A Pyrene-Anchored Nickel N-Heterocyclic Carbene–Isoquinoline Complex Promotes CO₂ Reduction

Abstract

In this study, on the basis of a previous report, a pyrene-anchored nickel complex was designed and synthesized via five steps. The NMR spectra of the synthesized complex were found to exhibit significant proton and carbon chemical shift anisotropy. Cyclic voltammetry spectra showed that the introduction of pyrene slightly influenced the onset potential of CO₂ reduction. Lastly, controlled-potential electrolysis experiments disclosed that a pyrene-anchored nickel carbene–isoquinoline (Ni−2) complex selectively converted CO₂ into CH₄ with a TON value of 2.3 h⁻¹.

1. Introduction

The design of molecule catalysts for converting carbon dioxide into fuel has been a hot topic in industrial and academic research [1,2]. In the field of inorganic chemistry, extensive laboratory work has been conducted in the search for catalysts which are effective in converting carbon dioxide into storable molecules such as formic acid [3,4], oxalic acid [5], CO [6], and methanol [7].

In 2013, groups led by Thoi and Chang [8,9] developed biscarbene complexes (Scheme 1, Ni-1). Among these, isoquinine-derived carbene complexes have been shown to be highly effective, with low initial reductive potential and the ability to high-selectively convert carbon dioxide into carbon monoxide.

In addition, immobilization of homogenous catalysts on the surfaces of materials is an effective strategy for improving utilization efficiency as well as reducing costs. Carbon-based materials such as nanotube/graphene [10] are frequently used as heterocyclic materials. Anchoring pyrene onto a homogenous catalyst is a widely used strategy because of the π-π interaction between the pyrene and the graphene or nanotube. In addition to facilitating immobilization, pyrene can also indirectly alter the reductive product [11]. Furthermore, the catalytic activities of pyrene-extended complexes can be tuned by the interaction between the pyrene fragment and the metal core center [10].

In this study, a single pyrene-anchored nickel N-heterocyclic carbene (NHC)–isoquinoline complex (Ni−2) was designed and synthesized with the aim of achieving immobilization on a carbon nanotube, considering potential π-π interaction with a pyrene fragment [12,13].

2. Results

2.1. Synthesis

The synthesis pathway for pyrene-anchored nickel N-heterocyclic carbene (NHC)– isoquinoline (Ni−2) is shown in Scheme 2. Firstly, a pyrene-containing fragment was synthesized and forged in three steps, with a total resulting yield of 21%. Pyrene-1-carbaldehyde was used as the starting material. Subsequently, Knoevenagel condensation (2−1, 90%), LiCl-assisted NaBH₄ reduction (2−2, 73%), and, finally, bromidation (2−3, 32%) were carried out, producing 2-pyrene-1,3-dibromopropane. Imidazolium salt was formed smoothly according to a previously reported method [8] by reaction between 2−3 and 2.0 equivalent 3-(1H-imidazol-1-yl)isoquinoline [9] in THF solution, with easily isolable precipitation producing imidazolium salts in the form of a slightly yellow solid (2−4, 50%). In the final step, several procedures were carried out. Initially, we attempted a previously reported analogous complex synthesis routine, involving deprotonation by Ag₂O in acetonitrile solution under room-temperature conditions. However, this was not feasible due to the low solubility of imidazolium salts in acetonitrile. Finally, deprotonation was successfully achieved in hot DMSO solution (100 °C). With an extension of deprotonation time to 12 h, the formed silver carbene intermediate was able to transmetalate with Ni(DME)Cl₂ and finally deliver complex red-brown Ni−2 (for other tested conditions, please see Supporting Information for details). Furthermore, for the purpose of catalytic performance comparison, the nickel complex Ni−1 was also synthesized according to a method previously reported in the literature [9].

