Jarrod Milshtein

May 2013
Yttria-Stabilized Zirconia Membrane Stability In Fluoride Melts For The Magnesium SOM Process
Committee Members: Advisor: Professor Uday. B. Pal, MSE/ME; Professor Soumendra N. Basu, MSE/ME; Professor Srikanth Gopalan, MSE/ME

Abstract: One proposed industry method for the direct electrolysis of magnesium oxide for magnesium production is the Solid Oxide Membrane (SOM) process. The SOM process offers an energy efficient and low-carbon magnesium production alternative when compared to other methods of primary magnesium production. During the SOM process, MgO is dissolved in a molten CaF2-MgF2 flux. An yttria-stabilized zirconia (YSZ) membrane is submerged in the flux, and this membrane acts as an oxygen anion conducting SOM tube. The YSZ membrane separates the cathode and flux from the anode. When an electric potential is applied across the electrolysis cell, magnesium cations travel through the flux and are reduced at a stainless steel cathode. Oxygen anions simultaneously move through the YSZ membrane to a liquid silver anode, where the anions are oxidized.

The SOM process has been demonstrated successfully on the laboratory scale, but in order for the SOM process to be commercially viable, electrolysis cells must operate for thousands of hours. The stability of the YSZ membrane limits the operating life of the SOM electrolysis cell. This thesis determines YSZ membrane stability in oxy-fluoride fluxes for the SOM process so that membrane degradation can be better understood and controlled. One primary degradation pathway of YSZ in the SOM process has been determined to be yttria depletion out of the YSZ membrane. Yttria concentration profiles in YSZ membranes were determined using x-ray spectroscopy, and the concentration profiles were used to analyze the depletion process. The yttria depletion mechanism was determined to be chemical diffusion, and the diffusion process was modeled. A method of controlling the yttria depletion process by adding small concentrations of YF3 to

the flux is described, modeled, and experimentally proven. An optimal range of YF3 concentrations to add to the flux is determined for increasing YSZ membrane stability.

This study investigated the role of flux impurities on YSZ membrane stability. The effect of impurities on YSZ membrane stability had not been studied or described before this work. Impurities tested are common to magnesium ores: calcia, silica, sodium oxide, and sodium peroxide. Any degradation effects due to these impurities were analyzed, and methods to remove the negative effects of impurities were described when possible.