Materials scientists have overcome a longstanding challenge in balancing stiffness, strength, and toughness in thermosetting polyimide films through an innovative AI-assisted materials-genome approach. Traditional trial-and-error synthesis methods have proven slow, costly, and limited in exploring complex molecular spaces, but researchers from East China University of Science and Technology have developed a machine-learning model capable of predicting three key mechanical parameters across thousands of candidate structures.
The study published in Chinese Journal of Polymer Science introduces Gaussian process regression models trained on over 120 experimental datasets of polyimide films. By defining polymer substructures as molecular genes, the team screened more than 1,700 phenylethynyl-terminated polyimide candidates and identified one formulation, PPI-TB, with simultaneously high mechanical performance across all three key metrics. The research, detailed at https://doi.org/10.1007/s10118-025-3403-x, represents a significant advancement in materials design methodology that could transform multiple industrial sectors.
Molecular dynamics simulations validated the screening process, showing that PPI-TB exhibited superior modulus, toughness, and strength indicators compared with established benchmark systems. Subsequent laboratory experiments on representative polyimides confirmed strong consistency between predicted and measured data, demonstrating the reliability of the AI-driven approach. The models achieved high predictive accuracy for Young's modulus, tensile strength, and elongation at break, enabling comprehensive scoring of every candidate for mechanical performance.
Further analysis revealed key design principles that provide valuable insights for materials engineers. Conjugated aromatic structures enhance stiffness, heteroatoms and heterocycles strengthen molecular interactions, and flexible silicon- or sulfur-containing units improve elongation. These findings demonstrate how integrating AI predictions with molecular interpretation can uncover structure-property relationships and accelerate polymer innovation beyond what traditional methods can achieve.
The implications of this research extend across multiple industries where polyimide films are essential components. Aerospace applications benefit from lightweight, durable materials with thermal stability, while flexible electronics and micro-display technologies require materials that combine mechanical robustness with flexibility. The AI-driven materials-genome strategy provides a universal, scalable framework for designing polymers with targeted combinations of stiffness, strength, and flexibility traits critical for microelectronics, aerospace composites, and flexible circuit substrates.
By replacing years of experimental iteration with predictive modeling and virtual screening, this method drastically reduces development costs and timeframes for high-performance materials. The approach could be adapted for other polymer classes, guiding the creation of lightweight, durable, and thermally stable materials that power future electronic and aerospace technologies. The successful identification and validation of PPI-TB exemplifies how artificial intelligence can redefine the discovery process for next-generation high-temperature polymers, potentially transforming materials development timelines across multiple industrial sectors from months or years to weeks.
