ieee journal of vol. 65 special issue papers an...

IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. QE-23, NO. 6, JUNE 1987 65 1

Special Issue Papers An Introduction to the Development of the

Semiconductor Laser RUSSELL D. DUPUIS FELLOW, IEEE

Abstract-In late 1962, the first semiconductor injection lasers were reported. While earlier workers had considered the possibility of light amplification in semiconductors, the achievement of high-efficiency electroluminescence from forward-biased GaAs p-n junctions was the event that catalyzed and accelerated efforts to demonstrate a semiconductor laser. This paper will attempt to review the experimental and theoretical work that preceded the actual demonstration of the semiconductor diode laser.

I INTRODUCTION

N LATE 1962, four groups reported the operation of homojunction semiconductor injection lasers. These

first reports are listed below in chronological order of their received dates.

R. N. Hall, G. E. Fenner, J. D. Kingsley, T. J. Soltys, and R. 0. Carlson, “Coherent light emission from GaAs junctions,’’ Phys. Rev. Lett., vol. 9, pp. 366-368, Nov. 1, 1962. (Received Sept. 24, 1962.) M. I. Nathan, W. P. Dumke, G. Burns, F. H. Dill, Jr., and G. Lasher, “Stimulated emission of radiation from GaAs p-n junctions,” Appl. Phys. Lett. , vol. 1, pp. 62-64, Nov. 1, 1962. (Received Oct. 6, 1962.) N. Holonyak, Jr. and S. F. Bevacqua, “Coherent (visible) light .emission from Ga( As, junctions,’’ Appl. Phys. Lett., vol. 1, pp. 82-83, Dec. 15, 1962. (Received Oct. 17, 1962.) T. M. Quist, R. H. Rediker, R. J . Keyes, W. E. Krag, B. Lax, A. L. McWhorter, and H. J. Zeiger, “Semiconductor maser of GaAs,” Appl. Phys. Lett. , vol. 1, pp. 9 1-92, Dec. 1, 1962. (Received Oct. 23, 1962, in final form Nov. 5, 1962 .)

The work that led up to the achievement of injection laser operation at these laboratories is discussed else- where in this Special Issue of the IEEE JOURNAL OF QUANTUM ELECTRONICS. This paper will discuss the work concerning the semiconductor laser prior to the demonstration of injection laser diodes in 1962 and attempt to describe various early theoretical and experimental efforts

Manuscript received November 12, 1986. The author is with AT&T Bell Laboratories, Murray Hill, NJ 07974. IEEE Log Number 8714079.

to make the semiconductor laser a reality. The author re- alizes that many different perspectives exist on this subject. Since only a few historical review papers exist describing the efforts at various research labs, it is difficult to determine exactly what ideas were being pursued and when the idea of making a semiconductor laser first came into being. For this reason, the authors of the first four papers describing semiconductor laser operation have been asked to write a description of their early ideas and research which led to the achievement of semiconductor laser operation in their laboratories. These papers appear in this issue. Since the early semiconductor laser work of these four groups is discussed in depth in these Invited Papers, their work is not discussed in detail in this paper.

EARLY CONCEPTS OF THE SEMICONDUCTOR LASER While various ideas concerning the semiconductor laser

had been discussed previously by several workers, the first demonstrations of the operation of semiconductor lasers in 1962 were realized without the benefit of many viable suggestions from earlier (i.e., prior to 1962) theoretical and experimental treatments of the concept of the semiconductor laser. The work leading up to the realization of the semiconductor laser has been briefly discussed by several authors. (See for example, [l] and [2].) In addition, the IEEE Press has published a volume in the ‘‘Selected Reprint Series” (sponsored by the IEEE Quantum Elec- tronics and Applications Society, now the IEEE Lasers and Electro-optics Society) on semiconductor lasers [ 3 ] . This volume contains reprints of many of the important papers on semiconductor lasers from the period 1962- 1978. The first four papers in this book are the four reports listed above.

The first documented discussion of the possibility of light amplification by the use of stimulated emission in a semiconductor (to the best of our knowledge) was made in an unpublished manuscript written by John von Neu- mann in 1953 [4]. In this paper, von Neumann discussed using carrier injection across a p-n junction as one possible means of achieving stimulated emission in semiconductors. While von Neumann does not explicitly use the term “p-n junction” in this paper, he clearly has this idea in mind since in Sect. X of this manuscript, he calculates the radiative transition rates between two Brillouin zones,

0018-9197/87/0600-0651$01.00 0 1987 IEEE

652 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. QE-23. NO. 6, JUNE 1987

