Boron Radical Promoted Metal-Free Transformation of Nitroazobenzene into Benzotriazole

Abstract

Although the triazole skeleton is significant in biochemistry as a click reaction candidate, as well as in material chemistry due to its excellent absorption of UV light, the preparation of these compounds relies on multinitrogen reagents such as diazo and azido compounds. In this work, o-nitroazobenzenes are first used in a series of neat, fast, green, and efficient reactions for the synthesis of 2-aryl-2H-benzotriazoles under visible light, without RN₃ and metals. It is the visible light-induced boron radical that initiates the reaction by reducing the nitro group into a nitroso group, followed by a barrierless N–N coupling and a facile further deoxygenation by diboron ester to yield benzotriazoles as potential UV absorbers in excellent yields.

Introduction

Nowadays, the useful triazole compounds in organic synthesis, materials science, biochemistry, and other fields − have been widely noticed. Nonetheless, the construction of multinitrogen heterocycles, specifically the formation of triazole rings, still heavily relies on the use of diazo and azido compounds, as well as transition metal catalysts. , With the growing demand for such compounds, it has become increasingly important to develop more convenient methods to generate triazole compounds. Since the 1980s, triazole derivatives have been extensively studied for their antimicrobial and antiprotozoal properties. Additionally, benzotriazole compounds are widely used as UV protection agents in various materials, including plastics, rubber, resins, coatings for automobiles, wood, and textiles, due to their excellent light absorption capabilities. Notably, triazole compounds have been used in click reactions in biochemistry, which were given a Nobel Prize in 2022. However, the involvement of transition metals and azido compounds in triazole preparation prohibits its universal application. Furthermore, the green, convenient, and highly efficient synthetic methodology of N²-substituted benzotriazoles remains even challenging due to the lower electron density at the N² atom than that at the two terminal nitrogen atoms (N¹ and N³) of the triazole heterocycle. −

Currently, the synthesis of N²-substituted benzotriazoles via oxidative or reductive cyclization reactions using substituted azobenzenes as substrates has been reported as a competitive method with high regioselectivity. , However, challenges such as long reaction times, high temperatures, and the use of metal catalysts remain as drawbacks. Easy-to-operate, mildly conditioned, and environmentally friendly photochemical methods are the most promising ways to break through these disadvantages. Based on the photoisomerization of trans- to cis-azobenzene, we have successfully used diboron esters for the reduction of azobenzenes to generate phenylhydrazines. Azobenzenes are also good candidates as dinitrogen reagents, and azobenzenes have been used for nitrogen heterocycle construction. In our previous study, we have successfully used diboron esters for the reduction of nitroarenes, and o-carbonyl cis azobenzenes are found necessary for indazole production through diboron ester-induced deoxygenative coupling (Scheme a,b). However, these strategies are not workable for o-nitro cis-azobenzenes to construct the triazole skeleton. There is no report for diboron ester-assisted benzotriazole generation, even if the application of diboron reagents has been extensively explored by Aggarwal, Jiao, and other research groups. What gives us hope is the results of previous studies that pyridin-boryl radical chemistry has been established by Li and co-workers in 2016 to develop new reactions, and boron radical-induced nitrobenzene reduction has also been reported in 2023 by our research group.

1. Reduction Strategies for Multinitrogen Heterocycle Construction.

Driven by the thought that a N–N coupling between the nitro and azo groups of azobenzene will provide benzotriazole, we use an improved method to generate a boron radical, which, combined with the dioxygen capacity of extra diboron esters, enables a visible-light-induced metal-free intramolecular deoxygenative coupling that occurs between nitro and azo groups in o-nitroazobenzene as a new method for the efficient synthesis of N²-substituted benzotriazole (Scheme c). Importantly, mechanistic studies have shown that the reduction of nitro into the nitroso group is critical for the reaction, and a visible-light-produced boron radical is necessary for this step. Once a nitroso group is generated, the N–N coupling and further deoxygenation are facile to give N²-substituted benzotriazoles as the final product (Scheme c). This study paves the way for a new cyclization mechanism inspired by organic synthesis via radical initiation under visible light.

