) interface modes, as calculated via both macroscopic
and microscopic models for q = (qx , 0, 0), where qx = 0.05 Å−1 for the modes
displayed in the upper half of the figure and qx = 0.15 Å−1 for the modes displayed
in the lower half of the figure. The broken-and-dotted and solid lines represent the
potentials derived from the macroscopic model and from the microscopic model
respectively. As expected, the agreement between the macroscopic and microscopic
calculations is excellent. From Rücker et al. (1992), American Physical Society, with
permission.
236 Appendices
are taken to be qz = 0 and q = 0.15 Å−1 . Figure C.4 presents a detailed

comparison of potentials for the interface modes as calculated via both macroscopic
and microscopic models. As expected, the agreement between the macroscopic and
microscopic calculations is excellent (Rücker et al., 1992).
Appendix D: Optical phonon modes in single- and

double-heterointerface würtzite structures
In subsection 7.3.1 the interface optical phonon modes for a single-heterointerface
würtzite structure were derived on the basis of the dielectric continuum model and
Loudon’s model for uniaxial crystals. In this appendix, all the optical phonon modes
in single- and double-heterointerface würtzite-like uniaxial structures are formulated
on the basis of the Loudon model and the dielectric continuum model (Lee et al.,
1998). These structures are depicted in Figures D.1(a), (b) respectively.
D.1 Single-heterointerface uniaxial structures

First consider the half-space (HS) modes of a heterostructure with a single interface
separating two semi-infinite polar-semiconductor regions. The c-axis is taken to be
normal to the heterointerface. In the region where z < 0 the dielectric constants are
(ω)⊥(),2 and in the region where z > 0 the dielectric constants are (ω)⊥(),1 .
The half-space modes behave like the normal bulk modes as z → ±∞ and they
Figure D.1. (a) The basic

single-heterointerface
structure and (b) the basic
double-heterointerface
structure. From Lee et al.
(1998), American Physical
Appendices 237
satisfy the electrostatic boundary conditions at z = 0. Consider the case in which an

optical phonon mode propagates into region 2 from region 1 as depicted in Figure
D1(a). The wavevector in region 1 is denoted by (q, q1, ) and the angle between this
wavevector and the z-axis is θ1 . When the allowed frequencies in region 1 do not
overlap with the allowed frequencies of region 2, it follows that (ω),2 (ω)⊥,2 > 0
and the modes in region 2 decay exponentially. Following an analysis similar to that
of subsection 7.3.1,

ρ
A cos q,1 z + B sin q,1 z z > 0,
φ(r ) = e iq·ρ
× (D.1)
Ceκ2 z z < 0,
and in the absence of free charge, ∇ · D = 0 so that
(ω),1 q,1
2
+ (ω)⊥,1 q 2 = 0 (ω),1 (ω)⊥,1 > 0,
(D.2)
(ω),2 κ22 − (ω)⊥,2 q 2 = 0 (ω),2 (ω)⊥,2 > 0.
From the continuity of the tangential component of the electric field at z = 0 it
then follows that A = C. From the continuity of the normal component of the
electric displacement at z = 0, ,1 q,1 B = ,2 κ2 C; thus, C = 1 B with 1 =
,1 k1, /(,2 κ2 ). Thus

1 cos q,1 + sin q,1 z > 0,
φ(r ) = φ0 eiq·ρρ × κ
(D.3)
1 e 2 z z < 0.
Now E⊥ and E are given by the appropriate gradients of φ0 . The normalization
condition of subsection 7.3.3,
! "
1 h̄ 1 1 ∂(ω)⊥,n 2
1 1 ∂(ω),n 2

= E ⊥,n + E,n dz,
L 2 2ω 4π 2ω ∂ω 4π 2ω ∂ω
(D.4)
2 2
may be calculated in terms of E⊥,n and E,n through

L/2 L/2
E⊥,1 2 dz = φ 2 q 2 sin q,1 z + 1 cos q,1 z 2 dz
0
−L/2 0
1 L
= φ02 q 2 (1 + 21 ) , (D.5)
2 2

L/2 L/2
E,1 2 dz = φ 2 q 2 sin q,1 z + 1 cos q,1 z 2 dz
0 ,1
−L/2 0
2 1 L
= φ02 q,1 (1 + 21 ) , (D.6)
2 2
where the normalization length L in the z-direction is defined over the region
(−L/2, +L/2) and where the limit L/2 → ∞ will be taken at a later stage of
238 Appendices
the calculation. Evidently, in the limit L/2 → ∞ the exponentially decaying mode
in region 2 makes a negligible contribution to the normalization integrals. Thus, it
follows that

4π h̄ 4 ∂⊥,1 2 ∂,1 2 −1
φ0 = 3
2
q + q , (D.7)
L 1 + 21 ∂ω ∂ω 1,
and that

4π e2 h̄ L −3
HHS =
q q1, (∂/∂ω)(⊥,1 sin2 θ 1 + ,1 cos2 θ1 )
1 2
× # # eiq·ρρ (aq + a−q
†
)
q + q,1 ,1 q,1 + ,2 κ2
2 2 2 2 2 2

,1 q,1 cos q,1 z + ,2 κ2 sin q,1 z z > 0,
× (D.8)
κ
,1 q,1 e 2 z z < 0,
where the summation is taken only over values of q,1 consistent with the condition
(ω),2 (ω)⊥,2 > 0. The case in which an optical phonon mode propagates into
region 1 from region 2 may be treated in the same manner.
The angular dependence of the phonon frequencies in würtzite heterostructures
leads to the situation – in contrast to the typical case for zincblende heterostructures
– where there is an overlap in the allowed phonon frequencies in adjacent material
regions. This situation leads to the occurence of modes that propagate across the
heterointerface. Consider an optical phonon incident on region 2 from region 1. In
this case, the phonon will propagate in region 2 if the phonon frequency in region 1,
ω, is such that (ω),2 (ω)⊥,2 < 0. As before, the allowed frequencies in region 1

2 + (ω)
are the solutions of (ω),1 q,1 ⊥,1 q = 0 or q,1 =
2 −(ω)⊥,1 /(ω),1 q.
If (ω),2 (ω)⊥,2 < 0, there exists a real solution for q,2 which satisfies
(ω),2 q,1
2
+ (ω)⊥,2 q 2 = 0, (D.9)
or

−(ω)⊥,2 (ω),1 (ω)⊥,2
q,2 = q= q,1 . (D.10)
(ω),2 (ω)⊥,1 (ω),2
Writing the phonon potential as

!
Ae−iq,1 z + Beiq,1 z z > 0,
φ(r ) = eiq·ρρ × (D.11)
Ce−iq,2 z z < 0,
it follows that continuity of the tangential component of the electric field at the
heterointerface leads to A + B = C and continuity of the normal component of the
electric displacement at the heterointerface results in the condition
Appendices 239
,1 k1,
(A − B)2 = C where 2 = .
,2 k2,
Eliminating B from these two conditions, it follows that 2A = (1 + −1

2 )C so that
C = 2A2 /(1+ 2 ) and B = A(2 − 1)/(1+ 2 ). Then

 2 − 1 iq1, z
 e−iq1, z + e z > 0,
φ(r ) = φ0 eiq·ρρ × 1 + 2 (D.12)


22 −iq2, z
e z < 0.
1 + 2
Once again, the normalization condition of subsection 7.3.1,

! "
1 h̄ 1 1 ∂(ω)⊥,n 2
1 1 ∂(ω),n 2

= E⊥,n + E,n dz,
L 2 2ω 4π 2ω ∂ω 4π 2ω ∂ω
(D.13)
2 2
may be calculated in terms of E⊥,n and E,n through
L/2 2

E⊥,1 2 dz = φ 2 q 2 1 + 2 − 1 L
,
0
0 1 + 2 2
L/2 2

E,1 2 dz = φ 2 q 2 1 + 2 − 1 L
,
0 1,
0 1 + 2 2
0 2 (D.14)

E⊥,2 2 dz = φ 2 q 2 22 L
,
0
−L/2 1 + 2 2
0
22 2 L
E,2 2 dz = φ 2 q 2 .
0 1,
−L/2 1 + 2 2
These normalization integrals determine the propagating (PR) optical phonon modes
in a single-heterointerface würtzite structure. The corresponding electron–optical-
phonon interaction Hamiltonians are:
$ %1/2
S
HPR = 4π e2 h̄ L −3
q q1,
%−1/2
∂ $
× ⊥,1 (ω)q 2 + ,1 (ω)q1,
2
+ {1 ↔ 2}
∂ω

ρ †
2 cos q1, z z > 0,
× e (aq + a−q )
iq·ρ
(D.15)
2 cos(q2, z) z < 0,
for the quasi-symmetric modes and

240 Appendices
$ %1/2
A
HPR = 4π e2 h̄ L −3
q q1,
−1/2
∂
× {[⊥,1 (ω)q 2 + ,1 (ω)q1,
2 2
],2 2
q2, } + {1 ↔ 2}
∂ω

ρ †
,2 q2, sin q1, z z > 0,
× e (aq + a−q )
iq·ρ
(D.16)
,1 q1, sin q2, z z < 0,
for the quasi-antisymmetric modes (Lee et al., 1998).
D.2 Double-heterointerface uniaxial structures

First consider the interface modes of a uniaxial polar-semiconductor heterostructure
consisting of a layer of material 1, of thickness d, bounded by two semi-infinite
regions of material 2 occupying the regions z < −d/2 and z > d/2. The c-axis
is taken to be normal to the heterointerfaces. In the region |z| < d/2 the dielectric
constants are (ω)⊥(),1 and in the semi-infinite regions |z| > d/2 the dielectric
constants are (ω)⊥(),2 , as depicted in Figure D.1(b). As for the zincblende case
discussed in Chapter 7, the potential φ(r ) is taken to be of the form
 κ (z+d/2)
 Ce 2 z < −d/2,
ρ
φ(r ) = e iq·ρ
× A cosh κ1 z + B sinh κ1 z |z| < d/2, (D.17)

De−κ2 (z−d/2) z > d/2.
In the absence of free charge, ∇ · D = 0 so that the dispersion relations in the

two material regions are
(ω),1 κ12 − (ω)⊥,1 q 2 = 0 (ω),1 (ω)⊥,1 > 0,

(D.18)
(ω),2 κ22 − (ω)⊥,2 q 2 = 0 (ω),2 (ω)⊥,2 > 0,

so it follows that κ1 = (ω)⊥,1 /(ω),1 q and κ2 = (ω)⊥,2 /(ω),2 q. The
continuity of the tangential component of the electric field at the heterointerfaces is
satisfied if
C = A cosh κ1 d/2 − B sinh κ1 d/2 z = −d/2, (D.19)
and
D = A cosh κ1 d/2 + B sinh κ1 d/2 z = d/2. (D.20)
Likewise, the continuity of the normal component of electric displacement at the

heterointerfaces is satisfied if
(ω),2 κ2 C = (ω),1 κ1 (−A sinh κ1 d/2 + B cosh κ1 d/2) z = −d/2,

(D.21)
Appendices 241
and
−(ω),2 κ2 D = (ω),1 κ1 (A sinh κ1 d/2 + B cosh κ1 d/2) z = −d/2.

(D.22)
Eliminating C and D from these equations, it follows that
(ω),2 κ2 (A cosh κ1 d/2 − B sinh κ1 d/2)

= (ω),1 κ1 (−A sinh κ1 d/2 + B cosh κ1 d/2) (D.23)
and
−(ω),2 κ2 (A cosh κ1 d/2 + B sinh κ1 d/2)

= (ω),1 κ1 (A sinh κ1 d/2 + B cosh κ1 d/2). (D.24)
Accordingly, the condition for a non-trivial solution is

[(ω),1 κ1 sinh κ1 d/2 + (ω),2 κ2 cosh κ1 d/2]
= 0. (D.25)
[(ω),1 κ1 cosh κ1 d/2 + (ω),2 κ2 sinh κ1 d/2]
Clearly, two cases are possible. The case where
(ω),1 κ1 tanh κ1 d/2 + (ω),2 κ2 = 0 (D.26)
corresponds to B = 0, with the results that

C φ0
C = D = φ0 and A= = ; (D.27)
cosh κ1 d/2 cosh κ1 d/2
hence, the potential is symmetric and
 κ (z+d/2)
 −e 2 z < −d/2,
ρ
× (cosh κ1 z)/(cosh κ1 d/2) |z| < d/2, (D.28)
 −κ2 (z−d/2)
e z > d/2.
The case where
(ω),1 κ1 coth κ1 d/2 + (ω),2 κ2 = 0 (D.29)
corresponds to A = 0 with the results that

D φ0
−C = D = φ0 and B= = ; (D.30)
sinh κ1 d/2 sinh κ1 d/2
hence, the potential is antisymmetric and
 κ (z+d/2)
 −e 2 z < −d/2,
ρ
× (sinh κ1 z)/(sinh κ1 d/2) |z| < d/2, (D.31)
 −κ2 (z−d/2)
e z > d/2.
242 Appendices

