Radical–Radical Cross-Coupling Assisted N–S Bond Formation Using Alternating Current Protocol

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE5 Aug 2022Radical–Radical Cross-Coupling Assisted N–S Bond Formation Using Alternating Current Protocol Yong Yuan†, Jun-Chao Qi†, Dao-Xin Wang, Ziyue Chen, Hao Wan, Jing-Yun Zhu, Hong Yi, Abhishek Dutta Chowdhury and Aiwen Lei Yong Yuan† National Research Center for Carbohydrate Synthesis, Jiangxi Normal University, Nanchang 330022 College of Chemistry and Molecular Sciences, Institute for Advanced Studies (IAS), Engineering Research Center of Organosilicon Compounds & Materials (Ministry of Education), Wuhan University, Wuhan, Hubei 430072 Gansu International Scientific and Technological Cooperation Base of Water-Retention Chemical Functional Materials, College of Chemistry and Chemical Engineering, Northwest Normal University, Lanzhou, Gansu 730070 †Y. Yuan and J.-C. Qi contributed equally to this work.Google Scholar More articles by this author , Jun-Chao Qi† National Research Center for Carbohydrate Synthesis, Jiangxi Normal University, Nanchang 330022 †Y. Yuan and J.-C. Qi contributed equally to this work.Google Scholar More articles by this author , Dao-Xin Wang National Research Center for Carbohydrate Synthesis, Jiangxi Normal University, Nanchang 330022 Google Scholar More articles by this author , Ziyue Chen National Research Center for Carbohydrate Synthesis, Jiangxi Normal University, Nanchang 330022 Google Scholar More articles by this author , Hao Wan National Research Center for Carbohydrate Synthesis, Jiangxi Normal University, Nanchang 330022 Google Scholar More articles by this author , Jing-Yun Zhu National Research Center for Carbohydrate Synthesis, Jiangxi Normal University, Nanchang 330022 Google Scholar More articles by this author , Hong Yi College of Chemistry and Molecular Sciences, Institute for Advanced Studies (IAS), Engineering Research Center of Organosilicon Compounds & Materials (Ministry of Education), Wuhan University, Wuhan, Hubei 430072 Google Scholar More articles by this author , Abhishek Dutta Chowdhury College of Chemistry and Molecular Sciences, Institute for Advanced Studies (IAS), Engineering Research Center of Organosilicon Compounds & Materials (Ministry of Education), Wuhan University, Wuhan, Hubei 430072 Google Scholar More articles by this author and Aiwen Lei *Corresponding author: E-mail Address: [email protected] National Research Center for Carbohydrate Synthesis, Jiangxi Normal University, Nanchang 330022 College of Chemistry and Molecular Sciences, Institute for Advanced Studies (IAS), Engineering Research Center of Organosilicon Compounds & Materials (Ministry of Education), Wuhan University, Wuhan, Hubei 430072 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202101350 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Alternating current (AC) electrolysis is a promising, yet challenging, and under-developed protocol in organic synthesis. To achieve as high an atom-efficiency as possible and avoid the use of external oxidant, electrochemistry has become a standard organic synthesis tool. Herein, an AC-based protocol is a superior option than its counterpart, direct current (DC), especially for those reactions that cannot be accomplished by DC. With an aim to achieve this objective, we have employed AC electrolysis successfully to deliver the unprecedented cross-coupling of sulfonyl- or acyl-substituted aniline derivatives with thiophenols/thiols under exogenous-oxidant- and catalyst-free conditions. Numerous N–S bond-containing compounds (40 in total) have been prepared at up to 93% yield and good selectivity. Interestingly, no reactivity has been noted using DC electrolysis or traditional chemical approaches under our reaction conditions. Mechanistic studies indicated that disulfide is the key intermediate in this AC-mediated radical–radical cross-coupling reaction. Therefore, the current work does not only provide an efficient method for N–S bond formation but also paves the way to new sustainable avenues for innovation in