Results and Discussion

Reductive coupling of nitro by thermochemical approaches has been used to construct 2-aryl-2H-benzotriazoles from azobenzenes, , while metal-free photochemical methods have not been reported at this time. In our previous study, the utilization of visible-light-induced boron radicals for the successful reduction of nitro and carbonyl groups has been achieved. Nitro reduction coupling with the azo bond could be a promising strategy for triazole skeleton construction. To our delight, when we put o-nitroazobenzene 1a with bis (catecholato) diboron (B₂cat₂) and 4-cyanopyridine (4-CNpy) in THF under blue LED, benzotriazole was detected and confirmed as the final product in very high yield.

Encouraged by the initial results, we used compound 1a as a model substrate to optimize the reaction conditions (Table ). First, 95% of the desired product 2a was isolated with 4.0 equiv B₂cat₂ and 0.5 equiv 4-CNpy in THF (2.0 mL) under 450 nm light and a nitrogen atmosphere for 0.5 h (Table , entry 1). Light exposure is crucial for efficient conversion (entries 1–3). Boron radicals may be slowly generated by thermal activation under dark conditions compared to light-mediated conditions (see below), and trace products are generated. Specifically, we observed a 23% yield of 2a after 30 min of dark reaction, which gradually increased to 46% over a 24 h period. This result indicates that while the reaction can proceed without illumination, achieving meaningful yields necessitates an extended reaction duration. However, heating can increase the yield but to a limited extent (entry 4), similar to the results reported in others’ work about boron radical generation and related reactions. − , If B₂cat₂ or 4-CNpy is removed from the reaction (entries 5–6), the reaction would be aborted, indicating the essential role of B₂cat₂ and 4-CNpy in boron radical generation under visible light.

1. Optimization of the Reaction Conditions .

entry	variation from standard condition	yield (%)
1	none	96(95)
2	no light	23,46
3	sunlight instead of 450 nm	53
4	80 °C instead of 450 nm	72
5	no B2cat2	N.D
6	no 4-CNpy	N.D
7	3.0 equiv B2cat2 instead of 4.0 equiv	71
8	0.4 equiv 4-CNpy instead of 0.5 equiv	81
9	B2pin2 instead of B2cat2	trace
10	B2nep2 instead of B2cat2	36
11	DCM instead of THF	73
12	MeCN instead of THF	67
13	DMF instead of THF	43
14	toluene instead of THF	36
15	MeOH instead of THF	25
16	H2O instead of THF	N.D
17	air instead of N2	73

^aStandard condition: 0.2 mmol of 1a, 0.8 mmol of B₂cat₂, 0.1 mmol of 4-CNpy, 2.0 mL of THF, 450 nm LEDs, room temperature, 0.5 h.

^bYields were determined by ¹H NMR analysis using 1,3,5-trimethoxybenzene as an internal standard, and the yield in parentheses indicates the isolated yield.

^cReact for 24 h.

^dReact for 2 h.

Furthermore, we notice a decrease in the yield of the final product when the amounts of B₂cat₂ or 4-CNpy are reduced (entries 7–8). 0.5 equiv 4-CNpy is necessary for boron radical generation in our previous study. Bis (pinacolato) diboron (B₂pin₂) or bis (neopentyl glycolato) diboron (B₂nep₂) instead of B₂cat₂ gives a trace amount or low yield of 2a (entries 9 and 10), confirming the high reactivity of the boron radical formed from B₂cat₂ and 4-CNpy under visible light. We systematically investigate the influence of solvent effects on reactions (entries 11–16). The aqueous phase systems stop the reaction due to the complete insolubility of the substrate. In nonpolar toluene solvent, the yield of the reaction is only 36% due to insufficient initial solubility. The protic solvent MeOH can quench boron radicals (generating HBcat) or lead to the reduction of azobenzene into phenylhydrazine, thus reducing the yield of benzotriazole to 25%. DCM and MeCN show slightly lower yields than THF due to stable radical intermediates. Additionally, the strong complexation between DMF and B₂cat₂ might prohibit the participation of 4-cyanopyridine in the formation of the boron radical, − , thus hindering the reaction. Finally, exposure of the reaction to air leads to a decrease in yield (entry 17), attributed to the quenching of the radical by oxygen. Therefore, the optimized reaction conditions for the preparation of 2-aryl-2H-benzotriazole via o-nitroazobenzene (0.2 mmol) are determined as the combination of B₂cat₂ (4.0 equiv) with 4-CNpy (0.5 equiv) at 450 nm light, protected by N₂ for 30 min in THF.

