Session: Rising Stars of Mechanical Engineering Celebration & Showcase

Paper Number: 148559

148559 - Discovery of Multi-Functional Polyimides Through High-Throughput Screening Using Explainable Machine Learning 

Polyimides have been widely used in modern industries because of their excellent mechanical and thermal properties, e.g., high-temperature fuel cells, displays, and aerospace composites. However, it usually takes decades of experimental efforts to develop a successful product. Aiming to expedite the discovery of high-performance polyimides, we utilize computational methods of machine learning (ML) and molecular dynamics (MD) simulations. Our study provides compelling evidence for the effectiveness of a data-driven approach in discovering novel polyimides. We first build a comprehensive library of more than 8 million hypothetical polyimides based on the polycondensation of existing dianhydride and diamine/diisocyanate molecules. Then we establish multiple ML models for the thermal and mechanical properties of polyimides based on their experimentally reported values, including glass transition temperature, Young’s modulus, and tensile yield strength. The obtained ML models demonstrate excellent predictive performance in identifying the key chemical substructures influencing the thermal and mechanical properties of polyimides. The use of explainable machine learning describes the effect of chemical substructures on individual properties, from which human experts can understand the cause of the ML model decision. Applying the well-trained ML models, we obtain property predictions of the 8 million hypothetical polyimides. Then, we screen the whole hypothetical dataset and identify three (3) best-performing novel polyimides that have better-combined properties than existing ones through Pareto frontier analysis. For an easy query of the discovered high-performing polyimides, we also create an online platform that embeds the developed ML model with interactive visualization. Furthermore, we validate the ML predictions through all-atom MD simulations and examine their synthesizability. The MD simulations are in good agreement with the ML predictions and the three novel polyimides are predicted to be easy to synthesize via Schuffenhauer’s synthetic accessibility score. Following the proposed ML guidance, we successfully synthesized a novel polyimide and the experimentally obtained high glass transition/thermal decomposition temperature demonstrated its excellent thermal stability. This study discovers novel polyimides with promising thermomechanical properties and guides the further experimental synthesis of innovative polyimides. More importantly, our method of utilizing explainable ML techniques and high-fidelity MD simulations demonstrates an efficient way to deal with a daunting number of chemical structures. It is important to note that our proposed method is designed specifically to provide guidance for the selection of promising candidates and corresponding raw materials, rather than to evaluate the entire experimental synthesis process. While factors such as solvent selection, reaction time, temperature control, toxicity, and other conditions are undoubtedly critical aspects of the experimental synthesis process, they are outside the scope of this study. Nonetheless, we believe that our ML-assisted workflow represents a significant step forward in the field of polymer informatics and offers exciting possibilities for future research. By leveraging the power of advanced ML techniques and carefully designed workflows, we can rapidly identify promising candidates for further development and enhance the efficiency and effectiveness of materials discovery efforts.

Presenting Author: Ying Li University of Wisconsin-Madison

Presenting Author Biography: Dr. Li joined the University of Wisconsin-Madison in August 2022 as an Associate Professor of Mechanical Engineering. From 2015 to 2022, he was an Assistant Professor of Mechanical Engineering at the University of Connecticut and was promoted to Associate Professor. He received his Ph.D. in 2015 from Northwestern University, focusing on the multiscale modeling of soft matter and related biomedical applications. His current research interests are: multiscale modeling, computational materials design, mechanics and physics of polymers, and machine learning-accelerated polymer design. Dr. Li’s achievements in research have been widely recognized by fellowships and awards, including NSF CAREER Award (2021), Air Force’s Young Investigator Award (2020), 3M Non-Tenured Faculty Award (2020), ASME Haythornthwaite Young Investigator Award (2019), NSF CISE Research Initiation Initiative Award (2018) and multiple best paper awards from major conferences. He has authored and co-authored more than 100 peer-reviewed journal articles, including Physical Review Letters, ACS Nano, Biomaterials, Nanoscale, Macromolecules, Journal of Mechanics and Physics of Solids, and Journal of Fluid Mechanics, etc. He has been invited as a reviewer for more than 90 international journals, such as Nature Communications and Science Advances. Dr. Li’s lab is currently supported by multi-million-dollar grants and contracts from NSF, AFOSR, AFRL, ONR, DOE/National Nuclear Security Administration, DOE/National Alliance for Water Innovation, and industries.

Authors: