Group contribution method promoted correlations of glass transition temperature of deep eutectic solvents

Abstract

Glass transition temperature (T_g) being a crucial thermal property, linear models for predicting T_g of deep eutectic solvents (DES) are proposed. The group contribution method and genetic algorithm using an experimental dataset of 51 DESs are applied to obtain the group contribution values for each functional group present in DESs by considering their stereochemistry. By segregating the DESs into subclasses according to their molecular structures, linear regressions for each class are performed to develop the model. The framework is used to compute the T_g of all DESs taken in this study, and it shows an absolute average deviation of 2.7%.