Salen ligand complexes as electrocatalysts for direct electrochemical reduction of gaseous carbon dioxide to value added products

Abstract

CO₂, being a linear and centrosymmetric molecule, is very stable, and the electrochemical reduction of CO₂ requires energy. Moreover, the CO₂ electroreduction proceeds at a significantly higher applied voltage than thermodynamically required. In this concern, salen ligand, H₂L^NO₂, and its corresponding Ni(II) and Cu(II) complexes are synthesized and reported for the first time as electrocatalysts for CO₂ electroreduction. Both the metal complexes are active for CO₂ reduction and surprisingly result in formation of C₁ and C₂ hydrocarbons. The complexes are found to be very efficient to minimize overpotential required for the reaction as compared to their pure metal counterparts. Thus, the present study opens up a new class of metal complexes for the efficient electrochemical reduction of CO₂ having the potential to generate hydrocarbons at lower overpotentials.

---

1. Introduction

Utilization of CO₂ for curbing the global warming menace is one of the highly investigated fields in the present scenario. In this respect, the electrocatalytic reduction of CO₂ (ERC) may pave possibly a better way to mitigate and utilize CO₂, along with the generation of high calorific value fuels. Moreover, the use of renewable energy required for the CO₂ reduction can be stored in the form of fuels.^1–5 Eqn (1)–(3) represent formation of a few of the high calorific value products using electrochemical reduction of CO₂. The equations show a full cell reaction, where 2, 8, and 12 moles of electrons are utilized for the formation of one mole of CO, CH₄, and C₂H₄, respectively, at corresponding standard cell potentials. Thus ERC may be helpful not only for the reduction of CO₂ and generation of the conventional energy resources but also for the storage of renewable energy.

2CO₂ + H₂O → 2CO + O₂ + H₂O, E° = −1.33 V (1)

CO₂ + 2H₂O → CH₄ + 2O₂, E° = −1.06 V (2)

2CO₂ + 2H₂O → C₂H₄ + 3O₂, E° = −1.15 V (3)

Though the electrochemical reduction of CO₂ offers advantages for the formation of value added products but the process is quite energy intensive. Most of the energy is used up towards conversion of CO₂ by utilizing one-electron to form a ˙CO₂⁻ anion radical, which has bent geometry. Generally, metals, such as Cu, Ru, and Mo, have been studied for the ERC process because of their ability to undergo multi-electron reduction of CO₂, forming higher hydrocarbons and/or alcohols.^6–8 However, large activation overpotential is required for the formation of such reaction products using metals.⁹ Therefore, metal complexes are being studied in order to reduce the activation overpotential using well designed ligand and central metal atoms.^2,9,10 Macrocyclic metal complexes, such as Fe porphyrins; Mn, Cu, Zn phthalocyanines; and Ru bipyridines, have been extensively studied for CO₂ electroreduction.^11–17 Most of the systems resulted in the two-electron reduction of CO₂, forming CO and/or HCOO⁻. A few of the workers also reported six and eight-electron reduction mechanism of ERC resulting in CH₃OH and CH₄ formation using Mn, Zn porphyrins and Cu, Ga and Ti phthalocyanines, respectively.^18–20 Thus, the synergetics of suitable metal and ligand in metal complexes may have a greater role in the formation of high calorific value products as well as lowering down the overpotential. In the present study, for the first time, the use of salen ligand metal complex is reported as the electrocatalyst for direct electrochemical reduction of the gaseous CO₂.

