Electrochemistry of Molten Chlorides

Electrochemical processes in chloride melts are important because of their use for the industrial production of a series of active metals (Magnesium Production. M. Faraday, 1833, R. Bunsen, 1852; Sodium Production, C. Downs, 1924, Calcium Production, WWII USA, USSR). An electrochemical process in a chloride melt environment is simultaneously extremely difficult and possesses some unquestionable advantages, which makes it a real challenge for an explorer. On the one hand, the process Downs apparatushas to be performed at high temperature (typically, ~ 600-800 ºC) and in a very aggressive environment (Cl2, HCl, liquid alkaline metals), while both the reactants and the products are extremely sensitive to humidity and oxygen and no standard electrodes (like Ag|AgCl, SCE or NHE for aqueous systems) are known. On the other hand, the melt has high electric conductivity, the interaction of reactants and products with solvents is poor (and therefore it is typically easy to separate the products from the solvent after the completion of  the process), electrochemical windows are wide and potential differences for many processes are often larger than for aqueous systems.

Metal production in molten chloride requires sophisticated apparatuses be used to eliminate the risk of secondary reactions between products. For instance, sodium production by Dawns method required the physical separation of both the final products (sodium and chlorine) and the melts in the vicinity of each electrode.

Another application of chloride melt electrochemistry makes use of the larger differences between reduction potentials for some chemically similar metals. For instance, while in an aqueous system the difference between reduction potentials of lead and tin is only about 10 mV, which makes the separation practically impossibly, in molten KCl this difference is ~ 40 mV, which is completely achievable for the modern technologies. This feature is especially useful for the separation of lanthanides and actinides, which comprise large families of chemically very similar metals. For that reason molten chlorides have been widely used for the separation and enrichment of rare-earth and radioactive ores and for the treatment of nuclear wastes.


Metal Deposition from Chloride Melts

Most if not all applied electrochemical processes in chloride melts are those of metal ion reduction and deposition on liquid or solid electrodes. Such processes are successfully studied by various potentiodynamic techniques, primarily by Cyclic Voltammetry. A typical cyclic voltammogram for such process, namely iron(II) ion reduction on tungsten electrodes (image from Lugovskoy A., Zinigrad M., Aurbach D. Electrochemical Determination of Diffusion Coefficients of Iron (II) Ions in Chloride Melts at 700-750oC, Israel Journal of Chemistry, 2007, 47 (3-4), pp. 409-414) demonstrates a relatively simple mechanism of the quasireversible reduction of Fe2+ ions (peak C at ca. - 0.260 V vs. W electrode).


The process is almost ideally reversible (only moderate shift of peak potentials is observed as the scan-rate (velocities of potential change in the experiment) grows, because the tungsten electrode practically does not react with the iron deposit. This is readily confirmed by the microscopic study of iron deposits:

Well-shaped micronic hexagonal crystals are not attracted by the tungsten surface.

This is not the case for other systems. For instance, if platinum electrodes are taken for the same system instead of tungsten, the picture becomes completely different (A. Lugovskoy; M. Zinigrad; D. Aurbach; Z. Unger: Electrodeposition of iron(II) on platinum in chloride melts at 700-750 degrees C, Electrochimica Acta,  6,  54,  1904-1908,  2009). As the scan-rate grows, the reduction potential (Peak C) is shifted leftwards and the oxidation potential (Peak A) becomes more anodic. A plausible explanation is the formation of Pt-Fe alloy on the surface of the platinum cathode: iron is "caught" by the alloy and this causes the deviation from the reversibility. This explanation was confirmed by the analysis of iron deposits on the surface of platinum electrodes.

As seen from the scanning electron microscope image, iron nanoplatelets (1) "grow" on the surface of the electrode (3). EDS analysis shows that the surface contents of platinum and iron are almost equal, 60:40 atomic percents, respectfully.

These results illustrate a way to several applications of a mechanistic electrodeposition study. It is obvious, for instance, that iron and iron-platinum nanoplatelets can be readily synthesized by the deposition of iron ions on a platinum electrode from molten chlorides. On the other hand, ferromagnetic iron crystals are produced on tungsten. It deserves to be noted that these studies provide an interesting example of the paradox behavior of the two metals, when platinum, normally an inert metal, is more active than usually more reactive tungsten.