An Intramolecular Reaction between Pyrroles and Alkynes Leads to Pyrrole Dearomatization under Cooperative Actions of a Gold Catalyst and Isoxazole Cocatalysts

The gold-catalyzed one-pot synthesis of 3H-benzo[e]isoindoles (3) from a mixture of isoxazole (2) and diynamides (1) is described. This tandem catalysis involves two separate steps: (i) initial synthesis of 2-(3-pyrrolyl)-1-alkynylbenzenes and (ii) a novel alkyne/pyrrole coupling reaction through pyrrole dearomatization. Our control experiments reveal the cooperative action of the gold catalyst and isoxazole cocatalyst to enable the novel alkyne/pyrrole coupling leading to a 1,2-acyl shift.

I n organic synthesis, gold-catalyzed reactions have become a powerful tool for constructing many carbo-or heterocyclic frameworks. 1 Gold-catalyzed cycloisomerizations of enynes represent some of the most valuable reactions. 2Scheme 1, eq 1 shows one notable enyne cycloisomerization reported by Hashmi,3 who used an alkyne and substituted furans to afford substituted phenol derivatives in either inter-or intramolecular systems.This catalytic phenol synthesis is postulated to involve cyclopropylgold carbene Int-1 as the reaction intermediate (Scheme 1, eq 1).Unfortunately, such an aromatization/ rearrangement sequence fails to work with other heteroaromatic compounds, including pyrroles or indoles, which afforded single addition products instead (eq 2). 4 We sought new chemoselectivity for catalytic coupling reactions between pyrroles and alkynes.This work reports catalytic cascade reactions involving an intramolecular reaction between substituted pyrroles and alkynes as in intermediate Int-2 (eq 3), further affording 2,2-disubstituted-2H-pyrroles (3).This tandem catalysis involves two separate steps using one gold catalyst in a one-pot operation.The first step involves the catalytic formation of 2-alkynylphenyl-3-pyrroles (Int-2) from a mixture of diynamides (1) and isoxazoles (2), and the final step ends with catalytic cycloisomerizations between alkynes and their tethered substituted pyrroles, affording the observed 3H-benzo[e]isoindoles (3).In the second step, an intriguing 1,2-acyl shift occurs to induce pyrrole dearomatization.The key intermediates (Int-2) can be isolated efficiently, and according to our mechanistic analysis their pyrrole/alkyne reactions require the cooperative action of a gold catalyst and an isoxazole cocatalyst.
The diynamide scope was assessed using substrates 1 and isoxazole 2a under the optimized conditions in Table 1 (entry 3), and the results are summarized in Scheme 2. Notably, two atropisomers can be detected for large sulfonamides (R 1 = nbutyl, i-propyl, benzyl), rendering the conformation inflexible around the nitrogen center.We synthesized various diynamides 1b−1d bearing various tosylamides (R 1 = n-Bu, i-Pr, Bn; R 2 = tosyl).Their catalytic cyclizations with 3,5-dimethylisoxazole 2a afforded the corresponding products 3b−3d as two aptroisomers that were not separable by a silica column; their respective ratios were 1.2:1, 2.0:1, and 2.5:1 with combined yields in the 51−53% range.Further, we prepared diynamide 1e having R 1 = Me and R 2 = Ms, producing the desired product 3e as a single isomer, albeit in 33% yield.We examined the reactions on additional diynamides 1f and 1g with R 1 = n-Bu and Bn, respectively, bearing different mesylamides (R 2 = Ms), and the resulting products 3f and 3g were produced as two inseparable atropisomers in ratios of 2.2:1 and 2.3:1.We next tested the reactions on alkylcontaining internal alkynes (R 3 = i-Pr, n-Pr, n-hexyl, Me, Cy) as in diynamides 1h−1l, and their corresponding products 3h−3l were obtained in 45−60% yields.Small MeNT groups avoid forming atropisomers.We also tested the substituents of the central benzene cores.We prepared diynamides 1m and 1n bearing C( 4) substituents (R 4 = Me, Cl), further affording compounds 3m and 3n in 56% and 47% yields, respectively.For the analogous C(5) substituents (R 5 = Cl, Br) as in substrates 1o and 1p, compounds of the same type (3o and 3p) were obtained 52% and 30% yields, respectively.For diynamide 1q bearing a phenylethyne group (R 3 = Ph), the resulting product is contaminated with an impurity that cannot be removed by a silica column or crystallization.
The scope of 3,5-disubstituted isoxazoles has been studied, and the results are provided in Scheme 3. Notably, isoxazoles monosubstituted at either the C(3) or C(5) position (X or Y = H) are inapplicable substrates.Dialkyl-substituted isoxazoles performed better, possibly due to their better nucleophilicity.We prepared various 3-methylisoxazoles (Y = Me) containing different C(5)-alkyl substituents 2b−2e (X = Et, i-Pr, n-Bu, i-Bu), and their corresponding products were obtained in low yields (ca.27−31% yields).This trend is reasonable because the resulting products 3 are produced from the N-attack of isoxazoles at the C(1)-ynamide carbon, whereas these long alkyl groups will increase oxygen's nucleophilicity to facilitate the O-attack of the isoxazole.However, byproducts from the O-attack were absent here because such an attack only occurs with electron-deficient propiolates.We also tested the reactions on isoxazoles 2f and 2g bearing large alkyl groups (Y = Pr and Et) at C( 5), but the resulting products were obtained in 26% and 46% yields, respectively.We believe that a long alkyl at C(5) as in isoxazoles 2f and 2g impedes not only the attack of the isoxazole on the gold π-alkyne complex but also the 1,2-acyl shift in the second step.In the second step, highly hindered tertiary carbon N−CY(COX) is present on the C(2)-pyrrole ring of products 2f and 2g.These unfavorable factors on large X or Y substituents of isoxazoles are expected to give a limited scope of applicable isoxazoles.
