SummaryThe performance of the ThalesNano H-Cube®, a commercial packed bed flow hydrogenator, was evaluated in the context of small scale reaction screening and optimization. A model reaction, the reduction of styrene to ethylbenzene through a 10% Pd/C catalyst bed, was used to examine performance at various pressure settings, over sequential runs, and with commercial catalyst cartridges. In addition, the consistency of the hydrogen flow was indirectly measured by in-line UV spectroscopy. Finally, system contamination due to catalyst leaching, and the resolution of this issue, is described. The impact of these factors on the run-to-run reproducibility of the H-Cube® reactor for screening and reaction optimization is discussed.

A well known advantage of flow chemistry reactors in chemical synthesis is the ability to screen multiple catalysts and reaction parameters with optimal conditions scaled accordingly. This approach, however, consumes significant quantities of material as the reactor must be equilibrated with the reactants in a continuous, steady-state mode before the start of the reaction. In this work we explore a screening and reaction approach using bolus injections, which is more conducive to the lower material consumption that may be required in a drug discovery setting. A commercially available ThalesNano H-Cube® was evaluated to determine the practicality of this approach for heterogeneous hydrogenations. When working with boluses in flow systems, one of the biggest limitations can be the inherent dispersion of the reactant stream caused by the reactor. The dispersion on the H-Cube® was characterized to determine the minimum volume for the reactor to reach a steady-state. The H-Cube® fluidics and heating coil were found to generate significantly more dispersion than the reaction cartridge (CatCart®) itself, increasing the minimum volume of injection required to achieve steady-state. A 2 mL injection was found as a good compromise between maximizing material conservation and sufficient volume of reaction at steady-state condition. Conditions optimized at 2 mL screening scale were successfully scaled five-fold, while lower volume bolus injections were shown to be less predictable. A stacked injection protocol using lower volume boluses was found to be a reliable alternative to scale reactions while efficiently conserving material. This application of small bolus injections to flow reaction screening and scale-up provides a desirable alternative to traditional continuous flow approaches in the material-limited discovery setting.