Electrochemical Sulfenylation of 4-Hydroxycoumarins with Aryl Thiols Catalyzed by Potassium Iodide

Unless otherwise noted, all commercially available solvents and chemicals were purchased from commercial suppliers and were used without further purification. 4-Hydroxy-6-methyl-2H-chromen-2-one (1b) and 4-hydroxy-6-methoxy-2H-chromen-2-one (1c) were synthesized according to the procedure of literature. ^{37 1}H NMR and ¹³C NMR spectra were recorded with DMSO-d₆ as the solvent and tetramethylsilane (TMS) as the internal standard at a Bruker Avance III HD spectrometer at 500 MHz for ¹H NMR and 125 MHz for ¹³C NMR.

Cyclic voltammetry was performed in CH₃CN solution (15 ml) using NaBF₄ (0.1 mol l⁻¹) as the supporting electrolyte in 25 ml undivided cell to determine the redox properties of each compound. A three-electrode system was adopted in the experiments, containing L-type Pt electrode (3 mm in diameter) as the working electrode, platinum sheet (2.25 cm²) as the counter electrode, and Ag/Ag⁺ electrode (0.1 mol l⁻¹ AgNO₃ in CH₃CN) as the reference electrode. Potential was applied using CHI600e electrochemical workstation (CH Instrument Inc., Austin, TX, USA).

In situ FTIR spectroscopy experiments were carried out on Nicolet 670 FTIR spectrometer equipped with a MCT-A detector cooled by liquid nitrogen. The working electrode was a Pt disk (6 mm in diameter). A three-electrode spectro-electrochemical cell with CaF₂ window at the bottom was used. IR spectra was calculated as:

$$\begin{eqnarray}&&\displaystyle \frac{{\rm{\Delta }}R}{R}=\displaystyle \frac{\left[R\left({E}_{S}\right)-R\left({E}_{R}\right)\right]}{R({E}_{R})}\end{eqnarray} \tag{ 5 }$$

In the above equation, R (E _S ) and R (E _R ) were the single beam spectra separately received at the sample potential (E _S ) and the reference potential (E_R). ^38–40 The reference potential (E _R ) was selected at 0 mV. Each spectrum was collected at 8 cm⁻¹ resolution at different potentials. ΔR/R represented the potential difference of the single-beam spectra collected at different potentials.

4-Hydroxycoumarins had been prepared according to Scheme 2. The starting diverse substituted malonic acid mono phenol esters were synthesized by Meldrum's acid with different substituted phenol in high yield under solvent-free condition, then cyclized by Eaton's reagent in mild conditions to give corresponding 4-hydroxycoumarins. ³⁷

Zoom In Zoom Out Reset image size

Electrolysis experiments were performed on Vertex Potentiostat/Galvanostat. Both the working electrode and the counter electrode were platinum electrodes (1.5 cm × 1.5 cm). The reference electrode was Ag/Ag⁺ electrode (0.1 mol l⁻¹ AgNO₃ in CH₃CN). 4-Hydroxycoumarin (1a, 0.5 mmol, 81.1 mg), 4-chlorothiophenol (2a, 0.75 mmol, 108.5 mg), KI (0.05 mmol, 8.3 mg) and 0.1 mol l⁻¹ NaBF₄/CH₃CN solution (15 ml) were added into 25 ml undivided cell. The electrolysis reactions were operated at 0.4 V under 60 °C for 6 h. The resulting mixture was concentrated under reduced pressure and purified by column chromatography on silica gel using petroleum ether/ethyl acetate (5:1) as eluent to afford 3-((4-chlorophenyl)thio)-4-hydroxy-2H-chromen-2-one (3aa) as a white solid in 94% yield.

4-Hydroxy-6-methyl-2H-chromen-2-one ( 1 b). —Yield: 71%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 12.43 (s, 1H), 7.58 (d, J = 1.15 Hz, 1H), 7.42–7.40 (m, 1H), 7.22 (d, J = 8.45 Hz, 1H), 5.56 (s, 1H), 2.35 (s, 3H); ¹³C NMR (125 MHz, DMSO-d₆) δ 165.6, 162.0, 151.7, 133.4, 133.1, 122.8, 116.1, 115.5, 90.9, 20.3.

