A Novel Approach to Molecular Modeling of Transport through Inorganic Nanoporous Membranes

Introduction Numerous growth and permeation studies of nanoporous films have led to their attractiveness for application in separating gas mixtures with permselectivities approaching unprecedented subnanometer molecular resolution. The understanding and ability required for predicting macroscopic transport characteristics for diffusion of interacting molecular species through the complex lattices of these nanoporous inorganic membranes is key to the ultimate development of both traditional applications of this technology such as integrated reaction and separations devices, and more novel ones, such as substrates for growth of nanowires and chemical sensors. Molecular simulations, specifically molecular dynamics and Monte Carlo algorithms, have emerged in the past few decades as preeminent computational tools for science and engineering research. Despite their widespread use and unprecedented insights into microscopic dynamics, such molecular simulations are computationally limited to short length and time scales, while inorganic membranes, such as zeolite films, invoke much larger scales (e.g., [1]) by way of macroscopic heterogeneities and imposed gradients. On the opposite end of the modeling spectrum, continuum, Maxwell-Stefan models have been proposed in attempts to describe experimentally observed diffusion behavior. While these models have proven beneficial in fitting single and multi-component diffusion data, their phenomenological derivation does not capture the underlying microscopic diffusion dynamics and restricts their applicability to unrealistically simple diffusion lattices. Consequently, their accuracy as a priori predictive models is severely limited. A major obstacle in addressing this multiscale modeling challenge is the lack of a rigorous mathematical and computational framework directly linking atomistic simulations and scales to complex mesoscopic and macroscopic phenomena dictated by microscopic intermolecular forces and material microstructure. A new mathematical framework has been introduced and validated through comparison with gradient Monte Carlo simulations, for modeling diffusion of interacting species in nanoporous materials over large length scales while retaining molecular scale information typically captured only by molecular simulations 2, .