(208f) Rufinamide/1,2,3-Triazole Preparation with a Functionalized Micronano System: A Combined Simulation and Experimental Approach

An entirely new reactor concept for multi-step organic reactions and particularly homogeneous-bio catalysis cascades is presented based on micro-flow continuous processing. Functional multi-phase solvent and nanoparticle combinations are developed to provide a compartmentalized flow reactor/separator system with ‘horizontal hierarchy’ – as opposed to the ‘vertical hierarchy’ of common multi-step flow syntheses (or batches) with their consecutive reactors-separators. Such flow cascade processing ideally needs just one reactor passage (‘ONE-FLOW’; www.one-flow.org). This ‘Green Spaciant Factory’ will fluidically open and close interim reaction compartments. The tasks are (a) orthogonality during reaction, (b) recycling of catalysts and reactants, (c) purification of products, (d) enable high-c processing, (e) ensure activity and stability of the catalysts.

This work aims to develop a novel synthetic route to prepare a Rufinamide and its precursor 1,2,3-triazole in (shown as Scheme 1) by introducing functionalized polymersomes as nanoreactors and metal catalyst separators into a continuous flow process. As for the current technologies of the 1,2,3-triazoles preparation, multiple steps are required to get the purified product; meanwhile, these processes are mostly high-energy consuming, such as high temperature and high pressure. [1,2] To achieve the multiple step reaction as a one-flow process under environmentally-friendly conditions, both functional reaction solvents and nanoreactors for catalyst compartmentalization are investigated.

For the reaction solvent selection, we used a combined COSMO-RS and experimental methodology to screen out the solvent candidates, seen in Scheme 2. Firstly, the reactant and product solubilities at different temperatures in a large number of solvents were determined separately by an auxiliary batch-processing program in COSMOthermX. Then, as the key scenario, the solubility variation between reactants and product were calculated and used as the solubility constraint. Afterwards. the reactivity potential of different families of solvents with the reactants was considered to narrow the screening space. Furthermore, relevant physical properties and the environmental/economic effect of solvents were taken into account to screen out the suitable solvents. Afterwards acetonitrile was recognized as the top solvent and validatedby experiments. With a good agreement with COSMO-RS results, acetonitrile is proved with a high selectivity between reactants and product at room temperature, which can dissolve all reactants except the final product. But with the temperature increasing, the selectivity gap decreasing which can guarantee the possible homogeneous catalysis during the reactions. Notably, without COSMO-RS calculation, the selected acetonitrile also totally dissolved the benzyl azide which is an intermediate synthesized from benzyl chloride. It is mainly due to the similar polarities with these two molecules.

The Rufinamide synthesis consists of two parts (scheme 1). First the benzyl chloride is azidated, followed by a Huisgen cycloaddition reaction catalyzed by Cu(I) (CuAAC). For the first part of the reaction, resin-N₃ immobilization was performed through ion exchange in batch, resulting in an azide loading capacity of 3.9 mmol/g. Then, the synthesis of the 2,6-difluorobenzyl azide from 2,6-difluorobenzyl chloride was optimized with the Resin-N₃ in both batch and flow process. Under these conditions the Resin-N₃ reactor reached 95% conversion within 1 hour reaction time in flow. The resin could also be recovered easily. For the second reaction, the CuAAC, a nanoreactor based on Cu-bis(oxazoline) loaded polymersomes [1] as employed. This was successfully used before for Cu(I) catalyzed reactions in batch, but has never been exploited in a microflow system, and whether or not it could work effectively in a pure organic solvent environment toward the synthesis of a pharmaceutical remains unknown. CuI crosslinked polymersomes were prepared as previously reported and characterized by TEM, EDX, XPS to check the polymersome morphology, Cu concentration and oxidation state under reaction conditions (Figure 1).

The Huisgen cycloaddition was first performed in batch using CuI powder, CuI ligand, and CuI crosslinked polymersomes with the 2,6-difluorobenzyl azide and phenylacetylene as substrates. All three CuI compunds as catalyst were fully converted after 2 hours and after cooling down, the triazole product spontaneously precipitated. Moreover, in contrast to the other two CuI catalysts, the sample with CuI crosslinked polymersomes formed a turbid organic layer, which contained as a recycle flow residual reactant and catalyst (Figure 2). Currently, the two-step process is optimized in flow.

Aceknowledgement

The research work is supported by FET-Open EU project ONE-FLOW (grant no. 737266).

Reference

[1] Borukhova S, Noël T, Metten B, et al. Green Chemistry, 2016, 18(18): 4947-4953.

[2] Britton J, Raston C L. Chemical Society Reviews, 2017, 46(5): 1250-1271.

[3] van Oers M. C. M., Abdelmohsen, L. K., van Hest J. C. M., Rutjes F. P. J. T. Chemical Communications, 2014, 50(31), 4040-4043.