Electrochemical Site-Selective Alkylation of Azobenzenes with (Thio)Xanthenes

Herein, we first report an electrochemical methodology for the site-selective alkylation of azobenzenes with (thio)xanthenes in the absence of any transition metal catalyst or external oxidant. A variety of groups are compatible with this electrochemical alkylation, which furnishes the products in moderate to good yields.


Introduction
Azobenzenes are a class of unique aromatic compounds that have been broadly applied in numerous fields, including biomedicine, solar thermal fuels and organic synthesis [1][2][3]. The azo unit is always considered as a privileged scaffold in the design of polymer and chiral catalysts as it readily undergoes cis/trans isomerization upon irradiation under UV or visible light [4][5][6][7][8]. Particularly, azobenzenes have also received increasing attention because of their powerful ability to manipulate organic molecules in synthetic chemistry. Accordingly, various synthetic methods that provide access to, and direct functionalization of, azobenzenes have become an area of interest within the fields of organic synthesis [9][10][11].
To the best of our knowledge, transition metal-catalyzed inert C-H activation, assisted by a directing group, is the most reliable method for chemical bond formation, and has proven to be indispensable for organic synthesis [12][13][14]. More specifically, azobenzenes, containing an "N=N" unit, can readily coordinate with a suitable transition metal, such as Pd, Ru, Rh, Co, Mn and some others, enabling the activation and late-stage functionalization of an ortho-position C-H bond (Scheme 1A) [15][16][17][18][19][20][21][22][23][24][25]. For example, our group and Ellman's group early in 2013 completed indazole synthesis from the inert ortho C-H activation of azobenzenes enabled by Pd and Rh catalysis, respectively [16,17]. A number of excellent works on the functionalization of azobenzenes have been reported, using the same strategies. Recently, our group found that aryne chemistry could realize the mild transformation of azobenzenes into carbazole derivatives under sunlight irradiation, which bypasses the use of toxic transition metal catalysts and oxidants (Scheme 1B) [26]. Most precedents mainly focus on the ortho-C-H, while those involving meta-or para-position C-H functionalization remain relatively scarce [27,28]. Yang, Li and coworkers in 2017 reported Ru-catalyzed C Ar -H (di)alkylation reactions of azobenzenes with various types of alkyl bromides, in which meta-/ortho-selectivity could be well controlled and achieved (Scheme 1C) [27]. Furthermore, advancement on the para-position C-H activation and functionalization of azobenzenes has just been achieved. Very recently, Su's group first reported a cobalt-catalyzed para-selective amination of azobenzenes with a variety of secondary amine compounds, in which the presence of a ligand is crucial for the transformation be well controlled and achieved (Scheme 1C) [27]. Furthermore, advancement on the paraposition C-H activation and functionalization of azobenzenes has just been achieved. Very recently, Su's group first reported a cobalt-catalyzed para-selective amination of azobenzenes with a variety of secondary amine compounds, in which the presence of a ligand is crucial for the transformation (Scheme 1D) [28]. Remarkably, most of the previously reported works on the functionalization of azobenzenes suffered from the use of transition metal catalysts, toxic oxidants and high reaction temperatures, which have severely restricted their further application in synthetic chemistry. Currently, the development of a simple and mild method for the diverse functionalization of azobenzenes is highly desirable. In recent years, electrochemical synthesis has received increasing attention for its powerful ability to forge chemical bonds, presumably due to the advantages of no external stoichiometric chemical oxidants or reductants and milder conditions over the conventional approaches [29][30][31][32][33][34][35][36][37][38][39][40]. As a result, we speculate that electrochemistry maybe provides a unique opportunity to facilitate the functionalization of azobenzene. In a recent study, we disclosed an electrochemical formal [3 + 2] cycloaddition of azobenzenes with hexahydro-1,3,5-triazines, which afforded 1,2,4-triazolidine derivatives in an efficient Scheme 1. Strategies for the C-H functionalization of azobenzenes.
