Rhodium(III)-Catalyzed Redox-Neutral Cascade [3 + 2] Annulation of

7 mins ago - †Shanghai Key Laboratory for Molecular Engineering of Chiral Drugs & School of Chemistry and Chemical Engineering, ‡Key Laboratory fo...
0 downloads 0 Views 1MB Size
Letter Cite This: Org. Lett. 2018, 20, 7131−7136

pubs.acs.org/OrgLett

Rhodium(III)-Catalyzed Redox-Neutral Cascade [3 + 2] Annulation of N‑Phenoxyacetamides with Propiolates via C−H Functionalization/ Isomerization/Lactonization Jin-Long Pan,†,⊥ Peipei Xie,§,⊥ Chao Chen,† Yu Hao,† Chang Liu,† He-Yuan Bai,† Jun Ding,† Li-Ren Wang,¶ Yuanzhi Xia,*,§ and Shu-Yu Zhang*,†,‡

Downloaded via KAOHSIUNG MEDICAL UNIV on November 16, 2018 at 12:55:14 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.



Shanghai Key Laboratory for Molecular Engineering of Chiral Drugs & School of Chemistry and Chemical Engineering, ‡Key Laboratory for Thin Film and Microfabrication of Ministry of Education, and ¶Zhiyuan College, Shanghai Jiao Tong University, Shanghai 200240, P. R. China § College of Chemistry and Materials Engineering, Wenzhou University, Wenzhou 325035, P. R. China S Supporting Information *

ABSTRACT: A Rh(III)-catalyzed cascade [3 + 2] annulation of Nphenoxyacetamides with propiolates under mild conditions using the internal oxidative O−N bond as the directing group has been achieved. This catalytic system provides a regio- and stereoselective access to benzofuran-2(3H)-ones bearing exocyclic enamino motifs with exclusive Z configuration selectivity, acceptable to good yields and good functional group compatibility. Mechanistic investigations by experimental and density functional theory studies suggest that a consecutive process of C−H functionalization/isomerization/lactonization is likely to be involved in the reaction.

T

Scheme 1. Transition-Metal-Catalyzed C−H Activation and Annulation with Alkynes

ransition-metal-catalyzed C−H bond functionalization/ annulation cascade reactions continue to capture the attention of chemists and provide a powerful and convenient alternative to synthesize heteroatom-containing compounds in an ecofriendly and step-economical manner.1 In the past decade, transition-metal-catalyzed C−H functionalization/[n + 2] annulation of aromatic substrates bearing directing groups (DG) with internal alkynes, acting as twocarbon synthons, to synthesize heterocycles has been well documented.2 In addition to these methods, increasing attention has been devoted to transition-metal-catalyzed C−H functionalization reactions using N-phenoxyacetamides as the privileged substrates. They offer opportunities and possibilities to develop new transformations by utilizing the unique reactivity of O− NHAc acting as a redox-active DG, which avoids the use of stoichiometric amounts of external oxidants.3−10 In 2013, Liu and Lu pioneered an efficient synthesis of substituted benzofurans and ortho-hydroxyaryl enamides via a Cp*Rh(III)catalyzed coupling between N-phenoxyacetamides and alkynes with tunable selectivity under mild reaction conditions (Scheme 1, eq 1).3a Followed by this pioneering work, the research groups of Liu,3b−e Wang,4 Zhao,5 Yi,6 You,7 Glorius,8 Zhou,9 and others10 have made great advances in transition-metal-catalyzed C−H functionalization reactions of N-phenoxyacetamides. In contrast, few examples were reported involving internal alkynes, serving as one-carbon coupling partners, in transition-metalcatalyzed C−H activation/[n + 1] annulations.11 In 2015, Chang disclosed the first example of Cp*Rh(III)-catalyzed redoxneutral [4 + 1] annulation of arylnitrones with internal alkynes to afford indolines using the nitrone group as the internal oxidative © 2018 American Chemical Society

directing group accompanied by oxygen atom transfer (Scheme 1, eq 2).11a More recently, Gooßen demonstrated an example of Ru(II)-catalyzed formal [3 + 3] annulation of benzoic acids with propargylic alcohols through oxidative C−H hydroarylation and subsequent intramolecular lactonization to provide γ-alkylideneδ-lactones bearing exocyclic C−C double bonds, in which only one of two alkyne carbon atoms was installed in the lactone ring Received: September 26, 2018 Published: November 8, 2018 7131

