Background: The presence of water in organic solvents and biofuels can complicate their production and reuse because many hydrophilic solvents form difficult-to-separate mixtures with water (e.g., azeotropes). Pervaporation (PV) and vapor permeation (V⋅P) remove water from such mixtures via selective solution-diffusion transport through a membrane material. A recent article reviewed design factors that impact the effectiveness of PV/V⋅P solvent dehydration processes (J. Chem. Technol. Biotechnol 95: 495-512 (2020)). For the sake of simplicity, the earlier work assumed constant membrane permeabilities. The objective here is to explore the impact of variable permeabilities on predictions of PV/V⋅P system performance.
Results: A multiparameter expression relating permeability to process conditions was incorporated into the spreadsheet calculators from the previous work. Use of the expression was demonstrated with literature ethanol/water PV data for a NaA zeolite material and two poly (vinyl alcohol) (PVA) membranes. The variable permeabilities of the membranes yielded membrane area requirements that were 20-30% different from those calculated using permeances fixed at either end of the target water range. The impact of composition-dependent permeabilities was most pronounced on the fraction of ethanol transferred to the permeate for the NaA membrane.
Conclusion: The inclusion of membrane permeabilities that vary with fluid composition and temperature noticeably altered predictions of the membrane area required to carry out water removal from ethanol by PV and of the transfer of ethanol to the permeate stream. Unless a PV/V⋅P process is expected to operate at a constant temperature and in a narrow concentration range, process performance estimates would be improved by inclusion of concentration- and temperature-dependent permeabilities or permeances.
Keywords: dehydration; membrane permeability; membrane processes; membrane-based separation; pervaporation; solvent reclamation; vapor permeation.