Salen ligand complexes are well known for their ability to chemically catalyze various reactions.^21–24 Therefore, Gambarotta et al. studied the bifunctional salen ligand complex containing nucleophilic cobalt(I) and an alkali cation for CO₂ capture. They have found very encouraging results, where acidic–basic (bifunctional) sites present in the metal complex played a role in binding the CO₂.²⁵ In fact, this inculcated the idea to bind the CO₂ at the site and simultaneously reduce it at the electrode. Moreover, there is no study on such complexes for CO₂ electroreduction. Hence, the potential of salen complexes needs to be explored in the field of ERC. In this effort, the primary objective of the paper is to evaluate the salen metal complexes for the electrochemical reduction of CO₂ for the formation of value added products. To establish and evaluate the ERC, Ni and Cu were used to form salen metal complex electrocatalysts. Cu is well known to facilitate the formation of hydrocarbons, whereas Ni mostly results in undesired H₂ production with meager formation of CO.^26–28 Therefore, these metals were chosen to understand the effect of salen metal complex over ERC. It may be noted that the direct use of gaseous CO₂ for the electrochemical reduction of CO₂ (dERC) is being used.²⁹

2. Experimental

2.1 Materials

4-Nitro-o-phenylenediamine was procured from Alfa Aesar. Potassium bicarbonate (KHCO₃), N,N-dimethylformamide (DMF), NiSO₄·6H₂O and CuSO₄·5H₂O were purchased from Loba Chemie. Acetonitrile (CH₃CN), NiCl₂·6H₂O, CuCl₂·2H₂O, triethylamine (Et₃N), ethanol (EtOH) and salicylaldehyde were procured from Merck. CDCl₃ for NMR spectroscopy was purchased from Sigma Aldrich. Pt/C (40 wt% Pt) was purchased from ElectroChem, Inc. USA. All the chemicals were used without further purification and de-ionized water was used in all the experiments unless stated otherwise. Gas diffusion layer from CeTech was used as a backing layer for electrocatalyst film to form an electrode. Nafion membrane was procured from DuPont, USA.

2.2 Method

2.2.1 Characterization techniques. The synthesized ligand and metal complexes were characterized using Fourier transform infrared spectroscopy (FTIR), nuclear magnetic resonance spectroscopy (NMR), high-resolution mass spectrometry (HRMS), CHN analysis, and electrochemical half-cell studies. FTIR spectra were recorded in the range of 500–4000 cm⁻¹ using a Perkin Elmer FTIR spectrometer by grinding the sample with KBr (IR grade). The NMR spectrum was recorded at a 400 MHz frequency using a Varian, Mercury Plus nuclear magnetic resonance spectrometer. To know the mass of the ligand, the spectrum was obtained using an Agilent high-resolution mass spectrometer. CHN analysis was carried out using a EuroEA elemental analyser. Electrochemical tests (cyclic voltammetry (CV) and chronoamperometry) were performed using a potentiostat/galvanostat (CH Instruments Inc.). Product analysis was carried out using gas chromatography (GC) (Nucon 5765).

2.2.2 Synthesis of the salen ligand (H₂L^NO₂) and corresponding complexes. 4-Nitro-o-phenylenediamine (A) contains a highly electron withdrawing –NO₂ group at the –para position of one of the two amine groups (Scheme 1), and hence, the –NO₂ group reduces the nucleophilic character of the amine group. Therefore, the ligand, H₂L^NO₂, was synthesized by reacting a large excess (6 equivalents) of salicylaldehyde with A in ethanol under refluxing condition. The ligand upon reacting with equivalent amounts of NiCl₂·6H₂O/CuCl₂·2H₂O in acetonitrile in the presence of Et₃N provided Ni complex/Cu complex, respectively. The synthesis route of the ligand and the metal complexes is shown in Scheme 1.

Scheme 1 Synthesis route to formation of the ligand and corresponding Ni(II) and Cu(II) complexes (A = 4-nitro-o-phenylenediamine).