DOI: 10.1021/acs.orglett.8b03082 Org. Lett. 2018, 20, 7131−7136

Letter

Organic Letters (Scheme 1, eq 3).12 Inspired by the above elegant work, we envisioned that a combination of C−H activation and intramolecular lactonization using N-phenoxyacteamides and propiolates would be highly interesting and promising. In line with our interest in transition-metal-catalyzed C−H activations,13 herein, we present our new development of rhodium(III)-catalyzed redox-neutral cascade [3 + 2] annulation of readily available N-phenoxyacetamides and propiolates to furnish a straightforward approach to benzofuran-2-(3H)-one skeletons bearing exocyclic enamino motifs, by means of O− NHAc serving as an internal oxidative directing group (Scheme 1, eq 4). Interestingly, benzofuran-2(3H)-ones represent an important class of valuable synthons and privileged structural cores in numerous natural products, drug candidates, and biologically active compounds.14 We evaluated our hypothesis with N-phenoxyacetamide 1a and 3-phenylpropiolate 2 (Table 1). To our delight, upon the

reaction was sluggish when 1,4-dioxane was employed as the solvent (entries 2−4). Interestingly, 3-phenylpropiolates with ethyl and benzyl substituents bonded on oxygen afforded 3aa in similar yields (entries 5 and 6). However, lower yields were obtained with phenyl, iso-propyl, and tert-butyl groups (entries 7−9). This may due to the steric effect, which is not favorable to the lactonization step. Next, a survey of additives indicated that KOAc was more practical and provided better results (entries 10−12). Finally, we were pleased to find that the desired product 3aa could be isolated in 70% yield by slightly increasing catalyst loading (entries 13 and 14). Control experiments showed that Cp*Rh(III) was crucial toward the success of this reaction as its omission led to no formation of 3aa. Other representative Cp*M catalysts (M = Ir, Ru, or Co) did not show any catalytic activity. Increasing the amount of 1a did not give a higher yield. Moreover, N-phenoxypivalamide (PhONHPiv) did not afford the corresponding product. Gratifyingly, the reaction was not sensitive to the air atmosphere and moisture. Thus, we obtained the optimal catalytic system consisting of 1a (0.2 mmol), 2a (1.0 equiv), [Cp*RhCl2]2 (3.5 mol %), and KOAc (0.25 equiv) in MeCN at room temperature for 24 h. After establishing the optimized reaction conditions, we sought to examine the scope and limitations of N-phenoxyacetamides in the present Cp*Rh(III)-catalyzed redox-neutral C−H functionalization/isomerization/lactonization cascade reactions. As shown in Scheme 2, para-substituted Nphenoxyacetamides containing alkyl and phenyl groups led to the construction of the desired benzofuran-2-(3H)-one skeletons bearing exocyclic enamino motifs in reasonable yields (3ba−3fa). Moreover, 4-vinyl and 4-halo groups were compatible, the corresponding products could be generated

Table 1. Optimization of Reaction Conditionsa,b

entry

solvent

additive

2

yield (%)

1 2 3 4 5 6 7 8 9 10 11 12 13c 14e

MeOH DCE 1,4-dioxane MeCN MeCN MeCN MeCN MeCN MeCN MeCN MeCN MeCN MeCN MeCN

CsOAc CsOAc CsOAc CsOAc CsOAc CsOAc CsOAc CsOAc CsOAc CsOPiv KOAc NaOAc KOAc KOAc

2a 2a 2a 2a 2a-2 2a-3 2a-4 2a-5 2a-6 2a 2a 2a 2a 2a

58 61 29 66 66 64 47 49 13 68 69 58 76 (70)d 65

Scheme 2. Substrate Scopea,b

a

Reaction conditions: 1a (0.20 mmol), 2 (0.20 mmol), [Cp*RhCl2]2 (2.5 mol %), and additive (0.25 equiv) in solvent (2.0 mL) at room temperature for 24 h. DCE = 1,2-dichloroethane. bYields were determined by 1H NMR analysis with 1,3,5-trimethoxybenzene as an internal standard. c[Cp*RhCl2]2 (3.5 mol %) was used. dIsolated yield was shown in parentheses. e[Cp*RhCl2]2 (2.0 mol %) was used.

