Qingjing
Yang
a,
Pui Ying
Choy
a,
Yinuo
Wu
*b,
Baomin
Fan
*c and
Fuk Yee
Kwong
*a
aState Key Laboratory of Chirosciences and Department of Applied Biology and Chemical Technology, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China. E-mail: fuk-yee.kwong@polyu.edu.hk
bSchool of Pharmaceutical Sciences, Sun Yat-sen University, Guangzhou, People's Republic of China. E-mail: wuyinuo3@mail.sysu.edu.cn
cYMU-HKBU Joint Laboratory of Traditional Natural Medicine, Yunnan Minzu University, Kunming 650500, People's Republic of China. E-mail: adams.bmf@hotmail.com
First published on 28th January 2016
Cross-dehydrogenative-coupling (CDC) between C–H/C–H bonds of indoles and cyclic ethers/cycloalkanes is made viable through a simple transition-metal-free pathway. With the aid of only di-tert-butyl peroxide, a number of inactive cyclic ethers and cycloalkanes can be directly coupled with indole derivatives in satisfactory yields.
The indolyl framework is a common sub-unit in various pharmaceutically attractive and naturally occurring products.4 Particularly, the relevant indolyl motifs exhibit unique anti-inflammatory and phosphodiesterase inhibitory activities.5 Therefore, facile convergent approaches for accessing an array of C-3 or C-2 alkylated indole derivatives are in high demand. Nevertheless, the reported procedures are sometimes cumbersome and/or require costly catalysts/substrates (Scheme 1A).6 In fact, the direct C–C bond construction of α-substituted cyclic ethers (e.g. five- or six-membered ring) or cycloalkanes to the C-3 or C-2 position of indoles is still rare by virtue of the low reactivity of these C–H bonds. The transition-metal-catalyzed C–C bond formation of ethers/cycloalkanes with heterocycles via cross-dehydrogenative-coupling (CDC) reactions has received considerable attention and advanced remarkably (Scheme 1B).7 However, C–C coupling of indoles with ethers/cycloalkanes is far less explored.8 In 2015, Cai reported a nickel-catalyzed regioselective CDC of inactive C(sp3)–H bonds with indole derivatives.9 Regiospecific C-3 or C-2 coupling of indoles with 1,4-dioxane was shown with the aid of a Ni(acac)2 or NiF2 complex (Scheme 1B). Apart from the transition metal-catalyzed pathway, the CDC reaction would be further attractive if this reaction can proceed in a metal-free manner.10 Nevertheless, a simple metal-free oxidative coupling between indoles and ethers/cycloalkanes remains sporadically studied. Cai showed that di-tert-butyl peroxide (DTBP) mediated oxidative coupling of isochroman with indole derivatives.11 Tian and Li reported a KOt-Bu-mediated coupling of indoles and [60]fullerene with high regioselectivity at the C-3 position of indoles.12 Continuing our research interest on metal-free coupling reactions,13 C–H functionalization of heteroarene14 and CDC reaction of ortho-acylaniline synthesis,15 herein we demonstrate a metal-free protocol for coupling of cyclic ethers or cycloalkanes with the C-3 or C-2 position of indole derivatives using di-tert-butyl peroxide (DTBP) as the oxidant (Scheme 1C).
We first started our investigation by using indole 1a and 1,4-dioxane (2a) as the model substrates (Table 1). A screening of oxidants showed that no reaction occurred when commonly used oxidants such as tert-butyl hydroperoxide (TBHP), benzoyl peroxide (BPO) and potassium persulfate (K2S2O8) were employed (Table 1, entries 3–5). To our delight, di-tert-butyl peroxide (DTBP) and tert-butylperoxybenzoate (TBPB) gave the desired product 3a in 69% and 45% yields, respectively (entries 1 and 2). There were no improvements when CuI, KI, Bu4NI or Pd(OAc)2 was added as the catalyst (entries 7–10). Control experiments revealed that the reaction did not proceed without DTBP, which suggested the crucial importance of peroxide in this transformation (entry 6). Increasing the amount of DTBP or lowering the reaction temperature led to the decrease of the product yield (entries 11 vs. 12 vs. 14). The optimal stoichiometry of DTBP employed was 1.5 equivalents with respect to 2a (entry 1). The best yield was obtained under a nitrogen atmosphere (entry 13).
