they talked about vitamins, about which Gold was knowledgeable. (In 1943, Gold and a friend had applied unsuccessfully to the Corn Exchange Bank in Philadelphia for a loan to start a vitamin assay laboratory, one of Gold's many unfulfilled dreams.) Gold told them he was a biochemist from Pittsburgh with a wife and two children. Though he never said so, he presumably learned during his visit to the Heinemans that Klaus Fuchs was still at Los Alamos.
Whether Fuchs passed information to Lona Cohen in New Mexico in the winter and spring of 1945–1946 as well as to another courier, as he confessed, “soon after he returned to the United Kingdom,” much of what he passed added further detail to his previous communications about implosion, plutonium metallurgy, initiator design and bomb effects. He attended talks at Los Alamos on experimental data developed on levitated implosion and the composite core on March 11, 1946; on the possibility of thermonuclear reactions in water and air on March 12; on nuclear breeder and power reactors on March 21; on the processing of plutonium from nitrate to metal on April 1, all information of value to Soviet research. But information about the thermonuclear — Teller's superbomb — became available to Fuchs as well before he left Los Alamos in June 1946. Fuchs confessed to communicating some of what he learned to Harry Gold and to his postwar espionage contact in Britain. He is unlikely to have passed less than everything he knew.
The idea of a superbomb exploiting the fusion of light elements as well as the fission of heavy elements was a logical extension of basic ideas in nuclear physics known to physicists throughout the world. As Oppenheimer drummed into the Acheson-Lilienthal board of consultants (until Herbert Marks could recite it by heart), “Only in reactions of very light nuclei, and in reactions of the very heaviest, has there ever been, to the best of our knowledge, any large-scale release of atomic energy.” In May 1941, University of Kyoto physicist Tokutaro Hagiwara, in a lecture on “Super-explosive U235,” had commented that the fissionable uranium isotope “has a great possibility of becoming useful as the initiating matter for a quantity of hydrogen.” Hagiwara was the first scientist on record to notice that an explosive fission chain reaction might generate enough energy to force hydrogen to fuse to helium, with the potential for producing a far larger nuclear explosion than fission could yield alone.
Ernest Rutherford and two of his younger colleagues at Cambridge, Marcus Oliphant and Paul Harteck, had discovered the hydrogen fusion reaction in 1934. In a paper titled
“This was another of that long catalogue of scientific papers which came before their time,” writes historian David Irving. “In retrospect, [this] paper can be seen to have been of little less moment that Hahn and Strassmann's 1939 paper on the fission of the uranium nucleus.” Because the fusion reaction depended on heating the nuclei until their thermal motion overcame their electrical repulsion, the reaction came to be called “thermonuclear fusion.” It could be created a few nuclei at a time in particle accelerators such as Cambridge's Cockcroft-Walton generator or Berkeley's cyclotron. In 1938, Hans Bethe identified a sequence of thermonuclear reactions proceeding from hydrogen to carbon as the source of the energy that lit the sun and stars. But until a fission chain reaction became feasible, no one had imagined that a large-scale thermonuclear fusion reaction could be kindled on earth. (“In the center of an exploding fission bomb,” notes theoretical physicist Herbert York, “temperatures substantially exceeding 100,000,000 degrees are produced, and so at least one of the conditions necessary for igniting a thermonuclear reaction under the control of man seemed to be within reach.”)
If Hagiwara was the first, his insight fell on fallow ground — Japan in wartime lacked the resources even to develop an atomic bomb, much less to explore a thermonuclear. But the same idea occurred to Enrico Fermi at Columbia University and he passed it along to young Edward Teller in September 1941. Fermi wondered if an atomic bomb might serve to heat a mass of deuterium sufficiently to kindle a full-scale thermonuclear reaction. If so, then cheap deuterium distilled from seawater could be added to a critical mass of expensive U235 or plutonium. Each gram of deuterium converted to helium should release energy equivalent to about 150 tons of TNT, 100 million times as much as a gram of ordinary chemical explosive and eight times as much as a gram of U235; theoretically, twelve kilograms of liquid deuterium ignited by one atomic bomb would explode with a force equivalent to one million tons of TNT — one megaton; a cubic meter of liquid deuterium would yield ten megatons. Teller made the realization of Fermi's idea the focus of his life.
At a secret seminar on atomic-bomb development that Robert Oppenhei-mer chaired and Teller, Bethe, Robert Serber and other theoretical physicists attended at Berkeley in the summer of 1942, the possibility of a thermonuclear bomb was discussed at length. During that discussion one of the participants, a young theoretician from Indiana University named Emil Konopinski, suggested mixing another isotope of hydrogen, radioactive hydrogens, tritium (one proton, two neutrons), into the thermonuclear fuel. Tritium (T) was much rarer than deuterium (D) but because of its nuclear characteristics ought to kindle thermonuclear reactions at a far lower ignition temperature, 40 million degrees rather than 400 million. The cross section for fusion (a measure of probability) of D+T turned out to be one hundred times greater than the cross section for fusion of D + D. Teller began systematic theoretical studies of the thermonuclear at Los Alamos in autumn 1943, devoting his full time to the project with Oppenheimer's approval in the last year of the war. Among others in his group, Teller signed on a Polish mathematician named Stanislaw Ulam, whom John von Neumann had recommended, to help with the work; on Ulam's first day on the job, late in 1943, Teller asked him to study the exchange of energy between free electrons and radiation in a hot gas, one process that might cool a fusion reaction sufficiently to prevent it from propagating. Oppenheimer arranged for a small quantity of tritium to be bred in an Oak Ridge reactor for cross- section measurements and other research. Manhattan Project heavy-water production was intended for thermonuclear studies as well as for reactor research.
Since kindling a thermonuclear explosion required setting off an atomic bomb, its progress could not be studied in the laboratory, as fission had been studied with tabletop near-critical assemblies and at full criticality with diluted cores in the notorious Dragon critical-mass experiments. Instead of real experiments, Teller and his group had to depend on mathematical calculations of unprecedented complexity which nevertheless greatly oversimplified the phenomena they modeled. To establish the initial conditions for the thermonuclear explosion, Los Alamos needed to understand the fission explosion that preceded it in great detail: the behavior of the immense flux of neutrons which the fission explosion produced (the neutronics), of the immense flux of heat released (the thermodynamics) and of the fluid flow of particles and radiation released in the explosion (the hydrodynamics). These fission calculations had been started by Richard Feynman and others during the war as part of the work of implosion research, using IBM punch-card machines to automate the thousands of necessary repetitions. The calculations were repetitive because they followed the histories of dozens or hundreds or thousands of individual particles through cross-sectional slices of time as the explosion bloomed — like catching the successive positions of a hall full of dancers with the quick pulses of a strobe.
Understanding the fission explosion was only the first step in thermonuclear explosion calculations, however, and by the end of the war even that step had been advanced only tentatively and crudely, for a small sample of particles through relatively thick slices of time. Thermonuclear calculations added significantly higher levels of complexity, Stanislaw Ulam writes:
All the questions of behavior of the material as it heated and expanded — the changing time rate of the reaction; the hydrodynamics of the motion of the material; and the interaction with the radiation field, which “energy-wise” would be of perhaps equal importance to that of the thermal content of the expanding mass — had to be formulated and calculated…
To realize… the magnitude of the problems involved, one should remember that, even only mathematically, the problem of the start and explosion of a mass of deuterium combined a considerable number of separate problems. Each of these was of great difficulty in itself, and they were all strongly interconnected. The “chemistry” of the reaction, i.e., the production, by fusion, of new elements not originally present and the appearance of tritium,
