Electrochemical reduction of carbon dioxide with a molecular polypyridyl nickel complex

Abstract

The synthesis and reactivity of a molecular nickel(II) complex 1 with the polypyridyl ligand framework N,N,N′,N′-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN) under electrochemically reducing conditions in the presence of CO₂ is reported. Cyclic voltammetry (CV), infrared spectroelectrochemistry (IR-SEC) and electrolysis experiments suggest this Ni complex is competent at mediating the two-electron reduction of CO₂ to CO and H₂O with phenol as an added proton donor, but is subsequently prone to rapid degradation as Ni(CO)₄. This deleterious pathway was shown to be mitigated by the inclusion of the CO scavenger [Ni(TMC)]²⁺, although this requires stoichiometric inclusion for each catalyst turnover and does not significantly improve the observed catalytic current densities.

---

Introduction

The continuing increase of the world's population poses significant challenges for both sourcing new fuels and controlling global emissions due to the corresponding increase in energy demand. The current primary source of energy, fossil fuel, is nonrenewable and its large-scale use to meet current global energy demands has led to an increase in anthropogenic CO₂ emissions (current atmospheric levels >400 ppm) with related negative environmental impacts such as loss of polar ice and increasingly energetic weather patterns.^1,2 Based on this steadily increasing demand and a global desire to diminish its negative effects, an interest towards utilizing carbon dioxide (CO₂) as a precursor towards liquid fuels (e.g. methanol or formate/formic acid) or fuel precursors continues to generate interest.^3–5 However, the thermodynamic stability of CO₂ poses a challenge towards its reduction.⁵ Although many catalysts which efficiently reduce CO₂ exist, some are based on non-abundant precious metals.^6–14 As a result, there continues to be profound interest in the development of molecular electrocatalysts with earth-abundant transition metals for the electrocatalytic reduction of CO₂.^3,15–27 The use of earth-abundant transition metal electrocatalysts offers a more cost-effective alternative for energy research.^15–27

Herein, the electrochemical behavior of a Ni-based complex with a polypyridyl framework, N,N,N′,N′-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN), with respect to added CO₂ and phenol (PhOH) is reported. Polypyridyl catalysts have previously been employed as a part of electrocatalytic platforms for hydrogen evolution, CO₂ reduction, and thermal activation of oxygen.^{15,22,23,28–40} Polydentate ligand platforms with flexible linkers can accommodate redox-induced structural changes readily, resulting in complexes which are stable across a wide range of potentials.^{10,18,41–47} Although a hexadentate ligand, TPEN has been reported to have variable coordination numbers with first-row transition metals, suggesting that substrate interactions can still occur.⁴⁸ Indeed, cyclic voltammetry and infrared spectroelectrochemistry establish that [Ni(TPEN)](PF₆)₂ reacts quite readily with CO₂ upon a one-electron reduction to a Ni(I) state, generating CO. Prolonged electrolysis results in substantial formation of the degradation product Ni(CO)₄; this deleterious pathway can be circumvented through the use of a CO scavenger, [Ni(1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane)](PF₆)₂ (1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane = TMC), prolonging the lifetime of [Ni(TPEN)](PF₆)₂.⁴⁹

Materials and methods

General

All chemicals and solvents (ACS or HPLC grade) were commercially available and used as received unless otherwise indicated. For all air-sensitive reactions and electrochemical experiments, solvents were obtained as anhydrous and air-free from a PPT Glass Contour Solvent Purification System. Gas cylinders were obtained from Praxair (Ar as 5.0; CO₂ as 4.0) and passed through molecular sieves prior to use. Gas mixing for variable concentration experiments was accomplished using a gas proportioning rotameter from Omega Engineering; concentration values were determined according to Henry's law using a saturation concentration of 0.28 M for CO₂ in MeCN.⁵⁰ NMR spectra were obtained on either a Varian 600 MHz or 500 MHz instrument and referenced to the residual solvent signal. IR absorbance spectra were obtained on a Vertex V80 IR instrument from Bruker and UV-vis absorbance spectra on a Cary 60 from Agilent. GC experiments were performed using an Agilent 7890B Gas Chromatograph with an Agilent J&W Select Permanent Gases/CO₂ column; eluent retention times and product characterization were determined by standard injections. HRMS data were obtained by the Mass Spectrometry Lab at the University of Illinois at Urbana-Champaign and elemental analyses were performed by Midwest Microlab.

Synthetic procedures

N,N,N′,N′-Tetrakis(2-pyridylmethyl)ethylenediamine (TPEN). The ligand framework, TPEN, was prepared according to a modified literature procedure.⁵¹ A round-bottom flask equipped with a stir bar was charged with 2-picolyl chloride hydrochloride (15.99 g, 0.098 mol) and ethylenediamine (1.6 mL, 1.44 g, 0.0239 mol). The resulting solution was heated to 70 °C and NaOH (aq) (10 M, 85 mL) was added dropwise over the course of an hour. During this addition, the solution turned a deep red color. The solution was allowed to cool to room temperature and extracted with dichloromethane (3 × 25 mL). The organic layers were combined, dried over MgSO₄, and filtered before the solvent was removed under reduced pressure, leaving a red oil. An aliquot of hexanes (20 mL) was added to the oil and the resulting suspension was stored overnight at 0 °C. After this time, a red brown precipitate was collected via vacuum filtration; 8.38 g isolated (20% yield). ¹H NMR (600 MHz; CD₃CN; Me₄Si), δ 8.43 (4H, d, ArH), δ 7.62 (4H, t, ArH), δ 7.44 (4H, d, ArH), δ 7.16 (4H, t, ArH), δ 3.70 (8H, s, –CH₂–), δ 2.68 (4H, s, –CH₂–); Fig. S1–S3.† CHN Analysis; found: C, 71.20; H, 6.46; N, 19.24. Calc. for C₂₆H₂₈N₆: C, 73.56; H, 6.65; N, 19.80; ESI-MS (m/z) calc'd (M + H⁺)⁺: 424.2454 found: 425.2451.