4-Hydroxy-6-methoxy-2H-chromen-2-one ( 1 c). — Yield: 64%; Light red solid; ¹H NMR (500 MHz, DMSO-d₆) δ 12.49 (s, 1H), 7.29 (d, J = 9.3 Hz, 1H), 7.22–7.19 (m, 2H), 5.59 (s, 1H), 3.80 (s, 3H); ¹³C NMR (125 MHz, DMSO-d₆) δ 165.3, 162.0, 155.3, 147.9, 120.3, 117.6, 116.2, 105.0, 91.3, 55.6.

3-((4-Chlorophenyl)thio)-4-hydroxy-2H-chromen-2-one ( 3 aa). — Yield: 94%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.98–7.96 (m, 1H), 7.74–7.70 (m, 1H), 7.44–7.39 (m, 2H), 7.34–7.32 (m, 2H), 7.24–7.21 (m, 2H); ¹³C NMR (125 MHz, DMSO-d₆) δ 168.6, 160.8, 153.1, 135.2, 133.8, 130.2, 129.0, 128.0, 124.4, 124.3, 116.5, 115.8, 94.1.

3-((4-Fluorophenyl)thio)-4-hydroxy-2H-chromen-2-one ( 3 ab). — Yield: 90%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.97–7.96 (m, 1H), 7.72–7.69 (m, 1H), 7.42–7.39 (m, 2H), 7.30–7.28 (m, 2H), 7.15–7.12 (m, 2H); ¹³C NMR (125 MHz, DMSO-d₆) δ 168.7, 161.4, 160.7 (d, J = 201.3 Hz), 153.5, 134.1, 131.8 (d, J = 2.5 Hz), 129.5 (d, J = 6.3 Hz), 124.9, 124.7, 116.9, 116.5 (d, J = 18.8 Hz), 116.2, 95.5.

3-((4-Bromophenyl)thio)-4-hydroxy-2H-chromen-2-one ( 3 ac). — Yield: 73%; Light yellow solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.98–7.96 (m, 1H), 7.73–7.70 (m, 1H), 7.46–7.39 (m, 4H), 7.16 (d, J = 7.2 Hz, 2H); ¹³C NMR (125 MHz, DMSO-d₆) δ 169.2, 161.3, 153.6, 136.3, 134.2, 132.3, 128.8, 124.9, 124.7, 118.8, 117.0, 116.3, 94.3.

4-Hydroxy-3-(p-tolylthio)-2H-chromen-2-one ( 3 ad). — Yield: 71%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.96–7.94 (m, 1H), 7.71–7.67 (m, 1H), 7.42–7.37 (m, 2H), 7.13–7.08 (m, 4H), 2.23 (s, 3H); ¹³C NMR (125 MHz, DMSO-d₆) δ 167.9, 160.8, 152.9, 135.2, 133.5, 132.2, 129.7, 127.0, 124.3, 124.2, 116.4, 115.6, 95.3, 20.4.

4-Hydroxy-3-((4-methoxyphenyl)thio)-2H-chromen-2-one ( 3 ae). — Yield: 71%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.95–7.94 (m, 1H), 7.69–7.67 (m, 1H), 7.40–7.37 (m, 2H), 7.29–7.26 (m, 2H), 6.90–6.87 (m, 2H), 3.71 (s, 3H); ¹³C NMR (125 MHz, DMSO-d₆) δ 167.8, 161.3, 158.7, 153.3, 134.0, 130.5, 126.3, 124.8, 124.7, 116.9, 116.1, 115.3, 97.2, 55.7.

3-((2-Chlorophenyl)thio)-4-hydroxy-2H-chromen-2-one ( 3 af). — Yield: 75%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.99–7.98 (m, 1H), 7.75–7.72 (m, 1H), 7.47–7.40 (m, 3H), 7.22–7.15 (m, 2H), 6.97–6.96 (m, 1H); ¹³C NMR (125 MHz, DMSO-d₆) δ 169.5, 160.8, 153.2, 135.2, 133.8, 129.5, 127.7, 126.3, 125.8, 124.4, 124.2, 116.6, 116.0, 92.1.

