A PFAS-selective semifluoroalkylated polyvinylimidazolium resin for reversible adsorption of anionic fluorosurfactants

Abstract

Per- and polyfluoroalkyl substances (PFAS) in drinking-water sources demand sorbents that capture both long- and short-chain species and can be efficiently regenerated. A hydrolytically robust fluorinated polymer is reported, in which cationic heteroaromatic charges and perfluoroalkyl moieties are co‑localized within each operative functionality. Thus, dual-mode binding sites matching electrostatic and fluorophilic attraction were evaluated as sorbent design for anionic fluorosurfactants. The fluorinated polymer-based anion exchange resin (FP-AER) was synthesized via UV-initiated copolymerization of a perfluoroalkyl-tethered vinylimidazolium monomer with a bis-vinylimidazolium crosslinker and comprehensively characterized by physicochemical methods. Batch adsorption assays versus powdered activated carbon (PAC) quantified single- and multicomponent isotherms, kinetics, co-solute and ionic-strength effects, regeneration, and performance in spiked aqueous film-forming foam (AFFF)-impacted groundwater. FP-AER achieved Langmuir capacities of 597 mg g−1 (PFOA) and 143 mg g−1 (GenX); PFBA exhibited near-linear Freundlich behaviour. At a benchmark equilibrium concentration of 50 μg L−1, FP-AER outperformed PAC by 3–9-fold. The adsorbent was selective in the presence of organic micropollutants (OMPs) and showed pronounced sensitivity to chloride. Regeneration using methanol–salt eluents yielded >90 % recoveries across at least six cycles. In AFFF-impacted groundwater spiked at 0.5 μg L−1 per PFAS, batch tests confirm high PFAS removal by FP-AER relative to PAC. Exploratory rapid small‑scale column tests (RSSCTs) demonstrated feasibility under continuous-flow conditions. A focused mini-review and comparative benchmarking contextualize FP-AER against established AERs and emerging fluorinated sorbents, highlighting the remarkable capacity of the developed dual-mode sorbent; future work will target lower influent levels and optimized column operation.