2.2.2.1 Synthesis of salen ligand (H₂L^NO₂). A suspension of 4-nitro-o-phenylenediamine (1.53 g, 10 mmol) was prepared in ethanol solution (20 mL). Salicylaldehyde (7.33 g, 60 mmol) was added dropwise into the prepared suspension to form a mixture. The resulting mixture was heated to reflux for 20 hours. A yellow precipitate appeared after cooling the mixture at 30 °C. The precipitate was filtered, washed thoroughly with EtOH, and finally air dried to get H₂L^NO₂. Yield: 78%. FTIR (KBr pellet cm⁻¹): 3447, 1615, 1566, 1511, 1478, 1341, 1277, 1194, 1153, 910, 843, 765. ¹H NMR (CDCl₃, 400 MHz): δ 6.97 (td, J = 7.6 Hz, 1.2 Hz, 2H), 7.05 (t, J = 8.0 Hz, 2H), 7.33 (d, J = 8.8 Hz, 1H), 7.41–7.46 (m, 4H), 8.15 (d, J = 2.4 Hz, 1H), 8.18 (dd, J = 8.8 Hz, 2.4 Hz, 1H), 8.65 (s, 1H), 8.73 (s, 1H), 12.50 (s, 1H), 12.56 (s, 1H). HRMS (+) m/z for [C₂₀H₁₅N₃O₄ + H]: calcd, 362.1135; found, 362.1115.
2.2.2.2 Synthesis of metal complexes [Ni^IIL^NO₂ and Cu^IIL^NO₂]. In a CH₃CN solution (10 mL) of H₂L^NO₂ (0.180 g, 0.5 mmol), NiCl₂·6H₂O (0.120 g, 0.50 mmol)/CuCl₂·2H₂O (0.090 g, 0.50 mmol) and triethyl amine (0.2 mL) were added, and the resulting solution was stirred at 30 °C for five hours. During this time the color of the solution changed to reddish-brown (for Ni complex)/yellowish-brown (for Cu complex) with concomitant precipitation of the solid with the abovementioned color. Furthermore, the solid was filtered, washed thoroughly with CH₃CN, and air dried.
Ni complex. Yield: 80%. FTIR (KBr pellet cm⁻¹): 3433, 1610, 1576, 1509, 1335, 1279, 1248, 1198, 1155, 914, 836, 768. Anal. calcd for [C₂₀H₁₃N₃O₄Ni]: %C, 57.55; %H, 3.14; %N, 10.07. Found: %C, 59.46; %H, 3.14; %N, 9.66.
Cu complex. Yield: 82%. FTIR (KBr pellet cm⁻¹): 3320, 1612, 1575, 1509, 1334, 1259, 1189, 1154, 875, 818, 760. Anal. calcd for [C₂₀H₁₃N₃O₄Cu·CH₃CN·0.5H₂O]: %C, 55.93; %H, 3.63; %N, 11.87. Found: %C, 56.04; %H, 3.26; %N, 11.92.

2.2.3 Preparation of electrodes and membrane electrode assembly. Pt/C (40 wt% Pt), with Nafion dispersion (5 wt%) as binder, was used to prepare the anode electrocatalyst ink. The ink was coated over the gas diffusion layer using spraying technique and dried to get the anode. For the cathode (or working electrode), the ink was prepared by dispersing the metal complex electrocatalyst in an appropriate solvent along with Nafion dispersion. The electrocatalyst ink was then coated over the gas diffusion layer followed by drying. For the preparation of pure metal electrocatalysts (cathode) of Ni and Cu, electroplating was done over the gas diffusion layer using electroplating baths of NiSO₄·6H₂O and CuSO₄·5H₂O respectively. Suitable additives were added in respective baths to get the smooth deposition of metals over the gas diffusion layer. The electrocatalyst loading for anode and cathode was kept as 0.5 mg cm⁻² and 2 mg cm⁻², respectively. The cation exchange membrane (Nafion-117) was treated using the standard procedure³⁰ before the experiments. The membrane electrode assembly was prepared by sandwiching the Nafion membrane in between anode and cathode at a particular temperature and pressure.

