(366a) Modeling Gas and Vapor Transport Mechanisms in Mesoporous Membranes Using Dynamic Mean Field Theory

Authors: 
Rathi, A., University of Massachusetts Amherst
Ford, D., University of Massachusetts Amherst
Monson, P. A., University of Massachusetts Amherst
Mesoporous membranes with technologically relevant pore sizes provide an energy efficient means for separations such as CO2 from effluents and volatile organic compounds from air. Continuous progress in material synthesis over the years has led to techniques which allow reasonable fine tuning of geometry and surface chemistry of mesopores within the membranes. However, absence of design rules linking mesopore properties with confinement and flow mechanisms which determine permeation and separation efficiency have proven a hindrance towards designing better materials. Towards development of such design rules through the study of underlying transport processes, we have investigated the problem of permeation of pure gas and condensable vapor as well as their mixtures under pressure gradients, through the application of dynamic mean field theory (DMFT), a lattice-based theory. We initially modeled permporometry, an experimental technique where light gas permeates in presence of a condensable vapor in the pore. Interestingly, the maximum flux of light gas was found adjacent to the strongly adsorbed surface layer. We then studied separation of a condensable vapor from its mixture with light gas under significant pressure gradient. We discovered capillary condensation confined to the high pressure (feed side) of the pore. Experimentally relevant model parameters were used to investigate effects of such a localization on separation efficiency. Effects of modification in geometry and orientation were studied through ink-bottle like structures while effects of surface chemistry were explored through modification in interactions. To gain a perspective on the results, experimental studies from the literature, and computationally expensive dual control volume â?? grand canonical molecular dynamics (DCV-GCMD) atomistic simulations, were used for comparison.