Liquid fluoride thorium reactor
A liquid fluoride thorium reactor or LFTR is a variant of the molten salt reactor that breeds fissile uranium-233 fuel from thorium in its operation. As the name indicates, the reactor uses molten fluoride salts to both contain the nuclear fuel and to transfer heat from the core. In the proposed designs for LFTR, molten fluoride salts of lithium and beryllium (called FLiBe) serve as solvent for fissile uranium and fertile thorium salts, neutron moderator to promote fission of uranium-233, and heat transfer medium in a separate coolant loop.
A single fluid LFTR uses a large fluoride-resistant reactor vessel with graphite moderator rods. Although relatively simple to build, the low rate of breeding means that fissile uranium may need to be added periodically.
A two fluid LFTR physically separates the core salt containing fissile uranium-233 from a blanket salt containing fertile thorium, but uses the core salt's neutron emissions to breed uranium-233 from the thorium in the blanket. This has a better breeding rate and can theoretically sustain itself on thorium fluoride once started, but adds complexity to the reactor design and potentially opens up a way to create very pure uranium-233, which raises proliferation concerns. In order to mitigate complexity, a hybrid of the single and two fluid LFTRs, or a "1.5 fluid" LFTR, has been suggested, which simplifies the reactor design and allows for the higher breeding ratio of the two fluid LFTR, at the cost of more complex fuel processing systems.
Combining the thorium fuel cycle with a molten salt reactor in this way confers a number of advantages over current reactor technology and even the MSR itself. First, because no uranium-238 is used (unless the LFTR is "denatured"; see Variants), there is a much lower probability of transuranic (plutonium, americium, curium) formation. In fact, the probability of plutonium-238 formation is much higher, and this is far more useful in peaceful applications like deep-space probes. Second, uranium-233 has the largest fission cross-section of any fissile material, fissioning 92% of the time on neutron capture. Third, if the reactor design does not isolate protactinium-233 from neutron flux, the resulting uranium-232 contamination is a very effective proliferation deterrent; the IAEA recognizes a ratio of uranium-232 to uranium-233 as low as 1% to be effective. Finally, the design can potentially consume abundant thorium at a far higher efficiency than solid fuel reactors, meaning a very small amount of thorium can provide a very large amount of energy-- this can solve the thorium problem for US rare earth mining, and satisfy the high energy requirements of fabricating monocrystalline silicon, helping the production of renewable technologies.
Because of the lapse in research, however, challenges remain. Even with isotopic separation of lithium in the FLiBe salt, radioactive tritium can be formed under neutron flux, and so the reactor must account for hydrogen fluoride formation outside of the fuel processing area, the radiation from the tritium, and the formation of helium-3 (which, incidentally, is a substance of interest in nuclear fusion.) While the Molten Salt Reactor Experiment tested many scenarios involving reactor-grade graphite and Hastelloy-N, their performance in the two-salt or 1.5-salt configurations at larger scales and high temperatures is yet to be tested.
The denatured LFTR is a single-fluid theoretical design initially proposed by Oak Ridge National Laboratory. In the original specification, it is a once-through design that incorporates uranium-238 for proliferation deterrence. This eliminates much of the need to concentrate the fuel salt using a chemical processing unit, but requires enriched uranium in the fuel and will produce more transuranic elements than a LFTR sustained by thorium alone, and breeds less uranium-233.
While fissile material such as uranium-235 would be ideal in starting the nuclear reactions in the core salt, some designs (among them cited by Dr. Takashi Kamei of Japan) can prepare the reactor's uranium fuel by irradiating thorium with a particle accelerator. This accelerator-driven LFTR design is an example of a subcritical reactor, and fission will stop soon after the accelerator is deactivated, providing an additional layer of safety.