Electrocatalytic hydrogenation of cyanoarenes, nitroarenes, quinolines, and pyridines under mild conditions with a proton-exchange membrane reactor

Introduction

Nitrogen-containing molecules are important bioactive compounds and intermediates in chemical synthesis. Therefore, the chemical transformations of nitrogen-containing compounds have been widely studied in the field of organic synthesis . For instance, the reduction of cyanoarenes is a straightforward and powerful method for the synthesis of primary amines , and the reduction of nitroarenes is useful for the synthesis of aniline derivatives . Nitrogen-containing aliphatic heterocycles, such as piperidines and tetrahydroquinolines, are key motifs in pharmaceuticals, and the reductive syntheses of these heterocycles from pyridines and quinolines have been well studied . Although these transformations have been studied intensively, such reductive reactions usually require harsh reaction conditions such as high reaction temperatures and high pressure of hydrogen .

Meanwhile, electrochemical systems using solid polymer electrolytes (SPEs) have recently attracted significant attention . Among these, proton-exchange membrane (PEM) reactors are powerful tools for hydrogenation . The PEM reactor included a membrane electrode assembly (MEA) consisting of a PEM and an electro-catalyst supported on carbon (Figure 1). Humidified hydrogen gas (H2) or H2O was injected into the anodic chamber and the substrate passed through the cathodic chamber. The hydrogen (H2) or H2O were oxidized at the anode to form protons (H+) that moved to the cathodic chamber, and the protons were reduced to monoatomic hydrogen species (absorbed hydrogen, Had). Thus-generated Had reduced the substrate passed through the cathodic chamber. MEA eliminates the need for a supporting electrolyte, which is necessary for conventional organic electrolysis, reduces the environmental impact, and facilitates product purification. In addition, using nanoparticles in the catalyst layer, which serve as the electrode, results in a large specific surface area and efficient reactions. As PEM reactors are flow reactors, they have an advantage over batch reactors in terms of continuous production.

Figure 1: Schematic of (a) a PEM reactor and (b) MEA.

Several reductive transformations taking advantage of the characteristics of the PEM reactor have been reported in recent years. For example, Atobe et al. reported the electrocatalytic semihydrogenation of alkynes to form Z-alkenes using a PEM reactor . The Pd/C catalyst was essential for the reaction. They recently found that a PEM reactor with a Rh/C catalyst was effective for the stereoselective reduction of cyclic ketones . Nagaki et al. reported the electrochemical deuteration of aryl halides . PEM reactors are also gaining industrial attention. Weber et al. reported a series of large-scale syntheses using PEM reactors .

We studied electrochemical transformations and recently reported the selective reduction of enones using a PEM reactor . As mentioned above, there have been several reports on reductive reactions using PEM reactors; however, the application of PEM reactors for precise chemical transformations remains limited. To the best of our knowledge, no reports are available on the efficient reduction of cyanoarenes, quinolines, and pyridines using PEM reactors .

In this context, we have focused on the synthesis of nitrogen-containing molecules using a PEM reactor. Herein, we report the application of a PEM reactor for the reduction of cyanoarenes, nitroarenes, quinolines, and pyridines. These reductions proceeded smoothly to afford benzylamines, anilines, tetrahydroquinolines, and piperidines using a PEM reactor under ambient conditions.

Results and Discussion Reduction of cyanoarenes to benzylamines

Benzonitrile (1a) was chosen as the model substrate, and the electroreductive hydrogenation of 1a was performed with a PEM reactor (Table 1). Humidified hydrogen was used as a proton source, and 4.0 F mol−1 of electricity was passed to a circulated solution of 1a (for the details, see the Supporting Information File 1). When Pd/C was used as the cathode catalyst, benzylamine (2a) was not obtained (Table 1, entry 1). Ru/C, Pt/C, and an alloy catalyst PtRu/C were also ineffective (Table 1, entries 2–4). Further screening revealed that trace amounts of 2a were obtained when an alloy catalyst PtPd/C was used as the cathode catalyst (Table 1, entry 5). Although the desired compound 2a was obtained using PtPd/C, undesired dibenzylamine (3a), was obtained as a major product . To suppress the generation of 3a, we examined the effect of solvent (Table 1, entries 6 and 7). When the reaction was performed in a mixed solvent consisting of CH2Cl2 and 1,1,1,3,3,3-hexafluoropropan-2-ol (HFIP), the yield of 2a increased; however, the generation of 3a was not completely suppressed (Table 1, entry 6). The use of CH2Cl2/2,2,2-trifluoroethanol (TFE) gave similar results (Table 1, entry 7).

