Aniruddha
Biswas
,
Satabdi
Bera
,
Puja
Poddar
,
Dibakar
Dhara
and
Rajarshi
Samanta
*
Department of Chemistry, Indian Institute of Technology Kharagpur, Kharagpur 721302, India. E-mail: rsamanta@chem.iitkgp.ac.in
First published on 20th December 2019
An efficient Rh(III)-catalyzed straightforward strategy was developed for the tandem C4 arylamination/annulation of indole derivatives with anthranil to provide indoloquinoline moieties. This method is simple and regioselective with a wide scope and functional group tolerance. Mechanistic studies revealed the important role of the newly installed azacycle in the conversion of O-protected aldoximes to their cyano derivatives. Studies were carried out to explore the promising photophysical properties of the obtained indoloquinoline derivatives.
Retrosynthetically, C4-selective direct tandem arylamination with a proximal carbonyl group and its subsequent annulation might provide the desired indolo[4,5-b]quinolines. Recently, transition metal-catalyzed site-selective functionalization of an indole core especially at its C4 position has become an important challenge for synthetic chemists.5–8 Subsequently, various C–C bond formations at the C4 position were accomplished under transition metal catalyzed conditions.6 However, C–X bond formation at this place has very rarely been reported in the literature.7,8 In their pioneering study, the Prabhu and You groups independently established the C4-amidation of the indole moiety under Ir(III) catalysis using sulfonyl azides as coupling partners (Scheme 1i).7 Recently, we also developed Rh(III)-catalyzed C4-thiolation of indole scaffolds.8
Benzo[c]isoxazole (anthranil) was considered as a bifunctional arylaminating reagent under redox-neutral conditions for the direct arylamination of sp2 and sp3 C–H bonds9,10 and annulation to construct N-PAHs.11–13 Hashmi and co-workers elegantly explored Au(III)-catalyzed construction of various structurally diverse polyazaheterocycles.11 In the relevant advancement, Li's group and Wang's group independently described the tandem Rh(III)-catalyzed C2-arylamination/annulation of indoles using anthranils (Scheme 1ii).12,13 Intrigued by our previous studies on indole C4 functionalization8 and step-economic synthesis of N-PAHs using anthranil,14 we hypothesized that the use of an appropriate guiding group at the indole's C3 position and anthranil as an arylaminating agent might help in providing our desired indoloquinoline structure. Herein, we wish to disclose an efficient Rh(III)-catalyzed tandem C4 arylamination/annulation of C3-oxime tethered indole derivatives with anthranil.
We embarked on our investigation with different directing groups at the C-3 position of indoles. Directing groups like formyl, acetyl, pivaloyl, esters, nitriles, carboxylic acids, and amides did not produce the annulated product under [Cp*RhCl2]2/AgNTf2 catalyzed conditions (ESI,† Table S1, entries 1–6).
Intrigued by our previous studies,6e,8 acetyl-oxime and aldoxime were also screened (ESI,† Table S1, entries 7 and 8). Notably, an aldoxime directing group provided annulated compound 3a′ in 10% isolated yield (ESI,† Table S1, entry 7). However, keto-oxime and other related directing groups did not provide any desired annulated product. Other relevant transition metal catalysts like [Cp*IrCl2]2, [Cp*Co(CO)I2] and [Ru(p-cymene)Cl2]2 did not provide the desired annulated product (see the ESI,† Table S2, entries 1–3). Next, in the presence of NaOAc (30 mol%) as an additive, product 3a was obtained in very poor yield (Table 1, entry 2). Other similar additives like Zn(OAc)2, CsOAc, Cu(OAc)2, and LiOAc·H2O (see the ESI,† Table S2) did not furnish 3a. Further, the yield of 3a was marginally increased to ∼30% via the addition of higher mol% NaOAc (Table 1, entries 3 and 4). Further, bulkier NaOPiv (20 mol%) improved the yield of 3a to 65% (Table 1, entry 6). Further tuning of the NaOPiv concentration to 50 mol% provided the best yield of 3a (75%; Table 1, entry 7). For more improvement, different silver salts were checked but without much success (Table 1, entries 9–11). Other tested common solvents were not much useful (Table 1, entries 13–15).
