Genius in the Shadows

by William Lanouette

  “You wouldn’t put boron into your graphite, or would you?” Szilard asked his guests. The two men looked at each other in embarrassed silence.

  “As a matter of fact,” one said, “samples of graphite which come from one of our factories contain boron, because it so happens that we manufacture in that factory graphite electrodes for electric arcs, into which boron is customarily put.” Had the Columbia team relied on the National Bureau of Standards for graphite supplies instead of using Szilard’s half-joking and eclectic procurement methods, it is likely that US researchers would have followed their German counterparts and rejected graphite as a moderator. Speaking of this discovery years later, physicist Hans Bethe noted that Szilard “finally persuaded one of the chief manufacturers to make some graphite without the use of boron and that graphite turned out, by Fermi’s experiments, to be suitable as a reactor medium. So Szilard contributed in a very major way to the early success of perhaps the most important branch of the Manhattan Project.”27

  Still fearing the Germans’ progress and not knowing about their failure with impure graphite moderators, Szilard warned other American researchers to keep silent about their work. To Ernest O. Lawrence, whose colleagues would be the first to create plutonium in their Berkeley cyclotron, Szilard wrote about Turner’s letter, saying in July 1940 that “if element 239 [plutonium] shows fission for thermal [slow] neutrons, it would be highly advisable to keep this a closely guarded secret” and repeating in August that “information about it should not leak out in the newspapers or otherwise. . . .”28

  When French physicist Bertrand Goldschmidt, a refugee from German occupation, came to Columbia in the summer of 1940 seeking to help the anti-Nazi cause, Szilard met him at the King’s Crown Hotel and took him to lunch at the Men’s Faculty Club. Szilard asked Goldschmidt to work with him on purifying uranium and tried through Pegram to arrange an appointment. But now that research at Columbia was funded by the government, the military was suspicious of “enemy aliens,” and despite months of trying, Szilard could not maneuver around the new federal restrictions.29

  The US military’s security checks were not confined to foreign visitors, however, and in one of the most ironic episodes of the war effort, Fermi and Szilard were nearly barred from working on atomic research. Had this happened, it is safe to say there would have been no controlled nuclear chain reaction by December 1942 (the first “neutronic reactor” was codesigned by Fermi and Szilard) and no A-bomb by July 1945. An army report gave this information about Fermi and Szilard, then being considered for defense-contract work at Columbia:

  (1) ENRICO FERMI. Department of Physics, Columbia University, New York City, is one of the most prominent scientists in the world in the field of physics. He is especially noted for breaking down the atom. He has been in the United States for about eighteen months. He is an Italian by birth and came here from Rome. He is supposed to have left Italy because of the fact that his wife is Jewish. He has been a Nobel Prize winner. His associates like him personally and greatly admire his intellectual ability. He is undoubtedly a Fascist. It is suggested that, before employing him on matters of a secret nature, a much more careful investigation be made. Employment of this person on secret work is not recommended.

  (2) MR. SZELARD. It is believed that this man’s name is SZILLARD. He is not on the staff of Columbia University, nor is he connected with the Department of Physics in any official capacity. He is a Jewish refugee from Hungary. It is understood that his family were wealthy merchants in Hungary and were able to come to the United States with most of their money. He is an inventor, and is stated to be very pro-German, and to have remarked on many occasions that he thinks the Germans will win the war. It is suggested that, before employing him on matters of a secret nature, a much more careful investigation be made. Employment of this person on secret work is not recommended.30

  Although said to be based on “highly reliable sources,” the report is wrong on several points. Fermi was certainly not a Fascist, nor was Szilard a refugee from Hungary or his family wealthy merchants. These errors were minor compared to later ones that would be made about Szilard’s maverick life, but on one point the report was prescient. Szilard did believe that Germany might win the war, a fear that drove him to work day and night to develop the chain reaction first.