Table 1: Effect of cathode catalyst for the electrochemical reduction of 1a using a PEM reactora.

entry cathode catalyst 2a (%)b 3a (%)b recovered 1a (%)b 1 Pd/C N.D.c N.D. 100 2 Ru/C N.D. N.D. 100 3 Pt/C N.D. N.D. 100 4 PtRu/C N.D. 4 85 5 PtPd/C 4 18 78 6d PtPd/C 11 7 79 7e PtPd/C 13 19 65

aReaction conditions: anode catalyst, Pt/C; 1a, 2.5 mmol; solvent, CH2Cl2/EtOH (4:1, 0.5 M); flow rate of the solution of 1a, 0.25 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2. The solution was circulated until the passage of 4.0 F mol−1 (1 h 20 min). bArea ratio determined by gas chromatography analysis. cNot detected. dPerformed using HFIP instead of EtOH. ePerformed using TFE instead of EtOH.

A plausible mechanism for the reduction of 1a is shown in Scheme 1. First, the reduction of 1a afforded phenylmethanimine (A). Further reduction of A afforded the desired benzylamine (2a). However, nucleophilic attack of 2a on A, followed in situ by reduction, proceeded competitively to form dibenzylamine (3a). We considered that by protonating 2a to form 2a-H+, its nucleophilic nature could be suppressed, thereby inhibiting the formation of 3a.

Scheme 1: Plausible mechanism for the reduction of 1a leading to benzylamine 2a and dibenzylamine 3a.

Based on this hypothesis, we examined the reduction of 1a in the presence of several acids (Table 2). First, electrochemical reduction was performed with 0.2 or 1.0 equiv of acetic acid, but the yield of 3a did not decrease, suggesting that 2a could not be trapped by acetic acid (pKa = 4.75). To ensure the capture of 2a, we performed reduction with phosphoric acid (pKa = 2.12). Although the generation of 3a was suppressed, only a trace amount of 2a was obtained, and almost 1a was recovered (Table 2, entry 4). This was probably because the presence of water inhibited the reaction. Therefore, it was necessary to perform the reaction under anhydrous conditions. Hence, we used ethyl phosphate (mono- and di-mixture) (pKa = 1.42), which reacts easily under anhydrous conditions. As expected, the generation of 3a was suppressed and 2a was selectively obtained (Table 2, entry 5). With the increase of the electricity to 16.0 F mol−1, 2a was obtained selectively in 88% yield (Table 2, entry 6).

Table 2: Electrochemical reduction of 1a in the presence of acidsa.

entry acid (equiv) 2a (%)b 3a (%)b recovered 1a (%)b 1 none 4 18 78 2 CH3COOH (0.2) 2 15 82 3 CH3COOH (1.0) 3 22 74 4 H3PO4 (1.0) 1 N.D.c 99 5 ethyl phosphated (1.0) 18 trace 82 6e ethyl phosphated (1.0) 88f tracef 7f

aReaction conditions: anode catalyst, Pt/C; cathode catalyst, PtPd/C; 1a, 2.5 mmol; solvent, CH2Cl2/EtOH (4:1, 0.5 M); flow rate of the solution of 1a, 0.25 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2. The solution was circulated until the passage of 4.0 F mol−1 (1 h 20 min). bArea ratio determined by gas chromatography analysis. cNot detected. dMono- and di- ester mixture. e16.0 F mol−1 (5 h 20 min). fGC yield determined by GC analysis using dodecane as an internal standard.

Next, we examined the scope of the electrochemical reduction of cyanoarenes under optimal conditions (Scheme 2). The reactions of cyanoarenes bearing electron-donating methyl, hydroxy, and methoxy groups proceeded smoothly to afford the corresponding benzylamines 2b–d in moderate-to-good yields. Electron-withdrawing groups, such as esters, can be tolerated under these conditions. Unfortunately, the electrochemical reduction of 1,4-dicyanobenzene (1f) did not give the desired product 2f probably due to the oligomerization of the substrate.

