Nanoporous, Metal Carbide, Surface Diffusion Membranes for High Temperature Hydrogen Separations

Abstract

Colorado School of Mines (CSM) developed high temperature, hydrogen permeable membranes that contain no platinum group metals with the goal of separating hydrogen from gas mixtures representative of gasification of carbon feedstocks such as coal or biomass in order to meet DOE NETL 2015 hydrogen membrane performance targets. We employed a dual synthesis strategy centered on transition metal carbides. In the first approach, novel, high temperature, surface diffusion membranes based on nanoporous Mo 2 C were fabricated on ceramic supports. These were produced in a two step process that consisted of molybdenum oxide deposition followed by thermal carburization. Our best Mo 2 C surface diffusion membrane achieved a pure hydrogen flux of 367 SCFH/ft 2 at a feed pressure of only 20 psig. The highest H 2 /N 2 selectivity obtained with this approach was 4.9. A transport model using "dusty gas" theory was derived to describe the hydrogen transport in the Mo 2 C coated, surface diffusion membranes. The second class of membranes developed were dense metal foils of BCC metals such as vanadium coated with thin (< 60 nm) Mo 2 C catalyst layers. We have fabricated a Mo 2 C/V composite membrane that in pure gas testing delivered a H 2 flux of 238 SCFH/ft 2 at 600 °C and 100 psig, with no detectable He permeance. This exceeds the 2010 DOE Target flux. This flux is 2.8 times that of pure Pd at the same membrane thickness and test conditions and over 79% of the 2015 flux target. In mixed gas testing we achieved a permeate purity of ≥99.99%, satisfying the permeate purity milestone, but the hydrogen permeance was low, ~0.2 SCFH/ft 2 .psi. However, during testing of a Mo 2 C coated Pd alloy membrane with DOE 1 feed gas mixture a hydrogen permeance of >2 SCFH/ft 2 .psi was obtained which was stable during the entire test, meeting the permeance associated with the 2010 DOE target flux. Lastly, the Mo 2 C/V composite membranes were shown to be stable for at least 168 hours = one week, including cycling at high temperature and alternating He/H 2 exposure.4

Application of TiC in Vanadium-Based Hydrogen Membranes

“…An important note was the appearance of boundary layer effects at the membrane interface indicated by a decrease in hydrogen permeance in H 2 /He mixtures. Computational fluid dynamic (CFD) simulations revealed the boundary layer may be a result of module design . However, the He mixtures provided a metric of comparison for hydrogen permeation for other gaseous mixtures.…”

“…Computational fluid dynamic (CFD) simulations revealed the boundary layer may be a result of module design. 44 However, the He mixtures provided a metric of comparison for hydrogen permeation for other gaseous mixtures. Exposure to N 2 terminated H 2 flux through the membrane.…”