1 h̄ 1 1 ∂(ω)⊥,n 2 1 1 ∂(ω),n 2
= E⊥,n + E,n dz,
2
L 2ω 4π 2ω ∂ω 4π 2ω ∂ω
(D.32)
2
may be calculated in terms of the appropriate integrals of E⊥,n = q 2 |φ(z)|2 and
2
E,n = |∂φ(z)/∂z|2 . For the symmetric case, these integrals are
−d/2
∞ ∞
E⊥,2 2 dz = E⊥,2 2 dz = φ 2 q 2 e−2κ2 (z−d/2) dz
0
−∞ d/2 d/2
q2
= φ02 ,
2κ2

d/2 1 d/2
E⊥,1 2 dz = φ 2 q 2 cosh2 κ1 z dz
0
−d/2 cosh2 κ1 d/2 −d/2

2φ02 q 2 1 κ1 d κ1 d d
= sinh cosh + ,
cosh2 κ1 d/2 2κ1 2 2 4
−d/2 ∞ ∞

E,2 2 dz = E,2 2 dz = φ 2 κ 2 e−2κ2 (z−d/2) dz
0 2
−∞ d/2 d/2
κ2
= φ02 ,
2

d/2 1 d/2
E,1 2 dz = φ 2 κ 2 sinh 2 (κ1 z)dz
0 1
−d/2 cosh (κ1 d/2) −d/2
2

2φ02 κ12 1 κ1 d κ1 d d
= sinh cosh − .
cosh2 κ1 d/2 2κ1 2 2 4
(D.33)
For the antisymmetric case, these integrals are

−d/2 ∞
∞
E⊥,2 2 dz = E⊥,2 2 dz = φ 2 q 2 e−2κ2 (z−d/2) dz
0
−∞ d/2 d/2
q2
= φ02 ,
2κ2
d/2
d/2 1
E⊥,1 2 dz = φ 2 q 2 sinh 2 (κ1 z)dz
0
−d/2 sinh 2 (κ1 d/2) −d/2

2φ02 q 2 1 κ1 d κ1 d d
= sinh cosh − ,
sinh 2 (κ1 d/2) 2κ1 2 2 4
Appendices 243
−d/2
∞ ∞
E,2 2 dz = E,2 2 dz = φ 2 κ 2 e−2κ2 (z−d/2) dz
0 2
−∞ d/2 d/2
κ2
= φ02 ,
2
d/2
d/2 1
E,1 2 dz = φ 2 κ 2 cosh 2 κ1 z dz
0 1
−d/2 sinh 2 (κ1 d/2) −d/2

2φ02 κ12 1 κ1 d κ1 d d
= sinh cosh + .
sinh 2 κ1 d/2 2κ1 2 2 4
(D.34)
It then follows that the carrier–optical-phonon interaction Hamiltonian for the

symmetric modes is
! "−1/2
∂ √ √
S
HIF = 4π e2 h̄ L −2 [ ⊥,1 ,1 tanh( ⊥,1 /,1 qd/2) − ⊥,2 ,2 ]
q ∂ω


 cosh( ⊥,1 /,1 qz)
1  |z| < d/2,
× √ eiq·ρρ (aq + a−q †
) cosh( ⊥,1 /,1 qd/2) (D.35)

2q  −√⊥,2 /,2 q(|z|−d/2)

e |z| > d/2,
where the frequency is the solution of the transcendental equation
√ √
⊥,1 ,1 tanh( ⊥,1 /,1 qd/2) − ⊥,2 ,2 = 0 (D.36)
with the range of frequencies determined by ,1 ,2 < 0 and, as discussed
previously, ⊥,1 ,1 > 0 and ⊥,2 ,2 > 0. For the antisymmetric mode,
! "−1/2
−2 ∂ √ √
HIF =
A 2
4π e h̄ L [ ⊥,1 ,1 coth( ⊥,1 /,1 qd/2) − ⊥,2 ,2 ]
q ∂ω
1
× √ eiq·ρρ (aq + a−q†
)
2q


 sinh( ⊥,1 /,1 qz)
 |z| < d/2,
× sinh( ⊥,1 /,1 qd/2) (D.37)

 √

sgn(z) e− ⊥,2 /,2 q(|z|−d/2) |z| > d/2,
where the frequency is determined by
√ √
⊥,1 ,1 coth( ⊥,1 /,1 qd/2) − ⊥,2 ,2 = 0, with ,1 ,2 < 0.
The dispersion relations for these interface modes are displayed in Figure D.2 for a
würtzite AlN/GaN/AlN heterostructure having a quantum well of thickness d.
244 Appendices
For the symmetric confined modes in the double-heterointerface structure the

potential, φ(r ), is taken to be of the form
 κ (z+d/2)
 cos q,1 d/2 e 2 z < −d/2,
ρ
× cos q,1 z |z| < d/2, (D.38)

cos q,1 d/2 e−κ2 (z−d/2) z > d/2.
In the absence of free charge, ∇ · D = 0 so that the dispersion relations in the
two material regions are
(ω),1 q,1
2
− (ω)⊥,1 q 2 = 0 (ω),1 (ω)⊥,1 < 0 for |z| < d/2,
(ω),2 q,2
2
− (ω)⊥,2 q 2 = 0 (ω),2 (ω)⊥,2 > 0 for |z| > d/2.
(D.39)
Clearly, the continuity of the tangential component of the electric field at the
heterointerfaces is ensured by the choice of φ(r ). The continuity of the normal com-
ponent of the electric displacement at the heterointerfaces leads to the requirement
that
(ω),1 q,1 sin q1, d/2 − (ω),2 κ2 cos q,1 d/2 = 0. (D.40)
Figure D.2. Interface phonon dispersion relations for an AlN/GaN/AlN

double-interface heterostructure with interfaces separated by a distance d. From Lee
Appendices 245

1 h̄ 1 1 ∂(ω)⊥,n 2 1 1 ∂(ω),n 2
= E⊥,n + E,n dz,
L 2 2ω 4π 2ω ∂ω 4π 2ω ∂ω
(D.41)
2
2
E,n = |∂φ(z)/∂z|2 . For the symmetric case, these integrals are
2
+∞ ∂(ω)
⊥,n E⊥,n 2
n=1 −∞
∂ω

∂(ω)⊥,1 d 1 q,1 d q,1 d
= q 2 φ02 + sin cos
∂ω 2 q,1 2 2

1 ∂(ω)⊥,2 q,1 d
+ q 2 φ02 cos2 (D.42)
κ2 ∂ω 2
and
2
+∞ ∂(ω)
,n E⊥,n 2
n=1 −∞
∂ω

∂(ω),1 d 1 q,1 d q,1 d
= φ02 − sin cos
∂ω 2 q,1 2 2

∂(ω),2 q ,1 d
+ φ02 κ2 cos2 ; (D.43)
∂ω 2
thus
!
1 h̄ 1 1 2 ∂(ω)⊥,1 2 ∂(ω),1 2 d
= φ q + q,1
L 2 2ω 4π 2ω 0 ∂ω ∂ω 2
/ 0
∂(ω)⊥,1 q 2 ∂(ω),1 q,1 d q,1 d
+ − q,1 sin cos
∂ω q,1 ∂ω 2 2
/ 0
∂(ω)⊥,2 q 2 ∂(ω),2 q1, d
+ + κ2 cos2
∂ω κ2 ∂ω 2
! "
1 1 2 ∂(ω)⊥,1 2 ∂(ω),1 2 d ∂ fs q,1 d
= φ q + q,1 − 2q cos
4π 2ω 0 ∂ω ∂ω 2 ∂ω 2
! 2 3 "
1 1 2 ∂ d ∂ fs q,1 d
= φ (ω)⊥,1 q + (ω),1 q,1
2 2
− 2q cos ,
4π 2ω 0 ∂ω 2 ∂ω 2
(D.44)
where f s (ω) is defined by

q,1 d
f s (ω) = sgn[(ω),1 ] −(ω)⊥,1 (ω),1 sin
2
q1, d
− sgn[(ω),2 ] (ω)⊥,2 (ω),2 cos . (D.45)
2
246 Appendices
Thus,
! "
∂ 2 3d ∂ fs q,1 d −1
φ02 = 4π h̄ L −2 (ω)⊥,1 q 2 + (ω),1 q,1
2
− 2q cos
∂ω 2 ∂ω 2
(D.46)
and
!
∂ 2 3d
HCS = eiq·ρρ (aq + a−q
†
)4π h̄ L −2 (ω)⊥,1 q 2 + (ω),1 qm,1
2
q m ∂ω 2
"
∂ qm,1 d −1/2
× −2q f s (ω) cos
∂ω 2

cos qm,1 z |z| < d/2,
× (D.47)
−κ
cos qm,1 d/2 e 2 (z−d/2) z > d/2,
where the discrete confined wavevectors q1,m are determined from
(ω),1 qm,1 sin qm,1 d/2 − (ω),2 κ2 cos qm,1 d/2 = 0,
f s (ω) is evaluated at the discrete values of q,1 , qm,1 , and
2mπ/d < qm,1 < 2(m + 1)π/d. (D.48)

As defined previously κ2 = (ω)⊥,2 /(ω),2 q. Through an analogous derivation,
it follows for the antisymmetric modes that

HCA = eiq·ρρ (aq + a−q
†
)
q m
!
∂ 2 3d
× 4π h̄ L −2 (ω)⊥,1 q 2 + (ω),1 qm,1
2
∂ω 2
"−1/2
∂ qm,1 d
− 2q f a (ω) sin
∂ω 2

sin q1,m z |z| < d/2,
× (D.49)
sgn(z) sin qm,1 d/2 e−κ2 (z−d/2) z > d/2,
where
q,1 d
f a (ω) = sgn[(ω),1 ] −(ω)⊥,1 (ω),1 cos
2
q,1 d
+ sgn[(ω),2 ] (ω)⊥,2 (ω),2 sin (D.50)
2
and where the discrete confined wavevectors q1,m are determined from
(ω),1 qm,1 cos qm,1 d/2 + (ω),2 κ2 sin qm,1 d/2 = 0 (D.51)

with 2(m − 1)π/d < qm,1 < 2(m + 1)π/d and κ2 = (ω)⊥,2 /(ω),2 q.
Appendices 247
The half-space modes in a double-heterointerface structure may be considered

usefully in terms of symmetric and antisymmetric modes since such a heterostruc-
ture is symmetric about z = 0. For the symmetric modes take
φ(r ) = eiq·ρρ


 A cos[q,1 (z + d/2)] − B sin[q,1 (z + d/2)] z < −d/2,
× C cosh κ1 z |z| < d/2, (D.52)

 A cos[q (z − d/2)] + B sin[q (z − d/2)]
,1 ,1 z > d/2.
Then, continuity of the tangential component of the electric field at the heteroin-
terfaces yields A = C cosh κ1 d/2 and continuity of the normal component of
the electric displacement at the heterointerfaces requires that (ω),2 q,2 B =
(ω),1 κ2 C sinh κ1 d/2. Since ∇ ·D = 0 it follows that (ω),2 q,2
2 +(ω)
⊥,2 q = 0.
2
Defining 3 = (ω),2 q,2 /(ω),1 κ1 it follows that A = B3 coth κ1 d/2. Then,
with B ≡ φ0 ,
φ(r ) = φ0 eiq·ρρ


 − sin[q,1 (z + d/2)] + 3 coth κ1 d/2 cos[q,1 (z + d/2)] z < d/2,
× 3 (cosh κ1 z)/(sinh κ1 d/2) |z| < d/2,

 sin[q (z − d/2)] + coth κ d/2 cos[q (z − d/2)]
,1 3 1 ,1 z > d/2,
(D.53)
or

 sin[q,1 (|z| − d/2)]
φ(r ) = φ0 eiq·ρρ × + 3 coth κ1 d/2 cos[q,1 (|z| − d/2)] |z| > −d/2,

3 (cosh κ1 z)/(sinh κ1 d/2) |z| < d/2.
(D.54)
For the antisymmetric modes take

 −A cos[q,1 (z + d/2)] + B sin[q,1 (z + d/2)] z < −d/2,
φ(r ) = eiq·ρρ × C sinh κ1 z |z| < d/2,

A cos[q,1 (z − d/2)] + B sin[q,1 (z − d/2)] z > d/2.
(D.55)
Then, continuity of the tangential component of the electric field at the heteroin-
terfaces yields A = C sinh κ1 d/2 and continuity of the normal component of
the electric displacement at the heterointerfaces requires that (ω),2 q,2 B =
(ω),1 κ1 C cosh κ1 d/2. Thus, it follows that A = B3 tanh κ1 d/2. Then, with
B ≡ φ0 ,


 sin[q,2 (z + d/2)] − 3 tanh κ1 d/2 cos[q,2 (z + d/2)] z < −d/2,
× 3 sinh κ1 z/(cosh κ1 d/2) |z| < d/2,

 sin[q (z − d/2)] + tanh κ d/2 cos[q (z − d/2)]
,2 3 1 ,2 z > d/2,
(D.56)
248 Appendices
or


 sgn(z) sin[q,2 (|z| − d/2)]
× + 3 tanh κ1 d/2 cos[q,2 (|z| − d/2)] |z| > −d/2, (D.57)

 (sinh κ z)/(cosh κ d/2)
3 1 1 |z| < d/2.

