Machine Work
As Tuve observed when shelving the almost finished Carnegie cyclotron in favor of defense work, many other directors of cyclotron laboratories were imagining how their instruments might be commissioned in the war effort. Those with half-built machines worried that otherwise they would not be able to acquire needed material; those with functioning production machines wanted to keep their staffs together and their clients supplied. But what purpose central to the national defense could a cyclotron serve? In the fall of 1939, J. Stuart Foster, casting about frantically for arguments to go forward with long-laid plans for a cyclotron at Toronto, proposed radioactive labelling of foods, metals, and strategic materials, to discover what was being shipped to enemy territory, and of documents, to detect them when stolen; radioactive beacons, to demarcate combat zones, batteries, and so on. Lawrence's reply to Foster's fancies suggests that he had not thought about military uses of his invention. "It is difficult for me to suggest concrete practical applications of the cyclotron in warfare, but it seems to me there are possibilities along the lines you have suggested."[107]
As late as June 1941, directors of cyclotron laboratories not privy to the doings of the uranium committee saw only mundane uses for their equipment. A.L. Hughes, of Washington University in Saint Louis, whose cyclotron neared completion, to cyclotron headquarters: "I suppose that Defense application of the cyclotron falls into two classes, the production of radioactive yttrium as a substitute for radium in the examination of castings and the production of tracer elements to help chemists solve their problems." Lawrence replied encouragingly. Hughes's machine got its first beam on December 10, 1941, just after Bush and Conant had decided to pursue bombs made of U235 and perhaps also bombs of element 94. Washington's deuteron beam was very large, almost 0.5 mA. It went to work producing samples of plutonium, in
which it outdid the Crocker cracker. News gets around. At the end of December, A.C.G. Mitchell of Indiana, who had inquired over a year earlier for defense work on his cyclotron to keep Kurie and Laslett, asked Lawrence whether there might now be a "possibility that we could get a project at Indiana to use our magnet for a defense job?"[108]
Karl Compton, a man very much in the know, tried to solve the financial problems of the MIT and Harvard cyclotrons and to further the purposes of OSRD by having them declared "stand by defense group[s]." He arrived at this strategy in November 1941, after discussion with Lawrence. Compton's reasons: both laboratories needed high-priority items for supplies; both faced the breakup of their groups unless their members could "feel that they are engaged on recognized national defense work;" and lack of money. In fact, the Harvard cyclotron had received a top priority rating in August 1941, along with a contract from the NDRC to make radioisotopes for other NDRC projects; but no requests had come for over seven months.[109] Harvard got a little business—some radiosilver, some radioarsenic—in February 1942, following a directive from NDRC to use Harvard rather than Berkeley unless "very active material is required." Insufficient demand closed it down, all set up; in which condition it was later carted to Los Alamos, where it saw important service in fast-fission work under the command of Robert Wilson.[110]
The Berkeley cyclotrons early entered into business relations with the NDRC. A contract for production of miscellaneous isotopes existed by January 1941. It supported five research assistants. The Laboratory charged the government $25.25 an hour for
overhead computed at 50 percent of direct costs. For a time in the spring, the machine ran 24 hours a day (it then dropped back to a more normal 12) to supply other NDRC contractors and the usual internal and external users. These orders could mount up. In the fall of 1941, the Laboratory billed $2,700 for 104 hours of cyclotron time to make 500 µCi of radioiodine and 30 mCi of radioarsenic for experiments at Caltech.[111] The contracts also provided new equipment for chemical separations, mechanisms for remote handling of hot isotopes, and so on, permanent gains from evanescent products. Kamen oversaw most of the production, under "terrific tension."[112]
The NDRC also picked up the cost of the work on element 94. Early in January 1941, Seaborg prepared at Lawrence's request an account of bombardment time and materials expended. He understood the purpose: "I believe that he is considering the possibility of having the government finance this work as an official project." In accounts rendered on March 11, 1941, $835 was charged to making element 94 by deuterons and $8,310 to the test of its fissionability by slow neutrons, exclusive of salaries. By far the largest item in both cases was cyclotron time, 315 hours completed or estimated, most of it at the 60-inch.[113] As we know, Lawrence acquired further support through the uranium committee when he advised Briggs later that March. The demonstration in May by Kennedy, Seaborg, Segrè, and Wahl that 94239 fissions with slow neutrons and the follow-up in July by Segrè and Seaborg, which showed that it does so with fast ones as well, and much more readily than U235 , brought a good deal more from the NDRC, some $21,500.[114]
The demonstrated fissionability of 94235 with fast neutrons confirmed its candidacy as an explosive. It might therefore appear odd that the third, and definitive NAS report commissioned by Bush, the report of November 1941 promoted by Lawrence, did not mention 94. But that was to follow the lead of Bush and the British, who had considered 94 chiefly in relation to a power generator. Joseph Rotblat of Chadwick's group at Liverpool had guessed at the fissionability of 94 in June or July 1940 and Bretscher had done the same by November, reasoning from the theory of Bohr and Wheeler. To go further, Bretscher told the MAUD Committee, he would need a sample procurable only at Berkeley; and he agitated "whether we should ask for facilities to enable us to work there."[115] When Oliphant visited the Laboratory in September 1941, he saw what could be accomplished with good funding, practiced investigators, and a big cyclotron. "When I saw at Berkeley the work going on there on 93 and 94 and saw activities of more than one curie of 93 I felt that you [Bretscher] were struggling against very difficult circumstances!" On this testimonial, Bretscher asked again whether he might not have samples of 93 and 94, "for which one requires neutron sources of the strength available in California."[116] But by then the new intensity of the S-1 program had put a premium on every atom of 94239 made in the Berkeley cyclotrons.
