Solid oxide membrane process for the direct reduction of magnesium from magnesium oxide.

摘要

The Solid Oxide Membrane (SOM) process is an emerging generic technology for the environmentally friendly extraction of high-energy-content metals directly from their oxides. This process has the potential to offer a viable, cost effective and cleaner alternative to existing state of the art primary magnesium extraction processes.; The SOM process in principle uses a tubular yttria stabilized-zirconia-based solid oxide fuel cell with liquid metal (copper or tin) as an anode in the temperature range of 1100--1300°C. Magnesium oxide is dissolved in a molten ionic flux and oxygen ions are pumped out of the flux through the zirconia membrane and are oxidized at, the liquid metal anode. Magnesium vapor evolves at the cathode and is condensed in a separate chamber (condenser).; The proof of concept for the SOM process was initially demonstrated at 1300°C using a magnesium fluoride-based flux. Since the membrane is the most expensive part of the process, its long-term stability is critical to the scale up and eventual commercialization of the process. Temperature, flux chemistry and cell operating conditions have been identified as key process parameters for membrane stability.; A new low temperature flux based on the eutectic: magnesium fluoride-calcium fluoride system, has been developed which has lowered the operating temperature of the SOM cell to 1150°C. Additionally, a minor addition of yttrium fluoride to the flux minimized yttria diffusion from the yttria-stabilized-zirconia membrane, thereby further enhancing membrane stability. Important thermo-physical properties of the selected flux compositions critical to the process such as viscosity, density, volatility, solubility and electrical conductivity have been measured. The SOM cell has been electrochemically characterized and concepts related to MgO dissociation voltage, observed leakage current and mass transfer in the SOM cell are explained. The viability of the SOM process has been demonstrated by the operation of a 100 g scale reactor. Preliminary cost estimates and design of a commercial reactor based on this investigation are also discussed.