Autonomous reaction self-optimization using in-line high-field NMR spectroscopy

Abstract

Autonomous self-optimization in flow is a powerful approach to efficiently optimize chemical transformations in a high dimensional space. Self-optimizing flow reactors combine automated flow devices with feedback optimization algorithms, which are powered by process analytical technology. In this contribution, we introduce the concept of autonomous self-optimizing flow reactors guided by in-line high-field NMR spectroscopy. We designed an autonomous experimental setup, combining an automated flow reactor with a high-field NMR spectrometer and a feedback optimization algorithm. User-friendly interfaces were developed for straightforward input of experimental parameters and precise control of equipment. Using 1D ¹H NMR spectroscopy with a solvent suppression method, we achieved accurate quantitative measurements. Self-optimization utilizing the Nelder–Mead algorithm to maximize either the yield or the throughput of a formal [3 + 3] cycloaddition was conducted through the fine-tuning of the residence time, stoichiometry, and catalyst loading as input variables. The integration of high-field NMR within autonomous flow systems promises enhanced precision and efficiency in chemical synthesis optimization, particularly for complex reaction mixtures, setting the stage for advances in chemical synthesis.