A density‐functional theory approach to the separation of K₂ZrF₆ and K₂HfF₆ via fractional crystallization

Abstract

Because of the low absorption cross-section for thermal neutrons of zirconium (Zr) as opposed to hafnium (Hf), Zr-metal must essentially be Hf-free (<100 ppm Hf) to be suitable for use in nuclear reactors. However, Zr and Hf always occur together in nature, and due to very similar chemical and physical properties, their separation is particularly difficult. Separation can be achieved by traditional liquid-liquid extraction or extractive distillation processes, using Zr(Hf)Cl 4 as feedstock. However, the production of K 2 Zr(Hf)F 6 via the plasma dissociation route, developed by the South African Nuclear Energy Corporation Limited (Necsa), could facilitate the development of an alternative separation process. In this theoretical study, the results of density-functional theory (DFT) simulations of K 2 Zr (1-z) Hf z F 6 solid solutions [using Cambridge Serial Total Energy Package (CASTEP)] are presented, for which the supercell approach was applied in an attempt to determine whether solid solution formation during crystallization from aqueous solutions (fractional crystallization) is thermodynamically possible, which would hinder the separation efficiency of this method. Consequently, the calculated thermodynamic properties of mixing were used to evaluate the separation efficiency of Zr and Hf by fractional crystallization using a thermodynamic model to calculate the relative distribution coefficients. The small mixing enthalpies that were calculated from the DFT results, indicates that lattice substitution of Zr(IV) by Hf(IV) in K 2 ZrF 6 could occur with relative ease. This is not surprising, considering the close similarities between Zr and Hf, and it was therefore concluded that K 2 Zr (1-z) Hf z F 6 solid solution formation might well restrict the separation efficiency of Zr and Hf by fractional crystallization of K 2 Zr(Hf)F 6 .

Influence of KF and HF on the Selectivity of Zr and Hf Separation by Fractional Crystallization of K2Zr(Hf)F6

“…In Figure A, the sample pattern corresponds quite well with the low-angle peaks (10° < 2θ < 50°) of all the possible compounds, whereas the low-intensity peaks of the sample at higher angles might indicate the presence of orthorhombic K 2 ZrF 6 . The results therefore show that, in addition to the orthorhombic crystal structure that we have used in our previous work to model solid solutions of K 2 Zr (1− z ) Hf z F 6 to evaluate the role of the solid-state thermodynamics on the selectivity, occurrence of the monoclinic structures is also possible (Figure A). As shown in Figure B, occurrence of the cubic heptafluorides is possible at higher KF concentrations (sample peaks at 2θ values of 28, 44, and 50° correspond with that of K 3 ZrF 7 and K 3 HfF 7 ), which is probably the reason for the significant change in the appearance of the crystals shown in Figure D.…”

“…Due to the chemical similarities of Zr and Hf, K 2 ZrF 6 , and K 2 HfF 6 , like other crystalline solids of these two elements, have isomorphic crystal structures which can lead to solid solution formation during crystallization from a solution that contains both entities, causing poor selectivity, as has been observed for solution crystallization of other isomorphic compounds. − In our previous work, we have shown, by using molecular modeling, that solid solution formation by lattice substitution of Hf(IV) into the crystal structure of K 2 ZrF 6 seems to be thermodynamically feasible. Thus, solid solution formation might be the cause of the relatively low selectivity that was reported in the past for fractional crystallization.…”