One can establish a hypothesis that a boron radical is generated from B₂cat₂ and 4-CNpy in THF under blue LED, similar to our previous studies, and this reaction may involve radicals due to the condition optimization experiments. As a result, the TEMPO reagent almost totally quenches the reaction (eq a), but it should be noted that the TEMPO reagent could quench the reaction by consuming B₂cat₂. The presence of 1,1-diphenylethylene significantly reduces the reaction yield (eq b), which is consistent with the results from entry 17 (Table ). Furthermore, electron paramagnetic resonance (EPR) experiments confirm the formation of radicals between B₂cat₂ and 4-CNpy (Figure S1). Combining all of these results, it is obvious that the reaction involves radicals. Therefore, the desired product yield can be maximized only if the reaction is carried out under an inert atmosphere.

Based on the experimental results obtained under different conditions, we propose the reaction mechanism, as shown in Scheme . Initially, B₂cat₂ and 4-CNpy form a colored electron donor–acceptor (EDA) complex. Upon blue light irradiation, this complex undergoes single-electron transfer from the donor to the acceptor, generating a boron radical. Subsequently, the boron radicals attack the oxygen of the nitro group to give 3, from which an energy downhill BcatOBcat molecule is released to give nitroso intermediate 4. This process has been studied through density functional theory (DFT) methods in nitrobenzene reduction by boron radicals in our previous study. Afterward, the lone pair electrons of the azo group attack the electron-deficient nitroso to form a more stable benzotriazole oxide intermediate 5. Obviously, the anionic oxygen in 5 can be easily stripped off by B₂cat₂ to yield the final product 2a.

2. Proposed Reaction Mechanism from Condition Optimization Experiments.

With optimal reaction conditions and the proposed reaction mechanism, we have gained a deep understanding of the reaction and believe this is a promising synthetic method due to its mild and convenient reaction conditions. The driving force of the reaction is assumed to be the initiation of the boron radical, which is formed via light irradiation, as well as the subsequent generation of stable BcatOBcat to make the reaction exothermic. The BcatOBcat molecule can be converted into the useful reagent HOBcat, which renders the reaction recyclable. This meets the requirements of green chemistry, including safe chemical synthesis, safe solvents and auxiliaries, renewable feedstocks, no ecotoxicity, and harmless derivatives. Therefore, we further examine the applicability of the methodology. To demonstrate the potential utility of this methodology in macro production, we conduct the reaction on a 5 mmol scale: the yield reaches 75% after 30 min and 91% after extending the reaction time to 2 h. Although the yields in the scale-up experiments are somewhat reduced, the reactions are still attractive and applicable due to the short reaction time as well as mild and convenient reaction conditions.

Thus, we also examined the scope of the substrates. First, substrates with substituents at one of the benzene rings are investigated (Table ). Unsubstituted 2a generates 2-aryl-2H-benzotriazole in 95% yield. Substituents on the same ring substituted by a nitro group lower down the yields, either for electron-donating groups (2b–2f) or for electron-withdrawing groups (2g–2i), affecting the electronic effect of both the azo and nitro groups. The significant steric hindrance of groups adjacent to the nitro group (1i–1k) is tolerable (2i–2k), but the yield decreases with increasing steric hindrance. It should be noted that the yields of other fluorinated analogues also decreased, possibly due to the stronger electron-withdrawing ability of the fluorine group. Increasing the electron-donating ability of the groups on the benzyl ring without NO₂ can enhance the yield up to 98% (2n, 2t, 2u, 2w). Thus, the improved nucleophilicity of the azo group facilitates the cyclization according to the mechanism in Scheme . Conversely, the presence of electron-withdrawing groups on that ring reduces the yield due to a decrease in the nucleophilicity of the azo group (2p–2s, 2v).

