Transient N-Aziridinyl Radicals in Olefin Functionalization

Abstract

Aziridines are the smallest nitrogen-containing heterocycles and are responsible for the biological activity of aziridine-natural products and active pharmaceutical ingredients (APIs). Classically, aziridines are prepared from acyclic precursors via 1) [2+1] cycloaddition of a nitrene equivalent with an olefin, 2) [2+1] cycloaddition of a carbene equivalent with an imine, or 3) via intramolecular cyclization of β-functionalized amines. In comparison, introduction of intact aziridines is an uncommon disconnection. Here, we highlight the recent development of N-aziridinyl radicals as novel intermediates in synthetic chemistry that enable olefin hydroxyaziridination. These intermediates, generated by photoredox activation of N-pyridinium aziridine precursors, afford access to products of a heretofore unknown epoxide opening with N-aziridine nucleophiles and introduce new disconnections of olefin aziridination chemistry.

Graphical Abstract

1. Introduction

Aziridines are important synthetic intermediates and are pharmacophores in various bioactive small molecules.^[1] The activity of aziridines results from strain-enhanced ring-opening chemistry with a wide variety of nucleophiles.^[2] In contrast to epoxides, which are widespread in natural products, aziridines are much less frequently encountered. While epoxide-containing natural products can be generated by direct oxygen-atom transfer from reactive metal oxo species to olefinic substrates by enzyme-catalyzed processes, analogous nitrogen-group transfer reactions to generate aziridines are not endemic.^[3] As a result, the discovery, optimization, and application of bioactive aziridines demands the continued development of innovative synthetic methods.

2. Classical Aziridination Methods

While a variety of disconnections are possible for the construction of aziridines, the term “olefin aziridination” typically evokes cycloaddition chemistry between a nitrene equivalent and an olefinic substrate (Figure 1b). A wide variety of metal-catalyzed and metal-free processes have been developed to access aziridines via this disconnection.^[4] Historical challenges in olefin aziridination chemistry included nitrene transfer to unactivated (i.e., aliphatic) olefins and limited methods to access diverse families of exocyclic N-substituents.^[5] Recent progress has begun to address both challenges.^[6] Beyond olefin aziridination, carbene transfer to imines and cyclization of β-functionalized amine derivatives are commonly encountered disconnections for the assembly of aziridines^{[1b, 4a, 7]} Finally, biosynthesis of aziridines may proceed via β-C─H activation of amines, but this disconnection has not yet been translated to the synthetic milieu^[8]

Figure 1.

(a) Selected aziridine-containing natural products. (b) General strategies for aziridine synthesis via cycloaddition chemistry and comparison with the N-aziridinyl radical transfer chemistry highlighted here.

Inspired by the burgeoning synthetic literature of N-centered radicals to olefins,^[9],[10] we envisioned an alternate “olefin aziridination” disconnection based not on constructing aziridines from olefins, but by addition of an intact aziridine to olefins. Such a transformation could enable the introduction of aziridine fragments into complex organic molecules, could enable construction of novel families of azirdine-based structures, and could provide the chemical tools to rapidly evaluate the potential biological activity of aziridine-based small molecules. Access to this alternate aziridination paradigm required the development of appropriate precursors for N-aziridinyl radicals. Given the relative bond dissociation energies (BDEs) of the C─N (~54 kcal/mol)^[11] and N─H bonds (~92 kcal/mol) in aziridines, we did not view N─H aziridines as attractive N-aziridinyl radical precursors.^[12] Instead, we envisioned generation of N-aziridinyl radicals by activation of N-functionalized aziridines bearing a labile N-substituent. Here, we highlight our first foray into this chemistry, which was made possible by photoactivation of N-pyridinium aziridine precursors.^[13]

3. Aziridine Group Transfer

We initiated our discovery campaign by examining the activation of N-pyridinium aziridine^{[4b, 6c]} 1a (onset potential – 0.85 V vs. Fc⁺/Fc; tpp = triphenylpyridinium)^[14] by photolysis of a MeCN solution in the presence of Ir(ppy)₃ (Ir(III)*/Ir(IV) = −1.73 V vs. Fc⁺/Fc) and triethyl amine. We employed radical acceptor 2 in order to trap the putative N-aziridinyl radical intermediate. Under these conditions, we received compound 3, the product of aziridine transfer in 39% yield (Figure 2a). Support for the intermediacy of an N-aziridinyl radical was obtained by replacing 2 with N-tert-butyl-α-phenyl nitrone (PBN), a commonly encountered spin trap for electron paramagnetic resonance (EPR) spectroscopy. Under these conditions, persistent radical 4 was obtained and characterized both by EPR and high-resolution mass spectrometrty (Figure 2b).

