The Man Behind the Microchip

by Leslie Berlin

  In the end, however, the desire to write on solid-state electronics outweighed all other considerations. Noyce asked Nottingham to advise a dissertation to be called “A Photoelectric Investigation of Surface States on Insulators.” Since early in the twentieth century, physicists had known they could study the movement of electrons in a solid by shining a light on the material. The light would excite the electrons within the bulk of the material, and as electrons made transitions from one area (called an energy band) within the material to another, they either absorbed or emitted a minute amount of energy that could be measured. These photoelectric measurements would indicate where and at what density within the solid the electrons congregated, and also how they moved under the stimulus of light.

  By the end of the Second World War, the photoelectric findings had given scientists a relatively good understanding of electrons inside solids. Inside a crystal, for example, every atom is firmly connected to every other. One might picture a neighborhood in which houses (atoms) are connected to each other by a clearly defined pattern of walkways that radiate from the front, back, and side yards of each house. Any electron wanting to move from one house to another needs to stick to the established pathways.33

  But what about the last row of houses in the neighborhood, the houses at the “surface” of the neighborhood, so to speak? These atoms have no back-door neighbors to connect to, no paths from their back yards for an electron to follow. Any electrons that made it out the back door of these houses would encounter a no man’s land where the yard ends and whatever-comes-next begins. The “surface states” that Noyce proposed to investigate can be thought of as this area where the end of the neighborhood meets the whatever-comes-next. In this surface state, electrons liberated from the established paths would not be as predictable as their counterparts in the interior of the neighborhood (the bulk of the solid). The electrons in the surface state would have a lot of room to move about, and in these wide open spaces they might even meet a few foreign intruders.

  Surface states remained an area of great confusion in the early 1950s. What quantum mechanical laws governed the behavior of electrons in the surface state? How many could squeeze into this space at one time? What could move an electron in or out of the surface state? How did the presence of impurities on the surface of a material affect its electrical properties? On the most fundamental level, for most materials, the existence of these “surface states” had yet to be demonstrated at all. Noyce wanted to see if they existed in two particular insulators: quartz and magnesium oxide.

  Noyce’s interest in surface states came directly out of the Bell Labs transistor research. The point-contact transistor did its most important work at the surface of the germanium, specifically at that point on the surface where the gold wires contacted the P-N junction. The device had been developed in the course of investigations into surface states in semiconductors; all three inventors wrote papers on the subject. For a while, Bell had even referred to its transistor research by the code name “Surface States Project.”

  The connection between the transistor and Noyce’s proposed surface-state research on insulators was tenuous at best, however. The Bell Labs surface-state research concerned semiconductors, not the insulators that Noyce proposed to study. Moreover, transistor research had taken a turn away from surface states in the years between the device’s invention at the end of 1947 and Noyce’s proposal in 1951. William Shockley, the leader of the Bell Labs transistor group, had invented a completely different type of device, called a junction transistor, a few weeks after the patents on the point-contact transistor were filed. The junction transistor did its work not on the surface, but in the middle of the semiconductor, which is where its P-N junction was located—picture a microscopically thin layer of N-type peanut butter between two P-type pieces of bread. This junction transistor, made public in July 1951, was more reliable, more easily produced, and capable of amplifying signals nearly a million times more efficiently than its point-contact predecessor. It was a marvelous device, but it had nothing to do with surface states.34

  Noyce undoubtedly knew about the shift away from surface states in transistor research. In fact, William Shockley spoke at a departmental tea around the time Noyce was writing his proposal. Noyce nonetheless understood that he needed to devise a thesis project that he could complete and that Nottingham could advise. Moreover, scientists continued to debate the merits of point-contact versus junction transistors. In any case, surface states were still a worthwhile subject of study, and investigating them would still enable Noyce to research the electrons, holes, quantum mechanics, and other topics he would need to understand if he wanted to work with transistors after he graduated.

