A team of researchers at the University of Nottingham has developed new artificial intelligence tools to encourage chemists and chemical engineers to make their processes more sustainable.

Led by Professor Jonathan Hirst, the Royal Academy of Engineering Chair in Emerging Technologies, the team has enhanced a well-known AI algorithm to improve predictions for retrosynthesis. Retrosynthesis is a framework used by chemists to determine the best way to synthesize molecules from readily available starting materials. Their research findings were published this week in the Journal of Chemical Information and Modeling.

This method is also accessible through the open-source AI4Green electronic laboratory notebook at https://ai4green.app.

Hirst stated, "Chemical processes are critical to the invention of new molecules and materials, driving innovation across industries—from creating new effective drug therapies to providing new batteries for sustainable energy. However, these processes can be toxic and environmentally damaging, so chemists must find new ways of working to achieve more sustainable outcomes. AI is providing shareable solutions for this, and we hope everyone working in the chemical field can use this new framework."

Earlier this year, the AI4Green team published a new interactive machine learning method in ACS Sustainable Chemistry & Engineering, which helps chemists select greener solvents—the medium in which chemical reactions occur. Choosing more sustainable solvents is a key factor in reducing the environmental impact of chemical processes.

Training the next generation of chemists to use AI tools and digital research environments is crucial. The AI4Green team, in collaboration with the Physical Sciences Data Infrastructure, has developed a teaching version of the AI4Green electronic laboratory notebook.

This electronic laboratory notebook is being used in the School of Chemistry's teaching laboratories for third-year undergraduate students, aiming to enhance their awareness of sustainable chemistry. The research findings have been published in this month's issue of the Journal of Chemical Education.