Nov 09, 2023
A Brief History of the MOS transistor, Part 2: Fairchild
No company was better equipped and better positioned to capitalize on the development of the first MOSFET than Fairchild Semiconductor. Founded in 1957 to work on silicon transistors, Jean Hoerni
No company was better equipped and better positioned to capitalize on the development of the first MOSFET than Fairchild Semiconductor. Founded in 1957 to work on silicon transistors, Jean Hoerni developed the planar process and Robert Noyce developed the ideas for the first practical integrated circuit (IC) based on Hoerni’s planar process just months before Atalla and Kahng got the first MOSFET to work at Bell Labs. Like the two keys needed to open a safety deposit box in a bank vault, the planar semiconductor process technology and the planar IC were the two keys needed to unlock the MOSFET’s full potential.
Fairchild possessed these keys, and although Fairchild researchers contributed significantly to the development and improvement of the MOSFET, the company failed to create a successful MOS IC product line. As a result, the company watched its early lead in ICs – bipolar ICs – evaporate as Moore’s Law drove IC device densities beyond the reach of bipolar transistor technology and into the enticing MOSFET domain.
William Shockley left Bell Labs in 1953 because he felt passed over for promotions and recognition. He moved back to California, took a position at Caltech, cut a deal with Caltech professor and high-tech entrepreneur Arnold Beckman, and started Shockley Transistor Laboratory in 1955. Initially, Shockley thought he could raid Bell Labs for personnel, but no one at his former employer wanted to work with him. He was forced to look elsewhere, and he managed to assemble a superlative team of young and freshly graduated scientists and engineers, luring them to California’s superlative weather. He also promised that they’d be developing the Holy Grail du jour, the silicon transistor.
The following year, Shockley shared the Nobel prize in physics with John Bardeen and Walter Brattain for the invention of the point-contact transistor. Around that time, Shockley became intensely interested in the 4-layer diode, a semiconductor switch that would have been of huge interest to the Bell System. However, it was not the device he’d promised to his researchers, and they weren’t pleased. Shockley’s autocratic management style and ego fractured his team, causing a showdown on May 29, 1957. The research team’s demand was for the “Shockley Problem” to be solved. It wasn’t, and eight members of Shockley’s research team – which became known as the Traitorous Eight – left in September 1957. That core group cut a deal with Sherman Fairchild and founded Fairchild Semiconductor on October 1, 1957. Fairchild Semiconductor would quickly become the most important semiconductor company in the world and the company most likely to elevate the MOS transistor to its full potential.
The first important step towards achieving Fairchild’s destiny was the invention of the planar semiconductor process. On December 1, 1957, just two months after Fairchild’s founding, Hoerni was hit with a flash of inspiration. He knew about work on silicon dioxide passivation, photolithography, and etching taking place at Bell Labs because Shockley had discussed it with his research team earlier that year, before the Traitorous Eight’s departure. Hoerni needed just two pages in his lab notebook to describe the planar process. His innovation was to leave the thermally grown silicon dioxide on the semiconductor wafer after diffusion to protect the circuitry beneath. Bell Labs thought this oxide was too dirty to leave in place but Hoerni realized that a sufficiently clean insulating layer would prevent contamination from dust, dirt, and water. With few changes, Hoerni filed for a patent on the planar process on January 14, 1959.
The second important step towards the semiconductor breakthrough that was needed to allow MOSFETs to achieve their destiny occurred on January 23, 1959. That was the day that Fairchild Semiconductor’s founder Robert Noyce wrote down the ideas for a monolithic integrated circuit in his lab notebook. He’d been spurred to think of ways to use Hoerni’s planar process to make more than discrete transistors. He realized that the silicon dioxide layer was a perfect insulator and allowed metal interconnect to be deposited on top to complete the connections among multiple devices on the IC. With that flash of insight, Noyce changed the electronics industry forever and transformed soldering and wiring into a high-tech printing process.
These two ideas, Hoerni’s planar process and Noyce’s IC concept, lit Fairchild’s fuse. First, Fairchild used the planar process to make better, more stable transistors than any of its competitors. Fairchild’s transistors quickly became the gold standard. Within two years, Fairchild announced the world’s first IC product family, dubbed Micrologic. It was a family of bipolar logic ICs. For six months after the introduction of the Micrologic family, Fairchild swept the field. None of Fairchild’s competitors had anything like Fairchild’s IC to offer to their customers. Even Texas Instruments, officially the co-inventor of the IC, was forced to license Fairchild’s IC patents just to compete.