  The project Noyce set for himself—to measure the presence of electrons at the surface of quartz and magnesium oxide—is today absolutely trivial, the near-instantaneous work of a $600 digital electrometer. In the early 1950s, however, Noyce found obtaining this information to be an ordeal that lasted more than a year.

  He needed perfectly clean samples, since impurities on the surface could cause confusing readings that would render his data useless. This meant he needed to conduct the experiment in a vacuum, where not even gas or oxygen could contaminate the sample. Noyce, like many of Nottingham’s students who were also conducting research that needed to run at high vacuum, built his experiment in a glass vessel that he hooked to a vacuum pump. The whole apparatus was then put into an oven and cooked at as high a temperature as the glass could withstand. As Noyce’s work progressed, the lab table assigned to him gradually disappeared beneath a layer of glass vessels, glass jars, glass tubes, glass pipes, and glass nozzles that he was using or had used at various points in his work. The other tables in the room looked the same way. The overall effect to the untrained eye was of a workshop belonging to a glassblower gone mad.

  “Oh, Noyce had a hell of a time,” chuckles one physicist who reviewed Noyce’s doctoral work some 50 years after its publication. Noyce needed to heat his sample to about 1,000° C, but it took him several tries to find a sufficiently powerful heating device. He tried two different light sources before finding a third that was bright enough to generate a measurable current. Twice the samples he was using shattered in the process of trying to clean them. Another time, a short developed between the heater and a back electrode, rendering the sample unusable. Each of these failures meant completely reconstructing the experiment from the beginning: requesting new equipment from the glassblowers who worked with Nottingham (or occasionally, if the job was easy enough, blowing his own glass), building the experiment inside this new equipment, pumping it out, and re-calibrating his instruments.35

  The instruments presented their own problems. The Compton electrometer Noyce had planned to use to conduct his measurements proved insufficiently sensitive, so he switched to a vibrating-reed electrometer, a higher-performing device that he needed to teach himself how to use. He soon discovered, though, that the device was so sensitive and the currents he wanted to measure so small that he was detecting interference from stray fields in the lab. He tried wrapping the leads to the electrometer with polystyrene to reduce interference, but this did not help. Finally he discovered that if he rid the experimental apparatus of any insulators other than two—the sample he was studying and the glass press of the vacuum tube—he could cut the interference from stray electric fields to nearly zero simply by slightly altering the relative positions of the tube and the electrometer from time to time. This was not an ideal solution, however, because every change of position entailed a wait of several hours while the background currents settled down.

  Noyce’s work was interrupted when he suffered a bad spiral fracture of the humerus from falling hard on his right elbow while practicing his ski jumping near Nottingham’s property. Noyce was in traction for two weeks—and in agony because a nerve just above his elbow was pinched between two pieces of bone. Gaylord, Maurice Newstein, and George Clark came up from Boston to drive Noyce to Hano
ver, where the doctors at Dartmouth surgically removed a piece of bone. Then a blood clot set in. By the time he was fully recovered and back at MIT, Noyce had lost almost two months of work. His arm would bother him for the rest of his life.36

  By the summer of 1953, his dissertation was nearly finished. In the end, it had proven a bit of a disappointment. Noyce had not been able to find evidence of surface states in the insulators he studied. To make matters worse, he had to admit he did not know whether his results meant that surface states did not exist in these materials or simply that his experimental methodology had been poor. The one consolation came from the work with magnesium oxide. He had managed to characterize its electrical properties, and the 11 curves he drew to demonstrate his findings were a small but real contribution to knowledge in his field.

  The most important lesson that Noyce learned from his dissertation could not easily be translated onto paper. He had developed outstanding laboratory skills as a solid-state experimental physicist. His early false starts had taught him how to prepare materials and how to keep them from contamination. He also understood photoelectric emissions, electrons, holes, quantum states, and the physical properties of solids.

