REVIEWS OF MODERN PHYSICS, VOLUME 73, JULY 2001

Nobel Lecture: The double heterostructure concept and its applications

in physics, electronics, and technology*
Zhores I. Alferov
A. F. Ioffe Physico-Technical Institute, Russian Academy of Sciences,
St. Petersburg 194021, Russian Federation
(Published 22 October 2001)

I. INTRODUCTION enon (Frenkel and Ioffe, 1932; Zhuze and Kurchatov,

1932a, 1932b). In 1931 and 1936 Frenkel published his
It is impossible to imagine now modern solid-state famous articles where he predicted, gave the name, and
physics without semiconductor heterostructures. Semi- developed the theory of excitons in semiconductors, and
conductor heterostructures and, particularly, double het- E. F. Gross experimentally discovered excitons in 1951
erostructures, including quantum wells, wires, and dots, (Frenkel, 1931, 1936; Gross and Karryev, 1952a, 1952b).
are today the subject of research of two-thirds of the The first diffusion theory of p-n heterojunction rectifi-
semiconductor physics community. cation, which became the base for W. Shockley’s p-n
The ability to control the type of conductivity of a junction theory, was published by B. I. Davydov in 1939
semiconductor material by doping with various impuri- (Davydov, 1939). Because of Ioffe’s initiative in the late
ties and the idea of injecting nonequilibrium charge car- 1940s at the Physico-Technical Institute, research into
riers could be said to be the seeds from which semicon- intermetallic compounds was begun. Theoretical predic-
ductor electronics developed. Heterostructures tion of semiconductor properties in A 3 B 5 compounds
developed from these beginnings, making it possible to and their subsequent experimental discovery were done
solve the considerably more general problem of control- independently by H. Welker and (on the example of
ling the fundamental parameters inside the semiconduc- InSb) N. A. Gorunova and A. R. Regel at the Physico-
Technical Institute (Goryunova, 1951; Blum et al., 1952;
tor crystals and devices: band gaps, effective masses of
Welker, 1953). We benefited a lot from the high degree
the charge carriers and the mobilities, refractive indices,
of theoretical, technological, and experimental expertise
electron energy spectrum, etc.
in this area at the Ioffe Institute at that time.
Development of the physics and technology of semi-
conductor heterostructures has resulted in remarkable
changes in our everyday life. Heterostructure electronics II. CLASSICAL HETEROSTRUCTURE
are widely used in many areas of human civilization. It is
hardly possible to imagine our recent life without double The idea of using heterojunctions in semiconductor
heterostructure (DHS) laser-based telecommunication electronics was put forward at the very dawn of the elec-
systems, heterostructure-based light-emitting diodes tronic era. In the first patent concerned with p-n junc-
(LED’s), heterostructure bipolar transistors, or low- tion transistors, Shockley (1951) proposed a wide-gap
noise high-electron-mobility transistors for high- emitter to obtain unidirectional injection. A. I. Gubanov
frequency applications including, for example, satellite at our Institute first theoretically analyzed current-
television. Double-heterostructure lasers now enter voltage characteristics of isotype and anisotype hetero-
practically every house with CD players. Heterostruc- junctions (Gubanov, 1950, 1951) but the important
ture solar cells have been widely used for space and ter- theoretical considerations at this early stage of hetero-
restrial applications. structure research were put forward by H. Kroemer,
Our interest in semiconductor heterostructures was who introduced the concept of quasielectric and quasi-
not occasional. Systematic studies of semiconductors magnetic fields in a graded heterojunction and made an
were started in the early 1930s at the Physico-Technical assumption that heterojunctions might exhibit extremely
Institute under the direct leadership of its founder, high injection efficiencies in comparison to homojunc-
Abraham Ioffe. V. P. Zhuze and B. V. Kurchatov studied tions (Kroemer, 1957a, 1957b). In the same period there
the intrinsic and impurity conductivity of semiconduc- were various suggestions about applying heterostruc-
tors in 1932, and the same year Ioffe and Ya. I. Frenkel tures in semiconductor solar cells.
created a theory of rectification in a metal- The proposal of p-n junction semiconductor lasers
semiconductor contact based on the tunneling phenom- (Basov et al., 1961), the experimental observation of ef-
fective radiative recombination in GaAs p-n structure
with a possible stimulated emission (Nasledov et al.,
*The 2000 Nobel Prize in Physics was shared by Zhores I. 1962), and the creation of p-n junction lasers and LED’s
Alferov, Jack Kilby, and Herbert Kroemer. This lecture is the (Hall et al., 1962; Holonyak and Bevacgua, 1962; Nathan
text of Professor Alferov’s address on the occasion of the et al., 1962) were the seeds from which semiconductor
award. optoelectronics started to grow. However, lasers were

0034-6861/2001/73(3)/767(16)/$23.20 767 © The Nobel Foundation 2000

768 Zhores I. Alferov: The double heterostructure concept

not efficient because of high optical and electrical losses.

The threshold currents were very high, and low tem-
perature was necessary for lasing. The efficiency of
LED’s was very low, as well, due to high internal losses.
The important step was made immediatly after the
creation of p-n junction lasers when the concept of the
double heterostructure laser was formulated indepen-
dently by us and Kroemer (Alferov and Kazarinov, 1963;
Kroemer, 1963). In his article Kroemer proposed to use
the double heterostructures for carrier confinement in
the active region. He proposed that ‘‘laser action should
be obtainable in many of the indirect gap semiconduc-
tors and improved in the direct gap ones, if [it] is pos-
sible to supply them with a pair of heterojunction
injectors.’’
In our patent we also outlined the possibility of
achieving a high density of injected carriers and inverse
population by ‘‘double’’ injection. We especially pointed
out that homojunction lasers ‘‘do not provide cw at el- FIG. 1. Main physical phenomena in classical heterostructures:
evated temperatures,’’ and an additional advantage of (a) One-side injection and superinjection; (b) diffusion in
DH lasers that we considered was the possibility ‘‘to en- built-in quasielectric field; (c) electron and optical confine-
large the emitting surface and to use new materials in ment; (d) wide-gap window effect; (e) diagonal tunneling
various regions of the spectrum.’’ through a heterostructure interface.
Initially the theoretical progress was much faster than
experimental realization. In 1966 (Alferov et al., 1966), double and single classical heterostructures are shown in
we predicted that the density of injected carriers could Fig. 1. Then it was only necessary to find heterostruc-
exceed the carrier density in a wide-gap emitter by sev- tures in which these phenomena could be realized.
eral orders of magnitude (the ‘‘superjunction’’ effect). At that time general skepticism existed with respect to
The same year, in a paper submitted to a new Soviet the possibility of creating the ‘‘ideal’’ heterojunction
journal, Fizika i Tekhnika Poluprovodnikov (Soviet with a defect-free interface and first of all with theoret-
Physics Semiconductors), I summarized our understand- ical injection properties. Even a very pioneering study of
ing of the main advantages of the double heterostruc- the first lattice-matched epitaxially grown single-crystal
ture for different devices, especially for lasers and high- heterojunctions Ge-GaAs by R. L. Anderson (1960,
power rectifiers: 1962) did not give any proof of the injection of nonequi-
librium carriers in heterostructures. Actual realization of
‘‘The recombination, light-emitting, and population an efficient wide-gap emitter was considered to be next
inversion zones coincide and are concentrated in the to impossible, and the patent for the double-
middle layer. Due to potential barriers at the bound- heterostructure laser was often referred to as a ‘‘paper
aries of semiconductors having forbidden bands of dif- patent.’’
ferent width, the through currents of electrons and Mostly due to this general skepticism there existed
holes are completely absent, even under strong for- only a few groups trying to find the ‘‘ideal couple,’’
ward voltages, and there is no recombination in the which was, naturally, a difficult problem. Many condi-
emitters (in contrast to p-i-n, p-n-n ⫹ , n-p-p ⫹ ho- tions of compatibility needed to be met between ther-
mostructures, in which the recombination plays the mal, electrical, and crystallochemical properties and be-
dominant role) . . . . Because of a considerable differ- tween the crystal and the band structure of the
ence between the permittivities, the light is com- contacting materials.
pletely concentrated in the middle layer, which acts as A lucky combination of a number of properties, i.e., a
a high-grade waveguide, and thus there are no light small effective mass and wide energy gap, effective ra-
losses in the passive regions (emitters)’’ (Alferov, diative recombination, and a sharp optical absorption
1966). edge due to the ‘‘direct’’ band structure, a high mobility
Here are the most important peculiarities of semicon- at the absolute minimum of the conduction band, and its
ductor heterostructures we emphasized at that time: (i) strong reduction of the nearest minimum at the (100)
superinjection of carriers, (ii) optical confinement, and point ensured for GaAs even at that time a place of
(iii) electron confinement. honor in semiconductor physics and electronics. Since
The realization of the wide-gap window effect was the maximum effect is obtained by using heterojunctions
very important for photodetectors, solar cells, and LED between the semiconductor serving as the active region
applications. It permitted us to broaden considerably and a more wideband material, the most promising sys-
and to control precisely the spectral region for solar cells tems looked at in that time were GaP-GaAs and AlAs-
and photodetectors and to improve drastically the effi- GaAs. To be compatible, materials of the ‘‘couple’’
ciency of LED’s. The main physical phenomena in should have, as the first and the most important condi-

