True Genius: The Life and Science of John Bardeen: The Only Winner of Two Nobel Prizes in Physics (2002)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

9 The Break from Bell I n the days following their invention, Bardeen and Brattain must have had an overwhelming urge to continue working, despite the holidays that had crept up on them. Their families wanted them home to help celebrate. Nature sided with their families. An unusually heavy snow began to fall on Friday, December 26. All staff members were sent home after more than a foot of snow had accumulated at Murray Hill. Across the Hudson, New York City had been “utterly unpre- pared” for the storm that the New York Times reported as having “swept in from the Atlantic at a point where there are no weather observers.” By Saturday morning, after a night in which the storm dropped 25.8 inches of snow, transportation had ground to a halt. “The city’s towers wore tremendous tufts and beards of snow.” Worse than the blizzard of 1888, which had set New York City’s previous snowfall record of 20.9 inches, the blizzard of 1947 pro- duced “the greatest snowfall in the city’s recorded history.” From Summit, Jane wrote to her mother that they were having “one hell of a storm.” The snow had turned to freezing rain, com- plicating New Year’s Eve celebrations. Most residents were with- out power for nearly a week. The Bardeens had to borrow a neighbor’s gas stove to prepare meals. Jane described it as “great sport (?) stumbling through the snow from their back door to ours with a tray of hot kettles.” She said that the members of her hungry 142 

The Break from Bell 143 family were “all ready, with forks in hand, around the kitchen table when I came in. I told them I felt like the zoo keeper who throws fish to the seals!” Shockley worked at home all through the storm and even after the sun finally came out on Sunday, December 27. The Christmas holidays of 1947 could not have been joyous for Jean Shockley or for Bill’s and Jean’s two young children, Billy and Alison. Shockley stewed in the knowledge that Brattain and Bardeen had made an important, perhaps revolutionary, discovery and that he had not been part of it. Because the work had been motivated by his designs, Shockley, as Bardeen later reported, “thought he would have the basic patent on the field effect principle.” Shockley resolved to explain to Bell’s attorneys why the patent should be written in his name. While the children of New York and New Jersey dug tunnels in the snow, Shockley struggled alone to develop his idea for a new kind of bipolar transistor that would be a significant improvement over Bardeen and Brattain’s design. Unlike the cumbersome point- contact device, his he expected would be commercially viable. It would use junctions of p- and n-type silicon or germanium in place of the electrically noisy point contacts. This idea had been working its way through Shockley’s mind at odd moments over the last month while the point-contact transistor was born. He had kept his work hidden from the semiconductor group. On Monday, December 28, Shockley ran his designs over to Murray Hill and got an experimentalist, Richard Haynes, to wit- ness the documents. Then he took the train to Manhattan and boarded the Twentieth Century Limited to Chicago. There Shockley attended a meeting of the American Physical Society on December 30 and 31. Afterwards, he returned to his room at the Bismarck Hotel and worked for almost a week on his new design: a three-layer “sandwich” in which the “bread” was of p-type semi- conductor and the “meat” was of n-type. One of the outside layers would supply the holes; the other would collect them. The n-type layer acted as the grid, the “faucet” that controlled the current by creating an electrical barrier whose height he later realized could be raised or lowered by changing the voltage. The idea would evolve into the junction transistor. Back at Murray Hill, Kelly clamped down on communicating with the outside about the Bardeen and Brattain discovery. Imitating the secrecy procedures of the Manhattan Project, he 

144 TRUE GENIUS declared their invention “Bell Labs confidential.” He code named the work the “Surface States Project.” Jane had already written to her parents about “John’s invention.” Now she hurriedly wrote back to ask that they keep the news confidential “until the patent arrangements have been made. The lawyers are working on that now.” She added proudly that “all the top ‘brass’ of Bell has been out to see it, so it must be good.” Kelly also insisted that Bardeen and Brattain drop everything else and focus on drafting patent applications. In effect, by tying up their time, he was offering Shockley an open field to work on his new junction device. Bardeen and Brattain worked closely with the patent attorney, Harry Hart, who came to think of the two as one combined individual. Years later he wrote a joint letter to them, because he always “thought of you as a pair.” He added, “Your personalities are as different as those of Jack Sprat and his wife. Perhaps that is the reason you were able, between you, to lick the platter so clean.” When Shockley returned to Murray Hill from Chicago, he took a further step that ruined any possibility of future teamwork with Bardeen or Brattain. He called each of them separately into his office and told them, as Brattain recalled, “that he could write a patent—starting with the field effect—on the whole damn thing.” Bardeen and Brattain were stunned. Shockley added the final insult when he said, “Sometimes the people who do the work don’t get the credit for it.” But Shockley’s plan failed. The attorneys discovered that the field effect idea was not original after all, but had “turned up in a number of forms in the patent literature.” In 1930 the idea had been patented by the Polish-American inventor, Julius E. Lilienfeld, then a professor of physics at Leipzig. Lilienfeld’s 1926 patent application had described the field effect concept so well that “there was no way that Shockley would get a patent on the field effect principle,” Bardeen later explained. It did not matter whether Lilienfeld’s device actually worked. Bell Labs could not risk the rejection of a claim based on the already patented field effect idea. Bell therefore based its application on Bardeen and Brattain’s bipolar design. That idea was new because holes—rather than electric fields—were functioning as the grid. Shockley must have felt that he was losing on all sides. He continued to work in solitude on his junction device, while Bardeen 

