ACS PRF | ACS
All e-Annual Reports
44783-G1
Lewis Acid Activation of Oxaziridines
Overview. The interaction between a drug and its target protein is typically mediated by hydrogen bonds between oxygen- and nitrogen-containing functional groups on the drug and receptors on the protein. The specificity and strength of this interaction is dictated by the arrangement of the drug's functional groups in space. The ultimate starting materials for most synthetic organic compounds, however, are simple petrochemical hydrocarbons (e.g., alkanes, alkenes, and arenes) that do not bear functional groups, cannot hydrogen bond, and are not stereochemically well-defined. Thus, a fundamental challenge in synthetic organic chemistry is the development of powerful, general methods for the selective installation of functional groups onto unfunctionalized hydrocarbon substrates with precise control over the position and spatial arrangement of these new functionalities.
Initially, we attempted to address this important problem by examining the ability of transition metal catalysts to control the oxidative functionalization of hydrocarbons. However, the instability of dioxiranes towards isolation and the aqueous conditions under which they are generated impeded our early investigations. We have elected instead to examine the reactions of structurally analogous oxaziridines, which perform many of the same reactions as dioxiranes. In contrast to dioxiranes, however, oxaziridines are stable towards isolation, and the ability to investigate the chemistry of oxaziridines under controlled conditions encouraged us to examine their reactivity as a model system.
General strategy. At the outset of our research into the chemistry of oxaziridines, we focused upon two central challenges: (1) Could we develop electron-deficient catalysts that would increase the overall reactivity of oxaziridines towards hydrocarbon substrates? (2) Can catalysts fundamentally alter the way in which oxaziridines and hydrocarbons engage one another in order to install both oxygen and nitrogen functional groups from the oxaziridine into the final products?
Our initial investigations are summarized in Table 1. As expected, the uncatalyzed reaction of styrene (1) with N-sulfonyl oxaziridine 2 proceeds sluggishly to produce trace quantities of styrene oxide (3). Upon addition of 20 mol% of AcOH, the rate of epoxidation was increased four-fold, which served as a valuable validation of our hypothesis that electron-deficient catalysts could increase the reactivity of oxaziridines. Upon addition of 2 mol% of Cu(TFA)2, however, we observed the regioselective formation of aminal 4 instead of the epoxide. Finally, the use of 20 mol% of TiCl4 as the catalyst produced isoxazolidine 5 instead of either the previously observed epoxide or aminal products. This study, therefore, constituted an unusual example of a system exhibiting divergent, catalyst-controlled reactivity.
Copper-catalyzed aminohydroxylation. 1,2-Aminoalcohols are ubiquitous substructures in biologically active organic compounds, and the osmium-catalyzed Sharpless aminohydroxylation has been the method of choice for the construction of this important structural motif for over a decade. Nevertheless, the Sharpless method is often poorly regioselective, and the cost and toxicity associated with the use of osmium catalysts has limited the practical utility of this process in large-scale applications.
Our new copper-catalyzed process offers a promising alternative to the powerful Sharpless method. Reactions using a variety of electron-rich, electron-poor, polysubstituted, cyclic, and acyclic styrenes are high-yielding, and in all cases product a single regioisomer of the protected 1,2-aminoalcohol. Two features of this new methodology are particularly notable: (1) Unlike osmium catalysts, copper(II) salts are inexpensive, abundant, and environmentally innocuous. (2) The Sharpless method tends to give mixtures of regioisomers, particularly in aminohydroxylations of styrenes, which reduces the yield of the reaction and necessitates a tedious separation of the isomers. In contrast, we have only observed single regioisomers in all reactions of styrenes we have conducted to date.
[3+2]-Dipolar cycloaddition. We have determined that 1,2-isoxazolidine 5 arises from TiCl4-catalyzed rearrangement of the oxaziridine to a transient N-sulfonyl nitrone, which undergoes cycloaddition with the olefin to produce 5. Further optimization of the reaction showed that the yield and stereoselectivity of the reaction could be significantly improved using N-nosyl oxaziridines (Ns = 4-NO2-PhSO2). A variety of olefins and substituted oxaziridines participate in this new cycloaddition, produce a range of structurally diverse 1,2-isoxazolidines in good yields and selectivities.
The sensitive nitrogen-oxygen bond of isoxazolidines can be easily cleaved to reveal synthetically valuable 1,3-aminoalcohols. Interest in the ability to access this important substitution pattern has produced an enormous body of literature regarding cycloadditions of conventional N-alkyl nitrones. More recently, isoxazolidines themselves have begun to attract attention as potential anti-cancer agents, anti-retroviral drugs, and biological probes; however, exploration of their biological activity has been hampered by the inability to remove N-alkyl groups from isoxazolidines without cleavage of the sensitive N-O bond. In contrast, the N-nosyl group can selectively be removed under mild conditions, easily producing the otherwise inaccessible N-unsubstituted isoxazolidines. The N-sulfonyl oxaziridines from which these are derived constitute a new class of highly electron-deficient nitrones bearing electron-withdrawing nitrogen substituents, and our methodology provides the first general method for their synthesis.
