Prediction of Vapor Recovery from Light Gas with Mesoporous Membranes Using Dynamic Mean Field Theory

Mesoporous membranes have great potential for the separation of condensable species such as in the recovery of volatile organic compounds from air or the removal of carbon dioxide from flue gas. Selectivity and permeance of membranes in these processes are influenced by mechanisms like preferential adsorption, surface flow, and capillary condensation, but predictive correlations relating these mechanisms to mesopore properties such as geometry and surface chemistry are absent. Atomistic simulations provide a pathway to such correlations but are computationally intensive. Dynamic mean field theory (DMFT), a lattice-based density functional theory that successfully describes adsorption equilibrium and uptake dynamics in mesoporous materials, is a less intensive approach that provides qualitative and semi-quantitative insight. We have recently adapted DMFT to study non-equilibrium steady-state transport in mesopores and published a computational analysis of a permporometry experiment, in which a light gas permeates through a membrane in the presence of an essentially static condensable vapor. In this work, we use DMFT to study the recovery of a condensable vapor from a light gas, a situation in which the vapor is under a substantial partial pressure gradient. Model parameters are chosen to be relevant to experimental work in the literature on recovery of volatile organic compounds (VOCs) from air. Even in a simple slit pore, the results show interesting new phenomena like capillary condensation that is confined to the part of the pore closest to high pressure (feed side) of the membrane. We explore how modifying the geometry of the pore influences the selectivity and permeance, which will be useful to researchers designing mesoporous membrane materials to perform such separations.