Communication pubs.acs.org/JACS
Divergent Rhodium-Catalyzed Cyclization Reactions of Enoldiazoacetamides with Nitrosoarenes Qing-Qing Cheng, Marianne Lankelma, Daniel Wherritt, Hadi Arman, and Michael P. Doyle* Department of Chemistry, The University of Texas at San Antonio, San Antonio, Texas 78249, United States S Supporting Information *
Scheme 1. Divergent Cyclization Reactions of Enoldiazo Compounds and Nitrosoarenes
ABSTRACT: The first cyclization reactions of enoldiazo compounds with nitrosoarenes have been developed. Under the catalysis of rhodium(II) octanoate, [3 + 2]cyclization between enoldiazoacetamides and nitrosoarenes occurred through cleavages of the enol double bond and the amide bond, thus furnishing fully substituted 5isoxazolone derivatives. Upon changing the catalyst to rhodium(II) caprolactamate, the reaction pathway switched to an unprecedented formal [5 + 1]-cyclization that provided multifunctionalized 1,3-oxazin-4-ones with near exclusivity under otherwise identical conditions. Mechanistic studies uncovered distinct catalytic activities and reaction intermediates, which plausibly rationalized the novel reactivity and catalyst-controlled chemodivergence. Furthermore, a mechanism-inspired enantioselective rhodium-catalyzed reaction of γ-substituted enoldiazoacetamide with nitrosobenzene produced highly enantioenriched heterocycle-linked trialkylamine.
result in 1,3-dipolar cycloaddition of the metallo-enolcarbene2b,c onto the NO bond of nitrosoarenes,4c−e but the actual outcome was completely unexpected. Catalyst-controlled divergent cyclization reactions between enoldiazoacetamides and nitrosoarenes occurred with profound structural rearrangements: catalytic cleavages of the enol double bond and the amide bond allowed the construction of structurally rearranged [3 + 2]-cycloadducts (Scheme 1e), and with only a change of the catalyst, the reaction pathway switched to an unprecedented formal [5 + 1]-cyclization (Scheme 1f). Moreover, the heterocycles formed through both cyclization pathways, namely 5-isoxazolones6 and 1,3-oxazin-4-ones,7 are widespread functionalities in numerous biologically active natural products and pharmaceuticals. In our initial survey of the reaction, tert-butyldimethylsilyl (TBS)-protected enoldiazoacetamide 1a was combined with nitrosobenzene 2a in chloroform at room temperature, and a selection of commercially available metal complexes were evaluated as catalysts (Table 1). With copper(I) tetrafluoroborate, a mixture of products was obtained, from which 3aa and 4aa were isolated in 31% and 22% yields, respectively (entry 1). Spectral analyses informed us that these compounds were not those expected from simple cycloaddition; their unpredictable structures were established by NMR and X-ray diffraction analyses [Figures S2−S7, Supporting Information (SI)]. Reactions catalyzed by copper(I) triflate and palladium(II) chloride also furnished mixtures of the two products (entries 2 and 5), whereas copper(II) triflate and silver triflate
M
etal catalysis, which has dramatically improved the efficiency of organic synthesis, has made possible tremendous advances in cyclization reactions.1 Numerous dipolar organic reagents, including diazo and nitroso compounds, have been successfully employed in cyclization processes that produce a diversity of carbocyclic and heterocyclic compounds. In previous studies, enoldiazo compounds served as efficient three-carbon dipolar synthons in various formal [3 + n]-cycloaddition reactions (n = 1−5), mainly through metallo-enolcarbene formation, thus furnishing four- to eight-membered cyclic products (Scheme 1a).2 Alternatively, donor−acceptor-substituted cyclopropenes that were catalytically generated from enoldiazo compounds underwent [2 + m]-annulations (m = 3, 4) to provide cyclopropane-fused ring systems (Scheme 1b).3 Nitrosoarenes have also become ideal dipolar substrates for selected cyclization processes, forming N−O-containing heterocycles via 1,3-dipolar cycloaddition and hetero-Diels−Alder reactions (Scheme 1c).4 Additionally, nitrosoarenes have been employed in annulation reactions involving ortho-C(sp2)−H functionalization to afford indole and quinoline derivatives (Scheme 1d).5 Despite considerable attention given to cycloaddition reactions of enoldiazo compounds2 and to nitrosoarenes in their cyclization processes,4,5 reactions between enoldiazo compounds and nitrosoarenes have remained unknown. Based on prior reports,2,4 we thought that the combination of these two reactants in the presence of a suitable catalyst would © 2017 American Chemical Society