The Break from Bell 145 and Brattain worked on the patents for their bipolar device. Then one day Shockley suddenly revealed his secret work. On February 18, experimentalist John Shive demonstrated a design for a transistor to his Bell Labs colleagues in which the points were on opposites of a sliver of germanium. The fact that Shive’s device worked indicated that holes were indeed traveling through the sample. Shockley knew that Bardeen would instantly recognize the im- plications of Shive’s demonstration, if indeed he had not already done so. Shockley could no longer afford to keep his work quiet, for Bardeen could easily scoop him. So at the demonstration Shockley abruptly rose to present his design for a new type of bipolar ampli- fier based on p-n junctions. The design resembled his New Year’s Eve conception, except that it was an n-p-n rather than a p-n-p sandwich, in which electrons, rather than holes, carried most of the current. In addition, electrical contacts were attached to all three layers, allowing better control of the current. Brattain was appalled that Shockley “went off by himself and worked at home” and especially that he had “ceased being a member of the research team.” Bardeen also fumed, but he, charac- teristically, said little about it. Jane and a few others, including Brattain, knew how much Shockley’s behavior disturbed Bardeen when he would mutter through his teeth that Shockley had “jumped in with both feet.” Bardeen and Brattain had other worries, too. With Kelly’s secrecy order in place, they could not publish or even speak about their invention during the first six months of 1948. They also were anxious about being scooped, particularly by the physicists work- ing under Karl Lark-Horowitz. Ralph Bray, a member of the Lark-Horowitz group, had reported a “spreading resistance” in germanium. As a graduate student dur- ing the Second World War, Bray had observed that when small pulses of positive voltage are applied to points in contact with germanium, the electrical current spreads out under the points, lowering the resistance. He did not yet realize that injected holes were causing the phenomenon, but Brattain and Bardeen felt that sooner or later Bray or another member of his group would figure this out. In January 1948, at an American Physical Society meeting in New York, Brattain discussed Bray’s experiments in the corridor 

146 TRUE GENIUS with Seymour Benzer, another member of the Lark-Horowitz group. Benzer said, “I think if somebody put another point contact down on the surface, close to this point, and measured the distribution of potential about the point, then we might be able to understand what this is all about.” Brattain responded uneasily before walking quickly away, “Yes, I think maybe that would be a very good experiment.” Another group was indeed working on a similar device in Paris. Heinrich Welker and Herbert Mataré were creating a germanium device similar to the Bardeen-Brattain point-contact transistor. They filed a patent for it on August 13, 1948. Historian of science Kai Handel, who has written about Welker and Mataré’s work, asserts that “it is probable that in any laboratory concerned with high purity Germanium and with experience in experiments with point contacts, somebody would perform experiments with two point contacts close to each other.” Like the Purdue group, the Paris team did not have the benefit of a theorist like Bardeen with a thorough understanding of the phenomenon. Handel adds, “It is certain however that they didn’t—unlike Bardeen and Brattain— succeed in understanding how their device really worked.” Military classification was another concern. Bell Labs felt obliged to get a military release because of the amplifier’s possible military applications. While they waited for that release, Bardeen and Brattain sent a letter about their invention to the Physical Review, asking the editor to hold back publication until Bell Labs had officially heard that the invention would not be classified. Again, Brattain had difficulty containing his impatience. He felt “very strongly that most restrictions done in the name of national security turn out to be foolish.” Not until June did they hear, with considerable relief, that their device would not be classified. In the meantime it was crucial to find a good name for the invention. Suggestions included semiconductor triode, surface- state amplifier, crystal triode, and iotatron (to emphasize its small- ness). Many names were dropped because Brattain and Bardeen didn’t like words ending in ‘tron.’ In May 1948, a Bell Labs electri- cal engineer found the best name. John Pierce happened to stop by Brattain’s office. “John, you’re just the man I want to see,” said Brattain. “Sit down.” Pierce had a way with words. He would later write science fiction stories under the pen name of J. J. Coupling. As the two men discussed the nam- 

The Break from Bell 147 ing issue, Pierce pointed out that the vacuum tube worked because of its transconductance, while the new device worked because of its transresistance; it was the electrical dual of the vacuum tube. Brattain said that he thought the name should fit with the other names used for solid-state devices, such as “varistor” and “ther- mistor.” Pierce suggested “transistor.” “Pierce, that is it!” Brattain exclaimed. All the suggested names were circulated in a Bell Labs memo- randum. Those on the distribution list were asked to vote on the best name for the new device. “Transistor” was the overwhelming favorite. The invention was finally released to the public on June 30, 1948. May Shockley, Bill’s mother, flew in from California and joined Jean, Keren, Jane, and their husbands for an elegant lunch in the executive dining room on the top floor of the West Street build- ing. The women watched while the men were photographed. A few pictures were taken of them as well. Then all were called into the auditorium. Ralph Bown, then the Bell Labs’ director of research, was pre- paring to speak. He was smartly dressed in a tailored gray suit with a colorful bow tie. “What we have to show you today represents a fine example of teamwork,” he began. He stressed “the value of basic research in an industrial framework.” Then he went on to explain what “this cylindrical object which I am holding up” actu- ally did: We have called it the Transistor, T-R-A-N-S-I-S-T-O-R, because it is a resistor, or semiconductor device which can amplify electrical sig- nals as they are transferred through it from input to output termi- nals. It is, if you will, the electrical equivalent of a vacuum tube amplifier. But there the similarity ceases. It has no vacuum, no fila- ment, no glass tube. It is composed entirely of cold solid substances. Bown demonstrated the device by amplifying his own voice. He spoke through a telephone handset to members of the audience who were holding receivers into which he inserted, and then removed, one of the new transistorized amplifiers. He emphasized that the new transistors, unlike vacuum tubes, needed no warm- up. After the demonstration, Shockley fielded questions. Bardeen did not initially notice the shadow that Bown’s dem- onstration cast on his and Brattain’s role in the development of the transistor. “Everything went well. I thought it was a very success- 