B1 and B2, with electrons in excess of holes in B, (i.e., n- type) and holes in excess of electrons in B2 (i.e., p-type) 141. von Neumann sent this manuscript to Edward Teller along with a letter dated September 19, 1953 [4]. This letter and von Neumann’s previously unpublished manuscript titled “Notes on the photon-disequilibrium-amplification scheme” appear in this issue of the IEEE JOUR- NAL OF QUANTUM ELECTRONICS [4]. A summary of the von Neumann manuscript was included in Vol. 5 of the collected works of John von Neumann, which was published in 1963 [ 5 ] . John Bardeen summarized von Neumann’s ideas concerning a semiconductor laser in the following way: “By various methods, for example by injection of minority carriers from a p-n junction, it is possible to up- set the equilibrium concentrations of electrons in the conduction band and holes in the valence band. Recombina- tion of excess carriers may occur primarily by radiation. . . The rate of radiation may be enhanced by incident radiation of the same frequency in such a way as to make an amplifier” [ 5 ] . It is clear that von Neumann had the idea of Light Amplification by the Stimulated Emission of Radiation (i.e., the concept of a LASER) about a year before the operation of a MASER was first reported by Gordon, Zeiger, and Townes in July 1954

As far as we know at present, the first suggestion in an international public forum that it might be possible to obtain coherent light from a semiconductor was made by Pierre Aigrain of the Ecole Nornde Superieure in an unpublished paper given in June 1958 [7]. Aigrain had pri- vately discussed the possibility of making a semiconductor laser with Jacques Pankove of RCA Laboratories in 1956 171. At about the same time, at the Lebedev Institute in the U.S.S.R. (July 1958), Basov, Krokhin, and Popov independently came up with a similar idea [SI. While these workers suggested that it was theoretically possible to achieve population inversion in a semiconductor, it was still uncertain exactly what semiconductor to use and what types of electronic transitions should be exploited. In addition, the concept of optical feedback and the need for an electromagnetic cavity were not contained in these early papers.

Casey and Panish 121 discuss the history of the injection laser in Ch. 1 of their book and state that the first theoretical understanding of the requirements for the realization of stimulated emission in a semiconductor was developed by Bernard and Duraffourg of CNET and published in 1961 [9]. Bernard and Duraffourg used the idea of “quasi-Fermi levels” to correctly derive the equations which predict the onset of stimulated emission in a semiconductor 191. They suggested looking for stimulated emission involving direct transitions in the materials InAs or InSb or transitions between the conduction band and the acceptor states of Zn and In in the indirect semiconductors Ge and Si. Although the idea of achieving population inversion in indirect semiconductors was in error, Bernard and Duraffourg were the first to publish a clear account of the requirements for population inversion in a

161.

semiconductor p-n junction. However, they did.not discuss the idea of optical feedback and thus one important element in achieving laser operation was missing from their work. Their work did have an influence on at least one of the groups that demonstrated laser operation in 1962. In fact, in his historical review paper published in 1976, Mall mentions discussions he had with Bernard on various concepts concerning semiconductor lasers [IO].

Another theoretical treatment of the semiconductor laser was published in 1961 by Benoit a la Guillaume and Tric of the Ecole Normale Superieure [I I]. In this work, the authors consider three possible types of semiconductor laser transitions: 1) stimulated emission from an excited state to a ground state of an impurity level in an indirect semiconductor, e.g., the impurity As in Ge, 2) laser operation of a degenerate semiconductor having a direct bandgap (the example of InSb is considered), and 3) lasers involving the annihilation of excitons in indirect semiconductors. These transitions result in the emission of two bosons: a photon and a phonon (Si and Ge are considered as examples) [ 1 I]. These authors also pointed out the importance of using an optical cavity for longitudinal mode selection and describe the use of a p-i-n junction for carrier injection with the edges of the semiconductor serving as the mirrors of the optical resonator.

Basov and his group at the Lebedev Institute in Moscow were also considering how to make a semiconductor laser [8]. Basov et al. published a paper in 1959 concerning the possibility of obtaining a semiconductor laser using “negative temperature states’ ’ (population inversion) generated by impact impurity ionization under a pulsed electric field [8]. These workers later discarded this idea in favor of the use of a degenerate p-n junction. In 1961, Basov and co-workers discussed the idea of negative temperature states (using the quasi-Fermi level concept) in semiconductor p-n junctions. [I21 These workers considered the case of heavily-degenerate junctions (they refer to tunnel diodes). Also, they stated that “the high density of the majority carriers surrounding the region of negative temperature can, apparently, serve as reflecting surfaces, i.e., a “resonating cavity” is formed,” indicating that they understood the waveguiding properties due to refrac- tive index changes resulting from differences in the carrier concentrations at the edges of the active region [ 121. Ba- sov et al. suggested the existence of negative temperature states in a semiconductor would be indicated by changes in the I-V curve when the sample is illuminated with light of “suitable frequency” [l2]. In addition, they discussed the achievement of negative temperature states in indirect semiconductors 1121, [ 131 and they also referred to Pan- kove’s work on luminescence from Ge in a discussion of negative temperature states, and the possibility of achieving laser operation in Ge. It appears that as of about 1961, some of the ideas of Basov and co-workers concerning what to look for in a semiconductor injection laser and in what material it should be built were incorrect.