In recent years, electrochemical synthesis has received increasing attention for its powerful ability to forge chemical bonds, presumably due to the advantages of no external stoichiometric chemical oxidants or reductants and milder conditions over the conventional approaches [29][30][31][32][33][34][35][36][37][38][39][40]. As a result, we speculate that electrochemistry maybe provides a unique opportunity to facilitate the functionalization of azobenzene. In a recent study, we disclosed an electrochemical formal [3 + 2] cycloaddition of azobenzenes with hexahydro-1,3,5-triazines, which afforded 1,2,4-triazolidine derivatives in an efficient fashion [41]. Based on these works, and our recent findings in electrochemical synthesis [41][42][43], we continue our effort to address the problem of para-position C-H functionalization in the azobenzenes with the electrochemical method. Herein, we report a catalyst-free alkylation of azobenzenes with (thio)xanthenes enabled by electrochemistry, which affords a series of azobenzenes derivatives with high regioselectivity (Scheme 1E).

Results and Discussion
Initially, (E)-1,2-Diphenyldiazene (1a) and xanthene (2a) were chosen as the model substrates to optimize the reaction conditions for the electrochemical alkylation reaction (Table 1). The reaction system was conducted with two carbon rods as the anode and cathode, n Bu 4 NPF 6 as an electrolyte, MeOH as a solvent, at constant current of 9 mA and room temperature for 4 h, generating the desired product 3a in 76% yield (Table 1, entry 1). Meanwhile, the faradaic efficiency for the electrochemical alkylation of azobenzene was determined as 33.9% (For details, see the electronic Supporting Information). Replacing the electrolyte n Bu 4 NPF 6 with some other commonly used electrolytes, such as n Bu 4 NBF 4 , n Bu 4 NI and LiClO 4 , led to the formation of 3a in decreasing yields (entries 2-4). It was found that choice of electrode materials proved to be crucial for this alkylation reaction. Employment of Pt(+)|Pt(−) as an electrode did not promote the model reaction (entry 5). Lower yields of 3a were obtained when carbon with Pt was used as either the anode or cathode (Table 1, entries 6 and 7). Graphite felt (GF) or Ni electrodes could not improve the yield (entries 8-10). Next, a variety of solvents, including DCE, CH 3 CN, THF, DMF and acetone, were screened, and the result showed that MeOH was the best solvent (entries [11][12][13][14][15]. Decreasing or increasing the reaction time did not improve the yield of 3a (entries [16][17]. Subsequently, changing the intensity of constant current from 8 mA to 10 mA also failed to enhance the yield of 3a (entries [18][19]. The control experiment demonstrated that the reaction could not proceed without electric current (entry 20). Furthermore, the reaction performed under N 2 atmosphere had no obvious effect on the yield of 3a (entry 21).  [41][42][43], we continue our effort to address the problem of para-position C−H functionalization in the azobenzenes with the electrochemical method. Herein, we report a catalyst-free alkylation of azobenzenes with (thio)xanthenes enabled by electrochemistry, which affords a series of azobenzenes derivatives with high regioselectivity (Scheme 1E).

