Instrumentation also needed to be considered. To monitor the reactions, Fermi’s favorite iridium foils could be placed deep inside the pile, removed, and tested for radioactivity. The problem with this process was that it was cumbersome and unsuitable for monitoring the reactions while they were taking place. Volney Wilson, a long-time Compton collaborator, would be in charge of the team responsible for developing this new instrumentation. Working with Wilson were Herb Anderson and Leona Libby. The new instruments would click loudly when either a neutron or a gamma ray was detected and would drive an electronic pen to graph the level of neutron activity on paper tacked to a circular drum, much the way an earthquake detector traces seismic motion.
Another matter deeply concerned Fermi: safety. His studies in New York told him that cadmium was a highly efficient neutron absorber. To control the fission reactions and make sure the pile did not run out of control, perhaps leading to an explosion (the term meltdown had not yet been invented), he decided to insert cadmium-covered wooden strips at strategic points throughout the pile. With all the cadmium strips in place, the pile could not go “critical,” their term for a sustained chain reaction. There would also be a fail-safe mechanism: if during the course of operation the pile became too reactive and at risk of blowing up, a rope could be cut that would allow all the cadmium strips to drop back into the pile simultaneously, bringing the reaction to an abrupt halt. This mechanism would come to be called SCRAM. Though debate continues as to what the acronym came from, its meaning is painfully obvious.
Once a critical mass of material had been assembled—once the reactor had a reproduction rate that grew exponentially—the way to turn the reactor on would be to remove all the cadmium strips but one and then slowly pull out the final strip, the “control rod,” just enough to reach criticality. If the reaction looked like it would grow out of control, all that was needed was to slip the control rod back in place. The reaction would die down almost immediately, as the cadmium absorbed neutrons from the chain reaction. As an added precaution there would be a small group of intrepid souls standing on top of the pile with buckets of a cadmium solution, prepared to douse the whole pile if for some reason SCRAM did not work. That would surely stop the reaction instantly but would also render the entire apparatus useless for further research.
By the end of summer 1942, the general plan was clear. The pile would be built out of town, in an area west of Chicago called Argonne Forest where Compton and his wife enjoyed horseback riding on weekends. Its distance from central Chicago made Argonne ideal. If an accident occurred, it would be far away from the densely populated urban area. Its distance also made it easy to isolate and maintain the kind of secrecy that was required. Work began on a facility for the pile using Compton’s engineers of choice, Stone & Webster, a Massachusetts firm that worked under the auspices of the Army Corps of Engineers. In mid-September a colonel in the corps, Leslie Groves, was promoted to brigadier general and assigned to oversee the entire Manhattan Project, effectively putting Compton, Fermi, Lawrence, Oppenheimer, and the hundreds of other physicists who were drawn into the project over the previous year under military authority. The plan was to finish the facility by October 20, 1942, at which point the entire Met Lab would move there and complete the pile.
Beginning in September 1942, Fermi gave a series of lectures to the Met Lab scientists regarding the theory behind the pile, focusing on calculations of the reproduction factor. Notes of these lectures were taken by Anderson, Libby, and others, and they include some of Fermi’s more endearing uses of American slang. In describing what to do if the reproduction factor ended up much greater than one—in other words, if the chain reaction got out of control—he said, “run quick-like behind a hill many miles away.” The lectures were an important part of the program, designed to ensure that those working so hard on the pile maintained confidence in the science underlying it. They also allowed Fermi to indulge yet again in one of his favorite pastimes, teaching.
GROVES, A BULL OF A MAN WITH A GRUFF MANNER AND AN ABILITY to get things done, had just finished successfully overseeing the construction of the Pentagon, the world’s largest office building. He wanted his next assignment to be overseas and only agreed to the Manhattan Project assignment reluctantly in exchange for two assurances: first, that he would be promoted to general officer rank, and second, that he would have first priority for men and materiel, without restriction. He got both.
Groves visited Chicago in early October 1942 and met with the senior scientists on the project. He was impressed by the “crackpots,” as he liked to call them, who were making progress toward the first chain reaction, and the meeting was productive. Owing to the new authority that Groves had extracted from his masters, graphite and uranium of increasingly higher quality now arrived in Chicago in truly massive quantities. Everyone concerned expected to make the move to Argonne and start work on what would become the first working nuclear reactor.
