IN VIA PANISPERNA, CORBINO’S BOYS DISCOVERED THAT LIGHT nuclei could slow down neutrons. The paraffin block experiment was a perfect demonstration of this phenomenon. At the time the Rome team did not know they were splitting atoms and thus had no interest in the neutrons that might be emitted from fission reactions. Now, as he worked to create a controlled chain reaction, the behavior of neutrons themselves mattered greatly to Fermi.
If hydrogen was not suitable for sustaining the chain reaction, what light nucleus would work? The periodic table of elements is organized from lightest to heaviest—hydrogen, with an atomic number of 1, is first and, in 1939, uranium, the heaviest naturally occurring element, ended the table with an atomic number of 92. It was quite natural to examine elements sequentially along the periodic table from hydrogen onward for the next best neutron moderator, especially for someone as methodical as Enrico Fermi.
Hydrogen has two heavy isotopes, deuterium and tritium, both of which would be less likely to capture neutrons, but these isotopes are rare in nature and difficult to manufacture. Moving up the periodic table, helium is naturally found as a gas, and liquid helium is so cold that it needs special handling. Lithium, beryllium, and boron are the next in line, but the first two are relatively dangerous to work with and boron was not readily available—the major source of all boron is Turkey. Fermi later discovered that boron absorbs neutrons and would not make a suitable moderator. Next after boron is carbon.
Carbon is safe, plentiful, and comes in a variety of solid forms. Coal, of course, is one such source, but it is too soft to be machined with precision. Diamond is another form, but it is far rarer, and as the hardest substance on the planet it is virtually impossible to machine. Graphite, a crystalline form of carbon, is not quite as plentiful as coal but still easily obtained in nature in large quantities and quite easy to machine with precision. Every common pencil contains a piece of graphite that has been machined down to a thin rod. Making graphite bricks is relatively easy.
Fermi and Szilard came to the idea of substituting graphite for water and spent much of the summer corresponding about the possibility of using graphite as a moderator. Carbon atoms are about twelve times heavier than hydrogen atoms, but both scientists believed that carbon might absorb a sufficient amount of a neutron’s kinetic energy to slow it down for the purpose of a fission chain reaction. Perhaps it would not grab neutrons out of the chain reaction, the way hydrogen did. The scheme that Fermi had in mind would require a lot of graphite, and Szilard was a man who knew how to get it.
WHILE HE WAS CORRESPONDING WITH FERMI, SZILARD WAS ALSO scheming with his old friends Edward Teller and Eugene Wigner to kick-start the US government’s interest in uranium chain reaction research. The story of how Szilard and his fellow Hungarians persuaded the most famous scientist in the world, Albert Einstein, to sign a letter to President Roosevelt on August 2, 1939, urging the president to initiate large-scale research into the possibility of a nuclear weapon is well known. The image of a carload of Hungarian geniuses, chauffeured by a New York investment banker friend of Szilard, hunting for Einstein’s house in the wilds of Long Island’s North Fork is one of the more vivid of this entire period. The letter galvanized American research in fission weapons. Fermi was neither directly involved in writing the letter nor in getting it signed by the great man, but he was mentioned in the famous first paragraph, drafted by Szilard himself:
Some recent work by E. Fermi and L. Szilard, which has been communicated to me in manuscript, leads me to expect that the element uranium may be turned into a new and important source of energy in the immediate future. Certain aspects of the situation which has arisen seem to call for watchfulness and if necessary, quick action on the part of the Administration. I believe therefore that it is my duty to bring to your attention the following facts and recommendations.
Events that summer in Europe only served to heighten the sense of urgency. In diplomatic, military, and intelligence circles, rumors swirled that Germany was poised to invade Poland in a lightning attack. The negotiations between Germany and the Soviet Union that led to the Molotov-Ribbentrop Pact remained secret for much of August—they were underway when Heisenberg dismissed out of hand Laporte’s speculative question at the Ann Arbor party about a Nazi-Soviet alliance—but the two governments sprang their surprise alliance on an incredulous world on August 23, 1939. Hostilities began a week later. The United States was not yet involved, and many influential politicians and public figures remained opposed to US involvement, but Roosevelt was already engaged in quiet efforts to bring American public opinion and industrial might in line behind eventual engagement in a European war on the side of the Allies.