  On August 22, 1940, the War Department sent this report to Lt. Col. J. Edgar Hoover, director of the Federal Bureau of Investigation (FBI), requesting an FBI check of Fermi and Szilard “to verify their loyalty to the United States.” The FBI agents repeated almost verbatim the army’s details on Fermi and Szilard and continued filing reports about Szilard to Hoover after the two physicists began work for Bush’s NDRC at Columbia in November.31 Had the army investigators’ advice been followed in the fall of 1940, neither Fermi nor Szilard would have been hired by the NDRC to work on the A-bomb. So in this case it appears that FBI surveillance, strongly influenced by a last-minute interview with Einstein, allowed the two physicists to continue building and designing their uranium-graphite piles. Ironically, when the army officers worried about “secret work” in 1940, they did not realize that the only secrets then worth protecting were not in government files but solely in the minds of “enemy aliens” Fermi and Szilard.

  Szilard’s mind centered on his secret work again in August 1940, after Germany began its bombing “blitz” of London. When Bela’s wife, Elizabeth, lamented the destruction one evening, Leo suddenly stared at her intensely, thought for a pained moment, and said quietly: “Before this war is over there will be bombs thousands of times more powerful than those in the blitz.”32

  By joining the staff of the NDRC, Szilard accepted his first full-time job in more than a decade. But he complained that the $4,000 a year salary was such a “low figure” because he was “instrumental in inducing the government to assume expenditures” for testing chain reactions, a pursuit that “might come to nothing.”33 Still restless about the slow pace of their work at Columbia, Szilard turned to the John Simon Guggenheim Memorial Foundation for a fellowship. He wanted “to find out whether or not a nuclear chain reaction can be maintained by means of thermal [slow] neutrons or by means of fast neutrons in a system that contains uranium,” but his application was rejected.34

  When the work pace around Columbia quickened in the spring of 1941, Szilard realized he needed an assistant to conduct tedious but important calculations on the neutron absorption of various elements. He heard that a young graduate student, Bernard T. Feld, was interested. At Columbia, Feld later recalled, “Szilard was to us a mystery. He occasionally appeared at colloquia, and when he intervened, one could be sure that the comment or question would be particularly incisive.” But Szilard remained for Feld “a remote and exotic figure” until one spring morning when the telephone rang in Isidor I. Rabi’s laboratory. A call for Feld.

  “This is Leo Szilard,” a voice on the phone cracked. “Can you have lunch with me today at the King’s Crown Hotel at twelve-thirty?” Feld accepted “with puzzled pleasure.” Over lunch Szilard came right to the point. He needed help with his uranium-fission studies. Szilard explained what he and Fermi were doing. For their uranium-graphite pile, Szilard said, he concluded the uranium should be compacted into lumps. Then the neutrons released in fission would be slowed down by the surrounding graphite; as slow neutrons they would be most efficient in creating more fissions and more neutrons—thus maintaining the chain reaction.

  “I listened with mounting excitement,” recalled Feld, and when he realized that Szilard had anticipated a way to accommodate Feld’s graduate studies, he accepted. After lunch, Szilard led Feld up to his room, dug out his notebooks, and spent that afternoon explaining in detail the calculations that covered the pages in his hasty scrawl. Szilard explained to Feld the theory and data behind these numbers. He wanted Feld to carry these calculations further. Their first session ended when Szilard handed Feld a fresh notebook and a twenty-inch slide rule, then the most reliable calculator available.


  “Until I can make better arrangements,” said Szilard, “you are free to work here. I will be away for the next few weeks.” And he disappeared, in Feld’s words “rushing up to Boston to cajole the carbon manufacturers into producing more, denser, and purer graphite at more reasonable prices; dashing down to Washington to extract greater support; stopping off at Princeton to consult with Einstein and the Wigner group.” And as Szilard’s “proxy hands at Columbia,” Feld also got to stack graphite bricks.35

  Szilard’s rushing around seemed frantic to many colleagues, driven as it was by his fear of German A-bombs. That fear was renewed in April 1941 when physicist Rudolph Ladenburg at Princeton wrote to Lyman Briggs, reporting that a scientist had just come from Berlin via Lisbon with news that German physicists under Heisenberg were working on a bomb. Werner Heisenberg, the eminent German physicist, was trying to delay the effort, the traveler said, but all were under orders to make a weapon. Hurry up!36

  Eager to avoid bureaucratic delays, Szilard used $400 of his own money to rent fifty pounds of thorium, hoping to test if it fissioned. Later it was discovered to be the only natural element besides uranium that sustains a chain reaction.