Received: June 6, 2017 Published: July 11, 2017 9839
DOI: 10.1021/jacs.7b05840 J. Am. Chem. Soc. 2017, 139, 9839−9842
Communication
Journal of the American Chemical Society
Besides catalysts, other reaction conditions (e.g., solvents) were carefully examined (Tables S2 and S3, SI). Among all the tested solvents, chloroform stood out as the optimal choice for both reaction pathways. The addition of 4 Å molecular sieves had a positive influence on the product yields, especially for the Rh2(oct)4-catalyzed process. With the optimized conditions in hand, we sought to explore the substrate scope of these divergent cyclization reactions, and the results are presented in Table 2. Enoldiazoacetamides bearing different amide or silyl moieties were all suitable reagents for both cyclization pathways: switchable chemoselectivities were effectively achieved by the choice of rhodium catalysts, albeit with slightly diminished yields in the case with a triisopropylsilyl (TIPS) protecting group (3aa−3ca and 4aa− 4ca). It should be noted that, although quinoline derivatives had been formed from [3 + 3]-cycloaddition between nitrosoarenes and alkenyldiazoacetates (Scheme 1d, n = 3),5d,8 and nitrone species had been generated via condensation reactions of nitrosoarenes with styryldiazoacetates5d or other diazo compounds,4b,5e these potential competing reaction products were not detected in the present investigation with enoldiazoacetamides. A variety of nitrosoarenes were then examined for their suitability in these cyclization reactions. In general, parasubstituted nitrosobenzenes performed well under standard reaction conditions, and various functional groups, such as bromo, fluoro, and ethylcarboxylate, were well tolerated by both catalytic systems (3ab−3ah and 4ab−4ah). Also, metaand ortho-substituted nitrosobenzenes smoothly underwent both cyclization processes to furnish the corresponding heterocyclic products in good yields with excellent chemoselectivities (3ai−3ak and 4ai−4ak). Notably, Rh2(oct)4catalyzed cyclizations exhibited consistently higher reaction rates than those catalyzed by Rh2(cap)4: the former were generally completed within 6 h, whereas the latter required overnight stirring. Moreover, to demonstrate the synthetic practicality of this methodology, both cyclization products 3aa and 4aa were successfully prepared on gram scale (Scheme S1, SI). Intrigued with the unprecedented reactivity and catalystcontrolled chemodivergence, we continued our study to gain more insight into the reaction mechanism.9 Coordination of
Table 1. Divergent Metal-Catalyzed Cyclization Reactions of Enoldiazoacetamide 1a with Nitrosobenzene 2a: Screening of Catalystsa
a Reaction conditions: catalyst/1a/2a = 0.002x:0.3:0.2 (mmol), with 4 Å MS (60 mg) in CHCl3 (3 mL) at rt for 24 h. bIsolated yields. n.d. = not detectable by 1H NMR and TLC of the reaction mixture.
decomposed the enoldiazoacetamide, but allowed recovery of nitrosobenzene (entries 3 and 4). To our delight, dirhodium(II) catalysts not only drastically increased the product yields but also provided switchable selectivity between the two cyclization pathways (entries 6−10). By using rhodium(II) octanoate [Rh2(oct)4] as the catalyst, fully substituted 5isoxazolone 3aa was isolated as the sole product in 81% yield (entry 7). When the catalyst was changed to rhodium(II) caprolactamate [Rh2(cap)4], multifunctionalized 1,3-oxazin-4one 4aa was obtained with near exclusivity under otherwise identical conditions (entry 10).
