Periodic Reporting for period 1 - DESYPHOR (CATALYTIC ENANTIOSELECTIVE DESYMMETRIZATION OF PHOSPHORUS (V) CENTERS)
Période du rapport: 2021-08-15 au 2023-08-14
Compounds containing one or more phosphorous atoms in the P(V) oxidation state are important to chemistry, biology and medicine. These include marketed antiviral drugs such as Remdesivir, and Sofosbuvir; the former garnering interest as a potential treatment for COVID-19, and the latter being on the WHO list of essential medicines for the treatment of Hepatitis C (Figure 1). Accordingly, new and improved methods for the efficient synthesis of P(V) containing compounds, especially in an enantioselective fashion are essential. Although promising protocols are beginning to arise for the synthesis of racemic P(V) compounds, new strategic approaches for the stereoselective synthesis of P-stereogenic centers are limited and catalytic enantioselective approaches remain largely unknown. Despite these advances, direct enantioselective catalytic desymmetrization protocols involving reactivity directly at the P(V) remains unexplored. We envisioned a two-stage desymmetrization–derivatization strategy (Figure 2) by which enantiotopic phenolic leaving groups on a prochiral phosphonate ester are enantiodiscriminated by a suitable nucleophile under the control of a chiral catalyst, generating a new P−O bond. The resulting intermediate would retain suitable reactivity for sequential substitution of the remaining leaving group. With appropriate stereocontrol, this approach would overcome the key restrictions of previously developed protocols. Given the tunability and high basicity of our bifunctional iminophosphorane catalysts (BIMP), we expected that they would provide sufficient activation and could be suitably adapted to obtain high levels of enantiopurity in the products. Objectives: Therefore, the overall objectives of this project can be summarized as follows: a) To develop and explore the scope of the catalytic enantioselective nucleophilic desymmetrization reaction at P(V) with phenol pronucleophiles, including downstream derivatizations. b) To develop and explore the 2nd generation desymmetrization platform using thiazolidinone leaving group and application to a wide range of P(V) derivatives. Conclusion: In conclusion, a two-stage strategy for the synthesis of stereogenic P(V) compounds through an unprecedented enantioselective nucleophilic desymmetrization and subsequent enantiospecific derivatization, was developed (Scheme 1). A BIMP catalyst provided a unique chiral environment and sufficient pronucleophile/substrate activation to allow the desymmetrization to proceed with excellent yield and enantioselectivity. Through judicious choice of leaving group, facile downstream diversification of the desymmetrized P(V) ester with very high enantiospecificity was allowed. A 2nd generation enantioselective desymmetrisation at P(V) was developed (Scheme 2). Through the synergistic use of thiazolidinone leaving groups on P and fine tuning of the BIMP catalyst numerous drawbacks of previously established methods have been overcome. Now, phenols with a wide array of substitution patterns can be readily employed as nucleophiles in the desymmetrisation step with both aryl and alkyl derived P(V) electrophiles as substrates. The resulting enantioenriched intermediates could be converted to an even greater range of distinct classes P(V) of compounds.
To explore the proposed reaction design, we began our investigations by identifying an appropriate model system, utilizing 2,4-dimethylphenol as the nucleophile and a BIMP superbase catalyst. After a comprehensive investigation of potential leaving groups, 2-methyl-6-nitrophenol gave optimal enantioselectivity (97:3 e.r.) albeit with a drop in reactivity. By conducting the reaction at room temperature in fluorobenzene and fine-tuning the basicity of the catalyst, the yield of the desymmetrized product was increased to 84% and its e.r. was found to be 96:4. With the optimized conditions in hand, we proceeded to explore the scope of this transformation. We were pleased to find that a variety of phenols possessing ortho-substituents were well tolerated and high yields and enantioselectivities were maintained. Bulky, naturally occurring phenols totarol and thymol were also competent nucleophiles. Nucleophiles bearing no ortho-substituent were found to afford the corresponding products in good yields but poor enantioselectivity. Next, we proceeded to vary the P-linked carbon substituent on the phosphonate ester electrophile. Such changes were broadly tolerated with aryl, methyl, higher alkyl and β-branched substituents maintaining high reaction efficiency and enantioselectivities. The reaction was also successful on a gram scale with no loss in yield or enantioselectivity. Having established the scope of the desymmetrization for both the nucleophilic and electrophilic components, we proceeded to assess the crucial 2nd stage of the strategy: the enantiospecific nucleophilic substitution of the remaining leaving group. We found that Lewis acid activation of the P(V) species, was highly efficient in promoting the second nucleophilic displacement. All four DNA nucleosides, hepatitis C treatment sofosbuvir analog as well as acetal-protected uridine and adenosine were successfully phosphorylated at the 5′-OH with moderate to good yields and >95:5 d.r. However, when using n-propanethiol as the nucleophile a different reaction conditions had to be used in order to access desired enantioenriched thiophosphonate ester. We then turned our attention to N-centred nucleophiles and were thrilled to find that when using N-Boc benzylamine as the nucleophile we could obtain phosphonamidite ester in 84% yield and 100% e.s. Cognizant of the strengths and limitations of our 1st generation strategy, the need for an improved 2nd generation method became clear. To achieve this, a full re-optimization of our 1st generation method was required. After an extensive evaluation of new leaving groups and BIMP catalysts, the use of thiazolidinone as a leaving group was found to be optimal. To our delight, our 2nd generation protocol could be carried out on a gram scale employing just 2.5 mol% of catalyst. Finally, the remaining leaving group can be readily displaced by medicinally relevant alkyl and aryl alcohols, amines, Grignards and thiols, providing divergent access to a wide range of enantioenriched P(V) species.
This fellowship has enabled great improvement in the state-of-the-art of active organic molecule synthesis through a powerful and effective synthetic strategy, raising the standing of EU chemistry within this field on a global scale. The study represents a clear strategic departure from previously established catalytic methods that rely on substrate engineering and do not easily allow for facile downstream modification to medicinally attractive molecules. Owing to the abundance of biologically relevant chiral phosphorous (V) compounds across medicinal and agrochemical sectors and the likely increase in the demand for such compounds over the coming years, our newly developed catalytic synthetic approaches would likely find numerous applications in large scale synthesis, library generation, late stage functionalisation, and drug molecule synthesis. Therefore, this fellowship has made a significant contribution to the field, and will be of great benefit to organic and medicinal chemists in the pharmaceutical, agrochemical and fine-chemicals industries in the EU, who have already shown interest for the projects developed in this fellowship.