| The primary objective of this study is to develop a mechanistic understanding of the rejection of organic micropollutants by high-pressure membranes, based on an integrated framework of compound properties, membrane properties, and operational conditions. High-pressure membranes, encompassing reverse osmosis (RO), low-pressure RO (LPRO), and nanofiltration (NF), may provide an effective treatment barrier for representative trace organic compounds including disinfection by-products (DBPs; e.g., trichloroacetic acid, bromoform), halogenated solvents (e.g., trichloroethylene), and pharmaceutically active compounds (PhACs; e.g., ibuprofen). These compounds are emphasized in this research for comparison purposes, based on considerations of compound properties, occurrence, and health effects.; Six RO membranes (e.g., LE-440, XLE-440) and two NF membrane (e.g., NF-90), provided by several manufacturers, were characterized according to pure water permeability (PWP), molecular weight cutoff (MWCO), hydrophobicity (contact angle), and surface charge (zeta potential). In bench-scale experiments, solute rejections were determined with pure water (Milli-Q) as well as synthetic feedwaters with adjusted pH and ionic strength (IS).; The experimental approach of bench-scale cross-flow tests with flat-sheet specimens involved determining rejections from synthetic waters over a range of J0/k ratios and/or recoveries (R). In addition, stirred cell tests have been performed as dynamic adsorption tests, with results compared against static (isotherm) adsorption tests to describe solute partitioning into the membrane. It is noteworthy that rejections of compounds of intermediate hydrophobicity by candidate membranes were observed to be less than salt rejections reported for these membranes, suggesting that transport of these solutes through the membranes is facilitated by solute-membrane interactions. Diffusion cell measurements have been performed using actual membrane specimens to determine "hindered" or "facilitated" diffusion coefficients that, when compared to solute diffusion coefficients in water, describe hindered or facilitated solute transport through a membrane.; Data derived from cross-flow and diffusion cell tests have been used as a basis in formulating a solute transport model, delineating transport by convection versus diffusion. Greater rejections of (negatively) charged compounds than neutral compounds have been observed, and greater rejections of non-polar than polar compounds. In addition, the role of hydrophobic-hydrophobic interactions and the influence of membrane fouling have been further explored. It has been observed that fouling of membranes changes their rejection properties.; Single-elements tests, employing 6.25 cm x 100 cm modules, have also been performed on a recirculation basis to confirm bench-scale results at a larger scale and to address scale-up issures. In these experiments, given the large water volume requirements, deionized water was spiked with a cocktail of several compounds. In selected experiments, results from bench-scale versus single element tests were compared to ascertain how adsorption affects solute rejection. |