2.2.4 Electrochemical tests. The half-cell studies were conducted in a three-electrode cell assembly using Ag/AgCl as a reference electrode and Pt wire as a counter electrode. The working electrode was prepared using a 2 mg cm⁻² electrocatalyst loading supported over the gas diffusion layer. The cyclic voltammetry test was carried out in aqueous (0.5 M KHCO₃) and non-aqueous (dimethylformamide, DMF) electrolytes at the scan rate of 0.01 V s⁻¹. In all the half-cell studies, the potential is reported with respect to standard hydrogen electrode (SHE), unless stated otherwise. The full cell studies were carried out in an electrochemical reactor as described elsewhere.^31,32 The electrode area of 3 × 3 cm² was used for both anode and cathode of the electrochemical reactor. The cathode side of the reactor was fed with CO₂ gas (99.995%) with the flow rate of 20 mL min⁻¹ adjusted by a rotameter. Deionized water was used as proton source at the anode with the flow rate of 1.5 mL min⁻¹. The experiment was conducted using chronoamperometry technique (pulse width 5 s) at 8 different electrode potentials and the reaction products were collected at the cathode outlet and analyzed by GC. The electrochemical experiments were carried out at 30 °C and at atmospheric pressure.

3. Results and discussion

3.1 Metal complex characterization

The salen ligand and metal complexes were characterized using FTIR spectroscopy along with other characterization techniques as described in Section 2.2.2. However, an ESI-MS measurement for the complexes was not possible due to insoluble nature of the complexes in the measurement friendly solvents. The infrared spectrum of ligand H₂L^NO₂ showed a broad band centered at 3447 cm⁻¹ due to ν(O–H) stretching of two –OH groups. The broadness of the band indicated hydrogen bonding in the ligand. The aldamine C=N stretch appeared at 1615 cm⁻¹. A sharp band at 1566 cm⁻¹ was found because of a phenyl asymmetric ν(C=C) stretch. The existence of bands at 1511 cm⁻¹ and 1341 cm⁻¹ consolidated the presence of –NO₂ functional group in the ligand. The phenolic ν(C–O) band appeared at 1277 cm⁻¹. In Ni and Cu complexes, the aldamine C=N stretch shifted slightly to a lower value and appeared at 1610 cm⁻¹ and 1612 cm⁻¹, respectively. After complexation, an asymmetric ν(C=C) stretching band for the phenyl ring appeared at 1576 cm⁻¹ and 1575 cm⁻¹ for the Ni complex and Cu complex, respectively, and the value was ∼10 cm⁻¹ higher compared to the value observed in the free ligand. This indicated more aromatic character inside the phenyl rings in the complexes compared to the ligand. In addition, a lower phenolic ν(C–O) stretching value in the complexes (1248 cm⁻¹ in Ni complex and 1259 cm⁻¹ in Cu complex) compared to free ligand (1277 cm⁻¹) supported further the higher aromatic character of the phenolate moiety. In the IR spectra of both the complexes, no appreciable changes were observed in 1511 cm⁻¹ and 1341 cm⁻¹ stretching bands. This indicated an undisturbed and non-interacting nature of the –NO₂ group in both Ni and Cu complexes. Interestingly, in Cu complex a relatively weak and broad band appeared at 3321 cm⁻¹. The band is attributed to the ν(O–H) stretching of H-bonded lattice water. Microanalysis data further supported the presence of lattice water.

3.2 Half-cell studies

To find out the feasibility of the synthesized metal complexes for electrochemical reduction of CO₂, half-cell studies were carried out. Metal complexes were studied in both aqueous and non-aqueous medium. In aqueous medium, the test was conducted in 0.5 M KHCO₃ solution, which was saturated with N₂ or CO₂ gases, to compare the voltammograms in inert and CO₂ environments, respectively. The pH of the saturated solution with N₂ and CO₂ was 8.5 and 7.5, respectively. The cyclic voltammograms (CVs) of the Ni and Cu complexes in the potential range of 0 to −3 V vs. SHE are shown in Fig. 1. The CVs were recorded in inert gas (N₂) as well as in CO₂ environment for both the complexes. It can be seen that a significant amount of current density is observed in both environments. However, the current density in the CO₂ saturated system is higher than the corresponding current density in the N₂ saturated system for both complexes. The current density observed in the inert atmosphere is due to the reduction of water into H₂ gas. The onset potential (potential at which 0.1 mA cm⁻² current density is observed) was found to be ∼1 V for both complexes. It can be seen that with the increase in the applied voltage beyond onset potential, the current density increases (Fig. 1). It can be seen that the current density was found to be more, up to some extent, in the CO₂ environment as compared to the inert environment for both complexes. The higher current density in the CO₂ saturated environment may signify that electrocatalysts are able to reduce CO₂. However, in aqueous medium the increase in current density is not only from the CO₂ reduction but also due to the reduction of H+ at the cathode, which is evident from Fig. 1. Moreover, the different pH level of the saturated solutions also affect the cathodic process as low pH favours the water reduction to hydrogen gas.³³ Therefore, efficacy of the metal complex electrocatalysts may not be truly ascertained by the abovementioned study. It may also be noted that the CV was conducted, and the products were evaluated using only the salen ligand (without metal). It was found that only hydrogen gas was evolved, which rules out the possibility that the salen ligands themselves are decomposing under reducing conditions to give the value added products. Hence, CO₂ electroreduction was also carried out in non-aqueous medium to get more information about the electrocatalysts.

