pK_a Data-Driven Insights into Multiple Linear Regression Hydrolysis QSARs: Applicability to Perfluorinated Alkyl...

Area of Science:

• Environmental Chemistry
• Chemical Kinetics
• Predictive Modeling

Background:

• Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are widely used but can transform into persistent byproducts in aquatic environments.
• Hydrolysis of PFAS esters generates perfluorinated carboxylic acids (PFCAs) and fluorotelomer alcohols (FTOHs), impacting environmental persistence.
• Accurate prediction of PFAS transformation rates is essential for environmental risk assessment.

Purpose of the Study:

• To assess the predictive performance of existing quantitative structure activity relationships (QSARs) for perfluorinated alkyl ester hydrolysis.
• To identify key chemical descriptors and properties, such as pKa values, that influence hydrolysis rate predictions for PFAS.
• To evaluate the accuracy of cheminformatic tools in estimating pKa values for PFCAs and FTOHs.

Main Methods:

• Compiled and analyzed experimental hydrolysis data for various perfluorinated alkyl esters.
• Assessed the predictive capabilities of established QSAR models using this compiled data.
• Evaluated the performance of multiple cheminformatic applications (ChemAxon, SPARC, pkasolver, MolGpka, OPERA) for estimating pKa values of PFCAs and FTOHs.

Main Results:

• QSAR models demonstrate capability in estimating half-lives for perfluorinated alkyl esters across different chain lengths.
• Predictive accuracy of hydrolysis rates can be enhanced by improving calculated chemical descriptor values for PFCA precursors.
• Accurate estimation of pKa values is identified as a critical factor for improving hydrolysis rate predictions, particularly for longer-chain PFAS.

Conclusions:

• Existing QSAR models offer a foundational ability to predict PFAS ester hydrolysis rates.
• Improvements in calculating pKa values for PFCA precursors are necessary for more reliable hydrolysis predictions.
• Further development is needed in cheminformatic tools to accurately predict pKa values for complex PFAS structures, especially longer-chain compounds.