The original planning for the 184-inch cyclotron had called for the magnet’s construction to be completed by November, but that deadline had slipped with the arrival of war. Now Ernest launched a crash program to convert it into a gigantic mass spectrograph containing several calutrons, each with multiple ion sources and collectors. His goal was to demonstrate that electromagnetic separation could produce enriched uranium on an industrial scale.
Ernest estimated that the conversion program would cost about $60,000, mostly for labor on round-the-clock shifts. Vannevar Bush balked at the sum, complaining that it was too much for the government to spend on equipment that would remain the property of the University of California. Lawrence turned instead to his newest and wealthiest private patron, the Rockefeller Foundation. He solicited an emergency meeting in Washington with Warren Weaver, refusing to divulge the subject in advance but offering to arrange government clearance for Weaver if he desired a full discussion. Weaver demurred. “Don’t tell me anything secret,” he said. He could guess the reason for the meeting, anyway. “What else could you want from me but money?”
Lawrence’s appeal to Weaver resembled his recruiting pitch for the MIT Rad Lab: he needed the foundation’s help with a project of the utmost importance that he could not divulge; his reputation and his record would have to do. Weaver brought the blind request to Raymond Fosdick, disguised as a grant application “for expediting the construction of the giant cyclotron, and for the purchase of certain associated equipment.” Trustees who inquired why Berkeley needed more money so soon after receiving one of the biggest grants in the foundation’s history were told “the need was urgent but could not be stated in specific terms,” Fosdick would recall. “And so the $60,000 was voted as a matter of faith.”
Only after the bomb was dropped did Lawrence reveal to Fosdick what the latter recalled as “the vital part played by the Rockefeller Foundation in the development of the atomic bomb.” The trustees’ discovery that they had unwittingly funded a weapon of mass destruction was not regarded in the boardroom as grounds for pride. “In the whole history of the Foundation, no grant had ever been made for a destructive purpose,” Fosdick observed unhappily in the foundation’s official history, “let alone such a lethal weapon as this.”
The expedited work was finished by the end of May, and a vast circular building near Strawberry Canyon originally designed to house the world’s biggest cyclotron became the headquarters of an entirely different project. Compared with its predecessor facilities on campus, the new building was unimaginably spacious, but the ceaseless activity inside could make it seem cramped. Machine shops and arrays of heavy-duty electrical equipment were situated on the ground floor along the curved inner walls, facing the two-story arch of the magnet. A second-floor deck held offices and conference rooms. Between the magnet’s fifteen-foot-wide pole faces yawned a six-foot gap, wide enough to accommodate at least two C-shaped vacuum tanks arranged back-to-back. The arms of the Cs faced out so the ion sources, electrodes, heaters, and collectors could be moved, tilted, pushed in, or drawn out by operators seated within the concave space, which was fitted with vacuum-proof windows through which they could monitor the beams.
Work on the calutrons proceeded twenty-four hours a day even while the magnet was energized, its powerful field a constant presence in the building. “We all knew not to wear a watch or carry keys,” recalled Bill Parkins, a young recruit from Cornell, but the magnet’s tug on the nails in his shoes produced the sensation of slogging through mud. The workers’ tools were forged from a nonmagnetic alloy of beryllium and copper, though occasionally an iron nail fell within the magnetic field and shot to a pole face like a stray bullet, striking its target with a metallic ping. Large ferrous appliances, such as the wheeled metal canisters for liquid nitrogen known as Dewar flasks, had to be chained in place. Perched atop the magnet housing was a sofa on which technicians working long shifts could catch catnaps, the nighttime chill of the Berkeley hillside chased off by the heat generated by the humming behemoth below them.
