PREDICTING REVERSE OSMOSIS REMOVAL OF TOXICOLOGICALLY RELEVANT UNIQUE ORGANICS

Climate change, aging infrastructure, increasing population, and shifts in land use and urbanization present challenges for managing water resources in urban areas. Water reuse is receiving increasing attention as a strategy that can improve reliability and drought resistance of water supplies. Direct potable reuse (DPR) is being considered to purify water directly into a potable water supply distribution system. Reverse Osmosis (RO) is an attractive and promising membrane-based treatment process that separates dissolved compounds from water by forcing water through the membrane under pressure. RO is implemented in wastewater treatment processes after filtration and before advanced oxidation processes. However, RO will not completely remove some emerging contaminants. Rejection of contaminants by RO membranes depends on membrane properties, physical chemical properties of organics, and process operational parameters. This calls for a need to develop a comprehensive and practical model that can be used to predict the rejection efficiencies of RO for a wide variety of organics possibly present in wastewater reclamation process. We developed a comprehensive predictive model that was supported by experimental work conducted the team of Dr. Kerry Howe in University of New Mexico. The selection of organics was based on the availability of analytical methods, physical chemical properties and past rejection performance of RO found in the literature. The compounds include a wide variety of groups of organics such as saturated and unsaturated chlorinated hydrocarbons, aromatic hydrocarbons, and chlorinated aromatics. The bench-top laboratory experiments were performed to examine the mass transfer coefficients of both water and targeted organics at one operational condition at around neutral pH because these determine the concentration of organics in the permeate solution. The thermodynamic property related to the partition coefficient that reflects solute-membrane interaction and the kinetic property of diffusion coefficient that reflects the steric hindrance effect are the two dominant factors of mass transfer of organics through RO membrane. We theoretically calculated fundamental intrinsic interaction parameter, aqueous-phase free energy of interaction between organics and membrane (), which is proportional to the natural logarithm of partition coefficient (K). The aqueous-phase diffusion coefficient was estimated by using Hayduk-Laudie’s equation to reflect the kinetic property of the solute, which has been verified has correlation with the diffusion coefficient in membrane phase. The conventional molecular descriptors, such as solubility, Henry’s constant, and molecular dimensions, were also considered. Then, we developed empirical QSAR models to predict the ks of organics through the reference RO membrane of ESPA2-LD with the governing molecular descriptors