Abstrakt: |
A set of 74 small molecules was employed to generate a quantitative structure-property relationship (QSPR) model for predicting gas-phase Gibb’s free energy changes (ΔGgas) on the simplified molecular input line entry system (SMILES). To obtain the models, the Monte Carlo method (MCM) was applied to calculate the descriptors. The best proposed model of them provides an excellent statistical result of r² = 0.9866, q² = 0.9857, s = 2.48 and F = 4346 for the training set, and r² = 0.9340, q² = 0.8418, s = 5.80, F = 149 for the test set. Consequently, the gas-phase Gibb’s free energy changes were also calculated in DFT (MPW1PW91/6-311+G(2d,2p)), CBS-QB3, G1, G2 and G2MP2 methods which showed a very good agreement with experimental values. [ABSTRACT FROM AUTHOR] |