Quinoxalines are key structural components of many natural products and exhibit privileged pharmacological and biological activities.[1] For instance, AG1295 (Figure [1]) is a PDGF receptor tyrosine kinase inhibitor[2] and chloroquinoxaline sulfonamide is a halogenated heterocyclic sulfanilamide, identified as an active agent against various human solid tumors in an in vitro human tumor colony-forming assay.[3] Moreover, riboflavin (vitamin B2) possesses a ribose-derived quinoxaline core fused to a uracil.[4] Several recent studies have demonstrated that incorporating quinoxalines into luminescent materials can provide unique characteristics.[5]
Figure 1 Some important quinoxaline derivatives
Various methods have been developed to exploit the unique biological activities of quinoxalines by employing general, efficient, and sustainable synthetic strategies with readily available building blocks. Despite these advances, challenges persist in synthesizing quinoxaline derivatives, including severe reaction conditions, limited raw-material availability, expensive reagents, multistep synthetic processes, and stoichiometric waste.[6] In 2017, Zhang et al. reported an oxidative synthesis of quinoxalines from primary amines under mild conditions with an ortho-quinone catalyst as the terminal oxidant (Scheme [1a]).[7] In 2022, Guo and co-workers reported a method for the annulation of terminal alkynes and o-phenylenediamines, employing oxygen as the terminal oxidant and a cobalt catalyst (Scheme [1b]).[8] Chaubey et al. developed an iridium-catalyzed [4+2] annulation of β-ketosulfoxonium ylides with o-phenylenediamines to synthesize quinoxaline derivatives (Scheme [1c]).[2a] Nguyen et al. conducted an oxidative condensation of o-phenylenediamines with aryl alkyl ketones in the presence of sulfur to synthesize quinoxaline derivatives (Scheme [1d]).[9] However, there remains a high demand for strategies that permit the high-efficiency synthesis of the quinoxaline skeleton under mild conditions and from readily available starting materials.
Scheme 1 Synthetic strategy and routes to 2-substituted benzoxazoles
Developing novel, metal-free, and versatile synthetic strategies for synthesizing quinoxaline derivatives is crucial in academia and in the pharmaceutical sector to discover bioactive compounds.[10]
[11] Electrochemical synthesis has recently emerged as a green and sustainable strategy for preparing nontraditional and innovative molecules, with the advantage of using electrons as traceless reagents instead of dangerous and toxic redox reagents.[12] The cyclization of aryl alkyl ketones and o-phenylenediamines is a direct and convenient method for preparing quinoxalines, considering the availability of the chemicals and its high atom economy.
Table 1 Optimization of the reaction conditionsa
|
|
Entry
|
Variation from the standard conditions
|
Yieldb (%)
|
|
1
|
none
|
68
|
|
2
|
HCO2H instead of H2C2O4·2H2O
|
64
|
|
3
|
AcOH instead of H2C2O4·2H2O
|
53
|
|
4
|
TFA instead of H2C2O4·2H2O
|
50
|
|
5
|
TsOH instead of H2C2O4·2H2O
|
43
|
|
6
|
H2C2O4·2H2O (0.5 equiv)
|
46
|
|
7
|
H2C2O4·2H2O (1.5 equiv)
|
57
|
|
8
|
No KI
|
32
|
|
9
|
KI (1 equiv)
|
45
|
|
10
|
NaI or NH4I instead of KI
|
35, 54
|
|
11
|
Temp 50 °C or 120 °C
|
–, 60
|
|
12
|
Current 10 mA or 18 mA
|
43, 65
|
|
13
|
DMSO instead of DMA
|
60
|
|
14
|
nBu4NBF4 instead of nBu4NI
|
32
|
|
15
|
nBu4NClO4 instead of nBu4NI
|
26
|
|
16
|
KPF6 instead of nBu4NI
|
trace
|
|
17
|
no nBu4NI
|
46
|
|
18
|
nBu4NI (1.0 equiv)
|
56
|
|
19
|
Ni cathode
|
trace
|
|
20
|
Pt anode
|
trace
|
|
21
|
no electric current
|
–
|
a Standard reaction conditions: 1a (1 equiv), 2a (1 equiv), nBu4NI (0.5 equiv), H2C2O4·2H2O (1 equiv),
KI (0.5 equiv), graphite felt anode, Pt cathode, constant current (15 mA), DMA, 100 °C, 24 h.
b Isolated yield.