Gordon Moore learned of Bell Labs’ successful development of the first MOS transistor shortly after Mohamed Atalla and Dawon Kahng got their first device working in 1960. Moore took over as director of Fairchild’s Research and Development department in 1959 after Robert Noyce became the company’s General Manager after the man hired for that role, Edward Baldwin, departed abruptly with five other Fairchild staff members to found Rheem Semiconductor.
The week following Baldwin’s departure, Fairchild tested the first bipolar transistor made with Hoerni’s planar process. The transistor worked well. There’s a myth that says Hoerni spat on the transistor during the test to prove that the planar process made transistors impervious to contamination. Perhaps the actual test performed was not so dramatic as to use a sample of Hoerni’s saliva, but the planar process indeed delivered the desired results. Fairchild co-founder Jay Last recalls saying, “Gee, it’s too bad Baldwin had to leave last week.” Fairchild had been making mesa transistors and, within months, moved production to the far superior and far more stable planar transistor design.
With the new planar bipolar transistors and Micrologic ICs, Fairchild had its hands full just making the planar process reliable and stable, learning to implement the precision lithography needed to make ICs, and developing new types of bipolar transistors and Micrologic chips. No Fairchild customers were demanding or even asking about MOS transistors, so none were in development. Gordon Moore’s Research and Development department’s focus was on long-range projects, which did not include MOS transistors. Fairchild’s indifference to MOSFETs changed in 1962 when the company hired a newly minted physics PhD from the University of Utah named Frank Wanlass. When he joined the company, Wanlass was already obsessed with MOS transistors, and he knew that Fairchild’s planar process was the way to make them.
Wanlass was hired into Fairchild’s Research and Development department, and his assignment gave him sufficient leeway to work on MOSFETs, even if Fairchild wasn’t making them at the time, because the MOS – metal-oxide-semiconductor – structure was an integral part of the planar process whether it was used to make bipolar or MOS transistors. He derived a lot of his freedom from his ability to do nearly everything himself. As a physicist, Wanlass understood the physics of the MOS structure and could therefore design MOSFETs himself. He understood electronics, so he designed not just the transistors but the circuitry that went into a MOS IC. One of his first designs was an integrated MOS flip-flop that yielded more than 80%. In February, 1963, Wanlass and his manager C.T. Sah presented a paper at ISSCC that revealed that Wanlass had conceived of circuits that combined p- and n-channel MOSFETs on the same IC. He’d invented CMOS as a mere byproduct of his work.
Along the way, Wanlass dealt with the MOSFET’s inherent stability problems and with Fairchild’s indifference to MOSFETs in general. The company was doing far too well making bipolar semiconductor products to devote much energy to the slow MOSFET. Although Gordon Moore’s Research and Development department devoted a lot of effort to analyzing and modeling MOS physics as a way to improve the planar process, those improvements targeted bipolar transistor manufacturing. By December 1963, Wanlass grew frustrated and jumped ship. He joined General Microelectronics (GME), which was started in the summer of 1963 by members of Fairchild’s Micrologic group for the specific purpose of developing MOS ICs.
With the loss of several key members of the Micrologic group and Wanlass, the wind went out of Fairchild’s sails and the company never developed a MOS IC product line. Eventually, Moore himself would realize that Fairchild wasn’t going to fully realize the MOSFET’s potential. When he left Fairchild with Robert Noyce in 1968 to found Intel, it would be to create a semiconductor company devoted to manufacturing MOS ICs – memory ICs to be specific – but that event was nearly five years in the future.
To the Digital Age: Research Labs, Start-Up Companies, and the Rise of MOS Technology, Ross Knox Bassett, 2002
Moore’s Law: The Life of Gordon Moore, Silicon Valley’s Quiet Revolutionary, Arnold Thackray, David C. Brock, and Rachel Jones, 2015
Michael Riordan, “The Silicon Dioxide Solution,” IEEE Spectrum, December 2007, pp 51-56
Michael Riordan, “From Bell Labs to Silicon Valley: A Saga of Semiconductor Technology Transfer, 1955-61,” The Electrochemical Society Interface, Fall 2007, pp 36-41
Bruce E. Deal, “A Scientist’s Perspective on the Early Days of MOS Technology,” The Electrochemical Society Interface, Fall 2007, pp 42-45
Ross Knox Bassett, “MOS Technology, 1963-1974: A Dozen Crucial Years,” The Electrochemical Society Interface, Fall 2007, pp 46-50