Roosevelt finally received the letter in October 1939 and authorized work on fission as an immediate priority. The national effort launched by Roosevelt eventually evolved into the largest, most complex military-scientific program ever conceived. At that moment, however, the Manhattan Project was limited to Fermi’s work at Columbia’s Pupin Labs.
BEGINNING IN THE FALL OF 1939, FERMI WROTE FORTY-SEVEN papers describing the experimental work leading to the creation of the world’s first controlled self-sustaining chain reaction on December 2, 1942. The work was methodical, demanding, and sometimes dangerous and involved a growing group of physicists. Fermi continued relying on Szilard’s skill in obtaining increasingly pure batches of graphite. Szilard also served as a sounding board and a sometimes irritating cheerleader for the project. Pegram, as head of the physics department at Columbia and a dean of the college, threw his considerable weight and sound judgment behind the project. Anderson and Zinn lent their experimental knowhow and attention to detail, and Fermi added a bright young Columbia graduate student, George Weil, to his team. Other students were to follow, including Albert Wattenberg and Bernard Feld.
Given the importance of the chain reaction experiments, one might assume that they completely preoccupied Fermi, but he continued to pursue other research interests along the way. For example, he lectured on the geophysics of iron in the core of the earth at the 1940 Washington Conference, the one subsequent to the conference at which Bohr and Fermi sprang fission on an unsuspecting world. In the spring of 1940, Fermi went to Berkeley to give the annual, highly prestigious Hitchcock Lecture on “High Energies and Small Distances in Modern Physics.” His archival notebook demonstrates the effort he made to prepare for these lectures. His Ann Arbor work on cosmic rays passing through gases and solids continued into the fall at Columbia, resulting in laborious and frustrating calculations on the relationship between the density of a medium and the speed with which an ionized particle slows down. Apropos of this, Fermi quipped to Anderson that “he could calculate almost anything to an accuracy of ten percent in less than a day, but to improve the accuracy by a factor of three might take him six months.” Colleagues sometimes noted his tendency to get frustrated if he could not immediately solve a problem. This was a good example.
He also continued with a full teaching load, including courses on geophysics and quantum theory.
Like all Americans, he followed the war in Europe. On the first of every month, he and a group of faculty colleagues—a “Society of Prophets” Laura called them—would meet at the faculty club to predict developments over the coming month, writing down answers to ten yes-or-no questions. Laura reports that by the time the society dissolved, when the Fermis departed for Chicago in the summer of 1942, Enrico had established himself as the Prophet, having predicted successfully 97 percent of the time. He did this, she writes, using the most conservative algorithm imaginable: the next month would look almost exactly like the previous month. He did, however, miss one prediction—the surprise German invasion of the Soviet Union. The game was ideal for someone of Fermi’s temperament, invariably conservative and skeptical of any predictions of quick or revolutionary change.
In the lab, the bulk of his efforts focused on fission and chain reactions and, within a few months of returning from Ann Arbor, he and his team were making progress.
THE CONCEPT OF A PILE DIFFERS LITTLE CONCEPTUALLY FROM THE configuration of the water tank experiments that Fermi, Szilard, and Anderson carried out during the winter and spring of 1939. Effectively, Fermi decided to substitute graphite bricks for the water and to build up, as well as out.
