Operation and Optimization of Microwave-Heated Continuous-Flow Microfluidics

Microwave (MW) technology can be powerful for electrification and process intensification but limited fundamental understanding of scalability and design principles hinders its effective use. In this work, we build a continuous-flow microreactor inside a commercial single-mode MW applicator and the corresponding computational fluid dynamics model to simulate the temperature profile. The model is in good agreement with experiments for various microreactor dimensions and operating conditions. The model indicates that MW heating is greatly influenced by reactor geometry as well as the operating parameters. We observe a strong correlation between parameters and develop a gradient boost regression tree model to predict the outlet temperature accurately. This model is then applied to optimize the dimensions and operating conditions to maximize the outlet temperature and energy efficiency, resulting in a Pareto optimal. We demonstrate computationally and experimentally that it is possible to surpass the Pareto optimal and achieve an energy efficiency of \~90% or greater at temperatures relevant for liquid-phase chemistry via salting of the solvent. The present methodology can be applied to other complex MW reactors. The combined numerical and experimental approach provides insights into and a framework for scale-up and optimization.