148 TRUE GENIUS ful demonstration,” Bardeen said. The Bardeens and the Brattains had “received VIP treatment and were very thrilled to be there,” Jane gushed to her family. She mentioned that “the final perfect touch was being chauffeured home through the blazing heat in a seven-passenger Packard.” To the public, however, Bown’s presen- tation and Shockley’s articulate handling of the audience questions gave the impression that the transistor’s development had been orchestrated from above. The invention was presented as the result of teamwork in a group directed by one of Bell Labs’ most brilliant physicists. Their spirits dampened, recalled Bardeen, when “we read the New York Times the next day” and found “they had just a few words in the radio column about it, and that’s all.” The Times reported that “a device called a transistor, which has several appli- cations in radio where a vacuum tube ordinarily is employed, was demonstrated for the first time yesterday at Bell Telephone Labora- tories.” The account, buried on page 46, near the end of the paper’s regular News of Radio column, was given lower priority than the notice that on Monday evenings during the summer CBS would air the popular radio comedy Our Miss Brooks in place of its regular Monday evening feature, Radio Theater. The New York Herald Tribune was more upbeat, saying that “engineers believe it will cause a revolution in the electronics industry!” Radio buffs responded enthusiastically to the news. Instead of waiting for transistors to hit the market, they “would just buy a germanium dial and make themselves a transistor,” reported Brattain. “A radio set without vacuum tubes,” Jane marveled in a letter home, “just John’s gadget and a speaker. Won’t that be some- thing!” Engineers were perhaps the most excited of all. In September 1948 Electronics magazine featured a carefully posed image of Bardeen, Brattain, and Shockley on its cover. They had been photo- graphed in the laboratory wearing dress shirts and ties. But neither of the two inventors was at the center of the photo. Instead Shockley was posed at Brattain’s workbench, while Brattain and Bardeen looked on from behind. The portrait was staged to convey a message consistent with Bown’s press conference. “Walter sure hates this picture,” Bardeen told a friend. All three men now hit the lecture circuit. Bardeen spoke in Chicago at an American Physical Society meeting, at Purdue in 

The Break from Bell 149 Indiana, and at the Oak Ridge Laboratory in Tennessee on the way home. In January 1949, Bell Labs, now welcoming publicity, sent Life magazine photographers to the Bardeens’ home for a planned article on the transistor. The day before, John went downtown to “get a haircut so he’ll look pretty.” Jane worried that John was exhausting himself. In late October she wrote home that he “has been under considerable pressure lately writing a paper for Physical Review on the transistor and giving lectures. He spoke twice last week at the labs and on Friday at Columbia.” He was planning to be gone most of the month of November giving talks in San Francisco, Los Angeles, and Chicago and would “not be home again until the weekend after Thanks- giving.” Bell Labs prepared illustrations and demonstrations for use in the talks. In 1949, to highlight the transistor’s portability, an engi- neer at West Street built three music boxes. Each contained an oscillator-amplifier circuit and two of the early point-contact transistors. Bardeen and Brattain were each given a box; the other was for Bell Labs. Bardeen’s box (now preserved at the Spurlock Museum of the University of Illinois) was the only one that did not disappear over the years. “Dad was very proud of this music box,” said his son Jim. John kept his box in his office safe and it still worked more than fifty years later. Bardeen would use the box in lecture demonstrations to illus- trate the transistor’s use as both an oscillator and an amplifier. Referring to a small piece of paper taped on top, he would key in the tune “How Dry I Am,” which could be heard over a loud- speaker. Then he would amplify the sound by switching the second point-contact transistor into the circuit. The Prohibition-era drinking song would usually produce laughter in the audience, easing John’s way into his lecture. Jane was given one of the first transistorized hearing aids. The earlier-generation vacuum tube hearing aids were large devices, usually worn around the waist or sometimes held in a pocket. Con- sumers were willing to pay the high cost of improving them. They were limited by the expense and weight of the battery needed to power them. Transistorized hearing aids would eventually become more affordable and less cumbersome, but the early transistors cost about $200 each and were about the size of a pencil eraser. In recog- nition of Alexander Graham Bell’s commitment to helping the deaf 

150 TRUE GENIUS and hearing impaired, AT&T offered hearing aid manufacturers roy- alty-free licenses. Throughout 1949 and 1950 Bardeen grappled unhappily with the stress of his continuing lectures and travel, and especially from the deterioration of his relationship with Shockley. Other staff at Bell Labs could see that Shockley “wanted to have everything under his control,” limiting the freedom of those who worked under him. Shockley was breaking with the enlightened philosophy of research that had drawn Bardeen to Bell Labs in the first place, but Kelly and Bown did not intervene. Conyers Herring was among those who noticed that Shockley had “a great measure of confidence from the higher management,” who were impressed “with his mental brilliance.” That confidence from above meant a lot to Shockley, who, Herring sensed, “felt all alone.” Shockley wanted recognition and ultimately to “become a big name in the world of business.” He thought he “was brighter than anyone else, and in many ways he was.” Contributing to his growing isolation, Shockley would let others know that he did not consider most of them intelligent enough to make good company for him. Bardeen and Brattain were distressed when Shockley banned them from the work on the junction transistor, a project they saw as the continuation of their work on the bipolar transistor. In the fall of 1948, Shockley went so far as to assign Bardeen and Brattain to offices on a different floor from his in the newly completed Build- ing 2. Shockley directed them to work on the surface effects of point-contact transistors, an area he “probably knew was a blind alley.” When Bardeen complained, Shockley proposed that Bardeen could go and work in development with Jack Morton or in radio engineering with Russell Ohl, “but these solutions naturally did not appeal to me,” Bardeen later wrote. Like Shockley, Bardeen had come to Bell Labs because of the promise to work on his own basic research. Shockley’s policy “meant that I could not play an active role in the semiconductor research program.” Bardeen was directed to work only on problems “in which Shockley was not interested himself.” Bardeen went to speak with Ralph Bown, but Bown was reluc- 

