SMiPoly: Generation of a Synthesizable Polymer Virtual Library Using Rule-Based Polymerization Reactions

Area of Science:

• Polymer Chemistry
• Computational Chemistry
• Materials Science

Background:

• Machine learning accelerates de novo molecular design in polymer research.
• Current methods struggle with identifying synthetic routes for designed polymers.
• Molecular generators are crucial for creating and modifying polymer structures.

Purpose of the Study:

• To develop a computational tool for generating synthesizable polymers.
• To address the unresolved challenge of predicting synthetic accessibility for designed polymers.
• To integrate virtual polymer generation into molecular design workflows.

Main Methods:

• Developed the Small Molecules into Polymers (SMiPoly) Python library.
• Implemented 22 chemical rules for common polymerization reactions.
• Generated polymers from 1083 readily available monomers, covering seven molecular types.

Main Results:

• Generated 169,347 unique polymers, including polyolefins, polyesters, polyamides, and more.
• Achieved 48% coverage and 53% novelty compared to existing synthesized polymers.
• Demonstrated SMiPoly's ability to create an exhaustive list of potentially synthesizable polymers.

Conclusions:

• SMiPoly facilitates the virtual generation of synthesizable polymers.
• The library enhances de novo polymer design by predicting synthetic routes.
• Incorporating SMiPoly accelerates the selection of viable candidate polymers for synthesis.