Table 2. Divergent Rhodium-Catalyzed Cyclization Reactions of Enoldiazoacetamides 1 with Nitrosoarenes 2: Substrate Scopea,b
a
Reaction conditions: Rh2(oct)4 or Rh2(cap)4/1/2 = 0.004:0.3:0.2 (mmol), with 4 Å MS (60 mg) in CHCl3 (3 mL) at rt for 6 h or 24 h. bIsolated yields, ratios between 3 and 4 determined by 1H NMR of the reaction mixtures. 9840
DOI: 10.1021/jacs.7b05840 J. Am. Chem. Soc. 2017, 139, 9839−9842
Communication
Journal of the American Chemical Society
capability to decompose enoldiazoacetamides (Scheme 2a). The participation of nitrosobenzene dramatically promoted the conversion of enoldiazoacetamide 1a (Scheme 2d, 90%, 16 h vs Scheme 2a, 18%, 16 h), but no cyclopropene 5a was observed during NMR monitoring (Scheme 2d). Instead, a 2acylaziridine intermediate 6aa accumulated and was consumed in the reaction mixture (Scheme 2d). Notably, 2-acylaziridine derivatives were previously reported to be generated from 4isoxazolines via rearrangement, and they could further undergo intramolecular transformations upon aziridine ring opening.11 In addition, control experiments with 2-acylaziridine 6aa demonstrated its uncatalyzed rearrangement to deliver product 4aa (Scheme 2e). On the basis of these results and previous reports,2d,3,10,11 a plausible mechanism for the divergent cyclization reactions is proposed (Scheme 3). Rh2(oct)4 not only catalyzes a tandem
nitrosobenzene to both Rh2 (oct) 4 and Rh 2(cap) 4 was demonstrated by NMR analyses (Scheme S2, SI).10 However, these two rhodium catalysts exhibited different activities with enoldiazoacetamides (Scheme 2a). Rh2(oct)4 efficiently cataScheme 2. Experimental Studies of Mechanism
Scheme 3. Proposed Mechanism of Divergent RhodiumCatalyzed Cyclization Reactions
dinitrogen extrusion/metallo-enolcarbene formation/intramolecular cyclization sequence with enoldiazoacetamide 1 to produce donor−acceptor-substituted cyclopropene 5,3 but this catalyst also activates nitrosoarene 2 for a subsequent azaMichael addition to the cyclopropene, thus furnishing intermediate III that further proceeds through a consecutive five-membered ring closure/cyclopropane opening/dialkylamino migration process to deliver product 3 (Scheme 3, top). Rather than directly decomposing the enoldiazoacetamide, Rh2(cap)4 activates the nitrosoarene for electrophilic attack at the diazo carbon, thus furnishing diazonium intermediate V that undergoes intramolecular nucleophilic addition to the vinylogous position.2d The resulting 4-isoxazoline intermediate VI rapidly rearranges to generate 2-acylaziridine 6;11 subsequent ring expansion, which is triggered by aziridine ring cleavage followed by silyl group migration and six-membered ring closure, defines the overall [5 + 1]-cyclization (Scheme 3, bottom). Inspired by the mechanistic study, especially by the proposed dialkylamino migration process that rationalized the amide bond cleavage (Scheme 3, IV → 3), we envisioned that by utilizing chiral catalysts and γ-substituted enoldiazoacetamides,