The water tank experiments were not designed to analyze the way neutrons diffused within the moderating medium. Realizing how crucial it would be to understand this diffusion process with graphite, Fermi devised a series of experiments to do just that. In rooms at Pupin and later in the basement of nearby Schermerhorn Hall, he and his colleagues—with the occasional help of burly members of the Columbia football team who were press-ganged by Pegram—stacked graphite bricks into square columns several feet thick, placed rhodium foil at key locations throughout the stacks, set a neutron source at the top of the stacks, and studied how neutrons made their way through the pile. Once the foils were exposed, they were quickly extricated from the stacks and run down a corridor so the radioactivity could be measured by Geiger counters. Fermi and Anderson raced down the corridors to take advantage of rhodium’s short, forty-four-second half-life, re-creating scenes from Via Panisperna in 1934, most likely with Fermi in the lead. They built one stack after another, getting covered in fine black graphite dust that made them look more like coal miners than experimental physicists. As the stacks grew in height, to over ten feet, the physicists required ladders to get to the top of the stacks and place the neutron source. Many years later, when Fermi recalled these experiments at a public lecture, he drew a laugh from the audience as he described it as “the first time when I started climbing on top of my equipment because it was just too tall—I’m not a tall man.” These diffusion experiments were critical in establishing how neutrons slowed down during their voyage through the graphite and gave some sense of how often neutrons might be absorbed by the graphite.
These experiments began in the spring of 1940 and continued throughout much of the rest of the year. Anderson, chasing after Fermi running down the corridor with rhodium foil, played Amaldi’s role from six years earlier at Via Panisperna. Szilard kept up with the increasing demand for ever larger quantities of graphite, playing the same procurement role as Segrè did in Rome. Szilard had help from the “Committee on Uranium,” a group of senior scientists and military officers established by Roosevelt to provide guidance and coordination for the new effort authorized by the president. It was an “all-American” group. As foreign nationals, Fermi, Szilard, Teller, and Wigner were formally excluded but met frequently with the committee to provide input into their decisions. One of their first decisions was to allocate $6,000 to buy what Fermi described as “a huge amount” of graphite.
As the diffusion experiments continued through 1940, Fermi and Szilard began to suspect that impurities in the graphite were absorbing neutrons at a rate that would reduce the probability of a successful chain reaction. A visit to one of the manufacturers, National Carbon in Cleveland, resulted in the identification of trace quantities of boron in the graphite as the culprit. In 1941, Szilard worked closely with the engineers at National Carbon to develop a processed form of graphite with fewer impurities.
A second experimental project took place alongside the diffusion studies, the purpose of which was to test out Fermi’s idea of placing the uranium in lumps throughout the graphite pile, to reduce the likelihood that neutrons would be absorbed by U-238.
These experiments pushed materials science to new frontiers. Industrial graphite had too many impurities to make it useful as a moderator for the chain reaction. Szilard and the engineers at National Carbon worked hard to develop methods to remove boron and produce a suitable graphite moderator. In doing so, they created the world’s first “nuclear graphite,” a form of pure graphite now used throughout the world in graphite-moderated nuclear reactors.
Another aspect of materials science that got a boost through these experiments involved the production of uranium. Before the war, uranium was valuable only insofar as it was used in scientific experiments, and the most readily available form of the element was uranium oxide, not ideal for the purposes of the “exponential” pile.* There being little economic need for uranium metal, production was still relatively primitive. The best that industry could do was to produce the metal in a powder form, which had a tendency to spontaneously combust when exposed to air. Monitoring these experiments with great interest, the Committee on Uranium worked with a variety of manufacturers to improve uranium metal production techniques.
During this period, the team continued to grow, with Wigner, Wheeler, and others from Princeton joining Fermi’s team.
BY SEPTEMBER 1941, FERMI BELIEVED THAT THE TEAM HAD MADE sufficient progress to build a true working exponential pile. The Committee on Uranium allocated the considerable sum of $40,000 for the purchase of massive quantities of uranium and graphite, and Szilard negotiated the purchases with industrial vendors. The Columbia football players were called back into service, and the pile began taking shape in the basement of Schermerhorn Hall. Fermi later spoke with awed amusement at the ease with which the Columbia athletes packed heavy cans of uranium oxide—uranium was and remains one of the heaviest elements found in nature—and hoisted them into a lattice whose structure Fermi determined as he balanced theoretical calculations and educated guesswork with the practicalities of working with the materials at hand. Large-scale electronic computers not having been invented, it was impossible to do a full-blown calculation as to what size the lattice would have to be to produce a self-sustaining chain reaction. What was possible to measure was the performance of a pile of a specific size and specification and extrapolate whether such a structure, if extended infinitely, would produce such a chain reaction.
