Molten salt is one of the widely used liquid media for a selective extraction through electrolysis. In the nuclear industry, pyroprocessing which utilizes LiCl-KCl molten salt in its core unit processes is suggested as a candidate of non-proliferative fuel recycling process. In an electrorefiner, the uranium metal is selectively collected on a solid electrode through the electrochemical reduction of uranium ion dissolved in LiCl-KCl. The deposit shows a dendritic characteristic which dramatically increases an electroactive surface area of the working electrode. As the surface area increases, current density decreases, and the overall electrochemical condition including potential gradient and exchange current along the diffusion layer is, therefore, changed as well in a real time. Despite of its critical influence on the precision of process simulation, there is no model which reflects this transient change of the electrode. In this work, cerium is selected as a surrogate of uranium. Its behaviors of electrodeposition at various conditions (time, current, concentration, and overpotential) are investigated, and a methodology for quantitative prediction of deposit formation on the working electrode is suggested.

All sample preparation and experiments were carried out in a glove box filled with high purity argon gas (99.9999 % Ar, [H₂O] and [O₂] < 1ppm). Oxidation or hydrolysis was completely avoided in advance. 44 wt.% of LiCl and 56 wt.% of KCl (both 99 % purity, anhydrous, Sigma-Aldrich) were mixed for the preparation of LiCl-KCl eutectic. This eutectic and CeCl₃ (99.99 % purity, anhydrous, Sigma-Aldrich) were put together into an alumina crucible. This sample was placed in a high temperature furnace which sustained the temperature 773 ± 2 K. Two 1 mm diameter tungsten rods were used as working and counter electrodes, and a silver wire was immersed in 1 wt.% AgCl-LiCl-KCl as the reference electrode. All electrochemical experiments were performed with a potentiostat (Autolab, PGSTAT302N).

Prior to the electrodeposition, cyclic voltammetry (CV) was measured to obtain redox peak of cerium and to check the existence of impurity in a system. Then chronoamperometry (CA) was carried out at a fixed potential of -2.05 V (vs. 1 wt.% Ag/AgCl), which is slightly more negative value than that of the reduction peak for successful electrodeposition. As shown in Figure 1, CA plots showed three different characteristics as the concentration of Ce(III) increased. At low concentration (< 0.45 wt.%, black plots), almost constant or gradual increase of current flow was observed. On the other hand, at high concentration (0.48 wt.% >, red plots), the magnitude of current flow increased drastically with illustrating clear linear slope. Therefore, it can be interpreted that, under the same condition of applied potential, the larger the concentration, the greater the current flow due to more electroactive species around the electrode, and thus the larger the growth of the surface area with the larger amount of electrodeposition. At the concentration between 0.45 wt.% and 0.48 wt.% of Ce(III), the CA plot expressed an intermediate behavior which can be shifted from one to the other characteristic at any time depending on the formation of deposit. Even though the CA plots showed different characteristics based on the concentration parameter, they all sketched linear trend. The amount of charge transferred which corresponded to the amount of metal deposit was able to be expressed as a function of time with slope and y-intercept. The slopes of CA plots at high concentration condition were directly proportional to the concentration. The y-intercept, the first plateau of CA plot, could be defined from the current of reduction peak of CV, which is proportional to the concentration. The current signal of reduction peak was affected by immersed depth of working electrode and overpotential. In conclusion, under assumption of no physical detachment or loss of electrodeposit, the preliminary prediction of the amount of electrodeposit was possible with three parameters (time, concentration, and overpotential) which were substantially related to the determination of induced current during CA experiment. This prediction should be achieved in a scrupulous way with consideration of concentration parameter because of different behaviors of electrodeposition.