The Man Behind the Microchip Read online

  Bob Noyce led a group focused on transistors, paying special attention to the work and results Bell Labs had reported on the diffusion process. Noyce had ample opportunity to do the collaborative good science he loved, preparing highly technical papers with his colleagues on topics such as “Carrier Generation and Recombination in the Space-Charge Region of a P-N Junction,” and “Localized Radiation Damage as a Source for Interstitials or Vacancies.”

  On August 14, 1956, Noyce noted an idea for a “negative resistance diode” in his lab notebook. With most diodes, the current flow increases with increases in voltage—the more voltage applied to the device, the more current passes through it. Noyce, however, made a startling prediction. He imagined doping a semiconductor with roughly a thousand times more impurities than was standard. When the voltage applied to this heavily doped diode increased from zero, Noyce predicted, current would also initially increase (as in any other diode). But, he said, as the voltage increased even further, “current must drop” because the high impurity density would make it possible not only for electrons, but also for holes, to transfer across the P-N junction—a phenomenon called tunneling. If one continued to increase the voltage through this period of “negative resistance,” Noyce theorized, the amount of current passing through the device would begin to rise, and the diode would resume normal behavior.38

  Noyce’s musings about a negative resistance diode excited him. They indicated that an important concept of quantum mechanics—“tunneling,” which existed only as a theoretical postulate—could be demonstrated in a simple P-N junction. If one thinks of conduction electrons in a semiconductor as balls bouncing against a wall (a wall built from an insulator or other potential barrier), quantum tunneling would predict that every once in a while, a ball would not bounce off the wall but would instead tunnel right through it.39

  Noyce brought his lab book entry to Shockley, fully expecting him to be impressed. Instead, “the boss showed no interest in the idea.” The lab was not equipped to do anything profitable with Noyce’s thoughts, and besides, Shockley did not like his employees to chart their own investigative paths. Disappointed, Noyce closed his lab book and “went on to other projects.” He later commented that this exchange with Shockley taught him that “the message of ‘no interest’ is certainly a powerful demotivator.’”40

  Noyce may have given up his ideas about a negative resistance diode, but the device soon reappeared in his life. On January 15, 1958, almost exactly 17 months after Noyce noted his ideas, a Japanese scientist named Leo Esaki published an article in the prestigious Physical Review describing the same negative resistance diode. The article caused quite a sensation in the electronics community. The audience for Esaki’s presentation at an international physics conference was filled to overflowing.41

  After reading the Esaki article, Noyce found his friend Gordon Moore. Noyce’s lab book pages and Esaki’s foundational paper are strikingly similar—they use almost identical illustrations, for example. There was, however, one important difference. Noyce predicted the drop in current (the evidence of tunneling) would occur. Esaki, who actually built a device to demonstrate his ideas, showed that it would. This difference is crucial—many good ideas die en route from the mind to the lab bench—and it is almost certainly a direct result of William Shockley’s discouraging comments to Noyce, who was an experimentalist at heart, in 1956. Noyce was irritated, primarily with himself for not pursuing his ideas even after Shockley dismissed them. “If I had gone one step further,” he told Moore, “I would have done it.” In 1973, Leo Esaki shared the Nobel Prize for physics for his work on the negative-resistance, or tunnel, diode.42

  None of Noyce’s colleagues aside from Moore knew about Noyce’s near-miss on the tunnel diode, but they had seen ample evidence of his advanced understanding of semiconductor physics, which Moore describes as “an equivalent amount of knowledge as the rest of us combined.” Noyce filed for four patents while he worked at Shockley; two of these patents listed William Shockley himself as Noyce’s co-inventor. When Shockley asked the staff to rank its senior members in terms of technical leadership, Noyce emerged as the top choice.43