1 h̄ 1 1 ∂(ω)⊥,n 2
+ 1 1 ∂(ω),n E,n 2 dz,

= E ⊥,n
L 2 2ω 4π 2ω ∂ω 4π 2ω ∂ω
(D.58)
2

E,n 2 = |∂φ(z)/∂z|2 . For the symmetric case, the integrals that make contributions
in the limit L −→ ∞ are
L/2 $ %

E⊥,2 2 dz = φ 2 q 2 1 1 + 2 coth2 κ1 d/2) L ,
0 3
d/2 2 2
L/2 (D.59)
2 3
E,2 2 dz = φ 2 q 2 1 1 + 2 coth2 (κ1 d/2) L .
0 ,2 3
d/2 2 2
For the antisymmetric case, the integrals that make contributions in the limit L →
∞ are
−d/2 L/2 $ %

E⊥,2 2 dz = E⊥,2 2 dz = φ 2 q 2 1 1 + 2 tanh 2 κ1 d/2 L ,
0 3
−L/2 d/2 2 2
−d/2 L/2 $ %

E,2 2 dz = E,2 2 dz = φ 2 q 2 1 1 + 2 tanh 2 κ1 d/2 L .
0 ,2 3
−L/2 d/2 2 2
(D.60)
Thus

4π h̄ ∂(ω)⊥,2 2 ∂(ω)⊥,2 2 −1
φ02 = 3 ×2 q + q,2
L ∂ω ∂ω

(1 + 23 coth2 κ1 d/2)−1 symmetric case,
× (D.61)
(1 + 23 tanh2 κ1 d/2)−1 antisymmetric case.
The interaction Hamiltonian for the symmetric HS modes is then
−1/2
−3 ∂
HHS =
S 2
4π e h̄ L (⊥,2 sin θ2 + ,2 cos θ2 )
2 2
eiq·ρρ (aq + a−q
†
)
q q,2 ∂ω
√
2 −1/2 2
−1/2
× 2 q 2 + q,2 ,1 κ12 sinh2 κ1 d/2 + ,2
2 2
q,2 cosh2 κ1 d/2


 {,1 κ1 sinh κ1 d/2 sin[q,2 (|z| − d/2)]
× + ,2 q,2 cosh κ1 d/2 cos[q,2 (|z| − d/2)]} |z| > d/2, (D.62)

 q cosh κ z
,2 ,2 1 |z| < d/2,
Appendices 249
where the sum is over q,2 and is taken only over those values (D.62) such that
(ω),1 (ω)⊥,1 > 0. Likewise, the interaction Hamiltonian for the antisymmetric
HS modes is
−1/2
∂
HHS =
A 2
4π e h̄ L −3
(⊥,2 sin θ 2 + ,2 cos θ2 )
2 2
eiq·ρρ (aq + a−q
†
)
q q,2 ∂ω
√ 2 −1/2 2
× 2(q 2 + q,2 ) (,1 κ12 cosh2 κ1 d/2 + ,2
2 2
q,2 sinh2 κ1 d/2)−1/2

 sgn(z) ,1 κ1 cosh κ1 d/2 sin[q,2 (|z| − d/2)]

× + ,2 q,2 sinh κ1 d/2 cos[q,2 (|z| − d/2)] |z| > d/2,

 (D.63)
,2 q,2 sinh κ1 z |z| < d/2.
The interaction Hamiltonians for the propagating (PR) modes may be obtained
straightforwardly from those of the HS modes by making the substitution κ1 →
iq,1 . Indeed, this substitution results in 3 = −i3 with
(ω),2 q,2
3 = ,
(ω),1 q,1
cosh κ1 z → cos q,1 z,
sinh κ1 z → i sin q,1 z, (D.64)
coth κ1 d/2 → −i cot q,1 d/2,
tanh κ1 d/2 → −i tan q,1 d/2.
The interaction Hamiltonian for the symmetric PR modes is then
−1/2
∂
S
HPR = 4π e2 h̄ L −3 (⊥,2 q 2 + ,2 q,2
2
) eiq·ρρ (aq + a−q
†
)
q q,2 ∂ω
√ 2 2 −1/2
× 2 ,1 q,1 sin2 q,1 d/2 + ,2
2 2
q,2 cos2 q,1 d/2

 ,2 q,2 cos q,1 d/2 cos[q,2 (|z| − d/2)]


× − ,1 q,1 sin q,1 d/2 sin[q,2 (|z| − d/2)] |z| > d/2,

 (D.65)
,2 q,2 cos q,1 z |z| < d/2,
where the sum is over q,2 and is taken only over these values such that
(ω),1 (ω)⊥,1 < 0. Likewise, the interaction Hamiltonian for the antisymmetric
PR modes is
−1/2
∂
A
HPR = 4π e2 h̄ L −3 (⊥,2 q 2 + ,2 q,2
2
) eiq·ρρ (aq + a−q
†
)
q q,2 ∂ω
√ 2 2 −1/2
× 2 ,1 q,1 cos2 q,1 d/2 + ,22 2
q,2 sin2 q,1 d/2


 sgn(z) ,2 q,2 sin q,1 d/2 cos[q,2 (|z| − d/2)]
× + ,1 q,1 cos q,1 d/2 sin[q,2 (|z| − d/2)] |z| > d/2,

 q sin q z (D.66)
,2 ,2 ,1 |z| < d/2.
250 Appendices
Appendix E: Fermi golden rule

The Fermi golden rule provides a convenient perturbative result for estimating
carrier–phonon scattering rates. In calculating these rates in solids it is in fact usually
possible to treat the carrier–phonon interactions as perturbations to the Hamiltonian
describing the system in the absence of phonon interactions. Taking the unperturbed
system to be described by the Hamiltonian H and the carrier–phonon interaction
Hamiltonian to be Hc−p (t), the wavefunction of the system, ψ(x, t), obeys
h̄ ∂ψ(x, t)
[H + Hc−p (t)]ψ(x, t) = − , (E.1)
i ∂t
where a one-dimensional system is considered for the sake of simplicity. The final
result for the Fermi golden rule will be independent of the assumption that the
system is one dimensional. Taking the wavefunction to be of the form

ψ(x, t) = cn (t)φn (x) e−i E n t/h̄ , (E.2)
n
where φn (x) is the eigenstate of the unperturbed system with quantum number n, it
follows that,

[H + Hc−p (t)]ψ(x, t) = cn (t)E n φn (x) e−i E n t/h̄
n

+ cn (t) e−i E n t/h̄ Hc−p (t)φn (x) (E.3)
n
and
h̄ ∂ψ(x, t) h̄ dcn (t)
− =− φn (x) e−i E n t/h̄
i ∂t i n dt

+ cn (t)E n φn (x) e−i E n t/h̄ . (E.4)
n
Thus, without approximation,

h̄ dcn (t)
− φn (x) e−i E n t/h̄ = cn (t) e−i E n t/h̄ Hc−p (t)φn (x). (E.5)
i n dt n
Since the spectrum of eigenstates φn for a quantum mechanical system is complete,

φm |φn = δmn . Thus, upon multiplying both sides of the last equation by φm∗ and
integrating over x, it follows that
dcm (t) 5 7
i h̄ = φm Hc−p (t) φn e−i(E n −E m )t/h̄ cn (t). (E.6)
dt n
Now if it is assumed that the system is initially in state n, so that cn (0) = 1 and all
other cm (0) are zero, the time evolution of state m is approximated by
Appendices 251

i t 5 7
cm (t) − dt φm Hc−p (t) φn e−i(E n −E m )t/h̄ . (E.7)
h̄ 0
In the case where Hc−p (t) is a step function in time,
5 7 e−i(E n −E m )t/h̄ − 1
cm (t) = φm Hc−p φn (E.8)
En − Em
or
5 72 sin2 [(E n − E m )t/(2h̄)]
|cm (t)|2 = φm Hc−p φn . (E.9)
[(E n − E m )/2]2
In the case of nearly degenerate states E n E m , and where the transition from
state n is to a dense group of final states centered around E m , the probability of a
transition, |cm (t)|2 , takes the form
E n +E/2
5 72 sin2 [(E n − E)t/(2h̄)]
P= d Eρ(E) φm Hc−p φn
E n −E/2 [(E n − E)/2]2
E n +E/2
5 72 sin2 [(E n − E)t/(2h̄)]
4ρ(E) φm Hc−p φn dE , (E.10)
E n −E/2 (E n − E)2
5 72
where the last result follows since ρ(E) φm Hc−p φn varies slowly with energy
relative to [sin2 (E n − E)t/(2h̄)]/(E n − E)2 . Taking x = (E − E n )t/(2h̄),

2t 5 72 tE/4h̄ sin2 x
P = ρ(E) φm Hc−p φn dx . (E.11)
h̄ −tE/4h̄ x2
In the limit of large t,

2π 5 72
P→ ρ(E) φm Hc−p φn t; (E.12)
h̄
the transition rate is given by
dP 2π 5 72
= ρ(E) φm Hc−p φn . (E.13)
dt h̄
This result is one of the expressions known as the Fermi golden rule.
An alternative form of the Fermi golden rule is obtained by returning to the case
of a transition between two specific states (n and m), by taking the limit of large t,
2 2h̄
t = (E.14)
ωmn Em − En
and using the relation
sin2 ωmn t/2 2π h̄

→ 2π δ(ωmn t) = δ(h̄ωmn ) (E.15)
(ωmn t/2) 2 t
252 Appendices
for t 2/ωmn ; it then follows that for t 2/ωmn

|cm (t)|2 2π 5 72
= φm Hc−p φn δ(h̄ωmn )
t h̄

2π 5 72
= φm Hc−p φn δ(E m − E n ). (E.16)
h̄
In the case where Hc−p is associated with a phonon field and thus has a harmonic
time dependence, e±iωphonon t , the derivation of the last result is modified by replacing
ωmn by ωmn ∓ωphonon , the upper sign corresponds to phonon emission and the lower
sign is associated with phonon absorption; thus,

|cm (t)|2 2π 5 72
P= = φm Hc−p φn δ(E m − E n ∓ h̄ωphonon ). (E.17)
t h̄
This form of the Fermi golden rule is used throughout this book. Although the
derivation of the Fermi golden rule is based on the assumption that t 2/ωmn ,
there are many instances for nanoscale devices where this condition is not met.
Indeed, for nanoscale devices, transit times are frequently 0.1 ps or less and energy
differences are a few to tens of meV so that 2/ωmn 0.1 ps. Hence, t ≈ 2/ωmn
and t 2/ωmn is not satisfied. Nevertheless, the Fermi golden rule is used
routinely in such situations and there are few cases where its failure to yield
a reasonable approximation has been demonstrated convincingly. Moreover, the
alternative means of making non-perturbative calculations such as those based on
the Feynman path integral (Register, et al., 1988; Komirenko, et al., 2000c) are so
computationally intense that they are rarely used in practice.
Appendix F: Screening effects in a two-dimensional

electron gas
In this appendix, the role of screening is examined by the Lindhart method for
the case of a two-dimensional electron gas (Bastard, 1988). The phonon field is
taken to give rise to a potential phonon (r). In general, this potential will induce
a charge distribution ρ induced (r), which produces a potential, induced (r). The
total electrostatic potential is then the sum of that produced by the phonon field,
phonon (r), and that associated with the induced charge density, induced (r):
total (r) = phonon (r) + induced (r). (F.1)
Thus, the total interaction potential energy, V̄ (r), is the sum of that associated
with the phonons in the absence of the Coulomb screening, −ephonon (r), and that
induced by the Coulomb screening effects, −einduced (r):
V̄ (r) = −e[phonon (r) + induced (r)] = −etotal (r). (F.2)

Appendices 253
Taking this interaction to be a perturbation to the Hamiltonian of the carrier–phonon

system without phonon interactions or screening, H0 , it follows from Appendix E
that
dci (t) 5 7
i h̄ = φi V̄ φk ck (t) e−i(Ei −E k −h̄ω)t/h̄ eγ t , (F.3)
dt k
where the time-dependent carrier–phonon interaction with screening effects in-

cluded is taken as
Hc−p (t) = V̄ (r) e−iωt+γ t . (F.4)
As in Appendix E, the system is assumed to be initially in state |φk ; then

dci (t) 5 7 −i(Ei −E k −h̄ω)t/h̄ γ t
i h̄ = φi V̄ φk e e , (F.5)
dt
and

i 5 7 t
ci (t) − φi V̄ φk dt e−i(Ei −E k −h̄ω)t/h̄ eγ t
h̄ −∞
5 7
φi V̄ φk
=− e−i(Ei −E k −h̄ω)t/h̄ . (F.6)
E i − E k − h̄ω − iγ
The unperturbed wavefunction is taken to be that of a two-dimensional electron gas
in the lowest subband,
1
φk (r, t) = √ χ1 (z) eik·ρρ −i E k t/h̄ , (F.7)
S
where S is the area of the two-dimensional electron gas, χ1 (z) describes the
dependence of the wavefunction normal to the x y-plane of translational invariance,
and eik·ρρ is the wavefunction for a plane wave in the x y-plane with ρ = (x, y) and
k = (k x , k y ). The perturbed wavefunction is then given by
1
k (r, t) = √ χ1 (z) eik·ρρ −i E k t/h̄
S
1 φk |V |φk eik ·ρρ
+ √ χ1 (z) e−i(E k +h̄ω)t/h̄
S E k − E k + h̄ω + iγ
k

1 φk |V ∗ | φk eik ·ρρ
+ √ χ1 (z) e−i(E k −h̄ω)t/h̄ , (F.8)
S E k − E k − h̄ω + iγ
k
where
V̄ = V e−iωt+γ t + V ∗ eiωt+γ t ,

