Mémoire, Partim A, COLLÉGIALITÉ

Abstract

Cancer is one of the leading causes of death worldwide. Thus, the administration of the appropriate treatment is of crucial importance to guarantee the recovery of millions of people each year. Cyclophosphamide, mentioned on the WHO List of Essential Medicines, is a phosphorus-based chemotherapy drug frequently used in the treatment of a vast array of cancers. It targets the DNA of cancerous cells as an alkylating nitrogen mustard derivative. It is also an immunosuppressive agent for various disorders of the immune system. The classical strategies for the batch synthesis of cyclophosphamide involve two steps which feature significantly long reaction times, hazardous and highly sensitive derivatives such as a nitrogen mustard derivative. The numerous assets of continuous flow, such as safety improvements and enhancement of reaction efficiency, could offer an alternative to the actual laborious production of cyclophosphamide. Herein, we report the flow synthesis of a non-chlorinated and safer analog of cyclophosphamide, which could be directly translated to the production of cyclophosphamide itself. First, we focused on its batch synthesis to obtain the analog in high purity as a reference in order to generate spectral data of novel compounds. We assessed 3 pathways: the traditional synthesis through a non-cyclic intermediate, the synthesis through a cyclic intermediate, and lastly a one-pot synthesis. Only the first route led to the desired analog in high purity. Next, a flow setup was established for this classical pathway. Both steps were conducted and optimized separately. N-methylimidazole and acetonitrile were first employed as the base and the solvent, respectively, to allow homogenous conditions through the formation of an ionic liquid, a necessary condition to enable smooth operation in continuous flow. The reaction conditions were modulated to reach optimal conversion and selectivity, however, this appeared insufficient. Subsequently, a batch screening of a multitude of bases and solvents was performed and led to an unprecedented process involving tributylamine, acetonitrile, and 2-methyltetrahydrofuran, which ensured homogenous conditions. Finally, we sought to concatenate the two optimized steps and achieved the synthesis of the cyclophosphamide analog in high conversion (94-99%) and selectivity (82-84%) in a fully concatenated system. Two other strategies, namely a continuous flow synthesis through a cyclic intermediate and a continuous flow process relying on a P(III) derivative, were assessed as well but did not lead to satisfactory results.