Fig. 1 Cyclic voltammograms of the Ni complex (a) and Cu complex (b) in N₂ (broken line) and CO₂ (solid line) saturated in aqueous solution.

To further confirm the electrocatalytic activity of the complexes towards CO₂ reduction, the half-cell tests were conducted in a non-aqueous environment (DMF solvent) to avoid any effect due to H⁺ reduction to form H₂ gas.³⁴ The CV scans in a non-aqueous environment for both the metal complexes are shown in Fig. 2. It can be seen that the scan was recorded in absence of CO₂, and thus, only the charging current could be observed as there is no possibility of electron transfer between the electrolyte and electrode in the absence of any redox species (CO₂). Furthermore, CO₂ was bubbled in DMF, and the pH was maintained so as to have the same pH when CO₂ was saturated in the aqueous electrolyte.³⁴ Fig. 2 shows that the current density increases in the presence of CO₂ because of the transfer of electrons from the electrode surface to the dissolved CO₂. This increased current density in CO₂ environment confirms the reduction of CO₂, as there is no other electroactive species present in the medium. Hence, the increase in current density is solely attributed to the CO₂ reduction. It was also seen that there is no increment in current density after −2.8 V vs. SHE and −2.6 V vs. SHE in CV of Ni complex and Cu complex, respectively. Moreover, on comparing the corresponding current densities in Fig. 1 and 2, it can be seen that the magnitude of the current density is significantly low in the case of Fig. 2. In the case of CO₂ saturated aqueous system (Fig. 1) a significant amount of current density was due to the reduction of water into hydrogen gas at the cathode apart from the CO₂ reduction. In the case of DMF (Fig. 2), the CO₂ was bubbled only up to a certain level to maintain the pH instead of CO₂ saturation in the DMF. Moreover, the electroreduction might not be completed due to the absence of other reactants to combine with the reduced or activated CO₂. In turn, the low level of CO₂ in the electrolyte along with increased mass transport resistance and scarcity of the co-reactant reduces the current density, which does not increase even with the increase in the applied voltage. Apart from other reasons, the difference (in the order of 10³) in electrical conductivities of the media (mS cm⁻¹ in case of KHCO₃, whereas μS cm⁻¹ in case of DMF) also affects the performance.³⁴ Therefore, these studies confirm qualitatively the efficacy of the metal complexes, however, cannot be used for quantifying the role of the electrocatalysts. Therefore, further study was conducted to evaluate and quantify the products obtained out of the dERC using Ni and Cu complexes.

Fig. 2 Cyclic voltammograms of the Ni complex (a) and Cu complex (b) in the absence (broken line) and presence of CO₂ (solid line) in dimethyl formamide.

3.3 Electrochemical reduction of CO₂ using metal complexes in electrochemical reactor

Once the activity of the synthesized metal complexes was proved for the electrochemical reduction of CO₂ in preliminary half-cell studies, the full cell studies were attempted for the qualitative and quantitative information regarding the products formed in dERC using the complexes. The reactions were carried out in aqueous solution and above the onset potential range, which is ∼1 V in all the cases. The products formed were collected at definite time intervals from the cathode outlet and analyzed using gas chromatography. It may be noted that the experiments without the cathode (electrocatalyst) (that is, with only carbon paper) hardly produce any product.