Results and Discussion
Initially, (E)-1,2-Diphenyldiazene (1a) and xanthene (2a) were chosen as the model substrates to optimize the reaction conditions for the electrochemical alkylation reaction ( Table 1). The reaction system was conducted with two carbon rods as the anode and cathode, n Bu4NPF6 as an electrolyte, MeOH as a solvent, at constant current of 9 mA and room temperature for 4 h, generating the desired product 3a in 76% yield (Table 1, entry 1). Meanwhile, the faradaic efficiency for the electrochemical alkylation of azobenzene was determined as 33.9% (For details, see the electronic Supporting Information). Replacing the electrolyte n Bu4NPF6 with some other commonly used electrolytes, such as n Bu4NBF4, n Bu4NI and LiClO4, led to the formation of 3a in decreasing yields (entries 2-4). It was found that choice of electrode materials proved to be crucial for this alkylation reaction. Employment of Pt(+)|Pt(−) as an electrode did not promote the model reaction (entry 5). Lower yields of 3a were obtained when carbon with Pt was used as either the anode or cathode (Table 1, entries 6 and 7). Graphite felt (GF) or Ni electrodes could not improve the yield (entries 8-10). Next, a variety of solvents, including DCE, CH3CN, THF, DMF and acetone, were screened, and the result showed that MeOH was the best solvent (entries [11][12][13][14][15]. Decreasing or increasing the reaction time did not improve the yield of 3a (entries [16][17]. Subsequently, changing the intensity of constant current from 8 mA to 10 mA also failed to enhance the yield of 3a (entries [18][19]. The control experiment demonstrated that the reaction could not proceed without electric current (entry 20). Furthermore, the reaction performed under N2 atmosphere had no obvious effect on the yield of 3a (entry 21). LiClO4 instead of n Bu4NPF6 n.d.  With the established optimal reaction conditions, we set out to investigate the substrate scope of azobenzenes (Scheme 2). In general, azobenzenes bearing electron-donating and electron-withdrawing groups are well compatible with this reaction. First, a variety of mono-substituted azobenzenes were examined under the optimized conditions. For the 4-substituted azobenzenes, the reaction took place specifically on the 4'-position (3b-3i). Alkyl substituents, including Me, Et, i-Pr and t-Bu, were well tolerated in the electrochemical system, and generated the desired product 3b-3e in good yields. Gratifyingly, we further determined the exact structure of 3d by single-crystal analysis [44]. In addition, we found that incorporation of OCF 3 on the 4-position of azobenzene gave the product 3f a 63% yield. Azobenzenes bearing strong electron-withdrawing groups, such as acetyl, cyano and trifluoromethyl group, could interact well with xanthene to form the products 3g-3i in moderate yields by prolonging reaction time, which indicated that the electron-withdrawing group could reduce the reactivity of the substrates. Unfortunately, introduction of an ester group failed to cause a reaction with xanthene (3j). For the 3-substituted azobenzene, the reaction randomly happened on both the 4-positon and the 4'-position of the aromatic ring, affording the mixture of 3k and 3k' in 72% total yield. Next, we examined a variety of disubstituted azobenzenes bearing 2,3-dimethyl,2,4-dimethyl,3,4-dimethyl and 3,5-dimethyl substituents, and all of them worked well under the reaction conditions to form the corresponding products 3l-3o in good yields and regioselectivity. Notably, the alkylation reaction selectively happened on the 4'-position of these disubstituted unsymmetrical azobenzenes. In addition, some symmetrical azobenzenes were also tested. It was found that both (E)-1,2-di-m-tolyldiazene and (E)-1,2-bis(2-isopropylphenyl)diazene proceeded smoothly to generate the products 3p-3q in moderate yields.
We next continued to explore the dialkylation of azobenzenes with xanthene 2a (Scheme 3). By increasing the amount of 2a to 2.2 equivalents and prolonging the reaction time to 6 h, the dialkylation of azobenzenes proceeded well under the modified reaction conditions. For instance, some unsymmetrical azobenzenes, bearing 2-Me and 2-iPr substituents, reacted with xanthene to generate the corresponding products 4a and 4b in 68% and 63% yields, respectively. Additionally, we also found that symmetrical azobenzene (E)-1,2-bis (2-isopropylphenyl)diazene was demonstrated to be a suitable substrate and resulted in the formation of 4c in 57% yield.
We then turned our attention to the tolerance of the reaction towards functional groups on the xanthenes and thioxanthenes, and the results are listed in Scheme 4. More specifically, methyl, methoxy, phenyl on the different position of xanthene were well tolerated (5a-c). Benzoxanthene and derivatives, such as 12H-benzo[a]xanthene, 7H-benzo[c]xanthene, 10-methyl-12H-benzo[a]xanthene, reacted well with azobenzene, generating the products 5d-5f in acceptable yields. Furthermore, some simple thioxanthenes were also examined, and products 5g-5i were achieved in 58-66% yields.