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

 Zhores I. Alferov: The double heterostructure concept 769

tion, close values of the lattice constants; therefore het-

erojunctions in the system AlAs-GaAs were preferable.
However, prior to starting work on the preparation and
study of these heterojunctions one had to overcome a
certain psychological barrier. AlAs had been synthesized
long ago (Natta and Passerini, 1928; Goldschmidt, 1929),
but many properties of this compound remained unstud-
ied, since AlAs was known to be chemically unstable
and to decompose in moist air. The possibility of prepar-
ing stable and adequate applications of heterojunctions
in this system seemed to be not very promising.
Initially, our attempts to create double heterostruc-
tures were related to a lattice-mismatched GaAsP sys-
tem. And we succeeded in fabricating by vapor-phase
epitaxy the first DHS lasers in this system. However, due
to lattice mismatch, the lasing, like that in homojunction
lasers, occurred only at liquid-nitrogen temperatures
(Alferov, Garbuzov, et al., 1967). I would like to mention
that, curiously, it was the first practical result obtained
for a lattice-mismatched, even partially relaxed, system.
Our experience, which we got from studying the FIG. 2. Emission spectrum of the first low-threshold Alx Ga1⫺x
GaAsP system, was very important for understanding double heterostructure 1 J th⫽4300 A/cm2. The current rises
many specific heterojunction physical properties and the (a) from (1) 0.7 A to (2) 8.3 A and then to (3) 13.6 A; s⫽2.2
basics of heteroepitaxy. The development of the multi- ⫻10⫺3 cm2.
chamber vapor-phase epitaxy method for the GaAsP
system permitted us to create in 1970 superlattice struc- Alx Ga1⫺x As-GaAsx heterojunction, and carefully stud-
tures with a 200-Å period and to demonstrate the split- ied luminescence properties, diffusion of carriers in a
ting of the conduction band (Alferov, Zhilyaev, and graded heterostructure, and very interesting peculiari-
Shmartsev, 1971). ties of the current flow through the heterojunction. The
But from the general point of view at the end of 1966 current flow is similar, for instance, to diagonal
we came to a conclusion that even a small lattice mis- tunneling-recombination transitions directly between
match in heterostructures GaP0.15As0.85-GaAs did not holes of the narrow-band and electrons of the wide-band
permit us to realize potential advantages of the double heterojunction components (Alferov, Andreev,
heterostructure. At that time my co-worker D. N. Korol’kov, Portnoi, and Tret’yakov, 1969; Alferov, An-
Tret’yakov told me that some small crystals of dreev, Korol’kov, Portnoi, and Yakovenko, 1969a; Alf-
Alx Ga1⫺x As solid solutions of different compositions, erov, Garbuzov, et al., 1969; Alferov, 1970).
which had been prepared two years ago by cooling from At the same time, we created some important devices
a melt, were put in a desk drawer by Dr. A. S. Bort- that realized the main advantages of the heterostructure
shevsky and nothing happened to them. It immediately concepts:
became clear that Alx Ga1⫺x As solid solutions turned —DHS lasers with low threshold at room tempera-
out to be chemically stable and suitable for the prepara- ture. (Fig. 2) (Alferov, Andreev, Portnoy, and Trukan,
tion of durable heterostructures and devices. Studies of 1969)
phase diagrams and the growth kinetics in this system —Highly effective LED’s using semiconductor hetero-
and development of the liquid-phase epitaxy method, structures and double heterostructures (Alferov, An-
especially for heterostructure growth, soon resulted in dreev, Korol’kov, Portnoi, and Yakovenko, 1969b)
fabrication of the first lattice-matched AlGaAs hetero- —Heterostructure solar cells (Alferov, Andreev, Ka-
structures. When we published the first paper on this gan, et al., 1970)
subject, we felt lucky to be the first to find a unique, —Heterostructure bipolar transistor (Alferov, Ahme-
practically ideal lattice-matched system for GaAs, but as dov, et al., 1973)
frequently happened, simultaneously and independently —Heterostructure p-n-p-n switching devices (Alf-
the same results were achieved by H. Rupprecht and J. erov, Andreev, Korol’kov, Nikitin, and Yakovenko,
Woodall at the T. Watson IBM Research Center (Alf- 1970).
erov, Andreev, et al., 1967; Rupprecht et al., 1967). One of the first successful applications in industrial-
From then on, progress in the semiconductor hetero- scale production in our country was heterostructure so-
structure area was very rapid. First of all, we experimen- lar cells in space research. We transferred our technol-
tally proved the unique injection properties of the wide- ogy to the Quant company and, since 1974, GaAlAs
gap emitters and superinjection effect (Alferov, solar cells have been installed on many of our sputniks.
Andreev, et al., 1968a) and the stimulated emission in Our space station Mir (Fig. 3) used them for 15 years.
AlGaAs double heterostructures (Alferov, Andreev, Most of these results were achieved afterwards in
et al., 1968b), established the band diagram of the other laboratories in one to two years and in some cases

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

770 Zhores I. Alferov: The double heterostructure concept

FIG. 5. Energy gaps vs lattice constant for III-V semiconduc-

tors. Lattice-matched heterojunctions: Ge-GaAs—1959. From
Anderson, 1960, 1962. AlGaAs—1967. From Alferov et al.,
FIG. 3. Space station Mir equipped with heterostructure solar
1967; Rupprecht et al., 1967. Quaternary heterostructure (In-
cells.
GaAsP and AlGaAsSb): Proposal—1970. From Alferov, An-
dreev, Konnikov, et al., 1971; First experiment—1972. From
even later. But in 1970 the international competition be-
Antipas et al., 1973.
came very strong. Later on, one of our main competi-
tors, Izuo Hayashi, who was working together with M. nology for applications in the USSR; Bell Telephone,
Panish at Bell Telephone Laboratories in Murray Hill, the D. Sarnoff RCA Research Center, and T. J. Watson
wrote: IBM Research Center in the U.S.), by the beginning of
‘‘In September 1969 Zhores Alferov of the Ioffe Insti- 1971 many universities and industrial labs in the U.S.,
tute in Leningrad visited our laboratory. We realized the USSR, the United Kingdom, Japan, and even Brazil
(300) and Poland had begun investigations of III-V hetero-
he was already getting a J th of 4.3 kA/cm2 with a
structures and heterostructure devices.
DH. We had not realized that the competition was so
At this early stage in the development of heterostruc-
close and redoubled our efforts . . . Room-
ture physics and technology it became clear that we
temperature cw operation was reported in May
needed to look for new lattice-matched heterostructures
1970 . . . ’’ (Hayashi, 1984).
in order to cover a broad area of the energy spectrum.
In our paper published in 1970 (Alferov, Andreev, The first important step was taken in our laboratory in
Garbuzov, et al., 1970), cw lasing was realized in stripe- 1970: in our paper (Alferov, Andreev, Konnikov, et al.,
geometry lasers formed by photolithography and 1971) we reported that various lattice-matched hetero-
mounted on copper plates covered by silver (Fig. 4). The junctions based on quaternary III-V solid solutions were
lowest J th density at 300 K was 940 A/cm2 for broad-area possible, which permitted independent variation be-
lasers and 2.7 kA/cm2 for stripe lasers. Independently, tween lattice constant and band gap. Later on G. Anti-
cw operation in DHS lasers was reported by Hayashi pas and co-workers came to the same conclusions (An-
and Panish (Hayashi et al., 1970; for broad-area lasers tipas et al., 1973). As a practical example utilizing this
with diamond heat sinks) in a paper submitted only one idea we considered different InGaAsP compositions,
month later than our work. Achievement of continuous and soon this material was recognized as being among
waves at room temperature produced an explosion of the most important ones, for many different practical
interest in the physics and technology of semiconductor applications, including photocathodes (James et al.,
heterostructures. Whereas in 1969 AlGaAs heterostruc- 1973) and especially lasers in the infrared region for
tures were studied in just a few laboratories, mostly in fiber-optic communications (Bogatov et al., 1974) and
the USSR and U.S. (A. F. Ioffe Institute, Polyus, and the visible (Alferov, Arsent’ev, et al., 1975a, 1975b;
Quant—industrial labs where we transferred our tech- Hitchins et al., 1975).
The early 1970s ‘‘world map’’ of ideal lattice-matched
heterostructures is shown in Fig. 5. Only a decade later
this ‘‘world map’’ was drastically changed (Fig. 6).
Nowadays, it is necessary to add III nitrides.
The main ideas of a semiconductor distributed-
feedback laser were formulated by us in our 1971 patent
(Alferov, Andreev, Kazarinov, et al., 1971). The same
year H. Kogelnilk and C. V. Shank considered the pos-
sibility of replacing the Fabry-Perot or similar types of
resonator in dye lasers with volume periodic inhomoge-
neities (Kogelnik and Shank, 1971). It is necessary to
FIG. 4. Schematic view of the structure of the first injection note that their approach is not applicable to semicon-
DHS laser operating in the cw regime at room temperature. ductor lasers, and all laboratories that carried out re-

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

 Zhores I. Alferov: The double heterostructure concept 771

practical applications of these structures are still ham-

pered by a poor understanding of their fundamental
properties and scarcity of actual systems, which have
been studied experimentally up to now.
To summarize this brief review of classical hetero-
structure development, we can classify the most impor-
tant as follows:

CLASSICAL HETEROSTRUCTURES
I. Fundamental physical phenomena (Fig. 1)
• One-side injection
• Superinjection
• Diffusion in built-in quasielectric fields
• Electron confinement
• Optical confinement
FIG. 6. Energy gaps vs lattice constants for semiconductors of • Wide-gap window effect
IV elements, III-V and II(IV)-VI compounds, and magnetic • Diagonal tunneling through heterostructure inter-
materials in parentheses. Lines connecting the semiconductors, face
solid for the III-V’s and dashed for the others, indicate quan- II. Important applications in electronics
tum heterostructures that have been investigated.
• Semiconductor lasers—Low threshold and continuous
waves at room temperature, distributed-feedback and
search in distributed-feedback and distributed-Bragg- distributed-Bragg-reflector lasers, vertical surface emit-
reflector semiconductor lasers used the ideas formulated ting lasers, IR type-II heterostructure lasers
in Alferov, Andreev, Kazarinov, et al. (1971): • High-efficiently LED’s
(1) Diffraction grating created not in volume, but on a • Solar cells and photodetectors, based on wide-gap
surface waveguide layer. window effect
(2) Interaction of waveguide modes with a surface dif- • Semiconductor integrated optics, based on semicon-
fraction grating, which gives not only distributed feed- ductor distributed-feedback and distributed-Bragg-
back but also highly collimated light output. reflector lasers
A detailed theoretical analysis of the semiconductor • Bipolar wide-gap transistors
laser with surface diffraction grating was published in • Transistors, thyristors, and dynistors with photonic
1972 (Kazarinov and Suris, 1972b). In this paper the au- signal transmission
thors established the way to obtain single-frequency • High-power diodes and thyristors
generation. The first semiconductor lasers using surface • Infrared to visible converters
diffraction gratings and distributed feedback were real- • Efficient cold cathodes
ized practically simultaneously at the Physico-Technical
Institute (Alferov et al., 1974, 1975), Caltech (Nakamura III. Important technological peculiarities
et al., 1973), and Xerox Laboratory in Palo Alto (Scifres • Lattice-matched structures are necessary in prin-
et al., 1974). ciple
In the early 1980s Kroemer and G. Griffiths (1983) • Multicomponent solid solutions are used for lattice
published a paper that stimulated strong interest in matching
staggered-lineup heterostructures (type-II heterojunc- • Epitaxial growth technology is needed in principle
tion). Spatial separation of electrons and holes at the Concluding this concise summary of the early devel-
interface results in tunability of their optical properties opment of bulk heterostructures, one may say that the
(Alferov, Garbuzov, et al., 1969; Baranov et al., 1986). invention of an ‘‘ideal’’ heterojunction and the introduc-
Staggered band alignment makes it possible to realize tion of the heterostructure concept into semiconductor
optical emission with a photon energy much smaller physics and technology have led to the discovery of new
than the band-gap energy of each of the semiconductors physical effects, pivotal improvement in the characteris-
forming a heterojunction. The demonstration of an in- tics of practically all known semiconductor devices, and
jection laser based on a type-II GaInAsSb-GaSb hetero- the invention of new ones.
junction (Baranov et al., 1986) showed promise for the
creation of effective coherent light sources in the infra-
red optical range. Radiation in such a device is due to III. HETEROSTRUCTURE QUANTUM WELLS AND
the recombination of electrons and holes localized in SUPERLATTICES
self-consistent potential wells at different sides of the
heterointerface. Thus type-II heterostructures open pos- Owing to electron confinement, the double-
sibilities both for fundamental physics and for device heterostructure laser became an important precursor of
applications, which cannot be realized with type-I het- the quantum well structure: when a middle layer had a
erostructures in the III-V material systems. However, thickness of some hundred angstroms, the electron lev-