3.3.1 Effect of applied voltage on current density and Faradaic efficiency towards total CO₂ reduction. For both the metal complex electrocatalysts, the experimental results obtained are presented in Fig. 3. Fig. 3 shows the current density and Faradaic efficiency corresponding to various applied voltages at the terminals of the electrochemical reactor. It can be seen that the current density increases with the increase in the applied voltage for both the complexes. It depicts that with increase in the applied voltage the current density increases, which in turn signifies the increased rate of various electrochemical reactions including H₂ gas generation at the electrode surface.

Fig. 3 Effect of applied voltage on current density (a) and Faradaic efficiency (b) for dERC using metal complex electrocatalysts.

Fig. 3(b) shows the Faradaic efficiency for electrochemical reduction of CO₂ using metal complexes. It can be seen that the Faradaic efficiency for the Ni complex is quite high viz. 74% at 1.5 V. On further increasing the voltage, the Faradaic efficiency decreases sharply. The Faradaic efficiency for the Cu complex increases initially and attains maxima at 1.8 V and then subsides gradually with the increase in the voltage. The highest Faradaic efficiency of 25% is achieved at 1.8 V for the Cu complex electrocatalyst. It may be noted that the Faradaic efficiency using the complexes is the result of cumulative efficiencies for the formation of various products from electroreduction of CO₂. To determine the energy utilized (Faradaic efficiency) for the various reaction products, it is necessary to quantify the reaction products at the cathode of the electrochemical reactor.

3.3.2 Effect of applied potential on Faradaic efficiency of individual products. To get more insight into the metal complex performance for the formation of various products, studies were conducted to find the Faradaic efficiencies of individual products of dERC as well as hydrogen production. Fig. 4 shows the Faradaic efficiency for the various products formed for dERC using Ni and Cu complexes.

Fig. 4 Effect of applied potential on Faradaic efficiency of the products using Ni-complex (a) and Cu-complex (b).

It can be seen in Fig. 4 that using metal complex electrocatalysts, the main products of dERC are hydrocarbons and carbon monoxide. The Ni complex resulted in better Faradaic efficiency than the Cu complex. In the case of Ni complex, Faradaic efficiency for CH₄ and C₂H₆ is considerably high with the maximum for CO formation (Fig. 4a). It is to be noted that CO, CH₄ and C₂H₆ show the highest Faradaic efficiency at 1.5 V. However, C₂H₄, which was minor, formed at 1.4 V with Faradaic efficiency of 2.1%. In the case of Cu complex, CH₄ was the major product along with considerable Faradaic efficiency of CO (Fig. 4b). Moreover, C₂H₄ was formed feebly with Faradaic efficiency of 1%. It can be seen that the Faradaic efficiency (23%) for CH₄ is highest at 1.5 V; however, CO attains its maxima at 1.8 V with Faradaic efficiency of 10.5%. Apart from the products resulting from CO₂ reduction, H₂ formation was also found in both the complexes, which approximately compensate for rest of the Faradaic efficiency at any particular voltage.

