Characterization of Thin Film Composite Reverse Osmosis Membranes Using Porosimetry and X-Ray Computed Tomography

Forward osmosis (FO) and pressure-retarded osmosis (PRO) are emerging processes for sustainable water and power production respectively. The flux performances in these processes depend greatly on the transport of solutes and solvents through, and hence structure of, the membrane used. Equations used to predict performance indicate the importance of membrane structural parameters, like thickness, porosity and tortuosity in flux performance. It has been postulated that, in these osmotically-driven processes, a phenomenon called internal concentration polarization (ICP) is the most important factor limiting performance. Severity of ICP is directly related to, and influenced by, the membrane’s structural parameters. So far, all FO and PRO studies have calculated S by fitting it to a model equation using experimental flux values and it has never been independently calculated. This approach could result in incorrect estimates of the flux but due to the inherent difficulties in characterizing the structural metrics, particularly, the porosity and tortuosity, it has never been attempted so far. In this study, the structural parameters have been independently calculated using analytical and imaging techniques. Two commercially available thin film composite reverse osmosis membranes, BW30 and SW30-XLE from Dow Water and Process Solutions have been used as model membranes in this study. Analytical measurements were done using mercury intrusion porosimetry from which porosity and pore diameter distributions were obtained. A novel 3-D imaging technique, x-ray computed tomography, was used to obtain porosity, tortuosity and also porosity distribution as a function of distance through the membrane. The structural parameter was then independently calculated, for the first time, and the estimates were compared to that from experimental flux measurements. Analysis of the two sets of values showed that the structural parameter is more complex than currently assumed and that deconvoluting it is essential in order to understand its effect on FO and PRO performance.