Hyatt Gibbs
Over and above the fact that it is extremely interesting science, research in fundamental semiconductor optics has provided a knowledge base from which the optoelectronics industry has sprung. Nowadays, when trends in research funding threaten the continued existence of such research, it is perhaps useful to look at the role of fundamental research in developing the vertical-cavity surface-emitting laser (VCSEL).
Shih-Yuan Wang at Hewlett-Packard Laboratories (Palo Alto, CA) has mused that VCSEL technology followed an unusual route to market, in that the idea was conceived in Japan and became a product in the USA when so many other technologies have gone in the other direction (see "Next generation technology awaits second VCSEL decade," by senior editor Hassaun Jones-Bey, Laser Focus World, Dec. 1998, p. 73). I would add that the Japanese would have eventually made a low-threshold VCSEL without the US development effort, just as US researchers (due to a better environment for fundamental research at the time) would have conceptualized the VCSEL without the Japanese.
"My work on VCSELs is an outgrowth of the work on optical bistability that I began as a grad student . . . ," wrote Jack Jewell (founder of Pico light; Boulder, CO) in Oscillations (July 1, 1998), the Optical Sciences Center (OSC; University of Arizona) newsletter. "The transition from optical bistability to VCSELs was natural and took place at Bell Labs. Interestingly, around 1990, four of the world's foremost VCSEL pioneers were all at Bell Labs and had worked on optical bistability. . . at OSC. This underscores one of the great values of education at OSC: a very broad background in optics."
The creative curiosity of Sam McCall was a major factor throughout the hundreds of references on bistable Fabry-Pérot etalons that evolved into VCSELs. We first saw GaAs optical bistability at Bell Labs at low temperature in a molecular-beam-epitaxy (MBE) grown 4-µm-thick piece coated with dielectric mirrors (90% reflectivity). That same etalon lased when pumped with modelocked pulses. Note that when we saw lasing in 1978, we had never heard of the Japanese re searcher K. Iga or the often cited work of his group.
Because lasing of our bistable etalon was rather brute force, and the group of Holonyak had demonstrated a vertically emitting 2-µm-long CdSe laser way back in 1967, we did not think much about it. Soon after the bistability research was moved to Arizona (with Art Gossard and Bill Wiegmann still doing the MBE growth at Bell Labs), we saw optical bistability at room temperature-first in a multiple-quantum-well etalon and then in the original bulk-GaAs etalon. Further improvements in the etalon design resulted in very-low-energy gates. Growing interest in parallel processing led us to demonstrate large arrays of bistable pixels. This was the precursor to the beautiful scanning electron micrograph showing arrays of different-size low-threshold VCSELs made famous by Jack Jewell.
Shortly afterward, Jewell and his colleagues at Bell Labs demonstrated a high-finesse monolithic Fabry-Perot microcavity, grown by Florez and Harbison at Bellcore; this was an essential step toward a useful VCSEL. It was based on a fundamental contribution of van der Ziel and Ilegems much earlier, namely the demonstration of a monolithic MBE-grown distributed Bragg mirror. Encouraging results on monolithic structures, including high-threshold lasing, were also reported by Gourley and Drummond. Not long after the high-finesse "empty VCSEL," the first optically and electrically pumped low-threshold VCSELs were reported.
Jones-Bey reported in Laser Focus World (Dec. 1998, p. 73) that "a lot of work was performed on the concept in many different places, and the technology finally arrived in 1989 at Bell Labs (Murray Hill, NJ) as the fallout from a quixotic effort to develop optical cavities of high enough quality to enable optical computers." In deed the whole optical-computing craze of the 1980s was quixotic, but something that is becoming technologically important, namely the VCSEL, came from that rather far-out research.
Of course, we spent much more effort understanding the basic physics and operating characteristics of semiconductor etalons than we did on optical computing. It didn't take long to see little future for optical computing, except for optical interconnects. In fact, research on the fundamental properties of semiconductor Fabry-Perot interferometers continues with an emphasis on nonperturbative normal-mode coupling. Next will be the drive toward the quantum entanglement regime where a single photon changes the optical properties for a second photon.
Supporting basic research
The bottom line is that the VCSEL didn't come just from edge-emitter researchers trying to improve or modify their diode lasers. Often advances come quite indirectly and unexpectedly, while having fun "turning over stones," as eloquently described by Charles Townes in How the Laser Hap pened (Oxford University Press, NY, 1999). The important thing is to push the limits every way we can; something new and significant is likely to turn up.
Members of Congress and government program managers should be encouraged to direct 5% of the total research budget to curiosity-driven basic research from which comes many of the technological advances according to history. The management of Bell Labs in the "good old days" was wise enough to do that. That is, if we want to stay competitive economically, not to mention scientifically, we simply cannot afford not to do so! Otherwise more and more enabling breakthroughs will occur elsewhere. Despite major contributions, if present trends continue, research in fundamental semiconductor optics will dwindle to nothing, even though it is extremely interesting fundamental science and the basis for the whole optoelectronics industry. o
REFERENCES
For an extensive list of references, contact Hyatt Gibbs at [email protected].
HYATT GIBBS is a professor in the Optical Sciences Center at the University of Arizona, Tucson, AZ 85724; e-mail: [email protected].