At about the same time as the paper by Basov et a!. concerning negative-temperature states in p-n junctioas

DUPUIS: DEVELOPMENT OF THE SEMICONDUCTOR LASER 65 3

was submitted to Zh. Eskp. Theo. Fiz. (April 18, 1961), a paper by Adirovich and Kuznetsova (also of the Le- bedev Institute in Moscow) was submitted to Fiz. Tver. Tela (received on May 27, 1961) in which these authors discussed the conditions for population inversion in degenerate semiconductors 1141. In this paper, the authors state “. . .one method of realizing an inverted distribution between bands might be above-the-barrier injection through the p-n junction in tunnel diodes. ” It is clear that these authors also had the idea of the use of a p-n junction in mind. Adirovich and Kuznetsova determined the low- temperature threshold current densities required for population inversion in p-n junctions in Si and Ge for three different acceptor doping Conditions in the p-type material. They estimated that under optimal conditions, laser operation of Ge could occur at current densities much less than 1 A /cm2 at 20 K [ 141. They reported no experimental work in this paper.

In April of 1962, Nasledov and co-workers at the Ioffe Physicotechnical Institute in Leningrad published a paper (received January 1 1, 1962) describing a slight narrowing of the “intrinsic recombination radiation” of a GaAs p-n junction [15]. The 77 K spectrum of this diode was observed to narrow from a FWHM of -255 A at 10 A/cm2 to a FWHM of - 200 A at a current density of 1500 A /cm2. We know now that this was clearly not a laser. Nasledov et al. proposed two possible explanations for this narrowing: 1) stimulated emission, and 2) effects due to internal absorption at shorter wavelengths. These workers state that “the second explanation above is more likely” [15]. While Nasledov and co-workers did not achieve stimulated emission, they did realize that line narrowing was one way of determining if stimulated emission was occurring. The paper by Nasledov et al. does not discuss the quantum efficiency of the GaAs diodes so it is impossible to tell if these light-emitting diodes were of high quality or not.

Another quite different idea for using a semiconductor to produce coherent radiation was proposed by Benjamin Lax of Lincoln Laboratory in 1959 [16]. He suggested that a semiconductor maser in the far-infrared and submillimeter wavelength region could be realized by using the cyclotron resonance transitions of electrons in conduction band states that had been split by the application of a large magnetic field [ 161. Lax proposed that a non- equilibrium population of electrons could be created in the conduction band by absorption of optical pump radiation. The applied magnetic field would then be tuned to yield energy spacings that corresponded to the modes of the cavity in which the semiconductor had been placed. This would result in cyclotron resonance transitions and the electrons would emit photons of energy corresponding to the field-induced splitting of the energy states. Lax speculated that the most favorable transitions for this cyclotron resonance maser action would be the direct [ 11 11 transitions at the L point of the Brillouin zones in Ge or InSb [ 161. In the “Conclusion” section of his 1959 paper on cyclotron resonance masers [16], Lax also discussed

semiconductor masers emitting in the near-infrared and referred to unpublished calculations by H. J. Zeiger that predicted a very small matrix element for band-to-band radiative recombination in an indirect semiconductor. From these calculations, Lax concluded that radiative recombination across the bandgap in a direct semiconductor is much more efficient than in an indirect semiconductor. Lax stated that the realization of the semiconductor maser is “perhaps . . . just a question of finding a suitable material” [16].‘

In addition to the above workers who discussed stimulated emission in semiconductors, there were other workers who patented concepts concerning the laser operation of semiconductors. The earliest known examples are the patents of Yasushi Watanabe and Jun-ichi Nishizawa that were filed in Japan in 1957 [ 171. In these patents, the authors describe several schemes of making a semiconductor maser. The authors use an analog of the microwave maser cavity for the electromagnetic resonator and the active semiconductor (Cl-doped Te or B-doped Si are suggested) is inserted into this cavity with electrodes con- nected to it to inject free carriers. The patent also discusses the use of a p-n junction for increasing the injection efficiency of carriers (generated by the absorption of light from an optical pump) into “the high energy state” of the semiconductor [ 171.

Another early patent concerning a semiconductor laser is that of William Boyle and David Thomas of Bell Lab- oratories which was filed January 1 1, 1960 [ 181. This patent discusses laser operation of Si, Ge, Gap, and CdS using stimulated emission from exciton states at low temperatures. They state that “the pumping can be either by the injection of minority carriers across a p-n junction within the semiconductor or by any incident ionizing energy suitable for producing hole-electron pairs in the semiconductor. . . .” [18]. Boyle and Thomas also discuss the requirements of using an optical cavity to achieve optical maser operation and suggest employing “the semiconductor wafer as a mode isolator by treating the surface of the wafer so as to favor selectively the growth of a longitudinal mode” [ 181. It is clear that Boyle and Thomas had the concept of injection of minority carriers by the use of an p-n junction and also the concept of forming the optical cavity by using the semiconductor it- self as the optical resonator.