2.2. NMR Characterization

The formation of Ni−2 complexes was supported by spectroscopic data (¹H NMR and ¹³C NMR. In contrast to pseudo-C₂ symmetric Ni−1, an asymmetric pyrene fragment was introduced into complex Ni−2. Symmetric protons or carbons in precursor imidazolium salts presented an anisotropic chemical shift in complex Ni−2, and this was induced not only by the asymmetric anti-shielding or shielding field caused by the pyrenyl fragment. Further details of these findings may be stated as follows:

In the case of proton NMR, the disappearance of typical imidazolium C2 protons (δ = 9.94 ppm) (Figure 1A) meant that imidazolium salts were deprotonated, a finding similar to results previously reported for a metal N-heterocyclic carbene complex [2,14,15,16]. The C2 symmetry was broken as a result of free rotation being prevented after coordination [17,18]. Two significant pieces of evidence were noted. First, the gem-methylene protons on the flexible propanyl link of Ni−2 exhibited magnetic anisotropy [19,20,21] (Figure 1B); second, the isoquinoline standard proton exhibited two independent peaks (9.69 ppm, 9.59 ppm).

In ¹³C NMR spectra, the aromatic ring signals of two chemical equal isoquinoline–imidazolium fragments in ligand 2−4 were not chemically equal after coordination. Each isoquinoline–carbene fragment exhibited an independent carbon signal. In total, 40 aromatic carbon signals were recorded for nickel complex Ni−2 (Figure 2A), consistent with ¹H NMR spectra. Furthermore, the corresponding symmetric alkyl carbon in ligand 2−4 presented a different carbon NMR signal in the coordinative complex ¹³C NMR (Figure 2B). Specially, two independent nickel–carbene chemical shift signals were observed at 159.34 and 158.75 ppm, corresponding with a position about 22 ppm downfield from the parent imidazolium salts (136.74 ppm) [22].

2.3. Evaluation of Ni−2 for Electrocatalytic CO₂ Reduction

With Ni-1 and the newly synthesized pyrenyl containing nickel complex (Ni−2) in hand, the carbon dioxide reduction performances of the two nickel complexes were evaluated (Figure 3). For comparison purposes, the differences in catalytic performance between Ni−1 and the pyrene-anchored complex Ni−2 were considered. Cyclic voltammetry was used to investigate their electrochemical behavior under an argon or carbon-dioxide atmosphere.

For comparison purposes, the electrochemical CO₂ reduction performances of the nickel complex (Ni−1) and the pyrene-anchored nickel complex (Ni−2) were investigated under conditions of Ar atmosphere and CO₂-saturated acetonitrile solution. In the case of Ni−1, two reversible reduction peaks at E_1/2 = −0.92 and −1.37 V vs. SCE were observed, highly consistent with the results reported by Thoi et al. [9]. In comparison, the pyrene-anchored Ni−2 exhibited irreversible reductions at −0.82 and −1.53 V. Considering that the pyrenyl fragment exhibited reductive potential at −2.4 ~ −2.6 V vs. SCE [23,24], it was deduced that the introduction of pyrene resulted in a shift in Ni^II/Ni^I redox in a positive direction (+100 mV) and a shift in Ni^I/Ni⁰ redox in a negative direction (−160 mV). Next, we replaced the Ar atmosphere with CO₂, which was allowed to bubble for 30 min. Investigation of subsequent electrochemical behavior in the CO₂ atmosphere revealed the following: Ni−1 exhibited enhanced current at E_onset = −1.20 V vs. SCE, similar to findings reported in the literature [9], while Ni−2 showed enhanced current (E_onset) at −1.25 V, i.e., 50 mV lower than Ni−1. Both of these results were indicative of electrocatalysis.

With these results in hand, controlled-potential electrolysis (CPE) experiments were performed in which a corresponding onset potential was applied. Electrolysis was applied in the DMF/H₂O mixture solution (100/1 v/v) to test its selectivity for CO₂ over proton reduction. Based on the gas chromatography analysis, we observed that both of the two complexes had good selectivity for reducing CO₂ over proton reduction with a C1 reductive product, a result reasonably consistent with the literature [8,9]. Profiles of CO₂-reducing products are listed in Figure 4. Complex Ni−1 selectively catalytically generated CO product, with a turnover number (TON) of 13 being recorded. In contrast, the pyrenyl-modified complex, Ni−2, detected only CH₄-reducing product in its capped gas, with a TON value of 4.6. To our knowledge, such high selectivity in CO₂ conversion to methane has rarely been reported for a homogenous nickel catalyst [25,26,27]. The difference in the reduction products may have been because the additional pyrene fragment accelerated electron transfer from nickel center to carbon dioxide and promoted the eight-electron-reduced product [10]. However, further experiments are necessary to obtain a more detailed understanding of the mechanism involved. Furthermore, we originally sought to investigate immobilization of nickel complex (Ni−2) on the nanotube; however, after several attempts, we failed to demonstrate Ni−2 immobilization on the nanotube.