  SHOCKLEY’S MANAGERIAL METHODS came as somewhat of a surprise to his employees. He had been careful to test his recruits for compatibility with each other, but no one had tested them for compatibility with their boss. Shockley spent hours in the machine shop telling Kleiner and Blank how to build the equipment, even going so far, in one case, as to redesign the bolts Kleiner wanted to use on the crystal puller used to grow silicon ingots. He sent Hoerni to work alone in an apartment near the lab, ostensibly because he did not want him distracted, but more likely because he felt threatened by the brilliant young theoretician with his clipped accent and pair of doctorates. Angry and lonely, Hoerni managed to talk his way back into the Quonset hut after only a few days’ banishment.44

  The work preoccupying the hot minds was fundamental. Physics, chemistry, optics, thermodynamics, fluid mechanics: every one of these disciplines had to work in concert to build a functional semiconductor device. Many of the basic properties of silicon were still not well understood. How did heat diffuse off a specific point? What happens if silicon is bombarded with atoms of argon? No one knew how to build production quantities of silicon transistors. What type of furnace would work best? How best to measure the thickness of the layers etched on the silicon in the manufacturing process? How efficient were semiconductors with melting temperatures lower than silicon’s?

  Answers to these questions would serve as the foundation for transistor manufacture throughout the world, for Shockley was not the only man to see the future in diffused silicon transistors. Indeed, by the time Shockley Semiconductor Labs began operation, researchers at Bell Labs, Texas Instruments, and Philco were already exploring the same issues that preoccupied the Shockley team. Closer to home, Hewlett-Packard had one man working in a similar vein. “We used to go back and forth with [silicon] ingots, like borrowing cups of sugar from one another,” laughs Julius Blank. “The blind leading the blind.”45

  Although there was some official reporting structure at Shockley, in reality, everyone worked with everyone else in the cavernous room, constantly comparing results and asking for feedback. What about this way of doing things? Do you see anything wrong? If I build a furnace, what do you want it to do? Any ideas about why I might have gotten this result from this experiment? It was a welcome change for Noyce, who was accustomed to the formal management hierarchy and equipment requisitioning procedures that Philco had adopted as part of its military contract work. At one point over lunch with his friend Julius Blank, Noyce started sketching out his ideas for a piece of equipment he wished he had. “I need a bell jar.… I want to expose the contacts …” Nothing more specific than that. A few days later, Blank handed him a rough model. “We just got through talking here!” Noyce shook his head in astonishment. “If we wanted the same thing at Philco, it would have taken six months!”46

  The scientists arranged informal technical seminars for themselves, each man familiarizing the others with his particular field of expertise. Shockley gave them each a copy of the textbook he had written, which was universally acknowledged as the world’s best semiconductor physics text. Shockley himself was an excellent teacher, regularly cutting to the heart of a problem in less time than it took other people to formulate it. “Shockley has this marvelous ability of going all the way back to first principles and making the right simplifying assumptions so you can wade through the mathematics and do it relatively simply,” Noyce explained in 1982. “If you were going to try to put wave equations down for every electron in a solid material, you would be in such a mess you couldn’t ever do anything. So you have to find representations that are manageable.” Shockley’s physical intuition was so great that his employees claimed that he could actually see electrons. He could speak on even the most technical topic with remarkable clarity. His best-known book on semiconductor theory, for example, starts by co
mparing the movement of electrons to cars looking for a parking space in a full garage.47

  The young men got to work by 9:00 and generally left around 6:30. At lunch time, groups of three or four would head for a burger at Kirk’s, the nearby greasy spoon, or if they were lucky, Shockley might take them to his favorite waiters-and-menus restaurant, the Black Forest, in his green Jaguar convertible. Several of the men shared an interest in hiking and mountain climbing. Their wives and children had become friends. In general, it was a happy time.48

  THE MORNING of Thursday, November 1, 1956, found the lab abuzz with excitement. Shortly after seven that morning, William Shockley had received a phone call notifying him that he, along with his Bell Labs colleagues Walter Brattain and John Bardeen, had won the Nobel Prize for Physics for their invention of the transistor. The workday began with champagne toasts to Shockley, who had not yet come into the lab. No matter. The young men were as much celebrating their own good fortune at working with a Nobel Prizewinner as they were happy for their boss. The award confirmed everything they had told themselves before they joined the company. They were indeed playing in the big leagues. Arnold Beckman, who certainly was feeling the same way, flew up from Southern California to offer his congratulations in person.49