Scheme 2: Electrochemical reduction of cyanoarenes under optimal conditions. Reaction conditions: anode catalyst, Pt/C; cathode catalyst, PtPd/C; 1, 2.5 mmol; solvent, CH2Cl2/EtOH (4:1); flow rate of the solution of 1, 0.25 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2. The solution was circulated until the passage of 16.0 F mol−1 (5 h 20 min). aDetermined by GC analysis using n-dodecane as an internal standard. bDetermined by 1H NMR analysis using 1,1,2,2-tetrachloroethane as an internal standard. cNot detected.

Reduction of nitroarenes to anilines

Next, we reduced nitroarenes using a PEM reactor. First, the electrocatalyst and solvent were optimized (Table 3). While 6.0 F mol−1 of electricity should be required for the reduction of nitrobenzene (4a) to aniline (5a), the charge for screening was set to 2.0 F mol−1 for rapid evaluation, and several cathode catalysts were examined. When Pd/C was used as the cathode, the reaction did not proceed (Table 3, entry 1). The desired reduction proceeded with Ru/C and 5a was obtained in 79% of current efficiency (Table 3, entry 2). Pt/C afforded the best result (90% current efficiency, Table 3, entry 3). To increase the yield, the reaction was carried out until 4a was consumed. After 7 h of electrolysis (23.2 F mol−1), 4a was completely consumed and 5a was obtained in 82% yield. Although Ir/C was inefficient (Table 3, entry 4), Rh/C was as efficient as Pt/C (Table 3, entry 5). As Rh is more expensive than Pt, Pt/C was selected as the best cathode catalyst.

Table 3: Electrochemical reduction of 4a with several cathode catalystsa.

entry cathode catalyst 5a (%)b current efficiency (%)b recovered 4a (%)b 1 Pd/C N.D.c – – 2 Ru/C 25 79 65 3 Pt/C 30 (82)d 90 (20)d 76 (N.D.)d 4 Ir/C 6 18 70 5 Rh/C 29 90 59

aReaction conditions: anode catalyst, Pt/C; 4a, 5 mmol; solvent, CH2Cl2 (1.0 M); flow rate of the solution of 4a, 0.25 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2. The solution was circulated until the passage of 2.0 F mol−1 (1 h 20 min). bDetermined by GC using n-dodecane as an internal standard. cNot detected. dPerformed until 4a was consumed using 2.25 mmol of 4a (7 h, 23.2 F mol−1).

Although 5a was obtained in good yield with Pt/C, the current efficiency was low. We assumed that this was due to the recombination of Had to form hydrogen. To suppress hydrogen generation, we varied the flow rate of the reaction solution (Table 4). The reaction time was set to 2.5 h (2.25 mmol, 8.3 F mol–1) and the flow rate of the reaction solution was changed from 0.25 to 1.0 mL min−1. As expected, increasing the flow rate increased the current efficiency and yield of 5a. With 0.75 mL min−1 of flow rate, 5a was obtained in 88% yield with 62% of current efficiency (Table 4, entry 3). In contrast to the previous report on electrocatalytic hydrogenation of nitrobenzene using a PEM reactor , cyclohexylamine was not observed in each reaction.

Table 4: Effect of flow rate in the electrochemical reduction of 4aa.

entry flow rate (mL min-1) 5a (%)b current efficiency (%)b recovered 4a (%)b 1 0.25 75 53 14 2 0.50 78 56 7 3 0.75 88 62 3 4 1.0 88 63 2

aReaction conditions: anode catalyst, Pt/C; cathode catalyst, Pt/C; 4a, 2.25 mmol; solvent, CH2Cl2 (0.5 M); flow rate of the solution of 4a, 0.25–1.0 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2. The solution was circulated for 2.5 h (8.3 F mol−1). bDetermined by GC analysis using n-dodecane as an internal standard.

Next, the scope of the nitroarene electro-reduction was explored (Scheme 3). To obtain products in high yields, the electrolysis was performed until the substrates were consumed. First, nitroarenes bearing electron-donating groups were investigated. Nitroarenes 4b–d bearing methyl groups gave the corresponding anilines 5b–d in 70–76% yield. p-Methoxyaniline (5e) was obtained in 83% yield. Nitroarenes bearing electron-withdrawing groups are also useful. Acetyl and cyano groups were tolerated under the reaction conditions and anilines 5f and 5g were obtained in 85% and 81% yields, respectively. 1-Naphthylamine, a more π-extended aniline was easily obtained in a high yield.