With total control of the program, Groves imposed military secrecy and ordered that key personnel—Fermi included—could no longer travel by air. The risks of losing essential assets in an air accident were simply too great. He also insisted that a handful of scientists—once again, including Fermi—travel at all times with bodyguards. The bodyguard assigned to Fermi was a suitably large former Chicago policeman named John Baudino. The two eventually became good friends, and Fermi joked that Baudino grew into a decent physicist in the process of sticking by Fermi’s side for the duration of the war.
Finally, in the name of military security, key scientists were not to travel under their own names. They were to use code names suitably chosen so that the scientists would have no trouble remembering them. Fermi liked his new name, Henry Farmer. It sounded very American, even though his pronunciation of his new name sounded distinctly Italian.
Fermi had a sense of humor about his code name. All the senior scientists had one—Eugene Wigner was “Gene Wagner,” Niels Bohr was “Nicholas Baker.” One evening at Los Alamos, after a screening of a forgettable 1943 movie about the life of Madame Curie, Fermi could not resist approaching Bohr: “Mr. Baker, I’ve just seen a grand movie, Madam Cooper.”
On Columbus Day 1942, Roosevelt repealed the enemy alien status of Italians in the United States, even though the country remained at war with Italy. Ironically, just as Fermi became free to travel as he wished, he was restricted by Groves to traveling only by train or car and with Baudino following him everywhere. The bodyguard was supposed to chauffeur his ward whenever the need arose, but on this point Fermi stood his ground. No one would be driving him. Baudino might accompany him, but the bodyguard would be in the passenger seat.
THE PREPARATION FOR FERMI’S PILE EXPERIMENT IN CHICAGO WAS not the only, nor even the main, focus of activity for the Manhattan Project during the summer and fall of 1942. For many participants the experiment Fermi was preparing in Chicago was a foregone conclusion. Planning moved ahead under the presumption that the pile would work according to expectations. Such was the priority of getting a workable bomb in the shortest conceivable amount of time that a number of tracks that depended crucially on each other’s success were moving forward simultaneously. One track was theoretical work on fast-neutron fission, conducted by Oppenheimer and a team based in Berkeley.
Related to this was crucial work on ever more accurate initial calculations of what the critical mass of the uranium bomb would be. More work was also being done on plutonium as a possible fission material. Studies at Berkeley and Chicago indicated plutonium could be used to fuel a fission chain reaction, but the more the new element was studied, the more concern there was about its stability in the quantity necessary for a working bomb.
Groves also directed work to begin on a variety of schemes to separate U-235 from U-238. He chose a location fifteen miles west of Knoxville, Tennessee, a site later known as Oak Ridge. Vast isotope separation plants rose on this site as the Manhattan Project progressed. Oak Ridge was also the site of the first small
plutonium production reactor, built once the Chicago experiment proved the concept. Oak Ridge was primarily a research reactor to produce just enough plutonium to begin a more serious set of experiments to determine the new element’s physical properties.
As 1942 drew to a close, Groves selected a large, desolate desert area of southeastern Washington State, eventually known as Hanford, for the top secret location of giant plutonium production reactors. The Columbia River would provide cold fresh water for cooling purposes and the area was easily secured because it was so remote from any urban centers. It was by far the largest facility in the Manhattan Project, some 586 square miles in area.
The work at Berkeley and the selection of large sites in Tennessee and Washington proceeded under the assumption that Fermi’s pile would succeed. Though few physicists who knew about the project doubted that in principle it could work, Fermi and the team were aware that unforeseen difficulties might arise, including issues of safety.
SOON AFTER GROVES’S VISIT TO THE MET LAB IN EARLY OCTOBER 1942, a labor dispute arose at the new lab facilities at Argonne, and by mid-October construction work stopped. Compton had a schedule to meet. Fermi had a pile to build. The two discussed the situation and Fermi suggested finding space on the campus in which to build the pile. Compton thought about it and decided on his own authority—without consulting the university’s president, who almost certainly would have vetoed the idea on safety grounds—to authorize a change in plans. They would build the pile in a squash court under the west stands of Stagg Field, the abandoned football stadium.