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

772 Zhores I. Alferov: The double heterostructure concept

els would split due to the quantum size effect. The de- pendence, and photoconductivity were explained by the
velopment of heterostructure growth techniques made it splitting of the conduction band due to the one-
possibile to fabricate high-quality-double heterostruc- dimensional periodic potential of the superlattice. These
tures with ultrathin layers. Two main methods of growth first superlattices were also the first strained-layer super-
with very precise control of thickness, planarity, compo- lattices. E. Blakeslee and J. Matthews, who were work-
sitions, etc. were developed in the 1970s. A modern ing with Esaki and Tsu at IBM, succeeded in the mid-
molecular-beam epitaxy method became practically im- 1970s in growing strained-layer superlattices with a very
portant for III-V heterostructure technology due first of low concentration of defects. But many years later, after
all to the pioneering work of A. Cho (1971a, 1971b). G. Osbourn’s (1982) theoretical study at Sandia Lab and
Metallo-organic chemical vapor deposition originated the first successful preparation of a high-quality
from the early work of H. Manasevit (1968) and found strained-layer superlattice GaAs-In0.2Ga0.8As by M. Lu-
broad application in III-V heterostructure research after dowise at Varian, N. Holonyak at the University of Illi-
R. Dupuis and P. Dapkus (1977) reported the room- nois achieved on those structures a cw room-
temperature injection of AlGaAs DH lasers which had temperature laser action (Ludowise et al., 1983). It
been grown by the metal-organic chemical-vapor depo- became clear that in a strained-layer superlattice the lat-
sition method. tice strain became an additional degree of freedom, and
Clear manifestation of the quantum size effect in op- by varying the layer thicknesses and compositions one
tical spectra of GaAs-AlGaAs semiconductor hetero- could vary continuously and independently of one an-
structures with ultrathin GaAs layers (quantum wells) other the forbidden gap, lattice constant, of the overall
was demonstrated by Raymond Dingle et al. (1974). The superlattice.
authors observed a characteristic steplike behavior in In the early 1970s, Esaki and co-workers moved to
absorption spectra and systematic shifts of the character- molecular-beam epitaxy technology in AlGaAs systems
istic energies with a quantum well width decrease. (Chang et al., 1973), and in March 1974 they submitted a
Studies of superlattices were launched by the work of paper on resonant tunneling (Chang et al., 1974). It was
L. Esaki and R. Tsu (1970), who considered the electron the first experimental demonstration of quantum well
transport in a superlattice, i.e., at an additional periodic heterostructure physics. They measured the tunneling
potential created by doping or changing the composition current and conductance as a function of an applied
of semiconductor materials with the period bigger than, voltage in GaAs-GaAlAs double barriers and found cur-
but comparable to, the lattice constant of a crystal. In rent maxima associated with resonant tunneling. Later
this ‘‘man-made crystal,’’ as Esaki called it, a parabolic in the same year Esaki and Chang (1974) observed reso-
band would break into minibands separated by small nant tunneling in a superlattice. The strong interest in
forbidden gaps and having Brillouin zones determined resonant tunneling obviously was connected with its po-
by this period. Similar ideas were described by L. V. tential applications in high-speed electronics. In the late
Keldysh (1962) when considering the periodic potential 1980s picosecond operation was achieved in a double-
produced on a semiconductor surface by an intense ul- resonant tunnel diode, and oscillations up to 420 GHz
trasonic wave. At the Physico-Technical Institute, R. Ka- were reported in a GaAs resonant tunnel diode at room
zarinov and R. Suris theoretically considered the current temperature.
flow in superlattice structures in the early 1970s (Ka- The restriction of electron motion to two dimensions
zarinov and Suris, 1971, 1972a, 1973). It was shown that in field-effect transistors had been recognized long ago
the current between wells determined by tunneling (Shriffer, 1957) and that for trapped electrons in inver-
through the potential barriers separated the wells, and sion layers was first verified in a magnetoconductance
the authors predicted very important phenomena: tun- experiment by A. B. Fowler et al. (1966). Spectral effects
neling under the electric field when the ground state of a due to spatial quantization were observed in thin bis-
well coincided with the excited state of the next well, muth films in 1968 by V. N. Lutskii and L. A. Kulik
and stimulated emission resulting from photon-assisted (Lutskii, 1970).
tunneling between the ground state of one well and the The pioneering work of Dingle et al. (1978) on
excited state of a neighboring well, which is lower in modulation-doped superlattices demonstrated a mobil-
energy due to an applied electric field. At that time ity enhancement with respect to the bulk crystal and
Esaki and Tsu independently considered resonant tun- stimulated research on the application of high-mobility
neling in superlattice structures (Tsu and Esaki, 1973). two-dimensional electron gas for microwave amplifica-
The pioneering experimental studies of the superlat- tion. In France and Japan practically simultaneously
tice structures were carried out by Esaki and Tsu: the new types of transistors based on a single
superlattices were grown by vapor-phase epitaxy in the nAlGaAs-nGaAs modulation-doped heterostructure
system GaPx As1⫺x -GaAs. At the same time, in our were created that were labeled TEGFET’s (two-
laboratory we developed the first multichamber appara- dimensional electron-gas field-effect transistors) in
tus and, as was mentioned before, prepared a superlat- France (Delagebeaudeuf et al., 1980) and HEMT’s
tice structure GaP0.3Al0.7-GaAs with the thickness of (high-electron-mobility transistors) in Japan (Mimura
each layer 100 Å and a total of 200 layers (Alferov, Zhi- et al., 1980).
lyaev, and Shmartsev, 1971). Observed peculiarities of The first quantum well laser operation was demon-
the current-voltage characteristics, their temperature de- strated by J. P. van der Ziel et al. (1975), but the lasing

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

 Zhores I. Alferov: The double heterostructure concept 773

parameters were much worse than for average DHS la-

sers. In 1978 Dupuis and Dapkus, in collaboration with
Holonyak, first reported on a quantum well laser with
parameters comparable to those of conventional DHS
lasers (Dupuis et al., 1978). The name ‘‘quantum well’’
was used in that paper. The real advantages of quantum
well lasers were demonstrated much later by W. T. Tsang
at Bell Telephone Laboratories. Thanks to many im-
provements in molecular-beam epitaxy growth technol-
ogy and the introduction of an optimized structure
(GRIN SCH), he found threshold currents as low as
160 A/cm2 (Tsang, 1982).
We started to develop molecular-beam epitaxy and
FIG. 7. Time evolution of a double heterostructure active re-
metal-organic chemical vapor deposition methods for
gion under high-level photoexcitation: (a) AlGaAs/GaAs, (b)
growing III-V heterostructures only in the late 1970s. InGaAsP/GaAs. Diameter of Kr⫹-laser excitation beam ⫺40
First of all, we stimulated the design and construction of ␮m. Excitation level (a) 104 W/cm2, (b) 105 W/cm2.
the first Soviet molecular-beam epitaxy machine in our
electronic industry. In a few years three generations of tance for InGaAsP laser heterostructures was the cre-
these machines were developed, and the last, which had ation of a record threshold current density for InGaAsP/
the name ‘‘Cna,’’ was good enough for our goals. (It was InP (␭⫽1.3 and 1.55 ␮m) and for InGaAsP/GaAs (␭
named for the nice river not very far from Ryazan, the ⫽0.65– 0.9 ␮ m) single quantum well separate confine-
city where NITI—the industrial laboratory of the Elec- ment lasers (Alferov et al., 1986b, 1987). For high-power
tronic Industry—was located; NITI carried out develop- InGaAsP/GaAs (␭⫽0.8 ␮ m) lasers a total efficiency of
ment of the molecular-beam epitaxy machine.) In paral- 66% with cw power 5 W for 100 ␮m width, a stripe-
lel, later on, we started to develop a molecular-beam geometry laser was achieved (Alferov et al., 1988a; Gar-
epitaxy system with NTO AN—the scientific instru- buzov et al., 1988). In this laser the effective cooling of a
ments company of the Academy of Sciences in Lenin- semiconductor power device by recombination radiation
grad. In the mid-1980s we got a few systems of this ver- was for the first time realized as had been predicted
sion. Both types of molecular-beam epitaxy systems are much earlier (Alferov, 1966). Garbuzov et al. (1990)
still working at the Ioffe Institute and other noted an important characteristic of the InGaAsP het-
laboratories. erostructure, its unusual resistance to multiplication of
We developed the metal-organic chemical vapor dislocations and defects (Fig. 7). It was this research
deposition systems in our Institute and later, in the that initiated the broad application of Al-free hetero-
1980s, a Swedish company, Epiquip, specially designed structures.
with our participation two systems for our Institute, A most complicated quantum well laser structure
which are still used in our research. (Fig. 8), which combined a single quantum well with
The strong interest in the experimental study of low-
dimensional structures and the lack of equipment for
molecular beam epitaxy and metal-organic chemical-
vapor deposition growth technology stimulated our re-
search on the development of liquid-phase epitaxy suit-
able for quantum well heterostructures.
However, until the late 1970s it seemed impossible to
grow III-V heterostructures with an active-region thick-
ness of less than 500 Å by liquid-phase epitaxy because
of the existence near the heterojunction of extended in-
terface regions with varying chemical compositions.
The situation was changed due to the work of Holo-
nyak et al. (Rezek et al., 1977) for superlatticelike
InGaAsP structures, using a rotating boat system. In our
laboratory we developed a new liquid-phase epitaxy
method with the usual translational motion in a standard
horizontal system for InGaAsP heterostructures (Alf-
erov et al., 1985) and a low-temperature liquid-phase ep-
itaxy method for AlGaAs heterostructures (Alferov
et al., 1986a). These methods permitted us to prepare
practically any kind of excellent-quality quantum well
heterostructures with an active-region thickness of up to
20 Å and with the size of the interface regions compa- FIG. 8. Quantum well heterostructure laser, using short-period
rable to one lattice constant. Of great practical impor- superlattices grown by molecular-beam epitaxy.