3-((3-Chlorophenyl)thio)-4-hydroxy-2H-chromen-2-one ( 3 ag). — Yield: 94%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.97–7.95 (m, 1H), 7.73–7.69 (m, 1H), 7.43–7.38 (m, 2H), 7.32–7.27 (m, 1H), 7.23–7.15 (m, 3H); ¹³C NMR (125 MHz, DMSO-d₆) δ 168.9, 160.9, 153.1, 138.8, 133.8, 133.7, 130.1, 125.5, 125.4, 124.8, 124.5, 124.2, 116.5, 115.8, 93.5.

4-Hydroxy-3-(naphthalen-2-ylthio)-2H-chromen-2-one ( 3 ah). — Yield: 95%; Light yellow solid; ¹H NMR (500 MHz, DMSO-d₆) δ 8.01–7.99 (m, 1H), 7.85–7.80 (m, 3H), 7.74–7.70 (m, 2H), 7.47–7.40 (m, 4H), 7.38–7.36 (m, 1H); ¹³C NMR (125 MHz, DMSO-d₆) δ 168.5, 160.9, 153.1, 133.6, 133.5, 133.4, 131.1, 128.6, 127.6, 126.9, 126.6, 125.5, 124.9, 124.4, 124.2, 123.9, 116.5, 115.8, 94.4.

4-Hydroxy-3-(thiophen-2-ylthio)-2H-chromen-2-one ( 3 ai). — Yield: 39%; Light yellow solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.94–7.92 (m, 1H), 7.68–7.64 (m, 1H), 7.52–7.51 (m, 1H), 7.38–7.35 (m, 2H), 7.29–7.28 (m, 1H), 6.99–6.98 (m, 1H); ¹³C NMR (125 MHz, DMSO-d₆) δ 166.8, 160.7, 152.8, 133.6, 133.2, 132.2, 129.4, 127.5, 124.4, 124.3, 116.4, 115.6, 98.2.

4-Hydroxy-6-methyl-3-((4-chlorophenyl)thio)-2H-chromen-2-one ( 3 ba). — Yield: 77%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.75 (d, J = 1.0 Hz, 1H), 7.52–7.51 (m, 1H), 7.34–7.30 (m, 3H), 7.22–7.20 (m, 2H), 2.39 (s, 3H); ¹³C NMR (125 MHz, DMSO-d₆) δ 169.0, 161.4, 151.7, 135.7, 135.1, 134.1, 130.6, 129.4, 128.4, 124.4, 116.8, 115.8, 94.4, 20.8.

4-Hydroxy-6-methoxy-3-((4-chlorophenyl)thio)-2H-chromen-2-one ( 3 ca). — Yield: 74%; White solid; ¹H NMR (500 MHz, DMSO-d₆) δ 7.38–7.36 (m, 2H), 7.34–7.33 (m, 2H), 7.31–7.28 (m, 1H), 7.22 (d, J = 1.7 Hz, 1H), 7.21 (d, J = 1.6 Hz, 1H), 3.83 (s, 3H); ¹³C NMR (125 MHz, DMSO-d₆) δ 168.8, 161.4, 156.0, 147.9, 135.6, 130.6, 129.4, 128.5, 122.1, 118.3, 116.5, 106.4, 94.8, 56.2.

