EARLY TRANSISTOR AND DIODE HISTORY AT BELL LABS Art Uhlir Jr.
Oral History – Art Uhlir Jr. (Continued)	Oral History – Art Uhlir Jr. (Continued)
I was much earlier interested in anodizing, having done a senior project on phosphoric acid anodizing of aluminum, while working at a wartime aircraft plant. But in the case of germanium, we made no attempt to try anodizing on the grown-junction channel problem because we were persuaded at the outset that the solubility of the oxide made anodizing unlikely to be even as useful as the red lead. However, the alloy junction transistor could be made even more attractive if the base region geometry could be shaped. We quickly learned that we could localize electrolytic etching of N-type germanium by injection and depletion of holes at a junction. Light would also inject holes. Additional localization could be provided with “virtual-cathode” electrolytic micromachining. Reversing the current made it possible to plate spots of metal. The techniques were at hand for making much higher frequency germanium transistors. But the silicon diffused-base transistor had arrived and we were redeployed to the urgent microwave-diode project. However, someone had just given us a reject P-type silicon crystal – too small to yield wafers for the transistor line and anyway not the N-type used for NPN's. We thought we ought to at least see if electrolytic polishing could remove saw damage. And P-type had plenty of holes without illumination.	We started with low current as the best way to measure the etching valence with less chance of interfering secondary reactions. Even so, bubbles appeared at the anode. The first thought (from anodizing experience) was that some of the current was being waste by electrolysis of water to give oxygen. But the apparent valence turned out to be a little more than 2 instead of the 4.0 that germanium gave in agreement with the periodic chart. This was not smooth and shinny, but a matt brown substance that furthermore slowly evolved bubbles after being rinsed and placed in distilled water. These peculiarities encouraged us to collect the anode gas from mass spectrometer analysis. That was easy enough to do with an inverted test tube so we thought we could risk giving the first specimen the glowing-ember test for hydrogen and got a convincing “pop.” We thought that was enough weirdness to be worth mentioning in our wrap-up report on electrolytic shaping of germanium, for research to contemplate if they would. We also did enough testing to report that glycol addition and higher current would produce polishing. The report was eventually published in 1956, by which time we also published our solution of the transmod diode problem. Go To Uhlir Oral History, Page 14
COPYRIGHT © 2004 by Jack Ward. All Rights Reserved. http://www.transistormuseum.com/ PAGE 13