Visible-Light-Mediated Dehydrogenative Cross-Coupling of Azaarenes and Ethers

Abstract

Heteroaromatic motifs are prevalent in natural products and numerous high-value drug molecules. Consequently, the construction of alkylated heterocyclic frameworks holds significant importance. The Minisci reaction of heteroarenes has evolved into a flexible technique for the synthesis of substituted heterocyclic derivatives. However, the use of strong oxidants and external acid is inevitable during the reaction process. Herein, we present a versatile and accessible method for achieving cross dehydrogenation coupling between quinoline derivatives and inactive ether. This strategy utilizes inexpensive NaI and PPh₃ to support the reaction, obviating the need for metal complexes or sacrificial oxidants, and enables the straightforward synthesis of a diverse library of alkyl-substituted N-heteroarenes. Additionally, radical trapping experiments and fluorescence quenching experiments have been conducted to gain a more comprehensive understanding of the reaction mechanism.

1. Introduction

Heteroarenes, especially N–heteroarenes, are widely distributed in natural products and find extensive applications in chemical synthesis, medicine, agriculture, and other fields [1,2]. In recent years, an increasing number of novel drug molecules featuring alkyl-substituted heteroarene motifs, which play a crucial role in human health, have been developed [3,4,5,6], as illustrated in Scheme 1. Therefore, the development of efficient synthesis methods for introducing alkyl functional groups into N–heteroarenes to prepare corresponding products has consistently been a research hotspot.

Over the past half-century, various free-radical alkylation reaction methods for heteroarenes and their derivatives have been developed based on the Minisci reaction model [7,8,9]. These include different transition-metal-catalyzed strategies [10,11,12,13,14] and metal-free alkyl-radical addition approaches [15,16,17,18,19]. Cleavage of C–C bonds [20,21,22], C–N bonds [23], C–X bonds [24,25,26,27], C–S bonds [28,29], and C–B bonds [30,31] has been reported to generate alkyl radicals. However, the simplest and most convenient route remains the activation of C(sp³)–H bonds to generate alkyl radicals [32,33,34]. This cross dehydrogenative coupling (CDC) method not only exhibits excellent atom economy compared to traditional Minisci reactions but also shortens the reaction steps [35], enhancing the overall reaction efficiency [36].

In recent years, photocatalysis [37,38,39,40,41,42,43], electrocatalysis [44,45,46,47,48], and photoelectrochemical catalysis [49,50,51,52] of Minisci reactions have been reported. For example, MacMillan [53] employed a catalytic system with an Iridium catalyst as a photocatalyst and Na₂S₂O₈ as an oxidant, in combination with trifluoroacetic acid (TFA) to activate heteroarenes (Scheme 2A). Under the irradiation of a compact fluorescent light (CFL), the CDC reaction of N–heteroarenes with ether compounds could be achieved. Nevertheless, traditional strong oxidants are still indispensable. Recently, the Minisci reaction of heteroarenes and alkanes has been accomplished without external chemical oxidants [38,54,55,56,57,58,59,60]. For instance, Fu and co-workers [61] reported that the combined reaction system of sodium iodide (NaI) and triphenylphosphine (PPh₃) under 456 nm blue LED light could accelerate chemical reactions through light-induced intermolecular donor–acceptor charge transfer, rather than directly exciting the photocatalytic cycle of the catalyst and substrate. This approach avoids the use of expensive dye photosensitizers. The Minisci–type alkylation of N–heterocycles was accelerated by the phosphine/iodide-based photoredox system (Scheme 2A), which could be integrated with chiral phosphoric acid to achieve high enantioselectivity in the reaction. Gonzalez Gomez’s group [62] also achieved visible-light-induced CDC of azaarenes with unactivated alkanes using the combination of a 9–arylacridine photocatalyst and pyridine N–oxide as a dual-catalyst system without sacrificial oxidants (Scheme 2A). Yao’s group successfully developed a novel photocatalytic material based on an Au/TiO₂ nanocomposite system [63]. Under 365 nm LED irradiation, this hybrid catalyst enabled the Minisci-type oxidative coupling reaction between heteroaromatic compounds and cyclic ethers through the combined use of oxygen as a green oxidant and a catalytic amount of trifluoroacetic acid (TFA) for acid modulation. In addition, an environmentally benign and highly efficient methodology was established for the photoinitiated CDC between quinolines and alcohols/ethers by Fang, employing HCl as a sustainable catalyst while eliminating the requirement for external oxidants or transition metal-based additives [64]. In 2021, our research group developed a photoinduced Minisci reaction of quinoline and ethers using a reagent system comprising NaI, PPh₃, and tetrafluorothianthrenium salt [65]. However, further investigation into the substrate scope and reaction mechanism was lacking.

