Gunther Fenner (left),
Robert Hall and Jack
Kingsley in their lab with
the first diode laser.
Courtesy of General Electric
Ga As diode by Robert J. Keyes and Thomas M. Quist of MIT
Lincoln Laboratories at the early-summer Solid State Device
Research Conference in Durham, N.H., to launch serious
efforts to make diode lasers.
Robert N. Hall of General Electric’s Schenectady research
center had worked on semiconductor devices from germanium transistors to III-V tunnel diodes. He had been skeptical
about junction lasers because he expected their light emission
to be very inefficient, perhaps only 0.01 percent. The Lincoln
Lab data changed his mind, and on the train home he began
planning how to make a diode laser. Back in Schenectady,
he worked out more details, such as cutting wafers into chips
and polishing their edges. He enlisted the help of four other
GE scientists.
“We had a good deal of freedom to move from one project
to another,” he recalled in 1985. He asked his boss, Roy Apker,
to let the group work part-time on a laser for a few months.
It was far from a sure thing, he wrote in 1987, estimating
about a 20 percent chance of success. “...But even if we never
produced any coherent light, we were bound to learn a good
deal about the high-efficiency junction luminescence that had
been reported,” he said. Hall felt that that in itself ought to be
enough justification to try it. “And of course, if we did succeed
in creating a laser, there would be plenty of scientific excite-
ment and all the rest. There was no need for a selling job; Roy
liked the idea and gave it his blessing on the spot.”
Knowing that Lincoln Lab and RCA were already work-
ing on GaAs made the project urgent, so they divided tasks.
Semiconductor fabrication expert Ted Soltys made diodes.
Gunther Fenner built a test rig and tested diodes. Jack King-
sley contributed his laser experience. Dick Carlson studied
diffusion of zinc impurities into the semiconductor, a crucial
issue in material performance. Hall looked at the big picture,
William E. Engler and Marvin Garfinkel with Bob Hall’s first diode laser.
Courtesy of General Electric
and worked on issues such as obtaining materials and verifying
laser operation.
About two months into the project, one diode emitted an
odd horizontal line, which Hall could never explain. The next
day Fenner saw strong interference lines in the far field of
another diode, as expected from an edge-emitting laser. After
weeks of collecting data, filing a patent application and writing
a paper, their report, submitted September 24, 1962, appeared
in the November 1 Physical Review Letters. Other diode lasers
were reported in short order by Marshall Nathan’s group at the
IBM Watson Research Center, Lincoln Labs and Nick Holo-nyak at GE’s Syracuse research lab.
The near photo-finish showed what the big science labs of
the 1960s could achieve at their best. Looking back in 1987,
Hall wrote, “The working climate in our laboratory was highly
conducive to that kind of investigation. The contract load was
relatively modest and there were not many project-oriented
programs so it was fairly easy to rearrange priorities in order
to pursue a promising new idea. We moved rapidly because we
had the enthusiasm that comes from knowing that we were
working on our own idea and that it was a good one, and we
were encouraged to work on it as hard as we wanted to.”
Improving diode lasers
Nonetheless, those first diode lasers required high-current
pulses and cryogenic cooling to reach threshold. The problem
was that they were homojunction devices, with nothing to
keep carriers near the junction. The solution came in 1963
from German-born physicist Herbert Kroemer, then at Varian
Associates in Palo Alto, Calif. After hearing a seminar speaker
say carrier diffusion would prevent diode lasers from sustaining
a population inversion at room temperature, he recalled, “I
immediately protested: ‘But that’s a pile of ... ; all you have to
