|Home | About | Journals | Submit | Contact Us | Français|
A variety of N-aryl β-nitroenamines were effectively transformed into 3-nitroindoles in good yields and with complete regioselectivity via a rapid microwave (μW) assisted intramolecular arene-alkene coupling reaction. This report further demonstrates the versatility of this method by constructing 3-carboalkoxy- and 3-cyanoindoles. Optimization data, substrate scope, and applications are discussed.
The indole scaffold, a ubiquitous core structure present in numerous natural products and synthetic pharmaceuticals, has been found to present antitumor, anticancer, antimicrobial, antibacterial, anti-inflammatory activities1 and dozens of methods for the construction of indoles have been developed.2 Belonging to this privileged heterocyclic class, substituted nitroindoles are versatile intermediates in the synthesis of important biologically active molecules. For example, 3-nitroindoles are useful precursors in the syntheses of the novel anti-diabetic agent N-(1H-indol-3-yl)-guanidine3 (1) as well as the cruciferous phytoalexins isocyalexin A (2) and rapalexin A (3) (naturally occurring isocyanide and isothiocyanate indoles, respectively; see Figure 1).4 A 3-nitroindole is also the key building block in the synthesis of a 3-acylamino-2-aminopropionic acid derivative (4),5 which has proven to be a potent ligand for the glycine coagonist site on the N-methyl-D-aspartate (NMDA) receptor involved in memory and developmental processes.
The synthetic utility of 3-nitroindoles creates a demand for reliable and versatile routes to these heterocycles from inexpensive precursors. Classic indole nitration methods – utilizing strongly acidic conditions,6a nitrous acid,6b or benzoyl nitrate6c (generated in situ from benzoyl chloride and silver nitrate) – can be effective, but their harsh conditions generally result in low functional group tolerance, low yield, and/or lack of regioselectivity.6
Experimentally, we observed that the benzoyl nitrate method leads to the formation of significant quantities of tarry side-products, reducing reaction yield and complicating purification. This method also requires stoichiometric silver nitrate, which creates heavy metal waste disposal issues. These shortcomings, in conjunction with the inherent limitations of direct nitration, prompted our investigation into more effective routes to 3-nitroindoles.
In the past several years, the emergence of numerous methods for indole synthesis utilizing organometallic catalyzed cross-coupling or metal-free oxidative coupling C-N/C-C bond formation have attracted considerable attention.2j–m,7,8 In that context, Nazaré’s methodology is particularly interesting as it represents an effective route to substituted indoles from o-chloroanilines and ketones.2k Driver et al. reported a route to 3-nitroindoles from β-nitro styryl azides by Rh2(II)-catalyzed nitro-group migration;9 an effective method, but somewhat limited by its use of the expensive Rh2(esp)2 catalyst as well as the multiple synthetic steps required to prepare the precursor nitro styryl azides. 3-Nitroindoles prepared by this method were unfunctionalized at the 2-position. To the best of our knowledge, there has been only one report of a substituted 3-nitroindole synthesized by C-H activation cyclization (a modest yield of 62% was reported).10
Building on our recently reported multicomponent method for indole synthesis utilizing coupled palladium-catalyzed coupling reactions,11 we believed an appropriate methodology could be developed to construct substituted 3-nitroindoles. Herein, we report the versatile palladium-catalyzed cyclization of N-aryl β-nitroenamines as a route to functionalized 3-nitroindoles; a method also applicable (vide infra) to the synthesis of indoles bearing other electron-withdrawing groups (–CO2R/–CN) at C3.
Since its first introduction in the mid 1980s, microwave (μW) irradiation has been widely employed as an effective reaction acceleration protocol, resulting in rapid, clean, and high-yielding transformations.12 Indeed, there is ample literature precedent proving microwave irradiation can mediate, for example, the Heck reaction.13 It is also well established that elevated pressure also assists Pd-mediated coupling reaction by accelerating oxidative addition and improving the catalyst lifetime and turnover by enhancing ligand association-dissociation.14
Our initial studies focused on the intramolecular arene-alkene coupling of enamine 5a [(Z)-2-bromo-N-(1-nitro-prop-1-en-2-yl)aniline; Table 1], which is available from the condensation of o-bromoaniline with 1-nitropropan-2-one.15 This o-bromoaniline-based enamine was chosen as a model substrate since bromoarenes are known to undergo palladium cross-coupling reaction faster than, for example, chloroarenes; in addition, o-bromoaniline derivatives are more readily available than iodoanilines. As illustrated in Table 1, we began by assessing the catalytic activities of various sources of palladium(0) (entries 1–3).
While Pd2(dba)3/DPPF/Cs2CO3/toluene/130 °C/48h (Table 1, entry 1) and Pd(PPh3)4/CuI/Et3N/toluene/130 °C 18h (entry 2) systems did not promote product formation, 5 mol % Pd(PPh3)4/Et3N/pyridine/140 °C/48h (entry 3) did effect the desired transformation. However, the inter-conversion proceeded slowly and in low yield (25%) under conventional oil bath heating conditions.
