A model for the separation of complex liquid mixtures with glassy polymer membranes: A thermodynamic perspective

Abstract

In this work, a new model is developed to predict the separation of organic solvent mixtures with glassy polymer membranes. Particular attention is paid to thermodynamic contributions such as guest solubility and guest-induced polymer swelling, and how these contributions affect the flux in multi-component mixtures. The thermodynamic contributions are complicated by the fact that the membranes are glassy polymers. The thermodynamic aspects of the diffusion problem are calculated using DGRPT (dry glass reference perturbation theory), which extends the non-equilibrium thermodynamics of glassy polymers (NETGP) such that polymer swelling is calculated self-consistently in the theory. The diffusion model for pure fluids diffusing through the membrane is obtained through Maxwell-Stefan equations where the polymer membrane is stationary. The mixture diffusion model is then constructed using a simple diffusivity averaging scheme on this pure fluid result. The model is then shown to accurately predict binary and complex mixture separations in SBAD-1, STARMEM-240, and Torlon membranes. We highlight the ability of the model accurately predict membrane based fractionation of crude oil, using only the sorption isotherms of toluene and octane as input to the model. • A new membrane model was proposed for multicomponent liquid mixtures in glassy polymer membranes. • The role of pressure in liquid phase reverse osmosis separations is discussed. • The model correctly predicts the separation of crude oil using only two experimental sorption isotherms.