Ernest and his crews tinkered with the setup through the spring and into the summer while Conant and Bush pressed for faster progress. On May 23 Conant summoned to Washington all the program chiefs—Lawrence, Briggs, Compton, Murphree, and Urey—in the hope of settling on one or two separation methods and scrapping the others. A few days earlier, he had outlined for Bush the ultimatum he would deliver to his cadre of highly competitive experts: they had to decide whether “the possession of the new weapon in sufficient quantities would be a determining factor in the war.” If it was, speed would be of the essence, and funding on a heroic scale would flow. If not—if even a couple of dozen atomic bombs would be “not in reality determining but only supplemental” to a successful war effort—then there would be no rush, no need for hasty spending, and possibly no need for a wartime bomb program at all.
To a certain extent, Conant was bluffing, for the rising concern in Washington and the scientific community about a possible German bomb made an allied effort seem imperative, regardless of the prospects of success. The concern was provoked in part by news that the Nazis had commandeered a Norwegian heavy-water plant, which could only mean that the enemy was building an atomic pile moderated by heavy water—one of the options considered by Fermi before he settled on graphite to control neutrons. That suggested in turn that the Nazis had discovered plutonium, which they intended, like the Allies, to breed in a reactor. Earlier that spring, Eugene Wigner had spun a ghastly scenario for Compton: if the Nazis knew about element 94, it might take them only two months to produce 6 kilograms of it in a heavy-water pile. That would be sufficient to build six bombs by the end of 1942, beating Compton’s timeline for an Allied bomb by two years. As was learned later, the Nazi bomb program never got off the ground; but obviously this could not be known at the time.
Despite the urgency, Conant was doomed to be disappointed in his hope for a firm decision on a separation process. None of the four options for uranium enrichment—electromagnetic or thermal separation, gaseous diffusion, or the centrifuge—had clearly outdistanced the others, so none could be ruled out. Fermi had yet to achieve a chain reaction, so the availability of plutonium was still in question. Yet proceeding at full speed with all the options meant an unnerving commitment of hundreds of millions of dollars.
Still, one method did show distinct promise: Lawrence’s. With his customary bravado, he reported that the first calutron in the big magnet would start separating uranium isotopes within three weeks. By September, he promised, he would be producing 4 grams a day of U-235. He believed that a process capable of producing 100 grams a day was technically within reach. Assuming that the calutron could be made to enrich its output to 80 percent U-235 (though that was far beyond its capabilities at the moment), there could be enough material for one 30-kilogram bomb core in a year.
Conant knew Lawrence well enough to have confidence he was not drawing castles in the air. He toyed briefly with casting the government’s die in favor of the calutron. But this was tantamount to calling Lawrence’s bluff: at its current level of development, Ernest acknowledged, the calutron could produce enriched uranium fast but only in very small quantities. Whether it could shoulder all the demands of a full-scale bomb program was still in doubt. Conant, unhappy with his scientists’ inability to settle on a single separation process, penciled out a proposal to spend $870 million for four individual pilot plants.
Over that summer, the Rad Lab continued to improve the calutron design. The most important advances involved manipulating the magnetic field to sharpen the ion beams; the problem harkened back to the imperfection of the magnetic field in the earliest generations of cyclotrons, and the solution proved to be the same: arrangements of metal shims to “shape” the field, much as an optical lens focuses light entering a camera to produce a sharp image. On August 13, at an S-1 meeting on his home turf in Berkeley, Ernest delivered another bravura presentation on the Rad Lab’s progress. The remaining p
roblems in calutron design were merely engineering, he declared, a cat he had skinned dozens of times. Electromagnetic separation now had widened its lead in development over the other methods. Gaseous diffusion and the centrifuge appeared to be feasible in theory for the large-scale production of U-235, but both were hobbled by unsolved technical problems, and neither had yet produced any U-235 to speak of. The main obstacle for the diffusion process being developed by Urey was its reliance on uranium hexafluoride gas, the detestable “hex,” which lived up to its nickname by severely corroding every permeable barrier Urey tried. The centrifuge, which was being pursued by Lawrence’s old friend Jesse Beams at the University of Virginia, persistently underperformed its expected enrichment yield; it would be the first method abandoned outright. Progress on the atomic pile gave reason for optimism, but Fermi—unhappily relocated to Chicago on Compton’s orders—was still building only prototypes.