[Ni(TPEN)](PF₆)₂1 (Fig. 1). A round-bottom flask equipped with a stir bar was charged with nickel(II) chloride hexahydrate (0.123 g, 0.473 mmol), TPEN (0.203 g, 0.478 mmol), and ethanol (50 mL). The resulting mixture was stirred for 30 min at room temperature. A portion of ammonium hexafluorophosphate (0.309 g, 1.9 mmol) was then added to the mixture with continued stirring for an additional 30 minutes, during which time a precipitate formed from the suspension. The precipitate was collected via vacuum filtration as a pale pink solid; 0.357 g isolated (97%). ¹H NMR was paramagnetic in CD₃CN (Fig. S4†). CHN Analysis; found C, 40.08; H, 3.73; N, 10.39. Calc. for C₂₆H₂₈N₆: C, 40.39; H, 3.65; N, 10.87. ESI-MS (m/z) calc'd (M − 2[PF₆])²⁺: 241.0865 found: 241.0868 (Fig. S5†).

Fig. 1 Molecular structures of the molecular Ni species relevant to this report. Two hexafluorophosphate (PF₆⁻) counteranions (not pictured) are present for both species [Ni(TPEN)](PF₆)₂1 and [Ni(TMC)](PF6)₂2.

Synthesis of 1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane (tetramethylcyclam; TMC). TMC was synthesized according to a modified literature procedure.^52,53 A round bottom flask equipped with a stir bar was charged with a solution of 1,4,8,11-tetraazacylcotetradecane (1.500 g, 7.5 mmol), formaldehyde (6.6 mL, 37% w/w aq solution), formic acid (8.5 mL) and de-ionized water (5 mL) and brought to reflux (130 °C) for 24 hours. While still hot, the reaction mixture was diluted with de-ionized water (15 mL) and cooled in an ice water bath. The reaction mixture was placed in an ice bath and monitored to maintain temperatures below 22 °C while a concentrated solution of sodium hydroxide (11.85 g in 50 mL, 5.93 M) was slowly added dropwise over thirty minutes (the solution pH > 12 after this addition). Next, the solution was extracted with chloroform (5 × 50 mL), the organic fractions combined and dried over magnesium sulfate. The solution was concentrated under reduced pressure and stored overnight at 0 °C; an oily yellow-brown solid was isolated, 0.890 g yield (46%). ¹H NMR (600 MHz; d₂-CD₂Cl₂; Me₄Si), δ 2.38 (16H, m, –CH₂–), δ 2.15 (12H, m, CH₃), and δ 1.57 (4H, m, –CH₂–); Fig. S6.† CHN Analysis; found C, 65.44; H, 12.52; N, 21.84. Calc. for C₁₄H₃₂N₄: C, 65.57; H, 12.58 N; 21.85; ESI-MS (m/z) calc'd (M + H)⁺: 257.270 found: 257.2706 (Fig. S7†).

Synthesis of [Ni(TMC)](PF₆)₂ (Fig. 1). [Ni(TMC)](PF₆)₂ was synthesized according to a literature procedure.^49,54 A warmed (40 °C) solution of nickel(II) chloride hexahydrate (0.249 g, 0.1 M) in ethanol (10 mL) was added to a solution of TMC ligand (0.21 M, 0.269 g) in ethanol (5 mL) in a round-bottom flask equipped with a stir bar. The reaction mixture was stirred at room temperature for thirty minutes. Next, the solvent was removed under reduced pressure and the precipitate was dissolved in de-ionized (DI) water (20 mL). To induce precipitation of the PF₆ salt, 4 equivalents of ammonium hexafluorophosphate were added (0.6489 g, 4 mol) and the solution stirred for an additional thirty minutes. The resulting suspension was filtered to isolate a pink solid which was washed with ether (2 × 5 mL); 0.3869 g (61%). ¹H NMR was paramagnetic in CD₃CN (Fig. S8†). CHN analysis; found C: 27.69 H: 9.36 N: 5.23. Calc. for C₁₄H₃₂N₄: C, 27.79; H, 9.26; N, 5.33. ESI-MS (m/z) calc'd 349.1669 (M − 2[PF₆] + Cl)⁺: found: 349.1666 (Fig. S9†).

