|
|
Oral History – Jacques Pankove (Continued)
|
|
Since each new circuit was
patentable, RCA speeded up the disclosure process by having someone pick up
Lloyd’s notebook every other day and sort out the most valuable gems to be
transformed into patents. Sometimes Lloyd was unhappy with the performance
of a transistor and would try anything to improve it. One day, his
inspiration led him to spread a drop of saliva over a point contact
transistor. He turned up the current causing the spittle to bubble. When
the bubbling stopped, the surface was black. He used a toothpick to clean
off the black scum and checked the performance of the transistor. The
current gain had increased significantly! Lloyd Barton had discovered
electrolytic etching.
Returning to the alloy junction
transistor consisting of two indium dots on opposite sides of a thin
germanium wafer, I hypothesized that if one were to collect most of the
injected holes, the collector electrode should be much larger than the
emitter. To verify this hypothesis, I made a set of transistors having
different sizes on indium dots on opposite sides. When the current gain
was plotted versus the ratio of collector to emitter area, it became
evident that the larger this ratio was, the higher the current gain. This
concept was patented (US Patent #3,005,132) and produced an endearing
nickname for me within the patent department: “Big-dot-small-dot Pankove”.
I also tried successfully the electrolytic etching of the surface using
various electrolytes.
|
|
|
Oral History – Jacques Pankove (Continued)
|
|
I also found that oxidation treatments were
useful, such as evaporating SiO2 on the surface of germanium around the
indium electrodes (See Patent J. Pankove and G. Ellis, 2,796,562). Today,
the surfaces of a transistor are always oxidized even in silicon
transistors.
The next interesting development was stimulated
by our rivalry with GE. They had also started to make transistors with
indium dots, but called their devices diffused junction transistors while
we called ours alloy junction transistors. The question then was which
name is correct? I answered this question by the following experiment. I
immersed a small cylinder of germanium into a pot of indium at 600 degrees
C overnight. The next day, I let the indium cool off to room temperature
and dissolved the indium. Surprise! A beautiful crystalline structure
appeared at the end of the germanium cylinder that had been immersed in hot
indium. Evidently, germanium had dissolved into the indium and, on
cooling, had recrystallized on the germanium seed, forming p-type germanium
due to trapping of indium acceptor atoms in the growing crystal. (J.I.
Pankove, et al, Proc. IRE 40, 1352 (1952), C.W. Mueller and J.I. Pankove,
RCA Rev. 14, 536 (1954) and Proc. IRE 42, 336 (1954).
Go
To Pankove Oral History, Page 4
|
|