An Informatics Framework for the Design of Sustainable, Chemically Recyclable, Synthetically-Accessible and Durable Polymers

Joseph Kern,Yongliang Su,Will Gutekunst,Rampi Ramprasad

2024-09-14

Abstract:We present a novel approach to design durable and chemically recyclable ring-opening polymerization (ROP) class polymers. This approach employs digital reactions using virtual forward synthesis (VFS) to generate over 7 million ROP polymers and machine learning techniques to rapidly predict thermal, thermodynamic and mechanical properties crucial for application-specific performance and recyclability. This combined methodology enables the generation and evaluation of millions of hypothetical ROP polymers from known and commercially available molecules, guiding the selection of approximately 35,000 candidates with optimal features for sustainability and practical utility. Three of these recommended candidates have passed validation tests in the physical lab - two of the three by others, as published previously elsewhere, and one of them is a new thiocane polymer synthesized, tested and reported here. This paper presents the framework, methodology, and initial findings of our study, highlighting the potential of VFS and machine learning to enable a large-scale search of the polymer universe and advance the development of recyclable and environmentally benign polymers.

Chemical Physics,Materials Science,Soft Condensed Matter

What problem does this paper attempt to address?

The paper aims to address the following issues:

1. **Designing sustainable and chemically recyclable polymers**: The paper proposes a new method to design durable and chemically recyclable ring-opening polymerization (ROP) polymers. This method utilizes virtual forward synthesis (VFS) to generate over 7 million ROP polymers and employs machine learning techniques to quickly predict key properties such as thermodynamics, thermal, and mechanical performance to meet specific application needs and recyclability.

2. **Replacing polystyrene (PS) materials**: Specifically, the research focuses on finding new polymers suitable for replacing polystyrene (PS) as container materials. Although polystyrene holds a significant position in plastic production, its recyclability and environmental health issues make it a focal point of research.

3. **Large-scale search of the polymer universe**: Through VFS and machine learning techniques, this method can efficiently generate and evaluate millions of hypothetical ROP polymers, guiding the selection of approximately 35,000 candidate polymers with optimal properties to meet sustainability and practicality needs.

4. **Experimental validation**: The paper also introduces three experimentally validated design schemes, two of which have been published by other researchers, and one is a newly synthesized sulfur ring polymer, reported for the first time in this paper.

In summary, the paper aims to develop an informatics framework by combining virtual synthesis and machine learning methods to design sustainable and chemically recyclable polymers, particularly for applications replacing polystyrene.