organic synthesis. Download figure Download PowerPoint Introduction Organic electrosynthesis, which can realize redox transformations in the absence of exogenous oxidant or reductant, has proved to be one of the most sustainable approaches for constructing complex molecules.1–6 Over the past two decades, tremendous efforts have been put into this attractive field, and considerable achievements have been made.7–13 Nevertheless, in most developed electrochemical reactions, only either the oxidation or reduction process can result in the expected product, while the concomitant half-reaction that takes place at the counter electrode is essentially a sacrificial reaction (Figures 1a and 1b), which makes it less sustainable. In recent years, chemists have turned their attention to developing convergently-paired electrolysis, in which both oxidation and reduction processes synergistically produce the target product.14 Although this method is theoretically promising and energy-efficient, convergently-paired electrolysis (Figure 1c) still remains largely under-developed,15–17 probably because of the lower yield due to the conflict between slow interelectrode mass transport rate and short survival time of intermediates. Figure 1 | Strategies for organic electrosynthesis. (a) Oxidation reaction. (b) Reduction reaction. (c) Convergently-paired electrolysis. (d) AC electrolysis. (e) Selective data for screening of reaction conditions. (f) 1H NMR spectra of selective data for screening of reaction conditions. Download figure Download PowerPoint In contrast to the direct current (DC) circuits in which electric current always flows in one direction, the voltage in alternating current (AC) circuits periodically reverses along with the current. Considering that both oxidation and reduction processes can occur at the same electrode and even almost the same time along with the electrode polarity reversal (Figure 1d), we envisioned that AC electrolysis might be the ideal choice for convergently-paired electrolysis. However, the strategic utilization of the concept of AC electrolysis in organic synthesis is scarce and remains a significant challenge.18–24,a In the search for an exclusive application of AC protocols in organic synthesis, we discovered that radical–radical cross-coupling to form an N–S bond represents a suitable class of reaction that cannot be delivered via DC electrolysis (Figures 1e and 1f). N–S bonds are versatile structural components, because they are ubiquitous, structural motifs present in many valuable molecules (Figure 2a),25–33 and reagents in organic transformations.34–40 Over the past decades, methods for N–S bond-containing compound synthesis have been largely developed.41–50 Although the nucleophilic substitution of amines with various sulfenylation reagents is proven as the most classical method (Figure 2b, top), most of these reactions require additional processes to synthesize sulfenylation reagents that often lead to large amounts of undesired by-products. Oxidative cross-coupling of commercially available thiophenols/thiols with amines has emerged as an increasingly popular method for N–S bond formation (Figure 2b, bottom). However, some drawbacks must still be improved, such as the over-oxidation of thiophenols/thiols due to the stoichiometric requirements of stoichiometric chemical oxidants. Figure 2 | State of the art for the formation of N–S bonds. (a) Important surrogates of N–S bond-containing compounds. (b) Previous methods for N–S bond formation. (c) Our design for N–S bond formation. Download figure Download PowerPoint Based on our idea of exploring AC electrolysis protocol and the urgent demand for developing new methods to construct N–S bonds, we demonstrate an AC-promoted radical–radical cross-coupling reaction under exogenous-oxidant-free conditions. As shown in Figure 2c, the electrochemical oxidation and following deprotonation of the amine (R2R3NH) form a nitrogen radical (R2R3N•). The disulfide intermediate is subsequently reduced51 at the same electrode to give the thiyl radical (R1S•) and thiolate anion (R1S−) intermediates upon electrode polarity reversal. After generating a nitrogen radical (R2R3N•) and thiyl radical (R1S•), the radical–radical cross-coupling produces the N–S coupled product. Experimental Methods The general procedure for the AC-promoted radical–radical cross-coupling reaction between thiophenols/thiols and anilines: In an oven-dried, undivided three-necked bottle equipped with a stir bar, nBu4NBF4 (0.2 mmol), thiophenol or thiol (0.3 mmol), a substituted aniline (0.6 mmol) were combined and added. The bottle was equipped with two platinum plate electrodes (15 × 15 × 0.3 mm). Under atmospheric conditions, 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP, 8 mL) and 1,2-dichloroethane (DCE, 3 mL) were added. The reaction mixture was stirred and electrolyzed at a constant voltage of 5 V at room temperature (RT) for 10 h. After completion of the reaction, the product was identified by thin-layer chromatography (TLC). The solvent was removed under reduced pressure by an aspirator, and then the pure product was obtained by flash column chromatography on silica gel. Results and Discussion At the initial stage, we selected 4-chlorothiophenol ( 1) and N-Ts-4-methoxyaniline ( 2) as model substrates to examine our designed protocol since sulfonamides are valuable structural motifs in drugs52–54 as well as the cross-coupling between N-Ts substituted anilines and thiophenols/thiols has not yet been reported. After a thorough optimization, we observed that when the electrolysis was performed with 0.2 mmol of nBu4NBF4 as the electrolyte, the mixture of DCE and HFIP (DCE/HFIP = 3/8) as the co-solvent, and a household transformer (sine waveform, 50 Hz) as the AC power supply, the desired N–S coupled product 3 could be isolated in 93% yield with an output voltage of 5 V for 10 h (Table 1, entry 1). Control experiments showed that AC is crucial for obtaining product 3 in excellent yield. In conditions without electricity, no reaction occurred (Table 1, entry 2). Using the strategy of DC electrolysis, the corresponding N–H/S–H cross-couplings showed poor selectivity and led to expected product 3 in only trace yields (Table 1, entries 3–6). For comparative purposes, a range of commonly used chemical oxidants were examined in the absence of electricity (Table 1, entries 7–14). However, all chemical oxidants used could not deliver the desired cross-coupling product 3, although disulfides could be detected in most cases. Platinum plate electrodes were significant for producing an N–S coupled product. Neither graphite plates nor carbon cloths could produce product 3 (Table 1, entries 15 and 16). Increasing the output voltage to 7.5 V furnished the product 3 in 89% yield, whereas when the reaction was conducted with an output voltage of 2.5 V for 10 h, no desired cross-coupling product was detected by TLC (Table 1, entries 17 and 18). Three milliliters of DCE and 8 mL of HFIP were found to be optimal for this N–H/S–H cross-coupling reaction. Performing the electrolysis with either 5 mL DCE/6 mL HFIP or 1 mL DCE/10 mL HFIP led to N–S coupled product 3 in 91% and 93% yields (Table 1, entries 19 and 20), respectively. In contrast, when the reaction was conducted in a mixture of DCE/2,2,2-trifluoroethanol (TFE) and MeCN/HFIP, only 17% and 50% yields of product 3 were obtained (Table 1, entries 21 and 22), respectively. The effect of supporting electrolytes was then examined. Both nBu4NPF6 and nBu4NClO4 displayed lower effectiveness than nBu4NBF4 (Table 1, entries 23 and 24). Finally, we evaluated the effect of O2 in the reaction. Conducting the reaction under an N2 atmosphere could still produce the cross-coupling product in excellent yields (Table 1, entry 25), ruling out the promotional effect of O2 to the reaction. Table 1 | Screening of Reaction Conditionsa Entry Variation from the Standard Conditions Yield (%) 1 None 93 2 No electricity n.d. 3 DC, 3 V, 10 h Trace 4 DC, 4 V, 10 h Trace 5 DC, 5 V, 10 h Trace 6 DC, 7 V, 10 h Trace 7 Using DDQ (2.0 equiv) instead of electricity n.d. 8 Using Mn(OAc)3 (2.0 equiv) instead of electricity n.d. 9 