THE DEMONSTRATION OF THE SEMICONDUCTOR LASER In many respects, the papers discussing the theoretical

aspects of semiconductor lasers published up to and during early 1962 were useful only in indicating what conditions might be required to observe stimulated emission in a semiconductor. In general, these papers treat direct and in’direct semiconductors more or less equally and im- plied that laser operation of either type of semiconductor was possible. Also, these theoretical papers failed to clearly identify the changes in device characteristics that are expected to occur when stimulated emission occurs in

654 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. QE-23, NO. 6. JUNE 1987

a semiconductor. As a result, there was no clear understanding of what to e.xpect experimentally when laser operation’ was achieved.

The principal event that led to the rush for injection laser operation of semiconductor p-n junction devices in 1962 was the achievement of high-efficiency electroluminescence from GaAs p-n junctions. The first report of high-efficiency spontaneous emission from GaAs p-n junctions was made in early 1962 by Sumner Mayburg and co-workers of GTE Laboratories at the “March” Meeting of the American Physical Society [ 191. Although Mayburg presented his data in a post-deadline paper at this conference, his results were apparently not published until January 1963 [ 191. Of course, by the time Mayburg and his co-workers’ paper was published, the semiconductor laser was a reality and, as a result, the GTE work got little recognition.

The data presented by Mayburg concerning high-efficiency GaAs p-n junction injection luminescence were similar to the results which were presented later by R. J. Keyes and T. M. Quist, and also independently by J. Pan- kove at the July 1962 Solid State Device Research Con- ference (SSDRC). Keyes and Quist from Lincoln Labo- ratory reported on GaAs diodes that had calculated internal quantum efficiencies of 48-85 percent at 77 K [20]. Pan- kove from RCA Laboratories reported high-speed modu- lation of a GaAs diode at 200 MHz [21]. Pankove’s diodes had an estimated total internal quantum efficiency of from 0.50 to 1 .OO photon /electron at 77 K , but had an external quantum efficiency of only about 1 percent [21].

After the SSDRC, Hall at the GE Research Lab in Schenectady , NY [ 101, and Holonyak at the GE Ad- vanced Semiconductor Lab in Syracuse, NY [22], went back to their labs and began an intensive study of how to obtain stimulated emission from p-n junctions and also began to think about what it would take to make an electromagnetic cavity to provide feedback into the active region of the diode. Of course, the group of Quist et al. at Lincoln Laboratory was also interested in getting their high-efficiency GaAs p-n junction diodes to lase [23]. In addition, Pankove at RCA was working on making an injection laser in GaAs. (See the reference to Pankove at the end of [7].)

The IBM work on the semiconductor injection laser had started in 1961 but was intensified after the January 1962 visit by Sumner Mayburg to the IBM T. J. Watson Re- search Center. Mayburg was working on GaAs p-n junctions at GTE and told the IBM workers of the results of his experiments that indicated that the internal efficiency of GaAs p-n junction diodes could be as high as 100 percent [24]. These results excited many of the people at IBM who were interested in the semiconductor laser problem, including William Dumke [24]. Dumke had already been considering the theoretical aspects of semiconductor lasers and had concluded that it might be possible to observe laser action in GaAs [24], [25]. The IBM research- ers then started a concerted effort to explore the possibility

of making a semiconductor laser in GaAs [24]. The IBM group’s work on GaAs p-n junction lasers was accelerated in response to the reports at the 1962 SSDRC of high- efficiency injection luminescence from GaAs diodes [24].

SUMMARY OF HISTORICAL LITERATURE

It is evident from the above discussion that many claims can be made concerning the j r s t person to have the idea of making a semiconductor laser. It would be difficult to determine exactly what merit there is to each of these claims since none of the early ideas concerning the semiconductor laser were entirely correct and none of the early workers had the technological problems completely solved. The first documented ideas concerning the requirements of making a semiconductor laser were written by von Neumann in 1953 [5]. Nishizawa’s “semiconductor maser” patents were filed in Japan in 1957 [17]. Ai- grain appears to have considered the concept of a semiconductor laser as early as 1956 [7]. Both Aigrain’s first unpublished statements in public [7] and Basov and his co-worker’s first documented ideas [8] concerning population inversion in semiconductors come somewhat later in June and July of 1958, respectively. Boyle and Thomas patented a semiconductor laser in 1960 [18].

With regard to the actual achievement of the operation of semiconductor lasers, none of this’early work mattered much. The use of the direct semiconductors GaAs and GaAsP as the materials of choice for the semiconductor injection laser was a direct result of the reports in 1962 of high-efficiency electroluminescence from fonvard- biased GaAs p-n junctions [ 191-[21). In most of the theoretical and experimental papers published up to late 1961, there were no obvious materials of choice for the fabrication of semiconductor lasers. In fact, some re- searchers continued to work on laser operation of Ge even after the papers reporting the operation of GaAs injection lasers were published in 1962 [7], [26]. The early (prior to 1962) theoretical and experimental work on semiconductor lasers is briefly summarized below.

von Neumann It seems safe to say that as far as we know at this time,

the first person to propose the concept of stimulated emission in a semiconductor was von Neumann, who developed a theoretical analysis of the use of a semiconductor material as a light amplifier using a p-n junction for the injection of carriers. von Neumann worked on the theoretical aspects of the problem as early as the summer of 1953. In September of 1953, he sent a detailed analysis of the semiconductor light amplifier to Teller. A summary of his unpublished manuscript appeared in the book John von Neumann, Collected Works, Vol. 5 , which was published in 1963 [ 5 ] . von Neumann’s theoretical work is quite detailed and discusses some of the practical aspects of the problem. (See von Neumann’s complete manuscript, published in this issue.)