Electrochemical sulfenylation of 4-hydroxycoumarin (1a, 0.5 mmol, 81.1 mg) with 4-chlorothiophenol (2a, 0.75 mmol, 108.5 mg) catalyzed by KI (0.05 mmol, 8.3 mg) in NaBF₄/CH₃CN solution (0.1 mol l⁻¹, 15 ml) was studied by cyclic voltammetry. Cyclic voltammgrams of 1a, 2a and KI in the solution of NaBF₄/CH₃CN were collected respectively with potentials ranged from −0.6 V to 0.8 V vs Ag/Ag⁺ electrode at the scan rate of 50 mV s⁻¹, and results were shown in Fig. 1. Two typical oxidation peaks of iodide ions were appeared at −0.05 V and 0.32 V in curve b, which corresponded to the oxidation of I⁻ to I₃ ⁻ and I₃ ⁻ to I₂, respectively. ⁴¹ As shown in curve a, no obvious redox peak was observed, indicating that 1a was difficult to be oxidized in the potential range of −0.6 V–0.8 V. While oxidation current of 2a started to increase gradually at about 0.38 V (curve c).

Zoom In Zoom Out Reset image size

Fig. 2 (A) explored the reaction of 2a in the presence of KI in NaBF₄/CH₃CN solution. The oxidation current for the mixture of 2a and KI (curve e) increased sharply compared with curves b and c, and the potential of oxidation peak moved from 0.32 V to 0.54 V. It indicated I₂ had high catalytic activity and 2a could be rapidly oxidized in the presence of I₂.

Zoom In Zoom Out Reset image size

In situ FTIR spectroscopy was an effective technique to study the surface reaction characteristics of the electrode surface. To further explore the consumption of reactants and the generation of intermediates or products in the reaction process, in situ FTIR spectroscopy was introduced. In situ FTIR spectra collected during the reaction of 2a and KI in 0.1 mol l⁻¹ NaBF₄/CH₃CN solution at the potential of 300 mV were shown in Fig. 2 (B). The positive-going bands represented the consumption of reactants, and the negative-going bands represented the formation of intermediates or products. ^42–46 The mixture of 2a and KI in NaBF₄/CH₃CN solution (curve ⅱ) showed obvious infrared signals compared with 2a in NaBF₄/CH₃CN solution (curve ⅰ). The positive-going bands at 2562 cm⁻¹ and 1485 cm⁻¹ were respectively attributed to the S-H stretching vibration and the aromatic ring skeleton vibration of 2a, which indicated that 2a was consumed. The negative-going band 1472 cm⁻¹ could be attributed to the characteristic peak of the aromatic ring skeleton vibration of 1,2-bis(4-chlorophenyl)disulfane produced by 2a. It showed that the corresponding 1,2-bis(4-chlorophenyl)disulfane was generated during the reaction. ⁴⁷

When KI was added into the NaBF₄/CH₃CN solution with 1a, the corresponding cyclic voltammetric curve was shown in Fig. 3 (A). As compared with curve b, the oxidation current of the mixture of 1a and KI (curve d) had a slight increase, which indicated that 1a had a small voltammetric response at the electrode in the presence of I₂. To further verify the result of cyclic voltammetry, we collected in situ FTIR spectra of 1a and KI in 0.1 mol l⁻¹ NaBF₄/CH₃CN solution with the time. In order to clearly observe the change of spectra signals, the in situ FTIR spectra was collected at 500 mV. As shown in Fig. 3 (B), with the increase of scanning time, the negative-going band of 977 cm⁻¹ could be clearly observed which was the signal of C-I bond generated from the reaction of 1a and I₂. The reaction of 4-hydroxycoumarin and I₂ to form 4-hydroxy-3-iodo-2H-chromen-2-one was shown as the Eq. 6 ⁴⁸ :

Zoom In Zoom Out Reset image size

When 1a, 2a and KI were added into the NaBF₄/CH₃CN solution, oxidation peak current at about 0.5 V (curve f in Fig. 4 (A)) increased compared with curve e, but the oxidation peak current at about −0.1 V (curve f) did not increase. It suggested that the electrochemical oxidation process of I₂ and 4-hydroxycoumarin was restrained when 2a was added to the solution of 1a and KI, and the product was mainly formed through the intermediate disulfane instead of 4-hydroxy-3-iodo-2H-chromen-2-one.