2. Substrate Scope of o-Nitroazobenzene with Substituents at One Benzene Ring .

^aReaction condition: 0.2 mmol of 1, 0.8 mmol of B₂cat₂, 0.1 mmol of 4-CNpy, 2.0 mL of THF, 450 nm LEDs, 30 min. Isolated yields are shown.

^bIsolated yield of 5 mmol scale reaction, reaction condition: 5 mmol of 1, 20 mmol of B₂cat₂, 2.5 mmol of 4-CNpy, 50 mL of THF, 450 nm LEDs, 30 min.

^cThe reaction time was extended to 2 h based on the b reaction conditions.

Notably, the reaction does not undergo dehalogenation, which facilitates further functionalization. To our surprise, the reaction conditions are applicable to compounds containing very sensitive functional groups such as −OOCCH₃, −CN, although the yield was as low as 42% due to the very strong electron-withdrawing ability of −CN (2x). Unfortunately, when the substituent on the benzene ring is a hydroxyl group, the reaction is completely inhibited. We speculate that it is the coordination between the hydroxy and the boron reagent that blocks the reaction. However, by utilizing the hydroxyl protection–deprotection strategy reported in the literature, we have successfully synthesized compounds with benzotriazole containing hydroxyl groups (in 90% yield for 2w to 2OH, SI), which provides a feasible solution to the generation of hydroxyl-substituted benzotriazoles as potential UV absorbers.

We also investigated the suitability of o-nitroazobenzenes with substituents on both benzyl rings as substrates (Table ). The substrate with two electron-donating groups at the para position of azo maintains excellent yields for desired products (98% for 2y and 91% for 2z), although larger groups slightly decrease the yields (2aa, 2ab). o-Nitroazobenzenes containing electron-withdrawing groups at the para position of the azo group had good yields (2ac–2ag). Either electron-donating or withdrawing groups at the meta position of the azo group also obtain a good yield (2ah–2aj). This discovery seems to highlight an intriguing mechanism contradiction for 2ag since electron-withdrawing substituents still give good yield, suggesting the existence of unrecognized reaction pathways. Nonetheless, the overall experimental results support the proposed reaction mechanism, and CO₂Et could be a weak electron-withdrawing group in 2ag. One can still conclude that it is a nucleophilic attack of azo nitrogen toward electron-deficient nitroso nitrogen that facilitates the cyclization.

3. Substrate Scope of Substrates with Substituents on Both Benzene Rings .

^aReaction condition: 0.2 mmol of 1, 0.8 mmol of B₂cat₂, 0.1 mmol of 4-CNpy, 2.0 mL of THF, 450 nm LEDs, 30 min. Isolated yields are shown.

To deeply understand the reaction mechanism, DFT calculations are performed (computational details in the SI). The computational results strongly support our proposed reaction pathway (Scheme ). In the reaction pathway, the boron radical produced by B₂cat₂ and 4-CNpy under blue LED is attracted by nitro oxygen through the interaction of the lone pair electrons in the p orbital of oxygen with the empty p orbital of the boron radical, forming a stable intermediate INT3 with a total reaction barrier of less than 20 kcal/mol. From INT3, one molecule of BcatOBcat is formed to drive the reaction downhill a lot and form the nitroso intermediate INT4 with a reaction barrier of 19.1 kcal/mol. Once this nitroso group is obtained, the electron-rich azo group attacks the nitroso group through facile cyclization to generate the benzotriazole-1-oxide intermediate INT5. Further deoxygenation by the diboron ester gives benzotriazole readily. The overall reaction barrier of the whole reaction pathway is less than 20 kcal/mol (Figure ), consistent with the experimental observation that the reaction finishes in a short time at room temperature.

1.

