(421f) Fundamental Origin of Flux Non-Linearity in Organic Solvent Reverse Osmosis and Organic Solvent Nanofiltration

Organic solvent reverse osmosis (OSRO) and nanofiltration (OSN) are attracting interest due to compactness, modularity and superior energy efficiency. However, while recent research efforts mostly focused on the synthesis and fabrication of solvent resistant composite polymer membranes, fundamental understanding of chemical and physical aspects that govern solvent and solute transport in OSRO-OSN membranes remains entirely unexplored. The transport mechanism itself is poorly understood: some researchers have hypothesized a solution-diffusion mechanism, others a pore-flow mechanism. Finally, others have considered a combination of the previous two mechanisms. A peculiar OSRO-OSN feature is the flux non-linearity vs. Dp, that is, a negative departure of flux from linearity is observed starting from Dp = 10 atm, which has been the subject of a long standing debate in the literature. Despite membrane compaction has been invoked to explain this phenomenon, this hypothesis has no quantitative support.

In this study, we critically discuss the hypothesis of membrane compaction. To demonstrate that the molecular origin of flux non-linearity is purely thermodynamic, we propose, for the first time, a thermodynamic-diffusion framework which describes solvent transport in OSRO-OSN membranes in terms of the concentration gradient produced by the applied pressure across the membrane. The model predictions agree very well with experimental data. Solvent diffusion coefficient in the membrane increases with increasing Dp, which further confirms that flux decline is not related to membrane compaction. The developed framework allows to quantify both frame of reference and non-ideal thermodynamic effects on solvent diffusion coefficients in OSN membranes.

This study demonstrates that the solution-diffusion model provides a satisfactory description of small molecule transport in OSN membranes, without the need to resort to pore-flow or more complicated transport models.