What and how can machine learning help to decipher mechanisms in molecular electrochemistry?

Abstract

Electrochemists have long been dedicated to heuristically analyzing electrochemical data with meticulous visual inspection and striving to deterministically assign reaction mechanisms. We contend that machine learning (ML) offers a new approach of mechanistic analysis with high data throughput and minimal human intervention. In this perspective, we propose that the deployment of ML in electrochemistry will enable a probability-driven mechanistic analysis amid the inevitable mechanistic ambiguity. We will discuss examples of ML deployment in electroanalysis, enlist current challenges for experimentalists, and discuss ML's prospects in molecular electroanalysis. We hope such a discussion will promote and advance ML-aided mechanistic deciphering for electrochemical systems in the long run.