φk |V |φk ≡ d 2ρ e−ik ·ρρ V eik·ρρ , (F.9)
5 ∗ 7∗
φk V φk ≡ φk |V |φk .
254 Appendices
The induced charge density is then proportional to the difference of the probability
densities of the perturbed and unperturbed wavefunctions, multiplied by the Fermi–
Dirac distribution function Fk :
! 1
"
ρ induced
(r) = −2e |φk (r, t)| − |χ1 (z)| Fk ,
2 2
(F.10)
k
S
where −e is the charge of a single carrier and the factor of two is introduced since
there are two spin states for every electronic quantum state. Thus,
5 7
χ 2 (z) k |V |k ei(k −k)·ρρ e−iωt
ρ induced
(r) = −2e 1
k
S k
E k − E k + h̄ω + iγ
5 7
k |V |k ∗ e−i(k −k)·ρρ eiωt
+
k
E k − E k + h̄ω − iγ
5 ∗ 7 i(k −k)·ρρ iωt
k |V | k e e
+
k
E k − E k − h̄ω + iγ
5 ∗ 7∗ −i(k −k)·ρρ −iωt
k |V | k e e
+ Fk
k
E k − E k − h̄ω − iγ
5 7
χ12 (z) k |V |k ei(k −k)·ρρ
= −2e
S k,k E k − E k + h̄ω + iγ
5 ∗ 7∗ −i(k −k)·ρρ
k |V | k e
+ e−iωt Fk + c.c.
E k − E k − h̄ω − iγ
χ12 (z) 5 7 Fk − Fk
= −2e k |V |k ei(k −k)·ρρ . (F.11)
5 7
Defining q = k −k, noting that the matrix element k |V |k depends on the modulus
5 7
of k − k, and defining V (|k − k|) ≡ k |V |k , it follows that
χ12 (z) Fk − Fk

ρ induced (r) = −2e V ( k − k) ei(k −k)·ρρ
(F.12)
or, alternatively,
χ12 (z) Fk+q − Fk

ρ induced (r) = −2e V (|q|) eiq·ρρ
S k q E k − E k+q + h̄ωq + iγ

= eχ12 (z) V (|q|) eiq·ρρ A(q, ω), (F.13)
q
where
Appendices 255
2 Fk+q − Fk
A(q, ω) ≡ − . (F.14)
S k E k − E k+q + h̄ωq + iγ
A(q, ω) represents the polarization of the two-dimensional electron gas. More-

over, ρ induced (r) is equivalent to the two-dimensional Fourier expansion of
ρ induced (q, z) = eχ12 (z)V (|q|)e A(q, ω). For the potential we have

dρρ dz ρ induced (r )
induced
(r) = 4π
ρ − ρ )2 + (z − z )2
(ρ

= induced (q, z) eiq·ρρ . (F.15)
q
Then
6 4 6 4
induced (q) = −A(q, ω) total (q) C(q), (F.16)
where the expectation values are of the form

6 4
induced (q) = dz χ12 (z)induced (q, z), (F.17)
and

dρρ eiq·ρρ
C(q) = 4πe dzdz χ12 (z)χ12 (z ) . (F.18)
ρ 2 + (z − z )2
The identity
1 2π 1 iq·(ρ −ρ ) −q |z−z |

= e e , (F.19)
|r − r | S q q
may be used to write C(q) as follows:

1
C(q) = 8π 2 e G(q), (F.20)
q
where

dz dz χ12 (z)χ12 (z ) e−q |z−z | .

G(q) = (F.21)
5 7 5 7 5 7
Noting that induced (q) = total (q) − phonon (q) , it follows that
6 4 6 4 6 4
total (q) − phonon (q) = −A(q, ω) total (q) C(q), (F.22)
so that
6 4 6 4
total (q) = phonon (q) /[1 − A(q, ω)C(q)]
6 4
= phonon (q) /κel (q), (F.23)
256 Appendices
where the electron permittivity, κel (q), is given by
κel (q) = 1 − A(q, ω)C(q)

