Thermodynamic Screening of Solvents for Ternary Cellulose Acetate Blend Membranes Using Hansen Solubility Parameters

Abstract

Solvent selection for phase inversion membranes still relies mostly on trial and error, even though the solvent governs how well a polymer blend forms into a stable, separation-capable structure. Few studies quantify that relationship before casting. This study calculates Hansen Solubility Parameters (HSP) for a ternary blend of Cellulose Acetate (CA), Polyethylene Glycol (PEG), and Polyvinyl Chloride (PVC) against six candidate organic solvents, with water as the nonsolvent, to test whether thermodynamic distance can predict membrane morphology ahead of fabrication. We calculated the interaction distances for all six solvents, then cast and visually examined membranes from three of them, dimethylacetamide (DMAc), dimethylformamide (DMF), and N-methyl-2-pyrrolidone (NMP), to check the calculations against real morphology. DMAc gave the smallest polymer-solvent distance (Δp−s = 1.623), ahead of DMF (2.394) and NMP (4.228), pointing to the strongest calculated affinity with the polymer blend among the six candidates. This ranking was consistent with the macroscopic appearance of the membranes: DMAc produced a smooth, glossy surface without visible defects, DMF a uniform matte surface without visible cracks, and NMP a wrinkled and deformed surface. Because morphology was assessed only by macroscopic photography, this agreement is qualitative and does not confirm skin-layer structure or selectivity. Nevertheless, the results suggest that HSP-based Δp−s calculation has potential as a simple, non-empirical preliminary screening step for solvent selection ahead of experimental fabrication. Membranes prepared with low-Δp−s solvents such as DMAc are hypothesized to show better separation performance, and salt-rejection testing and cross-sectional microscopy are proposed as future work.