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

774 Zhores I. Alferov: The double heterostructure concept

densities in a semiconductor, has shown that hetero-

structures can be used to model some very basic physical
effects. Recently many studies in this area have concen-
trated on understanding the condensation of electrons
and the search for Wigner crystallization.
To summarize, the development of the field of hetero-
structure quantum wells and superlattices can be out-
lined in the same manner as that of classical heterostruc-
ture research:

HETEROSTRUCTURE QUANTUM WELLS AND

SUPERLATTICES
I. Fundamental physical phenomena
• 2D electron gas
FIG. 9. Evolution of the threshold current of semiconductor • Steplike density-of-state function
lasers. • Quantum Hall effect
short-period superlattices, for the creation of GRIN • Fractional quantum Hall effect
SCH semiconductor heterostructures (the most favor- • Excitons at room temperature
able for the lowest J th) was demonstrated in our labora- • Resonant tunneling in double-barrier structure and
tory in 1988 (Alferov et al., 1988b). Using short-period superlattices
superlattices, we achieved not only the desirable profile • Energy spectrum in superlattices determined by
of a graded waveguide region, thus creating a barrier for choice of potential and strain
dislocation movement to the active layer, but also the • Stimulated emission at resonant tunneling in super-
possibility of growing different parts of the structure at lattices
sufficiently large differences in temperature. In this way, • Pseudomorphic growth of strained structures
we obtained both an excellent surface morphology and a II. Important consequences for applications
high internal quantum efficiency on a planar GaAs (100) • Shorter emission wavelength, reduced threshold cur-
surface. The lowest J th⫽52 A/cm2 and, after some small rent, larger differential gain, and reduced temperature
optimization, 40 A/cm2 was for a long time a world dependence of the threshold current for semiconductor
record for semiconductor injection lasers and a good lasers
demonstration of the application of quantum wells and • Infrared quantum cascade laser
superlattices in electronic devices. • Short-period superlattice quantum well laser
The idea of stimulated emission in superlattices which • Optimization of electron and light confinement and
had been suggested by Kazarinov and Suris (1971, waveguiding for semiconductor lasers
1972a, 1973) was realized nearly a quarter of a century • 2D electron-gas transistors (high-electron-mobility
later after a proposal by Federico Capasso (Faist et al., transistors)
1994). The proposed structure was strongly improved, • Resonant-tunneling diodes
and a cascade laser developed by Capasso gave rise to a • Precise resistance standards
new generation of unipolar lasers operating in the • SEED’s and electro-optical modulators
middle-infrared range. • Infrared photodetectors based on quantum size level
The history of semiconductor lasers is, from a certain absorption
point of view, the history of the evolution of the semi-
conductor laser current threshold, which is shown in Fig. III. Important technological peculiarities
9. The most dramatic changes took place just after the • Lattice match unnecessary
introduction of the DHS concept. The impact of short- • Low-growth-rate technology (MBE, MOCVD)
period superlattice quantum wells led to a theoretical needed in principle
limit on this most important parameter. What can hap- • Submonolayer growth techniques used
pen as a result of application of the new quantum wires • Blockading of mismatch dislocations during epitax-
and quantum dot structures will be discussed in the next ial growth
part of our paper. • Sharp increase in the variety of heterostructure
It may be that the crowning achievement of quantum components
well studies was the discovery of the quantum Hall ef-
fect (Klitzing et al., 1980). This discovery and its compre- IV. HETEROSTRUCTURE QUANTUM WIRES AND
hensive studies in AlGaAs-GaAs heterostructures, QUANTUM DOTS
which shortly led to the discovery of the fractional quan-
tum Hall effect (Tsui et al., 1982), had a huge impact on The principal advantage of using quantum-size het-
solid-state physics. Observation of the effect, which erostructures in lasers is the noticeable increase in the
deals only with fundamental quantities and does not rely density of states when the dimensionality of the electron
on peculiarities of the band structure, carrier mobility, or gas is reduced (Fig. 10).

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

 Zhores I. Alferov: The double heterostructure concept 775

FIG. 10. Density of states for charge carriers in structures with

different dimensionalities.

During the 1980s, progress in 2D quantum well het-

erostructure physics and its applications drew many sci-
entists to the study of systems of far less
dimensionality—quantum wires and quantum dots. In FIG. 12. Normalized temperature dependence of the threshold
contrast to quantum ‘‘wells,’’ where carriers are local- current for various DHS lasers: (a) bulk; (b) with quantum
ized in the direction perpendicular to the layers but wells; (c) with quantum wires; (d) with quantum dots.
move freely in the layer plane, in quantum ‘‘wires’’ car-
riers are localized in two directions and move freely are many theoretical and experimental papers in this
along the wire axis. With confinement in all three direc- field.
tions, quantum ‘‘dots’’—‘‘artificial atoms’’ with a totally The first semiconductor dots based on II-VI micro-
discrete energy spectrum—are created (Fig. 11). crystals in a glass matrix were proposed and demon-
Experimental work on the fabrication and investiga- strated by A. I. Ekimov and A. A. Onushchenko (1981).
tion of quantum wire and dot structures began more However, since the semiconductor quantum dots were
than 15 years ago. In 1982, Y. Arakawa and H. Sakaki introduced in an insulating glass matrix and the quality
(1982) theoretically considered some effects in lasers of the interface between glass and semiconductor dot
based on heterostructures with size quantization in one, was not high, both fundamental studies and device ap-
two, and three directions. They wrote: ‘‘Most important,
plications were limited. Much more exciting possibilities
the threshold current of such a laser is reported to be far
appeared after three-dimensional coherent quantum
less sensitive than that of [a] conventional laser reflect-
dots had been fabricated in a semiconductor matrix
ing the reduced dimensionality of [the] electronic state.’’
(Goldstein et al., 1985).
The authors performed experimental studies on a quan-
Several methods were proposed for the fabrication of
tum well laser placed in high magnetic fields directed
these structures. Indirect methods, such as postgrowth
perpendicular to the quantum well plane and demon-
lateral patterning of a 2D quantum well, often suffer
strated an increase in the characteristic temperature
from insufficient lateral resolution and interface damage
(T 0 ) describing the exponential growth of the threshold
caused by the patterning procedure. A more promising
current; the temperature increased in the magnetic field
approach is fabrication by direct methods, i.e., growth in
from 144 to 313 °C. They pointed to the possibility of
V grooves and on corrugated surfaces, which may result
weakening the threshold current dependence on tem-
in formation of quantum wires and dots. The groups at
perature for quantum wire lasers and full temperature
the Ioffe Institute and Berlin Technical University—who
stability for quantum dot lasers (Fig. 12). By now there
have carried out this research in close cooperation over
the past few years—have contributed significantly to the
latter type of fabrication.
These groups came to the conclusion that the most
exciting method for forming ordered arrays of quantum
wires and dots is use of self-organization phenomena on
crystal surfaces. Strain relaxation on step or facet edges
may result in the formation of ordered arrays of quan-
tum wires and dots for both lattice-matched and lattice-
mismatched growth. The first very uniform arrays of
three-dimensional quantum dots also exhibiting lateral
ordering were realized in the system InAs-GaAs by both
molecular-beam epitaxy and metal-organic chemical-
vapor deposition growth methods (Ledentsov et al.,
1995; Alferov, Gordeev, et al., 1996).
FIG. 11. Schematic representation of energy diagrams for a Elastic strain relaxation on facet edges and island in-
single atom (left), a bulk crystal (center), and a quantum dot teraction via the strained substrate are driving forces for
(right). the self-organization of ordered arrays of uniform, co-

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

776 Zhores I. Alferov: The double heterostructure concept

FIG. 13. Growth modes: (a) Frank–van der Merve; (b)

Volmer-Weber; (c) Stranski-Krastanow.