Inspired by these studies, we describe a novel blue light (455 nm)-induced protocol for activating C(sp³)–H bonds in non-activated alkanes. This protocol utilizes inexpensive NaI and PPh₃ without the addition of external oxidants, acid, noble metal complexes, or organic dyes, enabling the Minisci-type cross-dehydrogenation coupling of quinoline derivatives. This strategy not only features mild and efficient reaction conditions but also provides a simple and reliable method for constructing alkyl-functionalized heteroarenes (Scheme 2B). It is worth noting that (tetrafluoro)thianthrenium salt [66,67,68,69,70] is essential in this reaction, acting as a hydrogen–atom–transfer (HAT) reagent to facilitate the formation of alkyl radicals.

2. Results and Discussion

To verify our hypothesis, 4–methylquinoline (1a) was selected as the substrate, 1,4–dioxane (2a) was used as both the alkylating reagent and the solvent, tetrafluorothianthrenium salt (3a) was added as an additive, and NaI and PPh₃ were employed as supporting the reaction. The desired product 4a was obtained with excellent site selectivity and a yield of 95% (

Table 1. Optimization of the reaction conditions ^a.

Entry	Deviation from Standard Conditions	Yield (%) b
1	none	95
2	MeCN instead of 1,4–dioxane as solvent	65
3	DCM instead of 1,4–dioxane as solvent	33
4	EA instead of 1,4–dioxane as solvent	58
5	1,4–dioxane (1.0 mL)	62
6	1,4–dioxane (3.0 mL)	84
7	3b/3c/3d/3e/3f instead of 3a as additive	N.R.
8	10 mol% PPh3 instead of 20 mol% PPh3	34
9	20 mol% NaI instead of 10 mol% NaI	76
10	390 nm/550 nm light instead of 455 nm light	N.R.
11	Without NaI or PPh3	N.R.
12	Without light	N.R.
13	Without 3a	N.R.

^a Reaction conditions: 1a (0.1 mmol, 1.0 eq.), 2a (2.0 mL), 3a (0.3 mmol, 3.0 eq.), NaI (10.0 mol%), PPh₃ (20.0 mol%) at room temperature with 50 W Blue LEDs (455 nm) irradiation for 12 h under an argon atmosphere. ^b Isolated yield based on 1a.

, entry 1). Since 1,4-dioxane served as both the substrate and the solvent, 1.0 mL of MeCN, DCM, and EA were introduced into the reaction system. However, the product yield decreased significantly (Table 1, entries 2–4). Moreover, altering the amount of 2a, either increasing or decreasing it, led to a decrease in the yield (Table 1, entries 5 and 6). Substituting 3a with other (tetrafluoro)thianthrenium salts completely halted the reaction, which is speculated to be related to the coordinating ability of the oxygen atom (Table 1, entry 7). Compared with 3f, 3a was more likely to produce free radicals because of the stronger electronegativity of fluorine atoms. Reducing the loading of PPh₃ to 10 mol% resulted in a substantial decrease in the yield, while increasing the loading of NaI to 20 mol% caused a slight decrease (Table 1, entries 8 and 9). Changing the wavelength of the light source inhibited the reaction progress (Table 1, entry 10). Control experiments demonstrated that light irradiation was essential for the reaction, and the desired product could not be obtained without the addition of the catalyst or 3a (Table 1, entries 11–13).