Computational reaction mechanism. The boron radical initiated formation of 2-aryl-2H-benzotriazole from o-nitroazobenzene.

Radical reduction of the nitroso group has a higher barrier than deoxygenation by B₂cat₂ after N–N coupling from nitroso benzene. It should be noted that it is a thermodynamically unfavorable process to undergo a one-step β-scission to release the OBcat radical and yield INT4 (eq S1, SI). Meanwhile, we also consider that INT4 undergoes a one-step deoxidation and cyclization under the action of B₂cat₂ to obtain the final cyclization product, but it needs to overcome a greater energy barrier (25.0 kcal/mol, Figure S2, SI). At the same time, the radical cyclization pathway exhibits an even higher activation barrier (26.2 kcal/mol; Figure S3, SI). These results demonstrate that the reaction pathway in Figure holds a distinct energetic advantage over alternative routes (Figures S2 and S3).

Although we could not isolate the intermediate of benzotriazole-1-oxide in our experiments due to its rapid reaction with diboron ester, we discovered an alternative method to generate benzotriazole-1-oxide (3a) (synthesis of 3a in SI). Subsequently, in the absence of light, we are able to convert it into 2a with a high yield of 99% by using only B₂cat₂ (eq c). This conversion provides evidence that benzotriazole-1-oxide could serve as a key intermediate in the overall reaction, which is consistent with our computational results (Figure ).

2-(2-Hydroxyphenyl)-benzotriazole derivatives are one sort of efficient UVA-absorbing materials. Here, we synthesize 4-(2H-benzo­[d]­[1,2,3]­triazol-2-yl)­phenol (2OH) (synthesis of 2-OH in SI). The effect of substituent variation on the UVA absorbing capacity of benzotriazole chromophores is investigated by selecting 2a, 2z, 2v, 2h, 2p, 2o, and 2OH to detect the UV–vis absorption in THF (Figure ). The maximum absorption wavelengths of these molecules are between 250 and 350 nm, and the absorption strength is comparable to that of other benzotriazole derivatives. Importantly, the molar extinction coefficients of 2a and 2z are 23406 and 22516 L mol^–1 cm^–1, respectively, which are higher than the common commercial benzene UVA Tinuvin 326, but the hydroxyl-substituted 2OH (ε = 13,620 L mol^–1 cm^–1) is lower than Tinuvin 326. These results show that 2a and 2w could provide effective protection against UVR. Furthermore, this work appears to be a powerful strategy to develop highly effective UV absorbers via in situ visible light organic synthesis.

2.

Molar extinction coefficient of benzotriazole. UV–vis absorption of benzotriazole derivatives obtained in THF solvent (1 × 10^–5 mol/L) via a 10 mm quartz cell.

Conclusions

In conclusion, a convenient, efficient, and environmentally friendly method was developed for the synthesis of 2-aryl-2H-benzotriazoles. This method involves the visible-light-induced reduction of o-nitroazobenzenes, initiated by boron radicals, leading to spontaneous N–N coupling and cyclization. Subsequently, further deoxygenation results in high yields of the final products. Only commercially available reagents, such as B₂cat₂, 4-cyanopyridine, and blue light, are required to achieve excellent yields within 30 min. Notably, this process can be conducted at room temperature using visible light, even on a large scale. This is the first report of a metal-free, efficient artificial photosynthetic process for producing 2-aryl-2H-benzotriazoles, using o-nitroazobenzene as a substrate. It represents a significant breakthrough in the convenient preparation of benzotriazole derivatives as new UV absorber candidates. The mechanism study further inspires us to investigate other N–N coupling reactions between compounds containing nitroso and azo groups. Currently, our laboratory is actively engaged in further studies, aiming to reduce the amount of diboron reagents and expand this methodology to other multinitrogen heterocycle constructions.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.5c01067.

• Experimental details, characterizations, mechanism study, computational details, NMR spectra, and references (PDF)

†.

Y.C. and Y.W. contributed equally. CRediT: Yanqi chen data curation, methodology, writing - original draft.

The authors declare no competing financial interest.