treatment of 1a with 2a in the presence of dimeric [Cp*RhCl2]2 as catalyst precursor and CsOAc as an additive in MeOH at room temperature for 24 h, the unexpected annulation product 3aa was obtained with a NMR yield of 58% (entry 1). The unexpected Zconfiguration of the newly formed exocyclic double bond of 3aa was unambiguously confirmed by X-ray diffraction analysis, and it was speculated that the Z-geometry may be stabilized by the N−H···O type intramolecular interaction (hydrogen bond) (δH(NH) ≈ 11 ppm, and the distance between the hydrogen atom of amide and the oxygen atom of the ester carbonyl group was nearly 2.0 Å).15 Intrigued by the formation of 3aa, we then attempted to further extensively optimize the reaction parameters (see Table S3 in Supporting Information (SI) for detailed optimization studies). First, a screening of solvents revealed that MeCN was the optimal reaction medium, while the

a

Reaction conditions: 1 (0.20 mmol), 2 (0.20 mmol), [Cp*RhCl2]2 (3.5 mol %), and KOAc (0.25 equiv) in MeCN (2.0 mL) at rt for 24 h. bIsolated yield. cPerformed with [Cp*RhCl2]2 (5.0 mol %) at 50 °C. 7132

DOI: 10.1021/acs.orglett.8b03082 Org. Lett. 2018, 20, 7131−7136

Letter

Organic Letters

equivalent 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO), 1,1diphenylethylene, or butylated hydroxytoluene (BHT) as a radical scavenger, the desired product 3aa could still be isolated in maintained yields, implying that the reaction is most likely not based on a radical pathway (see the SI for more details). To probe the active intermediate, a stable five-membered rhodacycle B was prepared.5d The Cp*Rh(III) complex B reacted with equal equivalent 2a to give 3aa in a NMR yield of 43% in the presence of 2 equiv of HOAc, concomitantly generated in the catalytic cycle, which is essential for the formation of product. Moreover, the Cp*Rh(III) complex B was proven to be a robust catalyst for the cascade reaction of 1a and 2a to give 3aa without or with catalytic amount of KOAc in 43% and 62% yields, respectively. The above results demonstrated that the rhodacycle B is most likely involved in the catalytic cycle (Scheme 4a). Under the standard conditions

though higher reaction temperatures and increased catalyst loadings were required (3ga−3ia). For meta-substituted substrates, the reaction underwent smoothly and showed excellent regioselectivity occurring at the less sterically hindered position (3ja−3ma). Substrates bearing an alkyl group or a chlorine atom at the ortho-position or two methyl groups at the para- and meta-positions on the aryl rings gave the corresponding products in acceptable yields (3na−3qa) under the enhanced conditions. Substrates containing electron-withdrawing groups (such as CF3 and CO2Me) showed poor reactivity. To further illustrate the substrate scope, we subsequently explored a wide range of propiolates under standard conditions. To our satisfaction, various electron-donating and -withdrawing functionalities (such as alkyl, phenyl, formyl, acetyl, hydroxymethyl, trifluoromethyl, nitro, dimethylamino, ethers, and halides) on the para- or meta-positions of the aromatic rings were well-tolerated, giving the corresponding benzofuran-2(3H)-one derivatives in yields ranging from 60% to 77% (3ab− 3aq and 3at−3av). Propiolates with a substituent on the 2position of the aryl group showed low reactivity (3ar and 3as). Additionally, the reaction could be successfully extended to 3thiophen-2-yl and 3-thiophen-3-yl propiolates, delivering the desired products in 78% and 69% yields, respectively (3aw and 3ax). Disappointingly, 3-alkyl and 3-alkenyl propiolates gave unsatisfactory results, suggesting that the 3-(hetero)aryl groups were essential under our catalytic system. We then turned our attention to prove both the practicality and effectiveness of this method in organic synthesis. Notably, the cascade [3 + 2] annulation of 1a with 2a was scaled up to 6 mmol under the standard conditions to give access to 3aa in 66% yield. The resultant enaminoate moiety can be used to prepare heterocycles, and readily be converted to chiral β-amino acids by metal-catalyzed asymmetric hydrogenations.16 The obtained 3aa could be conveniently transformed to enamine 4, which could be further converted to the corresponding 2H-azirine 5 via intramolecular cyclization mediated by PhI(OAc)2 (Scheme 3a).17 Importantly, this method can be applied to late-stage

Scheme 4. Experimental Mechanistic Investigations

Scheme 3. Synthetic Applicability

in the presence of D2O, 1a was recovered and 42% deuterium incorporation was observed. In contrast, the reaction of 1a and 2a in the presence of D2O gave rise to the desired product 3aa in a maintained yield, and deuterium was not incorporated in the product (