In 2016, Zeng’s group reported an electrochemical protocol for the synthesis of α-amino ketones through the oxidative cross-dehydrogenative coupling of ketones with secondary amines.[13] The electrochemistry occurs in a simple undivided cell, utilizing NH4I as a redox catalyst and cheap graphite plates as electrodes under constant-current conditions. The reaction is proposed to proceed through an initial α-iodination of the ketone, followed by a nucleophilic substitution of the amine. Inspired by these works and by our program on electrochemical transformation, we have developed a highly selective intermolecular cyclization and dehydrogenation of o-phenylenediamines with aryl alkyl ketones, facilitating a versatile synthesis of quinoxalines (Scheme [1e]).[14]
Initially, we screened the intermolecular cyclization of acetophenone (1a) with 1,2-phenylenediamine (2a) (Table [1]). The reaction was performed using a graphite felt (GF) anode and a Pt cathode as the supporting electrodes, with 0.5 equivalents each of nBu4NI and KI in an undivided cell (a Schlenk tube). Treatment 1a and 2a in N,N-dimethylacetamide (DMA) (6 mL) in the presence of oxalic acid dihydrate (H2C2O4·2H2O; 1 equiv) at 100 °C for 24 hours yielded 2-phenylquinoxaline (3aa) in a 68% isolated yield (Table [1], entry 1). Replacement of H2C2O4·2H2O with HCOOH, AcOH, TFA, or TsOH decreased the yield to 64, 53, 50, and 43%, respectively (entries 2–5). Changing the concentration of H2C2O4·2H2O did not increase the yield (entries 6 and 7). Similarly, adjusting the KI dosage did not improve the yield (entries 8 and 9). However, use of the iodide salts NaI and NH4I resulted in lower yields (entry 10). The best yield was achieved at 100 °C: performing the reaction at different temperatures led to lower yields of 3aa (entry 11). Additionally, decreasing or increasing the constant current negatively influenced the reaction outcome (entry 12). When DMSO was employed as the solvent, the yield decreased to 60% (entry 13). Changing the electrolyte to nBu4NBF4 or nBu4NClO4 resulted in the formation of the desired products in isolated yields of 32, and 26% respectively, whereas KPF6 gave only a trace of the product (entries 14–16). Notably, the yield decreased when nBu4NI was not added or when the dosage of nBu4NI was decreased (entries 17 and 18). Additionally, Pt as anode and Ni as cathode proved unsuitable for this electrochemical reaction (entries 19 and 20). As expected, the reaction failed to give any product without an electric current (entry 21).
The scope of quinoxaline synthesis was then examined (Scheme [2]). 2-Arylquinoxalines 3aa–be with substituents of varying electronic nature were accessible. Acetophenones 1 bearing either electron-rich groups, such as methyl (3ab–ad, 3ai), methoxy (3ae), ethyl (3af), isopropyl (3ag), or tert-butyl (3ah), or electron-deficient substituents such as halo (3aj and 3am), trifluoromethyl (3an), or cyano (3ao) at various positions were all suitable substrates. Substrates containing an ester group or an amide group also reacted to give the corresponding quinoxalines 3ap and 3aq in moderate yields. Furthermore, substrates containing a 2-naphthyl (3ar), 2-furyl (3as), or 1,3-benzodioxol-5-yl (3at) group were also compatible with the reaction conditions. Subsequently, 1,2-diphenylethanone was found to react with 2a to produce the corresponding 2,3-diphenylquinoxaline (3au). Unfortunately, phenylacetone as a substrate was incompatible with the standard reaction conditions (3av).
Scheme 2 Investigation of the scope of the benzoxazole-2-carboxylate. Reaction conditions: 1 (0.5 mmol), 2 (0.75 mmol), nBu4NI (0.25 mmol), KI (0.25 mmol), H2C2O4·H2O (0.5 mmol), DMA (6 mL), constant current = 15 mA, undivided cell, GF anode (1.5 × 1.0 cm), Pt cathode (1.5 × 1.0 cm), 100 °C, 24 h. The isolated yields are reported.