herently strained islands on crystal surfaces (Shchukin

et al., 1995). In lattice-matched heteroepitaxial systems
the growth mode is determined solely by the relation
between the energies of two surfaces and the interface
energy. If the sum of the surface energy of epitaxial layer
␥ 2 and the energy of the interface ␥ 12 is lower than the
substrate surface energy, ␥ 2 ⫹ ␥ 12⬍ ␥ 1 , i.e., if the mate-
rial being deposited wets the substrate, then we have
Frank–van der Merve growth. Changing the ␥ 2 ⫹ ␥ 12
value may result in a transition from the Frank–van der
Merve mode to a Volmer-Weber one in which 3D islands
are formed on a bare substrate.
In a heteroepitaxial system with lattice mismatch be-
tween the material being deposited and the substrate,
the growth may initially proceed in a layer-by-layer
mode. However, a thicker layer has a higher elastic en-
ergy, and the elastic energy tends to be reduced via the
FIG. 15. Vertical and transverse ordering of coupled quantum
formation of isolated islands. In these islands the elastic dots in the system InAs-GaAs.
strains relax and, correspondingly, the elastic energy de-
creases. This results in a Stranski-Krastanow growth
mode (Fig. 13). The characteristic size of islands is de- sequences. Short-period alternation in deposition of
termined by the minimum in the energy of an array of strained materials leads to a splitting of quantum dots
3D coherently strained islands per unit surface area as a and to formation of vertically coupled QD superlattice
function of the island size (Fig. 14) (Shchukin et al., structures (Fig. 15; Alferov, Bert, et al., 1996). Ground-
1995). Interaction between islands via an elastically state quantum dot emission, absorption, and lasing ener-
strained substrate would result in lateral island ordering gies are found to coincide (Ledentsov et al., 1995). The
typical of a square lattice. observation of ultranarrow (⬍0.15 meV) luminescence
Experiments show in most cases a rather narrow size lines from single quantum dots (Ledentsov et al., 1995)
distribution of the islands, and on top of that coherent which do not exhibit broadening with temperature, is
islands of InAs form under certain conditions a quasi- proof of the formation of an electronic quantum dot
periodic square lattice (Fig. 15). The shape of quantum (Fig. 16).
dots can be significantly modified during regrowth or Quantum dot lasers are expected to have properties
postgrowth annealing, or by applying complex growth superior to those of conventional quantum well lasers.
High differential gain, ultralow threshold current den-
sity, and high-temperature stability of the threshold cur-
rent density are expected to occur simultaneously. Addi-
tionally, ordered arrays of scatterers formed in an optical
waveguide region may result in distributed feedback
and/or in stabilization of single-mode lasing. Intrinsically
buried quantum dot structures spatially localize carriers
and prevent them from recombining nonradiatively at
resonator faces. The overheating of facets, which is one
of the most important problems for the high-power and
high-efficiency operation of AlGaAs-GaAs and
AlGaAs-InGaAs lasers, may thus be avoidable.
FIG. 14. Energy of a sparse array of 3D coherently strained
Since the first realization of QD lasers (Kirstaedter
islands per unit surface area as a function of island size. The et al., 1994), it has become clear that the QD size unifor-
parameter ␣ is the ratio between the change in the surface mity was sufficient to achieve good device performance.
energy upon island formation and the contribution from island But even at that time, it was recognized that the main
edges to the elastic relaxation energy. When ␣ ⬎1, the system obstacle for QD heterostructure laser operation at room
tends thermodynamically toward island coalescence. When ␣ and elevated temperatures was the temperature-induced
⬍1, there exists an optimal island size and the system of is- evaporation of carriers from quantum dots. Different
lands is stable against coalescence. methods were developed to improve the laser perfor-

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

 Zhores I. Alferov: The double heterostructure concept 777

FIG. 16. (a) High-resolution cathodoluminescence spectrum of

InAs-GaA; (b) temperature dependence of the full width at
half-maximum of the cathode peak.

mance: (i) the increase of the density of quantum dots

by stacking them (Fig. 17); (ii) the insertion of quantum
dots into a quantum well sheet; (iii) the use of a matrix
material with a higher band-gap energy. As a result, we
FIG. 18. (a) Schematic view of the quantum dot vertical-cavity
got many parameters of QD heterostructure lasers bet-
surface-emitting laser structure. Basic advantages of quantum
ter than those for quantum well heterostructure lasers dots: (1) no interface recombination at oxide-defined aper-
based on the same materials. As an example, the world- tures; (2) reduced lateral spreading of carriers out of the aper-
record threshold current density of 19 A/cm2 was re- ture region. Single quantum dot laser at ultralow threshold
cently achieved (Park et al., 2000). Further, a cw output current is possible. (b) Dependence of the threshold current on
power up to 3.5–4.0 W (cw) for a 100-␮m strip width, a the aperture size in a QD vertical-cavity surface-emitting laser.
quantum efficiency of 95% and a wall-plug efficiency of (i) Low threshold current densities (170 A/cm2 at 300 K); (ii)
50% were recently obtained (Zhukov et al., 1999). low threshold currents at ultrasmall apertures; (iii) 1.3-␮m
Significant efforts towards a theoretical understanding range on GaAs substrate.
of QD lasers with realistic parameters have been under-
taken. For a QD size dispersion of about 10% and other practical structure parameters, the theory (Asryan and
Suris, 1996) predicts typical threshold current densities
of 5 A/cm2 at room temperature. The values of 10 A/cm2
at 77 K (Zhukov et al., 1997) and even 5 A/cm2 at 4 K
(Park et al., 1999) have been experimentally observed.
In view of advanced device applications of quantum
dots, the incorporation of quantum dots in vertical-
cavity surface-emitting lasers seems to be very impor-
tant. QD vertical-cavity surface-emitting lasers with pa-
rameters that fit to the best values for quantum well
devices of similar geometry have been demonstrated
(Fig. 18; Lott et al., 1977). Recently, very promising re-
sults have been obtained (Fig. 19; Lott et al., 2000) for
1.3-␮m QD vertical-cavity surface-emitting lasers on a
GaAs substrate to use in fiber-optic communications.
In a free-standing 3D island formed on a lattice-
mismatched substrate, the strains can relax elastically,
without the formation of dislocations. Thus a sufficiently
FIG. 17. Transmission electron microscopy image of the active large volume of a coherent narrow-gap QD material can
region of high-intensity laser. be realized. This makes it possible to cover a spectral

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

778 Zhores I. Alferov: The double heterostructure concept

selenides and III-V nitrides. The success in this research

was mostly determined by the application of hetero-
structure concepts and methods of growth which had
been developed for III-V quantum wells and superlat-
tices. The natural and most predictable trend is the ap-
plication of heterostructure concepts as well as techno-
logical methods and peculiarities of new materials.
Different III-V, II-VI, and IV-VI heterostructures,
developed in recent times, are good examples of this
statement.
But from a more fundamental point of view, hetero-
structures (including quantum wells and superlattices,
quantum wires, and quantum dots) offer a way to create
new types of materials—hetero-semiconductors. In the
FIG. 19. GaAs-based QD vertical-cavity surface-emitting laser
words of Leo Esaki, instead of ‘‘God-made crystals’’ we
emitting at 1.3 ␮m.
create ‘‘man-made crystals.’’
range of 1.3–1.5 ␮m using a GaAs substrate and to de- The study of classical heterostructures, quantum wells
velop wavelength-multiplexing systems on the basis of and superlattices is quite mature and we are now ex-
QD vertical-cavity surface-emitting lasers in the future. ploiting many of their unique properties. The study of
It is very important to emphasize that we were able to quantum wire and dot structures is still very young: ex-
realize the DHS concept for quantum wire and quantum citing discoveries and new unexpected applications are
dot structures because in both cases we have a narrow awaiting us. Even at this early stage, however, we can
band-gap material in a wide-gap matrix.
say that ordered equilibrium arrays of quantum dots
Let us summarize again as we did for other parts.
may be used in many devices: lasers, light modulators,
far-infrared detectors and emitters, etc. Resonant tun-
HETEROSTRUCTURE QUANTUM WIRES AND DOTS neling via semiconductor atoms introduced in larger
I. Fundamental physical phenomena band-gap layers may lead to significant improvement in
• 1D electron gas (wires) device characteristics. More generally speaking, QD
• Density-of-state function with sharp maximums structures will be developed both ‘‘in width’’ and ‘‘in
(wires) depth.’’
• 0D electron gas (dots) In width means new material systems to cover a new
• ␦-function type of density-of-state function (dots) energy spectrum. The lifetime limitations of the green
• Increasing binding energy of excitons and blue semiconductor lasers and even more general
II. Important applications in electronics problems of the creation of defect-free structures based
on wide-gap II-VI and III-V (nitrides) would be solved
• Reduced lasing threshold current and larger differ-
by using QD structures in these systems.
ential gain
As to in depth, it is necessary to mention that the
• Reduced temperature dependence of threshold cur-
degree of ordering depends on very complicated growth
rent (wires)
conditions, materials constants, and concrete values of
• Temperature stability of the threshold current (dots)
the surface free energy. To achieve resonant tunneling
• Discrete amplification spectrum and a possibility of
and single-electron devices, including optical ones, re-
obtaining performance characteristics similar to those of
quire a thorough investigation and evaluation of these
solid-state or gas lasers (dots)
parameters in order to obtain the maximal possible de-
• Higher modulation factor in electro-optical modula-
gree of ordering. In general, it is necessary to find stron-
tors
ger self-organization mechanisms for ordered arrays of
• Possibility of creating ‘‘single-electron’’ devices
quantum dots.
• A new possibility for the development of field-effect
In the early 1980s I was invited to deliver a lecture
transistors
about heterostructures and applications at the Amoco
III. Important technological peculiarities Photonic Center near Chicago.
• The application of self-organization effects for The summary of my lecture was as follows:
growth (1) Heterostructures—a new kind of semiconductor
• Epitaxial growth in V grooves (wires) material—expensive, complicated chemically and tech-
• High-resolution lithography of heterostructure nologically, but most efficient
quantum well lasers (2) Modern optoelectronics is based on heterostruc-
ture applications:
V. FUTURE TRENDS —The DHS laser—a key device of modern opto-
electronics
Recently very impressive results for short-wavelength —The heterostructure protodiode—the most effi-
light sources have been achieved on the basis of II-VI cient and high-speed photodiode

