Radical Redox-Relay Catalysis: Formal [3+2] Cycloaddition of N

Aug 21, 2017 - Providing Rapid, Low-Cost Analysis of Tap Water Using the 'Coffee-ring effect'. Determining chemical makeup of drinking water is an inc...
16 downloads 28 Views 1MB Size
Communication pubs.acs.org/JACS

Radical Redox-Relay Catalysis: Formal [3+2] Cycloaddition of N‑Acylaziridines and Alkenes Wei Hao, Xiangyu Wu, James Z. Sun, Juno C. Siu, Samantha N. MacMillan, and Song Lin* Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York 14853, United States S Supporting Information *

that adopt a single-electron transfer pathway favoring homolysis and functionalization of the more sterically hindered CN bond.10 In these examples, however, suitable substrates are primarily terminal and styrene-derived aziridines. In this work, we report a redox-metal-catalyzed radical [3+2] cycloaddition of N-acylaziridines and alkenes that leads to substituted pyrrolidines, structural motifs common in natural products, synthetic therapeutics, and small-molecule catalysts (Scheme 1C).

ABSTRACT: We report Ti-catalyzed radical formal [3+2] cycloadditions of N-acylaziridines and alkenes. This method provides an efficient approach to the synthesis of pyrrolidines, structural units prevalent in bioactive compounds and organocatalysts, from readily available starting materials. The overall redox-neutral reaction was achieved via a redox-relay mechanism, which harnesses radical intermediates for selective CN bond cleavage and formation.

Scheme 1. Ring Opening of Aziridines: Two-Electron versus Radical Redox-Relay Approach

O

wing to the presence of nitrogen-containing motifs in the vast majority of medicinally relevant synthetic targets, the development of efficient, selective, and sustainable technologies for constructing these organic structures is of critical importance. The ring opening of aziridines represents an attractive approach for the synthesis of novel nitrogenous molecules. The tendency of these strained heterocycles to rupture at two distinct reactive sites offers unique opportunities for the efficient introduction of new functionalities.1 Classical methods and recent developments in alkene aziridination have rendered these synthetic intermediates readily available from simple and abundant feedstocks.2 Current strategies for ringopening functionalization of aziridines rely primarily on closedshell, two-electron mechanisms mediated by acid, base, or transition metal catalysts (Scheme 1A,B).2 These protocols are restricted primarily to nucleophilic addition,3 insertion of unsaturated molecules,4,5 isomerization,6 and cross-coupling, however.7 Furthermore, the regioselectivity of these reactions usually favors cleavage and functionalization of the less substituted CN bond.8 As such, the N-containing structures accessible using existing approaches are limited. Radical-mediated chemistry offers a unique platform on which to expand the scope of transformations involving aziridines. Compared with their closed-shell counterparts, radicals usually exhibit exceptionally high reactivity and a distinct pattern of selectivity.9 Owing to the enhanced stability of carbon radicals with higher degrees of substitution, aziridine ring opening mediated by radical intermediates might display regioselectivity that differs from that of conventional electrophilic or nucleophilic activation. Therefore, a radical-mediated approach would enable the preferential cleavage and subsequent functionalization of the more substituted CN bond in the three-membered azacycle, thereby allowing access to a broader array of N-containing molecules with diverse structures starting from aziridines. Indeed, recent advances have shown that Ni catalysts enable aziridine-based cross-coupling reactions © 2017 American Chemical Society

We envisioned that the overall redox-neutral cycloaddition of aziridines and alkenes could be achieved in a tandem redoxrelay sequence consisting of single-electron reduction, radical addition, and radical oxidation (Scheme 2). Specifically, a redox-active metal complex could induce homolytic cleavage of the more substituted CN bond through coordination to Nacylaziridine I and subsequent single-electron transfer from the metal center to the organic substrate. This process would yield a carbon radical tethered to an azaenolate (II). Addition of the Received: June 28, 2017 Published: August 21, 2017 12141

DOI: 10.1021/jacs.7b06723 J. Am. Chem. Soc. 2017, 139, 12141−12144

Communication

Journal of the American Chemical Society Scheme 2. Proposed Catalytic Cycle

Table 1. Ti-Catalyzed [3+2] Cycloaddition under Various Reaction Conditions

a

radical intermediate to an alkene (III) would then result in the formation of a new CC bond. The incipient radical adduct IV would subsequently undergo cyclization to furnish a new CN bond (V) and thus complete the [3+2] cycloaddition. In this step, the metal complex would be reduced and released from the pyrrolidine product to regenerate the active catalyst. This proposed strategy was reminiscent of atom-transfer radical reactions in which a carbon−halogen bond is homolyzed and the resultant radical equivalents are added across an alkene C C π-bond.11 In this proposed redox-relay catalytic cycle, reductive aziridine ring opening and oxidative ring closure are the key steps that would likely determine the efficiency of the radical catalysis. We drew inspiration from recent research demonstrating that redox-active transition metals such as TiIII complexes enable facile cleavage of strong NH bonds in carbamates.12,13 Density functional theory calculations suggested that an analogous approach would also be feasible in facilitating the homolysis of the aziridine CN bond, as its bond dissociation free energy (BDFE) decreases substantially when bound to TiCl3 via the acyl O atom (ΔBDFE = 22.4 kcal/mol). Our scheme involved pyrrolidine formation via the addition of a carbon radical to an azaenolate (IV to V), a reaction step with no literature precedent. However, structurally analogous Tienolates have been shown excellent radical acceptors in CC bond-forming transformations.14,15 These seminal advances served as the foundation for our proposed redox-metalcatalyzed aziridine [3+2] cycloaddition. We first explored the ring opening of N-benzoyl-2,2dimethylaziridine (1) and its addition to t-butyl acrylate (2) to construct t-butyl N-benzoyl-4,4-dimethylprolinate (3). A series of redox-active complexes with oxophilic early transition metal centers were surveyed, and lead results were obtained with Ti catalysts bearing cyclopentadienyl (Cp) ligands. Multiple iterations of optimization resulted in a highly efficient reaction system comprising Cp*TiCl3 (Cp* = pentamethylcyclopentadienyl) (5 mol %) as the catalyst precursor and Zn dust (10 mol %) as the reductant to generate the active TiIII species. In toluene at ambient temperature, 3 was formed in 94% NMR yield with complete regioselectivity with respect to both aziridine ring opening and alkene addition (Table 1, entry 1).16 When less electron-rich CpTiCl3 (Ered = −0.82 V vs Fc+/Fc) was used instead of Cp*TiCl3 (Ered = −1.10 V) under identical

entry

variation from standard conditions

yield (%)a

1 2 3 4 5 6 7 8 9 10 11

none CpTiCl3 instead of Cp*TiCl3 Cp2TiCl2 instead of Cp*TiCl3 TiCl4 instead of Cp*TiCl3 without Zn dust Mn dust instead of Zn dust ZnCl2 instead of Cp*TiCl3 and Zn dust DCM instead of toluene THF or MeCN instead of toluene 1.0 equiv 2 4 or 5 instead of 1

94 20b