City of Light: The Story of Fiber Optics

RECIPES FOR GRAINS OF SALT 155
as light which increased laser efficiency and reduced threshold current. In
the Bell Labs lasers, this region was the plane of the junction across the entire
chip, 0.37 millimeter long and 0.08 millimeter wide. 33 Alferov concentrated
current flow further by covering parts of the laser chip with silica, an insulator,
so electrons had to flow through a 0.03-millimeter stripe in the 0.20
millimeter wide chip. 34 This extra confinement increased the likelihood the
electrons would recombine with holes—improving laser operation. Later diode
laser developers have followed the same philosophy, refining structures
to better confine current. Performance also improves when the layers confine
light by total internal reflection, trapping it in the active stripe like a cladding
traps light in an optical fiber.
Ironically, the narrow-stripe laser originated at Bell Labs in 1966. Jack
Dyment hoped that limiting current flow to a narrow stripe would reduce the
current needed to drive the laser. He didn’t make much progress toward that
goal; the real problem was the homojunction structure of his laser. But the
stripe geometry eased another problem of early semiconductor lasers—poor
beam quality. While gas lasers generated steady pencil-thin beams, the uneven
pulsed light from diode lasers spread out so fast that calling it a beam
was barely justified. When he added the stripe, Dyment says, ‘‘We could start
to see repeatable mode patterns coming out of these devices, the first time
anyone could see anything repeatable happening in diode lasers.’’ 35 Mode
control was an important advance, and stripe-geometry lasers helped focus
light into single-mode fibers. 36
Panish and Hayashi knew about stripe-geometry lasers, but they didn’t
bother with the extra step of adding a stripe to their laser. Alferov’s group
did, covering their wafer with a thin insulating layer of silicon dioxide, in
which they etched a 30-micrometer stripe for current to flow through. That
extra concentration of current flow did the job, and the Russians had the first
semiconductor laser to emit steadily at room temperature. 37
It probably didn’t last very long. The informal rules of the laser-making
game specify only that a laser has to emit a steady beam long enough to
measure its properties. It might last for only seconds or minutes. Hayashi was
a stickler and insisted that his operate repeatedly; it emitted for a couple of
hours altogether. The Russians left the precise lifetime discretely unmentioned.
The Russians also were vague on other details and announced their success
only in a Russian-language journal. By the time word trickled through
the Iron Curtain, Bell Labs had already claimed a first. The lack of details and
the penchant of Soviet officials for claiming their scientists had invented almost
everything left many Americans skeptical. You have to go back and
compare submission dates to discover that Alferov submitted his report on
May 6, a month before Hayashi submitted his. Neutral observers consider the
Russian claim credible. 38
Both groups realized they had something to learn from each other. In
October, Hayashi and Panish visited Alferov in Leningrad. In November, Alferov
came to America 39 to spend six months at the University of Illinois with