  In May 1941 came another stunning discovery, confirming Turner’s speculation. In Berkeley, chemist Glenn T. Seaborg demonstrated that plutonium239—the element made by bombarding uranium238 with neutrons—fissions as easily as uranium235. Now it was known that slow neutrons could change natural uranium to fuel for possible chain reactions—and possible bombs. Unaware of Seaborg’s work, the National Academy of Sciences’ committee that Bush had convened to study fission reported on May 17 that three phenomena might result if a chain reaction could be maintained. The three were in just the order Szilard had presented to Sachs more than eighteen months before: power for ships, and possibly submarines; radioactive poisons; and, perhaps, bombs. The committee recommended extending NDRC funding for another six months. This work then included chain-reaction piles and gaseous diffusion experiments to separate uranium235 at Columbia, and a gas centrifuge for isotope separation at the University of Virginia.

  “We were in the Columbia labs, worried about how to stack bricks to maximize neutron moderation,” recalled John Marshall, Jr., the young physicist Szilard hired from Rochester to work on the NDRC project, “and there was Szilard out in the halls. Pacing up and down. Worrying about reactor coolants and controls—problems we wouldn’t confront for another two years. His mind was always racing ahead.” Szilard’s speculations by this time included both laboratory experiments and a cooling apparatus for huge industrial-sized nuclear plants. He was thinking hard and thinking big.37

  The American public knew almost nothing about atomic energy at this time, and what it learned from the popular press was speculation. A July 22, 1941, issue of PIC magazine, a popular illustrated weekly, reported correctly about atomic research that “this war will be won or lost in the laboratory.” The weapon PIC predicted, however, was not an atomic bomb but radioactive “death dust,” and the use suggested for uranium235 was not to explode but to fuel aircraft. “A lump of this U-235 the size of an ordinary pack of cigarettes would supply power enough to run the greatest bomber in the world for three continuous years of unceasing flight,” PIC reported. “It would mean that there’d be no point on this planet which a bomber could not reach starting from any other point on Earth.”38

  As farfetched as the death dust and three-year bomber seem today, they conform to ideas in the second report the NAS review committee handed to Bush in mid-July 1941. In its main report the committee was still reluctant to recommend a full-scale program. Ship propulsion, pollution by radioactive debris (PIC’s “death dust”), and only possibly bombs still seemed the likely benefits of nuclear fission.

  By the time Bush reconvened the NAS committee a third time, word of a British calculation had been spread among its members by Mark Oliphant, a Cambridge researcher who toured American universities in the summer of 1941. A “fission bomb” might be made with between 2 and 100 kilograms (4.5 and 220 pounds) of uranium235, the third report to Bush noted. Working with Fermi in 1941, Szilard calculated roughly the amount of uranium needed to make a bomb. But it seemed enormous because of the arduous process needed to extract uranium235 from natural uranium ore. Elsewhere on Columbia’s campus, chemist Harold Urey worked with NDRC support to devise uranium-separation methods, and he soon realized that several pounds could be made if laboratory methods were expanded to industrial scale. But NDRC’s contract specified that Urey was not to discuss his work with Fermi and Szilard, two “enemy aliens.” As a result, Szilard later realized, “we were not able to put two and two together and come out with a simple statement that bombs could be made out of reasonable quantities of uranium-235.”39

  Oblivious to the strides made for separating uranium, Szilard continued to dash off memos on perfecting Columbia’s chain-reacting piles. With help from Feld and Marshall, Szilard wrote on possible ways to extract plutonium from natural ores, on ways to create “explosive chainreacting bodies,” on fast-neutron behavior in atomic piles, on how fission might be sustained by the neutrons that escape,40 and on how neutrons are captured during fission. Teller, who knew that in Berlin Szilard had the nickname Generaldirektor for his imperious ways, joked at Columbia that he had raised his rank in published reports by the choice of his collaborators—to”Feldmarschall Szilard.” But Wigner, who also enjoyed citing the Generaldirektor title, said after the war, “If the uranium project could have been run on ideas alone, no one but Leo Szilard would have been needed.”41

