Planning supply chains in the chemical industry can help to improve sustainability by including renewable feedstocks and by reusing and reducing waste products. Integrating waste steams back into the supply chain provides new chemical feedstocks. To use those efficiently within the industry, a wide range of possible reaction networks needs to be evaluated. The number of possible compounds and reactions in these networks is huge and can, thus, best be treated with computational approaches.

Alexei A. Lapkin, Cambridge Centre for Advanced Research and Education in Singapore and University of Cambridge, UK, and colleagues have screened a large chemical database with over 105 million compounds and identified “strategic” compounds which are promising for a circular economy. The team used graph theory to describe the molecules’ properties within the reaction network and used a specific search algorithm (an isolation forest search algorithm) to identify the strategic molecules. The team evaluated, e.g., the number of reaction “links” connecting the molecule to others, how central the molecule is within the reaction network, and the amount of the compound that is needed in the industry.

The strategic compounds (569 in this case) are well-connected to molecules from waste streams by less than three reaction steps and could, for example, be converted to pharmaceutically active molecules. Renewable feedstocks could be transformed into these strategic molecules to be optimally used in the industry. According to the researchers, their method extends the use of reaction network analysis to the general problem of developing new reaction pathways based on new chemical feedstocks.

• Identification of Strategic Molecules for Future Circular Supply Chains Using Large Reaction Networks,
  Alexei Lapkin, Pietro Lio’, Jana Weber,
  ChemRxiv 2019.
  https://doi.org/10.26434/chemrxiv.8488934.v1

  The research has been published as a preprint and has not yet been peer-reviewed.