Figure 2.

(a) Evidence of N-aziridinyl radical intermediates obtained by trapping with sulfone 2. (b) EPR data of PBN-trapped N-aziridinyl radical 4. (c) Addition of the N-aziridinyl radical derived from 1a to styrene effects olefin 1,2-hydroxyaziridination (tpp = 2,4,6-triphenyl pyridine). (d) Potential mechanism for aziridine-radical transfer. (e) Radical clock reaction with (1-cyclopropylvinyl)benzene. (f) Reaction performed with H₂¹⁸O to evaluate the source of hydroxyl group in the product.

To translate the developed conditions for the reliable generation of N-aziridinyl radicals to synthetic chemistry, we turned our attention to examining trapping these transient intermediates with olefinic substrates (Figure 2c). We envisioned that N-centered radical addition to an olefin would afford a C-centered radical. Trapping with O₂, followed by reduction of the incipient peroxide intermediate would ultimately afford 1,2-hydroxyaziridination products (Figure 2d). Optimization studies demonstrated the viability of the envisioned olefin hydroxyaziridination chemistry and identified important roles for both Et₃N and LiBr additives. Control experiments suggest that LiBr functions as a Lewis acid to enhance the electrophilicity of the N-aziridinyl radical. Application of the optimized conditions to styrene 5a afforded 6a in 65% yield (Figure 2c). Compound 6a can be envisioned as the product of epoxide opening with an aziridine nucleophile, which to our knowledge is an unknown transformation.

There is significant experimental evidence that olefin hydroxyaziridination chemistry proceeds via N-aziridinyl radical intermediates. First radical clock experiments were performed:^[15] Under an N₂ atmosphere, hydroxyaziridination of (1-cyclopropylvinyl)benzene afforded cyclopropyl ring opened product 7 (Figure 2e); in contrast, under an O₂ atmosphere, hydroxyaziridination of the same substrate afforded hydroxyaziridination product 8. The formation of 7 in the absence of O₂ suggests that following N-azirdinyl radical addition, cyclopropane opening and cyclization of the resulting primary C-radical is facile. Hydroxyaziridination of styrene (5a) in presence of H₂¹⁸O proceeded without significant ¹⁸O incorporation, which indicates that O₂ is the source of the hydroxyl group in this reaction. Stern-Volmer quenching studies were performed as a function of both [1a] and [Et₃N],^[16] and the obtained data indicated that 1a is the primary quencher for this reaction, consistent of reductive quenching of the excited state of Ir(ppy)₃ to generate N-aziridinyl radicals.

With optimized reactions conditions and significant evidence for the role of N-aziridinyl radicals in olefin 1,2-hydroxyaziridination, we evaluated the generality of this transformation. The tolerance for substitution of the olefinic partner (5) was queried using N-aziridinyl radical precursor 1a (Figure 3). para-Fluorinated (6b), para-chlorinated (6c), and ortho-brominated (6d) hydroxyaziridines are accessed in 56-74% yield. Electron-deficient substrates 5e-5g are well tolerated with the corresponding hydroxyaziridines being obtained in 56-64% yield; 3-nitrostyrene furnished the corresponding aziridine-addition product 6h in 81% isolated yield. α-Methyl- and phenyl-substituted styrene (5i and 5j) gave the hydroxylated products (6i and 6j) with moderate yields (41 and 44%, respectively). Though 1,1-disubstituted styrenes were competent substrates, 1,2-disubstituted derivatives were not productive coupling partners. Styrenes derived from important pharmaceutical such as ibuprofen provided the desired product 6k in 54% yield. Attempts to translate this chemistry to aliphatic olefins or enol ether derivatives have thus far been unsuccessful.