8π 2 e
=1− G(q)A(q, ω). (F.24)
q
In Section 10.6, this expression for the electron permittivity, κel (q), is used to take
into account the effect of electron screening.
References
Alexson, D., Bergman, L., Dutta, M., Kim, K.W., Komirenko, S., Nemanich, R.J., Lee, B.C.,
Stroscio, M.A., and Yu, S. (1999), Confined phonons and phonon-mode properties of
III-V nitrides with würtzite crystal structure, Physics, B263 and B264, 510–513.
Alexson, D., Bergman, L., Nemanich, R.J., Dutta, M., Stroscio, M.A., Parker, C.A., Bedair,
S.M., El-Masry, N.A., and Adar, F. (2000), UV Raman study of A1 (LO) and E2 phonons
in Inx Ga1−x N alloys, Applied Physics Letters, 89, 798–800.
Auld, B.A. (1973), Acoustic Fields and Waves in Solids. Wiley-Interscience, John Wiley &
Sons, New York.
Azuhata, T., Sota, T., Suzuki, K., and Nakamura, S. (1995), Polarized Raman spectra in GaN,
Journal of the Physics of Condensed Matter, 7, L129–L135.
Bannov, N., Mitin, V., and Stroscio, M. (1994a), Confined acoustic phonons in semiconductor
slabs and their interactions with electrons, Physica Status Solidi B, 183, 131–138.
Bannov, N., Mitin, V., Stroscio, M.A., and Aristov, V. (1994b), Confined acoustic phonons:
density of states, in Proceedings of the 185th Meeting of the Electrochemical Society, Vol.
94-1, 521; San Francisco, California, May 22–27, 1994.
Bannov, N., Aristov, V., Mitin, V., and Stroscio, M.A. (1995), Electron relaxation times due to
deformation-potential interaction of electrons with confined acoustic phonons in a
free-standing quantum well, Physical Review, B51, 9930–9938.
Bastard, G. (1988), Wave Mechanics Applied to Semiconductor Heterostructures. Halsted
Press, New York.
Behr, D., Niebuhr, R., Wagnar, J., Bachem, K.H., and Kaufmann, U. (1997), Resonant Raman
scattering in GaN/(AlGa)N single quantum wells, Applied Physics Letters, 70, 363–365.
Belenky, G., Dutta, M., Gorfinkel, V.B., Haddad, G.I., Iafrate, G.J., Kim, K.W., Kisin, M.,
Luryi, S., Stroscio, M.A., Sun, J.P., Teng, H.B., and Yu, S. (1999), Tailoring of optical
modes in nanoscale semiconductor structures: role of interface-optical phonons in
quantum-well lasers, Physica B, 263–264, 462–465.
257
258 References
Beltzer, A.I. (1988), Acoustics of Solids. Springer-Verlag, Berlin, Heidelberg, New York,
London, Paris, and Tokyo.
Bergman, L., Alexson, D., Murphy, P.L., Nemanich, R.J., Dutta, M., Stroscio, M.A., Balkas,
C., Shih, H., and Davis, R.F. (1999a), Raman analysis of phonon lifetimes in AlN and
GaN of würtzite structure, Physical Review, B59, 12 977–12 982.
Bergman, L., Dutta, M., Balkas, C., Davis, R.F., Christman, J.A., Alexon, D., and Nemanich,
R.J. (1999b), Raman analysis of the E1 and A1 quasi-longitudinal optical and
quasi-transverse optical modes in würtzite AlN, Journal of Applied Physics, 85,
3535–3539.
Bergman, L., Dutta, M., Kim, K.W., Klemens, P.G., Komirenko, S., and Stroscio, M.A. (2000),
Phonons, electron–phonon interactions, and phonon–phonon interactions in III-V nitrides,
Proceedings of the SPIE, Vol. 3914, 13–22.
Bhatt, A.R., Kim, K.W., Stroscio, M.A., Iafrate, G.J., Dutta, M., Grubin, H.L., Haque, R., and
Zhu, X.T. (1993a), Reduction of LO-phonon interface modes using metal–semiconductor
heterostructures, Journal of Applied Physics, 73, 2338–2342.
Bhatt, A.R., Kim, K.W., Stroscio, M.A., and Higman, J.M. (1993b), Simplified microscopic
model for electron–optical-phonon interactions in quantum wells, Physical Review, B48,
14 671–14 674.
Bhatt, A.R., Kim, K.W., and Stroscio, M.A. (1994), Theoretical calculation of the
longitudinal-optical lifetime in GaAs, Journal of Applied Physics, 76, 3905–3907.
Blakemore, J.S. (1985), Solid State Physics, 2nd edition. Cambridge University Press,
Cambridge.
Blatt, J.M. and Thompson, C.J. (1963), Shape resonances in superconducting films, Physical
Review Letters, 10, 332–334.
Borer, W.J., Mitra, S.S., and Namjoshi, K.V. (1971), Line shape and temperature dependence of
the first order Raman spectrum of diamond, Solid State Communications, 9, 1377–1381.
Born, M. and Huang, K. (1954), Dynamical Theory of Crystal Lattices. Oxford University
Press, Oxford.
Burns, G., Dacol, F., Marinace, J.C., and Scott, B.A. (1973), Raman scattering in thin-film
waveguides, Applied Physics Letters, 22, 356–357.
Campos, V.B., Das Sarma, S., and Stroscio, M.A. (1992), Hot carrier relaxation in
polar-semiconductor quantum wires: confined LO-phonon emission, Physical Review,
B46, 3849–3853.
Cardona, M. (1975), ed. Light Scattering in Solids. Springer-Verlag, New York.
Cardona, M. and Güntherodt, G., eds. (1982a), Light Scattering in Solids II. Springer-Verlag,
Berlin.
Cardona, M. and Güntherodt, G., eds. (1982b), Light Scattering in Solids III. Springer-Verlag,
Berlin.
Cardona, M. and Güntherodt, G., eds. (1984), Light Scattering in Solids IV. Springer-Verlag,
Berlin.
Cardona, M. and Güntherodt, G., eds. (1989), Light Scattering in Solids V. Springer-Verlag,
Berlin.
Cardona, M. and Güntherodt, G., eds. (1991), Light Scattering in Solids VI. Springer-Verlag,
New York.
References 259
Castro, G.R. and Cardona, M., eds. (1987), Lectures on Surface Science, Proceedings of the
Fourth Latin-American Symposium. Springer-Verlag, Berlin.
Chang, I.F. and Mitra, S.S. (1968), Application of modified random-element-isodisplacement
model to long-wavelength optic phonons in mixed crystals, Physical Review, 172,
924–933.
Cheng, T.K., Vidal, J., Zeiger, H.J., Dresselhaus, G., Dresselhaus, M.S., and Ippen, E.P. (1991),
Mechanism for displacive excitation of coherent phonons in Sb, Bi, Te, and Ti2 O3 ,
Applied Physics Letters, 59, 1923–1925.
Cho, G.C., Kutt, W., and Kurz, H. (1990), Subpicosecond time-resolved coherent-phonon
oscillations in GaAs, Physical Review Letters, 65, 764–766.
Choi, K.-K., Newman, P.G., and Iafrate, G.J. (1990), Quantum transport and phonon emission
of nonequilibrium hot electrons, Physical Review, B41, 10 250–10 253.
Choi, K.-K., Tidrow, M.Z., Chang, W.H. (1996), Electron energy relaxation in infrared
hot-electron transistors, Applied Physics Letters, 68, 358–360.
Cingolani, A., Ferrara, M., Lugara, M., and Scamarcio, G. (1986), First order Raman scattering
in GaN, Solid State Communications, 58, 823–824.
Cleland, A.N. and Roukes, M.L. (1996), Fabrication of high frequency nanometer scale
mechanical resonators from bulk Si crystals, Applied Physics Letters, 69, 2653–2655.
Colvard, C., Merlin, R., Klein, M.V., and Gossard, A.C. (1980), Observation of folded acoustic
phonons in semiconductor superlattices, Physical Review Letters, 45, 298–301.
Comas, F., Trallero-Giner, C., and Cardona, M. (1997), Continuum treatment of phonon
polaritons in semiconductor heterogeneous structures, Physical Review, B56, 4115–4127.
Constantinou, N.C. (1993), Interface optical phonons near perfectly conducting boundaries and
their coupling to electrons, Physical Review, B48, 11 931–11 935.
Cowley, R.A. (1963), Advances in Physics, 12, 421–480.
Cros, A., Angerer, H., Ambacher, O., Stutzmann, M., Hopler, R., and Metzger, T. (1997),
Raman study of optical phonons in Alx Ga1−x N alloys, Solid State Communications, 104,
35–39.
Das Sarma, S., Stroscio, M.A., and Kim, K.W. (1992), Confined phonon modes and
hot-electron relaxation in semiconductor microstructures, Semiconductor Science and
Technology, 7, B60–B66.
de la Cruz, R.M., Teitsworth, S.W., and Stroscio, M.A. (1993), Phonon bottleneck effects for
confined longitudinal optical phonons in quantum boxes, Superlattices and
Microstructures, 13, 481–486.
de la Cruz, R.M., Teitsworth, S.W., and Stroscio, M.A. (1995), Interface phonons in spherical
GaAs/Alx Ga1−x As quantum dots, Physical Review, B52, 1489–1492.
Debernardi, A. (1998), Phonon linewidth in III-V semiconductors from density-functional
perturbation theory, Physical Review, B57, 12 847–12 858.
Demangeot, F., Frandon, J., Renucci, M.A., Meny, C., Briot, O., and Aulombard, R.L. (1997),
Interplay of electrons and phonons in heavily doped GaN epilayers, Journal of Applied
Physics, 82, 1305–1309.
Demangeot, F., Groenen, J., Frandon, J., Renucci, M.A., Briot, O., Clur, S., and Aulombard,
R.L. (1998), Coupling of GaN- and AlN-like longitudinal optic phonons in Ga1−x Alx N
solid solutions, Applied Physics Letters, 72, 2674–2676.
260 References
Di Bartolo, B. (1969), Optical Interactions in Solids. John Wiley and Sons, New York.
Dutta, M. and Stroscio, M.A., eds. (1998), Quantum-based Electronic Devices and Systems.
World Scientific, Singapore, New Jersey, London, Hong Kong.
Dutta, M. and Stroscio, M.A., eds. (2000), Advances in Semiconductor Lasers and
Applications to Optoelectronics. World Scientific, Singapore, New Jersey, London, Hong
Kong.
Dutta, M., Grubin, H.L., Iafrate, G.J., Kim, K.W., and Stroscio, M.A. (1993),
Metal-encapsulated quantum wire for enhanced charge transport, US Patent Number
5 264 711 issued November 23, 1993.
Dutta, M., Stroscio, M.A., and Kim, K.W. (1998), Recent developments on electron–phonon
interactions in structures for electronic and optoelectronic devices, in Quantum-Based
Electronic Devices and Systems, Selected Topics in Electronics and Systems 14. World
Scientific, Singapore, New Jersey, London, Hong Kong.
Dutta, M., Stroscio, M.A., Bergman, L., Alexson, D., Nemanich, R.L., Dupuis, R. (2000),
Observation of interface optical phonons in würtzite GaN/AlN superlattices, to be
published.
Edgar, J.H., ed. (1994), Properties of Group III Nitrides. INSPEC, London.
Educato, J.L., Leburton, J.-P., Boucaud, P., Vagos, P., and Julien, F.H. (1993), Influence of
interface phonons on intersubband scattering in asymmetric coupled quantum wells,
Physical Review, B47, 12 949–12 952.
Empedocles, S.A., Norris, D.J., and Bawendi, M.G. (1996), Photoluminescence spectroscopy
of single CdSe nanocrystallite quantum dots, Physical Review Letters, 77, 3873–3876.
Engleman, R. and Ruppin, R. (1968a), Optical lattice vibrations in finite ionic crystals: I,
Journal of Physics C, Series 2, 1, 614–629.
Engleman, R. and Ruppin, R. (1968b), Optical Lattice vibrations in finite ionic crystals: II,
Engleman, R. and Ruppin, R. (1968c), Optical lattice vibrations in finite ionic crystals: III,
Esaki, L. and Tsu, R. (1970), Superlattice and negative differential conductivity in
semiconductors, IBM Journal of Research and Development, 14, 61–65.
Faist, J., Capasso, F., Sirtori, C., Sivco, D.L., Baillargeon, J.N., Hutchinson, L.A., Chu,
Sung-Nee G., and Cho, A. (1996a), High power mid-infrared quantum cascade lasers
operating above room temperature, Applied Physics Letters, 68, 3680–3682; also see Hu,
Q., and Feng, S. (1991), Feasibility of far-infrared lasers using multiple semiconductor
quantum wells, Applied Physics Letters, 59, 2923–2925.
Faist, J., Capasso, F., Sirtori, C., Sivco, D.L., Hutchnison, A.L., Hybertson, M.S., and Cho,
A.Y. (1996b), Quantum cascade lasers without intersubband population inversion,
Physical Review Letters, 76, 411–414.
Fasol, G., Tanaka, M., Sakaki, H., and Horikosh, Y. (1988), Interface roughness and dispersion
of confined LO phonons in GaAs/AlAs quantum wells, Physical Review, B38,
6056–6065.
Ferry, D.K. (1991), Semiconductors. Macmillan Co., New York.
Fuchs, R. and Kliewer, K.L. (1965), Optical modes of vibration in an ionic crystal slab,
Physical Review, 140, A2076–A2088.
References 261
Fuchs, R., Kliewer, K.L., and Pardee, W.J. (1966), Optical properties of an ionic crystal slab,
Physical Review, 150, 589–596.
Gauthier-Lafaye, O., Sauvage, S., Boucaud, P., Julien, F.H., Prazeres, R., Glotin, F., Ortega,
J.-M., Thierry-Mieg, V., Planel, R., Leburton, J.-P., and Berger, V. (1997), Intersubband
stimulated emission in GaAs/AlGaAs quantum wells: pump-probe experiments using a
two-color free electron laser, Physical Review Letters, 70, 3197–3200.
Gelmont, B., Gorfinkel, S., and Luryi, S. (1996), Theory of the spectral line shape and gain in
quantum wells with intersubband transitions and ionic crystals, Applied Physics Letters,
68, 2171–2173.
Giehler, M., Ramsteiner, M., Brandt, O., Yarg, H., and Ploog, K.H. (1995), Optical phonons of
hexagonal and cubic GaN studied by infrared transmission and Raman spectroscopy,
Gleize, J., Renucci, M.A., Frandon, J., and Demangeot, F. (1999), Anisotropy effects on polar
optical phonons in würtzite GaN/AlN superlattices, Physical Review, B60, 15 985–15 992.
Gleize J., Demangeot, F., Frandon J., Renucci, M.A., Kuball M., Grandjean N., and Massies, J.
(2000), Resonant Raman scattering in (Al,Ga)N/GaN quantum well structures, Thin Solid
Films, 364, 156–160.
Gmachl, C., Capasso, F., Tredicucci, D.L., Sivco, D.L., Hutchinson, A.L., Chu, S.N. G., and
Cho, A.Y. (1998), Noncascaded intersubband injection lasers at λ = 7.7 µm, Applied
Physics Letters, 73, 3830–3832.
Göppert M., Hetterich, M., Dinger A., Klingshorn C., and O’Donnell, K.P. (1998), Infrared
spectroscopy of confined optical and folded acoustical phonons in strained CdSe/CdS
superlattices, Physical Review, B57, 13 068–13 071.
Gorczyca, I., Christensen, N.E., Peltzer y Blanca, E.L., and Rodriguez, C.O. (1995), Optical
phonon modes in GaN and AlN, Physical Review, B51, 11 936–11 939.
Gorfinkel, V., Luryi, S., and Gelmont, B. (1996), Theory of gain spectra for quantum cascade
lasers and temperature dependence of their characteristics at low and moderate carrier
concentrations, IEEE Journal of Quantum Electronics, 32, 1995–2003.
Harima, H., Inoue, T., Nakashima, S., Okumura, H., Ishida, Y., Koiguari, T., Grille, H., and
Bechstedt, F. (1999), Raman studies on phonon modes in cubic AlGaN alloys, Applied
Hayashi, K., Itoh, K., Sawaki, N., and Akasaki, I. (1991), Raman scattering in Alx Ga1−x N
alloys, Solid State Communications, 77, 115–118.
Hayes, W. and Loudon, R. (1978), Scattering of Light by Crystals. John Wiley and Sons, New
York.
Hess, Karl (1999), Advanced Theory of Semiconductor Devices. IEEE Press, New Jersey.
Hewat, A.W. (1970), Lattice dynamics of ZnO and BeO, Solid State Communications, 8,
187–189.
Hoben, M.V. and Russell, J.P. (1964), The Raman spectrum of gallium phosphide, Physics
Letters, 13, 39–41.
Hon, D.T. and Faust, W.L. (1973), Dielectric parameterization of Raman lineshapes for GaP
with a plasma of charge carriers, Applied Physics Letters, 1, 241–256.
Hu, Q. and Feng, S. (1991), Feasibility of far-infrared lasers using multiple semiconductor
quantum wells, Applied Physics Letters, 59, 2923–2925.
262 References
Huang, K. (1951), On the interaction between the radiation field and ionic crystals,
Proceedings of the Royal Society A, 208, 352–365.
Huang, K. and Zhu, B. (1988), Dielectric continuum model and Fröhlich interaction in
superlattices, Physical Review, B38, 13 377–13 386.
Hwang, E.H., Das Sarma, S., and Stroscio, M.A. (2000), Role of confined phonons in thin-film
superconductivity, Physical Review, B61, 8659–8662.
Iafrate, Gerald J., and Stroscio, M.A. (1996), Application of quantum-based devices: trends
and challenges, IEEE Transactions on Electron Devices, 43, 1621–1625.
Iijima, S. (1991), Helical microtubules of graphitic carbon, Nature, 354, 56–57.
Irmer, G., Toporov, V.V., Bairamov, B.H., and Monecke, J. (1983), Determination of the charge
carrier concentration and mobility in n-GaP by Raman spectroscopy, Physics Status
Solidi, B119, 595–603.
Julien, F.H., Sa’ar, A., Wang, J., and Leburton, J.-P. (1995), Optically pumped intersubband
emission in quantum wells, Electronics Letters, 31, 838–839.
Jusserand, B. and Cardona M. (1991), Raman spectroscopy of vibrations in supertlattices, in
Light Scattering in Solids V, ed. M. Cardona, and G. Güntherodt. Springer-Verlag, Berlin.
Jusserand, B. and Paguet, D. (1986), Comment on ‘Resonance Raman Scattering by confined
LO and TO phonons in GaAs–AlAs superlattices’, Physical Review Letters, 56, 1752.
Kash, J.A. and Tsang, J.C. (1991), Light scattering and other secondary emission studies of
dynamical processes in semiconductors, in Light Scattering in Solids VI, M. Cardona, ed.
Springer-Verlag, Berlin.
Kash, J.A., Jha, S.S., and Tsang, J.C. (1987), Picosecond Raman studies of the Fröhlich
interaction in semiconductor alloys, Physical Review Letters, 58, 1869–1872; see also
Nash, K.J., and Skolnick, M.S. (1988), Comment on ‘Picosecond Raman studies of the
Fröhlich interaction in semiconductor alloys’, Physical Review Letters, 60, 863.
Kash, J.A., Jha, S.S., and Tsang, J.C. (1988), Kash, Jha, and Tsang reply, Physical Review
Letters, 60, 864.
Kash, J.A., Tsang, J.C., and Hvam, J.M. (1985), Subpicosecond time-resolved Raman
spectroscopy of LO phonons in GaAs, Physical Review Letters, 54, 2151–2154.
Keating, P.N. (1966), Theory of the third-order elastic constants of diamond-like crystals,
Physical Review, 149, 674–678.
Kim, K.W. and Stroscio, M.A. (1990), Electron–optical-phonon interaction in binary/ternary
heterostructures, Journal of Applied Physics, 68, 6289–6292.
Kim, K.W., Stroscio, M.A., Bhatt, A., Mickevicius, R., and Mitin, V.V. (1991),
Electron–optical-phonon scattering rates in a rectangular semiconductor quantum wire,
Journal of Applied Physics, 70, 319–327.
Kim, K.W., Bhatt, A.R., Stroscio, M.A., Turley, P.J., and Teitsworth, S.W. (1992), Effects of
interface phonon scattering in multi-heterointerface structures, Journal of Applied
Physics, 72, 2282–2287.
Kirillov, D., Lee, H., and Harris, J.S. (1996), Raman study of GaN films, Journal of Applied
Physics, 80, 4058–4062.
Kisin, M.V., Gorfinkel, V.B., Stroscio, M.A., Belenky, G., and Luryi, S. (1997), Influence of
complex phonon spectra on intersubband optical gain, Journal of Applied Physics, 82,
2031–2038.
References 263
Kisin, M.V., Stroscio, M.A., Belenky, G., and Luryi, S. (1998a), Electron–plasmon relaxation
in quantum wells with inverse subband occupation, Applied Physics Letters, 73 ,
2075–2077.
Kisin, M.V., Stroscio, M.A., Belenky, G., Gorfinkel, V.B., and Luryi, S. (1998b), Effects of
interface phonon scattering in three-interface heterostructures, Journal of Applied
Physics, 83, 4816–4822.
Kitaev, Yu. E., Limonov, M.F., Tronc, P., and Yushkin, G.N. (1998), Raman-active modes in
würtzite (GaN)m (AlN)n superlattices, Physical Review, B57, 14 209–14 212.
Kittel, C. (1976), Introduction to Solid State Physics, 5th edition. John Wiley & Sons, New
York.
Klein, M.V. (1975), Electronic Raman scattering, pp. 147–204, in Light Scattering in Solids,
M. Cardona, ed. Springer-Verlag, Heidelberg.
Klein, M.V. (1986), Phonons in semiconductor superlattices, IEEE Journal of Quantum
Electronics, QE-22, 1760–1770.
Klein, M.V., Ganguly, B.N., and Colwell, P.J. (1972), Theoretical and experimental study of
Raman scattering from coupled LO-phonon–plasmon modes in silicon carbide, Physical
Review, B6, 2380–2388.
Klemens, P.G. (1958), Thermal conductivity and lattice vibrational modes, in Solid State
Physics: Advances in Research and Applications, Vol. 7, eds. F. Seitz and D. Turnbull, pp.
1–98. Academic Press, New York.
Klemens, P.G. (1966), Anharmonic decay of optical phonons, Physical Review, 148, 845–848.
Klemens, P.G. (1975), Anharmonic decay of optical phonons in diamond, Physical Review,
B11, 3206–3207.
Kliewer, K.L. and Fuchs, R. (1966a), Optical modes of vibration in an ionic slab including
retardation. I.: Non-radiative region, Physical Review, 144, 495–503.
Kliewer, K.L. and Fuchs, R. (1966b), Optical modes of vibration in an ionic slab including
retardation. II.: Radiative region, Physical Review, 150, 573–588.
Knipp, P.A. and Reinecke, T.L. (1992), Interface phonons of quantum wires, Physical Review,
B45, 9091–9102.
Komirenko, S.M., Kim, K.W., Stroscio, M.A., and Dutta, M. (1999), Dispersion of polar
optical phonons in würtzite heterostructures, Physical Review, B59, 5013–5020.
Komirenko, S.M., Kim, K.W., Stroscio, M.A., and Dutta, M. (2000a), Energy-dependent
electron scattering via interaction with optical phonons in würtzite crystals and quantum
wells, Physical Review, B61, 2034–2040.
Komirenko, S.M., Kim, K.W., Demidenko, A.A., Kochelap, V.A., and Stroscio, M.A. (2000b),
Cerenkov generation of high-frequency confined acoustic phonons in quantum wells,
Komirenko, S.M., Kim, K.W., Stroscio, M.A., and Dutta, M. (2000c), Applicability of the
Fermi golden rule, Fröhlich polaron in the Feynman model and new perspectives for
possibility of quasiballistic transport in nitrides, to be published.
Kozawa, T., Kachi, T., Kano, H., Taga, Y., Hashimoto, M., Koide, N., and Manabe, K. (1998),
Raman scattering from LO-phonon–plasmon coupled modes in gallium nitride, Journal of
Applied Physics, 75, 1098–1101.
Krauss, Todd D. and Wise, Frank W. (1997), Coherent acoustic phonons in a quantum dot,
Physical Review Letters, 79, 5102–5105.
264 References
Kroto, H.W., Heath, J.R., O’Brien, S.C., Curl, R.F., and Smalley, R.E. (1985), C60 :
Buckminsterfullerine, Nature, 318, 162–163.
Kwon, H.J., Lee, Y.H., Miki, O., Yamano, H., and Yoshida, A. (1996), Raman scattering of
indium nitride thin films grown by microwave-excited metalorganic vapor phase epitaxy
on (0001) sapphire substrates, Applied Physics Letters, 69, 937–939.
Leburton, J.-P. (1984), Size effects on polar optical phonon scattering of 1D and 2D electron
gas in synthetic semiconductors, Journal of Applied Physics, 56, 2850–2855.
Leburton, J.-P. (1997), Dissipation and scattering time engineering in quantum devices, in
Proceedings of the International Conference on Quantum Devices, eds. K. Ismail, S.
Bandyopadhyay, and J.-P. Leburton, pp. 242–252. World Scientific, Singapore.
Lee, B.C., Kim, K.W., Dutta, M., and Stroscio, M.A. (1997), Electron–optical-phonon
scattering in würtzite crystals, Physical Review, B56, 997–1000.
Lee, B.C., Kim, K.W., Stroscio, M.A., and Dutta, M. (1998), Optical phonon confinement and
scattering in würtzite heterostructures, Physical Review, B58, 4860–4865.
Lee, I., Goodnick, S.M., Gulia, M., Molinari, E., and Lugli, P. (1995), Microscopic
calculations of the electron–optical-phonon interaction in ultrathin GaAs/Alx Ga1−x As
alloy quantum-well systems, Physical Review, B51, 7046–7057.
Lemos, V., Arguello, C.A., and Leite, R.C.C. (1972), Resonant Raman scattering of TO(A1 ),
TO(E1 ) and E2 optical phonons in GaN, Solid State Communications, 11, 1351–1353.
Licari, J.J. and Evrard R. (1977), Electron–phonon interaction in a dielectric slab: effect of the
electronic polarizability, Physical Review, B15, 2254–2264.
Loudon, R. (1964), The Raman effect in crystals Advances in Physics, 13, 423–482; erratum,
ibid. (1965), 14, 621.
Lucovsky, G. and Chen, M.F. (1970), Long wave optical phonons in the alloy systems:
Ga1−x Inx As, GaAs1−x Sbx , and InAs1−x Sbx , Solid State Communications, 8,
1397–1401.
Lyubomirsky, I. and Hu, Q. (1998), Energy level schemes for far-infrared quantum well lasers,
Applied Physics Letters, 73, 300–302; see also Dutta, M., and Stroscio, M.A. (1999),
Comment on ‘Energy level schemes for far-infrared quantum well lasers’, Applied
Physics Letters, 74, 2555.
McNeil, L.E., Grimsditch, M., and French, R.H. (1993), Vibrational spectroscopy of aluminum
nitride, Journal of the American Ceramics Society, 76, 1132–1136.
Manchon, D.D., Barker, A.S., Dean, P.J., and Zetterstrom, R.B. (1970), Optical studies of the
phonons and electrons in gallium nitride, Solid State Communications, 8, 1227–1231.
Marcatili, E.J. (1969), Dielectric rectangular waveguide and directional coupler for integrated
optics, Bell Systems Technical Journal, 48, 2071–2102.
M̌arkus, Štefan (1988), The Mechanics of Vibrations of Cylindrical Shells. Elsevier,
Amsterdam.
Menéndez, J. and Cardona, M. (1984), Temperature dependence of the first-order Raman
scattering by phonons in Si, Ge, and α-Sn: anharmonic effects, Physical Review, B29,
2051–2059.
Mitin, V.V., Kochelap, V.A., and Stroscio, M.A. (1999), Quantum Heterostructures:
Microelectronics and Optoelectronics. Cambridge University Press, Cambridge.
References 265
Miwa, K. and Fukumoto, A. (1993), First-principles calculation of the structural, electronic,

