Jean Hoerni
Jean Hoerni (1924–1997) was a Swiss-born physicist who invented the planar process, the method of building transistors under a protective oxide layer that made integrated circuits manufacturable. A theorist by training who worked in a field dominated by experimenters, he was one of the eight young scientists who left Shockley Semiconductor Laboratory in September 1957 to found Fairchild Semiconductor, and the technique he recorded in his notebooks there became the basis of semiconductor manufacturing for decades. He later founded or led several companies in the new industry and devoted his last years, and much of his wealth, to a very different passion: the mountains of Asia and the schools of the people who live in them.
Hoerni arrived in Silicon Valley as a physicist rather than an electrical engineer, and that background shaped how he solved problems. Born in Geneva in 1924, he earned doctorates from the University of Geneva and the University of Cambridge, worked in theoretical physics, and spent time at Caltech before joining William Shockley's new laboratory in Mountain View in 1956. At Shockley's lab he worked on the physics and theory of diffusion in silicon, a subject that later proved crucial, and he was among the people Shockley assembled to start a silicon device business. The management difficulties at Shockley Semiconductor led eight of its staff, including Hoerni, Robert Noyce, Gordon Moore, Jay Last, Julius Blank, Victor Grinich, Eugene Kleiner and Sheldon Roberts, to resign together in September 1957, the departure later nicknamed the Traitorous Eight.
His signature contribution was the planar process, and its story is one of persistent observation. At Fairchild the company's first products were mesa transistors, whose exposed junctions were vulnerable to contamination, and Hoerni, working in 1957 and 1958 on the problem, proposed leaving the oxide layer in place after diffusion so that it would protect each junction. His notebook entries dated December 1957 and January 1959 record the concept and its refinement, and a patent application followed in 1959. The technique only paid off fully when it combined with the oxide masking discoveries at Bell Labs and the photolithography of Lathrop and Nall. Noyce then used the flat, oxide-covered surface to propose planar interconnection for integrated circuits, so Hoerni's idea became the base on which Noyce's chip could stand.
Hoerni left Fairchild in the early 1960s and kept building companies, usually small and technical. He joined the Teledyne group's semiconductor venture Amelco, later worked at Union Carbide's electronics operation, and in 1967 founded Intersil, a company known for low-power integrated circuits and, later, for developing circuits for watches and calculators. This pattern of moving from invention to start-up to another start-up made him a prototype of the Silicon Valley technologist, an inventor who moved with ideas rather than staying with one employer, and it also meant that credit and wealth for the planar process were never concentrated in a single company. Fairchild was nevertheless the place where the process became a manufacturing standard.
Away from the semiconductor industry he was an accomplished mountaineer with a deep commitment to the Himalayas. Hoerni climbed and trekked in the great ranges, and after earning a fortune from his industry ventures he became a significant early supporter of efforts to build schools in remote Himalayan and Karakoram communities, including the Central Asia Institute, which was founded in the mid-1990s. His support helped establish the work in Pakistan and surrounding regions, and bequests following his death in 1997 continued it. A common misconception is that he was a founder of the semiconductor industry in the sense of a solo inventor; in reality he was a physicist whose single, very important idea relied on the context supplied by colleagues, notably Noyce, Moore and the Bell Labs researchers.
His legacy is visible in every integrated circuit, though his name rarely appears on one. The planar process remains the conceptual frame for the photolithographic fabrication of chips: grow oxide, pattern windows, introduce dopants, cover the surface and connect. Even the move from furnace diffusion to ion implantation, and from aluminium to copper interconnects, kept the logic that the flat oxide-covered silicon surface is the working platform. For students of transistor history, Hoerni is the figure who shows how a theorist's attention to a surface problem could reorganize an industry.
Read Jean Hoerni as the physicist who realized that leaving the oxide on, rather than removing it, would protect the junctions and let the whole chip industry be built flat.