Machine learning strategies for the structure-property relationship of copolymers

Lei Tao; John Byrnes; Vikas Varshney; Ying Li

doi:10.1016/j.isci.2022.104585

Machine learning strategies for the structure-property relationship of copolymers

iScience. 2022 Jun 10;25(7):104585. doi: 10.1016/j.isci.2022.104585. eCollection 2022 Jul 15.

Authors

Lei Tao¹, John Byrnes², Vikas Varshney³, Ying Li^{1

4}

Affiliations

¹ Department of Mechanical Engineering, University of Connecticut, Storrs, CT 06269, USA.
² SRI International, San Diego, CA 92131, USA.
³ Materials and Manufacturing Directorate, Air Force Research Laboratory, Wright-Patterson Air Force Base, Ohio 45433, USA.
⁴ Polymer Program, Institute of Materials Science, University of Connecticut, Storrs, CT 06269, USA.

Abstract

Establishing the structure-property relationship is extremely valuable for the molecular design of copolymers. However, machine learning (ML) models can incorporate both chemical composition and sequence distribution of monomers, and have the generalization ability to process various copolymer types (e.g., alternating, random, block, and gradient copolymers) with a unified approach are missing. To address this challenge, we formulate four different ML models for investigation, including a feedforward neural network (FFNN) model, a convolutional neural network (CNN) model, a recurrent neural network (RNN) model, and a combined FFNN/RNN (Fusion) model. We use various copolymer types to systematically validate the performance and generalizability of different models. We find that the RNN architecture that processes the monomer sequence information both forward and backward is a more suitable ML model for copolymers with better generalizability. As a supplement to polymer informatics, our proposed approach provides an efficient way for the evaluation of copolymers.

Keywords: Artificial intelligence; Materials science; Polymers.