3. Materials and Methods

Unless otherwise noted, all chemicals and materials were purchased from commercial suppliers and used without further purification. All the solvents were treated according to generally used methods. All ¹H NMR, ¹³C NMR, ¹⁹F NMR, and ³¹P NMR spectra were recorded on a 400 MHz Bruker AVANCE IIIHD spectrometer (400/101/376/162 MHz, respectively). Chemical shifts for protons were reported in parts per million downfield from tetramethylsilane and referenced to residual in the NMR solvent. Chemical shifts for carbon were reported in parts per million downfield from tetramethylsilane and referenced to the carbon resonances of the solvent (CDCl₃ = δ: 77.0; DMSO = δ: 39.5). The peak patterns were indicated as follows: s, singlet; d, doublet; t, triplet; m, multiplet; q, quartet; br, broad. The coupling constant J was reported in Hertz (Hz). High-resolution mass spectra were collected using Bruker MicroTOF II equipment. Products were purified by flash chromatography on 200–300 mesh silica gels (SiO₂). The cyclic voltammetry curve was produced on a CHI660E instrument. Copies of all NMR spectra are available in Supplementary Information (SI).

Synthesis of diethyl 2-(pyren-1-ylmethylene)malonate (2−1) was achieved as follows: Into a 100 mL round bottle, pyrene-1-aldehyde (2.3 g, 10 mmol), pyridine (0.1 mL), and toluene (30 mL) were placed. Next, 4A molecular sieves (100 mg) were added. The mixture was then stirred and refluxed for 24 h. After reaction had finished, the versatile solvent was removed and subjected to silica gel chromatography, which afforded the desired product (2−1) as yellow sticky oil (3.35 g, yield 90% The following data were obtained, consistent with the literature [28,29]: ¹H NMR (400 MHz, CDCl₃) δ 8.79 (s, 1H), 8.27 (d, J = 9.2 Hz, 1H), 8.21 (d, J = 7.5 Hz, 2H), 8.16–8.11 (m, 1H), 8.09 (t, J = 6.1 Hz, 3H), 8.03 (dd, J = 12.7, 6.5 Hz, 2H), 4.44 (q, J = 7.1 Hz, 2H), 4.23 (q, J = 7.1 Hz, 2H), 1.53–1.25 (m, 3H), 1.11 (t, J = 7.1 Hz, 3H); ¹³C NMR (101 MHz, CDCl₃) δ 166.56, 164.19, 141.27, 132.61, 131.12, 130.62, 129.72, 128.66, 128.58, 127.35, 127.19, 126.28, 126.05, 125.97, 125.75, 124.57, 124.36, 123.02, 61.79, 61.56, 14.25, 13.80.

Synthesis of 2-(pyren-1-ylmethyl)propane-1,3-diol (2−2) was achieved as follows. Into a 250 mL round bottle, diethyl 2-(pyren-1-ylmethylene)malonate (3.35 g, 7 mmol), anhydrous THF (37 mL), and ethanol (55 mL) were placed. The mixture was stirred and then cooled to 0 °C. Next, NaBH₄ (35 mmol + 35 mmol) was added in two portions. LiCl (2.0 mmol) was then added slowly, and the reaction solution stirred at room temperature. The reaction process was monitored by TLC. After 3 hours, the mixture was quenched by saturated ammonium chloride (100 mL) and extracted by ethyl acetate (20 mL × 5). The organic phase was combined and concentrated under reduced pressure, then subjected to silica gel chromatography (DCM/MeOH = 100/1) which afforded the target 2-(pyren-1-ylmethyl)propane-1,3-diol (2−2) as slightly yellow oil (1.5 g, 73%). The following data were obtained, consistent with the literature [30]: M.P. 123–125 °C; ¹H NMR (400 MHz, DMSO) δ 8.41 (t, J = 13.0 Hz, 1H), 8.29–8.23 (m, 2H), 8.20 (d, J = 7.9 Hz, 2H), 8.15–8.08 (m, 2H), 8.04 (t, J = 7.5 Hz, 1H), 7.93 (d, J = 7.7 Hz, 1H), 4.59 (t, J = 4.7 Hz, 2H), 3.51 (dt, J = 10.4, 5.2 Hz, 2H), 3.47–3.39 (m, 2H), 3.30 (d, J = 6.9 Hz, 2H), 1.98 (t, J = 5.6 Hz, 1H); ¹³C NMR (101 MHz, DMSO) δ 136.33, 131.31, 130.83, 129.67, 129.00, 128.87, 127.89, 127.48, 126.87, 126.54, 125.32, 125.19, 125.03, 124.71, 124.55, 124.23, 61.66, 46.55, 31.71.

