Prospects and challenges of iron pyroelectrolysis in magnesium aluminosilicate melts near minimum liquidus temperature

Abstract

Although steel production by molten oxide electrolysis offers potential economic and environmental advantages over classic extractive metallurgy, its feasibility is far from being convincingly demonstrated, mainly due to inherent experimental difficulties exerted by harsh conditions and lack of knowledge regarding relevant mechanisms and physico-chemical processes in the melts. The present work was intended to demonstrate the concept of pyroelectrolysis at very high temperature near the minimum liquidus point of magnesium aluminosilicate, being conducted under electron-blocking conditions using yttria-stabilized zirconia cells, and to provide a new insight into electrochemistry behind this process. Significant current yields are possible for pyroelectrolysis performed in electron-blocking mode using a solid electrolyte membrane to separate the anode and the molten electrolyte. Parasitic electrochemical processes rise gradually as the concentration of iron oxide dissolved in the molten electrolytes is depleted, impairing faradaic efficiency. Reduction of silica to metallic silicon was identified as a significant contribution to those parasitic currents, among other plausible processes. Direct pyroelectrolysis without electron blocking was found much less plausible, due to major limitations on faradaic efficiency imposed by electronic leakage and insufficient ionic conductivity of the aluminosilicate melt. Ohmic losses may consume an excessive fraction of the applied voltage, thus failing to sustain the Nernst potential required for reduction to metallic iron. The results suggest the need for further optimization of the molten electrolyte composition to promote ionic conductivity and to suppress electronic transport contribution, possibly, by tuning the Al/Si ratio and altering the network-forming/modifying behaviour of the iron cations.