Removal of Per- and Poly-fluoroalkyl Substances (PFAS) from Water: A Computational Approach

Per- and polyfluoroalkyl substances (PFAS) are persistent synthetic compounds widely detected in the environment due to the exceptional stability of the carbon–fluorine bond, posing significant risks to human health and ecosystems. This dissertation develops a multiscale framework for PFAS remediation and detection by integrating ab-initio dispersion-corrected density functional theory (DFT) calculations with data-driven machine learning (ML) approaches. DFT investigations reveal that PFAS adsorption on transition metal surfaces is primarily governed by interactions between functional head groups and the adsorbent, with protonated sulfonic groups exhibiting stronger adsorption than carboxylic groups. On Ni surfaces, adsorption is most favorable on the (110) facet and enhanced by vacancy defects, while pre-adsorbed hydrogen reduces adsorption stability. Zero-valent iron (Fe0) demonstrates superior and facet-independent adsorption affinity, particularly for sulfonated PFAS, and exhibits catalytic activity toward sulfonic group dissociation, highlighting its robustness for remediation. Studies of Cu, Pd, Pt, and Rh further show protonation-dependent adsorption energetics and charge-transfer behavior, suggesting that strategic matching of PFAS protonation states with transition metals can promote targeted redox pathways. Mechanistic analysis of perfluorooctanoic acid (PFOA) degradation on Fe⁰ surfaces indicates substantially reduced activation barriers, with successive defluorination emerging as the most kinetically favorable pathway. Complementary ML models developed for PFAS detection using o-phenylenediamine molecularly imprinted polymer electrochemical sensors demonstrate high classification accuracy, particularly with support vector machines enhanced through Bayesian model averaging. Collectively, this work provides fundamental mechanistic insights and practical strategies advancing PFAS adsorption, catalytic degradation, and electrochemical detection technologies.