(55f) Estimation of Deuterium Recovery from the Separation of Water-Isotope Mixture by Thermal Diffusion

ABSTRACT Nuclear energy perhaps will be supplied by thermonuclear fusion with deuterium as the nuclear fuel, instead of fission reactors, in the distant future. Thermal diffusion is a well-established method for separating isotopes, especially for hydrogen isotopes because of the large ratio in molecular weights. It was great achievement of Clusius and Dickel (Naturwiss. 1938;26:546-52) to point out that a horizontal temperature gradient produces not only thermal diffusion in the direction of the temperature gradient, but also natural convection of the fluid upward near the hot surface and downward near the cold surface. These convection currents produce a cascading effect analogous to the multistage effect of a countercurrent extraction or distillation, and as a result a considerably greater separation may be obtained. The first complete presentation of the separation theory of thermal diffusion was that of Furry and co-workers (Phys. Rev 1939;55:1083-95) The separation of water-isotope mixture (H2O-HDO-D2O) in a batch-type thermal diffusion column was done in the previous work (Chem. Eng. Sci 1984:39:1277-82). The separation theory, as well as the time histrories of the concentrions of HDO (C2)and D2O (C3), were derived with the assumption of local equilibrium (D2O+H2O°ê2HDO) within the column. In present study, the pseudo-product form of concentration, , was simplified, and the recovery of deuterium (D) is thus estimated by calculating the concentration difference of D at the bottom (CD,B)and top (CD,T) of the column as (CD,B-CD,T)=[(4/20)C3,B+(2/19)C2,B]-[(4/20)C3,T+(2/19)C2,T]. It is found that the experimental results are in fair agreement with prediction values.