Simple Approach for Reliable Prediction of Solubility of Polymers in Environmentally Compatible Solvents

Abstract

This work presents a simple approach for predicting the solubility parameter (δ) of polymers. It is based on the number of moles of hydrogen, nitrogen, oxygen, sulfur, fluorine, and silicon atoms per mole of repeating unit structures as well as two correcting functions based on the contributions of hydrogen bonding and polar groups. The largest available experimental data including the δ values of 163 amorphous polymers and amorphous phases of semicrystalline polymers are used to derive and test the new model. Different statistical parameters are used to confirm the high reliability of the novel method through internal and external validations as well as by comparing with two of the best existing quantitative structure–property relationship (QSPR) approaches containing complex descriptors. Since the available QSPR methods are restricted to polymers with repeating unit structures −(C1H2–C2R3R4)–, the results of the new method are compared with the outputs of QSPR methods for 97 polymers. The value of root-mean-square error (RMSE) of the novel approach for external validation of test data (57 different polymers) is 0.91 MPa1/2, which is close to RMSE values of the computed outputs of two complex QSPR methods, that is, 0.90 and 0.88 MPa1/2.