|Home | About | Journals | Submit | Contact Us | Français|
The transition-metal-catalyzed hydroamination reaction, that is, the addition of an N—H bond across a carbon-carbon multiple bond, has been widely studied.[1,2] We have recently reported that cationic gold(I) complexes,[3,4] supported by cyclic (alkyl)(amino)carbene (CAAC) ligands, readily catalyze the intermolecular hydroamination of alkynes with a variety of amines, including ammonia.[6c] Based on preliminary mechanistic studies, we postulated that the key step of the catalytic cycle was the formation of a tricoordinate gold complex (I), which was followed by inner-sphere C—N bond formation, as first postulated by Tanaka et al.,[7a] and Nishina and Yamamoto[7b,c] (Scheme 1). However, for other gold catalysts, Che et al.[9a] and Li et al.[9b] hypothesized an outer-sphere nucleophilic attack to the alkyne complex II, a mechanism widely accepted for palladium and platinum complexes. Herein, our attempts to isolate a gold(I) complex of type I have led to the structural characterization of two (CAAC)(η1-alkene)AuI complexes, and to the discovery of two catalytic reactions: the intramolecular hydroammoniumation using tertiary ammonium salts, and the aminomethylation of carbon–carbon triple bonds.
For obvious entropic reasons, intramolecular hydroamination reactions occur under much milder conditions than their intermolecular analogues, and therefore are better suited for characterizing reaction intermediates. We first chose N-methyl-2-(2-phenylethynyl)aniline (1), as the rigidity of the phenyl spacer group places both the amino and the alkyne moieties in the ideal positions to coordinate to the metal center. The reaction of 1 with a stoichiometric amount of cationic gold(I) complex B at room temperature was monitored by multinuclear NMR spectroscopy. The instantaneous disappearance of both 1 and B was observed, along with the formation of a new complex, which was isolated in near quantitative yield. However, the spectroscopic data suggested that it was not the desired tricoordinate gold(I) complex of type I, but rather complex 2, presumably formed from the coordination of the hydroamination product 3 to the metal (Scheme 2). Even by monitoring the reaction at −70°C, no trace of starting material 1 could be observed, the conversion into 2 being complete in a few seconds. The high catalytic activity was confirmed using 5 mol% of B; at room temperature, heterocycle 3 formed instantaneously in almost quantitative yield. As expected, addition of one equivalent of B to heterocycle 3 led cleanly to complex 2.
To prevent the hydroamination process, and thereby possibly characterize a putative tricoordinate gold(I) complex, we prepared the corresponding tertiary amine 4a. A stoichiometric amount of cationic gold(I) complex B was added at room temperature to a solution of 4a in CDCl3, and the reaction was monitored by NMR spectroscopy (Scheme 3). Once again, we observed the instantaneous disappearance of both 4a and B, and the clean formation of a new complex 5, which was isolated in 98% yield. Surprisingly, single-crystal X-ray diffraction revealed that 5 was a gold(I) η1-alkene complex resulting from the addition of the tertiary amino group to the coordinated alkyne (Figure 1, left). Complex 5 is structurally reminiscent of complexes recently isolated by Hammond et al, and Gagné et al., from the gold-promoted cyclization of allenoates, and the intramolecular hydroarylation of allenes, respectively. Interestingly, the formation of gold(I) η1-alkene complexes of type 5 is not limited to the rigid 2-alkynyl-benzenamine 4a. A mixture of (CAAC)AuCl(A)/AgOTf (1:1) instantaneously reacted stoichiometrically at room temperature with aliphatic N,N-diethylamino alkyne 6 to afford complex 7, which was isolated in 95% yield and fully characterized (Figure 1, right).
