QSPR study on refractive indices of solvents commonly used in polymers´s chemistry using flexible molecular descriptors

FIORESSI, S. E.; BACELO, D. E.; DUCHOWICZ, P. R.

Workshop; Primer Workshop Latinoamericano de Modelado Molecular y Simulación Computacional; 2016

A predictive Quantitative Structure-Property Relationship (QSPR) for the refractive indices of 370 solvents commonly used in the processing and analysis of polymers is presented, using as chemical information descriptors the simplified molecular input line entry system (SMILES). The model employs a flexible molecular descriptor and a conformation-independent approach. Various well-known techniques, such as the use of an external test set of compounds, the Cross-Validation method, and Y-Randomization were used to test and validate the established equations. The predicted values were finally compared with published results from the literature. The simple model proposed correlates the refractive index values with good accuracy, and it is not dependent on the 3D-molecular geometries.