While working on atomic bomb research a few months later, in September 1941, Enrico Fermi muses to Edward Teller ("out of the blue") whether a fission explosion could ignite a fusion reaction in deuterium. After some study Teller concluded that it is impossible and although no further work on the subject followed for awhile, this conversation began Teller's eventual obsession with fusion bombs.

[Historical footnote: During World War II, the idea occurred in Germany that convergent shock waves and collapsing shells might focus enough energy to allow conventional high explosives to ignite limited fusion reactions. This idea was probably inspired by Gudderly's work in converging shock waves, and certainly by the Allied attempts to destroy the heavy water plant at Vemork, Norway. Since German physicists considered fission weapons to be beyond reach during the current war, they concluded that the Allied interest in heavy water must be due its application in high explosive weapons. The Germans actually checked craters left by the British "Grand Slam", the largest conventional bomb dropped during the war, to discover whether its unusual power was due to fusion boosting. Polish researchers in the 60s and 70s reported actually generating fusion neutrons through convergent shock waves. Although the theoretical possibility remains, no one has apparently ever released significant amounts of energy this way.]

Research into the possibility of fusion weapons took an irregular and halting journey from the time of Fermi and Teller's conversation until bombs were actually built in the early 1950s. During WW II there was an initial surge of interest once fission bomb physics was fairly well understood. After preliminary theoretical investigation it was realized that much better experimental data was needed, and a fusion research program was included in the Manhattan Project at Los Alamos. Continuing theoretical investigations took repeated turns towards optimism, then pessimism, and back again. As the difficulty of the enterprise came clear, its priority was steadily downgraded. Teller on the other hand grew so captivated by the problem that he became unable to fulfill his duties at Los Alamos, was relieved of all technical leadership responsibilities, and was eventually transferred to a separate study group to prevent him from interfering in the work of others on the atomic bomb.

During July-September 1942 Oppenheimer's theoretical study group (Oppenheimer, Bethe, Teller, John Van Vleck, Felix Bloch, Robert Serber, and Emil Konopinski) in Berkeley examined the principles of atomic bomb design, and also considered the feasibility of fusion bombs. Megaton range fusion bombs were considered highly likely.

April, 1943 - During the initial organization effort at Los Alamos, Bethe is selected over Teller to head the Theoretical Division. Teller is soon placed in charge of lower priority research on fusion weapon design (designated the Super), but remains responsible for much theoretical work on the fission weapon as well.

February, 1944 - The Los Alamos Governing Board reevaluates deuterium fusion research and determines that tritium would be necessary to make an explosive reaction. Priority of fusion bomb work is further downgraded.

May, 1944 - Teller is removed entirely from the Theoretical Division to prevent his interference with fission bomb work. He is placed in charge of a small independent group for fusion research.

At the end of the war most of Los Alamos' scientific and technical talent, and virtually all of its leadership, left for civilian careers. Teller was among those who left. For a period of time very little progress on weapon research of any kind occurred. A conference, chaired by Teller, was held in April 1946 to review the wartime progress on the Super.

The design at that time was for a gun-type uranium fission bomb to be surrounded by about a cubic meter of liquid deuterium, with the whole assembly being encased in a heavy tamper. A large but undetermined amount of tritium would be required to ignite the reaction. If the amount of tritium required was too large, then the bomb would be impractical. Since the fusion of one T atom releases 8% the energy of the fission of a Pu-239 atom, with which it competes for neutrons in production reactors, the energy boost from D+D fusion must be considerably more than a factor of 10 greater than that released by the tritium starter fuel before the Super could be worthwhile.

The assessment at the time of the conference was that the Super was basically sound, but that more detailed calculations would be required verify it. Also present at the conference was Klaus Fuchs, who was spying for the Soviet Union. The Soviets thus were well informed about American interest and optimism about fusion weapons.

In mid-1946 Teller developed an idea he called the Alarm Clock. This involved the use of fusion fuel, specifically lithium-6 deuteride inside a uranium tamper of an implosion fission bomb. The idea was that the fission neutrons would breed tritium form the lithium, and fission energy would compress and heat the fusion fuel and ignite a reaction. A fusion-fission chain reaction would then proceed between the fusion fuel and tamper until the bomb disassembled.

By the end of 1946 Teller thought the Alarm Clock idea unpromising. In his September 1947 memorandum "On the Development of Thermonuclear Bombs" he was pessimistic about Alarm Clock's potential, but felt that it, like the Super were possible and required further study. Due to limitations in computing devices then available, he proposed delaying further work on both approaches for two years. If work had proceeded on the Alarm Clock design at this time the U.S. could probably have tested a device similar to Joe 4 (see below) before the end of 1949.

In the four years following the end of the war about 50% of the Los Alamos Theoretical Division's effort went into studying the Super, although its size and talent were much reduced from wartime levels. The absence of good calculating machines hampered the massive numerical computations that were required and greatly slowed progress.

By 1949 the Cold War was in full swing, with the Berlin blockade and Communist governments seizing control throughout Eastern Europe. This included Teller's homeland of Hungary, where much of his family still lived. Early in the summer of 1949 he thus rejoined Los Alamos to pursue the Super. On August 29 the first Soviet atomic bomb, code named RDS-1 and called Joe 1 by US intelligence, was exploded breaking the US nuclear weapon monopoly.

Up to this time the more detailed work on the classical Super design had showed that it was marginal at best. The large amounts of tritium required made it extremely expensive for the yield produced, and it was not even certain that the design would work at all. Teller remained optimistic however. During the next few months Robert Oppenheimer, as head of the Atomic Energy Commission's General Advisory Council (GAC) consistently opposed accelerating work on the Super due to its demonstrated shortcomings.

Despite this, on 31 January 1950, Pres. Truman announced that the US would proceed to develop hydrogen bombs. A couple of weeks after Truman's announcement, Teller issued a 72 page update of "On the Development of Thermonuclear Bombs". In this paper he again regarded both the Super and Alarm Clock as viable candidates for weapons development, but again proposed delaying decision on full scale development of either for another two years.

At this time Soviet research on the subject was already well underway, focusing on the Sakharov-Ginsberg version of the Alarm Clock concept which they called the Layer Cake. A special department was set up in March 1950 to proceed with actual Layer Cake weapon development.

By February 1950, immediately after Truman's decision, Stanislaw Ulam had discovered by hand calculation that even more immense amounts of tritium than previously believed would be necessary for the Super to have any chance of success. When Ulam and Cornelius Everett completed more detailed computations on June 16, the design even these huge amounts of tritium appeared to be inadequate. Additional analysis by Ulam and Enrico Fermi nailed the coffin shut on the classical Super. When John Von Neumann's newly invented ENIAC began doing extensive calculations on the problem later in the year, the negative results were simply piling more dirt on the grave. Until early 1951 real progress on hydrogen bomb development was impossible because no one knew how to proceed.

Information provided by: http://www.fas.org