  His knowledge and skills attracted offers to conduct research at Bell Labs at a starting salary of $7,500. IBM offered him a similar research position at $7,300 annual pay. Noyce spurned both these prestigious offers to accept a job at Philco, a Philadelphia-based company best known as a manufacturer of radios and televisions, for $6,900. Although Philco had 25,000 employees, only 30 worked in its semiconductor research group. Noyce liked the idea of working at a small company—perhaps, he thought, because he had grown up in a small town—and he also believed “[Philco] needed me very, very badly. They literally did not know what they were doing. I would be a necessary cog in that machine [whereas] in those larger, better-funded research organizations, I wasn’t going to have an essential role.” Noyce privately felt that the lack of a top-notch research team at Philco meant that he would have better opportunities to make a name for himself there than at either IBM or Bell Labs.37

  Meantime, he had to file the dissertation, a daunting task. In 1953, the MIT Physics Department required five pristine copies of the document, which meant a new set of carbon papers for each of Noyce’s 79 pages. A single typographical error, and the entire page was ruined. Noyce also hand-drew 22 figures included in the dissertation.

  IN AUGUST 1953, a few weeks before Noyce was slated to submit the dissertation, he placed an unprecedented person-to-person call to his father. As Harriet Noyce recalled it, the conversation was remarkably brief:

  “This is Bob, at Ephraim, Wisconsin. Will you marry us, Dad?”

  “You and who?”38

  THE GIRL WAS NAMED BETTY BOTTOMLEY, Noyce said. They were already en route from Boston, and they wanted to be married within the week.

  Bob Noyce had met Betty Bottomley while performing in a musical at Tufts College, for which Bottomley was the costume designer. (Noyce had accompanied a girlfriend to auditions for the play and ended up with a role himself.) Betty Bottomley was 22, small and slight, with short blonde curls that hugged her head. Bad eczema that left her skin flaky and bumpy kept her from conventional prettiness, but Noyce had not been drawn to her by her looks. Ever since high school, he had liked a woman to have a touch of acid in her heart and a “tounge [sic] as sharp as a razor,” as he once put it. Betty Bottomley certainly fit this description. A sickly child, asthmatic and unathletic in a family that spent a good deal of its time skiing and sailing, she had developed a slicing wit and an ability to comment on any situation with a jaw-dropping combination of humor and venom. She reminded more than one person of Dorothy Parker. Even her own family joked somewhat uneasily about the need to “sharpen our wits” when Betty was around. She seemed every inch the East Coast sophisticate that Noyce found attractive.39

  Betty Bottomley’s mother called her “a little human dynamo.” In raw energy, impulsivity, and strength of will, she was every inch Bob Noyce’s equal. She had graduated Tufts with a degree in English in 1952 and had spent the next year taking a graduate writing course, working in the public relations office of her alma mater, and writing for a literary magazine in addition to designing and sewing costumes for the Gilbert and Sullivan society. In her spare time, she enjoyed working complex word puzzles she found in magazines such as the Atlantic. She had also recently broken up with a boyfriend of several years, as Noyce made it his business to learn in relatively short order.40

  By the end of the play’s short run in May, the two were a couple. Three months later, Betty was typing Bob’s thesis for him. The shared sense of camaraderie and stress must have been exhilarating. The relationship certainly felt special to Betty, who brought Bob to meet her parents at their home near Providence, Rhode Island. She often invited groups of friends to stay, but inviting just one was unusual. Her parents noticed immediately.

  Betty’s family had more in common with the men Bob had served as a waiter at the Century Country Club than with the Noyces. Frank Bottomley was a vice president at the Sealall manufacturing company. Helen MacLaren Bottomley had raised four children (Betty was the youngest) while serving on committees at her club and volunteering at a school for handicapped boys from broken homes. The Bottomleys owned a large sailboat and spent so much time on the waters around their yacht club that they often said they might as well live in Narragansett Bay itself.