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

 Zhores I. Alferov: The double heterostructure concept 779

—Optoelectronic integrated circuits (OEIC’s)—will ⫽100 A/cm2 , efficience⫽59%),’’ Fiz. Tekh. Poluprovodn. 22,
solve problem of high information density of an optical 1031 [Sov. Phys. Semicond. 22, 650 (1988)].
communication system Alferov, Zh. I., et al., 1988b, ‘‘Reducing of the threshold cur-
(3) Future high-speed microelectronics will mostly rent in GaAs-AlGaAs DHS SCH quantum well lasers (j th
use heterostructures ⫽52 A/cm2 , T⫽300 K) with quantum well restriction by
(4) High-temperature, high-speed power short period superlattice of variable period,’’ Pis’ma Zh.
electronics—a new broad field of heterostructure appli- Tekh. Fiz. 14, 1803 [Sov. Tech. Phys. Lett. 14, 782 (1988)].
cations Alferov, Zh. I., F. A. Ahmedov, V. I. Korol’kov, and V. G.
(5) Heterostructures in solar energy conversion: the Nikitin, 1973, ‘‘Phototransistor utilizing a GaAs-AlAs hetero-
most expensive photocells and the cheapest solar elec- junction,’’ Fiz. Tekh. Poluprovodn. 7, 1159 [Sov. Phys. Semi-
tricity producer cond. 7, 780 (1973)].
(6) In the 21st century heterostructures in electronics Alferov, Zh. I., V. M. Andreev, D. Z. Garbuzov, Yu. V. Zhi-
will reserve only 1% for homojunctions lyaev, E. P. Morozov, E. L. Portnoi, and V. G. Trofim, 1970,
And 20 years later I see no reason to change this sum- ‘‘Investigation of the influence of the AlAs-GaAs hetero-
mary. structure parameters on the laser threshold current and real-
It is hardly possible to describe even the main direc- ization of continuous emission at room temperature,’’ Fiz.
tions of modern physics and the technology of semicon- Tekh. Poluprovodn. 4, 1826 [Sov. Phys. Semicond. 4, 1573
ductor heterostructures. There is much more than I have (1971)].
mentioned. Many scientists contributed to this tremen- Alferov, Zh. I., V. M. Andreev, M. B. Kagan, I. I. Protasov, and
dous progress, which defines to a great extent not only V. G. Trofim, 1970, ‘‘Solar-energy converters based on p-n
the future prospects of condensed-matter physics and Alx Ga1⫺x As-GaAs heterojunctions,’’ Fiz. Tekh. Polupro-
semiconductor laser and communication technology, but vodn. 4, 2378 [Sov. Phys. Semicond. 4, 2047 (1971)].
also, in a sense, the future of human society. I would like Alferov, Zh. I., V. M. Andreev, R. F. Kazarinov, E. L. Portnoi,
also to emphasize the impact of scientists of previous and R. A. Suris, 1971, ‘‘Semiconductor optical quantum gen-
generations who prepared our way. I am very happy that erator,’’ Inventor’s Certificate No. 392875 [in Russian], Appli-
I had a chance to work in this field from the very begin- cation No. 1677436, priority as of July 19, 1971.
ning. I am even happier that we can continue to contrib- Alferov, Zh. I., V. M. Andreev, S. G. Konnikov, V. G. Nikitin,
ute to progress in this area now. and D. N. Tret’yakov, 1971, ‘‘Heterojunctions on the base of
A IIIB V semiconducting and of their solid solutions,’’ in Pro-
ceedings of the International Conference on Physics and
REFERENCES Chemistry of Semiconductor Heterojunctions and Layer Struc-
tures, Budapest, October 1970, edited by G. Szigeti (Aca-
Alferov, Zh. I., 1966, ‘‘Possible development of a rectifier for demiai Kiado, Budapest), Vol. 1, p. 93.
very high current densities on the bases of a p-i-n Alferov, Zh. I., V. M. Andreev, V. I. Korol’kov, V. G. Nikitin,
(p-n-n ⫹ ,n-p-p ⫹ ) structure with heterojunctions,’’ Fiz. Tekh. and A. A. Yakovenko, 1970, ‘‘p-n-p-n structures based on
Poluprovodn. 1, 436 [Sov. Phys. Semicond. 1, 358 (1967)]. GaAs and on Alx Ga1⫺x As solid solutions,’’ Fiz. Tekh. Polu-
Alferov, Zh. I., 1970, ‘‘Electroluminescence of heavily-doped provodn. 4, 578 [Sov. Phys. Semicond. 4, 481 (1971)].
heterojunctions p Alx Ga1⫺x -nGaAs,’’ J. Lumin. 1, 869. Alferov, Zh. I., V. M. Andreev, V. I. Korol’kov, E. L. Portnoi,
Alferov, Zh. I., et al., 1974, ‘‘Laser with supersmall divergence and D. N. Tret’yakov, 1968a, ‘‘Injection properties of
of radiation,’’ Fiz. Tekh. Poluprovodn. 8, 832 [Sov. Phys. n-Alx Ga1⫺x As-p-GaAs heterojunctions,’’ Fiz. Tekh. Polu-
Semicond. 8, 541 (1974)]. provodn. 2, 1016 [Sov. Phys. Semicond. 2, 843 (1969)].
Alferov, Zh. I., et al., 1975, ‘‘Semiconductor laser with distrib- Alferov, Zh. I., V. M. Andreev, V. I. Korol’kov, E. L. Portnoi,
uted feedback in second order,’’ Pis’ma Zh. Tekh. Fiz. 1, 645 and D. N. Tret’yakov, 1968b, ‘‘Coherent radiation of epitaxial
[Sov. Tech. Phys. Lett. 1, 286 (1975)]. heterojunction structures in the AlAs-GaAs system,’’ Fiz.
Alferov, Zh. I., et al., 1985, ‘‘Auger-profiles and luminescence Tekh. Poluprovodn. 2, 1545 [Sov. Phys. Semicond. 2, 1289
investigations of LPE grown InGaAsP-heterostructures with (1969)].
active regions (1.5– 5)⫻10⫺6 cm,’’ Fiz. Tekh. Poluprovodn. Alferov, Zh. I., V. M. Andreev, V. I. Korol’kov, E. L. Portnoi,
19, 1108 [Sov. Phys. Semicond. 19, 679 (1985)]. and D. N. Tret’yakov, 1969, ‘‘Recombination radiation in ep-
Alferov, Zh. I., et al., 1986a, ‘‘AlGaAs-heterostructure quan- itaxial structures in the system AlAs-GaAs,’’ in Proceedings
tum wells grown by low temperature LPE,’’ Pis’ma Zh. Tekh. of the Ninth International Conference on Semiconductor
Fiz. 12, 1089 [Sov. Tech. Phys. Lett. 12, 450 (1986)]. Structures, Moscow, July 23–29, 1968 [in Russian] (Nauka,
Alferov, Zh. I., et al., 1986b, ‘‘Low threshold InGaAsP/InP Leningrad), Vol. 1, p. 534.
separate confinement lasers ␭⫽1.3 ␮ m and ␭⫽1.55 ␮ m (j th Alferov, Zh. I., V. M. Andreev, V. I. Korol’kov, E. L. Portnoi,
⫽600– 700 A.cm2 ),’’ Pis’ma Zh. Tekh. Fiz. 12, 210 [Sov. Tech. and A. A. Yakovenko, 1969a, ‘‘Recombination radiation in
Phys. Lett. 12, 87 (1986)]. Alx Ga1⫺x As solid solutions with variable forbidden gap,’’ Fiz.
Alferov, Zh. I., et al., 1987, ‘‘Low threshold quantum well Tekh. Poluprovodn. 3, 541 [Sov. Phys. Semicond. 3, 460
InGaAsP/GaAs separate confinement double heterostructure (1970)].
lasers formed by the LPE (␭⫽0.86 ␮ m, j th⫽90 A/cm2 , L Alferov, Zh. I., V. M. Andreev, V. I. Korol’kov, E. L. Portnoi,
⫽⬁; j th⫽165 A/cm2 , L⫽1150 ␮ m, 300 K),’’ Fiz. Tekh. Polu- and A. A. Yakovenko, 1969b, ‘‘Spontaneous radiation
provodn. 21, 824 [Sov. Phys. Semicond. 21, 503 (1987)]. sources based on structures with AlAs-GaAs heterojunc-
Alferov, Zh. I., et al., 1988a, ‘‘Quantum-well InGaAsP/GaAs tions,’’ Fiz. Tekh. Poluprovodn. 3, 930 [Sov. Phys. Semicond.
(␭⫽0.86:0.78 ␮ m) separate confinement lasers (j th 3, 785 (1970)].

