(373j) Polymeric Membranes in silico: Perspectives on Binary Separations and Sorbate Swelling

Microporous polymers are a steadily expanding class of adsorbent materials with promising applications for chemical storage, separation, and catalysis. In particular, polymers of intrinsic microporosity (PIMs) are an attractive family of adsorbent polymers that exhibit remarkable permeability while maintaining relatively high species selectivity. This work demonstrates the use of molecular simulations as an approach to explore the material design/performance space of PIMs for chemical separations of multi-component mixtures, which is essential to designing membrane materials for “real” applications where gas feeds contain numerous distinct components and impurities. In silico characterization tools are applied to study the amorphous PIM structures at an atomistic level of detail that would otherwise be difficult to obtain experimentally. Furthermore, to predict adsorption properties a combined Monte Carlo and molecular dynamics (MC/MD) simulation approach is used, which is an improved simulation technique that accounts for adsorbate-induced polymer chain mobility.

Through implementation of these approaches, this work seeks to provide detailed insight into three cases: 1) The evolution of model PIMs porous structure throughout the adsorption isotherm 2) simulated predictions of binary adsorption 3) the effect of chemical modification on membrane properties. In the first case the expansion of fractional free volume, broadening of the pore size distribution, and variation in accessible surface area is explored as a function of adsorbate single-component and mixture diversity. Regarding the second case, a matrix of diverse binary mixture selectivity values are produced alongside an array of 2-component isotherms that are used to describe the preferential adsorbate-adsorbent interactions with the PIM frameworks. Finally, in the interest of materials discovery, an SO₂ modified variant of PIM-1 is studied and its effect on enhancing adsorption selectivity, plasticization resistance, and porous character is examined. Overall, the work reported highlights the progress we have made towards understanding multi-component adsorption physics at an atomistic level of detail via a combination of MC/MD and in silico microporous material characterization. By directly simulating multi-component adsorption in a computational workflow that accounts for membrane swelling/plasticization we demonstrate that molecular simulations can be exploited to conceive new membrane materials with tailored uptake, selectivity, and/or swelling resistance for a specific diversity of adsorbate mixtures.