Entry | Solvent | Ag salt | Additive (mol%) | Yield 3ab (%) |
---|---|---|---|---|
a Reaction conditions: 1a (0.2 mmol), 2aa (0.1 mmol), [Cp*RhCl2]2 (3 mol%), AgNTf2 (12 mol%), carboxylate salt (50 mol%), solvent (0.1 M), 16–24 h, 80 °C. b Isolated yields were calculated with respect to anthranil. c Compound 3a′ isolated in 10% yield. | ||||
1c | DCE | AgNTf2 | — | — |
2 | DCE | AgNTf2 | NaOAc (30) | <10 |
3 | DCE | AgNTf2 | NaOAc (50) | 31 |
4 | DCE | AgNTf2 | NaOAc (100) | 30 |
5 | DCE | AgNTf2 | NaOTFA (50) | — |
6 | DCE | AgNTf2 | NaOPiv·H2O (20) | 65 |
7 | DCE | AgNTf 2 | NaOPiv·H 2 O (50) | 75 |
8 | DCE | AgNTf2 | NaOPiv·H2O (100) | 68 |
9 | DCE | AgSbF6 | NaOPiv·H2O (50) | 69 |
10 | DCE | AgPF6 | NaOPiv·H2O (50) | 15 |
11 | DCE | AgBF4 | NaOPiv·H2O (50) | — |
12 | DCE | — | NaOPiv·H2O (20) | — |
13 | MeOH | AgNTf2 | NaOPiv·H2O (50) | — |
14 | Dioxane | AgNTf2 | NaOPiv·H2O (50) | 37 |
15 | Toluene | AgNTf2 | NaOPiv·H2O (50) | 22 |
With the optimized conditions in hand, we thoroughly explored the substrate scope for C4,C5-annulated indoloquinolines. Different alkyl and aryl substitutions on the indole nitrogen center provided the desired products effortlessly with moderate to excellent yields (Scheme 2, 3a–3e). Strikingly, an N–H free indole 3-aldoxime also provided the desired annulated product (Scheme 2, 3f) in moderate yield.
Electronically different functional groups at the C2 position of the indole moiety produced the respective products with good to excellent yields (Scheme 2, 3g–3i). Further, a C6-halogen-substituted indole also afforded the desired product (Scheme 2, 3j). Next, the products with electron-donating substituents at the C7 position and more π-conjugation were achieved smoothly under the optimized conditions (3k–3o). Gladly, C7 halogenated substrates also survived the reaction (Scheme 2, 3p and 3q). Gratifyingly, the azaindole derivative also underwent site-selective annulation albeit in poor yield (Scheme 2, 3r). Significantly, more complex bis-indole di-aldoxime derivatives provided only monoannulated products under the optimized conditions. Subsequent oxime removal generated the corresponding products in synthetically acceptable yields (Scheme 2, 3s and 3t). To check the scalability, the reaction was carried out using a 3 mmol scale of 2aa and 6 mmol of 1a, which afforded product 3a in 65% isolated yield. Finally, the structure of 3a was unequivocally confirmed by single-crystal X-ray analysis (ESI,† Table S4, CCDC 1944062). Next, anthranils with electron-donating groups and the phenyl group afforded the desired products in good to excellent yields (Scheme 3, 4a–4e). To ensure the halogen compatibility, several halogen-substituted anthranils were tested successfully (Scheme 3, 4f–4h, 4j and 4k). Further, anthranil with an electron-withdrawing ester functionality was also explored (Scheme 3, 4i).
