Impact of manufacturing conditions of polyethersulfone membranes on final characteristics and fouling reduction

The use of membrane processes in drinking water treatment has increased exponentially in the last few years. However, one of the shortcomings that hinders their wider application is the propensity of membranes to become fouled, causing the permeate flux to decrease with time. One of the main membrane foulants is the natural organic matter (NOM) present in water sources. Membrane characteristics and performance are affected by membrane preparation conditions. The objective of this thesis was to investigate the impact of hydrophobic surface modifying macromolecule (SMM) addition and other manufacturing conditions on the membrane characteristics and performance of polyethersulfone (PES) membranes. As it was necessary to test numerous membrane coupons prepared under many different conditions, during the first stage of this study a membrane test protocol was established. Small test cells gave very similar results to those produced by a larger test cell, which is recommended by USEPA for membrane testing. The test protocol was relatively short in duration and required a small amount of test solution. Moreover, as membrane compaction rate depends on the material, it was important to incorporate a precompaction step into the test protocol in order to quantify the contribution of membrane compaction when fouling was studied. The second stage of the study evaluated the impact of different membrane preparation conditions (i.e., solvent evaporation time and PES concentration) and the impact of three different SMM formulations. The three SMMs differed only in the polyol used in their synthesis. SMM41, the most miscible SMM, proved to be the most promising SMM. In general, the increase of PES concentration and/or evaporation time, together with the surface modification, produced tighter membranes with improved TOC removals, decreased NOM deposition on top of the membranes and decreased flux reduction, but also resulted in a reduced permeation rate. The best SMM41-modified membranes performed as well as their unmodified counterparts in terms of pure water permeation rate, TOC removal, and final flux. SMM41 modification significantly reduced the amount of NOM deposited on the membrane surface, and in general reduced the percentage of flux reduction. Thus, the effect of surface modification on the membrane performance was positive but marginal. Membranes prepared with PVP had higher porosities and consequently, larger permeation rates than the membranes prepared without PVP, and similar TOC removals. In this case the effect of the SMM41-modification was statistically insignificant. When different fractions of NOM were used as feed solutions a significant effect was observed for all the response variables. Low-molecular-weight NOM, as that present in most groundwaters and coagulated surface waters, caused less fouling, having a higher permeate flux at the end of the experiment; however, smaller TOC removals were observed. In general, there is a compromise between the TOC removal that the membrane can achieve and the permeation rate. Higher TOC removal is achieved by membranes with smaller permeation rate