By the time of Pearl Harbor, the 37-inch and 60-inch machines had spent much of their time for almost a year on work that the NDRC deemed to be in the national interest. By then the 184-inch had also come under government protection, if not into government service. The commissioning occurred during the summer of 1941. Lawrence had been prompt in ordering the steel and copper, at the discounted prices that he and Loomis had negotiated with U.S. Steel (a savings of $30,000) and Phelps-Dodge; and also the power supply from GE, at its generous price without overhead charge. The contracts for the metal were let in July, and
for the construction of the magnet in September; the first installment of steel, some twenty-eight tons, a good chunk to be sure but less than 1 percent of the whole, went up the newly built road to the hilltop site on the last day of October 1940.[117] Problems with the unions in December occasioned delays in building, but design work went forward expeditiously. Around New Year the parameters had been set, or almost so, to obtain deuterons of 100 MeV: the minimum gap between the poles would be 40 inches, or even 4 feet, to allow sufficient clearance between the dees and the chamber walls to permit peak voltages of a million across the accelerating gap. By the middle of February 1941, Lawrence had succeeded in spending $534,550 on the he-man cyclotron.[118]
In the spring, however, just as workers were to start raising the cyclotron around the 30-foot tall magnet frame, construction met resistance that money alone could not overcome. To assure its supplies, the Laboratory tried to buy up all the copper and brass available locally before the government restricted or acquired the stock. In the summer, some contractors refused to ship equipment they had already made unless the 184-inch project could acquire a priority rating. In the fall, Salisbury, visiting GE's plant to enjoy the sight of the power transformers advertised as partially completed, learned that the completion referred only to the blueprints. "I was told that nothing would be done for at least a year unless we could get a priority." The same message came from Phelps-Dodge. "This unexpected turn of events made it necessary for us to appeal to the National Defense Research Council," Lawrence wrote the Rockefeller Foundation, in explanation of the admission of a new partner to their agreement. "Washington has been most cooperative. . . . [The] construction program is now going full steam ahead."[119] Lawrence viewed the intervention as a favor,
not as a measure for defense; most of the materials needed had already been shipped to Berkeley before priorities obtruded; the "blessing of the OSRD," Lawrence told the Research Corporation, had fallen on the project "in line with a wise and foresighted policy of encouraging steady scientific progress in the midst of the stress of war activities." He hoped to have the cyclotron ready for trials in 1943.[120]
If this encouragement were truly the purpose of the OSRD, virtue reaped its reward. The big magnet proved a fateful instrument of war. But no one saw it in the spring or summer of 1941. The uranium committee had paid for the electromagnetic separation of microgram samples of U235 only for determination of its nuclear parameters. No more than its British counterpart did the uranium committee or the first two NAS committees consider electromagnetism to be an option for large-scale separation. The third NAS committee—the one that reported on November 6—mentioned, without specifying, "other methods" than centrifugation and gaseous diffusion then under or needing investigation.[121] As an expert on large magnets, Lawrence took on responsibility within S-1 for electromagnetic separation of small samples of light uranium for experimental purposes. Nier came to Berkeley in November to help Brobeck convert the 37-inch cyclotron into a mass spectrograph. The principle of operation is represented in figure 10.3: gaseous uranium ions from the source traverse circular paths under the old Poulsen magnet, the lighter isotope following the tighter circle and ending (if all went well) in a collecting cup. The first rough test at the end of November gave encouraging results; the mechanism, which became known as the "calutron" after the institution that gave it birth, appeared likely to contribute samples of U235 for experimental purposes.[122]
This performance suggested that the chief technical difficulties in large-scale separation—the skimpiness of the beam and the menace of space charge that deflects the ions from their circles—
Fig. 10.3
Principle of the calutron. It is not easy to obtain the clean and copious
separation indicated. Smyth, Atomic energy , 164.
might be overcome. Already in October, Lawrence and Henry Smyth of Princeton had discussed the difficulties and persuaded each other to be optimistic. They had quite different corrections in mind. Lawrence thought to apply cyclotron technology and to build bigger; Smyth, who had some experience with Princeton's mass spectrometer, sought a new approach that would do without the collimating slits and so make use of an extended source. The ever-resourceful Robert Wilson found a way to help his colleague Smyth and challenge his teacher Lawrence. Wilson proposed to do without slits and magnetic fields by adapting the principle of the klystron: a wide beam of uranium ions would pass through a cavity oscillating at radio frequencies and emerge in two sets of bunches, one of each uranium isotope; at the cross section of the drift space where the bunches are best defined, a transverse rf field would work, phased to draw aside the heavier isotopes while allowing the lighter to pass on to a collector. The "isotron" (so named for no reason at all) as well as the calutron received funding at the meeting of the uranium committee on December 18, 1941.[123]
In January 1942, promising experimental versions of both machines existed. But the isotron scarcely had a chance. Lawrence's means were not limited to what the government chose to give him—Brobeck had converted the 37-inch to a calutron on money from the Research Corporation—or to local or unpracticed help. A call went out, and cyclotroneers came home—James Cork, J.R. Richardson, and Robert Thornton, among others—to join or rejoin men who had not left. (Tables 10.2 and 10.3 tell who ended where.) These men knew their business. They put poles on the frame of the 184-inch magnet and went to work squashing the bugs in calutron prototypes tested under the Rockefeller Foundation's investment in the higher aspirations of the human species. They squashed the isotron soon enough. "Ernest wanted to cannibalize our group," Wilson remembered. "We resorted [in vain] to every device of politics and rhetoric to forestall the takeover."[124] It took most of the rest of the war, most of the efforts of the Laboratory, and most of a billion dollars to make successful calutrons.