Zoom In Zoom Out Reset image size

Figure 4 (B) showed the changes of functional groups during the reaction of 1a, 2a and KI in 0.1 mol l⁻¹ NaBF₄/CH₃CN solution at potentials varied from 0 mV to 700 mV by in situ FTIR spectroscopy. With the increase of potential, starting from 300 mV, the infrared signals of positive-going bands (2562, 1726, 1637, 1571 cm⁻¹) and negative-going bands (1693, 1613, 1544, 1515, 1472, 1225, 1155, 1092 cm⁻¹) could be clearly observed. The positive-going band at 2562 cm⁻¹ was attributed to the S-H stretching vibration of 2a, which indicated that 2a was consumed. The positive-going bands at 1726 cm⁻¹ and 1637 cm⁻¹ could be attributed to the C=O stretching vibration and C=C stretching vibration of the six-membered lactone ring of 1a, respectively. In addition, the positive-going band located at 1571 cm⁻¹ could be attributed to a characteristic peak of the aromatic ring skeleton vibration of 1a. ^49,50

When ArS- was attached to 1a, due to the conjugation effect of benzene ring and lone pair electrons of sulfur, the infrared bands of the corresponding C=O and C=C stretching vibration would red-shift. ^51–54 Therefore, the negative-going band at 1693 cm⁻¹ was assigned to the C=O stretching vibrations of 3aa. However, it was not obvious because of the coverage by the positive-going band at 1725 cm⁻¹. The 1613 cm⁻¹ of the negative-going band could be attributed to the C=C stretching vibrations of 3aa. The signals of 1544 cm⁻¹, 1515 cm⁻¹ and 1472 cm⁻¹ could be attributed to the characteristic peaks of the aromatic ring skeleton vibration of 3aa. ^49,50 The signal located at 1092 cm⁻¹ might be caused by the synergistic vibration of C–S bond and C–Cl bond. Two strong negative-going bands around 1225 cm⁻¹ and 1155 cm⁻¹ could be assigned to the symmetric and asymmetric stretching of aryl ether group of 3aa. ^49,50 Therefore, 3aa was synthesized while 1a and 2a were consumed during the electrochemical process.

We commenced electrolysis experiments by employing 1a and 2a as the model substrates to optimize the reaction conditions (Table I). The initial studies began with a reaction of 1a (0.5 mmol, 81 mg) and 2a (0.75 mmol, 108 mg, 1.5 equiv.) in 0.1 mol l⁻¹ NaBF₄/CH₃CN solution containing 10 mol% KI at a constant voltage of 0.6 V at 60 °C in a 25 ml undivided cell. The reaction process was monitored by HPLC. It was found that 1a was almost completely consumed in 6 h. The desired C–S coupling product 3-((4-chlorophenyl)thio)-4-hydroxy-2H-chromen-2-one (3aa) was obtained in 89% yield (entry 1). The reaction temperature played an important role in this transformation. Decreasing the temperature from 60 °C to 45 °C and 25 °C, the desired product 3aa was given in 41% and 25% yields, respectively (entries 2 and 3). Therefore, to get a satisfactory yield, 60 °C was necessary. The effect of reaction potential was also investigated and the yield of desired product 3aa was increased from 89% to 94% when the reaction potential was reduced from 0.6 V to 0.4 V (entry 4). High potential would promote the side reaction of 1a and I₂ to form 4-hydroxy-3-iodo-2H-chromen-2-one. ⁴⁸ While the reaction potential was dropped to 0.2 V, the yield of 3aa was only 28% (entry 5). So, 0.4 V was suitable in the following experiments.

The effect of the amount of catalyst on the reaction was also studied. It was ideal to achieve high yield of 3aa with low amount of KI. However, when the amount of KI was decreased from 10 mol% to 5 mol%, the yield of 3aa was decreased from 94% to 62% (entry 6). Furthermore, 3aa could not be formed in the absence of KI (entry 7). Thus, 10 mol% KI was necessary for the next experiments. In addition, the loading of 2a had a significant impact on the reaction yield. When the loading of 2a was decreased from 1.5 equiv. to 1.2 equiv. or 1 equiv., the desired product was obtained in 80% and 74% yields, respectively (entries 8 and 9). It could be seen that excessive 2a was beneficial to the sulfenylation reaction. ⁵⁵ Hence, combining the above experimental results, 1a with 1.5 equiv. of 2a in the presence of 10 mol% KI in 0.1 mol l⁻¹ NaBF₄/CH₃CN solution at constant potential 0.4 V at 60 °C for 6 h were chosen as the optimized reaction conditions for the synthesis of 3aa.