The Break from Bell 151 tant to interfere. Shockley, Bown said, was “in a highly emotional state” and working seventy or more hours a week. Bown assured Bardeen that “the difficulties would be resolved in time. Bardeen deduced that “it was the desire of the administration to give Shockley a free hand.” In April 1948 Bardeen had seriously considered accepting a per- manent position at Oak Ridge, Tennessee, directing solid-state work in support of the reactor program of the Atomic Energy Labo- ratory. He thought about going there for six weeks, or possibly a year, to try the job out. Ultimately he decided not to accept the offer from Oak Ridge because he still hoped his position at Bell Labs would improve. “I am still interested in spending some time at Oak Ridge, but so far I haven’t been able to get out from under the work on the transistor,” Bardeen wrote to the director of the Oak Ridge laboratory, Alvin Weinberg. He held the door open. “If you want to give me a rain check, I will let you know when things let up here.” A year later, when Weinberg renewed his offer, the situation had not improved. Bardeen wrote back, “You make a very strong case for spending a period at Oak Ridge on the basis of a leave of absence from the Bell Laboratories. Although I am not certain of it, I believe that if the situation warranted it such a leave could be arranged.” With his Oak Ridge offer in hand, Bardeen wrote for advice to Jim Fisk, who was then at Harvard in the Department of Applied Physics. Bardeen explained that Weinberg “has been in contact with me, first in regard to a permanent position at Oak Ridge which I turned down, and now in regard to spending a period of a year or so there on leave from the Laboratories. He makes a strong case for the latter.” Bardeen understood that his position at Oak Ridge would be to “act as a senior advisor for the solid-state group.” His contributions “would be connected with fundamental investigations into the properties of matter subjected to nuclear bombardment.” Bardeen asked Fisk, “What do you think of the proposition?” Fisk, then slated to return to Bell Labs as director of physical research, advised Bardeen to wait the crisis out. Jane was secretly relieved about John’s decision to stay at Bell Labs. She enjoyed living in Summit. She also worried about how John would fare in a government position. “The rising anti- intellectual sentiment makes any government service by a professor 

152 TRUE GENIUS or scientist a risky business. Let’s hope we’re not heading toward a fascist state.” Committed to liberal principles, the Bardeens were opposed to the growing atmosphere of intolerance caused by the “red scare” and Senator Joseph McCarthy’s infamous hearings. The House Un-American Activities Committee had already attacked several scientists. Jane recommended that her family read a New Yorker article about William Remington, “one of the victims, last summer, of the House Un-American Activities Committee. I remember telling Betty that the Kittels knew him and were sure that he was not a Communist, nor guilty of giving information to the Communists as that crackpot E. Bentley charged.” At Bell Labs “conditions if anything got worse,” but Bardeen did find one project to have a little fun with. Collaborating with Herring, he wrote a paper criticizing some aspects of an earlier article by Fred Seitz on the Kirkendal effect in alloys. Bardeen had discussed the problem with Herring, telling him “he wanted to study diffusion from a rather fundamental point of view.” Herring had been working on the problem of diffusion from a similar per- spective. And as Bardeen “rather liked” Herring’s approach to the problem, he suggested they publish a joint paper. In it Bardeen dealt with the Kirkendal effect and Herring treated what became known as the correlation effect in diffusion. Herring had by this time become the Bell Labs’ authority on the latest research in solid-state physics. He was the one to whom people would usually come with questions about the physics literature. He took his role of guru seriously, but the field was grow- ing so rapidly that he felt he could not reliably retain all the infor- mation in his memory. So he began to carry around a made-to-order roller-skate box filled with 3 × 5 index cards containing notes and references to the latest work in the field. The box had begun as a notebook when Herring was at Princeton. After the war he replaced the notebook with a set of filing cards, initially kept in a file box, but the cards soon outgrew the box. One day, while attending a meeting of the American Physical Society in New York, he “stopped in at a luggage store there, and had them make me to order there a suitcase so big that I could put 3 × 5 cards in.” The box served Herring until the personal computer took over its function in the 1970s. But from the 1940s through the 1960s, everyone who worked on solid-state physics at Murray Hill depended on Herring to tote the institution’s solid-state database around in his roller- skate box. 