  Bush’s NAS committee issued a third report in the fall of 1941 that concluded atomic bombs were “possible,” perhaps within two to three years, at a cost of maybe $133 million. On November 27, 1941, Bush described in a memo to President Roosevelt an ambitious uranium program. He recommended turning over the work at the universities to the military.42 This memo required no decision by Roosevelt, and Bush began to implement it on December 6, at a meeting of the NDRC. Gathered at the Cosmos Club, on Lafayette Square across from the White House, Bush’s committee decided on an “all out” effort for military research on chain reactions.43 The next morning, Japanese warplanes attacked the US naval base at Pearl Harbor. Roosevelt went to Capitol Hill to denounce the day that would “live in infamy.” Congress declared war on Japan the same afternoon and on Germany and Italy three days later. More than two years after it had begun, the United States was in World War II. Consumed by his duties as commander in chief, Roosevelt did not respond formally to Bush’s November memorandum until January 1942, when, in his own hand, he wrote on a tiny memo pad: “V.B. OK—returned—I think you had best keep this in your safe FDR.”44

  Bush and his colleagues had wasted no time awaiting formal approval for their plan. In December 1941, he designated a planning board to oversee the A-bomb project. University of Chicago physicist Arthur Holly Compton became leader of a unified research and development effort. Increasingly the science leaders knew what they had to do. But with wartime secrecy in mind, they still had to decide where to do it.

  At Columbia, Szilard continued to write uranium studies: with Zinn on December 12 on neutron behavior in uranium and other heavy elements; with Feld on the day after Christmas on ways to shape uranium to maximize fission.45 Szilard’s thoughts about the bomb project were irrepressible. In his cluttered room at the King’s Crown he organized a survey of potential sites for the first chain-reacting pile, he devised new experiments, and he entertained schemes to organize not just the research and development but the entire project.46

  CHAPTER 16

  Chain Reaction Versus the Chain of Command

  1942–1944

  Had Leo Szilard had his way, the world’s first controlled, self-sustaining nuclear chain reaction would not have occurred in a squash court under a football stadium in Chicago. Instead, the historic site would have been an easy commute from Columbia University in Manhattan, where he and Enrico Fermi worked. With Fermi’s assi
stant, Herbert Anderson, Szilard had picked seven sites by January 1942, including a polo field and a blimp hangar in New Jersey and a golf course in Yonkers. Szilard didn’t have his way on this decision, or on many others, for his research to beat Germany to the A-bomb fell under federal patronage, then into the US Army’s control. Yet despite his waning authority, Szilard continued brainstorming to find that “narrow margin of hope” he believed would aid not only the progress of the Manhattan Project but also the fate of the world once nuclear weapons became a reality.

  Szilard lost the siting decision to intercampus politics, as physicist Arthur Holly Compton, named in January 1942 to direct US nuclear research, preferred his own domain at the University of Chicago. Szilard and Compton soon became friends, then allies against a mightier foe, Gen. Leslie R. Groves. Szilard’s feisty rationality clashed repeatedly with the headstrong efficiency of this West Point graduate, who had just directed construction of the Pentagon (then the world’s largest office building), ahead of schedule and under budget. Groves appeared as corpulent as Szilard, but with that their similarity ended. Before the war was over, Groves would try to have Szilard jailed as an “enemy alien,” order agents to follow him and to open his mail, and force him off the Manhattan Project payroll during a yearlong dispute over the chain-reaction patent.

  Szilard officially joined the Advisory Council of the S-l Physics Project, the code name for the Uranium Project, on New Year’s Day, 1942, at the same time becoming a member of the scientific staff, Division of National Research, at Columbia University. He was now part of a bureaucracy that expanded daily, under the watchful and energetic direction of the Carnegie Institution’s Vannevar Bush and James B. Conant, a chemist who at the time was also president of Harvard. As the S-l Physics Project’s new director, Compton came to Columbia in January 1942, met with Fermi, Szilard, and others, and after a tour of their graphite-uranium piles, sat down in the physics laboratories and spelled out his plans. “By July 1, 1942, to determine whether a chain reaction was possible. By January, 1943, to achieve the first chain reaction. By January, 1944, to extract the first element 94 [plutonium] from uranium. By January, 1945, to have a bomb.”1