(197br) Toward an End-to-End Computational Framework for Polymer Materials Design

Polymer materials are extensively used in a wide variety of domestic and industrial applications, including energy storage, membranes, biomedical applications, etc^1,2. Despite significant progress in new materials development over the past few decades, exploring the polymer space still holds strong potential due to the variability of monomer structure and polymerization methods, making the design space infinitely large. To meet the demand for better polymer materials with enhanced thermal, mechanical, optical, and electronic properties, as well as ease of synthesis and production, computational modeling has been used to simulate polymerization and polymer molecules, and can be used to support experimental efforts by providing guidance for more rational experiment design^3–5.

However, traditionally, polymerization modeling and polymer molecular simulation have been treated as separate subjects, resulting in suboptimal or infeasible design. Thus, an integrated computational framework for closed-loop polymer design is highly desirable, yet challenging to achieve due to the gaps in information and time scale of different simulation methods.

In this work, we present our efforts towards an integrated framework for designing polymerization recipes and targeting polymer material properties. We improved the kinetic Monte Carlo algorithm and combined it with an active learning algorithm to design the polymer chain sequence distribution efficiently. Coarse-grained molecular dynamic simulation was then used to connect the polymer chain information with the thermal and mechanical properties of the polymer material. Additionally, machine learning algorithms were employed to approximate molecular dynamic simulations and better unite the time scales of polymerization modeling and polymer molecular simulations. Our work is a first step towards an end-to-end computational framework for polymer materials design, with the potential to extend the scope of materials and connect with experiments.

Reference

(1) Ali, U.; Karim, K. J. B. A.; Buang, N. A. A Review of the Properties and Applications of Poly (Methyl Methacrylate) (PMMA). Polym. Rev. 2015, 55 (4), 678–705. https://doi.org/10.1080/15583724.2015.1031377.

(2) Frontiers in Polymer Science and Engineering. A Natl. Sci. Found. Spons. Work. 2016.

(3) de Pablo, J. J.; Jackson, N. E.; Webb, M. A.; Chen, L.-Q.; Moore, J. E.; Morgan, D.; Jacobs, R.; Pollock, T.; Schlom, D. G.; Toberer, E. S. New Frontiers for the Materials Genome Initiative. npj Comput. Mater. 2019, 5 (1), 41.

(4) Gao, H.; Waechter, A.; Konstantinov, I. A.; Arturo, S. G.; Broadbelt, L. J. Application and Comparison of Derivative-Free Optimization Algorithms to Control and Optimize Free Radical Polymerization Simulated Using the Kinetic Monte Carlo Method. Comput. Chem. Eng. 2018, 108, 268–275. https://doi.org/10.1016/j.compchemeng.2017.09.015.

(5) Gao, H.; Oakley, L. H.; Konstantinov, I. A.; Arturo, S. G.; Broadbelt, L. J. Acceleration of Kinetic Monte Carlo Method for the Simulation of Free Radical Copolymerization through Scaling. Ind. Eng. Chem. Res. 2015, 54 (48), 11975–11985. https://doi.org/10.1021/acs.iecr.5b03198.