Supported Molten Metal Membranes for Hydrogen Separation

Abstract

We describe here our results on the feasibility of a novel dense metal membrane for hydrogen separation: Supported Molten Metal Membrane, or SMMM.1 The goal in this work was to develop these new membranes based on supporting thin films of low-melting, non- precious group metals, e.g., tin (Sn), indium (In), gallium (Ga), or their alloys, to provide a flux and selectivity of hydrogen that rivals the conventional but substantially more expensive palladium (Pd) or Pd alloy membranes, which are susceptible to poisoning by the many species in the coal-derived syngas, and further possess inadequate stability and limited operating temperature range. The novelty of the technology presented numerous challenges during the course of this project, however, mainly in the selection of appropriate supports, and in the fabrication of a stable membrane. While the wetting instability of the SMMM remains an issue, we did develop an adequate understanding of the interaction between molten metal films with porous supports that we were able to find appropriate supports. Thus, our preliminary results indicate that the Ga/SiC SMMM at 550 oC has a permeance that is an order of magnitude higher than that of Pd, and exceeds the 2015 DOE target. To make practical SMM membranes, however, further improving the stability of the molten metal membrane is the next goal. For this, it is important to better understand the change in molten metal surface tension and contact angle as a function of temperature and gas-phase composition. A thermodynamic theory was, thus, developed, that is not only able to explain this change in the liquid-gas surface tension, but also the change in the solid-liquid surface tension as well as the contact angle. This fundamental understanding has allowed us to determine design characteristics to maintain stability in the face of changing gas composition. These designs are being developed. For further progress, it is also important to understand the nature of solution and permeation process in these molten metal membranes. For this, a comprehensive microkinetic model was developed for hydrogen permeation in dense metal membranes, and tested against data for Pd membrane over a broad range of temperatures.3 It is planned to obtain theoretical and experimental estimates of the parameters to corroborate the model against mental results for SMMM.