DUPUIS: DEVELOPMENT OF THE SEMICONDUCTOR LASER

Nishizawa Nishizawa claims to have invented the semiconductor

laser in 1957 [17]. This claim is based upon Japanese patents issued to Watanabe and Nishizawa [17]. This work was not brought to light outside of Japan until some time after the ,actual achievements of laser operation in 1962. There were apparently no published attempts to reduce this patent to practice.

Aigrain Aigrain made a verbal presentation of ideas about ob-

taining coherent light from a semiconductor at a conference in Brussels in 1958. As early as 1956, he had en- couraged Pankove to work on the problem of making a semiconductor laser [7]. He did not publish a paper describing these ideas in the proceedings of this conference, and his work on the subject of the semiconductor laser was not published until 1964. I can find no reports of any experimental work on this problem by Aigrain until 1964, when photoluminescence from InAs was reported [7].

Basov and Co- Workers Basov and his co-workers first published in 1958 their

ideas concerning ‘‘negative temperature states” (population inversion) in a semiconductor using impact ionization for the creation of electron-hole pairs [8]. They first proposed using a p-n junction in a paper published in June 1961 [ 121. While this group published many experimental and theoretical papers on light emission from semiconductors, most of their work was on Ge until February 1963 when they first reported a p-n junction laser in GaAs [26], ~271.

Lax Lax proposed in 1959 that a cyclotron resonance far-

infrared or submillimeter maser could be made using an optically pumped semiconductor placed in a high magnetic field [ 161. This device would emit coherent radiation in the far-infrared or submillimeter region. The device was never realized. However, using Zeiger’s unpublished calculations, Lax did suggest in his 1959 paper that direct semiconductors were probably a much better choice for a semiconductor infrared maser than were indirect semiconductors [ 161.

Boyle and Thomas In 1960, Boyle and Thomas patented a semiconductor

laser based upon radiative recombination of excitons at low temperatures using various schemes for injecting minority carriers, including the use of a p-n junction [ 181. They also described the formation of the optical resonator by using the plane-parallel edges of the semiconductor it- self for optical feedback.

Adirovich and Kuznetsova These workers discussed population inversion in a de-

generate semiconductor p-n junction in a paper published

in November 1961 [ 141. They of Si and Ge. No experimental paper.

Bernard and Durafourg Bernard and Duraffourg [9]

655

treated the specific cases work was reported in this

published two papers in 1961 that treated the theory of the semiconductor laser using the concept of the “quasi-Fermi level” and derived equations that predicted the onset of stimulated emission. They considered the case of both direct and indirect bandgap semiconductors.

Benoit a la Guillaume and Tric Benoit a la Guillaume and Tric [ 111 independently pub-

lished a theoretical paper in 1961 in which they discussed the theory of the semiconductor injection laser and suggested various embodiments of the concept in both direct and indirect band-gap semiconductor materials. They also noted the importance of optical feedback for the realization of a semiconductor laser.

SUMMARY At this time, we believe that John von Neumann in his

1953 manuscript was the first to treat the idea of a semiconductor light amplifier and to treat the idea substantially correctly. Several individuals did work independently on the concept of a semiconductor laser or maser during the middle and late 1950’s and early 1960’s and a number of theoretical and experimental papers were published on the subject of the injection laser prior to the actual demonstrations of, injection laser operation in 1962. However, the most important event that led to the drive to make a semiconductor laser from GaAs and direct alloys of GaAsP was the achievement of high-efficiency electroluminescence from GaAs p-n junctions reported in 1962.

ACKNOWLEDGMENTS The author wishes to thank Prof. N. Holonyak, Jr. for

interesting private conversations, a preprint of [22], and a copy of the von Neumann manuscript, and Dr. M. I . Nathan for fruitful private discussions and copies of the sources cited in [24]. He also wishes to thank Dr. R. N. Hall and Dr. R. H. Rediker for useful comments and pre- prints of their papers cited in [lo] and [23], respectively. Also, the author is grateful to Prof. J. Bardeen for infor- mation concerning the review he wrote of the von Neu- mann manuscript, and to Dr. H. Lockwood for comments on the manuscript and a copy of the GTE laser patent. Discussions with Dr. M. von Neumann Whitman, and also with Dr. P. Labalme of the Institute for Advanced Study are gratefully acknowledged.

REFERENCES [I] R. W. Campbell and F. M. Mims 111, Semi-Conductor Diode La-

sers. Indianapolis, IN: Howard w. Sams Co., 1972, pp. 15-16. [2] H. C. Casey and M. B. Panish, Heterostructure Lasers, Part A: Fun-

damental Principles. New York: Academic, 1978, pp. 1-4.