Electrochemistry

All electroanalytical experiments were performed using a Metrohm Autolab PGSTAT302N potentiostat or a BioLogic SP-50 potentiostat. Glassy carbon working (3 mm diameter) electrode and non-aqueous silver wire pseudoreference electrode separated by PTFE tip were obtained from CH Instruments. The Ag/AgCl pseudoreference electrode were generated by depositing chloride on the bare silver wire in 10% HCl at oxidizing potentials and were stored in a 0.1 M tetrabutylammonium hexafluorophosphate/acetonitrile solution prior to use. The counter electrode was Pt wire (Alfa Aesar, 99.95%, 0.5 mm diameter) or glassy carbon rod (2 mm diameter), as noted. All cyclic voltammetry (CV) experiments were performed in a modified scintillation vial (20 mL volume) as a single-chamber cell with a cap modified with ports for all electrodes and a sparging needle.

Controlled potential electrolysis experiments were performed in an H-Cell from Pine Research Instrumentation (two ∼25 mL chambers separated by glass frit), with the working electrode in one chamber separate from the pseudoreference and counter electrodes in the other chamber. Spectrographic grade carbon rods (Electron Microscopy Sciences, 6.35 mm diameter) were used as the working and counter electrodes with a silver/silver chloride pseudoreference electrode (CH Instruments) behind a PTFE frit. The cell was sealed with septa and electrical tape to allow head-space sampling and gas sparging via needles through the two septa using a y-splitter. Pressure balance was maintained by inserting a single section of AWG #14 standard wall PTFE tubing through each septum. Tetrabutylammonium hexafluorophosphate was purified by recrystallization from ethanol and dried in a vacuum oven overnight at 100 °C before being stored in a desiccator. All data were referenced to an internal ferrocene standard (ferrocene/ferricenium; Fc/Fc⁺) redox potential under stated conditions unless otherwise specified.

IR spectroelectrochemistry

All IR-SEC experiments were conducted using a custom cell based on a previously published design.^55–57 The three-electrode set-up consists of an inner glassy carbon working electrode disc (10 mm diameter), a central circular silver bare metal pseudoreference electrode, and an outer circular glassy carbon counter electrode embedded within a PEEK block. All data were referenced to an internal ferrocene standard (Fc/Fc⁺ reduction potential under stated conditions); obtained by taking a CV with the cell prior to injecting analyte for IR-SEC experiments unless otherwise specified. All spectra were processed by subtraction of a non-reactive/non-catalytic potential from those at which reactivity occurred. For experiments with CO₂, gas was sparged into the solution containing [Ni(TPEN)](PF₆)₂ for ∼30 s prior to injection into the assembled cell.

Computational methods

DFT calculations were performed on the Rivanna High-Performance Computing Cluster at the University of Virginia using ORCA 4.0.^58,59 Geometry optimizations were performed unrestricted with the B3LYP/G^60–62 functional and def2-TZVP^63–66 basis set with the RIJCOSX approximation,⁶⁷ D3BJ dispersion correction,^68,69 and CPCM⁷⁰ to model the acetonitrile solvent. Numerical frequency calculations at the same level of theory were performed to validate the optimized geometries as minima on the potential energy surface and to generate thermochemical data. Molecular graphics and analyses were performed with the UCSF Chimera package. Chimera is developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIGMS P41-GM103311).⁷¹

Results

Characterization of [Ni(TPEN)](PF₆)₂1

The ¹H NMR obtained for 1 in CD₃CN was consistent with a paramagnetic species (Fig. S4†); Evans method measurements gave an effective magnetic moment of 3.1 Bohr magnetons, consistent with a Ni(II) d⁸ octahedral compound (see ESI†).^72,73 Single crystals suitable for diffraction were obtained through vapor diffusion of CH₂Cl₂ into a solution of 1 in methanol (MeOH). Crystallographic studies revealed the molecular structure shown in Fig. 2, with an overall octahedral geometry defined by axial 2-methyl-pyridyl moieties and the tetrakis-2-methylpyridine-substituted ethylenediamine fragment in the equatorial plane. At the axial positions, Ni–N bond distances of 2.092(3) and 2.085(3) Å are observed. By comparison, the equatorial pyridyl fragments exhibit Ni–N distances of 2.077(3) and 2.085(3) Å. Slightly longer bond distances for the alkyl amine fragments of 2.102(3) and 2.098(3) Å are observed crystallographically. Based on these data, use of TPEN in an abbreviated coordination complex formulation refers to a κ⁶-coordination mode unless otherwise explicitly stated.

Fig. 2 Crystal structure of 1. Hydrogen atoms omitted for clarity; thermal ellipsoids at 50%. C = dark grey, blue = nitrogen, orange = phosphorus, light green = fluorine and dark green = Ni. CCDC 1816890.

Electrochemical characterization

The electrochemical properties of 1 were investigated using cyclic voltammetry (CV) under Ar saturation conditions in an MeCN solution with 0.1 M TBAPF₆ as supporting electrolyte. A reversible feature, assigned to the Ni(II)/Ni(I) reduction, is observed with an E_1/2 of −1.86 V vs. Fc/Fc⁺ and a peak current ratio (i_c/i_a) of 0.93 (Fig. 3 and Table S1†). CVs obtained with variable scan rate are consistent with a diffusion-controlled redox response (Fig. 3 and S10†).

Fig. 3 CVs of 1, obtained under Ar saturation conditions. Conditions: 1 mM analyte; glassy carbon working electrode, Pt wire counter electrode, Ag/AgCl pseudoreference electrode; varied scan rate; referenced to internal ferrocene standard.

