Determining Binary Diffusion Coefficients for Mixtures in Zeolites from PFG NMR, MD Simulation, and Theory

1. Introduction Practical applications of zeolites and other nanoporous materials always involve at least two different components, but understanding of multicomponent diffusion in zeolites is limited and experimental data are scarce. The elements of the multicomponent Fickian diffusion tensor for a given system may depend on the total loading of adsorbed molecules, the adsorbed-phase composition, and the temperature. Given the difficulties in measuring multicomponent diffusion coefficients, it would be helpful to have predictive models to calculate the multicomponent diffusion coefficients from single-component data. Several models have been proposed for this purpose. However, given the lack of experimental data, it has been hard to test how widely applicable they are. In this work, we have used atomistic molecular dynamics (MD) simulations to predict binary diffusion coefficients. The results have been used to test the model of Krishna and co-workers [1] for predicting multicomponent behaviour from single-component data. In addition, we report pulsed field gradient (PFG) NMR self-diffusivities for a systematic set of binary mixtures in NaX zeolite.