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

780 Zhores I. Alferov: The double heterostructure concept

Alferov, Zh. I., V. M. Andreev, V. I. Korol’kov, D. N. current,’’ Appl. Phys. Lett. 40, 939.
Trat’yakov, and V. M. Tuchkevich, 1967, ‘‘High-voltage p-n Asryan, L. V., and R. A. Suris, 1996, ‘‘Inhomogeneous line
junctions in Gax Al1⫺x As crystals,’’ Fiz. Tekh. Poluprovodn. 1, broadening and the threshold current density of a semicon-
1579 [Sov. Phys. Semicond. 1, 1313 (1968)]. ductor quantum dot laser,’’ Semicond. Sci. Technol. 11, 554.
Alferov, Zh. I., V. M. Andreev, E. L. Portnoi, and M. K. Tru- Baranov, A. N., et al., 1986, ‘‘Generation of coherent radiation
kan, 1969, ‘‘AlAs-GaAs heterojunction injection lasers with a in quantum-sized structure of the single heterojunction,’’ Fiz.
low room-temperature threshold,’’ Fiz. Tekh. Poluprovodn. 3, Tekh. Poluprovodn. 20, 2217 [Sov. Phys. Semicond. 20, 1385
1328 [Sov. Phys. Semicond. 3, 1107 (1970)]. (1986)].
Alferov, Zh. I., I. N. Arsent’ev, D. Z. Garbuzov, Basov, N. G., O. N. Krokhin, and Yu. M. Popov, 1961, ‘‘The
S. G. Konnikov, and V. D. Rumyantsev, 1975a, possibility of use of indirect transitions to obtain negative
‘‘Generation of coherent radiation in n Ga0.5 In0.5 P - temperature in semiconductors,’’ Sov. Phys. JETP 12, 1033.
p Gax⬃0.55 In1 ⫺ x Asy⬃0.10 P1 ⫺ y - n Ga0.5 In0.5 P , ’’ Pis’ma Zh. Blum, A. I., N. P. Mokrovsky, and A. R. Regel, 1952, ‘‘The
Tekh. Fiz. 1, 305 [Sov. Tech. Phys. Lett. 1, 147 (1975)]. study of the conductivity of the semiconductors and interme-
Alferov, Zh. I., I. N. Arsent’ev, D. Z. Garbuzov, S. G. Konni- tallic compounds in solid and liquid state,’’ Izv. Akad. Nauk
kov, and V. D. Rumyantsev, 1975b, ‘‘Red injection heterolas- SSSR, Ser. Fiz. 16, 139.
ers in the Ga-In-As-P system,’’ Pis’ma Zh. Tekh. Fiz. 1, 406 Bogatov, A. P., L. M. Dolginov, L. V. Druzhinina, P. G. Eliseev,
[Sov. Tech. Phys. Lett. 1, 191 (1975)]. L. N. Sverdlova, and E. G. Shevchenko, 1974, ‘‘Heterolasers
Alferov, Zh. I., N. A. Bert, A. Yu. Egorov, A. E. Zhukov, P. S.
on the base of solid solutions Gax In1⫺x As4P1⫺y and
Kop’ev, A. O. Kosogov, I. L. Krestnikov, N. N. Ledentsov, A.
Alx Ga1⫺x Sby As1⫺y ,’’ Kvant. Elektron. 1, 2294 [Sov. J. Quan-
V. Lunev, M. V. Maksimov, A. V. Sakharov, V. M. Ustinov, A.
F. Tsatsul’nikov, Yu. M. Shernyakov, and D. Bimberg, 1996, tum Electron. 1, 1281 (1974)].
‘‘An injection heterojunction laser based on arrays of verti- Chang, L. L., L. Esaki, W. E. Howard, and R. Ludke, 1973,
cally coupled InAs quantum dots in a GaAs matrix,’’ Fiz. ‘‘Growth of GaAs-GaAlAs superlattices,’’ J. Vac. Sci. Tech-
Tekh. Poluprovodn. 30, 351 [Sov. Phys. Semicond. 30, 194 nol. 10, 11.
(1996)]. Chang, L. L., L. Esaki, and R. Tsu, 1974, ‘‘Resonant tunneling
Alferov, Zh. I., D. Z. Garbuzov, V. S. Grigor’ev, Yu. V. Zhi- in semiconductor double barriers,’’ Appl. Phys. Lett. 24, 593.
lyaev, L. V. Kradinova, V. I. Korol’kov, E. P. Morozov, O. A. Cho, A. Y., 1971a, ‘‘Film deposition by molecular beam tech-
Ninua, E. L. Portnoy, V. D. Prochukhan, and M. K. Trukan, niques,’’ J. Vac. Sci. Technol. 8, 31.
1967, ‘‘Injection luminescence of epitaxial heterojunctions in Cho, A. Y., 1971b, ‘‘Growth of periodic structures by the
the GaP-GaAs system,’’ Sov. Phys. Solid State 9, 208. molecular-beam method,’’ Appl. Phys. Lett. 19, 467.
Alferov, Zh. I., D. Z. Garbuzov, E. P. Morozov, and E. L. Port- Davydov, B. I., 1939, ‘‘Contact resistance of semiconductors,’’
noi, 1969, ‘‘Diagonal tunneling and polarization of radiation Zh. Eksp. Teor. Fiz. 9, 451.
in Alx Ga1⫺x As-GaAs heterojunctions and in GaAs p-n junc- Delagebeaudeuf, D., et al., 1980, ‘‘Two-dimensional electron
tions,’’ Fiz. Tekh. Poluprovodn. 3, 1054 [Sov. Phys. Semicond. gas MESFET structure,’’ Electron. Lett. 16, 667.
3, 885 (1970)]. Dingle, R., H. L. Stormer, H. L. Gossard, and W. Wiegmann,
Alferov, Zh. I., N. Yu. Gordeev, S. V. Zaitsev, P. S. Kop’ev, I. V. 1978, ‘‘Electron mobilities in modulation-doped semiconduc-
Kochnev, V. V. Khomin, I. L. Krestnikov, N. N. Ledentsov, A. tor heterojunction superlattices,’’ Appl. Phys. Lett. 33, 665.
V. Lunev, M. V. Maksimov, S. S. Ruvimov, A. V. Sakharnov, Dingle, R., W. Wiegmann, and C. H. Henry, 1974, ‘‘Quantized
A. F. Tsatsul’nikov, Yu. M. Shernyakov, and D. Bimberg, states of confined carriers in very thin Alx Ga1⫺x As-
1996, ‘‘A low-threshold injection heterojunction laser based GaAs-Alx Ga1⫺x As heterostructures,’’ Phys. Rev. Lett. 33,
on quantum dots, produced by gas-phase epitaxy from orga- 827.
nometallic compounds,’’ Fiz. Tekh. Poluprovodn. 30, 357 Dupuis, R. D., and P. D. Dapkus, 1977, ‘‘Room temperature
[Sov. Phys. Semicond. 30, 197 (1996)]. operation of Ga1⫺x Alx As/GaAs double-heterostructure la-
Alferov, Zh. I., and R. F. Kazarinov, 1963, ‘‘Semiconductor la- sers grown by metalorganic chemical vapor deposition,’’
ser with electric pumping,’’ Inventor’s Certificate No. 181737 Appl. Phys. Lett. 31, 466.
[in Russian], Application No. 950840, priority as of March 30, Dupuis, R. D., P. D. Dapkus, N. Holonyak, Jr., E. A. Rezek,
1963. and R. Chin, 1978, ‘‘Room temperature operation of
Alferov, Zh. I., V. B. Khalfin, and R. F. Kazarinov, 1966, ‘‘A quantum-well Ga1⫺x Alx As-GaAs laser diodes grown by met-
characteristic feature of injection into heterojunctions,’’ Fiz. alorganic chemical vapor deposition,’’ Appl. Phys. Lett. 32,
Tverd. Tela 8, 3120 [Sov. Phys. Solid State 8, 2480 (1967)]. 295.
Alferov, Zh. I., Yu. V. Zhilyaev, and Yu. V. Shmartsev, 1971, Ekimov, A. I., and A. A. Onushchenko, 1981, ‘‘Quantum size
‘‘The splitting of the conduction band in ‘superlattice’ on the effect in three dimensional microscopic semiconductor crys-
base GaPx As1⫺x ,’’ Fiz. Tekh. Poluprovodn. 5, 196 [Sov. Phys. tals,’’ JETP Lett. 34, 345.
Semicond. 5, 174 (1971)]. Esaki, L., and L. L. Chang, 1974, ‘‘New transport phenomenon
Anderson, R. L., 1960, ‘‘Germanium-gallium arsenide hetero- in a semiconductor ‘superlattice,’ ’’ Phys. Rev. Lett. 33, 686.
junctions,’’ IBM J. Res. Dev. 4, 283. Esaki, L., and R. Tsu, 1970, ‘‘Superlattice and negative differ-
Anderson, R. L., 1962, ‘‘Experiments on Ge-GaAs heterojunc- ential conductivity,’’ IBM J. Res. Dev. 14, 61.
tions,’’ Solid-State Electron. 5, 341. Faist, J., F. Capasso, D. L. Sivco, C. Sirtori, A. L. Hutchinson,
Antipas, G. A., R. L. Moon, L. W. James, J. Edgecumbe, and and A. Y. Cho, 1994, ‘‘Quantum cascade laser,’’ Science 264,
R. L. Bell, 1973, ‘‘In gallium arsenide and related com- 553.
pounds,’’ Inst. Phys. Conf. Ser. 17, 48. Fowler, A. B., F. F. Fang, W. E. Howard, and P. J. Stiles, 1966,
Arakawa, Y., and H. Sakaki, 1982, ‘‘Multidimensional quan- ‘‘Magnetooscillatory conductance in silicon surfaces,’’ Phys.
tum well laser and temperature dependence of its threshold Rev. Lett. 16, 901.

