A review of automated and data-driven approaches for pathway determination and reaction monitoring in complex chemical systems

Abstract

In this work, we review the state of the art on approaches for the determination of reaction networks and the real-time monitoring of reactions in complex chemical systems consisting of multiple reactive components using automated and data-driven methods. This complexity of the system results in uncertainty about both the dominant species and reactions in the system. Automated approaches to reaction network or pathway determination include rule-based or algorithmically extracted methods, quantum mechanical simulations, and machine learning approaches. We also identify the effect of explicit pathway determination on the approach for reaction monitoring. Furthermore, we compare and contrast the automated and data-driven approaches for reaction pathway determination with some heuristics commonly used to develop reaction mechanisms in complex chemistries.