No abstract

An advanced synthesis protocol for the preparation of Thin-Film Nanocomposite (TFN) membranes is presented.

Thin film composite (TFC) polyamide membranes were prepared on a polysulfone support membrane and the effect of various synthesis conditions on the active layer morphology, the physicochemical properties and the membrane performance was investigated. The support membrane porosity factor had a significant effect on the TFC membrane performance. A polyamide top layer was formed within 15 s of reaction. Prolonging the reaction time, although resulting in a thicker active layer, only had a minor influence on the membrane performance. This highlights the importance of the incipient layer of the polyamide structure on its performance. The addition of both a surfactant and a base to the amine solution resulted in a change of the active layer morphology and an improved performance. The effect of additives was attributed to changes in the polymerization mechanism. In addition, it was demonstrated that curing at 50°C resulted in an improved membrane performance, due to more cross-linking of the active layer. Curing at higher temperatures deteriorated the structure of the support membrane. This research shows that the TFC membrane performance is well correlated with the changes in the active layer morphology, measured using SEM, AFM and TEM; whereas only minor changes in the physicochemical characteristics of the membranes were detected by zeta potential and ATR-FTIR spectroscopy when the same synthesis parameters were varied.

Membrane modification by concentration polarization (CP)-enhanced radical graft polymerization using a dilute aqueous solution of appropriate monomer was examined as a method for increasing rejection of boric acid by reverse osmosis (RO) membranes. On the basis of suggested physicochemical rationales a number of monomers were examined in order to determine those with the lowest affinity toward boric acid as compared to water. The improvement in the modified membrane performance was mainly attributed to sealing less selective areas ("defects") inherently present in the original low pressure RO (LPRO) membranes. However, the effect clearly differed for different monomers. Among the examined monomers glycidyl methacrylate (GMA) exhibited the lowest affinity and the largest improvement in removal of boric acid along with a moderate loss of permeability and slightly improved NaCl rejection. Modification of LPRO membrane thus resulted in a membrane with a permeability in the brackish water RO (BWRO) range but with removal of boric acid and salt superior to those reported for most commercial BWRO membranes.

Concentration polarization-enhanced radical graft polymerization, a facile surface modification technique, was examined as an approach to reduce bacterial deposition onto RO membranes and thus contribute to mitigation of biofouling. For this purpose an RO membrane ESPA-1 was surface-grafted with a zwitterionic and negatively and positively charged monomers. The low monomer concentrations and low degrees of grafting employed in modifications moderately reduced flux (by 20-40%) and did not affect salt rejection, yet produced substantial changes in surface chemistry, charge and hydrophilicity. The propensity to bacterial attachment of original and modified membranes was assessed using bacterial deposition tests carried out in a parallel plate flow setup using a fluorescent strain of Pseudomonas fluorescens. Compared to unmodified ESPA-1 the deposition (mass transfer) coefficient was significantly increased for modification with the positively charged monomer. On the other hand, a substantial reduction in bacterial deposition rates was observed for membranes modified with zwitterionic monomer and, still more, with very hydrophilic negatively charged monomers. This trend is well explained by the effects of surface charge (as measured by ζ-potential) and hydrophilicity (contact angle). It also well correlated with force distance measurements by AFM using surrogate spherical probes with a negative surface charge mimicking the bacterial surface. The positively charged surface showed a strong hysteresis with a large adhesion force, which was weaker for unmodified ESPA-1 and still weaker for zwitterionic surface, while negatively charged surface showed a long-range repulsion and negligible hysteresis. These results demonstrate the potential of using the proposed surface- modification approach for varying surface characteristics, charge and hydrophilicity, and thus minimizing bacterial deposition and potentially reducing propensity biofouling.