In a major step forward for materials science and polymer informatics, Damilola Ojedeji, a Ph.D. researcher in Chemical and Biomolecular Engineering at the University of Tennessee, Knoxville (UTK), has co-authored a study introducing a powerful new framework for quantitative structure–property relationships (QSPR) in polyester materials

The research, published in Macromolecular Theory and Simulations, leverages statistical learning to predict critical polymer properties such as glass transition temperature (T_g) and intrinsic viscosity with unprecedented accuracy. The approach, developed in collaboration with Eastman Chemical Company, Kingsport, Tennessee, USA, and UTK’s Department of Mathematics, significantly enhances the ability to design and optimize polymers for industrial applications.

“By combining chemical intuition with statistical learning, we’ve created a predictive model that is both accurate and adaptable,” said Ojedeji. “This work helps shorten the time from concept to material, which is critical for industry.”

Key Advantages of the Framework

• High accuracy: T_g predictions reached R² ≈ 0.96, outperforming traditional methods.
• Multi-property capability: Simultaneous modeling of T_g and viscosity strengthens prediction power.
• Uncertainty quantification: A Bayesian framework provides confidence intervals for each prediction.
• Scalability: Easily adapted to new polymers, blends, or alternative chemistries

The study also highlights how the framework compares favorably to established models such as Bicerano’s correlations, offering superior accuracy without the need for extensive molecular weight corrections.

Impact for Industry

This innovation provides companies with a powerful tool to accelerate polymer development, reducing reliance on costly trial-and-error experiments. Potential applications include packaging, coatings, and engineering plastics, as well as new opportunities in sustainable bio-based polymers and next-generation energy storage materials.

“This framework represents a transformative step for polymer R&D,” said Ojedeji. “It allows for faster, smarter material discovery – exactly what’s needed as industries push for sustainable, high-performance solutions.”

The research underscores the importance of industry-academia partnerships, with Eastman Chemical Company providing critical experimental data that validated the models.