|
1K - Tropospheric Fate (MO13/10) Modelling and Prediction of Atmospheric Persistence Index of VOCs by theoretical molecular descriptors. Gramatica, Paola1, Pilutti, Pamela1, Papa, Ester1, 1 University of Insubria, Varese, VA, Italy ABSTRACT- The tropospheric lifetime of most organic chemicals deriving from terrestrial emissions is mainly controlled by their degradation reaction with OH radicals and ozone during the daytime, and NO3 radicals at night. The limited availability, and the variability, of data related to atmospheric degradation reactions is a very relevant issue in studies related to environmental fate and chemical behaviour. A PCA model based on three principal degradation processes has been proposed to evaluate the overall tropospheric persistence of chemicals. The data set is composed of 399 VOCs for which the data for all the three studied oxidation reactions are either experimentally available or are obtained by our QSAR models. The PC1 score is proposed as an Atmospheric Persistence INdex (ATPIN) and is also modelled by theoretical molecular descriptors. These molecular descriptors are 1D-, 2D- and 3D-descriptors (constitutional, topological, WHIM, GETAWAY, quantum-chemical and others). The best descriptor subsets for the modelling were selected using the Genetic Algorithm-Variable Subset Selection strategy (GA-VSS) and model calculations were performed by Ordinary Least Squares regression (OLS). In order to obtain predictions for new chemicals, the best model was validated either internally by leave-one-out (Q2LOO=93%) and leave-many-out (50%) (Q2LMO=92%), or externally (by preliminary training-evaluation set splitting) (Q2EXT=95%), and scrambling the responses; the reliability of the predictions was checked by the leverage approach. Model predictivity appears good, thus this model can be used as an evaluative model to screen and rank chemicals just starting from chemical structure, thus also before synthesis, according to their tendency to degrade in the troposphere and their derived atmospheric persistence. Key words: Principal component analysis (PCA), VOC, Atmospheric persistence index, QSAR |