Synthesis of (1-(3-bromo-2-(bromomethyl)propyl)pyrene (2−3) was achieved as follows: In a 100 mL three-necked round bottle, compound 2-(pyren-1-ylmethyl)propane-1,3-diol (1.0 g, 3.44 mmol), PPh₃ (2.0 g, 7.56 mmol), and dichloromethane (50 mL) were layered. The mixture was cooled under ice bath conditions (−5 °C). Then, N-Bromosuccinimide (1.4 g, 7.56 mmol) was added in portions under nitrogen flow. The reaction process was monitored by TLC. After 3 h, the solution was concentrated and subjected to silica gel chromatography (PE/EA = 20/1) which afforded the target 2−3 as yellow solid (700 mg, 32%). The following data were obtained: M.P. 93–96 °C; ¹H NMR (400 MHz, CDCl₃) δ 8.37–7.89 (m, 10H), 3.66 (dd, J = 10.3, 4.3 Hz, 2H), 3.54 (t, J = 7.2 Hz, 4H), 2.68–2.50 (m, 1H). ¹³C NMR (101 MHz, CDCl₃) δ 132.19, 131.34, 130.74, 130.49, 128.93, 128.04, 127.88, 127.40, 127.16, 126.06, 125.28, 125.18, 125.09, 124.82, 122.86, 43.55, 36.32, 34.86. ESI-HRMS ([M]⁺) m/z calcd. for C₂₀H₁₇Br₂ 416.9671 (100.0%) gave a value of 416.9675 (100.0%).

Synthesis of 3,3′-(2-(pyren-1-ylmethyl)propane-1,3-diyl)bis(1-(isoquinolin-3-yl)-1H-imidazol-3-ium) bromide (^Pyrene-Prbimiq, 2−4) was carried out as follows: Into a 50 mL reaction tube, 2−3 (210 mg, 0.5 mmol), 3-(1H-imidazol-1-yl)isoquinoline (205 mg, 1.05 mmol), and dry THF (15 mL) were placed. The mixture was bubbled through nitrogen flow and then sealed. The tube was maintained at 120 °C in oil bath for 48 h. Precipitation was then collected through filtration, washed by ice-cooled THF (3 mL × 3), and dried under vacuum, which afforded the target compound 2−4 as yellow solid (280 mg, yield 70%). The following data were obtained: M.P. 155–157 °C; ¹H NMR (400 MHz, DMSO) δ 9.94 (s, 2H, -NCHN-), 9.20 (s, 2H), 8.37 (d, J = 9.2 Hz, 1H), 8.23 (d, J = 9.0 Hz, 4H), 8.13 (d, J = 9.2 Hz, 1H), 8.09 (d, J = 7.8 Hz, 1H), 8.03 (d, J = 11.4 Hz, 5H), 7.99 (d, J = 7.6 Hz, 1H), 7.95–7.89 (m, 4H), 7.88 (d, J = 6.1 Hz, 1H), 7.85 (s, 1H), 7.83 (s, 1H), 7.80 (d, J = 7.2 Hz, 1H), 7.74–7.66 (m, 2H), 4.69 (d, J = 5.8 Hz, 4H), 3.75 (dd, J = 17.9, 11.3 Hz, 1H), 3.67 (d, J = 6.8 Hz, 2H); ¹³C NMR (101 MHz, DMSO) δ 152.88, 141.34, 136.74(-NCHN-), 135.58, 132.85, 132.31, 130.80, 130.31, 129.88, 129.14, 128.65, 128.53, 128.38, 128.13, 127.37, 127.01, 126.17, 125.39, 125.14, 125.03, 124.43, 124.08, 123.91, 123.52, 119.05, 109.56, 52.02, 40.49, 33.85. Anal. calcd for C₄₄H₃₄Br₂N₆, C 65.52; H, 4.25; N, 10.42 found C, 65.45; H, 4.17; N, 10.47 [22].