To perform the reaction on a large scale, we used diynamide 1a on a 1.0 g (2.74 mmol) scale.The reaction produced 0.78 g (1.68 mmol) of compound 3a, corresponding to a 61% yield.For species 3a, removal of the tosyl group was achieved with HOTf (1.0 equiv) in DCM, affording compound 5a in 32% yield (Scheme 4, eq 4).The molecular structure of this new derivative was confirmed by an X-ray diffraction study. 6Shown in eq 5 are two control experiments to identify the reaction intermediates.We performed these gold-catalyzed reactions with two diynamides (1a and 1j) at room temperature; we hoped that these mild conditions could isolate tractable intermediates.After workup, two new products (6a and 6b) were isolated in 72% and 66% yields, respectively.The molecular structure of species 6b was verified by X-ray diffraction. 6In subsequent pyrrole/alkyne reactions using PPh 3 AuCl/AgNTf 2 (10 mol %) in hot DCE (20 h), both species 6a (R = n-butyl) and 6b (R = i-Pr), to our astonishment, only led to 65−76% recovery (eq 5).In a separate experiment (eq 6), we are sure that species 6a is the intermediate in this two-step catalysis; herein, we note that isoxazole 2b still remains under the initial condition 2a/1a = 1.5.Therefore, we tested the reaction on intermediate 6a with a mixture of the gold catalyst (10 mol %) and isoxazole (20 mol %), smoothly delivering product 3a in 74% yield (eq 7).This outcome confirms the cocatalyst role of isoxazole, which cooperates catalytically with the gold catalyst in this new pyrrole/alkyne coupling reaction.
To gain insight into the mechanism, we performed a crossover experiment involving a 1:1 mixture of intermediates 6a and 6c bearing two different sulfonamides and acyl groups.As shown in Scheme 5, we did not obtain any crossover product apart from expected products 3r and 3a.This information reveals that the 1,2-acyl migration proceeds in an intramolecular fashion.
Scheme 6 shows a plausible mechanism for the two separate steps.For the initial diynamide 1a, its gold π-alkyne complex preferably reacts with isoxazole 2a at the C(1)-ynamide carbon because it is highly electron-rich.This regioselectivity is expected to afford our isolable pyrrole intermediate 6a based on an early report from Ye et al. 7 A subsequent intramolecular cyclization of pyrrole intermediate 6a is only operable at high temperatures in the presence of both the gold catalyst and the isoxazole cocatalyst.A pyrrole addition at the gold π-alkyne in intermediate 6a is expected to form species B bearing a vinylgold substituent and also an iminium center.This electronic feature activates an intramolecular 1,2-acyl migration 8 to form gold carbene C, which is an ideal precursor for the formation of naphthylgold species D after a protonation reaction.A subsequent proto-deauration of this final intermediate D delivers the observed product 3a.In the second step, the loss of energy for pyrrole dearomatization is compensated for by the energy released in the formation of a new benzene.According to our control experiments in Scheme 4, eqs 7 and 8, the presence of isoxazole 2a is required to enable the pyrrole/alkyne coupling reaction.Deprotonation  and proto-deauration for intermediates C and D require a proton shuttle such as isoxazole (2a) to complete this process.
The ultimate goal of this catalysis is to explore the chemoselectivity of a new reaction between pyrroles and alkynes, which are reported to exclusively give alkyne addition products exclusively.This work reports the gold-catalyzed synthesis of 3H-benzo[e]isoindole (3) from a mixture of isoxazoles (2) and diynamides (1). 9This one-pot catalysis involves the initial formation of isolable 2-(3-pyrroly)-1alkynylbenzenes, followed by catalytic coupling of pyrrole/ alkyne functionalities.Herein, new chemoselectivity has been achieved with a pyrrole dearomatization accompanied by a 1,2acyl shift. 7Our control experiments reveal that this atypical alkyne/pyrrole coupling is enabled by the cooperative action of a gold catalyst and an isoxazole cocatalyst.Efforts to realize this reaction in the intermolecular system are under investigation.

■ ASSOCIATED CONTENT Data Availability Statement
The data underlying this study are available in the published article and its online Supporting Information.
Experimental procedures, characterization data, crystallography data, and 1 H NMR and 13 C NMR spectra for representative compounds (PDF) Accession Codes CCDC 2368287−2368288 and 2368293 contain the supplementary crystallographic data for this paper.These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/ cif, or by emailing data_request@ccdc.cam.ac.uk, or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

Scheme
Scheme 4. Control Experiments