Scheme 3: Scope of the electrochemical reduction of nitroarenes. Reaction conditions: anode catalyst, Pt/C; cathode catalyst, Pt/C; 4, 2.25 mmol; solvent, CH2Cl2 (0.5 M); flow rate of the solution of 4, 0.75 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2. The solution was circulated until the full conversion of 4. Isolated yield.

Reduction of quinolines to tetrahydroquinolines

The electrochemical reduction of quinolines was performed using a PEM reactor. First, several different cathode catalyst were examined for the reduction of quinoline (6a) (Table 5). Because 4.0 F mol−1 of electricity should be required ideally to reduce quinoline (6a) to 1,2,3,4-tetrahydroquinoline (7a), 4.0 F mol−1 of electricity was applied for the reactions. Pd/C, Ir/C, Ru/C, and Pt/C were used as cathode catalysts, and 3–5% yields of 7a were obtained by the use of each catalyst (Table 5, entries 1–4). We chose Pt/C, one of the most common catalysts used in fuel-cell reactors, and increased the charge to complete the reaction. With 50 F mol−1 of electricity, 6a was completely consumed and 7a was obtained in 96% yield (Table 5, entry 5). However, the electrochemical reaction of 6a with reused MEA did not give 7a, and MEA was torn after electrolysis, probably because 6a and/or 7a were trapped on the membrane having sulfonic acids, and the desired electrolysis was disturbed. We assumed that this could be resolved by adding a strong acid to liberate 6a and/or 7a from the membrane by the equilibrium (Figure 2).

Table 5: Electrochemical reduction of 6a with several cathode catalystsa.

entry cathode catalyst 7a (%)b recovered 6a (%)b 1 Pd/C 5 86 2 Ir/C 3 86 3 Ru/C 4 70 4 Pt/C 3 86 5c Pt/C 96 N.D.d 6e Pt/C N.D. N.D.

aReaction conditions: anode catalyst, Pt/C; 6a, 1.5 mmol; solvent, CH2Cl2 (0.5 M); flow rate of the solution of 6a, 0.75 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2. The solution was circulated until the passage of 4.0 F mol−1 (48 min). bDetermined by 1H NMR spectroscopy using 1,1,2,2-tetrachloroethane as an internal standard. c50 F mol−1 (1st run). dNot detected. e50 F mol−1 (2nd run).

Figure 2: Hypothesis of the trap of quinoline on membrane and tetrahydroquinoline and the effect of adding an acid.

Based on this hypothesis, we examined several acids and found that the addition of a catalytic amount (0.10 equiv) of p-toluenesulfonic acid (PTSA) was sufficient (Figure 3). The first run gave 7a in 88% yield and the second run with the MEA gave 7a in 85% yield. MEA was reused eight times, and 7a was obtained in high yield in each run. The addition of pyridinium p-toluenesulfonate (PPTS) was also efficient, and MEA was repeatedly used to afford 7a in high yield (Figure 4).

Figure 3: Recycled use of MEA for the electroreduction of 6a in the presence of PTSA (0.10 equiv). Reaction conditions: anode catalyst, Pt/C; cathode catalyst, Pt/C; 6a, 1.5 mmol; solvent, CH2Cl2 (0.5 M); flow rate of the solution of 6a, 0.75 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2. The solution was circulated until the passage of 50 F mol−1 (10 h). The yields were determined by 1H NMR analysis using 1,1,2,2-tetrachloroethane as an internal standard.

Figure 4: Recycled use of MEA for the electroreduction of 6a in the presence of PPTS (0.10 equiv). Reaction conditions: anode catalyst, Pt/C; cathode catalyst, Pt/C; 6a, 1.5 mmol; solvent, CH2Cl2 (0.5 M); flow rate of the solution of 6a, 0.75 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 50 mA cm−2 (10 h). The solution was circulated until the passage of 50 F mol−1. The yields were determined by 1H NMR analysis using 1,1,2,2-tetrachloroethane as an internal standard.