The key unsolved question was whether electromagnetic separation could produce more than a single uranium bomb core. Lawrence remained uncertain. Once again he advised Conant to keep his options open rather than “picking a horse for the long pull, for it seems to me quite likely that other methods will ultimately prove to be better.” Meanwhile, he worked out a design for a large-scale calutron plant employing an array of vacuum tanks arranged in large ovals between the poles of huge electromagnets. There would be ninety-six tanks to each oval, each holding two calutrons back to back, designed to beam uranium ions in 180-degree arcs four feet in radius from source to collector. At the Rad Lab, the arrangement was known, after its ovoid configuration, as the racetrack. Despite Lawrence’s own initial doubts, it would produce the fuel for the bomb that destroyed Hiroshima.
Chapter Thirteen
* * *
Oak Ridge
In June 1942 Vannevar Bush approved the purchase of a rural fifty-two-thousand-acre parcel about twenty-five miles from Knoxville, Tennessee. It was a long, flat valley shaded by wooded ridges on either side, sufficiently remote to serve as a secret facility for the manufacture of bomb-grade uranium but also close to the Tennessee Valley Authority electrical plants that would have to feed its prodigious appetite for power.
Preparing the site would require heavy construction, grading, and converting cow paths into four-lane highways. The job of finalizing the purchase and launching the work had been assigned to the US Army Corps of Engineers, but the Corps dithered, deciding that until the S-1 Committee settled on an enrichment technology, there was no need to acquire the site, much less get construction under way.
At the end of August 1942, Bush warned Army Chief of Staff George C. Marshall and Secretary of War Henry Stimson that the delay threatened to stifle a bomb program that had won the unanimous endorsement of “a group of men that I consider to be among the greatest scientists in the world.” His stern words got the project jump-started. General Brehon Somervell, the head of the Corps of Engineers, placed the program—then still known obliquely as the DSM Project, for “Development of Substitute Materials”—under the jurisdiction of the Corps’s Manhattan Engineer District, headquartered at 270 Broadway, and in the hands of his most trusted and efficient subordinate, Colonel Leslie R. Groves. Thus was born the Manhattan Project.
Groves, the son of an army chaplain, was a West Point graduate who tipped the scale at three hundred pounds and was not shy about throwing his weight around. The previous year, he had assumed command of construction of the Pentagon, then beleaguered by labor problems, material shortages, cost overruns, and miserable work conditions. He got the largest office building in the world completed within eight months. Groves joked later that when this complex and contentious job was done, “I was hoping to get to a war theater so I could find a little peace.”
Bush learned that he had a new teammate from the army only when Groves, who had been promoted to brigadier general upon taking over the bomb project, showed up at his office on September 17. Furious at being blindsided by the appointment, Bush dodged Groves’s brusque queries about the state of the project and dashed off a prickly letter to Stimson’s assistant, Harvey Bundy. “Having seen General Groves briefly, I doubted whether he had sufficient tact for such a job,” he wrote. “I fear we are in the soup.”
He soon changed his mind. The two most pressing necessities for the bomb program were the acquisition of the Tennessee site and the securing of a top-priority classification for the work, both of which had been in limbo for four months. Groves met both goals within forty-eight hours of taking over.
Three weeks later, Groves was in Berkeley. He arrived at the Rad Lab skeptical about electromagnetic separation, having been convinced of the superiority of gaseous diffusion by his technical advisors, who came from a petroleum industry familiar with that technology. But they had not reckoned with the horrors of hex—or with the galvanizing personality of Ernest Lawrence.
