From the late 1940s though the 1950s, the overwhelming concern of the U.S. Atomic Energy Commission (AEC)--and indeed of nuclear establishments worldwide--was the production of weapons-grade nuclear materials. This was done by enriching uranium to very high levels of the isotope uranium-235 or by forming plutonium-239 from abundant uranium-238 in a specialized nuclear reactor. In the U.S., massive gaseous diffusion plants for uranium enrichment at Oak Ridge, Tennessee, and later at Portsmouth, Ohio, were constructed to produce highly enriched uranium for weapons use. By 1957 roughly 7% of the entire electrical output of the U.S. was being devoted to uranium enrichment for weapons. Weapons-grade plutonium was produced at specialized military reactors at Hanford, Washington, and at the Savannah River plant in South Carolina. These plants produced no electrical power, but contained large facilities to extract plutonium from processed uranium.
For many years, reactor development for civilian power use was a low priority of the AEC. Even the projects the AEC pursued in civilian nuclear power had military applicability, with the most notable being the liquid-metal fast breeder reactor (LMFBR). The LMFBR was capable of simultaneously producing weapons-grade plutonium and producing electrical power, giving it a particular advantage over the reactors at Hanford and Savannah River in the eyes of the AEC. AEC funding of the LMFBR led to the first demonstration of electricity generation from nuclear power at the Experimental Breeder Reactor-1 in Arco, Idaho, on December 20, 1951.
Meanwhile, research into a nuclear-powered airplane led to a stunning conceptual breakthrough in nuclear reactor design: a reactor whose fuel was entirely dissolved in chemically stable fluoride salts. First proposed by R.C. Briant of Oak Ridge National Laboratory (ORNL) in 1951, the liquid fluoride reactor was radically different from other reactors that relied on solid fuel. The liquid fluoride reactor had tremendous safety and performance advantages over solid-fueled reactors, as well as a remarkable versatility in potential fuels. A proof-of-concept fluoride reactor was built and operated in 1954 at Oak Ridge. It was called the Aircraft Reactor Experiment (ARE), and it demonstrated that fluoride reactors had the chemical and nuclear stability that Briant and his colleagues had predicted. After the success of the ARE, the fluoride reactor was baselined for the nuclear aircraft project, but the advent of intercontinental ballistic missiles led to cancellation of the nuclear aircraft in 1960.
Dr. Alvin Weinberg, the director of ORNL and the inventor of the solid-fueled light water reactor (LWR), recognized the remarkable potential of the fluoride reactor and turned the attention of the fluoride reactor team from aircraft propulsion to terrestrial energy. He was particularly impressed with the ability of the fluoride reactor to safely and efficiently use thorium. Unlike any other reactor power source, a liquid form of thorium existed (thorium tetrafluoride, ThF4) that could be easily reprocessed to unlock thorium's potential.
In 1959, using his contacts in the AEC, Weinberg pushed for funding of a more advanced demonstration of fluoride reactor technology and was able to win funding for the second fluoride reactor: the Molten-Salt Reactor Experiment (MSRE), which was built and operated by ORNL from 1965 to 1969. The MSRE was a much-improved design over the ARE and led to the demonstration of reactor operation on different fuels, stable self-controlling operation without control rods, removal of reactor poisons online, and strong passive safety features.
Nevertheless, the MSRE and fluoride reactors in general could not fulfill the most important mission of the AEC in those days: the production of weapons-grade plutonium. The LMFBR could make the plutonium for the nuclear build up the AEC desired. Furthermore, the safety features of the fluoride reactor highlighted the safety risks of the LMFBR; after the first commercial LMFBR suffered a severe core meltdown in 1966, the meltdown-proof fluoride reactors offered a safe alternative to LMFBRs that proved politically embarrassing. With a great deal of money and political capital already invested in LMFBRs, the AEC moved to shut down all research on fluoride reactors at ORNL in the mid-1970s, and the fluoride reactor team was disbanded and assigned to other projects.