With the optimized reaction conditions in hand, the substrate scope and universality of this electrochemical sulfenylation with different 4-hydroxycoumarins 1 and aryl thiols 2 were explored (Scheme 3). Firstly, the influence of aryl thiols with different substituents on the reaction was studied. The reactions of 1a and aryl thiols with electron-withdrawing groups (Cl, F and Br) and electron-donating groups (CH₃ and OCH₃) at the 4-position were favorable and afforded the desired products 3aa–3ae in good to excellent yields. Specifically, the electron-withdrawing substituents worked with better yields than electron-donating substituents under the standard reaction protocol. Probably the presence of electron-withdrawing groups would reduce the electron cloud density of the thio radicals which facilitated the adding of thio radicals to the electron-rich heteroaromatic compound 1a. ⁵⁶ In addition, aryl thiols with substituents 2-Cl- or 3-Cl- successfully reacted with 1a to obtain the desired products 3af and 3ag in 75% and 94% yields, respectively. The relatively low yield of 3af might be due to the steric effect. Naphthalene-2-thiol could react well to give the corresponding product 3ah in 95% yield. To our delight, athiophene-2-thiol (3i) could also react with 1a successfully to give the 3ai in 39% yield.

Zoom In Zoom Out Reset image size

Subsequently, 4-hydroxycoumarins 1b and 1c were used as the substrates to react with 2a, and the corresponding sulfenylated 4-hydroxycoumarins 3ba and 3ca could be obtained in moderate yields in 10 h. It was worth noting that 1b and 1c could not be completely converted to their desired products in 10 h. By prolonging the reaction time, the yields of 3ba and 3ca could be further improved.

To ascertain the mechanism of this sulfenylation reaction, we conducted several control experiments (Scheme 4). It could be found that 2a was almost completely converted to 1,2-bis(4-chlorophenyl)disulfane (4) in the presence of KI under the standard reaction conditions (Eq. 7). The reaction of 1a and 4 under the standard conditions could still obtain the product 3aa in 96% isolated yield (Eq. 8). Encouraged by this result, we assumed that disulfide (4) should be the intermediate in the sulfenylation process. Furthermore, to check the involvement of radical intermediates in the reaction, the radical scavenger 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO, 2 equiv.) was employed in the reaction of 4 and 1a, resulting in the formation of 3aa in 47% yield (Eq. 9). Compared with Eq. 8, the yield of 3aa dropped sharply from 96% to 47%, indicating that this sulfenylation reaction existed in a radical process. However, the reaction could not be completely inhibited by TEMPO, so a non-radical pathway was still involved.

Zoom In Zoom Out Reset image size

Based on the above control experiment results and previous reports, ^32,57,58 two possible mechanisms for this electrochemical sulfenylation of 1a were proposed in Scheme 5. Pathway A, iodide ion lost two electrons to generate molecular iodine at the anode. And then molecular iodine oxidized 2a into intermediate 4. Subsequently, the free radical 5 was formed from 4 and reacted with 1a to produce intermediate product 7, which lost hydrogen atom under the action of molecular iodine to release the product 3aa. ⁵⁹ While hydrogen ions were reduced to H₂ at the cathode.

Zoom In Zoom Out Reset image size

Pathway B reflected the non-radical process of this reaction. Initially, the oxidation of 2a under the electrophilic molecular iodine afforded the disulfide 4. The reaction of 4 with I₂ formed an electrophilic species 4-Cl-PhSI (6). ^32,57,58 The intermediate 6 underwent the electrophilic attack of 1a to generate the intermediate 8. Finally, product 3aa was obtained by releasing HI from intermediate 8.