To demonstrate its practical applicability, compound 3a was debenzylated to provide compound 3f in 72% yield (Scheme 4i).8 Moreover, decyanation of 3a was accomplished under Ni(II)-catalyzed conditions to provide compound 5 (Scheme 4ii).15
To realize the plausible mechanism, several control experiments were executed. H/D scrambling at the C4 position of 1a was observed under the standard conditions in the absence of 2aa (Scheme 4iii). This result indicated that indole C4–H bond metalation is reversible in the absence of coupling partner 2aa. However, in the presence of 2aa, no deuteration was identified in the re-isolated 1a, indicating that the subsequent steps of C–H metalation were much faster (Scheme 4iv). Incidentally, there was no H/D scrambling observed at the C2 position under the optimized conditions. In a competition between methoxy and carboxylate attached anthranil, the coupling favored the electron-rich anthranil in a ∼2.75:
1 ratio (Scheme 4v). Further, a 3-cyanoindole derivative was explored under the optimized conditions with coupling partner 2aa. Additionally, 1a also did not transform into its 3-cyanoindole derivative under the optimized conditions. These results indicated that neither the first step was cyano group conversion nor the cyano group was directing the C4–H metalation (Scheme 5i and ii).16 In general, the reaction was not having any significant influence of oxygen (Scheme 5iii). Furthermore, an oxime attached indoloquinoline substrate (3a′) was prepared separately and subjected to various control experiments (Scheme 5 and Table 2) to realize the cyano group formation. The outcome revealed that the newly installed quinoline ring nitrogen center was essential to convert the oxime group into a cyano functionality under a [Cp*RhCl2]2/NaOPiv based catalytic system (Scheme 5 and Table 2, entry c). Based on the above experiments and previous studies,5,12–14 a plausible mechanistic pathway is proposed (Scheme 6). Initially, a cationic Rh(III)-species forms via ligand exchange. Next, Rh(III)-species furnishes a six-membered rhodacycle, A, presumably via oxime guided, pivalate assisted C4–H metalation.6b Next, the nitrogen center of anthranil (2aa) coordinates with A to generate intermediate B. Then migratory insertion of coordinated 2aa into the Rh–C bond affords intermediate C, which further provides intermediate Dvia protodemetalation with regeneration of the active Rh(III) catalyst.
[Cp*RhCl2]2 (3 mol%) | AgNTf2 (12 mol%) | NaOPiv·H2O (50 mol%) | Yield (3a) | |
---|---|---|---|---|
(a) | ⊕ | ⊖ | ⊖ | Traces |
(b) | ⊕ | ⊕ | ⊖ | 45% |
(c) | ⊕ | ⊖ | ⊕ | 93% |
(d) | ⊖ | ⊕ | ⊖ | nd |
(e) | ⊖ | ⊖ | ⊕ | nd |
(f) | ⊖ | ⊕ | ⊕ | nd |
(g) | ⊕ | ⊕ | ⊕ | 91% |
Intramolecular cyclization of D affords compound 3a′. Further, coordination of the oxime and quinoline nitrogen center with the Rh(III)-active catalyst followed by proton abstraction furnishes compound 3a. Importantly, some of these proposed intermediates were also realized by LCMS analysis (see the ESI†).
Next, we investigated the photophysical properties of these indoloquinolines (Fig. 2). An intense absorption band below 325 nm corresponding to the π → π* transition and a comparatively weak absorption band above 325 nm corresponding to the n → π* transition were observed. The values of λmax and molar absorption coefficients, ε (at λmax), for the absorption band of lower energy were calculated (ESI,† Table S5). Although the nature of the UV-vis spectra for all the compounds was similar, the presence of different substituents in the base structure influenced the spectra (Fig. S1–S7, ESI†). Besides their UV active properties, the indoloquinoline derivatives were also found to be fluorescent. The fluorescence excitation and emission spectra of compound 3r, as a representative example, are shown in Fig. 2. The compound is highly fluorescent, with a quantum yield value of 0.75. It is observed that the fluorescence excitation spectrum matches exactly with the corresponding absorption spectrum, confirming the purity of the compound (Fig. 2). The obtained fluorescence spectra and the quantum yield values (ESI,† Fig. S1–S7 and Table S5) implied that several of these compounds might be employed as emitting materials in OLED devices and as biosensors.
![]() | ||
Fig. 2 Absorption spectra of a set of compounds (left); and absorption (abs.), fluorescence excitation (ex.), and emission (emm.) spectra of compound 3r (right). |
In conclusion, we developed a cascade Rh(III)-catalyzed C4-arylamination/annulation of indole derivatives to afford indoloquinoline derivatives using anthranil as a coupling partner. This method is straightforward with a wide scope. Mechanistic studies determined the important role of the newly installed azacycle in the formation of a cyano functionality. Studies were carried out to explore the promising photophysical properties of the obtained indoloquinoline derivatives.
Financial support by the SERB, India (CRG/2018/000630), and DST, India (SR/FST/CSII-026/2013; for 500 MHz NMR), is gratefully acknowledged. AB, SB and PP thank CSIR and IIT Kharagpur for their fellowships.
Footnote |
† Electronic supplementary information (ESI) available. CCDC 1944062. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c9cc08372c |
This journal is © The Royal Society of Chemistry 2020 |