Prediction of Solubility of Fullerene C60 in Various Organic Solvents by Genetic Algorithm-Multiple Linear Regression

Abstract

Quantitative structure property relationship (QSPR) study is presented for modeling and predicting of solubility of fullerene (C60) in various solvents. A data set consisting of 36 benzene derivatives is used in this study. Various kinds of molecular descriptors were calculated to represent the molecular structures of compounds and the best-fitting descriptors were selected by using stepwise multiple linear regressions (SW-MLR) and a genetic algorithm (GA-MLR) the selection of variables. The models were validated using leave-one-out (LOO), leave-multiple-out (LMO) cross-validation, external test set and Y-randomization test. The outliers were also examined to understand better in which cases large errors were to be expected and to improve the predictive models. Comparison of the results obtained indicated the superiority of the genetic algorithm over the stepwise. Also electronegativity, dispersion interaction in solution and volume of molecule were the main independent factors contributing to the solubility of fullerenes in the studied solvents.