Using Phl(OAc)2 (2.0 equiv) instead of electricity n.d. 10 Using m-CPBA (2.0 equiv) instead of electricity n.d. 11 Using K2S2O8 (2.0 equiv) instead of electricity n.d. 12 Using NFSI (2.0 equiv) instead of electricity n.d. 13 Using O2 (balloon) instead of electricity n.d. 14 Using DDQ, Mn(OAc)3, Phl(OAc)2, m-CPBA, K2S2O8, NFSI, or O2 as oxidant, but at 60 °C n.d. 15 Graphite plates as electrodes n.d. 16 Carbon clothes as electrodes n.d. 17 AC, 2.5 V, 10 h n.d. 18 AC, 7.5 V, 10 h 89 19 DCE/HFIP = 5/6 91 20 DCE/HFIP = 1/10 93 21 DCE/TFE = 3/8 17 22 MeCN/HFIP = 3/8 50 23 nBu4NCIO4 instead of nBu4NBF4 82 24 nBu4NPF6 instead of nBu4NBF4 88 25 Under N2 atmosphere 93 Note: DDQ, 2,3-dichloro-5,6-dicyano-p-benzoquinone; m-CPBA, 3-chloroperoxybenzoic acid; NFSI, N-fluorobenzenesulfonimide. aReaction conditions: AC (50 Hz), two platinum plates as the electrodes, 5 V, 10 h, nBu4NBF4 (0.2 mmol), 1 (0.3 mmol), 2 (0.6 mmol), DCE (3 mL), HFIP (8 mL), RT, isolated yields. After the initial optimization, the substrate scope of this AC-promoted cross-coupling reaction has been evaluated (Table 2). First, a wide range of thiols was investigated. Thiophenols bearing either weak electron-withdrawing or -donating substituents delivered corresponding N–S coupled products in high to excellent yields (Table 2, 3–14). In contrast, thiophenols with strong electron-donating or -withdrawing group led to expected cross-coupling product in moderate to good yields (Table 2, 15 and 16). Note that halogen substituents, such as F, Cl, and Br, were well tolerated under the conditions of AC electrolysis (Table 2, 3–9), providing useful functionality for further modifications. Using unsubstituted thiophenol as the substrate, 93% yield of N–S coupled product could be provided (Table 2, 17). In addition to thiophenol derivatives, 2-naphthalenethiol was also tolerated, generating corresponding N–S bonds containing compounds in 81% yield (Table 2, 18). It is worth noting that the substrate scope of thiols could also be extended to aliphatic thiols. For example, using 2-propanethiol, 2-phenylethanethiol, 1-butanethiol, 3-methyl-2-butanethiol, cyclopentanethiol, or cyclohexanethiol as the reaction partners, the corresponding cross-coupling products could be obtained in 62–78% yields (Table 2, 19–24). Table 2 | Substrate Scope of Thiols and Aminesa aReaction conditions: AC (50 Hz), two platinum plates as the electrodes, 5 V, 10 h, nBu4NBF4 (0.2 mmol), thiols (0.3 mmol), amines (0.6 mmol), DCE (3 mL), HFIP (8 mL), RT, isolated yields. Next, various amines were employed to react with 4-chlorothiophenol. N–Ts substituted anilines bearing electron-donating groups (such as phenoxyl, methyl, and tert-butyl) delivered corresponding N–S coupled products in high to excellent yields (Table 2, 25–27). In contrast, when an electron-withdrawing group-substituted aniline was used as the coupling partner, the corresponding cross-coupling product was afforded in decreased yield (Table 2, 28). Ortho- and meta-substituted anilines were also compatible with the reaction conditions, generating corresponding cross-coupling products in 77–90% yields (Table 2, 30–32). Note that when N–Ts substituted aniline was used in the reaction, the corresponding cross-coupling product could be afforded in 65% yield (Table 2, 33). The benzenesulfonyl protecting groups were found to be important for developing desired products in high yields (Table 2, 34–37). For example, using 4-tert-butylbenzenesulfonyl, 4-methoxybenzenesulfonyl, 4-chlorobenzenesulfonyl, and benzenesulfonyl substituted aniline derivatives as the coupling partners, the corresponding cross-coupling reactions occurred smoothly and led to desired products in 78–83% yields (Table 2, 34–37). Note that besides sulfonyl groups, acyl-protecting groups were also tolerated. Either using aromatic or aliphatic acyl-protected anilines as reaction partners, the corresponding cross-coupling reactions occurred successfully and led to desired products in 60–81% yields (Table 2, 38–42). It is worth noting that the AC-promoted radical–radical cross-coupling reaction could