Next, various reactions of 4,5-dimethylbenzene-1,2-diamine with acetophenone derivatives were investigated under the optimized conditions, and gave moderate to high yields of the corresponding products 3ba and 3bb; remarkably, the anti-PDGF receptor agent AG 1295 (3ba) was obtained in a high yield under the optimized conditions. Notably, significant electronic effects were observed in the case of disubstituted 1,2-diaminoarenes containing electron-donating substituents (3ba), which exhibited a higher compatibility than those with electron-withdrawing substituents, such as a halo group (3bc). Furthermore, asymmetric 1,2-diaminoarenes as substrates gave two isomers of the corresponding product quinoxalines (3bd and 3be). These isomers could not be separated by flash chromatography on a silica gel column because of their similar polarities.
In addition, to further explore the applications of this method in constructing quinoxalines, this electro-redox process was evaluated through a 10 mM scale reaction. A gram-scale synthesis of 3aa was achieved by using 1a (0.6 g, 5.0 mM) and 2a (0.81 g, 7.5 mM), with a yield of 69% (0.71 g) (Scheme [3]A). Moreover, the menthol ester substrate 4 (0.151 g, 0.5 mM) could be dehydrogenated with 2a (0.081 g, 0.75 mM) for intermolecular cyclization, resulting in the formation of the complex molecule 5 in 42% yield (0.081 g) (Scheme [3]B).
Scheme 3 Gram-scale experiment and a derivatization
Cyclic voltammetry (CV) experiments were then conducted to further elucidate the electrochemical oxidation processes for this reaction. As shown in Figure [2], CV measurements for nBu4NI revealed an oxidative wave at 1.63 V (Curve a), whereas KI showed an oxidative wave at 1.41 V, indicating that KI was oxidized first (Curve b). Moreover, when all raw materials were mixed with KI and a blank, the redox response value was significantly increased. The concentration of proton (addition of acid) might play a role here, because it reduces the nucleophilicity of the amine thereby slowing the first substitution; however, as a tradeoff, it probably helps accelerate the cathodic reaction rate thereby lowering the overall resistance of the cell. (Curve c). Additionally, no obvious redox change was observed when the blank was mixed with the two raw materials. The blanks with 1a and 2a exhibited oxidation waves at 1.02 and 1.49 V, respectively (curves d and e). These results indicated that substrate 1a was initially more susceptible to oxidation.
Figure 2 Cyclic voltammograms of substrates 1a and 2a in 0.1 M
n
Bu4NI/DMA, using a Pt wire working electrode and glassy carbon and Ag/AgCl (0.1 M in DMA) as the counter and reference electrodes at 100 mV s–1 scan rate: (a) blank (0.1 M nBu4NI in DMA), (b) blank + KI (20 mM), (c) blank + 1a (20 mM) + 2a (20 mM) + KI (20 mM) + H2C2O4·2H2O (20 mM), (d) blank + 1a (20 mM), (e) blank + 2a (20 mM).
Scheme 4 Proposed mechanism
Based on these experimental results and previous reports,[15] a plausible mechanism for the reaction of 1a and 2a is shown in Scheme [4]. Initially, the iodide is oxidized to an iodine radical that then reacts with 1a to form 2-iodo-1-(4-methylphenyl)ethanone (C). The presence of an acid facilitates the condensation between the diamine 2a and iodinated intermediate C to form imine D. Subsequently, an intramolecular nucleophilic attack of the amino group on the C–I bond generates a cyclic intermediate E, and anodic oxidation potentially yields the desired aromatization product 3aa. Simultaneously, the reduction of protons at the cathode resulted in the release of hydrogen.
In summary, a practical and atom-economic electrocatalytic strategy has been developed for the intramolecular dehydrogenative cyclization of quinoxaline derivatives. This strategy resulted in the production of various 2-substituted quinoxalines in moderate to good yields.[16] The electrocatalytic approach modulated the reactivity of the starting materials to facilitate complex transformations in a single step. Oxidative dehydrogenation cyclization was performed to achieve this transformation without using transition metals or stoichiometric chemical oxidants, under mild conditions. Notably, the reaction, which releases H2 as a theoretical byproduct, has a high atom economy. Furthermore, the broad substrate scope and late-stage functionalization features of this method extend its adaptability. We expect that this novel approach will be used to construct complex quinoxaline derivatives as active ingredients.