Upon determining the optimal reaction conditions, the substrate scope was investigated. The alkylation reactions consistently occurred at the most electrophilic site on the substituted quinoline (Scheme 3). The corresponding products 4a and 4b were synthesized from electron-neutral substrates, such as 2–methylquinoline and 2–phenylquinoline, with yields of 95% and 59%, respectively. Meanwhile, products were obtained from electron-deficient substrates with yields ranging from 31% to 65% (4c and 4d). Subsequently, we explored the effect of substituents at different positions in quinoline on the reaction. When 2-substituents were present, the target C4 products could be smoothly obtained with yields of 34% and 40% (4e and 4f). When the substituent was located at positions 3, 5, 6, and 8 of the quinoline ring, the yield of the corresponding products ranged from 33% to 63% (4g–4j). Notably, isoquinoline also showed good compatibility with this reaction (4k and 4l). In addition, phenanthridine heterocycles could undergo the expected reaction (4m) with a yield of 82%.

Subsequently, tetrahydrofuran was reacted with quinoline and isoquinoline bearing different substituents to assess the reaction compatibility (Scheme 4). The target products 5a–5k were obtained in moderate to good yields. In most cases, the use of C-4 blocked quinoline afforded the corresponding target products in good yields. However, when C-2 blocked quinoline was employed as the substrate, the yield decreased significantly, whereas the scenario reported in the literature by Li et al. [63] still provided the target products in satisfactory yields. When both the C-2 and C-4 positions of quinoline were exposed, only the C-2 monosubstituted quinoline derivatives were obtained, in contrast to the disubstituted target products reported by Li et al. Interestingly, phenanthridine-type heterocycles were also compatible with this reaction, delivering the target product 5l in 68% yield. Additionally, other commercial ethers, such as tetrahydropyran and 7–oxanorbornane, could also react under the optimized reaction conditions, affording the corresponding products with yields of 41–49% (5m–5o). In addition, ether can also participate in the reaction to obtain the target product 5p with a yield of 50%.

To explore the reaction mechanism, radical-trapping experiments were carried out. When 3 equiv. of TEMPO (2,2,6,6–tetramethylpiperidinyloxy) was added, the reaction was completely inhibited, and the radical intermediate was captured by TEMPO and detected via high-resolution mass spectrum (HRMS) analysis (Scheme 5A). This indicated that the reaction system likely proceeded through a free-radical pathway. When 2a was omitted, the adduct of diphenyl ether and TEMPO could still be detected by HRMS, indicating that 3a was the radical donor. When d8-THF was used as the reaction substrate, the corresponding product and deuterated diphenyl ether was detected by HRMS as we predicted (Scheme 5B). As shown by the Stern–Volmer quenching experiments, 1a and tetrafluorothianthrenium salt 3a were capable of quenching the excited state of NaI and PPh₃, suggesting an oxidative quenching mechanism (Scheme 5C–E). UV-Vis absorption measurements on a solution matching the reaction mixture concentration showed no absorption in the visible region for NaI, PPh₃, or their combination (Scheme 5F). 3a showed absorption features only in the UV (<450 nm). When the NaI/PPh₃ component was mixed with 3a, a redshift was observed which indicating the formation of a charge transfer complex.

Based on the above–mentioned reaction data and Fu’s prior work [61], a plausible mechanism for the CDC reaction was proposed, as depicted in Scheme 6. Initially, NaI interacts with PPh₃ in the solvent through cation–π interaction. Subsequently, NaI and PPh₃ form a charge transfer complex B with 7a under Coulomb interaction. Under blue-light irradiation, complex B generates a diphenyl ether radical D, and simultaneously, complex B is oxidized to the intermediate PPh₃–I radical. The diphenyl ether radical D undergoes a single-electron transfer with 1,4–dioxane to generate an alkyl radical E. The generated alkyl radical E attacks the intermediate F, which was generated by the oxidation of 1a through a single-electron process, forming a carbon–carbon bond and producing intermediate G. Intermediate G yields the final product 4a.

3. Conclusions

In summary, we have developed a robust CDC-type Minisci reaction for N–heteroarenes without the need for external acid and sacrificial oxidants. This protocol features mild and efficient reaction conditions, high reaction selectivity, and enables the synthesis of a series of structurally diverse alkylated quinoline derivatives in moderate to good yields. Mechanistic studies reveal that the reaction proceeds via a photo-mediated Minisci process, in which the tetrafluorothianthrenium salt additive plays a crucial role as a HAT agent in combination with NaI and PPh₃. We firmly believe that this distinctive method can provide new opportunities for transition metal-free C–H functionalization and holds great potential for applications in organic and pharmaceutical chemistry.