be easily scaled up to a gram. For example, when the AC electrolysis of 1 and 2 on a 10 mmol scale was carried out with 2.0 mmol of nBu4NBF4 as the electrolyte and a mixture of DCE and HFIP (DCE/HFIP = 30/80) as the co-solvent, the desired N–S coupled product 3 could be isolated in 82% yield (Figure 3a, see details in Supporting Information Sections S2 and S3). Figure 3 | The utility of the AC-promoted radical–radical cross-coupling method. (a) Gram-scale synthesis. (b) The reaction of product 3 with GSH (43). Download figure Download PowerPoint As mentioned at the beginning of this research article, sulfonamides are valuable structural motifs in drugs. To further demonstrate the versatility of this AC promoted radical–radical cross-coupling reaction, the reaction of N–S coupled product 3 with glutathione (GSH) 43 (an abundant thiol in vivo) was conducted (Figure 3b). To our delight, both S–S coupled product 44, and the released sulfonamide product 2 could be detected (see details in Supporting Information Section S66). This result indicated that the N–S coupled products could react with GSH to release important sulfonamides into the blood using this method.26,32 That is, the prepared N–S coupled products may be used as potential prodrugs. To gain insight into the electrolysis of this AC promoted cross-coupling reaction, numerous experiments were planned (Figure 4). First, performing the electrolysis of 4-fluorothiophenol ( 45) in the absence of N-Ts-4-methoxyaniline ( 2) delivered the corresponding homo-coupling product 46 in 95% yield (Figure 4a). Second, using disulfide intermediate 46 as the cross-coupling partner to replace 4-fluorothiophenol ( 45) to react with N-Ts-4-methoxyaniline ( 2), the corresponding AC electrolysis occurred smoothly and led to desired cross-coupling product 5 in 74% yield (Figure 4b). These results suggest that disulfide might be the key intermediate for N–S bond formation. Third, running the electrolysis of N-Ts-4-methoxyaniline ( 2) in the absence of 4-chlorothiophenol ( 1) delivered the corresponding homo-coupling product 47 in 18% yield (Figure 4c). In addition, when 1.0 mmol of 1,1-diphenylethylene was added into the model reaction, 12% yield of intermolecular radical [3 + 2] annulation product 48 was afforded (Figure 4d).55 These results confirmed the formation of nitrogen radical intermediate. Cyclic voltammetry (CV) experiments were also carried out to clarify which substrate is preferentially oxidized (see details in Supporting Information Section S3). The first oxidation peak of 4-chlorothiophenol ( 1) was observed at 1.52 V, whereas N-Ts-4-methoxyaniline ( 2) showed the oxidation peak at 2.14 V. These results indicated that thiophenols are more easily oxidized than substituted anilines. Figure 4 | (a–d) Control experiments. Download figure Download PowerPoint Next, reactions between 4-fluorothiophenol ( 45) and N-Ts-4-methoxyaniline ( 2) with different reaction times were conducted. As shown in Figure 5, performing the electrolysis with an output voltage of 5 V for 2.5 h, product 5 could not be detected. Instead, disulfide intermediate 46 was obtained with 82% NMR yield (Figure 5a). Prolonging the electrolysis time to 5.0 h, in addition to 47% yield of 46, 36% yield of product 5 was also afforded (Figure 5b). Further prolonging the reaction time to 7.5 h, the yield of target product 5 increased to 85%, whereas the yield of disulfide intermediate 46 decreased to 3% (Figure 5c). It was found that when the electrolysis of 45 and 2 was conducted under an output voltage of 5 V for 10 h, the highest concentration of product 5 (91%) was obtained while disulfide intermediate 46 was hardly detected (Figure 5d). All together, these results indicated that disulfide is the key intermediate in this AC promoted radical–radical cross-coupling reaction. Figure 5 | (a–d) Electrolysis process studies. Reaction conditions: AC (50 Hz), two platinum plates as the electrodes, 5 V, 2.5–10 h, nBu4NBF4 (0.2 mmol), 45 (0.3 mmol), 2 (0.6 mmol), DCE (3 mL), HFIP (8 mL), RT, and the yields were determined by 19F NMR