Not surprisingly, treatment of complex 5 with one equivalent of trifluoromethanesulfonic acid induced immediate proto-deauration,[2,16] affording the corresponding gold-free cyclic ammonium salt 8a (Scheme 4). The stoichiometric two-step transformation of 4a into 8a (through 5) led to the question of whether this process could be catalytic for gold, or if the presence of triflic acid would preferentially protonate the basic tertiary amine,[7a,17] preventing the cyclization process. Addition of one equivalent of triflic acid to solution of alkynyl amine 4a in chloroform readily gave rise to the corresponding acyclic ammonium salt 9. In the absence of the gold catalyst, no cyclization occurred, even under heating a CDCl3 solution of 9 in a sealed tube at 120 °C for three days (Table 1, entry 1). In contrast, we were pleased to observe that in the presence of 5 mol% of a 1:1 mixture of (CAAC)AuCl(A)/AgOTf, derivative 9 underwent the desired cyclization to form 8a in 98% yield after only 3 h at 70°C (Table 1, entry 2).
Following a brief investigation into the scope of this hydroammoniumation reaction, we found that aryl or alkyl substituents were tolerated on the alkyne, and that the cyclization process occurred under milder conditions when a more weakly basic amine was used. For example, the hydroammoniumation of 4 f readily occurred in a few minutes at room temperature (Table 1, entry 7).
Examples in the literature of the direct carboamination of alkynes (the addition of a carbon–nitrogen bond to a carbon–carbon triple bond) are very rare. Yamamoto et al. have reported the platinum- and palladium-catalyzed intramolecular C—N bond addition of amides and N,O-acetals, and Cacchi et al. reported the palladium-catalyzed cyclization of 2-alkynyl-N-allyl-N′-trifluoroacetyl benzenamine. Although the cleavage of a relatively weak carbon–nitrogen bond was involved in both cases, these results prompted us to investigate the related methylamination reaction: in the presence of 10 mol% of a 1:1 mixture of (CAAC)AuCl(A)/KB(C6F5)4, 2-alkynyl-N,N′-dimethyl-benzenamines 4a–d was transformed into 2,3-disubstituted indoles 10a–d in good to excellent yields after 20 h at 160°C (Scheme 5). This rearrangement tolerates aryl or alkyl substituents on the alkyne. However, in the case of the N-ethyl-N′-methyl-benzenamine 4e, loss of ethylene was observed, and the 3-unsubstituted indole 3 was cleanly formed.
These results suggest that, at least for the intramolecular variant, the gold-catalyzed hydroamination of alkynes does not involve a tricoordinate gold complex of type I. Importantly, cationic gold(I) complexes supported by CAAC ligands promote the intramolecular hydroammoniumation and methylamination reactions of alkynes. The scope of these catalytic reactions is under investigation.
All manipulations were performed under an atmosphere of dry argon using standard Schlenk techniques.
A dried J-Young tube was loaded with gold complex B (0.10 g, 0.074 mmol), 2-alkynyl-N-methyl-benzenamine 1 (0.016 g, 0.075 mmol), and CD2Cl2 (0.5 mL). The reaction was monitored by NMR spectroscopy, and it had proceeded to completion after 5 min. Removal of the solvent under vacuum and washing with n-hexane gave gold complex 2 as a white solid (96% yield); m.p. 160°C; 13C NMR (126 MHz, CD2Cl2) δ =22.9 (CH3), 23.0 (CH3), 26.87 (CH3), 26.90 (CH3), 27.4 (CH), 28.4 (CH), 29.5 (CH × 2), 29.5 (CCH3), 29.6 (CCH3), 33.5 (NCH3), 34.6 (CH2 × 2), 35.4 (CH2), 35.9 (CH2), 37.39 (CH), 37.42 (CH), 39.1 (CH2), 48.6 (CH2), 64.6 (Cq), 67.1 (C(H)Au), 79.1 (Cq), 112.7 (CH), 122.1 (CH), 124.6 (br, Cq), 125.6 (CH), 125.8 (CH), 126.0 (CH), 126.2 (CH), 128.6 (Cq), 129.9 (CmH × 2 and CH × 2), 131.0 (CpH), 131.8 (CH), 134.0 (Cq), 135.7 (Ci), 136.9 (d, JCF = 242.8 Hz, Cq), 138.9 (d, JCF = 242.6 Hz, Cq), 140.8 (Cq), 145.1 (Co), 145.8 (Co), 148.8 (d, JCF = 238.3 Hz, Cq), 164.4 (Cq), 242.7 ppm (Ccarbene).
