Predicting low density polyethylene-water partition coefficients based on pp-LFER and QSPR models using molecular descriptors

Abstract

The partition coefficients of chemicals between sorbent material and water (KPE-w) are important to determine water phase chemical concentrations using passive sampling devices. Due to the wide application of low density polyethylene (LDPE) films, the polyparameter linear free energy relationship (pp-LFER) and quantitative structure property relationship (QSPR) models were established for predicting KPE-w values. With the pp-LFER model, V (McGowan's molar volume), B (hydrogen bond acceptor capacity) and A (hydrogen bond donor capacity) were screened as significant variables. However, this model had low correlation coefficient (R2) and cross-validation coefficient (Q2). To comprehensively study the mechanisms of the partition process, a QSPR model with 4 influential descriptors (CrippenLogP (Crippen octanol-water partition coefficient), CIC0 (neighborhood symmetry of 0-order), MATS3i (Moran autocorrelation-lag3/weighted by first ionization potential) and A (hydrogen bond donor capacity)) was developed. The values of R2 and Q2 ranged from 0.771 to 0.921 and 0.739 to 0.912, respectively, indicating that both models had good predictive capacity and robustness. Mechanism interpretation indicated that the elements influencing the distribution procedure between LDPE and water were the McGowan's molar volume and hydrophobic interactions. In the present study, a superb model was established to predict the log KPE-w values for a wide range of common hydrophobic organic pollutants.