656 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. QE-23, NO. 6 , JUNE 1987

[3] Semiconductor Injection Lasers, J . K. Butler, Ed. New York: IEEE Press, 1980, pp. 4-12.

[4] J . von Nelimann, “Notes on the photon-disequilibrium-amplification scheme,” an unpublished manuscript written before September 16, 1953. The typed original and two copies (one with additional nota- tions and changes written in pencil) are located in the “von Neumann Collection” of the Manuscript Division of the Library of Congress of the United States, James Madison Memorial Building, Room 101, Washington, DC, and are part of an extensive collection of von Neu- mann’s correspondence and other works. The collection was donated to the Library of Congress by Dr. M. von Neumann Whitmann and the Institute for Advanced Study in Princeton, NJ, where von Neu- mann had worked for several years before his death in 1957, von Neu- mann sent a copy of this manuscript, dated September 16, 1953, to E. Teller with a cover letter dated September 19. 1953. The file also contains a copy of the cover letter in which von Neumann states that the manuscript is “the long-promised write-up on amplification,” and that von Neumann had mentioned these ideas to Teller during the summer of 1953. von Neumann’s manuscript and the letter to Teller are reprinted in this issue. See J . von Neumann, “Notes on the photon-disequilibrium-amplification scheme (JvN), September 16, 1953,” this issue, pp. 659-673.

[5] John von Neumann Collected Works, Vol. 5 , A. H. Taub, Gen. Ed. New York: Macmillan 1963, p. 420. This short summary of the complete von Neumann manuscript is titled, “Notes on the photon- disequilibrium-amplification scheme.” and was edited from von Neu- mann’s unpublished notes by J . Bardeen. von Neumann’s calculations were also summarized in a letter Bardeen and Taub published in Scient. American, vol. 208, no. 2. p. 12, Feb. 1963. See also, [4].

[6] J. P. Gordon, H. J . Zeiger, and C. H. Townes, “Molecular microwave oscillator and new hyperfine structure in the microwave spectrum of NH,,”Phys. Rev., vol. 95, pp. 282-284, 1954.

[7] P. Aigrain, unpublished lecture at the Int. Con$ Solid Stute Phys. Electron., Telecommun., Brussels, 1958. This paper does not appear in Solid State Physics in Electronics and Telecornlnunications, Pro- ceedings of an International Conference held in Brussels, June 2-7, 1958, M. Desirant, Ed. New York: Academic 1960. Since this paper was never published, it is difficult to tell exactly what Aigrain discussed in his talk. His ideas are probably discussed in P. Aigrain, “Masers a semi-conducteurs” in Quantum Electronics, Proceedings of the Third International Congress, P. Grivet and N. Bloembergen, Eds. New York: Columbia Univ. Press, 1964, pp. 1761-1767. Aig- rain had discussions on the subject of semiconductor lasers with Jacques Pankove of RCA Laboratories in the fall of 1956 as described in the Introduction to the first Special Issue on Semiconductor Lasers of the IEEE Journal of Quantum Electronics: J. Pankove, “1967 semiconductor laser conference--Introduction,” IEEE J . Quantum Electron., vol. QE-4, pp. 109-1 10, 1968. It is interesting to note that in 1961, there were rumors that Aigrain had actually demonstrated a semiconductor maser using indirect transitions. See Lax’s comment on p. 506 of the second paper in [13] and Bromberg’s paper in [24].

[8] N. G. Basov, B. M. Vul, and Yu. M. Popov, in an unpublished paper registered with the Committee on Discoveries and Inventions of the Council of Ministers of the U.S.S.R. dated July 7, 1958; see also,

of electromagnetic oscillations,” Zh. Eskp. Theo. Fiz., vol. 37, pp. 587-588, 1959 (Sov. Phys. JETP, vol. 10, p. 416, 1959). A review of this work is given by N. G. Basov, “Quantum electronics at the P. N. Lebedev Physics Institute of the Academy of Sciences of the USSR (FIAN),” Usp. Fiz. Nauk, vol. 148, pp. 313-324, 1986 (Sov.

[9] M. G. A. Bernard and G. Duraffourg, “Laser conditions in semiconductors,” Phys. Stat. Sol . , vol. 1, pp. 699-703, 1961; see also, -, “Possibilites de lasers a semi-conducteurs,“ J . de Physique,

[lo] R. N. Hall, “Injection lasers,” IEEE Trans. Electron. Dev., vol. ED-23, pp. 700-704, 1976; see also, -, “Injection lasers,” this issue, pp. 674-678.

[ l l ] C. Benoit a la Guillaume and Mme. Tric, “Les semi-conducteurs et leur utilisation possible dans les lasers,” J . de Physique, vol 22, pp. 834-836, 1961.

[I21 N. G. Basov, 0. N. Krokhin, and Yu. M. Popov, “Production of negative-temperature states in P-N junctions of degenerate semiconductors,” Zh. Eskp. Theor. Fiz., vol. 40, pp. 1879-1880, 1961 (Sov. Phys. JETP, vol. 13, pp. 1320-1321, 1961).

