Shu-Jie Chen,
Guo-Ping Lu and
Chun Cai*
School of Chemical Engineering, Nanjing University of Science and Technology, Nanjing 210094, P. R. China. E-mail: c.cai@mail.njust.edu.cn; Fax: +86-25-8431-5030; Tel: +86-25-8431-5514
First published on 12th August 2015
Iridium-catalyzed methylation of indoles and pyrroles using methanol as the methylating agent was achieved. This transformation takes place via a borrowing hydrogen methodology under an air atmosphere, which constitutes a direct route to 3-methyl-indoles and methyl-pyrroles.
Based on the development of catalytic methanol dehydrogenation,10 we envisioned the possibility of a direct coupling of indoles and methanol to achieve the methylation of indoles. Although some pioneering works upon coupling of indoles with benzylic alcohols or aliphatic alcohols have been developed, to the best of our knowledge, the methylation of indoles with methanol has not been achieved via hydrogen autotransfer.11 For examples, the groups of Grigg,12 Shimizu,13 Piersanti,14 and Ohta15 demonstrated the C3-alkylation of indoles with alcohols in the presence of homogeneous and heterogeneous transition-metal catalysts.16 In addition, Beller and co-workers reported a N1-alkylation of indoles with alcohols through combination of Shvo's catalyst and p-toluenesulfonic acid.17 In 2013, Li and co-workers described an iridium catalytic system for the synthesis of 3,3′-bisindolylmethanes, which utilized methanol as a surrogate of formaldehyde and proceeded via an “interrupted-hydrogen-borrowing” process.18
In fact, there are two potential paths by which the intermediate A can be converted into different products. The first way is the Michael addition with another indole to give the 3,3′-bisindolylmethane; the latter route is a reduction by the metal hydride to afford the 3-methylindole (Scheme 1). We speculated that the lack of metal hydride suspended the “hydrogen-returning” process, so the Michael addition became dominated. Our idea for producing 3-methylindole was to interrupt the Michael addition by accelerating the consumption of indole and the reduction of the intermediate A. According to this train of thoughts, increasing the loading of the catalyst, which promoted the formation of metal hydride with the concomitant release of formaldehyde, was a possible way to achieve this goal.
In an initial series of experiments, we surveyed the effect of the catalyst loadings on the selectivity by using [Cp*IrCl2]2 as catalyst (Table 1, entries 1–3). In the presence of [Cp*IrCl2]2 (0.2 mol%) and KOtBu (1 equiv.), a 57% yield of methylated product 2a was obtained, along with bisindolylmethane 3a as a side product in 33% yield (entry 1). Notably, with a higher catalyst loading (0.5 mol%), a better yield of 2a was achieved, and the yield was improved to 84% upon increasing the catalyst loading to 1 mol% (entries 2 and 3). This is in line with our assumption mentioned above. It is worth mentioning that an atmosphere of oxygen was found to be beneficial to the methylation,6 and even under the air, the methylation could be accomplished in 90% yield (entries 4 and 5). A quick screening of common hydrogen transfer catalysts showed that [Cp*IrCl2]2 and it's Rh analogs were the optimal catalysts in both activity and selectivity among the testing catalysts (entries 6–11). With regard to the base, LiOtBu, KOtBu, and Cs2CO3 were found to be equally effective for this transformation (see Table S1, ESI†). A decrease in temperature or the amount of base led to unreacted indole, with inferior yield of the desired product (entries 12–14). Omission of the catalyst or base resulted in no desired product formation (entries 15 and 16).
