In order to develop a flow sheet for the purification procedure of a LiCl-KCl eutectic mixture with the underlying aim of avoiding Cl2 handling in an engineering-scale laboratory, the present work focuses on electrochemical investigations on vacuum dried eutectic mixtures LiCl-KCl + H2O and vacuum dried LiCl-KCl + H2O mixtures using cyclic voltammetry at 500°C. Complicated voltammetric features were observed that were attributed to the cathodic reduction of H2O to form OH, adsorption of OH on tungsten electrode, and cathodic reduction of OH to form O2−. The onset of cathodic current around −0.45 V (versus AgAg as reference) was due to both cathodic reductions of H2O and HCl, the latter being formed in melt due to high-temperature hydrolysis of LiCl, although it had limited solubility as compared to that of H2O. Cyclic voltammograms also showed an anodic peak at around −0.30 V attributable to the adsorption of O2− on the tungsten electrode.

A total of 19 different LiCl-KCl eutectic melts subjected to various vacuum drying conditions and moisture content were investigated in this work using cyclic voltammetry. The LiCl-KCl + H2O mixtures were prepared to simulate conditions when there is an ingress of moisture in LiCl-KCl mixtures during long storage to determine whether the mixture can be purified. A larger composition range of moisture in the LiCl-KCl eutectic mixtures was used that not only helped in the attribution of cathodic reduction peaks to reactions mentioned above but also served as references to investigate the influence of vacuum drying of moisture-containing eutectic mixtures.

A two-point criterion consisting of cathodic onset LiLi potential and residual cathodic current density estimated from cyclic voltammograms in the potential region −1.5 to −2.0 V at 500°C was used to quantify the purity of the eutectic melts. Former data were compared with the theoretical equilibrium potential of −3.637 V for the LiLi couple in LiCl-KCl eutectic melt at 500°C and those obtained from cyclic voltammograms of chlorinated melt. Cathodic reduction potentials for the above reactions were then compared with literature data where they were measured against Li-Al or NiNi reference electrodes in LiCl-KCl melts. Although reduction of HCl at Pt electrode in LiCl-KCl eutectic melts was known to be reversible from literature, it was not found to be so in the present work where a tungsten working electrode was used. The LiCl-KCl eutectic mixtures vacuum dried at 300°C were found to be closer in purity to those of chlorinated melt in that the onset LiLi potential and residual cathodic current density were similar. A lower residual cathodic current density for LiCl-KCl + 2 wt% H2O vacuum dried at 300°C was also achieved with the onset LiLi potential quite close to the theoretical equilibrium potential of LiCl (−3.637 V) in the eutectic melt.