In order to find the role of the salen ligand in the metal complex, the experiments were also carried out using pure Ni and Cu metal electrocatalysts and were compared with the metal complex electrocatalysts. The comparison of Faradaic efficiency of metal complexes with corresponding pure metals is shown in Fig. 5. Fig. 5 shows a few of the very interesting results. It can be seen that the Faradaic efficiency towards the generation of the undesired H₂ gas is reduced significantly using any of the metal complexes as compared to the pure metal electrocatalyst, which in turn increased the Faradaic efficiency towards other useful products. Moreover, the Faradaic efficiency of CO and hydrocarbons were significantly higher for the metal complexes as compared to the pure metal electrocatalysts. Thus, the metal complexes show low operating voltage values and formation of C₂ hydrocarbons as compared to pure metals, where only a meagre amount of CH₄ and CO could be formed at a high operating voltage. In the case of Ni metal, two products were formed with Faradaic efficiency of 1.4% for CO and 1.2% for CH₄. However, in the case of Cu metal, the Faradaic efficiency for CH₄ and CO was found to be 4.9% and 0.3%, respectively. C₂ hydrocarbons were not formed in any of the pure metals, and most of the supplied energy was inadvertently used for the H₂ generation. In contrast to this, metal complexes resulted in efficient CO₂ reduction with significantly low Faradaic efficiency towards H₂ generation. It can be said that the ligand has profound impact over the process as it is evident from the performance of Ni and Cu as pure metals and in the complexed form with salen ligand. Thus, drastic change was found in the efficiency of the metals and corresponding metal complexes. Table 1 shows the applied voltage for a few representative reaction products at maximum Faradaic efficiency using the metal (Ni or Cu) and metal complexes. It can be seen that the metal complexes use lower voltages corresponding to their metal counterparts for the product formation at their maximum Faradaic efficiency. It is very interesting and worth noting that the operating voltage of the reaction reduced significantly using the salen ligand metal complex as compared to the pure metal electrocatalyst. The reduction in operating voltage shows the reduction in overpotential of the reactions.

Fig. 5 Comparison of Faradaic efficiency of metal complexes (solid line) with corresponding pure metals (broken line); Ni complex vs. Ni metal (a) and Cu complex vs. Cu metal (b).

Table 1 Reduction in operating voltage for a few representative reaction products using the metal (Ni or Cu) and metal complexes

Electrocatalysts	Applied voltage (V) corresponding to maximum Faradaic efficiency of the product
	CH4	CO	C2H6
Ni	2	1.8	—
Ni-complex	1.5	1.5	1.5
Reduction in operating voltage (V)	0.5	0.3	—
Cu	2	2	—
Cu-complex	1.5	1.8	—
Reduction in operating voltage (V)	0.5	0.2	—

---

It is evident that both the Ni and Cu complexes were active towards CO₂ reduction and underwent multi-electron reduction of CO₂ resulting in C₁ and C₂ hydrocarbons, which was usually rare in case of other metal complexes. It happened because incorporation of the anchoring –NO₂ group in the ligand (H₂L^NO₂) backbone provides better interaction of salen complexes to the electrode material and hence facilitates an efficient electron transfer during the conversion of CO₂ to the products. In addition, during the conversion process, when Mⁿ⁺ reduces to M⁽ⁿ⁻¹⁾⁺ state (M stands for metal), the decrease in Lewis basicity of the central metal ion due to an increase in metal-to-ligand(imine) π-back donation is compensated by two strong σ donations by two phenolate donors as shown in Scheme 2. This is due to a more aromatic character inside the phenyl rings of the complexes, which was depicted in the FTIR spectra. Moreover, the operating voltage required for the reaction was quite low for both the metal complexes as compared to pure metals (Table 1). It is because of the initial reduction of the metal complex itself (Mⁿ⁺ to M⁽ⁿ⁻¹⁾⁺) and subsequent reaction with CO₂. Thus, metal complexes act as a redox couple, which help to reduce the operating voltage for the CO₂ electroreduction as well as help in multielectron transfer resulting in C₁ and C₂ hydrocarbons.³⁵

Scheme 2 Representation of redox behaviour of salen ligand complex.