and vibrational properties of gallium nitride and aluminum nitride, Physical Review, B48,
7897–7902.
Molinari, E., Bungaro, C., Gulia, M., Lugli, P., and Rücker, H. (1992), Electron–phonon
interactions in two-dimensional systems: a microscopic approach, Semiconductor Science
and Technology, 7, B67–B72.
Molinari, E., Bungaro, C., Rossi, F., Rota, L., and Lugli, P. (1993), Phonons in GaAs/AlAs
nanostructures: from two-dimensional to one-dimensional systems, in Phonons in
Semiconductor Nanostructures, eds. J.-P. Leburton, J. Pascual, and C. Sotomayer-Torres,
NATO Advanced Study Institutes, Series E, Vol. 236. Kluwer Academic Publishers,
Dordrecht, 39–48.
Mooradian, A., and Wright, G.B., eds. (1969), Light Scattering Spectra in Solids,
Springer-Verlag, New York, p. 297.
Mori, N. and Ando, T. (1989), Electron–optical phonon interaction in single and double
heterostructures, Physical Review, B40, 6175–6188.
Mori, N., Taniguchi, K., and Hamaguchi, C. (1992), Effects of electron–interface-phonon
interaction on resonant tunneling in double-barrier heterostructures, Semiconductor
Science and Technology, 7, B83–B87.
Morse, R.W. (1948), Dispersion of compressional waves in isotropic rods of rectangular cross
section, Journal of the Acoustical Society of America, 20, 833–838.
Morse, R.W. (1949), The dispersion of compressional waves in isotropic rods of rectangular
cross section, Ph.D.Thesis, Brown University, Providence, RI.
Morse, R.W. (1950), The velocity of compressional waves in rods of rectangular cross section,
Journal of the Acoustical Society of America, 22, 219–223.
Murugkar, S., Merlin, R., Botchkarev, A., Salvador, A., and Morkoç, H. (1995), Second order
Raman spectroscopy of the würtzite form of GaN, Journal of Applied Physics, 77,
6042–6043.
Nabity, J.C. and Wybourne, M.N. (1990a), Phonon trapping in thin metal films, Physical
Review, B42, 9714–9716.
Nabity, J.C. and Wybourne, M.N. (1990b), Evidence of two-dimensional phonons in thin metal
films, Physical Review, B44, 8990–8999.
Nash, K.J. (1992), Electron–phonon interactions and lattice dynamics of optic phonons in
semiconductor heterostructures, Physical Review, B46, 7723–7744.
Nipko, J.C. and Loong, C.K. (1998), Phonon excitations and related thermal properties of
aluminum nitride, Physical Review, B57, 10 055–10 554.
Nipko, J.C., Loong, C.K., Balkas, C.M., and Davis, R.F. (1998), Phonon density of states of
bulk gallium nitride, Applied Physics Letters, 73, 34–36.
Nusimovici, M.A. and Birman, J.L. (1967), Lattice dynamics of würtzite: CdS, Physical
Review, 156, 925–938.
Omar, M.A. (1975), Elementary Solid State Physics: Principles and Applications.
Addison-Wesley, Reading, MA.
Patel, C.K.N. and Slusher, R.E. (1968), Light scattering by plasmons and Landau levels of
electron gas in InAs, Physical Review, 167, 413–415.
Perlin, P., Polian, A., and Suski, T. (1993), Raman scattering studies of aluminum nitride at
high pressure, Physical Review, B47, 2874–2877.
266 References
Pfeifer, T., Kutt, W., Kurz, H., and Scholz, H. (1992), Generation and detection of coherent
optical phonons in germanium, Physical Review Letters, 69, 3248–3251.
Platzman, P.M. and Wolff, P.A. (1973), Waves and interactions in solid state plasmas, Solid
State Physics Series, Supplement 13, eds. H. Ehrenreich, Seitz, F., and Turnbull, D.
Academic Press, New York.
Ponce, F.A., Steeds, J.W., Dyer, C.D., and Pitt, G.D. (1996), Direct imaging of
impurity-induced Raman scattering in GaN, Applied Physics Letters, 69, 2650–2652.
Pötz, W. and Schroeder, W.A., eds. (1999), Coherent Control in Atoms, Molecules, and
Semiconductors. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Register, L.F., Stroscio, M.A., and Littlejohn, M.A. (1988), A highly efficient computer
algorithm for evaluating Feynman path-integrals, Superlattices and Microstructures, 6,
233–236.
Register, L.F., Stroscio, M.A., and Littlejohn, M.A. (1991), Conservation law for confined
polar-optical phonon influence functionals, Physical Review, B44, 3850–3857.
Richter, E. and Strauch, D. (1987), Lattice dynamics of GaAs/AlAs superlattices, Solid State
Communications, 64, 867–870.
Ridley, B.K. (1996), The LO phonon lifetime in GaN, Journal of Physics: Condensed Matter,
8, L511–L513.
Ridley, B.K. (1997), Electrons and Phonons in Semiconductor Multilayers. Cambridge
University Press, Cambridge.
Ridley, B.K. and Gupta, R. (1991), Nonelectronic scattering of longitudinal-optical phonons in
bulk polar semiconductors, Physical Review, B43, 4939–4944.
Rücker, H., Molinari, E., and Lugli, P. (1992), Microscopic calculation of the electron–phonon
interactions in quantum wells, Physical Review, B45, 6747–6756.
Ruppin, R., and Engleman, R. (1970), Optical phonons in small crystals, Reports on the
Progress of Physics, 33, 149–196.
Sakaki, H. (1989), Quantum wire superlattices and coupled quantum box arrays: a novel
method to suppress optical phonon scattering in semiconductors, Japanese Journal of
Applied Physics, 28, L314–L316.
Seyler, J. and Wybourne, M.N. (1992), Phonon subbands observed in electrically heated metal
wires, Physics of Condensed Matter, 4, L231–L236.
Scott, J.F., Leite, R.C.C., Damen, T.C., and Shah, J. (1969), Resonant Raman effects in
semiconductors, Physical Review, 188, 1285–1290.
Shah, J., Leite, R.C.C., and Scott, J.F. (1970), Photo-excited hot LO phonons in GaAs, Solid
State Communications, 9, 1089–1093.
Shapiro, S.M. and Axe, J.D. (1972), Raman scattering from polar phonons, Physical Review,
B6, 2420–2427.
Siegle, H., Kaczmarczyk, G., Filippidis, L., Litvinchuk, A.P., Hoffmann, A., and Thomsen, C.
(1997), Zone-boundary phonons in hexagonal and cubic GaN, Physical Review, B55,
7000–7004.
Singh, J. (1993), Physics of Semiconductors and Their Heterostructures. McGraw-Hill, New
York.
Sirenko, Y.M., Kim, K.W., and Stroscio, M.A. (1995), Elastic vibrations of biological and
artificial microtubules and filaments, Electrochemical Society Proceedings, 95-17,
260–271.
References 267
Sirenko, Y.M., Stroscio, M.A., and Kim, K.W. (1996a), Dynamics of cytoskeleton filaments,
Physical Review, E54, 1816–1823.
Sirenko, Y.M., Stroscio, M.A., and Kim, K.W. (1996b), Elastic vibrations of microtubules in a
fluid, Physical Review, E53, 1003–1011.
Sood, A.K., Menéndez, J., Cardona, M., and Ploog, K. (1985), Resonance Raman scattering by
confined LO and TO phonons, Physical Review Letters, 54, 2111–2114.
Sood, A.K., Menéndez, J., Cardona, M., and Ploog, K. (1986), Sood et al. respond, Physical
Review Letters, 56, 1753.
Stroscio, M.A. (1989), Interaction between longitudinal-optical phonon modes of a rectangular
quantum wire with charge carriers of a one-dimensional electron gas, Physical Review,
B40, 6428–6431.
Stroscio, M.A. (1996), Interface-phonon-assisted transitions in quantum well lasers, Journal of
Stroscio, M.A. and Kim, K.W. (1993), Piezoelectric scattering of carriers in confined acoustic
modes in cylindrical quantum wires, Physical Review, B48, 1936–1939.
Stroscio, M.A. and Dutta, M. (1999), Damping of nonequilibrium acoustic phonon modes in
semiconductor quantum dot, Physical Review, B60, 7722–7724.
Stroscio, M.A., Kim, K.W., and Littlejohn, M.A. (1990), Theory of optical-phonon interactions
in rectangular quantum wires, in Proceedings of the Society of Photo-optical
Instrumentation Engineers Conference on Physical Concepts of Materials for Novel
Optoelectronic Device Applications II: Device Physics and Applications, 1362 ed. M.
Razeghi, pp. 556–579.
Stroscio, M.A., Iafrate, G.J., Kim, K.W., Littlejohn, M.A., Goronkin, H., and Maracas, G.
(1991a), Transitions from LO-phonon to SO-phonon scattering in short-period
AlAs–GaAs superlattices, Applied Physics Letters, 59, 1093–1096.
Stroscio, M.A., Kim, K.W., and Rudin, S. (1991b), Boundary conditions for
electron–LO-phonon interaction in polar semiconductor quantum wires, Superlattices and
Microstructures, 10, 55–58.
Stroscio, M.A., Kim, K.W., Iafrate, G.J., Dutta, M., and Grubin, H.L. (1991c), Reduction and
control of inelastic longitudinal-optical phonon scattering in nanoscale and mesoscopic
device structures, in Proceedings of the 1991 International Device Research Symposium,
pp. 87–89, ISBN 1-880920-00-X, University of Virginia Engineering Academic Outreach
Publication, Charlottesville.
Stroscio, M.A., Kim, K.W., Iafrate, G.J., Dutta, M., and Grubin H. (1992), Dramatic reduction
of the longitudinal-optical phonon emission rate in polar-semiconductor quantum wires,
Philosophical Magazine Letters, 65, 173–176.
Stroscio, M.A., Iafrate, G.J., Kim, K.W., Yu, S., Mitin, V., and Bannov, N. (1993), Scattering
of carriers from confined acoustic modes in nanostructures, Proceedings of the 1993
International Device Research Symposium, pp. 873–875, ISBN 1-880920-02–6,
University of Virginia Engineering Academic Outreach Publication, Charlottesville.
Stroscio, M.A., Kim, K.W., Yu, S., and Ballato, A. (1994), Quantized acoustic phonon modes
in quantum wires and quantum dots, Journal of Applied Physics, 76, 4670–4673.
Stroscio, M.A., Sirenko, Yu. M., Yu, S., and Kim, K.W. (1996), Acoustic phonon quantization
in buried waveguides and resonators, Journal of Physics: Condensed Matter, 8,
2143–2151.
268 References
Stroscio, M.A., Kisin, M.V., Belenky, G., and Luryi, S. (1999), Phonon enhanced inverse
population in asymmetric double quantum wells, Applied Physics Letters, 75, 3258–3260.
Sun, C.-K., Liang, J.-C., Stanton, C.J., Abare, A., Coldren, L., and DenBaars, S. (1999), Large
coherent acoustic-phonon oscillations observed in InGaN/GaN multiple-quantum wells,
Sun, H.C., Davis, L., Sethi, S., Singh, J., and Bhattacharya, P. (1993), Properties of a tunneling
injection quantum-well laser: recipe for a ’cold’ device with a large modulation
bandwidth, IEEE Photonics Technology Letters 5, 870–872.
Sun, J.-P., Teng, H.B., Haddad, G.I., Stroscio, M.A., and Iafrate, G.J. (1997), Intersubband
relaxation in step quantum well structures, International Conference on Computational
Electronics, University of Notre Dame, 28–30 May 1997, private communication.
Sung, C.Y., Norris, T.B., Afzali-Kushaa, A., and Haddad, G.I. (1996), Femtosecond
intersubband relaxation and population inversion in stepped quantum well, Applied
Physics Letters, 68 , 435–437; also see the references in this paper.
Teng, H.B., Sun, J.P., Haddad, G.I., Stroscio, M.A., Yu, S., and Kim, K.W. (1998), Phonon
assisted intersubband transitions in step quantum well structures, Journal of Applied
Physics, 84, 2155–2164.
Thompson, C.J. and Blatt, J.M. (1963), Shape resonances in superconductors II: simplified
theory, Physics Letters, 5, 6–9.
Tsen, K.T. (1992), Picosecond time-resolved Raman studies of electron–optical phonon
interactions in ultrathin GaAs–AlAs multiple quantum well structures, Semiconductor
Science and Technology, 7, B191–B194.
Tsen, K.T. and Morkoç, H. (1988a), Picosecond Raman studies of the optical phonons in the
AlGaAs layers of GaAs–AlGaAs multiple-quantum-well structures, Physical Review,
B37, 7137–7139.
Tsen, K.T. and Morkoç, H. (1988b), Subpicosecond time-resolved Raman spectroscopy of LO
phonons in GaAs–Alx Ga1−x As multiple-quantum-well structures, Physical Review, B38,
5615–5616.
Tsen, K.T., Joshi, R.P., Ferry, D.K., and Morkoc, H. (1989), Time-resolved Raman scattering of
nonequilibrium LO phonons in GaAs quantum wells, Physical Review, B39, 1446–1449.
Tsen, K.T., Joshi, R.P., Ferry, D.K., Botchkarev, A., Sverdlov, B., Salvador, A., and Morkoc, H.
(1996), Nonequilibrium electron distributions and phonon dynamics in würtzite GaN,
Tsen, K.T., Ferry, D.K., Botchkarev, A., Sverdlov, B., Salvador, A., and Morkoc, H. (1997),
Direct measurement of electron-longitudinal optical phonon scattering rates in würtzite
GaN, Applied Physics Letters, 71, 1852–1853.
Tsen, K.T., Ferry, D.K., Botchkarev, A., Sverdlov, B., Salvador, A., and Morkoc, H. (1998),
Time-resolved Raman studies of the decay of the longitudinal optical phonons in würtzite
GaN, Applied Physics Letters, 72, 2132–2136.
Tua, P.F. (1981), Lifetime of high-frequency longitudinal-acoustic phonons in CaF2 at low
crystal temperatures, Ph.D. thesis, Indiana University.
Tua, P.F. and Mahan, G.D. (1982), Lifetime of high-frequency longitudinal-acoustic phonons
in CaF2 at low crystal temperatures, Physical Review, B26, 2208–2215.
Turley, P.J., Wallis, C.R., and Teitworth, S.W. (1991a), Phonon-assisted tunneling due to
localized modes in double-barrier structures, Physical Review, B44, 8181–8184.
References 269
Turley, P.J., Wallis, C.R., and Teitworth, S.W. (1991b), Electronic wave functions and
electron-confined-phonon matrix elements in GaAs/Alx Ga1−x As double-barrier resonant
tunneling structures, Physical Review, B44, 3199–3210.
Turley, P.J., Wallis, C.R., and Teitworth, S.W. (1991c), Effects of localized phonon modes on
magnetotunneling spectra in double-barrier structures, Physical Review, B44,
12 959–12 963.
Turley, P.J., Wallis, C.R., and Teitworth, S.W. (1992), Theory of localized phonon modes and
their effects on electron tunneling in double-barrier structures, Journal of Applied
Physics, 76, 2356–2366.
Turley, P.J., Wallis, C.R., Teitworth, S.W., Li, W., and Bhattacharya, P.K. (1993), Tunneling
measurements of symmetric-interface phonons in GaAs/AlAs double-barrier structures,
Physical Review, B47, 12 640–12 648.
Vogl, P. (1980), The electron–phonon interaction in semiconductors, in Physics of Nonlinear
Transport in Semiconductors, Proceedings of the NATO Advanced Study Institute
Seminar, eds. D.K. Ferry, J.R. Barker, and C. Jacoboni. Plenum, New York.
von der Linde, D., Kuhl, J., and Klingenberg, H. (1980), Raman scattering from
nonequilibrium LO phonons with picosecond resolution, Physical Review Letters, 44,
1505–1508.
Wang, Jin, Leburton, J.-P., Moussa, Z., Julien, F.H., and Sa’ar, A. (1996a), Simulation of
optically pumped mid-infrared intersubband semiconductor laser structures, Journal of
Wang, Jin, Leburton, J.-P., Julien, F.H., and Sa’ar, A. (1996b), Design and performance
optimization of optically-pumped mid-infrared intersubband semiconductor lasers, IEEE
Photonics Technology Letters, 8, 1001–1003.
Wang, X.F. and Lei, X.L. (1994), Polar-optic phonons and high-field electron transport in
cylindrical GaAs/AlAs quantum wires, Physical Review, B49, 4780–4789.
Waugh, J.L.T. and Dolling, G. (1963), Crystal dynamics of gallium arsenide, Physical Review,
132, 2410–2412.
Wendler, L. (1985), Electron–phonon interaction in dielectric bilayer systems: effects of the
electronic polarizability, Physica Status Solidi B, 129, 513–530.
Wendler, L. and Grigoryan, V.G. (1988), Acoustic interface waves in sandwich structures,
Surface Science, 206, 203–224.
Wetzel, C., Walukiewicz, W., Haller, E.E., Ager, III, J., Grzegory, I., Porowski, S., and Suski,
T. (1996), Carrier localization of as-grown n-type gallium nitride under large hydrostatic
pressure, Physical Review, B53, 1322–1326.
Wisniewski, P., Knap, W., Malzak, J.P., Camassel, J., Bremser, M.D., Davis, R.F., and Suski, T.
(1998), Investigation of optically active E1 transversal optic phonon modes in
Alx Ga1−x N layers deposited on 6H-SiC substrates using infrared reflectance, Applied
Worlock, J.M. (1985), Phonons in superlattices, in Proceedings of the Second International
Conference on Phonon Physics, J. Kollár, N. Kroó, N. Menyhárd, and T. Siklós., eds., pp.
506–520. World Scientific, Singapore.
Xu, B., Hu, Q., and Mellock, M.R. (1997), Electrically pumped tunable terahertz emitter based
on intersubband transition, Applied Physics Letters, 71, 440–442.
Yafet, Y. (1966), Raman scattering by carriers in Landau levels, Physical Review, 152,
858–862.
270 References
Yu, P.Y. and Cardona, M. (1996), Fundamentals of Semiconductors. Springer, Berlin.