The Break from Bell 153 Bardeen found some relief from his anxieties at work in family activities. John bought Jane a corsage for her birthday in April 1948. When the family went out to dinner at the William Pitt in Chatham, the men wore pansies in their lapels; Betsy wore a pansy corsage. In January 1949, John took nine-year-old Jimmy skating. “Dad and I went skating at Lake Surprise three times,” Jimmy wrote to Jane’s mother. “The first time I could just stand up. The second time I learned to skate.” In May 1949, to celebrate Jim’s tenth birthday, the family picnicked along the Musconetcong River. “Even had candles on the chocolate cake!” A few months later, Jane wrote to her mother, “John wants to take the kids to the zoo in New York City this weekend.” Betsy adored her father. Like him, she “could be very deter- mined.” At age four she refused to take off her new patent leather shoes until her daddy could see them. John, who never lost his love of sweets, abandoned his diet when almost-five-year-old Betsy of- fered him tidbits from a box of her own candy. The larger community of solid-state physicists formed another kind of supporting family. Bardeen became a father figure there too, one to whom other physicists came with questions of all kinds. He was always willing to help a young scientist. Nobel laureate Walter Kohn recalled his first interaction with Bardeen. Kohn was then working under Van Vleck on a project that used a table pub- lished by Wigner and Seitz. One day V2 [Van Vleck] remarked that he believed that through some misstep this table contained a wrong (perhaps preliminary) set of numbers and that I should write to John [Bardeen] who, V2 thought, had the right numbers. Of course, John was already very famous thanks to the creation of the transistor, while I was an absolute zero. With considerable trepidation I followed V2’s advice. Very promptly the reply came from John in a brown manila envelope. Contained therein was the correct table of numbers (V2 had been right!), accompanied by a long, handwritten letter which explained how the mix-up had occurred and described in detail a mathemati- cal procedure for making the best use of these numbers (I have used that procedure ever since). Kohn later wrote to Jane, explaining that “this—in some ways— small experience had a big effect on me.” Bardeen’s gesture “showed me a great scientist patiently giving a leg up to a raw beginner whom he had never even met.” 

154 TRUE GENIUS That kind of interaction with Bardeen became legendary—so much so that physicists at Bell Labs developed a number to describe it, the “Bardeen number.” Bell Labs physicist Ravin Bhatt (a student of Bardeen’s student William McMillan) described it as “the ratio of substance to self-advertisement.” Bhatt explained that, in general, if a scientist even made number one on that scale “you’re doing good. Most people these days are way below that.” But “Bardeen was infinity on that scale.” But while Bardeen’s interactions with his family and physics community often improved his mood, they did so only briefly. He grew increasingly unhappy throughout 1948, 1949, and into 1950. “Bardeen was fed up with Bell Labs—with a particular person at Bell Labs,” said Brattain. One day early in 1950 Bardeen just stopped working on the prob- lems that Shockley had relegated to him. With Fisk’s approval, Bardeen turned back to the problem he had been studying at the time he was called to Washington, D.C. Explaining superconduc- tivity was still the outstanding unsolved riddle of solid-state phys- ics. Bardeen had not thought extensively about superconductivity since 1941, but when he pulled out his decade-old notes it felt like coming home. In his prewar study of superconductivity, Bardeen had come to focus on the energy gap in the electronic structure of super- conducting materials, a notion highlighted in the work of the London brothers. Bardeen had tried to explain the gap using a single- electron quantum-mechanical model. He had explored whether applying a small periodic distortion to the lattice could lower the energy and introduce band gaps near the Fermi surface. That approach had failed. Still he felt certain that the energy gap was crucial to the phenomenon. He could not explain why. Nor could he account for the gap’s presence. Confronting this challenge dis- tracted him from the aggravation of working under Shockley. It also led him to his most important scientific work. Bardeen approached his return to superconductivity in the same way he approached every problem. He considered how to break the problem down into smaller parts and visited the library to see what others had done before him. He was pleased to find much new evidence supporting the Londons’ notion of the energy 

The Break from Bell 155 gap. By now many experimenters were working on superconduc- tivity. They included A. Brian Pippard, William Fairbank, Emanuel Maxwell, Paul Marcus, J. C. Daunt, Kurt Mendelssohn, B. Goodman, A. Brown, Mark Zemansky, Henry Boorse, W. Corak, Michael Tinkham, and Rolfe E. Glover. Pippard, David Shoenberg’s student, whom Bardeen had met in Cambridge in 1947, during his trip in Europe with Shockley, was now a professor at Cambridge. Pippard’s studies of the surface impedance of normal and superconducting tin supported the London theory, as did related experimental work in the United States by Fairbank at Yale and by the MIT team of Maxwell, Marcus, and John Slater. On May 15, 1950, Bardeen received a phone call that riveted his attention on superconductivity. Bernard Serin, an experimen- tal physicist in the Department of Physics at Rutgers University in nearby New Brunswick, New Jersey, wanted to discuss the impli- cations of certain recent findings of his group. In studying mercury isotopes Serin’s group had found a clear and definite “isotope effect”: the lighter the mass, the higher the temperature at which the material becomes superconducting. The stable and pure mer- cury isotopes they had used, having mass numbers between 198 and 202, had been prepared at Oak Ridge by bombarding gold with neutrons and separating the products using a mass spectrometer. They were available as a consequence of the wartime atomic bomb program. Emanuel Maxwell, then working at the National Bureau of Standards, had independently found the isotope effect studying isotopes from Los Alamos. Serin’s colleague, Henry Torrey, who during World War II had been nicknamed the “Crystal Crackin’ Papa” of the MIT Rad Lab’s crystal rectifier program, remembered “Serin coming to me when he got the results.” He “was very excited.” The fact that the mass was involved hinted that the lattice was too! As Torrey recalled, the observations “immediately suggested to Serin the relationship to the electron-phonon interaction.” That was what Serin wanted to discuss with Bardeen; Serin had heard that Bardeen “knew a lot about” superconductivity. The next day Bardeen scratched a note to himself about Serin’s call: “These results indicate that electron–lattice interactions are important in determining superconductivity. . . . It is important to include their effect on the free energy of the electrons.” Bardeen 

