Two new predictors combined with quantum chemical parameters for the selection of oxidants and degradation of organic contaminants: A QSAR modeling study

Abstract Oxidation is an attractive treatment method to effectively remove organic contaminants in water. In this study, degradation of 30 organic compounds in different oxidation systems was evaluated, including oxygen (O2), hydrogen peroxide (H2O2), ozone (O3) and hydroxyl radical (HO ). First, a quantitative structure-activity relationship (QSAR) model for oxidation-reduction potentials (ORPs) of organics was developed and exhibited a good performance to predict ORP values of organics with evaluation indices of squared correlation coefficient ( R 2 ) = 0.866, internal validation ( q 2 ) = 0.811 and external validation ( Q e x t 2 ) = 0.669. Four quantum parameters, including f ( + ) n , f ( − ) n , E H O M O and E B 3 L Y P dominate the ORP values. Subsequently, a relationship between reaction rates (k) and the difference of ORP for oxidants and organics ( Δ E o x i − o r g ) was established, however, which was limited ( R 2 = 0.697). Therefore, two new predictors (slopes and intercepts) are proposed based on the linear relationships between k values and ORPs of oxidants. These new predictors can be applied to estimate the reaction rates and minimum oxidation potential for organic compounds. Afterwards, to express the two predictors, QSAR models were established. The two optimal QSAR models fitted very well with experimental values and were demonstrated to be stable and accurate based on R 2 (0.982 and 0.965), q 2 (0.950 and 0.950) and Q e x t 2 (0.985 and 0.989). B O x , q ( H ) + and q ( C ) x were main factors influencing the slopes and intercepts. This study developed methods to predict ORPs of organics and established two new predictors to estimate the reaction rates undergoing different oxidation processes, offering new insights into the oxidant selection.