After further tunings, we found that the charge for electro-reduction of 6a could be reduced to 25 F mol−1 when decreasing the current density to 25 mA cm−2, and 7a was obtained in 90% yield (see the Supporting Information File 1). Next, the substrate scope was examined under optimal conditions (Scheme 4). Substrates bearing a methyl group afforded the corresponding products in high yields (7a–e, 7g, and 7h), except for 6f which contained a methyl group at the 4-position. The reaction of 2,3-dimethylquinoline (6i) gave the desired product 7i (cis/trans = 3:1) in 77% yield. Substrates bearing acetyl (6k), ester (6l), and amide (6m) groups were tolerated under these conditions and selectively afforded the desired products. A substrate with a methoxy group at the 8-position (6o) afforded the desired product 7o in high yield. However, 6p and 6q with a methoxy group at the 6- and 3-position gave only a small amount of the target product, and 6r with a methoxy group at the 4-position gave compound 7r′, in which the benzene ring was hydrogenated. Substrates with chloro groups produced the dechlorinated products (7t and 7u). Unfortunately, during the electrolysis of quinolines with cyano (6v), formyl (6w and 6x), nitro (6y), and amino (6z) groups, the flow path was clogged probably due to the decomposition of the substrates, and the desired products were not obtained.

Scheme 4: Scope of the electroreduction of 6 in the presence of PTSA (0.10 equiv). Reaction conditions: anode catalyst, Pt/C; cathode catalyst, Pt/C; 6, 1.5 mmol; solvent, CH2Cl2 (0.5 M); flow rate of the solution of 6, 0.75 mL min−1; flow rate of H2 gas, 100 mL min−1; reaction temperature, room temperature; current density, 25 mA cm−2. The solution was circulated until the passage of 25 F mol−1 (10 h). Isolated yield. aDetermined by 1H NMR analysis using 1,1,2,2-tetrachloroethane as an internal standard. b0.125 M. c0.25 M. d1,4-dioxane/H2O (7:1, 0.25 M).

This system can be applied to large-scale syntheses. A similar yield of 7a was obtained when the reaction was scaled up (Scheme 5a). The electrolysis of 2.32 g of 6a gave 2.12 g of 7a (88% yield). Next, we examined the electroreduction of 6a using an aqueous proton source instead of hydrogen. The use of DSE® as an anode and H2SO4 aq as an anolyte was effective, and 7a was obtained in 80% yield (Scheme 5b).

Scheme 5: a) Large scale synthesis of 7a and b) electoreduction of 6a using H2SO4 as a proton source.

Reduction of pyridines to piperidines

As mentioned previously, reduction of pyridines to piperidines is important for organic synthesis. Therefore, the electroreduction of pyridine (8a) was performed (Table 6). First, the electroreduction of 8a was performed with 0.1 equiv of PTSA, and 1.0 equiv of PTSA was added after electrolysis to determine the yield by 1H NMR analysis (Table 6, entry 1). The yield of 9a·PTSA (26% yield) was low and 8a·PTSA was obtained as the major product (60% yield), suggesting that the catalytic amount of PTSA was not sufficient because it would be completely trapped with 9a. These results suggest that stoichiometric amount of PTSA was required to liberate 8a from the membrane. As expected, the yield of 9a·PTSA increased upon increasing the amount of PTSA used for electrolysis (Table 6, entries 1–4). Electrolysis with 1 equiv of PTSA afforded 9a·PTSA quantitatively (Table 6, entry 4). MEA was used repeatedly, and the target compound was obtained quantitatively in each run (Table 6, entries 5 and 6). Finally, the reaction was examined using an aqueous proton source instead of humidified H2 gas. Similar to the reaction of quinoline (6a), 9a·PTSA was obtained in 85% yield (Table 6, entry 7) by the use of DSE® electrode and 2 M H2SO4 aq as a proton source.

Table 6: Electroreduction of pyridine (8a) using a PEM reactora.

entry x (equiv) y (equiv) 9a·PTSA (%)b 8a·PTSA (%)b 1 0.1 1 26 60 2 0.5 0.5 48 47 3 0.75 0.25 77 23 4 1 0 quant N.D.c

View original article

BEILSTEIN JOURNAL OF ORGANIC CHEMISTRY

分享书签

0 0 0 0 0 0 0

More from this channel

Electrocatalytic hydrogenation of cyanoarenes, nitroarenes, quinolines, and pyridines under mild conditions with a proton-exchange membrane reactor

留言 (0)