Entry | Cat. (mol%) | Atmosphere | Conversionb (%) | Yieldb (%) | |
---|---|---|---|---|---|
2a | 3a | ||||
a Indole 1a (0.3 mmol), methanol (1 mL), catalyst, KOtBu (1 equiv.), 140 °C for 17 h.b Determined by GC analysis.c 120 °C.d 100 °C, 24 h.e KOtBu (0.5 equiv.).f Base-free condition. | |||||
1 | [Cp*IrCl2]2 (0.2) | N2 | 100 | 57 | 33 |
2 | [Cp*IrCl2]2 (0.5) | N2 | 100 | 74 | 12 |
3 | [Cp*IrCl2]2 (1) | N2 | 100 | 84 | 2 |
4 | [Cp*IrCl2]2 (1) | O2 | 100 | 92 | 0 |
5 | [Cp*IrCl2]2 (1) | Air | 100 | 90 | 2 |
6 | IrCl3·3H2O (2) | Air | 100 | 76 | 14 |
7 | [Ir(cod)Cl]2 (1) | Air | 95 | 27 | 24 |
8 | [Ir(cod)Cl]2 (1)/PPh3 (2) | Air | 97 | 45 | 18 |
9 | (PPh3)2(CO)IrCl (2) | Air | 90 | 82 | 4 |
10 | [Ru(p-cymene)2Cl2]2 (1) | Air | 45 | 15 | 26 |
11 | [Cp*RhCl2]2 (1) | Air | 100 | 89 | 3 |
12c | [Cp*IrCl2]2 (1) | Air | 89 | 50 | 18 |
13d | [Cp*IrCl2]2 (1) | Air | 87 | 33 | 35 |
14e | [Cp*IrCl2]2 (1) | Air | 80 | 65 | 6 |
15 | — | Air | 0 | 0 | 0 |
16f | [Cp*IrCl2]2 (1) | Air | 0 | 0 | 0 |
To study the general applicability of the present system, the methylation of a variety of indoles with methanol was investigated under the optimized conditions (Table 1, entry 5), and the results are summarized in Table 2. The cross coupling of C-2 substituted indoles 1b-1c and methanol afforded the corresponding C-3 methylated products 2b-2c with 82–91% yields. Reactions of indoles bearing electron-donating or electron-withdrawing groups on the phenyl ring also proceeded smoothly. The 4-,7-methyl-indoles 1d-1e and 5-methoxyl-indole 1f participated in the reaction, and the corresponding products 2d–2f were obtained in 84–85% yields. The halide substituents such as fluoro-(1g-1h), chloro-(1i-1j) and bromo-(1k–1m), were tolerant under the conditions, affording the corresponding products in good yields (2g–2m). In the case of substrate bearing an ester group, a slightly longer reaction time was required to obtain good yield (2n). Notably, nitro group survived under the conditions, although it is easily reduced in the presence of alcohols (hydrogen donor) and transition metal catalysts (2o-2p). Similarly, the methylation of 5-cyanoindole (1n) gave the desired product 2n in 87% yield. The coupling was also applied to indole bearing an amino, except for C-3 methylation, N-methylation of amino was also observed (2r). In addition, 7-azaindole successfully yielded 62% of the desired product 2s. Consistent with prior observations, N-methyl indole 1t failed to undergo methylation, which indicated an indole anion involved mechanism.13
Except for indoles, we also investigated the reaction of pyrroles with methanol (Scheme 2). Unlike the indoles, pyrrole shown inferior selectivity, which afforded methylated products consisting of the mono-, di-, tri-, and tetra-methyl-pyrroles under the present conditions (eqn (1)). Although this, it was observed that the methylation occurred preferentially at the α position (C2 or C5) of pyrroles (eqn (2) and (3)).
An additional mechanistic experiment starting from the possible intermediate (1H-indol-3-yl)methanol gave 94% yield of the desired product, which again verified the reaction process outlined in Scheme 1 (Scheme 3, eqn (1)). Furthermore, this methodology also provides a simple and straightforward approach for the synthesis of d3-skatole, which is widely used in the deuterium isotope techniques, such as the metabolism kinetics (eqn (2)).19
In summary, we have developed a [Cp*IrCl2]2-catalyzed method for the direct methylation of indoles and pyrroles using the abundant and bio-renewable methanol as C1 feedstock. The methylation of indoles was selectively occurred at the C3 position, while the methylation of pyrroles was occurred at both the α position (C2 or C5) and the β position (C3 or C4).
Footnote |
† Electronic supplementary information (ESI) available: Experimental procedures and analytical data for products. See DOI: 10.1039/c5ra15822b |
This journal is © The Royal Society of Chemistry 2015 |