Under CO₂ saturation, the Ni(II/I) feature becomes irreversible and a slight increase in current density is observed (Fig. 4). This result is suggestive of a CO₂ binding and activation event, possibly catalytic in nature, with [Ni(TPEN)]⁺ at an applied potential. The addition of phenol (PhOH) as a Brønsted acid source caused an enhancement in the observed current at the Ni(II)/Ni(I) reduction, as well as a shift towards positive potentials (Fig. 5). The slope obtained from a plot of E_1/2 of the catalytic wave (taken at half-peak current density)⁷⁴ against the log of added PhOH concentration (Fig. S12†) shows a slope of 40 mV/log[PhOH], which suggests that this wave has a Nernstian dependence under these conditions. Furthermore, when plotting the log of the peak current density against the log of the concentration (M) of PhOH, a half-order rate dependence is observed (Fig. S11†). Similar kinetic studies suggest that with respect to both CO₂ and catalyst concentration, the reaction represented by the current increase has a first-order dependence (Fig. S13–S16†). Control CVs in the absence of 1 establish the homogeneous nature of the responses discussed above (Fig. S17†).

Fig. 4 CVs of 1, obtained under Ar saturation conditions (black) and CO₂ saturation conditions (red). Conditions: 1 mM analyte; glassy carbon working electrode, glassy carbon counter electrode, Ag/AgCl pseudoreference electrode; scan rate 100 mV s⁻¹; referenced to internal ferrocene standard.

Fig. 5 CVs of 1, obtained under CO₂ saturation conditions with variable PhOH concentration. Conditions: 1 mM analyte; 0.1 M TBAPF₆/MeCN, glassy carbon working electrode, glassy carbon counter electrode, Ag/AgCl pseudoreference electrode; scan rate 100 mV s⁻¹; referenced to internal ferrocene standard.

Foot-of-the-wave analysis (FOWA)

To determine the rate constant of the reaction, foot-of-the-wave analysis (FOWA) was employed to obtain the maximal turnover frequency (TOF_max) and catalytic rate constant (k_cat) for the electrocatalytic reduction of CO₂ by complex 1. FOWA uses the Tafel relationship of the initial part of the catalytic wave, which provides a better approximation than the use of peak current plateaus, where the consumption of substrate or long-lived intermediates can result in deviations from an ideal catalytic response.⁷⁵ Steady-state conditions could not be located by CV through variable scan-rate dependence studies with CO₂ saturation and 0.5 M PhOH up to 1 V s⁻¹ (Fig. S18–S21†).²³ Despite the appearance of current saturation in Fig. 5, FOWA suggests that the highest k_cat occurs at 2.50 M PhOH as 1.15 × 10¹⁰ M⁻² s⁻¹ with a TOF_max of 7.72 × 10⁸ s⁻¹ (Table S2†). The TOF_max values obtained by this method (Table S2†) show the effects of the Nernstian equilibrium observed by CV (E_1/2(cat) shifts 40 mV/log[PhOH]); a non-linear increase in TOF_max occurs at higher concentrations of [PhOH] as the Tafel relationship between catalytic and non-catalytic current shifts to increasingly positive potentials.

Characterization by controlled potential experiments

To establish a better mechanistic understanding of the behavior of 1, infrared spectroelectrochemistry (IR-SEC) was conducted. IR-SEC is a method whereby potential is varied as a function of time to extract speciation data from an electrochemical system.^56,76,77 Specifically, we were interested in observing changes in the IR absorbance bands of CO₂ and any presumptive reduction products with and without a proton source.⁴⁹

In the absence of a proton source, as the potential is increased from −1.6 to −2.1 V vs. Fc/Fc⁺, IR bands consistent with the formation of carbonate, CO₃²⁻, are observed to grow in intensity around 1671 and 1638 cm⁻¹ indicative of a reductive disproportionation reaction of two equivalents CO₂ to CO and CO₃²⁻ (Fig. 6B and S22†). As the potential is brought to −2.1 V vs. Fc/Fc⁺, a new IR band also grows in intensity at 1986 cm⁻¹, consistent with [Ni(κ⁵-TPEN)(CO)]⁺ formation, vide infra (Fig. 6A). Under the addition of a proton source, however, Ni(CO)₄ was observed (2043 cm⁻¹) beginning at −1.7 V vs. Fc/Fc⁺, suggesting that the formation of this degradation product was competitive with any catalytic reaction (Fig. 7A). Minor species are also observed near 1910 and 1888 cm⁻¹, which we attribute either to an intermediate in the pathway between 1 and Ni(CO)₄, different charge state of a [Ni(TPEN)(CO)]^x+ adduct, or a different conformer of Ni(TPEN)(CO); at prolonged exposure to potentials of −2.1 V vs. Fc/Fc⁺ and more negative, substantial accumulation of Ni(CO)₄ occurs (Fig. 7A).

Fig. 6 IR-SEC analysis of a 3 mM solution of 1 with CO₂ (A) Ni–CO stretch grows in at 1986 cm⁻¹ and a CO adduct grows in at 1888 cm⁻¹ from the applied potentials of −1.6 V to −2.1 V (B) the formation of CO₃²⁻ is observed at 1671 cm⁻¹ and 1638 cm⁻¹. Conditions: 0.1 M TBAPF₆/MeCN; the cell was referenced to internal ferrocene standard.