156 TRUE GENIUS then spent the next several days trying to use the new clue to re- vive his earlier theory of superconductivity. In his research at Minnesota, he had failed to achieve a lowering of the energy using a periodic lattice distortion. Now he tried lattice fluctuations. That also failed, but he was sure he was on the right track. To secure priority, he dashed off a letter about the work to the Physical Review. In the history of science, major discoveries often come in pairs or higher multiplets. The countless examples of simultaneous dis- covery of important scientific theories include the conservation of energy principle (Mayer, Joule, Kelvin, Helmholtz, and others), the entropy principle (Clausius, Kelvin), the calculus (Newton, Leibnitz), the theory of evolution (Darwin, Wallace), the math- ematical formulation of quantum mechanics (Heisenberg, Schrödinger), and quantum electrodynamics (Feynman, Schwinger, Tomonaga). What causes simultaneous discovery? Historians of science, including Thomas S. Kuhn, have found this question compelling. Is the coincidence the result of direct, or indirect, communication? Does it come from the “ripeness” of a field or of a body of research? Does it depend upon internal connections in the state of knowl- edge or thought—“something in the air”? Or is it but a matter of chance? Whatever the explanation, by the early 1950s several researchers, including Bardeen, had arrived at the hypothesis that electron–phonon interactions are crucial to superconductivity. Even before Maxwell and Serin had completed their experi- ments, Herbert Fröhlich, a British theoretical physicist on leave from Liverpool and spending the 1950 spring semester at Purdue University, had written a theory of superconductivity that pre- dicted the isotope effect. When Fröhlich learned about the con- firming experimental results a day or two after they appeared in the Physical Review, he promptly sent a letter to the Proceedings of the Royal Society to claim priority. The competition between Fröhlich and Bardeen was on. Fröhlich happened to visit Bell Labs a week after Bardeen sub- mitted his letter. “It was very common for people to come to Bell Labs and pass through,” Louise Herring recalled. On that particular visit, Fröhlich and Bardeen discussed their respective work on superconductivity, undoubtedly with a degree of tension. Without reference to any competition between them, Bardeen later wrote 

The Break from Bell 157 about their conversation, “Although there were mathematical dif- ficulties in both his method and mine, we were convinced that at last we were on the road to an explanation of superconductivity.” Although Bardeen’s and Fröhlich’s theory both explained the isotope effect, neither could explain superconductivity. The prob- lem was that both focused on individual electron energies. To explain superconductivity, it would be necessary to treat the energy that arises from the interaction of many electrons. The available mathematical formalism was still too limited. The basic problem on which both Bardeen and Fröhlich became stuck was to find an interaction that made the total energy of the superconducting state lower than that of the normal state. They realized that if the electron–lattice interactions were to accomplish such a lowering of the total energy, there had to be an attractive force. Moreover, this interaction had to dominate the ordinary elec- trical repulsion between electrons. Bardeen and Fröhlich’s theories in 1950 could not show this. Bardeen continued to feel isolated at Bell Labs as he struggled with the problem of superconductivity. He later explained to Kelly: “There are very few people in the laboratories who are interested in the problem.” Longing for colleagues with whom to discuss su- perconductivity, Bardeen asked Bown whether he could have his own theoretical group. Again the response was negative. When Bardeen brought up his problems with Shockley, Bown insisted that Shockley needed flexibility in directing his group. On one bright moment in this otherwise gloomy period they learned that the patent for the point-contact transistor had finally been approved. Bardeen had been concerned that it might not be, for the discovery of the Lilienfeld patents had brought rejection to two of the four transistor patent applications (the ones concerning the Brattain-Gibney-Bardeen field effect work in November 1947). Betsy, then six, remembered her father coming home one evening and announcing happily, “We got the patent!” She admitted, “I had no idea what he was talking about, but I remember hugging him around the knees.” The relief was temporary. In October 1950 Bardeen attended a conference on “Crystal Imper- fections and Grain Boundaries” held at a resort in the Pocono Mountains of western Pennsylvania. He and Shockley were both 

158 TRUE GENIUS scheduled to present papers. Also at the meeting was Fred Seitz, Bardeen’s old friend. Surrounded by the brilliant leaves of fall, Fred and John sat down for a heart-to-heart talk. John told Fred what the last two years had been like for him at Bell Labs. He spoke about his frustrations with Shockley and of his exclusion from the junction transistor work. He also told Fred about his new work on superconductivity and of his desire for col- leagues with whom to discuss the problem. “I am convinced I want to go back into academic work,” he concluded. “What is the job situation like?” Seitz was just the right person for Bardeen to speak with. He had known Shockley since 1932 and could easily imagine the situ- ation at Bell. He could also identify with the frustration of working in a nonreceptive environment. Just a year earlier Seitz had left his position at the Carnegie Institute of Technology and moved to the University of Illinois. Among other frustrations at the Carnegie Institute, an aide to the president had “decided to change the physics curriculum drastically without either our input or approval.” Seitz told Bardeen that he didn’t know of any academic jobs off-hand and would contact him if he heard of any. (He did not want to disappoint his friend.) But Fred, on his return to Illinois, went directly to William Everitt, the dean of engineering, and to F. Wheeler Loomis, the head of the Department of Physics. Everitt and Loomis were both enthusiastic about offering Bardeen an appointment. Unfortunately, there was no money available in the regular budgets for hiring new professors. Student enrollment was down because of low birth rates during the Depression, and the GI Bill had run its course. Everitt and Loomis agreed, however, that the opportunity to acquire Bardeen should not be missed. Everitt assured Seitz that he would find a way to piece together a package for Bardeen. When Seitz relayed the message from Everitt back to Bardeen, he told John that he should “keep cool, that the dean was working on it.” Bardeen’s response was, “Well, Illinois would be perfect. It’s the kind of place I’d like to be at.” With the help of Coleman Griffith, the provost of the univer- sity, Everitt pieced together parts of budgets and arranged a joint position for Bardeen, half in physics, half in electrical engineering. The offer included support for an applied group in engineering. Seitz 