spectroscopy, with 1-fluoronaphthalene (49) as the internal standard. Download figure Download PowerPoint Control experiments were also conducted to explore the effect of frequency (f) on reaction. The experimental results are shown in Figure 6. Increasing the frequency to 75 or 100 Hz delivered desired product 5 in decreased yields (56% or 12%, Figures 6b and 6c). Moreover, controlling the frequency at 500 Hz, product 5 could not even be detected (Figure 6a). In contrast, the cross-coupling reaction took place smoothly at 25 Hz and gave a corresponding N–S coupled product in 85% yield (Figure 6e). Even when the reaction time was shortened to 9 h, product 5 could be produced with a 90% NMR yield (Figure 6f). Although 25 Hz showed a faster reaction rate than 50 Hz (Figure 6d), 50 Hz AC power supply is much cheaper and easier to obtain. In addition, decreasing the frequency to 0 Hz, that is, performing the reaction with a DC, only 39% yield of disulfide intermediate 46 was obtained (Figure 6g), indicating that AC is essential to obtain the desired product. Figure 6 | (a–g) The effect of frequency on the reaction. Reaction conditions: AC (25–500 Hz), two platinum plates as the electrodes, 5 V, 9–10 h, nBu4NBF4 (0.2 mmol), 45 (0.3 mmol), 2 (0.6 mmol), DCE (3 mL), HFIP (8 mL), RT, the yields were determined by 19F NMR spectroscopy, with 1-fluoronaphthalene (49) as the internal standard. Download figure Download PowerPoint Based on experimental outcomes and literature reports,51 a possible mechanism for this N–H/S–H cross-coupling reaction is proposed in Figure 7. In the beginning, the electrochemical oxidation and following deprotonation of thiol (R1SH) lead to the formation of thiyl radical (R1S•), which is unstable and then quickly dimerized to produce disulfide (R1SSR1) intermediate. Upon the voltage polarity reversal, H+ is reduced at the same electrode to give H2. Accompanied by the continuous accumulation of disulfide (R1SSR1) and the release of H2, the electrolysis reaches the next stage. When the potential of electrode is negative enough to reduce the disulfide (R1SSR1), the disulfide intermediate will be reduced to access the corresponding disulfide radical anion and then thiyl radical (R1S•) and thiolate anion (R1S−). Upon the voltage polarity reversal, the electrochemical oxidation and following deprotonation of amine (R2R3NH) result in the formation of a nitrogen-centered radical (R2R3N•). The radical–radical cross-coupling of thiyl radical (R1S•) with nitrogen radical (R2R3N•) finally produces the N–S coupled product. Note that for the electrolysis with an AC, there is no difference between two electrodes and the same reaction also takes place on the other electrode. In addition, the generated thiolate anion (R1S−) is not a by-product, but an intermediate, and will be converted to the corresponding thiyl radical (R1S•) after losing an electron. Figure 7 | Proposed mechanism. Download figure Download PowerPoint Conclusions An AC-promoted N–H/S–H radical–radical cross-coupling reaction has been successfully developed. Mechanistic investigations indicated that disulfide is the crucial intermediate in this AC promoted cross-coupling reaction. The N–H/S–H cross-coupling method described herein reveals the unique advantages of AC electrolysis in realizing the reactions that cannot occur through DC electrolysis. We anticipate this work will stimulate research interest of chemists and pave new avenues to innovation in organic electrosynthesis. Footnote a During the preparation of this manuscript, Phil S. Baran, Yu Kawamata and and co-workers uploaded to ChemRxiv a manuscript also using the strategy of AC electrolysis. See ref 24. Supporting Information Supporting Information is available and includes the general information, experimental procedure, analytical data of products, references, and NMR spectra of products. Conflict of Interest There is no conflict of interest to report. Acknowledgments This work was supported by the National Natural Science Foundation of China (no. 22031008) and the Science Foundation of Wuhan (no. 2020010601012192).