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

 Zhores I. Alferov: The double heterostructure concept 781

Frenkel, Ya. I., 1931, ‘‘On the transformation of light into heat Kazarinov, R. F., and R. A. Suris, 1972b, ‘‘Injection heterolaser
in solids,’’ Phys. Rev. 37, 17; 37, 1276. with diffraction grating on contact surface,’’ Fiz. Tekh. Polu-
Frenkel, Ya. I., 1936, ‘‘Light absorption and electrons and provodn. 6, 1359 [Sov. Phys. Semicond. 6, 1184 (1973)].
holes sticking in dielectric crystals,’’ Zh. Eksp. Teor. Fiz. 6, Kazarinov, R. F., and R. A. Suris, 1973, ‘‘Theory of electrical
647. properties of semiconductors with superlattices,’’ Fiz. Tekh.
Frenkel, Ya. I., and A. Ioffe, 1932, ‘‘On the electrical and pho- Poluprovodn. 7, 488 [Sov. Phys. Semicond. 7, 347 (1973)].
toelectric properties of contacts between a metal and semi- Keldysh, L. V., 1962, ‘‘Effect of ultrasonics on the electron
conductor,’’ Phys. Z. Sowjetunion 1, 60. spectrum of crystals,’’ Fiz. Tverd. Tela. 4, 2265 [Sov. Phys.
Garbuzov, D. Z., et al., 1988, ‘‘Technical Digest CLEO,’’ Paper Semicond. 4, 1658 (1963)].
THU44, p. 396. Kirstaedter, N., et al., 1994, ‘‘Low threshold, large T 0 injection
Garbuzov, D. Z., et al., 1990, Conference Digest 12th Interna- laser emission from (InGa)As quantum dots,’’ Electron. Lett.
tional Semiconductor Laser Conference, Davos, Switzerland, 30, 1416.
Paper L-33, p. 238. Klitzing, K. V., G. Dorda, and M. Pepper, 1980, ‘‘New method
Goldschmidt, V. M., 1929, ‘‘Crystal structure and chemical con- for high-accuracy determination of the fine-structure constant
stitution,’’ Trans. Faraday Soc. 25, 253. based on quantized Hall resistance,’’ Phys. Rev. Lett. 45, 494.
Goldstein, L., F. Glas, J. Y. Marzin, M. N. Charasse, and G. Le Kogelnik, H., and C. V. Shank, 1971, ‘‘Stimulated emission in a
Roux, 1985, ‘‘Growth by molecular beam epitaxy and charac- periodic structure,’’ Appl. Phys. Lett. 18, 152.
terization of InAs/GaAs strained-layer superlattices,’’ Appl. Kroemer, H., 1957a, ‘‘Theory of a wide-gap emitter for transis-
Phys. Lett. 47, 1099. tors,’’ Proc. IRE 45, 1535.
Goryunova, N. A., 1951, ‘‘Seroe olovo’’ [Gray tin], Thesis Kroemer, H. 1957b, ‘‘Quasi-electric and quasi-magnetic fields
(Leningrad State University, Physico-Technical Institute). in a non-uniform semiconductor,’’ RCA Rev. 28, 332.
Gross, E. F., and N. A. Karryev, 1952a, ‘‘Light absorption by Kroemer, H., 1963, ‘‘A proposed class of heterojunction injec-
cuprous oxide crystal in infrared and visible spectrum,’’ Dokl. tion lasers,’’ Proc. IEEE 51, 1782.
Akad. Nauk SSSR 84, 261. Kroemer, H., and G. Griffiths, 1983, ‘‘Staggered-lineup hetero-
Gross, E. F., and N. A. Karryev, 1952b, ‘‘Exciton optical spec- junctions as sources of tunable below-gap radiation: Operat-
trum,’’ Dokl. Akad. Nauk SSSR 84, 471. ing principle and semiconductor selection,’’ IEEE Electron
Gubanov, A. I., 1950, ‘‘Theory of the contact between two Device Lett. EDL-4, 1, 20.
semiconductors with different types of conduction,’’ Zh. Ledentsov, N. N., M. Grundmann, N. Kirstaedter, J. Christen,
Tekh. Fiz. 20, 1287. R. Heitz, J. Bohrer, F. Heinrichsdorff, D. Bimberg, S. S. Ru-
Gubanov, A. I., 1951, ‘‘Theory of the contact of two semicon- vimov, P. Werner, U. Richter, U. Gösele, J. Heydenreich, V.
ductors of the same type of conductivity,’’ Zh. Tekh. Fiz. 21, M. Ustinov, A. Yu. Egorov, M. V. Maximov, P. S. Kop’ev, and
304. Zh. I. Alferov, 1995, ‘‘Luminescence and structural properties
Hall, R. H., G. E. Fenner, J. D. Kingsley, T. J. Soltys, and R. O. of (In,Ga)As-GaAs quantum dots,’’ in Proceedings of the
Carlson, 1962, ‘‘Coherent light emission from GaAs junc- 22nd International Conference on the Physics of Semiconduc-
tion,’’ Phys. Rev. Lett. 9, 366.
tors, Vancouver, Canada, 1994, edited by D. J. Lockwood
Hayashi, I., 1984, ‘‘Heterostructure lasers,’’ IEEE Trans. Elec-
(World Scientific, Singapore), Vol. 3, p. 1855.
tron Devices ED-31, 1630.
Lott, J. A., N. N. Ledentsov, V. M. Ustinov, A. Yu. Egorov, A.
Hayashi, I., M. B. Panish, P. W. Foy, and S. Sumski, 1970,
E. Zhukov, P. S. Kop’ev, Zh. I. Alferov, and D. Bimberg,
‘‘Junction lasers which operate continuously at room tem-
perature,’’ Appl. Phys. Lett. 17, 109. 1997, ‘‘Vertical cavity lasers based on vertically coupled quan-
Hitchens, W. R., N. Holonyak, Jr., P. D. Wright, and J. J. Cole- tum dots,’’ Electron. Lett. 33, 1150.
man, 1975, ‘‘Low-threshold LPE In1⫺x ⬘ Gax ⬘ P1⫺z ⬘ Asz ⬘ / Lott, J. A., N. N. Ledentsov, V. M. Ustinov, N. A. Maleev, A.
In1⫺x Gax P1⫺z Asz /In1⫺x ⬘ Gax ⬘ P1⫺z ⬘ Asz ⬘ yellow double- E. Zhukov, A. R. Kovsh, M. V. Maximov, B. V. Volovik, Zh. I.
heterojunction laser diodes (J⬍104 A/cm2 , I⫽5850 A, 77 Alferov, and D. Bimberg, 2000, ‘‘InAs-InGaAs quantum dot
K),’’ Appl. Phys. Lett. 27, 245. VCSELs on GaAs substrates emitting at 1.3 ␮m,’’ Electron.
Holonyak, N., Jr., and S. F. Bevacgua, 1962, ‘‘Coherent (vis- Lett. 36, 1384.
ible) light emission from Ga(As1⫺x Px ) junctions,’’ Appl. Ludowise, M., W. T. Dietze, C. K. Lewis, M. D. Camras, N.
Phys. Lett. 1, 83. Holonyak, Jr., B. K. Fuller, and M. A. Nixon, 1983, ‘‘Continu-
Hsieh, J. J., 1976, ‘‘Room-temperature operation GaInAsP/InP ous 300 K laser operation of strained superlattices,’’ Appl.
double-heterostructure diode lasers emitting at 1.1 ␮m,’’ Phys. Lett. 42, 487.
Appl. Phys. Lett. 28, 283. Lutskii, V. N., 1970, ‘‘Quantum-size effect-present state and
James, L., G. Antipas, R. Moon, J. Edgecumbe, and R. L. Bell, perspective on experimental investigations,’’ Phys. Status So-
1973, ‘‘Photoemission from cesium-oxide-activated In- lidi A 1, 199.
GaAsP,’’ Appl. Phys. Lett. 22, 270. Manasevit, H. M., 1968, ‘‘Single crystal GaAs on insolating
Kazarinov, R. F., and R. A. Suris, 1971, ‘‘Possibility of the am- substrates,’’ Appl. Phys. Lett. 12, 156.
plification of electromagnetic waves in a semiconductor with Mimura, T., S. Hiyamizu, T. Fuji, and K. Nanbu, 1980, ‘‘A new
a superlattice,’’ Fiz. Tekh. Poluprovodn. 5, 797 [Sov. Phys. field-effect transistor with selectively doped
Semicond. 5, 707 (1971)]. GaAs/n-Alx Ga1⫺x As heterojunctions,’’ Jpn. J. Appl. Phys.
Kazarinov, R. F., and R. A. Suris, 1972a, ‘‘Electric and electro- 19, L225.
magnetic properties of semiconductors with a superlattice,’’ Nakamura, M., A. Yariv, H. W. Yen, S. Somekh, and H. L.
Fiz. Tekh. Poluprovodn. 6, 148 [Sov. Phys. Semicond. 6, 120 Garvin, 1973, ‘‘Optically pumped GaAs surface laser with
(1972)]. corrugation feedback,’’ Appl. Phys. Lett. 22, 315.

Rev. Mod. Phys., Vol. 73, No. 3, July 2001

782 Zhores I. Alferov: The double heterostructure concept

Nasledov, D. N., A. A. Rogachev, S. M. Ryvkin, and B. V. ceedings of the Conference on the Physics of Semiconductor
Tsarenkov, 1962, ‘‘Recombination radiation of galiun ar- Surfaces, Philadelphia, June 1956, edited by R. H. Kingston
senic,’’ Fiz. Tverd. Tela 4, 1062 [Sov. Phys. Solid State 4, 782 (University of Pennsylvania, Philadelphia), p. 68.
(1962)]. Tsang, W. T., 1982, ‘‘Extremely low threshold (AlGa)As
Nathan, M. I., W. P. Dumke, G. Burns, F. H. Dill, Jr., and G. I. graded-index waveguide separate-confinement heterostruc-
Lasher, 1962, ‘‘Stimulated emission of radiation from GaAs ture lasers grown by molecular-beam epitaxy,’’ Appl. Phys.
p-n junctions,’’ Appl. Phys. Lett. 1, 62. Lett. 40, 217.
Natta, G., and L. Passerini, 1928, Gazz. Chim. Ital. 58, 458. Tsu, R., and L. Esaki, 1973, ‘‘Tunneling in finite superlattice,’’
Osbourn, G., 1982, ‘‘Strained-layer superlattices from lattice Appl. Phys. Lett. 22, 562.
mismatched materials,’’ J. Appl. Phys. 53, 1586. Tsui, D. C., H. L. Stormer, and A. C. Gossard, 1982, ‘‘Two
Park, G., O. B. Shchekin, S. Csutak, D. L. Huffaker, and D. dimensional magnetotransport in the extreme quantum
Deppe, 1999, ‘‘Room-temperature continuous-wave opera- limit,’’ Phys. Rev. Lett. 48, 1559.
tion of a single-layered 1.3 ␮m quantum dot laser,’’ Appl. van der Ziel, J. P., R. Dingle, R. C. Miller, W. Wiegmann, and
Phys. Lett. 75, 3267. W. A. Nordland, Jr., 1975, ‘‘Laser oscillations from quantum
Park, G., O. B. Shchekin, D. L. Huffaker, and D. G. Deppe, states in very thin GaAs-Al0.2Ga0.8As multilayer structures,’’
2000, ‘‘Low threshold oxide-confined 1.3 ␮m quantum dot Appl. Phys. Lett. 26, 463.
laser,’’ IEEE Photonics Technol. Lett. 33, 230. Welker, N. H., 1939, ‘‘Uber neue halblestende Verbindungen’’
Rezek, E., H. Shichijo, B. A. Vojak, and N. Holonyak, 1977, [New semiconductor compounds], Z. Naturforsch. A 7a, 744;
‘‘Confined-carrier luminescence of a thin In1⫺x Gax P1⫺z Asz 1953, Z. Naturforcsch A 8a, 248.
well (x 0 ⬃0.13, z⬃0.29, ⬃400 A) in an InP p-n junction,’’ Zhukov, A. E., A. R. Kovsh, S. S. Mikhrin, N. A. Maleev, V. M.
Appl. Phys. Lett. 31, 534. Ustinov, D. A. Lifshits, I. S. Tarasov, D. A. Bedarev, M. V.
Rupprecht, H. S., J. M. Woodall, and G. D. Pettit, 1967, ‘‘Ef- Maximov, A. F. Tsatsul’nikov, I. P. Soshnikov, P. S. Kop’ev,
ficient visible electroluminescence at 300 K from Ga1⫺x Alx As Zh. I. Alferov, N. N. Ledentsov, and D. Bimberg, 1999, ‘‘3.9
p-n junctions grown by liquid-phase epitaxy,’’ Appl. Phys. W CW power from submonolayer quantum dot diode laser,’’
Lett. 11, 81. Electron. Lett. 35, 1845.
Scifres, D. R., R. D. Burnham, and W. Streifer, 1974, Zhukov, A. E., V. M. Ustinov, A. Yu. Egorov, A. R. Kovsh, A.
‘‘Distributed-feedback single heterojunction GaAs diode la- F. Tsatsul’nikov, N. N. Ledenstov, S. V. Zaitsev, N. Yu.
ser,’’ Appl. Phys. Lett. 25, 203. Gordeev, P. S. Kop’ev, and Zh. I. Alferov, 1997, ‘‘Negative
Shchukin, V. A., N. N. Ledentsov, P. S. Kop’ev, and D. Bim- characteristic temperature of InGaAs quantum dot injection
berg, 1995, ‘‘Spontaneous ordering of arrays of coherent laser,’’ Jpn. J. Appl. Phys., Part 1 36, 4216.
strained islands,’’ Phys. Rev. Lett. 75, 2968. Zhuze, V. P., and I. V. Kurchatov, 1932a, ‘‘To question about
Shockley, W., 1951, ‘‘Circuit element utilizing semiconductor conductivity of cuprous oxide,’’ Zh. Eksp. Teor. Fiz. 2, 309.
material,’’ U.S. Patent 2269347, September 25. Zhuze, V. P., and I. V. Kurchatov, 1932a, ‘‘Zur electrischen Le-
Shriffer, J. R., 1957, ‘‘Semiconductor surface physics,’’ in Pro- itfahigkeit von Kupferoxydul,’’ Phys. Z. Sowjetunion 2, 453.

Rev. Mod. Phys., Vol. 73, No. 3, July 2001