The Break from Bell 159 felt the dean’s offer of $10,000 a year “was not quite enough for a person of John’s stature,” which had grown with the invention of the transistor. Fred told John, “Look, the dean has 10,000. If you hang on maybe you’ll get more.” John replied, “That’s enough.” The negotiations continued over the next five months. One drawback was that “moving at this stage of my life is a difficult and costly business. . . . I will have to start building up a pension fund over again.” Seitz wrote to Bardeen about his own recent experi- ence at Illinois. “I have found more than enough outlets for all of my interests with the greatest possible freedom. I think there is no doubt that you would have exactly the same experience.” He as- sured Bardeen that he would have colleagues to work with. “I need hardly add that the solids group would be more than thrilled to see you here.” Bardeen leaned toward accepting when both physics and electrical engineering assured him that his work would be self- directed. Late in March 1951, Bell Labs complicated Bardeen’s decision by finally taking some steps to improve his situation there. Fisk, the new Bell Labs director of physical research and thus Shockley’s new boss, recognized that Bell Labs might actually lose Bardeen. “One Friday,” recalled Brattain, he and Bardeen had “walked into Fisk’s office and told him that we did not wish to report to Shockley any longer.” And on “Monday morning we weren’t reporting to him.” A memorandum from Fisk, dated Wednesday, March 28, shows the Bell Labs solid-state group divided into two separate groups: the Physics of Solids, led by Stanley Morgan, and Transis- tor Physics, led by Shockley. Bardeen and Brattain were in the Mor- gan group. “I haven’t reached a final decision on whether to leave the Labo- ratories or not,” Bardeen wrote on April 6 to Gerald Almy, then the acting head of the Illinois physics department. “There was a reorganization here about a week ago which makes things much more favorable from my point of view, but I am still inclined toward Illinois.” When Seitz ran into Fisk at Los Alamos, he warned him that Bell Labs was in danger of losing Bardeen. “Oh, don’t you bother,” said Fisk. “We’ve got that under control.” The decisive factor was Dean Everitt’s letter to Bardeen on April 16 assuring him that he could lay out his own research program and “that in no case would you be asked to work on con- tracts which were not of your own choosing.” Everitt specified that 

160 TRUE GENIUS “the programs which we now have would not indicate in any way the type of work we would expect you to do here.” As for teaching, “You would be pretty much your own boss,” he wrote, “and would give such formal courses as you felt fitted into your program.” The regular course load for physics faculty was two courses per semester. However, “If you felt that only one course a semester was the best way for you to work, that would be quite all right.” Ten days later Seitz wrote to Bardeen again with assurance that “everyone I have spoken to is more than enthusiastic and will do everything within his power to make you enjoy life here.” Seitz described the collegial atmosphere and the opportunities for increasing one’s academic income—for instance, working with the military controls project for part or all of the summer. Seitz enlisted his wife Betty to write to Jane about housing opportunities. Betty described their recent house-shopping experience. They had found a comfortable, inexpensive, two-story stucco in a tree-lined neigh- borhood on Iowa Street, a short walk from the campus. Bardeen accepted the Illinois offer in a telegram sent on April 28. His initial salary of $10,000 would be evenly split between the physics and electrical engineering departments. On hearing from Everitt of Bardeen’s acceptance, Seitz immediately sent off a warm letter of welcome. Bardeen wanted Mervin Kelly to know why he was leaving Bell Labs, for when Kelly hired Bardeen, he had promised he would have full research freedom. Shockley had made a mockery of that promise. Now that Kelly was president of Bell Labs, Bardeen felt he deserved to know the full story. Bardeen’s carefully worded three-page letter, posted on May 24, began with a simple explanation of the Illinois offer. He empha- sized the university’s promise to be supportive of his research, no matter what direction it took. That was important, he explained, because he expected that his research would be on superconductivity “at least for the next year or two.” He also noted that the “financial prospects appear to be as good as those at the Laboratories.” Then he got to his main point: “I would not leave if I were not dissatis- fied with conditions here. In fact I would not have received the offer if I had not let it be known that I was considering leaving the Laboratories.” “My difficulties stem from the invention of the transistor.” He explained in detail how the work under Shockley had changed after 

The Break from Bell 161 his and Brattain’s invention. “Before that there was an excellent research atmosphere here.” Afterwards, Bardeen wrote, Shockley reserved all the interesting research for himself: Shockley at first refused to allow anyone else in the group (1170) to work on the problem (that is, aside from Brattain, Gibney and myself), and then did so only as he thought of problems of his own that he wanted investigated experimentally. In most cases these were problems in which he either already had done theoretical work or in which he wished to do the theoretical work himself in the future. In short, he used the group largely to exploit his own ideas. Each time Bardeen tried to go over Shockley’s head and gain entry into the research on the junction transistor, he had been re- buffed. Shockley “was well aware of the situation,” wrote Bardeen; it “was a deliberate policy.” Bardeen explained that he was in an “intolerable” position in which he “could not contribute to the experimental program unless he wanted to work in direct competi- tion with my supervisor.” Alluding to his Oak Ridge offer in the spring of 1949 Bardeen told Kelly, “I seriously considered leaving the Laboratories about two years ago under much less favorable circumstances.” Bardeen also told how his “intolerable” working conditions grew even worse after he decided to work on superconductivity, a new direc- tion he had “discussed at the time with Fisk.” Given the lack of support for his work on superconductivity, Bardeen explained that he “felt somewhat isolated.” But even if there were experimental work on superconductivity going on at Bell, “I feel that I can work on superconductivity more effectively in a university. The problem is one of more scientific than practical importance and there is great interest in it in academic circles.” Bardeen made it clear that he intended to continue research in semiconductor physics. Even in this area he believed he would be in a more “favorable” position at Illinois, given Shockley’s monopo- listic attitude about the transistor. Before making the decision to leave, I again explored the possibility of working on the semiconductor program with Shockley. His atti- tude had not changed. He felt that he could supply all the ideas required, that he would want the people in his group to work on his ideas, and that I would not be happy in this situation. It has been suggested at various times that an independent semiconductor group be set up under my direction, but I did not feel that this was a satis- 