  For all their wealth—which was not truly substantial but may have seemed so to Noyce—the Bottomleys were unpretentious. Frank Bottomley, who had emigrated from England and left school after eighth grade, considered himself a glorified draftsman. Helen Bottomley had let the wind and sun weather her face and the simple braids she wound around her head each morning. She called the luncheons she organized at the club “production lines for the entertainment committee.” Noyce’s final assessment: the Bottomleys were the sort who joined a yacht club for the sailing, not for the society connections. He approved of this attitude.41

  The Bottomleys liked how Bob helped Frank work on the boat, and they appreciated the eagerness with which he joined them around the piano to sing Gilbert and Sullivan tunes from memory. He was “pretty much all one could ask for” her parents told Betty, fully aware that their opinions would hold little sway over her. As Helen Bottomley once put it, weariness and pride equally evident in her voice, Betty “has made her own decisions for a long time.”42

  THIS ONE-WEEK NOTICE for an impending marriage was a bit much, even for Bob, whose impulsivity and lack of communication were legendary in his family. Both Don and Gaylord had introduced their future wives to the Noyces before proposing, of course, but Ralph and Harriet Noyce had never even heard of Betty Bottomley. Bob dated so many women that he had long ago stopped mentioning them in an effort to escape the inevitable deluge of questions from his mother. Indeed, he wrote home quite infrequently by the end of graduate school, perhaps only a few times per year, and then only to report truly dramatic news—that he had broken his arm, or won an award.

  What was Bob Noyce thinking? To be sure, it was quite common for men to marry within months of finishing their PhDs. While there were good reasons for the timing—post-graduate jobs, income, and location often were not settled until a few months before graduation—dorm wisdom also warned that the light at the end of the tunnel had an unfortunate habit of bathing the nearest female in an idealized glow.

  The prospect of heading off to a new place for a new job without any companionship undoubtedly led to a few hastily proffered proposals, but Noyce would have welcomed a completely unfettered beginning after graduate school. Although he was only 25 years old, he had already developed a habit of mentally tying up one part of his life, shelving it, and never looking back. Shortly after coming to MIT, he admitted to his mother that he wanted, more than anything else, to “be free.” He had almost nothing to do with his childhood friends after he graduated Grinnell High School. He rarely communicated with college buddi
es after those first few months in Cambridge when he had actively sought them out as an antidote to the academic and financial demands of MIT. After he left graduate school, none of his colleagues, including the ones he lived with for years, would see or hear from him more than a handful of times in the next four decades.43

  Even Gaylord, his beloved older brother, felt the effects of Bob’s need for constant new starts. The two brothers drifted apart as the years passed. “Old friends and family can slow you down,” Gaylord pointed out. “There are birthdays to remember, letters to write, and calls to accept even if you don’t want to talk. Bob was not the type to slow down for much of anything.”44

  Why then would Noyce marry Betty Bottomley, whom he had known only three months? Betty told her friends that it had been a spur-of-the-moment decision, but again, this seems an unlikely explanation. Noyce was impulsive, but he was shrewd when it mattered. He would not have married on a whim. Years later Bob told his daughter that he and Betty had feared she was pregnant in the summer of 1953. He did not want another abortion on his conscience. It is unclear whether Betty was pregnant and miscarried, or whether the fears were unfounded, but the couple bore no children for 15 months after the wedding.45

  Harriet Noyce worked herself into such a frenzy over the prospect of hosting a wedding and entertaining the wealthy Bottomleys that she had no time to worry about the marriage itself. “I felt it simply could not be true,” she later wrote. “Maybe he had only said ‘will you marry us’ without saying ‘this week.’” Betty and Bob wanted a simple ceremony, but the logistics were daunting nonetheless. Harriet and Ralph had moved to Richmond, Illinois, only a few months before—Reverend Noyce had taken another new job—and they were not yet at home in the community. To make matters worse, the family car was broken. Harriet ordered a wedding cake from the local grocery store, dressed a chicken herself, and called on friends from Sandwich to bake rolls, play the wedding march, house the Bottomleys (who were arriving in four days), and decorate the sanctuary with pink and white gladiolas donated by Ralph’s congregation.46