Data-Driven Insights into Resin Screening for Targeted Per- and Polyfluoroalkyl Substances Removal Using Machine Learning

• DOI: 10.1021/acs.est.4c14223
• Publication Date: 2025-02-11T20:42:06Z
• Authors: Jing Zhang; Kaixing Fu; Shifa Zhong; Jinming Luo

In this study, we address the challenge of screening resins and optimizing operation conditions for removal 43 perfluoroalkyl polyfluoroalkyl substances (PFASs), spanning both long- short-chain fluorocarbon variants, across diverse water matrices, using machine learning (ML) models. We first develop ML models that can accurately predict efficiency PFASs based on resin properties, conditions, matrix. The model performance is validated by a test set our own experimental tests. key features from matrix influencing PFAS as well their interaction effects are comprehensively investigated. finally target long-chain (e.g., PFOS, PFOA) PFBS, GenX), developed to inversely screen determine optimal under specified Experimental tests demonstrated ML-guided approach achieves desired (RE) these PFASs, with RE values reaching 86.56% PFBS 83.73% GenX, outperforming many reported resins. This work underscores potential methodologies in operational optimization enabling efficient structurally varied compounds.