162 TRUE GENIUS factory solution from my point of view. In the discussion with Shockley he indicated that he was unwilling to give up any signifi- cant part of the work. In closing, Bardeen offered an orderly recapitulation of his main points: To summarize, the invention of the transistor has led to the semi- conductor program being organized and directed in such a way that I could not take an effective part in it. I could work on superconduc- tivity, but I feel I could do this better in a university where it is of primary rather than secondary interest. I also feel that university work, in which one can set one’s own pace, becomes relatively more desirable as one gets older. Therefore I have decided to leave. Fisk and Kelly both tried to change Bardeen’s mind. They offered him a salary increase and his own theoretical group. But Bardeen was no longer interested. “And when Bardeen makes up his mind,” said Brattain, “there is no use doing anything about it. It is too late.” Bardeen’s resignation may have contributed to the rise of the semi- conductor industry in California. After his departure, Bell Labs executives recognized the hazards of entrusting an important research team to Shockley. By 1955 Shockley perceived that he had reached a “glass ceiling” on the Bell Labs ladder. He was also annoyed that his transistor work was not making him rich. He determined to strike out on his own. That year he moved to California. On the trip west, he stopped off in Urbana to see Seitz. With him was his new partner, Emmy Lanning, a psychiatric nurse Shockley had met a few years earlier in Washington, D.C. She had helped him work out a theory of productivity and creativity based on wartime theories of operations research and current theories of psychology. Bill and Emmy developed a close emotional and intel- lectual relationship in a time when Shockley’s marriage to Jean (then suffering from uterine cancer) was deteriorating. Bill invited Emmy to take part in his plans to create a pioneer- ing research company in California devoted to semiconductors and transistors. It would be staffed by scientists and engineers having a high “mental temperature.” She agreed. They eventually married. Seitz remembered the day “Bill called up and asked if they could stop here for a few days. And I said yes.” Shockley had told 

The Break from Bell 163 Seitz that he now “was on his way West.” He was returning home to Palo Alto. He had “already prepared the ground for receiving proposals.” More than twenty years had passed since Bill and Fred had traveled across the country together in Shockley’s 1929 DeSoto convertible. In 1932, as students, the two had been heading east, from Palo Alto to their East Coast schools. Seitz recalled the ad- venture of their earlier trip—and the uneasiness with which he had viewed the loaded pistol Shockley had stowed in the glove com- partment. On this visit in 1955, Seitz again felt considerable ambivalence in his interplay with Shockley, but it was of a different sort. Betty Seitz “distrusted” Shockley. She had agreed to the visit, but she wasn’t pleased about it. “She did it for me,” said Fred. “We couldn’t do other than treat them as guests.” The visit also caused some tension between the Seitzes and the Bardeens, but “not in any serious way,” said Seitz. Shockley had started out on the trip by flying to Columbus to pick up Emmy. They then drove in her car to Urbana and settled into the Seitz’s stucco on Iowa Street. Betty was dismayed when Bill tied up the phone for several days. “He just got on the phone and began talking to people up and down the coast,” recalled Fred. The Seitz’s teenaged son enjoyed the visit, for Bill “would amuse Jack with his sleight of hand tricks.” Seitz recalled that during one of Shockley’s countless telephone calls, he completed a deal with the wealthy chemist Arnold Beckman, who agreed to fund the new company that became Shockley Semiconductor Laboratory. It was the first semiconductor company to be established in the region later known as Silicon Valley. In California Shockley proved to be an even worse research director than he had been at Bell Labs. He assembled a brilliant research group, and insisted that all job applicants take creativity tests administered by psychologists in New York or San Francisco. Once they had been hired, he treated them like assistants. Most left after a short time. Two of the first eight who departed (Shockley called them “the traitorous eight”), Robert Noyce and Gordon Moore, went on to create Fairchild Semiconductor and later Intel Corporation. After Shockley’s company failed, he joined the fac- ulty at Stanford University. In the late 1960s, he focused his work on eugenics and achieved notoriety with his theories claiming a genetic correlation between race and intelligence. 

164 TRUE GENIUS After Bardeen’s decision to leave Bell Labs, the family got on with the business of transforming their lives. A possibly apocryphal, but nonetheless revealing story features John showing the house on Primrose Place to a prospective buyer. When the buyer made a good offer, Bardeen fell into one of his thoughtful trances. A period of time passed during which Bardeen said nothing. The buyer thought he had offered too little and raised the offer. Still Bardeen remained silent. The offer was raised again. Eventually the house sold for much more than the Bardeens had expected. The family bought a new car, a green DeSoto, for their move to the Midwest. Driving first to Madison, they spent some time renewing friendships and reinforcing family ties. At his alma mater Bardeen took part in a research seminar on low-temperature physics. He also discussed his evolving work on superconductivity with colleagues. John particularly enjoyed seeing Wigner, who visited Madison that summer. At the University of Wisconsin, Bardeen taught his first formal course on the electrical properties of solids. About 20 students enrolled in the summer course and several more audited. Shockley’s new book, Electrons and Holes, served as the text. Just inside the front cover of Bardeen’s copy Shockley had inscribed: “To John Bardeen, Who made a book like this a need. Bill Shockley, Dec. 1950.”