Entry | [Cat.] (10 mol%) | Oxidants (equiv.) | Yieldb (%) |
---|---|---|---|
a Reaction conditions: 1a (0.5 mmol), 2a (2 mL), [Cat.] (10% mol) and oxidant (0.75 mmol) were stirred at 140 °C in air for 20 hours. (DTBP = di-tert-butyl peroxide, TBPB = tert-butylperoxybenzoate, TBHP = tert-butyl hydroperoxide (70% in aqueous solution), BPO = benzoyl peroxide.) b Isolated yield. c Under a nitrogen atmosphere. d 120 °C was used. e 1 mL of 2a was used. | |||
1 | — | DTBP (1.5) | 69 |
2 | — | TBPB (1.5) | 45 |
3 | — | TBHP (1.5) | 0 |
4 | — | K2S2O8 (1.5) | 0 |
5 | — | BPO (1.5) | 0 |
6 | — | DTBP (0) | 0 |
7 | Pd(OAc)2 | DTBP (1.5) | 68 |
8 | CuI | DTBP (1.5) | 69 |
9 | KI | DTBP (1.5) | 65 |
10 | Bu4NI | DTBP (1.5) | 35 |
11 | — | DTBP (1.0) | 61 |
12 | — | DTBP (2.0) | 58 |
13 | — | DTBP (1.5) | 78 |
14 | — | DTBP (1.5) | 74c,d |
15 | — | DTBP (1.5) | 61e |
With the optimized reaction conditions in hand, we next tested the scope of this transformation (Table 2). A series of unprotected N–H indole derivatives coupled well with 1,4-dioxane to give the corresponding products in good yields. When the bulkiness of the ester-substituted group was increased, the yield of the desired product decreased (e.g., product 3aavs. product 3da). To the best of our knowledge, there have been no successful examples of indoles containing ketone and cyano groups in the oxidative C–C coupling with cyclic ethers. Gratifyingly, the reaction protocol displayed a good functional group tolerance, and bromo (e.g., product 3ga), keto (e.g., product 3ha, 3ia), and cyano groups (e.g., product 3ka) were found compatible under these conditions. Surprisingly, the aldehyde group remained intact under these reaction conditions with moderate yield (e.g., product 3ja). N-Protected indole derivatives were suitable substrates for this reaction with a lower yield possibly due to the steric effect (e.g., product 3la). Apart from the C–C bond formation at the C-2 position of indoles, C-3 target products could also be afforded under this protocol (e.g., products 3oa and 3pa).
a Reaction conditions: substituted indoles 1a–1p (0.5 mmol), 1,4-dioxane (2a) (2 mL), and DTBP (0.75 mmol, 1.5 equiv.) were stirred at 140 °C under N2 for 20 hours. Isolated yields are reported. Reaction times are not optimized for each substrate. |
---|
We next turned our attention to extend the substrate scope regarding the ether coupling partner and the results are compiled in Table 3. Apart from 1,4-dioxane, other ethereal entities were tested. Cyclic ethers such as tetrahydrofuran (2b), 1,3-dioxane (2c), and 1,3-dioxolane (2d) coupled smoothly with moderate yields (e.g., products 3ab, 3ac, and 3ad). Acyclic ethers including diethoxymethane (2e) and dibutyl ether (2f) also underwent the coupling and gave the corresponding coupled products 3ae and 3af in 50% and 74% isolated yields, respectively. In order to evaluate the scope of this system, we attempted to use cycloalkanes as the coupling partners instead of cyclic ethers. When cyclohexane was used, the desired coupling product was obtained in 60% yield (e.g., product 3ah). Cycloheptane and cyclooctane were also feasible coupling partners to yield the desired products (e.g., products 3ah and 3ai). It is noteworthy that the bulky adamantane also underwent the target reaction (e.g., product 3aj).
a Reaction conditions: substituted indole 1a or 1o (0.5 mmol), ether or cycloalkane 2a–j (2 mL), and DTBP (0.75 mmol, 1.5 equiv.) were stirred at 140 °C under N2 for 20 hours. Isolated yields are reported. Reaction times are not optimized for each substrate. b 2.0 mmol of DTBP was used. c 5 equivalents of 1j were used with 2 mL of xylene as a solvent. |
---|
To verify whether the reaction proceeded through a radical pathway, a radical trapping experiment was conducted. When 2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO) was added, a drastic suppression of the reaction resulted and only a trace amount of the desired product 3aa was detected whereas the 1,4-dioxane-TEMPO coupling product was isolated (Scheme 2a). These results suggested that this reaction likely proceeds via a free-radical intermediate. Meanwhile, a kinetic isotopic effect (KIE) experiment was also performed for probing the dependence of C–H bond cleavage (Scheme 2b). As is depicted, a significant KIE was observed with kH/kD = 3.3. This result indicated that the –C–H bond cleavage of 1,4-dioxane could be the kinetically-dependent rate-limiting step of this reaction. Based on the reported literature,8 it is believed that a tert-butoxyl radical was generated by homolytic cleavage through thermal decomposition of di-tert-butyl peroxide. Then, the tert-butoxyl radical undergoes a hydrogen abstraction of the α-proton adjacent to the oxygen atom of the cyclic ether to afford an ether radical. The radical species further undergoes addition to the indole derivative, followed by SET leading to the oxidative coupling final product.
Footnote |
† Electronic supplementary information (ESI) available: Detailed experimental procedures, characterization data and copies of the NMR spectra. See DOI: 10.1039/c6ob00076b |
This journal is © The Royal Society of Chemistry 2016 |