Figure 3.

Styrene scope. Conditions: 1a (0.1 mmol), 5a-5k (0.15 mmol), Et₃N (0.2 mmol), Ir(ppy)₃ (1.0 mol %), LiBr (0.1 mmol), blue LED, 23°C in MeCN:H₂O (1:1, 2.0 mL). Aziridine scope. Conditions: 1l–1u (0.1 mmol), 5a (0.15 mmol), Et₃N (0.2 mmol), Ir(ppy)₃ (1.0 mol %), LiBr(0.1 mmol), blue LED, 23 °C in MeCN:H₂O (1:1, 2.0 mL).

Using styrene as the olefinic partner, we next evaluated the diversity of N-aziridinyl radicals that could be engaged in olefin 1,2-hydroxyaziridination. These studies revealed that electrondeficient radical precursors (6l-6o) delivered addition products in moderate to good yields. However, N-pyridinium aziridines with electron-donating substituents, such as 1p and 1q, were less efficient, delivering hydoxyaziridines 6p and 6q in 46 and 42% yield, respectively. Aziridine 1r, derived from 2-vinyl benzothiophene is also compatible with the aziridine-transfer protocol, delivering 6r with 57% yield. Moreover, N-pyridinium aziridines derived from various pharmaceutical scaffolds such as indomethacin (1s), tufnil (1t) and probenecid (1u) all engage is efficient aziridine transfer chemistry (52-65% yields).

Consistent with the reductive quenching mechanism indicated by Stern-Volmer studies (vide supra) and the structure-dependent reduction potentials of N-pyridinium aziridines,^[14] N-aziridinyl radical synthesis from aziridines derived from aliphatic olefins is less efficient than those derived from styrenes (Figure 4). Photoactivation of cyclohexene-derived N-pyridinium aziridine 9 in the presence of styrene (5a) afforded hydroxyaziridinated product 10a in 42% yield. Blue-light irradiation of 9 in the presence of various styrene derivatives afforded products 10b-10d in 33-48% yield. Complex styrene derivatives, derived from indomethacin, could be engaged similarly, albeit the yield of the product 10e was isolated with only 18% yield. Finally, access to ethylene-derived N-pyridinium aziridine 11 provided the opportunity to evaluate the transfer of the simplest, completely unsubstituted N-aziridinyl radical. 1,2-Hydroxyaziridination of styrene with an unsubstituted aziridine radical affords 2-hydroxy-2-phenyl-1-aziridinoethane (12).

Figure 4.

(a) Conditions: 9 or 11 (0.1 mmol), 5 (0.15 mmol), Et₃N (0.2mmol), Ir(ppy)₃ (1.0 mol %), LiBr (0.1 mmol), blue LED, 23 °C in MeCN:H₂O (1:1, 2.0 mL). Isolated yields. Unsubstituted aziridine 11 was engaged to afford hydroxyazirinated product 12, which demonstrates transfer of the simplest, unsubstituted aziridinyl radical.

4. Conclusions

In summary, we have developed robust conditions to unveil N-aziridinyl radicals from N-pyridinium aziridine precursors. This discovery enabled the development of a new olefin aziridination disconnection in which an intact aziridine fragment is added to the olefinic substrate. We envision that access to N-aziridinyl radicals will unlock the development of a variety of other novel reaction pathways, from olefin hydroaziridination to aziridine transfer to aromatics and thus significantly expand the scope of substrates that can engage in productive aziridination chemistry.

Biography

David C. Powers received a B.A. from Franklin and Marshall College, a Ph.D. from Harvard University, and pursued pursued post-doctoral research at the Massachusetts Institute of Technology and Harvard University in the laboratory of Daniel G. Nocera. He joined the faculty at Texas A&M in 2015. His research interests include hypervalent iodine chemistry, reactive intermediates in C─H functionalization, and in crystallo photochemistry. Promita Biswas received a Ph.D. from IACS–Kolkata and pursued post-doctoral research at Texas A&M. She is currently a post-doctoral scholar at Purdue University in the laboratory of Christopher Uyeda.