Molecular Simulations of Gas Solution–Diffusion Processes in Mixed Matrix Membranes Comprising a Polymer of Intrinsic Microporosity Blended with Silica

Abstract

Polymers of intrinsic microporosity (PIMs) have recently attracted much attention as gas separation membranes in terms of scalability, low energy cost, and environmental friendliness. In the present study, we investigate the solution–diffusion processes of CO₂, N₂, and CH₄ molecules in PIM-1 based membranes (PIM-1 and PIM-1/a-SiO₂) using molecular dynamics (MD) and grand canonical Monte Carlo (GCMC) simulations. Long-time diffusion is characterized via the mean squared displacements of gas molecules, indicating that it takes them over 50 ns to reach normal diffusion in the PIM-1 based membranes. Specifically, systems with unimodal pore size distributions (PSDs) are likely to cause the gas molecules to reach normal diffusion within the time window of 50 ns, while systems with multimodal PSDs lead to sub- or super-diffusion of the molecules in that time window. Compared with bulk PIM-1 systems, the solubility of gas molecules is increased in PIM-1/a-SiO₂ systems; in particular, CO₂ solubility is significantly increased. This is because the amount of CO₂ molecules adsorbed in the vicinity of a PIM-1/a-SiO₂ interface increases due to an energetic affinity of CO₂ with surface OH groups. The diffusivity of N₂ and CH₄ is increased in the PIM-1/a-SiO₂ systems due to large voids created by the presence of the interface. Meanwhile, CO₂ diffusivity decreases due to an energetic affinity with the OH groups.