Yu, M., Strongin, M., and Paskin, A. (1976), Consistent calculation of boundary effects in thin
superconducting films, Physical Review, B14, 996–1001.
Yu, S., Kim, K.W., Stroscio, M.A., Iafrate, G.J., and Ballato, A. (1994a),
Electron–acoustic-phonon scattering rates in rectangular quantum wires, Physical Review,
B50, 1733–1738.
Yu, S., Kim, K.W., Stroscio, M.A., and Iafrate, G.J. (1994b), Electron–acoustic-phonon
scattering rates in cylindrical quantum wires, Physical Review, B51, 4695–4698.
Yu, S., Kim, K.W., Stroscio, M.A., Iafrate, G.J., and Ballato, A. (1996),
Electron–confined-acoustic-phonon scattering rates in cylindrical quantum wires via
deformation potential, Journal of Applied Physics, 80, 2815–2822.
Yu, S., Kim, K.W., Stroscio, M.A., Iafrate, G.J., Sun, J.P., and Haddad, G.I. (1997), Transfer
matrix method for interface optical-phonon modes in multiple-interface heterostructure
systems, Journal of Applied Physics, 82, 3363–3367.
Yu, SeGi, Kim, K.W., Bergman, L., Dutta, M., Stroscio, M.A., and Zavada, J.M. (1998),
Long-wavelength optical phonons in ternary nitride-based crystals, Physical Review, B58,
15 283–15 287.
Zhang, J.M., Ruf, T., Cardona, M., Ambacher, O., Stutzmann, M., Wagner, J.M., and
Bechstedt, F. (1997), Raman spectra of isotropic GaN, Physical Review, B56,
14 399–14 406.
Zhang, X., Gutierrez-Aitken, A.L., Klotzkin, D., Bhattacharya, P., Caneau, C., and Bhat, R.
(1996), 0.98-micrometer multi-quantum well tunneling injection lasers with ultra-high
modulation bandwidths, Electronics Letters, 32, 1715–1716.
Index
A1 mode, 17 canonical transformation, 40

