Measurement and QSPR modeling of Flory–Huggins parameter for solvent-swollen gels, and gel catalyst informatics

Background

The polymer–solvent interaction parameter (Flory–Huggins parameter), χ, is a quantitative measure of the degree of interaction between polymer and solvent molecules, reflecting the solvent’s affinity for the polymer. Predicting χ using a quantitative structure–property relationship (QSPR) model helps bridge data gaps and facilitates the development of new materials. We explore the application of materials informatics to accelerate research and development (R&D) activities in laboratory-scale projects.

Methods

We constructed a χ database for amphiphilic polymer gels by synthesizing ten types of gels and determining their χ values with various solvents through swelling tests. QSPR modeling was employed to predict these χ values using the random forest machine learning algorithm based on descriptors for each polymer and solvent serving as independent variables. Gel catalyst informatics was applied by developing copolymer gels bearing sulfo groups as acidic catalysts and evaluating their performance in solvent-free esterification reactions.

Findings

The catalytic activities of copolymer gels were evaluated in relation to the χ values predicted by the QSPR model. This study presents the first predictive model for the χ values of amphiphilic polymer gels using QSPR modeling combined with machine learning, establishing a new database and contributing to the advanced design of functional materials for chemical engineering applications.