A QSPR treatment for the thermal stabilities of second-order NLO chromophore molecules

Quantitative structure-property relationships were studied between descriptors representing the molecular structures and thermal decomposition temperatures (T(d)) for a diverse set of 90 second-order nonlinear optical (NLO) chromophores. A seven-parameter model was developed for the prediction of molar thermal decomposition function Y(d) (T(d) M, where M represents the molar weight) with R2 = 0.9642 and SEE = 14.01 by multilinear regression analysis. The mean relative error for the prediction of T(d) was 4.46%. The stability of the proposed model was validated using leave-one-out cross-validation. All descriptors involved in the model were derived solely from the chemical structures of the NLO chromophores.