QSAR study on the tropospheric degradation of organic compounds

In the environmental risk assessment of organic chemicals, persistence is o f particular importance as it may lead to adverse effects. The reaction of chemicals with OH and NO3 radicals and ozone are the main abiotic degradati on processes in the troposphere, so an upper limit of the atmospheric persi stence of chemicals is assessed by determining their reaction rate constant s with OH. and NO3. and O-3. Statistical models predicting the oxidation ra te constants with OH. and NO3. for many heterogeneous compounds have been d eveloped by the QSAR/QSPR (Quantitative Structure-Activity/Property Relatio nships) approach; the structural representation of the compounds was realis ed using different kinds of molecular descriptors (structural, topological, empirical and WHIM descriptors). In addition, Kohonen neural networks (K-A NN) and the GA-VSS (Genetic Algorithm Variable Subset Selection) strategy w ere respectively used to select the most representative training set and th e best descriptor subset. The predictive capability of the models on k(OH) and k(NO3) has been checked and appears to be satisfactory. Finally, the ox idation rate constants for some chemicals of concern were analysed in the P rincipal Component space in order to rank these chemicals according to thei r tropospheric degradability.