This work contributes to the development of pyroprocessing technology as an economically viable means of separating used nuclear fuel from fission products and cladding materials. Electrolytic oxide reduction is used as a head-end step before electrorefining to reduce oxide fuel to metallic form. The electrolytic medium used in this technique is molten LiCl-Li 2 O. Groups I and II fission products, such as cesium (Cs) and strontium (Sr), have been shown to partition from the fuel into the molten LiCl-Li 2 O.Various approaches of separating these fission products from the salt have been investigated by different research groups. One promising approach is based on a layer crystallization method studied at the Korea Atomic Energy Research Institute (KAERI). Despite successful demonstration of this basic approach, there are questions that remain, especially concerning the development of economical and scalable operating parameters based on a comprehensive understanding of heat and mass transfer. This research explores these parameters through a series of experiments in which LiCl is purified, by concentrating CsCl in a liquid phase as purified LiCl is crystallized and removed via an argon-cooled cold finger. All experiments were conducted in an inert argon atmosphere. The experimental LiCl-CsCl operating temperature was 650˚C with a cold finger temperature held at the freezing temperature of LiCl (≈605˚C). Molten salt and cold finger exhaust temperatures were measured throughout experiments and argon cooling gas flow was controlled manually via a needle valve. The varied parameters of interest were initial contaminant concentration, cold finger cooling rate, and separation time. CsCl concentration was measured via Inductively Coupled Plasma Mass Spectroscopy (ICP-MS) for purified LiCl crystals and concentrated LiCl-CsCl bulk.Preliminary results revealed that co-deposition of CsCl along with LiCl crystal formation due to slightly non-equilibrium mass transport is a problem that must be minimized. Analysis of experiments with initial 5 wt% CsCl concentration showed that for an experimental matrix of varying cold finger coolant flow rates (7.4, 9.8, 12.3, and 14.9 L/min) and separation times (5, 10, 15, and 30 min), the optimal cooling flow rate and separation time are 14.9 L/min and 15 min, respectively, producing a 0.33 wt% CsCl crystal purity at a production rate of 0.36 g/min. Experimental results for initial concentrations of 1, 3, and 7.5 wt% CsCl at the aforementioned coolant flow rates for separation times of 15 min were analyzed and showed iv that as initial bulk concentration increased from 1 to 7.5 wt% CsCl, the change in purified crystal concentration for experiments was less than 0.5 wt% CsCl for all but one data point.The optimal result of the initial 5 wt% CsCl experiments along with results of initial 1, 3, and 7.5 wt% CsCl bulk concentration experiments were used to predict a scale-up scenario that provided a potential separation rate and purity of 136 g/hr and 0.24 wt% CsCl, respectively, to provide in...