Fig. 7 (A) IR-SEC spectra of a 3 mM solution of 1 with the addition of 0.3 M PhOH with CO₂ sparged into solution for ∼30 s. (B) A 3 mM solution of 1 and 20 mM solution of [Ni(TMC)]²⁺ with 0.1 M PhOH and CO₂ sparged into solution for ∼30 s. Conditions: 0.1 M TBAPF₆/MeCN; the cell was referenced to internal ferrocene standard.

In order to confirm the identity of the proposed Ni–CO species, ¹³CO₂ was substituted for CO₂. The harmonic oscillator solution to Schrodinger's equation was used with the assumption that only the C–O bond is involved in the IR modes to determine the carbonyl shifts for the CO stretches of Ni–CO at 1986 cm⁻¹ and for Ni(CO)₄ at 2043 cm⁻¹. The experimentally observed shifts were consistent with those predicted by the harmonic oscillator (Fig. S23†), consistent with CO₂ being the source of all CO observed in these experiments. These experiments were also repeated using CO as the sparge gas. At −1.8 V vs. Fc/Fc⁺, a CO-saturated solution with 1 present showed the growth of a band consistent with the assigned [Ni(κ⁵-TPEN)(CO)]⁺ species at 1982 cm⁻¹, as well as the minor carbonyl-containing species at 1888 cm⁻¹ (Fig. S24†).

Indeed, attempts were made to analyze the catalytic efficiency of this system through controlled potential electrolysis (CPE) under CO₂ saturation with PhOH (0.5 M) present. Detectable amounts of CO and H₂ were not observed with prolonged electrolysis at −2.05 V vs. Fc/Fc⁺ under nominally anhydrous and added PhOH conditions (Fig. S25†), respectively; CO was only observed in trace amounts with 1 after 8.66 electron equivalents per molecule of catalyst were passed during 10 hours of electrolysis. We believe this is a result of the decomposition of the catalyst to Ni(CO)₄, which can be observed in the post bulk electrolysis solution via IR (Fig. S26†). This is consistent with the IR-SEC data showing degradation products accumulating (Fig. 7A). The control electrolysis conducted at the same potential under CO₂ saturation and 0.5 M PhOH without catalyst present was observed to generate H₂. A Student's t-test assuming equal variances (determined using an F-test) between this control electrolysis and 1 under catalytic conditions resulted in a P value of 1.15 × 10⁻⁶ between the two data sets, indicating that the control generated significantly more H₂ than 1 under comparable conditions.⁷⁸ These results suggest that the degradation of 1 occurs rapidly and prevents any substantial catalytic activity.

Addition of CO scavenger [Ni(TMC)](PF₆)₂

The absence of catalytic turnover during CPE, in addition to the observation of Ni(CO)₄, a well-known degradation product for Ni-based molecular electrocatalysts,^49,79 led us to infer that at an applied potential [Ni(TPEN)]²⁺ had a short experimental lifetime. In an attempt to verify a true catalytic response, we turned to the method developed by Kubiak and co-workers, who demonstrated that the use of [Ni(TMC)](PF₆)₂ as a CO sponge can enhance activity and prolong catalyst lifetime for the [Ni(cyclam)]²⁺ electrocatalyst.⁴⁹

Conducting CV experiments with 1 in the presence of varying concentrations of [Ni(TMC)](PF₆)₂ with PhOH as a proton source (Fig. S27†) and subsequent FOWA shows that k_cat is no longer proportional to PhOH concentration (Table S3†). The increase in k_cat with increasing concentrations of [Ni(TMC)](PF₆)₂ suggests that the degradation pathway is the dominant contributor to limiting the overall rate of 1 (Fig. S28 and Table S3†).

When IR-SEC experiments were repeated, the addition of the CO scavenger [Ni(TMC)]²⁺ to a CO₂ saturated solution of 1 suppressed formation of Ni(CO)₄ and the IR absorbance band of the CO capture product [Ni(TMC)(CO)]⁺ was clearly observed at 1967 cm⁻¹ at −1.6 V in Fig. 7B. If the potential is increased from −1.7 to −1.9 V vs. Fc/Fc⁺, a band associated with the Ni–CO adduct from 1 grew in as a shoulder at 1982 cm⁻¹ along with a new CO adduct at 1996 cm⁻¹ (Fig. S29†).

The CO scavenger [Ni(TMC)]²⁺ was added into the electrolysis solution to mitigate the effects of the degradation pathway and to determine if 1 was catalytically active using product analysis. However, the electrolysis experiment showed minimal CO (Fig. S30†). H₂ was detected and again determined to not be significant based on the P value, 5.86 × 10⁻⁵, obtained from Student's t-test assuming unequal variances with the control electrolysis.⁷⁸ Analysis of the post-electrolysis reaction solution showed minimal Ni(CO)₄ formation (Fig. S26†). Control CV experiments and IR-SEC experiments demonstrated that no additional reactivity is expected from [Ni(TMC)](PF₆)₂ under these conditions (Fig. S31 and S32†).