Then, the KIE experiments were carried out to gain insight into the reaction mechanism (Scheme 5). The competing reaction of xanthene 2a and deuterated xanthene 2a-D 2 (1:1) with azobenzene determined the KIE with K H /K D as 1.2, indicating that the cleavage of benzylic C(sp 3 )-H of xanthene was not the rate-determining step. In contrast, an obvious isotope effect (KIE = 2.2) was observed when performing the competing reaction of azobenzene 1a and deuterated azobenzene 1a-D 10 (1:1) with xanthene under standard conditions. These results showed that the cleavage of C Ar -H within azobenzene was presumably involved in the rate-determining step (For details, see Supplementary Materials). In addition, some cyclic voltammetry (CV) experiments were carried out to study the redox potential of the substrates (Figure 1). Remarkably, the oxidation potential of azobenzene 1a (E p = 2.2 V) was far higher than that of xanthene 2a (E p = 1.3, 1.7 V), demonstrating that the xanthene 2a should be preferentially oxidized in the electrochemical system. Based on the above mechanistic experiments and previous reports [41][42][43][45][46][47][48][49], a possible reaction mechanism is proposed in Scheme 6. Firstly, anodic oxidation of xanthene 2a led to the formation of intermediate I, which was further deprotonated to generate radical II, followed by an anode oxidation to form the cationic species III. Secondly, a possible Friedel-Crafts reaction of 1a with the cationic species III occurred to yield the intermediate IV.
Finally, deprotonation of IV gave the product 3a. We next continued to explore the dialkylation of azobenzenes with xanthene 2a (Scheme 3). By increasing the amount of 2a to 2.2 equivalents and prolonging the reaction time to 6 h, the dialkylation of azobenzenes proceeded well under the modified reaction conditions. For instance, some unsymmetrical azobenzenes, bearing 2-Me and 2-iPr substituents, reacted with xanthene to generate the corresponding products 4a and 4b in 68% and 63% yields, respectively. Additionally, we also found that symmetrical azobenzene (E)-1,2-bis (2-isopropylphenyl)diazene was demonstrated to be a suitable substrate and resulted in the formation of 4c in 57% yield. We then turned our attention to the tolerance of the reaction towards functional groups on the xanthenes and thioxanthenes, and the results are listed in Scheme 4. More specifically, methyl, methoxy, phenyl on the different position of xanthene were well tolerated (5a-c). Benzoxanthene and derivatives, such as 12H-benzo[a]xanthene, 7Hbenzo[c]xanthene, 10-methyl-12H-benzo[a]xanthene, reacted well with azobenzene, generating the products 5d-5f in acceptable yields. Furthermore, some simple thioxanthenes were also examined, and products 5g-5i were achieved in 58-66% yields.  We then turned our attention to the tolerance of the reaction towards functional groups on the xanthenes and thioxanthenes, and the results are listed in Scheme 4. More specifically, methyl, methoxy, phenyl on the different position of xanthene were well tolerated (5a-c). Benzoxanthene and derivatives, such as 12H-benzo[a]xanthene, 7Hbenzo[c]xanthene, 10-methyl-12H-benzo[a]xanthene, reacted well with azobenzene, generating the products 5d-5f in acceptable yields. Furthermore, some simple thioxanthenes were also examined, and products 5g-5i were achieved in 58-66% yields. Scheme 4. Scope of (thio)xanthenes a,b . a Reaction conditions: 1a (0.30 mmol), 2 (0.36 mmol), n Bu 4 NPF 6 (2.0 equiv), MeOH (5.0 mL), carbon rod anode (Φ 6 mm), carbon rod cathode (Φ 6 mm), rt, standard conditions. These results showed that the cleavage of CAr-H within azobenzene was presumably involved in the rate-determining step (For details, see Supplementary Materials). In addition, some cyclic voltammetry (CV) experiments were carried out to study the redox potential of the substrates (Figure 1). Remarkably, the oxidation potential of azobenzene 1a (Ep = 2.2 V) was far higher than that of xanthene 2a (Ep =1.3, 1.7 V), demonstrating that the xanthene 2a should be preferentially oxidized in the electrochemical system.  