Jerry H. Meldon, Tufts University, Chemical and Biological Engineering Department, Medford, MA 02155
Because most metals are permeable only to hydrogen, there has long been great interest in their deployment in membrane-based H2 separation processes. Palladium and its alloys have drawn the lion's share of attention because they offer reasonably high fluxes. Considerable R&D effort has been directed at supporting thin Pd-based layers on microropous materials including ceramics and dense materials like niobium and tantalum that are cheaper than Pd but form impervious surface oxide layers when unprotected. Judicious design and interpretation of the permeation measurements undertaken to characterize these materials is facilitated by a mathematical model of the hydrogen transport process, whose complex sequence of steps includes gas-phase mass transfer, reversible dissociative chemisorption, and interstitial diffusion. Determination of relevant rate parameters is accelerated by conducting transient permeation measurements – e.g., driven by step or sinusoidal variations in pressure. For deployment in conjunction with such experiments, we have developed a model of nonsteady-state H2 transport in thin Pd-based membranes and composites. The model is particularly convenient for deconvoluting data obtained when the step or sinusoidal perturbations are small – because it is then possible to linearize the governing ordinary and partial differential equations, solve them via Laplace transformation and, in some cases, obtain closed-form expressions for the time-dependence of the H2 flux in the immediate aftermath of the perturbation.