lyzed a rapid transformation from enoldiazoacetamide 1a to form donor−acceptor-substituted cyclopropene 5a, whether with (Scheme 2b) or without nitrosobenzene 2a (Scheme 2a). That cyclopropene 5a is a key intermediate was not only observed by NMR monitoring (Scheme 2b) but also verified by a subsequent experiment (Scheme 2c, x = 2), in which Rh2(oct)4 was found to significantly accelerate the generation of product 3aa from cyclopropene 5a (Scheme 2c, x = 0 vs x = 2). Compared to Rh2(oct)4, however, Rh2(cap)4 lacked the 9841
DOI: 10.1021/jacs.7b05840 J. Am. Chem. Soc. 2017, 139, 9839−9842
Communication
Journal of the American Chemical Society
(b) Smith, A. G.; Davies, H. M. L. J. Am. Chem. Soc. 2012, 134, 18241. (c) Jing, C.; Cheng, Q.-Q.; Deng, Y.; Arman, H.; Doyle, M. P. Org. Lett. 2016, 18, 4550. For [3 + 3], see: (d) Qian, Y.; Xu, X.; Wang, X.; Zavalij, P. Y.; Hu, W.; Doyle, M. P. Angew. Chem., Int. Ed. 2012, 51, 5900. (e) Xu, X.; Doyle, M. P. Acc. Chem. Res. 2014, 47, 1396. (f) Cheng, Q.-Q.; Yedoyan, J.; Arman, H.; Doyle, M. P. J. Am. Chem. Soc. 2016, 138, 44. For [3 + 4], see: (g) Reddy, R. P.; Davies, H. M. L. J. Am. Chem. Soc. 2007, 129, 10312. (h) Schwartz, B. D.; Denton, J. R.; Lian, Y.; Davies, H. M. L.; Williams, C. M. J. Am. Chem. Soc. 2009, 131, 8329. (i) Zhu, C.; Xu, G.; Sun, J. Angew. Chem., Int. Ed. 2016, 55, 11867. For [3 + 5], see: (j) Lee, D. J.; Ko, D.; Yoo, E. J. Angew. Chem., Int. Ed. 2015, 54, 13715. (3) For [2 + 3], see: (a) Cheng, Q.-Q.; Yedoyan, J.; Arman, H.; Doyle, M. P. Angew. Chem., Int. Ed. 2016, 55, 5573. For [2 + 4], see: (b) Davies, H. M. L.; Houser, J. H.; Thornley, C. J. Org. Chem. 1995, 60, 7529. (c) Deng, Y.; Pei, C.; Arman, H.; Dong, K.; Xu, X.; Doyle, M. P. Org. Lett. 2016, 18, 5884. (4) For [2 + 2], see: (a) Dochnahl, M.; Fu, G. C. Angew. Chem., Int. Ed. 2009, 48, 2391. For [2 + 3], see: (b) Xu, Z.-J.; Zhu, D.; Zeng, X.; Wang, F.; Tan, B.; Hou, Y.; Lv, Y.; Zhong, G. Chem. Commun. 2010, 46, 2504. (c) Chatterjee, I.; Fröhlich, R.; Studer, A. Angew. Chem., Int. Ed. 2011, 50, 11257. (d) Chen, C.-H.; Tsai, Y.-C.; Liu, R.-S. Angew. Chem., Int. Ed. 2013, 52, 4599. (e) Chakrabarty, S.; Chatterjee, I.; Wibbeling, B.; Daniliuc, C. G.; Studer, A. Angew. Chem., Int. Ed. 2014, 53, 5964. For [2 + 4], see: (f) Jana, C. K.; Studer, A. Angew. Chem., Int. Ed. 2007, 46, 6542. (g) Momiyama, N.; Yamamoto, Y.; Yamamoto, H. J. Am. Chem. Soc. 2007, 129, 1190. (h) Lu, M.; Zhu, D.; Lu, Y.; Hou, Y.; Tan, B.; Zhong, G. Angew. Chem., Int. Ed. 2008, 47, 10187. (i) Maji, B.; Yamamoto, H. J. Am. Chem. Soc. 2015, 137, 15957. (5) For [3 + 2], see: (a) Penoni, A.; Volkmann, J.; Nicholas, K. M. Org. Lett. 2002, 4, 699. (b) Murru, S.; Gallo, A. A.; Srivastava, R. S. ACS Catal. 2011, 1, 29. (c) Sharma, P.; Liu, R.-S. Org. Lett. 2016, 18, 412. For [3 + 3], see: (d) Pagar, V. V.; Jadhav, A. M.; Liu, R.-S. J. Am. Chem. Soc. 2011, 133, 20728. For [3 + 4], see: (e) Pagar, V. V.; Liu, R.-S. Angew. Chem., Int. Ed. 2015, 54, 4923. (6) (a) Eckhard, I. F.; Lehtonen, K.; Staub, T.; Summers, L. A. Aust. J. Chem. 1973, 26, 2705. (b) Iwama, T.; Nagai, Y.; Tamura, N.; Harada, S.; Nagaoka, A. Eur. J. Pharmacol. 1991, 197, 187. (c) Hall, I. H.; Izydore, R. A.; Zhou, X.; Daniels, D. L.; Woodard, T.; Debnath, M. L.; Tse, E.; Muhammad, R. A. Arch. Pharm. 1997, 330, 67. (d) Rozan, P.; Kuo, Y.-H.; Lambein, F. J. Agric. Food Chem. 2000, 48, 716. (e) Taha, M. O.; Dahabiyeh, L. A.; Bustanji, Y.; Zalloum, H.; Saleh, S. J. Med. Chem. 2008, 51, 6478. (7) (a) Basmajian, J. V.; Shankardass, K.; Russell, D.; Yucel, V. Arch. Phys. Med. Rehabil. 