Home – Home – Jornals – Journal of Structural Chemistry 2009 number 5

A quantitative structure-property relationship (QSPR) model obtained by using multiple linear stepwise regression analysis, with correlation coefficient R of 0.979 for the training set and 0.951 for the validation set, was developed to predict the glass transition temperature (T_g) values of polyacrylates. Three quantum chemical descriptors (the molecular average polarizability α, the entropy S, and the lowest unoccupied molecular orbital E_LUMO) obtained directly from polyacrylates monomer structure by density function theory (DFT) calculation, were used to produce the model. The result confirmed the role that quantum chemical descriptors play in studying QSPR of glass transition temperature for polymers.