Since CO loss from [Ni(TMC)(CO)]⁺ does not readily occur in solution, we attempted to assess the concentration observed during catalytic conditions. A simple calibration curve was constructed by varying the concentration of added [Ni(TMC)]²⁺ present under CO saturation, focusing on data points from the regime where a linear intensity response is obtained (Fig. S33 and S34†). Using the linear fit equation, if it is assumed that intensity is still in a region with a linear concentration dependence during catalytic IR-SEC experiments, the amount of CO generated can be approximated. For data obtained at −1.9 V vs. Fc/Fc⁺ after 60 s with 3 mM 1 (30 s CO₂ sparge, 0.1 M PhOH, 20 mM [Ni(TMC)]²⁺) an estimated concentration of 5.9 mM CO-capture product is present (Fig. S35†). This corresponds to slightly less than two turnovers per equivalent of 1.

DFT calculations and the proposed mechanism

To supplement the experimental studies, the electronic structures of probable intermediates from the proposed reaction mechanism were examined computationally. In the ground state, the calculated structure of the [Ni(TPEN)]²⁺ parent species is consistent with that determined by X-ray crystallography (Fig. S36†). Kohn–Sham orbital plots of the orbital energy levels with two unpaired electrons, as well as a visualization of unpaired spin density, are consistent with the high-spin Ni²⁺ metal center observed by the Evans method (Fig. S37†). Geometry optimizations of the one-electron reduction product only produced a species at a minimum on the potential energy surface when one of the axial 2-pyridylmethyl arms had dissociated from the Ni center [Ni(κ⁵-TPEN)]⁺2. Although there are two options for the dissociation of a 2-pyridylmethyl arm, the only pathway through which all presumptive intermediate species were verifiable minima through numerical frequency calculations (or lowest in energy according to this level of theory) was one where this dissociation occurred in the equatorial plane of the starting material, shown as Scheme 1.

Scheme 1 Proposed mechanism for the electrochemical reduction of CO₂ to CO as well as the formation of Ni(CO)₄.

The presumptive initial CO₂ adduct was calculated as a low-spin monocationic octahedral Ni(III) complex [Ni(κ⁵-TPEN)(η¹-CO₂)]⁺3 (Fig. 8). Further reduction and protonation is presumed to result in a species similar to [Ni(κ⁵-TPEN)(η¹-CO₂H)]⁺4. With further protonation and C–O bond cleavage, a monocationic monocarbonyl [Ni(κ⁴-TPEN)(CO)]⁺5 species would result (Fig. S38†). The predicted IR frequency of this species, 2005 cm⁻¹, qualitatively compares with that observed experimentally by IR-SEC (1986 cm⁻¹); error correction was achieved using calculated frequencies for Ni(CO)₄ at this level of theory compared to the experimentally observed value.

Fig. 8 Kohn–Sham orbital representations of [Ni(TPEN)(η¹-CO₂)]⁺3 (A); SOMO (B); LUMO (C); spin density (D). ORCA 4.0; B3LYP/G; ZORA; def2-TZVP; CPCM(Acetonitrile), 2S + 1 = 3.

The complete proposed mechanism for the reactivity of 1 with CO₂ at reducing potentials is shown in Scheme 1. Consistent with our experimental observations, there are two routes to the cationic Ni monocarbonyl species [Ni(κ⁴-TPEN)(CO)]⁺5 that are observed by IR-SEC under protic and aprotic conditions. Under aprotic conditions, further reduction and a second equivalent of CO₂ enable a reductive disproportionation reaction to occur; the growth in intensity of IR absorption bands consistent with carbonate are observed between 1600–1700 cm⁻¹, as well as the accumulation of [Ni(κ⁴-TPEN)(CO)]⁺5 and Ni(CO)₄. With PhOH present as a sacrificial proton donor, bicarbonate formation is greatly suppressed, suggesting that H₂O is forming as a co-product of CO under these conditions; the Ni monocarbonyl intermediate and tetracarbonyl degradation product are again observed. With the CO scavenger [Ni(TMC)]²⁺ present, Ni(CO)₄ formation is almost completely suppressed, with the stoichiometric CO capture product [Ni(TMC)(CO)]⁺ observed by IR-SEC.

Conclusions

The electrochemical reduction of CO₂ by a molecular nickel(II) complex has been characterized by IR-SEC, controlled potential electrolysis, and CV experiments. Under aprotic and protic conditions (where PhOH is a proton donor), CO is formed as a two-electron reduction product. Once CO is formed, degradation to Ni(CO)₄ occurs rapidly. This deleterious reaction can be suppressed by stoichiometric CO capture by the CO sponge [Ni(TMC)]²⁺. These results suggest that although CO₂ activation is rapid and efficient with this ligand framework, modification to improve stability is a requirement for practical application. Work on new ligand designs to improve the ligand association equilibrium constant K_a by several orders of magnitude at reducing potentials is currently underway.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The University of Virginia for generous funding and ARCS for maintaining the Rivanna HPC Cluster.