  Later that day, Shockley interrupted the rounds of interviews, telegrams (he received more than 200), and phone calls to take his staff—now 40 or 50 strong—to the restaurant at Rickey’s, the hotel where many of them had stayed during their job interviews. It was a luncheon in high style: white tablecloths, flowers, and candles graced the tables, while the heavy curtains lining the walls muffled the group’s animated chatter. After coffee, several of the men assembled around Shockley, who was seated at the head of one of the tables. A flashbulb popped. The oft-reprinted photo sparkles with raised wine glasses and grinning young men in open-neck shirts all but patting their beaming boss on the back. Noyce, standing in square-jawed profile at the center of the picture, is strikingly handsome, his wineglass held carefully to avoid blocking the camera’s view of his own or his neighbor’s face. You can almost hear strains of “For He’s a Jolly Good Fellow” rising from the blurred background of the black-and-white snapshot.

  A month later, Shockley left for the award ceremonies in Stockholm, accompanied by his mother and wife Emmy. He arrived a day after Brattain and Bardeen, who had flown together with their families and spent a celebratory night on the airplane catching up and exchanging stories of what happened when they got the news. Brattain had been greeted by a standing ovation when he reported for work at Bell Labs. Bardeen had been so surprised by the call from Sweden that he had dropped the pan of eggs he was cooking for breakfast.50

  The award ceremony was a glorious affair. Men in white tie and women in formal gowns rose to their feet in Stockholm’s elegant concert hall as King Gustav VI Adolph presented each man with his award. Bardeen, Brattain, and Shockley each gave a short talk. While the other laureates limited their comments to science, Shockley ended his with plugs for his own prescience—he had predicted a great future for “transistor electronics” as early as 1950, he reminded his audience—and for his “new organization in California.”51

  Shortly after midnight, when Shockley and his wife wandered into the elegant bar at the Grand Hotel, they saw Brattain and Bardeen already sharing a drink. Though the scientists had scarcely spoken to Shockley for more than five years, the occasion was too special to nurse old grudges. They invited him to join them.52

  SHOCKLEY RETURNED FROM STOCKHOLM at the end of December in spirits so high they verged on egomania in the eyes of his staff. He gathered the lab for a little speech. He said that when he received the Nobel Prize, he had felt like Churchill. He added, as an aside, that it was “about time” his contributions were appropriately recognized. At first the young scientists thought he was making a joke. He was quite serious.53

  Shockley’s behavior had grown increasingly erratic in the months leading up to his Nobel Prize. He had never been an easy man to like, even when he was trying to be likeable. When he gave a raise to Jay Last, the youngest scientist, for example, he did so with the admonishment that Last never should have agreed to work for the $675 he had previously been earning. “Jay, that will teach you never to sell yourself out cheap again.” Shockley seemed not to trust his hand-picked team. He infuriated Noyce by calling Bell Labs to double-check his interpretations of data. “Am I really needed here?” Noyce asked himself. “If he can call friends at Bell Labs and get answers to the same questions that I [am] trying to answer in the laboratory, my presence here isn’t that important.”54

  Shockley strongly adhered to a belief that in almost every aspect of life, there could be but one winner—and he wanted to be it. He had himself listed as a co-presenter and co-author on every one of the papers given by his employees at the elite American Physical Society conference in December 1956. Lunchtime workouts in the Stanford pool were always races for Shockley, who pushed himself to exhaustion if it appeared another man was swimming faster or farther. He once told an employee that although many people could write a paper together, patents should officially list only one inventor because “there’s only one light bulb to go on in somebody’s head…. The other [people] are mere helpers.” He nonetheless listed himself as a co-inventor on his employees’ most significant patent applications.55

  Shockley would attack people when they made mistakes—“reduce them almost to tears,” in the words of one former employee. One man who worked with Shockley in a different context described his behavior thus: “He could be helpful, but you had to go through a ritual humiliation first.” Where’d you go to school? Are you sure that you actually went to school? How could you not know something like this? In one instance, Shockley publicly fired a secretary who had not made travel arrangements in the manner he had specified. The other employees were horrified.56