Synthesis of [Ni(^Pyrene-Prbimiq-2HBr)](PF₆)₂ (Ni−2) was carried out as follows: Into a flame-dried Schlenk tube, ^Pyrene-Prbimiq (160 mg, 0.2 mmol), silver oxide (62 mg, 0.22 mmol), and anhydrous DMSO (10 mL) were placed. The tube was then evacuated and refilled with nitrogen three times. The resulting solution was stirred for 8 h at 100 °C. Ni(DME)Cl₂ (48 mg, 0.22 mmol) was then added under nitrogen flow, and the mixture was stirred at 100 °C for an additional 8 h. Next, NH₄PF₆ (163 mg, 1 mmol) was added and stirring carried out for a further 2 h at room temperature. The DMSO solvent was then removed under vacuum. The residual was washed by ionic water (3 × 5 mL) to remove inorganic salts. The left compound was dried under vacuum overnight (80 °C, 10 pa, 16 h), and further purified through crystallization (acetone/ethyl ether), which afforded the desired Ni(^Pyrene-Prbimiq-HBr)₂](PF₆)₂ (Ni−2) as red solid (130 mg, 65%). The following data were obtained: M.P. > 300 °C (decomposed); ¹H NMR (400 MHz, DMSO) δ 9.69 (s, 1H), 9.59 (s, 1H), 8.73 (s, 1H), 8.61 (s, 2H), 8.54 (t, J = 8.6 Hz, 2H), 8.43 (d, J = 13.0 Hz, 2H), 8.34 (dt, J = 9.0, 6.8 Hz, 4H), 8.24–8.18 (m, 3H), 8.12 (t, J = 7.5 Hz, 4H), 8.05 (t, J = 7.4 Hz, 1H), 7.93 (d, J = 7.8 Hz, 1H), 7.92–7.87 (m, 1H), 7.86–7.81 (m, 1H), 7.64 (s, 1H), 4.62 (d, J = 15.1 Hz, 1H), 4.55 (d, J = 13.2 Hz, 1H), 4.16 (d, J = 10.6 Hz, 1H), 3.98 (dd, J = 13.7 and 5.3 Hz, 1H), 3.94 (d, J =15.6 Hz, 1H), 3.21 (t, J = 12.1 Hz, 1H), 2.93 (m, 1H); ¹³C NMR (101 MHz, DMSO) δ 159.34 (Ni-C), 158.75 (Ni-C), 156.54, 154.64, 144.94, 144.75, 137.69, 137.45, 135.15, 134.56, 133.19, 131.36, 130.91, 130.45, 130.19, 129.21, 129.12, 128.92, 128.75, 128.10, 127.95, 127.90, 127.84, 127.76, 127.44, 127.25, 126.81, 126.34, 125.73, 125.55, 125.52, 125.04, 124.64, 124.06, 118.51, 117.17, 107.48, 107.09, 50.28, 42.00, 33.20, 31.14; ³¹P NMR (162 MHz, DMSO) δ −144.17 (hepta, J = 711.0 Hz); ¹⁹F NMR (376 MHz, DMSO) δ −70.13 (d, J = 711.0 Hz); HR-MS (ESI, positive ion trap) m/z calcd. for C₄₄H₃₂N₆Ni²⁺ 351.1015 (100.0%), 351.6032 (47.6%), 352.0993 (38.5%), 352.6009 (18.3%) found 351.1019 (100.0%), 351.6020 (47.6%), 352.0992 (38.5%), 352.5993 (18.3%); anal. calcd for C₄₄H₃₂F₁₂N₆NiP₂, C, 53.20; H, 3.25; N, 8.46 found C, 53.15; H, 3.18; N, 8.62.

4. Conclusions

In this study, a pyrene-anchored nickel N-heterocyclic carbene–isoquinoline complex was synthesized via five steps. Electrochemical studies showed that the introduction of pyrene altered the CO₂ reduction product from carbon monoxide to methane with a TON value of 4.6.