Notes and references

1. N. S. Lewis and D. G. Nocera, Proc. Natl. Acad. Sci., 2006, 103, 15729–15735.
2. J.-R. Petit, J. Jouzel, D. Raynaud, N. I. Barkov, J.-M. Barnola, I. Basile, M. Bender, J. Chappellaz, M. Davis and G. Delaygue, Nature, 1999, 399, 429–436.
3. E. E. Benson, C. P. Kubiak, A. J. Sathrum and J. M. Smieja, Chem. Soc. Rev., 2009, 38, 89–99.
4. J. Qiao, Y. Liu, F. Hong and J. Zhang, Chem. Soc. Rev., 2014, 43, 631–675.
5. N. S. Spinner, J. A. Vega and W. E. Mustain, Catal. Sci. Technol., 2012, 2, 19–28.
6. M. Aresta, A. Dibenedetto and A. Angelini, Chem. Rev., 2014, 114, 1709–1742.
7. J. Hawecker, J.-M. Lehn and R. Ziessel, J. Chem. Soc., Chem. Commun., 1984, 328–330,  10.1039/c39840000328.
8. Y. Kou, Y. Nabetani, D. Masui, T. Shimada, S. Takagi, H. Tachibana and H. Inoue, J. Am. Chem. Soc., 2014, 136, 6021–6030.
9. R. Kuriki, K. Sekizawa, O. Ishitani and K. Maeda, Angew. Chem., Int. Ed., 2015, 54, 2406–2409.
10. C. W. Machan, S. A. Chabolla, J. Yin, M. K. Gilson, F. A. Tezcan and C. P. Kubiak, J. Am. Chem. Soc., 2014, 136, 14598–14607.
11. S. Sato, T. Morikawa, T. Kajino and O. Ishitani, Angew. Chem., 2013, 125, 1022–1026.
12. S. Sato, T. Morikawa, T. Kajino and O. Ishitani, Angew. Chem., Int. Ed., 2013, 52, 988–992.
13. Y. Tamaki, T. Morimoto, K. Koike and O. Ishitani, Proc. Natl. Acad. Sci., 2012, 109, 15673–15678.
14. Y. Kitagawa, H. Takeda, K. Ohashi, T. Asatani, D. Kosumi, H. Hashimoto, O. Ishitani and H. Tamiaki, Chem. Lett., 2014, 43, 1383–1385.
15. B. J. Fisher and R. Eisenberg, J. Am. Chem. Soc., 1980, 102, 7361–7363.
16. M. Beley, J.-P. Collin, R. Ruppert and J.-P. Sauvage, J. Chem. Soc., Chem. Commun., 1984, 1315–1316,  10.1039/c39840001315.
17. G. Neri, I. M. Aldous, J. J. Walsh, L. J. Hardwick and A. J. Cowan, Chem. Sci., 2016, 7, 1521–1526.
18. M. Bourrez, F. Molton, S. Chardon-Noblat and A. Deronzier, Angew. Chem., Int. Ed., 2011, 50, 9903–9906.
19. M. D. Sampson, A. D. Nguyen, K. A. Grice, C. E. Moore, A. L. Rheingold and C. P. Kubiak, J. Am. Chem. Soc., 2014, 136, 5460–5471.
20. I. Azcarate, C. Costentin, M. Robert and J.-M. Savéant, J. Am. Chem. Soc., 2016, 138, 16639–16644.
21. C. Costentin, S. Drouet, M. Robert and J.-M. Savéant, Science, 2012, 338, 90–94.
22. S. Roy, B. Sharma, J. Pécaut, P. Simon, M. Fontecave, P. D. Tran, E. Derat and V. Artero, J. Am. Chem. Soc., 2017, 139, 3685–3696.
23. K. T. Ngo, M. McKinnon, B. Mahanti, R. Narayanan, D. C. Grills, M. Z. Ertem and J. Rochford, J. Am. Chem. Soc., 2017, 139, 2604–2618.
24. N. D. Loewen, E. J. Thompson, M. Kagan, C. L. Banales, T. W. Myers, J. C. Fettinger and L. A. Berben, Chem. Sci., 2016, 7, 2728–2735.
25. N. D. Loewen, T. V. Neelakantan and L. A. Berben, Acc. Chem. Res., 2017, 50, 2362–2370.
26. M. D. Sampson and C. P. Kubiak, J. Am. Chem. Soc., 2016, 138, 1386–1393.
27. N. Elgrishi, M. B. Chambers, X. Wang and M. Fontecave, Chem. Soc. Rev., 2017, 46, 761–796.
28. M. Beley, J. P. Collin, R. Ruppert and J. P. Sauvage, J. Am. Chem. Soc., 1986, 108, 7461–7467.
29. S. L.-F. Chan, T. L. Lam, C. Yang, S.-C. Yan and N. M. Cheng, Chem. Commun., 2015, 51, 7799–7801.
30. A. Chapovetsky, T. H. Do, R. Haiges, M. K. Takase and S. C. Marinescu, J. Am. Chem. Soc., 2016, 138, 5765–5768.
31. L. Chen, Z. Guo, X.-G. Wei, C. Gallenkamp, J. Bonin, E. Anxolabéhère-Mallart, K.-C. Lau, T.-C. Lau and M. Robert, J. Am. Chem. Soc., 2015, 137, 10918–10921.
32. C. Costentin, M. Robert and J.-M. Savéant, Acc. Chem. Res., 2015, 48, 2996–3006.
33. D. L. DeLaet, R. Del Rosario, P. E. Fanwick and C. P. Kubiak, J. Am. Chem. Soc., 1987, 109, 754–758.
34. N. Elgrishi, M. B. Chambers and M. Fontecave, Chem. Sci., 2015, 6, 2522–2531.
35. R. J. Haines, R. E. Wittrig and C. P. Kubiak, Inorg. Chem., 1994, 33, 4723–4728.
36. D. C. Lacy, C. C. L. McCrory and J. C. Peters, Inorg. Chem., 2014, 53, 4980–4988.
37. S. Lin, C. S. Diercks, Y.-B. Zhang, N. Kornienko, E. M. Nichols, Y. Zhao, A. R. Paris, D. Kim, P. Yang, O. M. Yaghi and C. J. Chang, Science, 2015, 349, 1208–1213.