Under microwave irradiation at the same bulk temperature, the transformation went to completion in 90 minutes and in significantly improved yield (57%; entry 4). As illustrated in Table 1/entry 5, switching to DMF, another highly polarizable solvent that, like pyridine, responds well to microwave irradiation, further improved the 3-nitroindole yield (57→81%). Finally, DIPEA and Et3N were both similarly effective, affording product 6a in 81% (Table 1, entries 5 and 6). A combination of Pd(OAc)2 (4 mol %) and PPh3 (8 mol %) gave the same results. We also screened more active σ-donor monodentate phosphine ligands (entries 7–10); while P(o-tol)3, P(t-Bu)3, tBuXPhos gave slightly higher yields, these ligands were not utilized in this work due to their air sensitivity and higher costs compared to Pd(PPh3)4.
With optimized conditions for 5a → 6a in hand (Table 1), we set out to more generally explore methodology for converting o-bromoanilines and α-nitroketones into the requisite 2-bromo-N-(1-nitroprop-1-en-2-yl)- and 2-bromo-N-(2-nitro-1-arylvinyl)aniline derivatives (5a–k; Scheme 1). These enamine substrates were synthesized in 30–88% yield by an acid-catalyzed condensation reaction between commercially available o-bromoanilines and a-nitroketones — in turn available by the C-acylation of nitromethane with N-acylimidazoles15 [prepared in situ by diimidazole (CDI) or acid chlorides with imidazole (Scheme 1)].16 Substrates 5a–k were obtained only as the Z-isomers, which presumably is a consequence of hydrogen bonding between the NH group and an oxygen on the nitro moiety.17
Under our optimized conditions, enamines 5a–k were successfully transformed into the corresponding 3-nitroindoles (6a–k) in moderate to good yields (Scheme 2). Substrates with electron donating or electron neutral substituents were converted to the desired indoles in higher yields than those with electron-withdrawing groups. Microwave heating provided better indole yields than conventional oil bath heating (6b,e). The presence of an electron-donating 5-OMe group in substrates 5d and 5h lowered the yields of the corresponding indoles. Aryl enamines (5e–g) led to slightly higher indole yields than did aliphatic enamines. Substitution at C4 (see 6k) is also tolerated, but the yield is diminished (42% vs. 81% for 6a).
In the case of heterocyclic (4′-thiazole, 5i–j) enamines, the yields were slightly lower, which could be due to poisoning of the catalyst. A chloro substituent on the aryl part of the enamine (5c,g) is tolerated in this transformation. We did not observe the formation of intermolecular coupling products with chloro-substituted aryl enamines. We also verified that this reaction does not involve the formation of soluble palladium nanoparticles as the catalytically active species by adding Hg(0) – this poisoning test on substrate 5b was negative.18 Finally, microwave irradiation provided such clean reaction mixtures that the isolation/purification protocol consisted of subjecting the crude reaction mixture to silica gel chromatography without the need for an aqueous work-up.
With these encouraging results in hand, we decided to further probe the reaction scope by employing other enamine substrates bearing electron-withdrawing substituents — such as cyano or carboalkoxy groups. N-Aryl-enamine carboxylate and N-arylaminoacrylonitrile substrates were obtained in good yields from the acid catalyzed condensation of o-bromoanilines with β-ketoesters or 3-oxo-3-aryl-propionitrile, respectively (Scheme 3).
Employing these enamines in our optimized reaction delivered the desired 3-carboalkoxy indoles and 3-cyanoindoles (Scheme 4) in shorter irradiation time. Our yields of 3-carboalkoxy indoles were comparable to similar examples synthesized by the C-H activation method reported by Glorius et al.19 With our o-bromoaniline protocol, we did not observed any nonselective formation of regioisomers; an issue with the Glorius method. N-Arylaminoacrylonitriles were rapidly converted to 3-cyanoindoles (8e,f) in excellent yield and modest catalyst loading. We note that 3-cyanation methods20 on preformed indoles utilize stoichiometric amounts of toxic cyanating agents.
Finally, we note that the aryl chloride moieties in 6c/g and 8e/f can be exploited in subsequent synthetic modification. For example, Suzuki-Miyaura cross coupling reactions can enable further diversification.21
In summary, we have developed a rapid and effective microwave mediated route to 3-nitroindoles from N-aryl β–nitroenamines by a palladium-catalyzed intramolecular arene-alkene coupling reaction. This method utilizes catalytic amounts of relatively inexpensive Pd(PPh3)4. The enamines required for this transformation are readily available in one synthetic step from commercially available o-bromoanilines. This method accesses indoles with functionalization at the 2-position — accommodating alkyl, aryl, or heterocyclic substituents. Substrate studies establish that this method affords good functional group tolerance and can also provide access to 3-carboalkoxy- and 3-cyanoindoles.
We thank the National Institutes of Health (GM089153) for generous financial support of this work. H.H.N. thanks Dr. John M. Knapp (UC Davis; Kurth Group) for helpful discussion. We also thank Ms. Kelli M. Farber (UC Davis; Kurth Group) for assistance in the collection of HRMS data.