In retrospect it was a remarkable decision, reflecting both Compton’s sense of urgency and the trust and confidence he had in his extraordinary Italian colleague. Fermi had persuaded Compton not only by outlining all the safety features he had envisioned but also by referring to the oddly comforting fact that some small percentage of the neutrons released in uranium fission would be emitted moments later than the initial prompt neutrons, giving Fermi additional time to put the control rods in place if the reaction looked like it might run out of control.
Compton was convinced. He understood the physics. He believed in Fermi. Now the work began in earnest.
* When she first met Fermi, Leona was unmarried and her maiden name was Woods. She would marry twice during her lifetime—first to physicist John Marshall, and then to chemist Willard Libby. Because her memoir was written during her marriage to Libby, she is referred to hereafter as Leona Libby, to avoid any confusion.
CHAPTER SEVENTEEN
“WE’RE COOKIN’!”
BY OCTOBER 1942, SHIPMENTS OF GRAPHITE BARS AND URANIUM in the form of uranium oxide powder and uranium metal eggs were arriving at a furious pace and activity was intense. Fermi placed Zinn in charge of what was effectively a high-pressure construction job without blueprints. Zinn and Anderson managed a team of young physicists and thirty-odd day laborers, drop-outs from the local high school—what young physicist Albert Wattenberg referred to as “Back-of-the-Yards” kids—to machine the graphite into proper shape and to bore holes for the insertion of the uranium slugs. The team also began the process of sintering the uranium powder, using the dilapidated press that Zinn and Marshall had used at Columbia. Working in three shifts of eight hours, the sintering team could produce some twelve hundred lumps a day, aiming for a total of twenty-two thousand in total. The team worked fast and made few mistakes. Volney Wilson’s instrumentation team also shifted into high gear.
As a first step in the construction, Fermi had Anderson approach the Goodyear Rubber Company for a heavy rubber “balloon,” shaped in a cube, large enough to surround a squash court. History does not record what the executives thought of Anderson’s request, although they were almost certainly assured that it was for the war effort. The balloon would, if necessary, play the role of the tin can that surrounded the last Columbia experiment, allowing Fermi to pump air out of the pile to enhance the chain reaction. Goodyear delivered a cubic balloon that would do the job if needed.
The balloon’s arrival on November 16, 1942, permitted final construction to begin. Because the pile would be built inside the balloon, the first task was to hang the balloon from the ceiling so the work could take place inside it. That done, the team worked in twelve-hour shifts, Zinn in charge of the day shift, Anderson managing the night shift. Teams drilled blocks of graphite to accommodate the uranium slugs and laid them layer by layer according to plans drawn up by Fermi. The wood frame rose alongside the graphite pile. After the completion of each layer Zinn and Anderson met at Eckhart Hall with Fermi, who then made a rough sketch of how the next layer should look.
The pile rose, two layers of uranium-embedded graphite interposed with a layer of pure graphite, resting on a layer of pure graphite set at the base. Fermi calculated that the internal uranium lattice would result in a fully operational exponential reactor when the pile rose to seventy-six layers, just below twenty-seven feet high.
The uneven quality of the graphite and uranium posed considerable challenges. To address these, Fermi decided to allocate the highest-quality material—the uranium metal and the purist graphite—to the center of the pile. This was to ensure that the highest reproduction factor would be deep within the pile, offering the best hope for achieving an exponential chain reaction.
The nonstop construction took its toll. Graphite dust filled the enclosed space of the squash court and the noise was incessant. When layer fifteen was completed, Fermi asked Wilson to start measuring neutron production within the pile. Every three layers, the team repeated the measurements and dutifully recorded the increase in reactivity. As the pile rose, they placed channels for the control rods and the instrumentation running deep into the pile. The horizontal control rods were managed by hand. A vertical control rod, the “zip” rod, ran right through the center of the pile, to be lifted out by a rope and tied off when the reactor was set to go critical. The zip rod was attached electronically to instrumentation that would reinsert it back into the reactor if the reactivity level rose above a certain point. The rope on which the rod was suspended could also be cut manually with an axe should the need arise to shut the pile down instantaneously.