  In the months after the Nobel Prize ceremony, recalls Jay Last, Shockley’s behavior deteriorated to the point that the lab came to resemble “a big psychiatric institute.” When he wanted people to leave the building, Shockley would quote T. S. Eliot’s “Love Song of J. Alfred Prufrock” (“Let us go then, you and I …”) rather than simply announcing an end to the day. When a secretary cut her hand on a tiny piece of metal protruding from her door, Shockley convinced himself that it was deliberate sabotage. He informed the lab that he planned to hire a private investigator and that they would all need to undergo lie detector tests to find the culprit. Sheldon Roberts spent the better part of a week proving to Shockley, with the help of a microscope, that the offending object was simply a thumbtack that had lost its protective plastic head.57

  Jay Last, for one, thought things were becoming intolerable. He and Hoerni found working with Shockley so frustrating that nearly every weekend they would drive south for hours and hike, mile after mile, complaining and “kicking Joshua trees” to vent their frustrations. Last had observed that Noyce’s relationship with Shockley seemed strong, and one day he described to Noyce his own troubles with the boss and asked for advice. Noyce could not offer much in the way of concrete suggestions to improve the situation, but Last left feeling that at least he had unburdened himself of his concerns. The next day, Shockley barreled up to him and began shouting: “What the hell did you say to Noyce?” Last was shocked. “Bob said he thought [telling Shockley] was ‘the best way to handle it.’ It sure wasn’t best for me.”58

  Shockley’s abrasive style, with its increasingly frequent dips into irrationality, was only one stress facing the young men working for him. “There was always this business about ‘how would Shockley respond to this?’” explains Vic Jones. Every idea, every potentially exciting new process—it all had to meet with Shockley’s approval, or it could not go further. The young scientists found themselves putting the brakes on their own ideas in anticipation of Shockley’s disapproval. It did not help that Shockley was a night owl who did some of his best work over cocktails and expected his favorites, includin
g Noyce, to join him at the bars.

  To make matters worse, far from raking in the profits and churning out the products that Shockley and Beckman had anticipated when they drafted their business agreement, Shockley Semiconductor was losing money and had yet to sell anything. By contrast, the data-processing research that Beckman Instruments funded at roughly the same time as Shockley Semiconductor had already yielded a prototype system and orders from leading firms such as Dupont, Westinghouse, General Electric, and General Motors.59

  The nub of the problem was simple. As a former employee put it, “Shockley ran the company for the benefit of his personality and his image, not for pure economic pay.” Succeeding as a businessman meant foregoing interesting explorations of basic physical phenomena for the more mundane task of building a usable product, but Shockley feared losing his status as one of the world’s leading solid-state physicists. One raised eyebrow from a scientific colleague or a few comments to the effect that he had not been publishing much lately, and Shockley would completely reorient his company, telling his employees to stop whatever they were doing and start writing up their findings for presentation at the next prestigious scientific meeting. If, in the course of one of these periods of intense work on basic science, someone from outside the company asked what was happening on the business front, Shockley was likely to reverse course again and throw the company into a flurry of product-oriented activity.60

  Further adding to the troubles was Shockley’s obsession with a device he had conceived while still at Bell Labs, the “four-layer diode”—a diode that, as its name implies, has one more layer of diffused semiconductor than a transistor. Shockley hypothesized a semiconductor diffused in four alternating layers doped P and N could do the work not only of a transistor, but also of a resistor. It would also be faster and cheaper than either conventional diodes or germanium transistors. Shockley was quite certain that Western Electric would want to buy thousands of these four-layer diodes to replace the hundreds of thousands of mechanical relays that switched and connected calls across the telephone grid, and he imagined that computer companies would be interested in the product, too. Shortly after his return from the Nobel Prize ceremonies, Shockley pulled a half-dozen senior scientists off the silicon transistor project to work on the four-layer diode.61