摘要

In this study, a novel model based on aromatic meta-substituent grouping was presented to predict the second order rate constants (k) for center dot OH oxidation of PCBs in gas phase. Since the oxidation kinetics are dependent on the chlorination degree and position, we hypothesized that it may be more accurate for k value prediction if we group PCB congeners based on substitution positions (i.e., ortho (o), meta (m), and para (p)). To test this hypothesis, we examined the correlation of polarizability (a), a quantum chemical based descriptor for k values, with an empirical Hammett constant (sigma(+)) on each substitution position. Our result shows that alpha is highly linearly correlated to Sigma sigma(+)(o,m,p) based on aromatic meta-substituents leading to the grouping based predictive model. With the new model, the calculated k values exhibited an excellent agreement with experimental measurements, and greater predictive power than the quantum chemical based quantitative structure activity relationship (QSAR) model. Further, the relationship of alpha and Sigma sigma(+)(o,m,p) for PCDDs congeners, together with highest occupied molecular orbital (HOMO) distribution, were used to validate the aromatic meta-substituent grouping method. This newly developed model features a combination of good predictability of quantum chemical based QSAR model and simplicity of Hammett relationship, showing a great potential for fast and computational tractable prediction of k values for gas phase center dot OH oxidation of polychlorinated compounds.