On the conference’s opening day, Fuchs and John von Neumann told Teller and Serber about a new idea they had for using an exploding atomic bomb to compress and ignite the Super’s thermonuclear fuel by implosion. A few weeks later, shortly before Fuchs returned to England, he and von Neumann jointly filed a classified patent on their invention.*92
In preparation for the conference, Teller and a half dozen of his lab colleagues had written a report—LA-551, “Prima Facie Proof of the Feasibility of the Super.” Prefaced by sixty pages of classified calculations and drawings, LA-551 concluded that a superbomb fueled by deuterium was indeed feasible. But the report’s recommendation—that “a large-scale theoretical and experimental program” be undertaken immediately to develop the weapon—amounted to Teller’s special pleading. Edward also urged, as the next logical step, a prompt start on making the tritium necessary for mass production of the Super.
The focus of the Los Alamos conference was upon one specific design, the so-called runaway Super that Teller had first proposed at the 1942 Berkeley summer seminar. His bomb consisted of a cubic meter of uncompressed liquid deuterium at the end of a cylinder attached to one or more high-yield, gun-type fission bombs. The design relied upon the energy from the detonation wave of the atomic explosion to trigger the thermonuclear reaction. A small amount of tritium gas was injected halfway down the tube, between the deuterium and atomic bombs, to start the deuterium burning. Theoretically, there was almost no limit to the power of the device; increasing the yield was simply a matter of adding more deuterium.
The question that dominated the second and third days of the conference—whether the thermonuclear flame would spread or simply go out—was the same problem that had bedeviled the summer meeting at Berkeley. The physics conundrum, which had almost caused the Super to be abandoned at wartime Los Alamos, returned to haunt Teller anew at the 1946 conference: as the energy from the exploding atomic bomb heated the mass of hydrogen fuel, energy was lost through radiation.93
Various schemes and fixes were proposed to get around this obstacle. But like a dog chasing its tale, the discussion went round and round; no solution was in sight.94
Serber claimed that Teller ultimately “solved” the problem by ignoring it. (Serber was at the meeting under protest; his Urbana colleague, physicist Phillip Morrison, had persuaded him to come, having predicted—correctly—that Teller would use the conference as a springboard to lobby for the Super.)95 Calculations done by von Neumann’s ENIAC computer at Princeton had deliberately left out the radiative cooling effect, since the complex hydrodynamics of thermonuclear burning were beyond the capabilities of the machine. Similarly, optimistic estimates were substituted for hard numbers in the critical opacity calculations, which remained unfinished by Mayer at Chicago.96
Teller summarized the results of the conference in a draft report, written that May. It conceded that definitive proof of the Super’s feasibility required the actual test of a finished device. He also acknowledged that a crash effort to develop the superbomb would necessarily draw away “a considerable fraction” of the resources needed to build a stockpile of atomic bombs in the coming years. Tritium, for example, could only be produced by Hanford’s reactors, which were already engaged full-time in making plutonium for the nuclear arsenal. Accordingly, a decision on whether to proceed with the Super had to be “part of the highest national policy,” Edward argued.
Shown the draft report by Teller, Serber thought it “incredibly optimistic,” protesting that the runaway Super was not a workable scheme, nor was a solution evident—much less “simple,” as Teller argued. Serber sat down with Teller later that month to write a revised version that was more realistic.97
Back in Chicago, Edward completed the final “Report of Conference on the Super,” LA-575, and sent it to Los Alamos. Two days later, on June 14, 1946, Fuchs left the New Mexico lab to return to England.98 Checking Teller’s report out of the Rad Lab library in Berkeley, Serber was upset to discover it essentially unchanged from the original version.99
* * *
Teller was back at Los Alamos a few weeks later with his family, following the end of classes at Chicago, Boasting that he had arrived at the lab “in the proper (or rather not proper but usual) crusader spirit,” Teller posed a rhetorical question to confidante Mayer: “Do you think there is any chance that I shall be somewhat less foolish than I have been?”100
In August, 1946, Teller proposed a new design for a hydrogen bomb, which he christened the “Alarm Clock” to distinguish it from the Super.101 (His Alarm Clock, Teller boasted, would wake up the world.) This device would use an atomic trigger to ignite alternating layers of enriched uranium and a mixture of deuterium and tritium arranged in concentric shells. Although more powerful than a conventional fission bomb, the Alarm Clock, unlike the runaway Super, could not be of unlimited yield. But Teller and the new leader of the lab’s Theoretical Division, Robert Richtmyer, hoped that it might be a practical alternative to the Super.102
By that fall, calculations by two of their colleagues at the lab—mathematicians Stanislaw Ulam and Nicholas Metropolis—showed that major unresolved problems remained with both the Super and the Alarm Clock. Despite the fact that these obstacles had yet to be surmounted, Teller urged Bradbury to begin scheduling tests of prototype thermonuclear weapons for the coming year.103
Bradbury simply ignored him.104 A seven-page letter that the lab director sent to the army in November, concerning various projects at Los Alamos, neglected even to mention the Super.105
It was not only Bradbury but Oppenheimer whom Teller now blamed for raising barriers to his pet project.
