It has been proposed to use liquid lithium as a plasma facing surface for the divertor and walls of a tokamak fusion reactor. There are numerous physics processes which suggest that such a reactor would be much more compact and in some sense simpler than existing conceptual designs for a tokamak reactor, which employ solid tungsten wall and divertor construction. A major stumbling block for this approach, however, is the need for removal of deuterium and especially tritium from the lithium inventory used to form the wall and divertor. Although the concentration of tritium, for example, in the liquid lithium would be low (<1%, perhaps approaching 0.1%), the limits on tritium inventory for a fusion reactor are likely to mandate frequent removal of tritium from the lithium. It is generally assumed that the inventory limit for tritium will be of order 1 kilogram. Since the amount of liquid lithium necessary to form the plasma facing surfaces is on the order of 100 kilograms to a few hundred kilograms, any removal system must be capable of dealing with low concentrations. In response to this issue, we have been developing concepts for the removal of tritium and deuterium (as well as hydrogen) from liquid lithium.
A disclosure was filed in 2015 for an electron beam–based evaporation system, which could liberate tritium and deuterium from liquid lithium. This approach would be much more efficient if the deuterium and tritium were present at higher concentrations – the efficiency could obviously be improved by an order of magnitude if the tritium and deuterium were present at the 10% level, rather than at concentrations of less than 1%. The use of centrifugal extraction to produce higher concentrations of lithium tritide (LiT) and lithium deuteride (LiD) is suggested here as a straightforward approach.
LiD and LiT will form from dilute solutions of deuterium and tritium in lithium when the temperature of the mix is reduced toward the melting point for pure lithium (182 C), since the solubility of hydrogen in lithium becomes very small as the temperature is reduced. This process has been cited by Masa Ono (PPPL) as a means to precipitate out LiT and LiD from lithium. Here we propose using the large density difference (LiD and LiT are nearly twice as dense as pure liquid lithium) as the basis for rapid centrifugal separation of an enriched slurry of lithium, LiD, and LiT, which can be subsequently broken down by electron beam heating.
We have further proposed that magnetic forces be employed to physically spin the liquid lithium without rotating the container. It may be difficult to avoid excessive mixing of the fluid with this approach. A carefully applied, slowly accelerated, rotating magnetic field may avoid excessive mixing; we are exploring this alternative approach to mechanical centrifuges.
In summary, this disclosure deals with the use of a centrifuge to concentrate impurities in liquid lithium, specifically deuterium and tritium.