The actual pile they built grew to a stack of graphite bricks eight feet on each side and eleven feet high. Within the stack of graphite, square tin cans of uranium oxide, eight inches on each side, were distributed in a three-dimensional matrix of some 288 cans. Slits were placed strategically to allow for insertion and removal of iridium foils to measure radioactivity. The neutron source would be placed in a bed of paraffin at the bottom of the pile.
This first pile perfectly reflected Fermi’s experimental style. Its design was partly a product of sophisticated theoretical considerations, in particular the lumping of uranium throughout the graphite in a “lattice” framework. Yet the design also reflected basic practicalities, such as the dimensions of the graphite bricks themselves. Fermi’s design proved easy to build and lent itself to systematic measurement and evaluation through a series of carefully controlled experiments. He played an active part in its construction, piling graphite bricks and cans of uranium oxide alongside the rest of the team. Unfortunately, however, the results of the experiment were disappointing. By Fermi’s calculation, even extending the structure they built into infinity, the performance would be 13 percent below what would be necessary for a self-sustaining chain reaction.
Undaunted, Fermi and his colleagues were convinced that adjustments in the structure of the lattice and improvements in the purity of the materials could squeeze more excess neutrons out of the process and deliver the desired result. The Committee on Uranium seemed to agree, as did the various bodies that were now coordinating and directing all national work on fission. Central to this effort was Vannevar Bush, an MIT-trained engineer with a skeptical, Yankee demeanor. Bush, an extraordinarily energetic administrator, reported directly to the president.
As the project grew in organizational complexity, research into fission became highly secretive. Secrecy in fission research had begun as an informal agreement among physicists working in the United States and Britain to avoid publishing experimental results that might help German scientists, but had morphed into a formal statutory edifice of security classification. Many of the scientists most deeply involved with this sensitive work—men like Fermi, Szilard, and Wigner—were foreign nationals and found themselves excluded from deliberations within the committees organized to guide and develop fission research. Political leaders all solicited their views, but the decisions took place behind closed doors without the foreign-born scientists.
Others may have treated the need for secrecy differently, but Fermi took it quite seriously and never discussed his work with his wife. Throughout the war, from New York to Chicago to Los Alamos, Fermi was silent about his activities and his role. Laura Fermi only learned anything substantive about her husband’s role in the Manhattan Project in August 1945 after the bomb was dropped, when Fermi handed her a copy of an unclassified US government report on the project.
The work at Columbia on the pile, as well as important technical progress elsewhere, gave Bush a sense of optimism that a fission bomb was indeed possible. On December 6, 1941, he announced an “all-out” effort to pursue a fission weapon. The next day Japan attacked Pearl Harbor and within a few days the United States was at war against Japan, Germany, and Fermi’s home country of Italy. Bush reorganized the project’s leadership once again, creating a new independent organization, called S-1, to replace the Committee on Uranium. Others on the executive committee included Harold Urey, Ernest Lawrence, and Arthur Compton. Lawrence, the inventor of the cyclotron and the leader of the experimental physics group at Berkeley, would be responsible for directing research on plutonium and for developing methods of isotope separation based on his cyclotron experience. Urey, the brilliant Columbia physical chemist who befriended Fermi in early 1939, would be responsible for chemical separation issues associated with plutonium production and would also pursue promising lines of work on isotope separation. Compton, who sat on various oversight committees during this period, would direct further research into the properties of uranium and plutonium under the aegis of the newly created Metallurgical Laboratory at the University of Chicago.
Eager to prove his loyalty to his new country, Fermi was in an awkward position. He was an enemy alien working at the heart of the US government’s most sensitive and secret military project. Change was afoot.
* In a “critical” pile the reactions are self-sustaining; in an “exponential” pile neutron production grows geometrically.