[I31 -, “Use of indirect transitions in semiconductors for the determ-

__ , “Quantum-mechanical semiconductor generators and amplifiers

Phys. USP., V O ~ . 29, pp. 179-185, 1986).

V O ~ . 22, pp. 836-837, 1961.

nation of states with negative absorption coefficients,” Zh. Eskp. Theor. Fiz., vol. 40, pp. 1203-1209, 1961 (Sov. Phys. JETP, vol. 13, pp. 845-849, 1961); see also, -, “Negative absorption coef- ficient at indirect transitions in semiconductors,” in Advances in Quantum Electronics, J . R. Singer, Ed. New York: Columbia Univ. Press, 1961, pp. 496-506.

[14] E. I. Adirovich and E. M. Kuznetsova, “Possible inverted distribution of electrons in degenerate semiconductors,” Fiz. Tver. Telu, vol. 3 , pp. 3339-3341, 1961 (Sov. Phys. Solid State, vol. 3 , pp. 2424- 2425, 1962). Note: 131 is missing from the English translation of this paper. Unfortunately, this is the reference the authors give to the use of injection in a p-n junction.

[I51 D. N. Nasledov, A. A. Rogachev, S . M. Ryvkin, and B. V. Tsar- enkov, “Recombination radiation of gallium arsenide,” Fiz. Tver. Tela, vol. 4 , pp. 1062-1065, 1962 (Sov. Phys.-Solid State, vol. 4 , pp. 782-784, 1962).

[I61 B. Lax, “Cyclotron resonance and impurity levels in semiconductors,” presented at the 1st Int. Conf. Quantum Electron., 1959; published in Quantum Electronics, C. H. Townes, Ed. New York: Co- lumbia Univ. Press, 1960, pp. 428-449. See also, -, “Cyclotron resonance maser,” presented at the 2nd Int. Conf. Quantum Elec- tron., 1961; published in Advances in Quantum Electronics, J . R. Singer, Ed. New York: Columbia Univ. Press, 1961, pp. 465-479. The calculations by Zeiger to which Lax refers to in the 1959 paper are presented in: H. J . Zeiger and W. E. Krag, “Semiconductor infrared maser studies,” Quarterly Progress Rep., Solid State Res. Lin- coln Lab., pp. 41-43, Oct. 1959 (unpublished). (Rep AFCRC-TN- 59-1018, also available from NTIS as No. AD231991.)

1171 This invention is referred to in Semiconductor Technologies, 1982, J. Nishizawa, Ed. Amsterdam, The Netherlands: North Holland, 198 I , p. ii; see also, Y. Watanabe and J. Nishizawa, Japanese patent no. 35-13787 (note: this appears to be a patent applicarion number, not a patent number) as cited in W. Susaki, “Progress in semiconductor lasers,” in Semiconductor Technologies, 1982, J . Nishizawa. Ed. Amsterdam, The Netherlands: North Holland, 1981, p. 174. There is another patent by Watanabe and Nishizawa (Japanese patent no. 273217) titled “Semiconductor Maser” filed in 1957 (application no. 32-9899) and granted in 1960.

[18] W. S . Boyle and D. G. Thomas, “Optical Maser,” U.S. Patent 3059117 filed January 11, 1960 and granted October 16, 1962. Reis- sued as patent RE. 25,632 on August 18, 1964; see also, D. G. Thomas and J. J . Hopfield, “Fluorescence in CdS and its possible use for an optical maser,” J . Appl. Phys., vol. 33. pp. 3243-3249, 1962.

[I91 J. Black, H. Lockwood, and S. Mayburg, “High efficiency electroluminescence in GaAs,” postdeadline paper P14, presented on March 28, 1962 by S . Mayburg at the 1962 “March” Meeting of the Amer- ican Physical Society, Baltimore, MD, Mar. 26-29, 1962. Mayburg’s postdeadline paper is mentioned in [4] of the 1962 laser paper by Nathan et al. It does not appear in the abstracts of this APS meeting given in Bull. Amer. Phys. Soc., series 11, vol. 7 , no. 3, Mar. 26, 1962. The data Mayburg discussed at this meeting are published in: J. Black. H. Lockwood, and S . Mayburg, “Recombination radiation in GaAs,” J . Appl. Phys., vol. 34, pp. 178-180, 1963, which was submitted July 16, 1962. In this paper, the authors report an estimated internal quantum efficiency near 100 percent for GaAs p-n junctions. It is worthy of note that intense injection luminescence from GaAs p-n junctions had been observed at GTE Labs before January 1962, as is mentioned in a proposal sent on January 5 , 1962 to the U.S. Army Procurement Office at Fort Monmouth, NJ (see J . L. Brom- berg’s paper cited in [24]. Birman and Mayburg were granted a patent concerning their ideas on the injection laser: J . L. Birman and S . Mayburg, “Optical Maser”, U.S. Patent 4083017, filed July 12, 1963 and granted April 4 , 1978. According to H. Lockwood (private com- munication, Nov. 24, 1986), GTE notebooks show that the laser concept, complete with resonant cavity, was documented during Decem- ber 5-14, 1961, and Mayburg and co-workers operated their first semiconductor laser during November 1962.

