Cyanomethylation of Aldehydes on an Electrochemical Microflow System and Utility of Machine Learning-Assisted Examination of the Reaction Conditions.

Cyanomethylation of carbonyl compounds affords a β-hydroxy nitrile, which is very useful in the field of synthetic organic chemistry. Although several electrochemical methods have also been reported to achieve cyanomethylation of carbonyl compounds, dehydration as a major side reaction is a significant issue that needs to be prevented. An electrochemical flow reactor, free from side reactions through rapid removal of the reaction solution from the electrode surface, enables the electrochemical cyanomethylation of aldehydes in acetonitrile without any dehydrated side products. To find the "best" reaction conditions for each starting material, we used machine learning models. In addition, comparing the machine learning models constructed using various molecular descriptors contributed to understanding the critical molecular properties to proceed the electrochemical cyanomethylation. Our cyanomethylation using a flow system went to completion with a semi-catalytic amount of electrical input, which is not consistent with the conventional basic mechanism, and DFT calculations suggest a radical mechanism.