During the last two weeks of November, Compton, who was monitoring the progress of the pile with great interest, was deep in negotiations with DuPont executives to handle the construction of all the plutonium-producing reactors planned for the project. Fermi, Szilard, and Wigner were already scoping out the design for the initial reactor in Oak Ridge. Seaborg had agreed to a series of experiments designed to separate the plutonium from the spent reactor fuel. This experience would guide larger processing plants to be built alongside the major plutonium production reactors at Hanford. DuPont executives were hesitant to commit to the project. The company had never worked in conjunction with the US military, had no knowledge of nuclear physics, and worried about the difficulties of coordinating with Groves’s Army engineers. To bring DuPont along, Compton convened a review committee, including the young, dynamic son-in-law of DuPont’s president, Crawford Greenewalt, to persuade the executives. Compton wanted Greenewalt to be present when the Chicago pile went critical. He hoped that Fermi’s performance that day would be sufficiently exciting to persuade the up-and-coming executive to commit the company to the project.
By late November 1942, Fermi had enough data to recalculate when the pile would go critical and determined the fifty-sixth layer of the pile would be the last one needed. He gave instructions to build the pile to the fifty-seventh layer as an insurance policy. So promising were the data that he decided not to use the giant cubic rubber balloon hanging from the ceiling. On the evening of December 1, 1942, layer fifty-seven was complete. With the last of some forty thousand graphite bricks set and with about nineteen thousand slugs of uranium sitting snugly in place, Anderson, on night watch, locked all the control rods into the pile and sat guard, waiting for dawn. Fermi had extracted a promise from him not to bring the pile to criticality by himself overnight. After al
most four years of work on the pile concept, after countless experiments and a beryllium powder accident that was destined to shorten his life, Anderson would not betray Fermi, tempting though it might have been to make history himself.
ON THE MORNING OF WEDNESDAY, DECEMBER 2, 1942, CHICAGO was in the grip of a cold snap. The previous day the high was thirty-two degrees Fahrenheit, but when Fermi awoke the next morning the temperature had dropped to zero. Leona Libby accompanied him to the pile, where they took some measurements of reactivity to compare with the measurements taken the night before. Anderson, who had been up late, arrived next, and the three made the short walk to Libby’s apartment, where she cooked pancakes. Then they returned to the squash court to begin the day’s historic work.
The process began about midmorning. The crowd overlooking from the balcony grew as the morning progressed and eventually included Zinn, Anderson, Szilard, Wigner, and several dozen other physicists who played a role in the pile’s construction. At 9:45 a.m., Fermi instructed three of the safety rods to be withdrawn. Immediately, the counters started clicking in response to neutron production, and Fermi watched as the production leveled off. Shortly after 10:00 a.m., having satisfied himself that his predictions to this point were correct, Fermi called out “Zip!” Zinn, who was responsible for the zip rod, now withdrew it completely and set it above the pile, hanging on its rope. Again the clicking of the counters began to race. Again the clicking leveled off.
Fermi instructed George Weil, who was manning the last control rod in the pile, to pull it “to thirteen feet,” halfway out of the pile. The rod had been marked carefully to allow its operator to know exactly how much of it remained inside the reactor. The counters rose dramatically in activity. Fermi not only knew that the pile was subcritical but also was able to point to the spot on the graph where the pen would begin to level off. Level off it did. After a few minutes of calculation, Fermi instructed Weil to withdraw the rod another foot. The counters picked up, but then leveled off again. Fermi fiddled with his slide rule, doing some quick calculations, and according to Wattenberg, “seemed pleased” that the neutron production was developing in the way Fermi predicted it would. Weil and Fermi repeated this process, six inches at a time. “Every time the intensity leveled off, it was at the values [Fermi] had anticipated for that position of the control rod,” Wattenberg later recalled. At 11:25 a.m., the intensity of the neutron production increased to the point at which an adjustment of the instrumentation scale was required, an adjustment Fermi oversaw with Wilson. As a test, Fermi asked for the safety rods to be reinserted in the pile, and the intensity dropped dramatically. He then asked Zinn to remove all the safety rods, and the reactor started up again, the counters ticking wildly for a moment before rather suddenly, at 11:35 a.m., a loud crash startled those watching. The instrumentation had recorded a level of intensity that tripped the mechanism holding a safety rod in place; the rod had come crashing down into the pile, bringing the reactivity to a complete halt. It was, however, a level of intensity that was still below criticality.
The Last Man Who Knew Everything Page 24