acoustic modes, localized, 105 carrier concentrations, 34
acoustic phonons carrier mobilities, 219
in cylindrical quantum wires, 175 carrier–optical-phonon interaction, 79
in cylindrical structures, 112 carrier–phonon scattering, 35
in cylindrical waveguide, 113 carrier relaxation, 48
in double-interface heterostructures, 100 carrier wavefunctions, 38
in free-standing layers, 97 Cerenkov emission, 216
in nanostructures, 39, 56, 60 Cerenkov-like generation, 212
in quantum dots, 124 classical acoustics, 177
in quantum-well structures, 212 clamped-surface boundary condition, 177
in rectangular quantum wires, 105 coherent phonons, 219
in superlattices, 95, 219 compressional solutions, 60
in thin metallic foils, 60 conduction bands, 27
in unconstrained layers, 96 confined-phonon effects, in thin film
in wires, 60 superconductors, 208
alloying, 28 continuum models of phonons, 52
anharmonic coupling, of phonons, 45 corner modes, in rectangular quantum wire, 154
anharmonic effects, in würtzite structures, 50 Coulomb interaction, 41
anharmonic interactions, 8, 30, 46 coupled radial–longitudinal modes, 120
anharmonic phonon decay, 215 creation operators, 35
anharmonic terms, 45 critical points in semiconductors, 27
anharmonic third-order potential, 47 current transfer ratio, 204
annihilation operators, 35 cylindrical quantum wire, 175
antisymmetric radial–axial modes, 115, 116 cylindrical shell, 112
antisymmetric shear vertical modes, 102 cylindrical shell, immersed in fluid, 122
antisymmetric torsional mode, 123 cylindrical waveguide, 112
cytoskeletal filaments, 123
Bardeen–Cooper–Schrieffer theory of crystal symmetry, 95
superconductivity, 3 cubic crystals, 6
binary polar semiconductor, 53
binary semiconductor layer, 66 damped spherical acoustic modes, 219
Brillouin zone, 9 decay of phonons, 30, 49
Boltzmann factor, 37 deformation potential constant, 172
Bose–Einstein condensation, 35 deformation potential interaction, 31, 43, 110
Bose–Einstein distribution, 38 in rectangular quantum wire, 181
Bose–Einstein occupation number, 47 deformation potential scattering, in bulk zincblende
breathing mode, in spherical quantum dot, 127, 128 structures, 172
bulk dispersion relations, 102 diatomic lattice, 7
271
272 Index
dielectric continuum model, 52, 60, 93 gain in intersubband laser, 195, 219
dielectric function, 13 Grüneisen constant, 47
dielectric polarizability, 224 guided modes, 225
dilatation, of medium, 58
dilatational modes, 99 half-space optical phonon modes, 152
dilatational solutions, 60 Hamiltonian for harmonic oscillator, 35
dilatational waves, 98 harmonic interactions, 8
dimensional confinement, 39 harmonic modes, 45
dispersion curves, in würtzites 32, 77, 78 harmonic oscillator, 35
dispersion relations, 8, 26 hexagonal würtzite structures, 53
displacement, 8 high-temperature electronics, 54
displacement eigenmodes, 98 Hooke’s law, 8, 56
displacement field, quantized, 40 hot electron distribution, 204
dissipative mechanisms, 220 hot-phonon-bottleneck effect, 164
distortional solutions, 60 hot phonon decay, 3
double-barrier heterostructure, 88 hot phonons, in polar quantum wires, 163
drifted Fermi distribution, 216 Huang–Born equations, 20, 54, 221
driven-oscillator equation, 14, 63 Huang–Zhu modes, 66, 228
dynamical screening, in polar quantum wires, 162,
165 impurity, 28
in-gap, 27
E 1 mode, 17 inhomogeneous broadening effects, 31
effective charge, 53 infrared-active modes, 20
elastic continuum model, 56, 176 infrared-active phonons, 138
elastic continuum theory, 60 LO-like, 138
elasticity, theory of, 47 TO-like, 138
electron–acoustic-phonon scattering interface disorder, 28
in cylindrical quantum wire, 176 interface modes
in rectangular quantum wire, 112 for optical phonons, 66, 68, 80, 83, 152
in slab, 61
electron permittivity, 215
optical phonon interaction Hamiltonian, 65
electron–phonon interaction, 30
intersubband lasers, 196, 219
for slab modes, 66
intersubband scattering rates, 91
electronic polarizability, 68
intrasubband transition rates, 187
electrostatic boundary conditions, 66, 79, 152
interwell phonon-assisted transition, 207
energy-conserving delta function, 142
intrasubband scattering rates, 91
energy loss rate, 48
inversion symmetry, 43
energy–wavevector relationship, 97
ionic bonding, 7
equivoluminal solutions, 60
irrotational solutions, 60
Euler–Lagrange equations, 226
isotopic mass, 33
excitonic states, 27
isotropic medium, 43
extraordinary waves, 19
Kane wavefunction, 196
face-centered cubic lattices, 7 Klemens’ channel, 2, 46, 163
femtosecond lasers, 219
Fermi golden rule, 39, 47 Lagrangian density, 225, 230
flexural modes, 100, 123 Lamé’s constants, 58, 101, 118
flexural thickness acoustic modes, in quantum dot, Landau fans, 203
128 lifetimes, longitudinal optical phonons, 3, 49
flexural waves, 98 linear-chain model, 7
folded acoustic modes, in semiconductor localized acoustic modes, 214
superlattices, 60 longitudinal acoustic mode, 9, 17
force equations, 59, 118 longitudinal electromagnetic disturbance, 66
Fourier decomposition, 38 longitudinal electromagnetic wave, 14
free-standing cylindrical structure, 180 longitudinal optical mode, 9, 17
free-surface boundary condition, 177 longitudinal solutions, 60
form factor, 232 longitudinal sound speed, 57, 59
Fröhlich Hamiltonian, for two-dimensional slab, 69 in thin plate, 119
Fröhlich interaction, 31, 40 Loudon model, 14, 15, 23, 54
Fröhlich interaction Hamiltonian, 41 Lyddane–Sachs–Teller relation, 13, 15, 52, 80, 221
for polar uniaxial crystal, 136
for quantum box, 168 macroscopic theory of polar modes, 14
for two-dimensional slab, 140 magnetotunneling spectra, 202
Index 273
metal–semiconductor heterointerfaces, 220 polarization vector, 38

metal–semiconductor structures, 165 population inversion, 205
micro-Raman techniques, 26 power loss, 164
microscopic models for phonons, 90, 91 pure torsional mode, 120
microtubules, 122, 218
mode normalization condition, 61 quantum-box arrays, 220
modified random-element isodisplacement model, quantum-cascade laser, 196
33 quantum-well laser, 2
quantum-wire superlattices, 220
nanometer-scale mechanical structures, 218 quantum wires, with circular and elliptical cross
neutron scattering measurements, 96 sections, 161
non-equilibrium phonons, 31, 48, 162, 163
normal-mode phonon displacement, 38 Raman-active modes, 95
normal vibrational modes, 16 Raman analysis, of anharmonic phonon decay, 51
normalization condition, 81, 226 Raman experiments, at ultraviolet wavelengths, 33
Raman scattering, 4, 26
in bulk zincblende structures, 26
occupation number representation, 35
in bulk würtzite structures, 26
one-mode behavior, 32
Raman scattering measurements, 66, 93, 95
optical phonons
Raman techniques, 26
in nanostructures, 39
Raman tensor, 26
in würtzite structures, 18
reduced-mass density, 51
ordinary waves, 19
reformulated dielectric continuum model, 61
reformulated modes, 66, 225
periodic boundary conditions, 52 resonant Raman studies, 95
phonon-assisted electron intersubband transition Ridley channel, 50
rates, 186 rotational vector-potential solutions, 60
phonon-assisted transition, 207
phonon-assisted tunneling, 202 saturation velocity, 2
phonon-assisted tunneling peak, 203 scalar potential, 59, 113
phonon bottleneck, 50 scattering rates, electron–LO-phonon
in quantum-well lasers, 50 in quantum dots, 167
phonon decay, in nitride materials, 51 in quantum wires, 146, 150
phonon decay time, 164 in würtzite quantum-well structures, 146
phonon eigenstates, 38 in würtzite semiconductors, 131
phonon emission, 204 in zincblende quantum-well structures, 141
phonon engineering, 219 in zincblende semiconductors, 131
phonon-enhanced population inversion, 205 scattering rates, interface phonon, in quantum
phonon lifetimes, 30 wires, 154
phonon linewidths, 31 screening of Coulomb-like interaction, 134
phonon matrix elements, 38 second-order susceptibilities, 28
phonon occupation number, 37, 38, 215 shear horizontal modes, 104
phonon operators, 38 shear solutions, 60
phonon potential, 82 shear waves, 98, 99
photocurrent transfer ratio, 204 short-range forces, 226
picosecond Raman spectroscopy, 31 short-wavelength optoelectronic devices, 54
piezoelectric coupling, 173 simple harmonic oscillator, 35
piezoelectric crystal, 44 slab modes, 61, 152, 225
piezoelectric interaction, 43 in confined würtzite structures, 71
piezoelectric interaction potential, 179, 181 space group, 29
piezoelectric polarization, 44, 173 space-group symmetry, 28
piezoelectric scattering, in bulk semiconductor strain, 28, 56, 103
structures, 173 stress, 56, 103
plasma frequency, 82 stress–strain relation, 57
plasmon emission, 204 stress tensor, 97
plasmon–phonon modes, 33 Sturm–Liouville equation, 227
plasmons, 33 sub-picosecond Raman spectroscopy, 31
Poisson ratio, 101, 118 surface charge density, 68
Poisson’s equation, 227 symmetric shear vertical modes, 102
polar-optical phonons, 40
polar semiconductors, 7 T2 representation, 28
polaritons, 11 ternary alloys, 32
polarization charge density, 68 ternary polar semiconductor, 53
274 Index
thermalization time, of carriers, 50 uniaxial polar materials, 54

thickness modes, in rectangular quantum wire, 107, uniaxial semiconductors, 71
109 unit cell, 16, 53
thin film superconductors, 5, 210, 219
third-order elastic coefficients, 47 valence bands, 27
III-V nitrides, 16 valley current, 202, 219
III-V nitride materials, 29 vector potential, 59
three-phonon decay process, 30 volume charge density, 68
time-resolved Raman scattering, 31
torsional mode, in spherical quantum dot, 127, 128
wavevector, 8, 35
traction force, 97
Wendler’s conditions, for two-layer system, 69
transfer matrix model, for multi-heterointerface
Wendler’s model, 222
structures, 79, 85
width modes, in rectangular quantum wire, 109
transverse acoustic mode, 17
würtzite nitride system, 95
transverse optical phonons, 11
würtzite nitrides, 32
transverse solutions, 60
würtzite structure, 6, 16
transverse sound speed, 59
würtzite superlattice, 76
tunneling currents, 202
tunneling injection lasers, 50
two-mode behavior, 32 Young’s modulus, 56, 101, 118
Umklapp process, 45 zincblende crystals, 6

uniaxial crystals, 18 zincblende structures, 16
uniaxial materials, 53 zone-center phonon frequency, 133

Phonons in Nanostructures-Cambridge University Press (2001)

Uploaded by

Copyright:

Available Formats

Phonons in Nanostructures-Cambridge University Press (2001)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Phonons in Nanostructures-Cambridge University Press (2001)

Uploaded by

Copyright:

Available Formats

This page intentionally left blank

This book focuses on the theory of phonon interactions in nanoscale structures

D R M ICHAEL A. S TROSCIO earned a Ph.D. in physics from Yale University

  

© Michael A. Stroscio and Mitra Dutta 2004

First published in printed format 2001

ISBN 0-511-03187-4 eBook (Adobe Reader)

Michael Stroscio dedicates this book to

Chapter 1 Phonons in nanostructures 1

1.1 Phonon effects: fundamental limits on carrier mobilities and dynamical

Chapter 2 Phonons in bulk cubic crystals 6

2.1 Cubic structure 6

Chapter 3 Phonons in bulk würtzite crystals 16

3.1 Basic properties of phonons in würtzite structure 16

Chapter 4 Raman properties of bulk phonons 26

4.1 Measurements of dispersion relations for bulk samples 26

4.2.1 Zincblende structures 28

Chapter 5 Occupation number representation 35

5.1 Phonon mode amplitudes and occupation numbers 35

Chapter 6 Anharmonic coupling of phonons 45

Chapter 7 Continuum models for phonons 52

7.1 Dielectric continuum model of phonons 52

Chapter 8 Carrier–LO-phonon scattering 131

8.1 Fröhlich potential for LO phonons in bulk zincblende and würtzite

Chapter 9 Carrier–acoustic-phonon scattering 172

9.1 Carrier–acoustic-phonon scattering in bulk zincblende structures 172

Chapter 10 Recent developments 186

10.1 Phonon effects in intersubband lasers 186

Chapter 11 Concluding considerations 218

11.1 Pervasive role of phonons in modern solid-state devices 218

This book describes a major aspect of the effort to understand nanostructures,

technology and made it possible to realize band-engineering and atomic-level

Research Laboratory; these leaders have placed a high priority on maintaining an

University and Howard Jackson, University of Cincinnati, her mentors in various

Michael Stroscio and Mitra Dutta

There are no such things as applied sciences, only applications of

1.1 Phonon effects: fundamental limits on carrier

range of temperatures and phonon wavevectors, the lifetimes of longitudinal optical

1.2 Tailoring phonon interactions in devices with

to those for an electron confined in a quantum well. Consider the well-known

A is the area of the heterointerface over which the electron wavefunction is

8, 9 and 10 of this book describe how the dimensional confinement of phonons in

The Creator, if He exists, has a special preference for beetles.

2.1 Cubic structure

2.2 Ionic bonding – polar semiconductors

2.3 Linear-chain model and macroscopic models

Figure 2.1. Zincblende

Figure 2.2. One-dimensional linear-chain representation of a diatomic lattice.

In the simple linear-chain model considered in this section, it is assumed that

2.3.1 Dispersion relations for high-frequency and

−mω2 u 2r = m(d 2 u 2r /dt 2 ) = α(u 2r +1 + u 2r −1 − 2u 2r ) + e∗ E

−Mω2 u 2r +1 = M(d 2 u 2r +1 /dt 2 ) = α(u 2r +2 + u 2r − 2u 2r +1 ) − e∗ E

Regarding the phonon displacements, in the long-wavelength limit there is no

−mω2 u 1 = 2α(u 2 − u 1 ) + e∗ E (2.10)

−Mω2 u 2 = 2α(u 1 − u 2 ) − e∗ E. (2.11)

Adding these equations demonstrates that −mω2 u 1 − Mω2 u 2 = 0 and it is clear

−(ω2 − ω02 )u 1 = e∗ E/m (2.14)

−(ω2 − ω02 )u 2 = −e∗ E/M (2.15)

where u = u 1 − u 2 , N is the number of pairs per unit volume, and e∗ is as

q 2 − ω2 /c2 [(0) − (∞)]ωTO

[110] K M [100] [001] A

Eliminating u⊥ and u in the first four of these equations yields

For the ordinary wave, E = 0, P = 0, and q · P = 0, so that the derivation of

For the ordinary mode it also follows that u = u⊥ = 0.

q · P = (q sin θ, q cos θ) · (P⊥ , P ) = q P⊥ sin θ + q P cos θ. (3.15)

q⊥ (q · P) = q 2 (P⊥ sin2 θ + P sin θ cos θ ),

= ⊥ (ω) sin2 θ + (ω) cos2 θ = 0 (3.20)

h̄ωi = h̄ωs ± h̄, (4.1)

Nq (aq† |Nq ) = (aq† Nq + aq† )|Nq = (n q + 1)aq† |Nq (5.6)

Nq (aq |Nq ) = (aq Nq − aq )|Nq = (n q − 1)aq |Nq . (5.7)

Nq ||Nq ≡ Nq |Nq = δ Nq ,Nq (5.10)