A Machine Learning Approach for Predicting Defluorination of Per- and Polyfluoroalkyl Substances (PFAS) for Their Efficient Treatment and Removal

Akber Raza,Sharmistha Bardhan,Lihua Xu,Sharma S. R. K. C. Yamijala,Chao Lian,Hyuna Kwon,Bryan M. Wong

DOI: https://doi.org/10.26434/chemrxiv.9756557.v1

IF: 11.558

2019-01-01

Environmental Science & Technology Letters

Abstract:Wepresent the first application of machine learning on per- and polyfluoroalkyl substances (PFAS) for predicting and rationalizing carbon-fluorine (C–F)bond dissociation energies to aid in their efficient treatment and removal. Usinga variety of machine learning algorithms (including Random Forest, LeastAbsolute Shrinkage and Selection Operator Regression, and Feed-forward NeuralNetworks), we were able to obtain extremely accurate predictions for C–F bond dissociation energies (withdeviations less than 0.70 kcal/mol) that are within chemical accuracy ofthe PFAS reference data. In addition, we show that our machine learningapproach is extremely efficient (requiring less than 10 minutes to train thedata and less than a second to predict the C–F bond dissociation energy of anew compound) and only needs knowledgeof the simple chemical connectivity in a PFAS structure to yield reliableresults – without recourse to a computationally expensive quantummechanical calculation or a three-dimensional structure. Finally, we present anunsupervised machine learning algorithm that can automatically classify andrationalize chemical trends in PFAS structures that would otherwise have beendifficult to humanly visualize/process manually. Collectively, these studies (1)comprise the first applicationof machine learning techniques for PFAS structures to predict/rationalize C–F bond dissociation energies and (2) show immensepromise for assisting experimentalists in the targeted defluorination ofspecific bonds in PFAS structures (or other unknown environmental contaminants)of increasing complexity.