Prof. Vasco Bonifácio (DBE, iBB) and collaborators from the Faculty of Pharmacy at the University of Lisbon, the Faculty of Sciences at the University of Porto, the University of Southern California, National Yang-Ming Chiao Tung University and Insilico Medicine Taiwan Ltd have published in Green Chemistry a pioneering method that employs active learning to optimise biomass valorisation. Under these optimised conditions, 3-acetamido-5-acetylfuran (3A5AF) was isolated from N-acetylglucosamine (NAG) in 66% yield, with the reaction medium being reused for up to eight cycles without any significant loss of efficiency.

Moreover, a one-step conversion of chitin (the NAG polymer), pre-processed in a planetary ball mill, afforded 3A5AF in 37% yield—approximately double that of the best method previously reported in the literature. Crucially, this mechanochemical pre-treatment rendered the process more efficient: 10.5 mg of 3A5AF were obtained per gram of shrimp shell, underscoring its industrial potential.

This study further demonstrates the decisive role of machine learning in uncovering new applications for biomass and in advancing sustainable chemistry. See more.