38. C. W. Machan and C. P. Kubiak, Dalton Trans., 2016, 45, 17179–17186.
39. M. Rakowski Dubois and D. L. Dubois, Acc. Chem. Res., 2009, 42, 1974–1982.
40. J. Schneider, H. Jia, J. T. Muckerman and E. Fujita, Chem. Soc. Rev., 2012, 41, 2036–2051.
41. C. M. Bolinger, B. P. Sullivan, D. Conrad, J. A. Gilbert, N. Story and T. J. Meyer, J. Chem. Soc., Chem. Commun., 1985, 796–797,  10.1039/c39850000796.
42. M. R. M. Bruce, E. Megehee, B. P. Sullivan, H. H. Thorp, T. R. O'Toole, A. Downard, J. R. Pugh and T. J. Meyer, Inorg. Chem., 1992, 31, 4864–4873.
43. N. Elgrishi, M. B. Chambers, V. Artero and M. Fontecave, Phys. Chem. Chem. Phys., 2014, 16, 13635–13644.
44. H. Ishida, H. Tanaka, K. Tanaka and T. Tanaka, J. Chem. Soc., Chem. Commun., 1987, 131–132,  10.1039/c39870000131.
45. J. M. Smieja and C. P. Kubiak, Inorg. Chem., 2010, 49, 9283–9289.
46. K.-Y. Wong, W.-H. Chung and C.-P. Lau, J. Electroanal. Chem., 1998, 453, 161–170.
47. T. C. Simpson and R. R. Durand, Electrochim. Acta, 1988, 33, 581–583.
48. C. A. Blindauer, M. T. Razi, S. Parsons and P. J. Sadler, Polyhedron, 2006, 25, 513–520.
49. J. D. Froehlich and C. P. Kubiak, J. Am. Chem. Soc., 2015, 137, 3565–3573.
50. A. Gennaro, A. A. Isse and E. Vianello, J. Electroanal. Chem. Interfacial Electrochem., 1990, 289, 203–215.
51. J. B. Mandel, C. Maricondi and B. E. Douglas, Inorg. Chem., 1988, 27, 2990–2996.
52. E. K. Barefield and F. Wagner, Inorg. Chem., 1973, 12, 2435–2439.
53. E. Evangelio, N. P. Rath and L. M. Mirica, Dalton Trans., 2012, 41, 8010–8021.
54. B. Bosnich, M. L. Tobe and G. A. Webb, Inorg. Chem., 1965, 4, 1109–1112.
55. I. S. Zavarine and C. P. Kubiak, J. Electroanal. Chem., 2001, 495, 106–109.
56. C. W. Machan, M. D. Sampson, S. A. Chabolla, T. Dang and C. P. Kubiak, Organometallics, 2014, 33, 4550–4559.
57. A. W. Nichols, S. Chatterjee, M. Sabat and C. W. Machan, Inorg. Chem., 2018, 57, 2111–2121.
58. F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2, 73–78.
59. F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2018, 8, e1327.
60. S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58, 1200–1211.
61. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785–789.
62. P. J. Stephens, F. J. Devlin, C. F. Chabalowski and M. J. Frisch, J. Phys. Chem., 1994, 98, 11623–11627.
63. A. Schäfer, H. Horn and R. Ahlrichs, J. Chem. Phys., 1992, 97, 2571–2577.
64. F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297–3305.
65. A. D. Becke, J. Chem. Phys., 1993, 98, 1372–1377.
66. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
67. R. Izsák and F. Neese, J. Chem. Phys., 2011, 135, 144105.
68. S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys., 2010, 132, 154104.
69. S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem., 2011, 32, 1456–1465.
70. M. Cossi, N. Rega, G. Scalmani and V. Barone, J. Comput. Chem., 2003, 24, 669–681.
71. E. F. Pettersen, T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng and T. E. Ferrin, J. Comput. Chem., 2004, 25, 1605–1612.
72. G. A. Bain and J. F. Berry, J. Chem. Educ., 2008, 85, 532.
73. C. Piguet, J. Chem. Educ., 1997, 74, 815.
74. A. M. Appel and M. L. Helm, ACS Catal., 2014, 4, 630–633.
75. C. Costentin, S. Drouet, M. Robert and J.-M. Savéant, J. Am. Chem. Soc., 2012, 134, 11235–11242.
76. K. Ashley and S. Pons, Chem. Rev., 1988, 88, 673–695.
77. W. Kaim and J. Fiedler, Chem. Soc. Rev., 2009, 38, 3373–3382.
78. D. C. Harris, Quantitative Chemical Analysis, W. H. Freeman and Company, New York, 7th edn, 2007.
79. G. B. Balazs and F. C. Anson, J. Electroanal. Chem., 1993, 361, 149–157.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details of the Evans method, cyclic voltammograms, foot-of-the-wave graphs and table summaries, kinetic data plots, and DFT coordinates. CCDC 1816890. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c8se00027a

---