Can a Simple Surrogate Model System Be Used to Develop a Continuous Flow Packed Bed Hydrogenation for a Complex Molecule?

Catalytic hydrogenations are key processes in the fine chemical and pharmaceutical industries, but the development of such processes is challenging due to aspects such as catalyst deactivation, metal leaching, mass transfer limitations, solubility issues, and the formation of side products. Processes are particularly difficult to develop when a substrate is a large molecule containing multiple functional groups. These difficulties are significant obstacles for the identification of robust operating conditions; thus, workflows are necessary to speed up development timelines. The use of a more cost-effective and commercially available surrogate in development is an alternative strategy to find the optimized conditions, which can then be subsequently validated on the real molecule only at a later stage in development. The approach we apply herein is designed to use less of the real compound while minimizing the perceived risk of failure when transferring the conditions to the complex molecule. In this article, we apply our workflow for the catalytic hydrogenolysis of a large glycopeptide molecule, Cbz-protected glycopeptide (Cbz-GP), in a packed bed reactor. As part of the workflow, we use a robustness screening approach, introduced by Collins and Glorius, to show that a surrogate molecule, Cbz-protected lysine (Cbz-Lys), in the presence of additives can mimic secondary functional groups present in Cbz-GP or represent residual impurities generated upstream in the synthesis of Cbz-GP. The data generated for Cbz-Lys enabled the identification of the operating conditions for the successful deprotection of Cbz-GP after minor modification. Gratifyingly, only a few additional experiments were necessary using the Cbz-protected GP molecule to modify the conditions to achieve >95% conversion under mild conditions and within <10 s of contact time for stable performance over >6 h operation time.