was presumably involved in the rate-determining step (For details, see Supplementary Materials). In addition, some cyclic voltammetry (CV) experiments were carried out to study the redox potential of the substrates (Figure 1). Remarkably, the oxidation potential of azobenzene 1a (Ep = 2.2 V) was far higher than that of xanthene 2a (Ep =1.3, 1.7 V), demonstrating that the xanthene 2a should be preferentially oxidized in the electrochemical system.  Based on the above mechanistic experiments and previous reports [41][42][43][45][46][47][48][49], a possible reaction mechanism is proposed in Scheme 6. Firstly, anodic oxidation of

General Considerations
All 1 H NMR and 13 C NMR spectra were recorded on a 600 MHz Bruker FT-NMR spectrometer (600 MHz and 151 MHz, respectively). All chemical shifts are given as δ

General Considerations
All 1 H NMR and 13 C NMR spectra were recorded on a 600 MHz Bruker FT-NMR spectrometer (600 MHz and 151 MHz, respectively). All chemical shifts are given as δ value (ppm) with reference to tetramethylsilane (TMS) as an internal standard. The peak patterns are indicated as follows: s, singlet; d, doublet; t, triplet; m, multiplet; q, quartet. The coupling constants, J, are reported in Hertz (Hz). High resolution mass spectroscopy data of the products were collected on an Agilent Technologies 6540 UHD Accurate-Mass Q-TOF LC/MS (ESI) and a Thermo Fisher Scientific LTQ FTICR-MS instrument. Melting points were determined in open capillary tube using WRS-1B digital melting point apparatus.
The starting materials, such as azobenzenes and xanthenes, were prepared according to the reported methods [42,43,50,51]. All the solvents are commercially available and directly used in this electrochemical system. Products were purified by flash chromatography on silica gels, eluting with petroleum ether/ethyl acetate (100:1 to 20:1).

Typical Procedure for the Synthesis of 3a
Azobenzene (1a, 0.30 mmol, 1.0 equiv), xanthene (2a, 0.36 mmol, 1.2 equiv), n Bu 4 NPF 6 (0.60 mmol, 2.0 equiv) and CH 3 OH (5.0 mL) were sequentially added into a 15.0 mL ovendried undivided single necked bottle that equipped with a magnetic stirrer bar and sealed with rubber plugs under air atmosphere. A carbon rod (Φ 6 mm) anode and a carbon rod (Φ 6 mm) were used as the cathode in the bottle. About 1.0 cm of the carbon rod was under the solution. The reaction mixture was stirred and electrolyzed at a constant current of 9 mA under air at room temperature for 4 h. After completion of the reaction, the solution was concentrated in vacuum. The resulting crude mixture was purified by flash column chromatography (petroleum ether/ethyl acetate = 100:1) to give the desired product 3a as an orange solid (82.6 mg, 76% yield).  152.3, 151.5, 151.1, 151.0, 149.0, 129.7, 129.1, 128.1, 127.1, 123.8, 123.3 13 13

Conclusions
In summary, we have established a mild protocol to access azobenzene derivatives through the electrochemical alkylation of simple azobenzenes with (thio)xanthenes. This electrochemical transformation proceeds well in the absence of any catalyst or external oxidant, and provides an atom-economic approach for the site-selective functionalization of azobenzenes. We postulate that this strategy can be extended to more challenging organic molecules akin to azobenzene for the development of sustainable electrochemical transformations.