Traditional chemical manufacturing significantly contributes to climate change and is difficult to decarbonize, relying heavily on fossil fuels for both feedstocks and energy. Electrocatalytic transformations of biomass-derived feedstocks offer a promising alternative, but implementing such technology in real settings requires a deep understanding and precise control of various reaction parameters that influence reaction selectivity and efficiency.

In the initial phase, the team surveyed how the parameters of current density, electrode texture, substrate concentration, and solvent composition affect the Faradaic efficiency for ADN production. They conducted over 100 parallel experiments using high-throughput combinatorial electrochemical testing and analysed the data with a Gaussian process regression model. “This was the first time we combined our liquid-handling robots with high-throughput electrochemical cells. We were able to perform more than 150 experiments in a week without even optimizing the protocol,” explains Modestino. “Regression models help us to identify trends that can be otherwise hard to spot and we visualize them on multi-dimensional reactivity maps.” The optimal combination of conditions that maximized Faradaic efficiency were found to be: high current density, high CPA concentration, low water content in the electrolyte, and the use of Pt electrodes.