1984, 65, 698. (b) Benedini, F.; Bertolini, G.; Cereda, R.; Donà, G.; Gromo, G.; Levi, S.; Mizrahi, J.; Sala, A. J. Med. Chem. 1995, 38, 130. (c) Minghetti, P.; Casiraghi, A.; Montanari, L.; Monzani, M. V. Eur. J. Pharm. Sci. 1999, 7, 231. (d) Mueller, R.; Rachwal, S.; Tedder, M. E.; Li, Y.-X.; Zhong, S.; Hampson, A.; Ulas, J.; Varney, M.; Nielsson, L.; Rogers, G. Bioorg. Med. Chem. Lett. 2011, 21, 3927. (e) Song, W.-W.; Zeng, G.-Z.; Peng, W.-W.; Chen, K.-X.; Tan, N.-H. Helv. Chim. Acta 2014, 97, 298. (8) Dirhodium(II), copper(I), and gold(I) catalysts were examined in ref 5d, all of which afforded the corresponding [3 + 3]-cycloadducts. (9) For further experimental details and high-resolution NMR spectra, see Section 7 of the Supporting Information. (10) (a) Lee, J.; Chen, L.; West, A. H.; Richter-Addo, G. B. Chem. Rev. 2002, 102, 1019. (b) Vasapollo, G.; Sacco, A.; Nobile, C. F.; Pellinghelli, M. A.; Lanfranchi, M. J. Organomet. Chem. 1988, 353, 119. (11) Freeman, J. P. Chem. Rev. 1983, 83, 241 and references therein.
chiral amines could be synthesized via the present approach. Accordingly, the reaction between γ-ethyl enoldiazoacetamide 7 and nitrosobenzene 2a was carried out with a series of chiral dirhodium(II) catalysts (Table S4, SI), among which Rh2(SPTTL)4 provided the highest level of enantiocontrol, thus producing heterocycle-linked trialkylamine 8 in 96% ee (Scheme 4). Scheme 4. Mechanism-Inspired Enantioselective RhodiumCatalyzed Reaction of Enoldiazoacetamide 7 with Nitrosobenzene 2a
In summary, the first cyclization reactions between enoldiazo compounds and nitrosoarenes have been developed. Rh2(oct)4 and Rh2(cap)4 selectively catalyze formal [3 + 2]- and [5 + 1]cyclizations to produce multifunctionalized 5-isoxazolones and 1,3-oxazin-4-ones, respectively. This methodology allows the efficient construction of two classes of biologically important heterocycles from easily accessible reagents with commercially available catalysts under exceptionally mild conditions. Furthermore, the novel reactivity and catalyst-controlled chemodivergence could provide inspiration for future discoveries in the field.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b05840. Experimental procedures and characterization data; Figures S1−S7, Tables S1−S5, and Schemes S1−S7 (PDF) Crystallographic data for 4aa (CIF)
■
AUTHOR INFORMATION
Corresponding Author
*
[email protected] ORCID
Qing-Qing Cheng: 0000-0001-6554-1789 Michael P. Doyle: 0000-0003-1386-3780 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Support for this research from the National Science Foundation (CHE-1464690) is gratefully acknowledged. This paper is dedicated to Professor Qi-Lin Zhou on the occasion of his 60th birthday.
■
REFERENCES
(1) For reviews, see: (a) Handbook of Cyclization Reactions; Ma, S., Ed.; Wiley-VCH: Weinheim, Germany, 2009. (b) Science of Synthesis: Metal-Catalyzed Cyclization Reactions; Ma, S., Gao, S., Eds.; Thieme: New York, 2017. (2) For [3 + 1], see: (a) Deng, Y.; Massey, L. A.; Zavalij, P. Y.; Doyle, M. P. Angew. Chem., Int. Ed. 2017, 56, 7479. For [3 + 2], see: 9842
DOI: 10.1021/jacs.7b05840 J. Am. Chem. Soc. 2017, 139, 9839−9842