[20] R. J . Keyes and T. M. Quist in a paper presented at the 1962 Solid State Dev. Res. Conf., Durham, NH (unpublished). See R. J . Keyes and T. M. Quist, “Recombination radiation emitted by gallium arsenide,” Proc. IRE, vol. 50, pp. 1822-1823, 1962.

[21] J. I . Pankove and J . E. Berkeyheiser in a paper presented at the 1962 Solid State Dev. Res. Conf., Durham, NH (unpublished). See -, “A light source modulated at microwave frequencies,” Proc. IRE, vol. 50, pp. 1976-1977, 1962. Pankove had been working on injection luminescence in GaAs as early as 1961: J. I. Pankove and M. J .

DUPUIS: DEVELOPMENT OF THE SEMICONDUCTOR LASER 657

Massoulie, “Injection luminescence from gallium arsenide,” presented at the 1962 Annual Meeting of the American Physical Society, January 24-27, 1962, paper ZA5; abstract published in Bull. Amer. Phys. Soc., series 11, vol. 7 , no. 1, p. 88, Jan. 24, 1962.

[22] N. Holonyak, Jr . , “Semiconductor alloy lasers-1962,’’ this issue, pp. 684-691.

[23] R. H . Rediker, “Research at Lincoln Laboratory leading up to the development of the injection laser in 1962,” this issue, pp. 692-695.

[24] M. I . Nathan, “Invention of the injection laser at IBM,” this issue, pp. 679-683. The IBM work is also discussed in: J . L. Bromberg, “The laser history project”, an invited paper given at the March 1985 Southwest Opt. Conf., unpublished. This work is part of the Laser History Project co-sponsored by the APS, IEEE-LEOS, LIA, and OSA. A volume is being prepared for publication.

[25] W. P. Dumke, “Interband transitions and maser action”, Phys. Rev., vol. 127, pp. 1559-1563, 1962.

[26] N. G. Basov, “Inverted populations in semi-conductors,’’ in Quan- tum Electronics, Proceedings of the Third International Congress, P. Grivet and N. Bloembergen, Eds. New York: Columbia Univ. Press, 1964, pp. 1769-1785. V. S . Baeaev. N. G. Basov, B. M. Vul, B. D. Kopylovskii, 0. N. Krokhin,-E. P. Markin, Yu. M. Popov, A. N. Khvoshchev, and A. P. Shotov, “Semiconductor quantum generator at a p-n junction in GaAs,” Dokl. Akad. Nauk SSSR, vol. 150, pp. 275-278, 1963 (Sov. Phys. Dokl., vol. 8, pp. 453-456, 1963); see also: V. C. Bagaev (sic), N . G . Basov, B. M. Vul, B. D. Kopylovsky (sic), 0. N. Kro- khin, Yu. M. Popov, E. P. Markin, A . N. Khvoshev (sic), and A. P. Shotov, “Semi-conductor laser on P-N junction in GaAs,” in Quan- tum Electronics, Proceedings of the Third International Congress, P. Grivet and N. Bloembergen, Eds. New York: Columbia Univ. Press, 1964, pp. 1891-1897.

Russell D. Dupuis (S’68-SM’84-F’87) received the B.S.E.E., M.S.E.E., and Ph.D.E.E. degrees in 1970, 1971, and 1973, respectively, from the University of Illinois at Urbana-Champaign. The subject of his Ph.D. dissertation was the luminescence properties of the N isoelectronic impurity in GaAsP and InGaP.

In 1973 he joined the Semiconductor Research and Development Laboratory, Texas Instruments, Dallas, TX, where he studied the liquid-phase ep- itaxial growth of GaP:N and GaP:ZnO for green

and red LED’s. In 1975 he joined the Electronics Research Center, Rock- well International, Anaheim, CA, where he developed the metalorganic chemical vapor deposition (MOCVD) process for the growth of AlGaAs- GaAs double heterostructure lasers and solar cells. In 1979, he joined the Solid State Electronics Research Laboratory, AT&T Bell Laboratories, Murray Hill, NJ, where he has continued his research on AIGaAs-GaAs lasers grown by MOCVD, as well as studied the growth of InP-InGaAsP lasers and photodetectors by the MOCVD process. He is currently studying the MOCVD growth of heteroepitaxial AIGaAs-GaAs lasers on Si sub- strates.

Dr. Dnpuis is a member of the American Physical Society and the Elec- trochemical Society, and a member and past Chair of the Technical Com- mittee on Semiconductor Lasers of the IEEE Lasers and Electro-optics Society. He received the IEEE Morris N. Liebmann Field Award in 1985 with Harold Manasevit for pioneering work in the metalorganic chemical vapor deposition process, and received the first Young Scientist Award of the Gallium Arsenide and Related Compounds Conference in 1986, for his work in MOCVD.

ieee journal of vol. 65 special issue papers an...

Documents