pKa Data-Driven Insights into Multiple Linear Regression Hydrolysis QSARs: Applicability to Perfluorinated Alkyl Esters

Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are used in many applications due to their attractive nonstick and fire-fighting properties. When released into aquatic environments, commercial PFAS and associated byproducts may be transformed by hydrolysis to form more persistent products. For example, hydrolysis of perfluorinated carboxylic acid esters leads to perfluorinated carboxylic acids (PFCAs) and fluorotelomer alcohols (FTOHs). Quantitative structure activity relationships (QSARs) previously developed for predicting carboxylic acid ester hydrolysis rates are assessed for their predictive performance for perfluorinated molecules using compiled perfluorinated alkyl ester hydrolysis data. The assessment indicates that the model(s) are capable of estimating half-lives of various chain-lengths of perfluorinated alkyl esters; however, predictive performance could be improved through more accurate calculated chemical descriptor values for precursors of PFCAs. In particular, more accurate estimated pKa values are needed to improve hydrolysis rate predictions. The performance of several available cheminformatic applications (ChemAxon, SPARC, pkasolver, MolGpka and OPERA) is assessed for estimating pKa values for PFCAs and FTOHs, highlighting the need for improvements in predicting pKa values for longer-chain PFAS with perfluorinated α-carbons adjacent to the OCO ester group.