ABSTRACT Automated flow platforms are well‐established in the context of chemical reaction optimization leveraging techniques such as Design of Experiments and self‐optimization. However, the development of such platforms in the context of kinetic investigations proves challenging, as an underlying mechanistic model needs to be identified. In order to address these challenges, we have developed an automated dynamic flow experimentation platform to automatically fit and identify the most accurate model. The effectiveness of this platform was successfully demonstrated on three complex transition metal catalyzed transformations (Buchwald‐Hartwig reaction, Re‐catalyzed oxygen atom transfer and Cu‐catalyzed C─H activation), automatically performing dynamic flow experiments, automatically fitting the kinetic parameters and independently identifying the appropriate kinetic model from a set of candidates. The obtained models were subsequently optimized using multi‐objective Bayesian optimization and both Pareto‐optimal and non‐Pareto‐optimal points from each of the models were seamlessly transferred to continuous flow to validate the workflows efficacy.
Wagner et al. (Fri,) studied this question.