Lawrence escorted Groves around the Rad Lab, showed him the calutron in operation, and explained its virtues so convincingly that the general was carried away. He measured Lawrence against the other program heads he already had met, and judged him the most capable. Lawrence’s experience as an impresario of Big Science gave him an instinctive grasp of how to scale up a process from prototype to production. That was a critical talent for a project facing such tight deadlines that immense factories would have to be designed and built for processes that could not be tested in advance. After a few days with Lawrence, Groves recognized that his process was the only one with a demonstrated capability to produce appreciable quantities of U-235. On November 5 he asked Lawrence to draw up specifications for a sprawling electromagnetic separation plant based on the existing calutron design, which was to be frozen, or made final, so that commercial manufacturers could begin turning out the tanks for installation. The plant, code-named Y-12, would be located on 825 acres at the southeastern end of the new federal reservation in the Tennessee Valley. Five miles away, a residential community would be built to house the staff. It would be known as Oak Ridge.
• • •
One other important event occurred during Groves’s visit to Berkeley: his first encounter with J. Robert Oppenheimer.
Since being tapped by Arthur Compton in June to supervise the theoretical work on bomb design, Oppie had been working out of his Berkeley office with a small group of trusted colleagues. One of them, Robert Serber, was present when Groves, fresh from his briefing by Lawrence, strode into the room trailed by his adjutant, Colonel Kenneth D. Nichols. “Groves walked in, unbuttoned his tunic, handed it to Nichols, and said, ‘Take this and find a dry cleaner and get it cleaned,’ ” Serber recalled. “Treating a colonel like an errand boy. That was Groves’s way.”
Oppenheimer, who could tell they were witnessing a theatrical display of military authority, was unfazed. He spent the next few hours in a collegial discussion with Groves about the challenges facing the bomb designers. As Groves recalled later, he had been advised by his consultants that the device could be “designed and fabricated in a very short time by a relatively small number of competent men”—even by twenty scientists in three months or less. He was looking for validation of his instinct that this was a “dangerously optimistic” view. Oppenheimer provided it, warning that a multitude of theoretical and technical questions still remained unresolved; therefore, design work should be initiated immediately if the weapon were to be ready in time to influence the war.
Groves asked Oppenheimer to join him in Chicago a week later for further discussion. There they boarded the New York–bound 20th Century Limited passenger train and, crammed into a rattling roomette with Nichols, talked about how to organize a bomb design lab: “setting up an organization, building the needed facilities . . . and dealing with such expected issues as recruiting scientists and confining them in a laboratory in a remote area,” Nichols recalled.
Groves originally had considered appointing Lawrence to head the bomb lab but decided to leave him in place overseeing the crucial electromagnetic separation wo
rk. Of the other possible candidates, Compton could not be spared from Chicago, and Harold Urey, a chemist, was not qualified to manage a physics lab. By the time Oppenheimer disembarked from the 20th Century in Buffalo to return west, Groves recognized that he was the best choice of all.
He did have several concerns, however. Oppenheimer had no administrative experience. Nor did he have a Nobel Prize. Groves, who thought the latter shortcoming deprived Oppie of the scientific prestige the project leader should possess, plainly was unaware of the scientist’s towering reputation in the physics community. Another concern was Oppenheimer’s leftist background. Citing Oppie’s habit of dabbling in politics, the Manhattan Project’s security apparatus, which was not yet entirely under Groves’s control, refused to clear him for war work. To overcome that obstacle, Groves sought help from Lawrence, who delivered an enthusiastic personal recommendation and provided Groves with a letter intended to vanquish all doubts: “I have known Professor J. Robert Oppenheimer for fourteen years as a faculty colleague and close personal friend,” Ernest wrote. “I am glad to recommend him in highest terms as a man of great intellectual caliber and of fine character and personality. There can be no question as to his integrity.”
Lawrence further assured Groves that, if Oppenheimer failed in his task, he would step in himself. With that escape clause in his pocket, Groves ordered the Manhattan Project’s security staff to award the necessary clearance “irrespective of the information which you have concerning Mr. Oppenheimer,” and appointed him to head the desert laboratory that would be known as Los Alamos.
Big Science Page 28