1940 · Silicon P-N Junction Photocell · Work 811 of 1,871
A small test piece of 1940 on a wooden laboratory bench in New Jersey, a clear acrylic box six centimetres long, inside it a curved half-cylindrical reflector of polished metal, across its middle a short dark grey silicon rod held between two small metal clamps, two brown cloth-covered cables leading out of the box, a hand in a white laboratory coat sleeve holding a chrome flashlight whose beam falls on the rod, beside it a voltmeter in a wooden case turned at a steep slant, two test probes on cloth-covered leads, a black Bakelite switch, window light from the left.
What it was used for
The cracked silicon rod is the test piece on which the pn junction was discovered at Bell Telephone Laboratories in 1940. Russell Ohl worked there on silicon rectifiers for short waves. Per Riordan and Hoddeson his notebook recorded on 23 February 1940 that near one end of the rod a crack indicated a change in the crystal structure. On 6 March 1940 Mervin Kelly called group leaders and two men from the radio department together and Ohl demonstrated the rod: in a flashlight beam the voltmeter needle rose to almost 0.5 volts, more than 10 times anything Brattain and Becker had ever observed with any other photocell; Brattain at first suspected a prank and later received from Ohl that piece or another cut from the same ingot to check the finding himself.
What was new
The names came out of this rod: in late March 1940 Ohl and Scaff coined the terms p-type and n-type from the direction of easy current flow and the sign of the photovoltage, and the boundary between the regions became the pn junction. Patent US 2,402,662, "Light-Sensitive Electric Device", was filed on 27 May 1941 and issued on 25 June 1946; per the Smithsonian the rod with the junction across its centre is shown there as Fig. 8. The barb: the boundary was not designed. A technician had unknowingly cut the rod across the line between purified and commercial silicon, and that accidental boundary was the barrier.
What the sources leave open: Riordan and Hoddeson do not say what caused the crack; they name only the cut across the material boundary. None of the sources opened gives an efficiency figure for the cell. The Computer History Museum spells the surname of the metallurgist Jack Scaff "Shaff" in one place.
Connections
Every connection names its kind. “Forerunner” on its own would be a claim.
Follows fromThe Cat's Whisker1906Ohl had turned to his old silicon detector for short-wave reception in the mid-1920s, when valves failed: "There I began to appreciate the power of the crystal detector". Silicon detectors dated from Pickard's patent of 1906. At Bell, Ohl guessed that their erratic behaviour came from impurities and in 1937 set out to purify silicon. The rod of 1940 came out of that work.Source: Michael Riordan, Lillian Hoddeson: The origins of the pn junction, IEEE Spectrum, June 1997