When samples containing perfluoroalkyl and polyfluoroalkyl substances (PFAS) are filtered prior to analysis, inadvertent adsorption onto membrane materials can result in concentration underestimations. Herein, we systematically examined the adsorption of six PFAS (PFOA, PFOS, PFNA, PFHxS, PFBS, GenX) onto 11 membrane syringe filters, differing in manufacturer, polymer material, diameter, and/or pore size under various experimental conditions. We perform a comprehensive characterization of the membrane materials to determine differences in morphology, zeta potential, porosity, surface area, and roughness. Evaluation of postfiltration PFAS recovery demonstrated significant impacts of filter material and surface area (related to pore size and diameter), initial PFAS concentration, water pH, and the co-occurrence of cations and anions. Longer-chain PFAS more readily adsorb than shorter-chain PFAS for all filter materials under all experimental conditions. Machine learning predictions of Abraham’s solute descriptors were used to qualitatively assess the predominant forces governing PFAS adsorption onto different materials. While differences in hydrogen-bonding ability exert influence, hydrophobicity and electrostatic interactions are the main drivers of adsorption, along with significant impacts of the water matrix. We recommend polypropylene, mixed cellulose ester, or glass fiber filters with larger pore sizes and smaller diameters to filter PFAS-containing samples, and we discourage the use of nylon-based filters. These findings offer important insights for optimizing PFAS sample preparation and improving membrane design for effective PFAS removal.
Reid et al. (Sat,) studied this question.