A dried J-Young tube was loaded with gold complex B (0.15 g, 0.112 mmol), amino alkyne 4a (0.025 g, 0.112 mmol), and CDCl3 (0.4 mL). After 5 min, the solvent was removed under vacuum, and the solid residue washed with n-hexane to give complex 5 as a white solid (98% yield). Single crystals were obtained by recrystallization from a CH2Cl2/n-hexane/toluene (10:20:1) solution at −20 °C; m.p. 243°C; 13C NMR (75 MHz, CDCl3): δ = 23.0 (CH3), 26.5 (CH3), 27.0 (CH), 27.8 (CH), 29.0 (CH), 29.1 (CH3), 34.2 (CH2), 35.2 (CH2), 37.1 (CH), 38.8 (CH2), 48.4 (CH2), 51.9 (CH3), 65.1 (Cq), 77.9 (Cq), 115.5 (CH), 125.0 (CH), 127.7 (CH), 128.0 (CH), 129.3 (CH), 129.6 (CH), 130.3 (CH), 130.7 (CH), 132.2 (CH), 135.4 (Cq), 136.3 (d, JCF = 246.6 Hz, Cq), 138.2 (d, JC–F = 245.4 Hz, Cq), 142.7 (Cq), 144.9 (Cq), 145.3 (Cq), 147.6 (Cq), 138.3 (d, JC–F = 239.5 Hz, Cq), 154.7 (Cq), 159.5 (Cq), 259.7 ppm (Ccarbene).
Complex A (0.043 g, 0.071 mmol) and AgOTf (0.021 g, 0.071 mmol) were loaded in a dried J-Young tube; CDCl3 (0.4 mL) and 6 (0.015 g, 0.071 mmol) were added, and the tube was sealed. The reaction was monitored by NMR spectroscopy. After removing the solvent under vacuum, the residue was washed with n-hexane to give gold complex 7 as a white solid (95% yield). Single crystals were obtained by recrystallization from a CHCl3/n-hexane (1:1) solution at −20°C; m.p. 199°C; 13C NMR (126 MHz, CDCl3): δ =9.7 (CH3), 22.9 (CH2), 23.4 (CH3), 26.1 (CH3), 27.2 (CH), 28.0 (CH), 29.1 (CH), 29.4 (CH3), 34.5 (CH2), 35.2 (CH2), 36,3 (CH2), 37.2 (CH), 39.0 (CH2), 49.0 (CH2), 60.6 (CH2), 63.1 (CH2), 64.7 (Cq), 77.9 (Cq), 121.0 (Cq, q, JCF = 320.5 Hz), 123.1 (CH), 124.1 (CH), 125.0 (CH), 128.3 (CH), 129.6 (CH), 133.0 (Cq), 136.2 (Cq), 145.2 (Cq), 145.3 (Cq), 161.5 (Cq), 258.5 ppm (Ccarbene).
CAAC(AuCl) complex A (0.008 g, 0.0125 mmol) and AgOTf (0.003 g, 0.0125 mmol) were loaded in a dried J-Young tube; CDCl3 (0.4 mL), the internal standard (benzyl methyl ether), derivatives 4 (0.25 mmol), and HOTf (37 mg, 0.25 mmol) were added, the tube was sealed, placed in an oil bath, and heated at the specified temperature (Table 1). The reaction was monitored by NMR spectroscopy.
CAAC(AuCl) complex A (0.030 g, 0.05 mmol) and KB(C6F5)4 (0.036 g, 0.05 mmol) were loaded in a dried J-Young-Tube. C6D6 (0.5 mL) and 4a–d (0.5 mmol) were then added. The tube was sealed, placed in an oil bath, and heated at 160°C for 20 h. Heterocycle 10a–d was purified by column chromatography on silica gel (ethyl acetate/n-hexane 5:95).
**We are grateful to the NIH (R01 GM 68825) and the DOE (DE-FG02-09ER16069) for the financial support of this work, and to the China Scholarship Council (X.Z.), and the Japan Society for the Promotion of Science (R.K.) for fellowships.