Overall, Ni complex was found to perform considerably better than Cu complex, which can be explained in terms of the stability of the Ni(I) and Cu(I) intermediate species. As it is known, the optimum stability or binding of metal–CO₂ bond is crucial for multi-electron reduction during dERC. The strong binding of metal–CO₂ may poison the electrocatalyst, whereas weak bonding may result in higher overpotential for the reaction along with dissociation of product from the reaction site by utilizing only a few electrons. During the CO₂ reduction process, Ni(II) and Cu(II) converts to Ni(I) and Cu(I) species, respectively. Ni(I) being less stable, coordinates with CO₂ yielding stable Ni–CO₂ complex³⁶ thus resulting in effective multi-electron reduction of CO₂ with high efficiency; however, Cu(I) is comparatively more stable and hence does not bind with CO₂ with the same efficacy, which in turn results in lower performance. Based on the products formed, an elementary reaction mechanism is proposed for the formation of hydrocarbons and CO as shown in Scheme 3. Scheme 3 shows that initially the metal complexed to the salen ligand is in +2 state and reduced to +1 state by taking up an electron from the electrode. The addition of CO₂ in the system receives the electron from the reduced metal, which was in +1 state, and binds with it. However, the stability of the system was maintained by the phenolate groups of the ligand. Simultaneously, H⁺ ions were transported through the membrane from the anode side to the cathode during the full cell reaction and reacted with the metal complex–CO₂ along with some internal rearrangements in the complex to form either the adsorbed CO (CO_ad) or free CO, which escapes from the cathode. The CO_ad may further participate in the reaction to form hydrocarbons^37,38 as shown in Scheme 3. Ultimately, the metal in the complex reverts to its previous oxidation state (+2). This redox behaviour of the salen metal complex is very peculiar and needs further investigation. The intermediate steps are not completely elucidated and are under investigation. Moreover, these metal complexes offer immense opportunity to study and improve the selective formation of hydrocarbons. Further work needs to be done to get more insights into the detailed reaction mechanism and to understand the salen metal complex.

Scheme 3 Proposed elementary mechanism for the formation of reaction products using salen metal complexes.

4. Conclusions

The electrocatalytic property of the salen ligand was unexplored in the field of ERC in spite of its ability to bind the CO₂. Hence, the salen ligand and its corresponding Ni and Cu complexes were synthesized and reported for their application as electrocatalysts in direct electrochemical reduction of CO₂ using half-cell and full cell studies. The complexes were found to be quite efficient for the process and resulted in the production of hydrocarbons (C₁ and C₂) and CO. The total Faradaic efficiency was found to be 74% for Ni complex and 25% for Cu complex at 1.5 V and 1.8 V, respectively. The better performance by the Ni complex as compared to the Cu complex was attributed to the low stability of Ni(I) as compared to Cu(I). It was also seen that Cu complex was more selective towards hydrocarbons as compared to Ni complex. Moreover, the Ni-complex showed better efficiency towards CO formation as compared to Cu-complex. The activity of complexes was also compared with the corresponding pure metals to understand the effect of the ligand. It was found that the redox behaviour of the metal complex is responsible for their better performance, and hence, the complexes were able to decrease the overpotential up to a great extent. The overpotential using metal-complexes was reduced significantly ranging from 10% to 25% against their pure metal counterparts for formation of different products. Moreover, a reaction pathway was proposed for formation of the various reaction products. It was observed that the formation of hydrocarbons using metal complexes is rather rare, and hence, these complexes offer an immense scope for future research in this field.

Acknowledgements

Authors are thankful to the ‘National Program on Carbon Sequestration Research’, Department of Science and Technology (DST), New Delhi for the financial support (project grant number DST/IS-STAC/CO₂-SR-139/12(G): Project Investigator: Anil Verma, IIT Delhi) and the Central Instruments Facility (CIF), IIT Guwahati, for providing the NMR facility for carrying out this work.

References

1. M. Jitaru, J. Univ. Chem. Technol. Metall., 2007, 42, 333–344.
2. E. E. Benson, C. P. Kubiak, A. J. Sathrum and J. M. Smieja, Chem. Soc. Rev., 2009, 38, 89–99.
3. G. Centi and S. Perathoner, Catal. Today, 2009, 148, 191–205.
4. N. S. Spinner, J. A. Vega and W. E. Mustain, Catal. Sci. Technol., 2012, 2, 19–28.
5. C. Costentin, M. Robert and J. M. Saveant, Chem. Soc. Rev., 2013, 42, 2423–2436.
6. K. P. Kuhl, E. R. Cave, D. N. Abram and T. F. Jaramillo, Energy Environ. Sci., 2012, 5, 7050–7059.
7. K. W. Frese Jr and S. Leach, J. Electrochem. Soc., 1985, 132, 259–260.
8. D. P. Summers, S. Leach and K. W. Frese Jr, J. Electroanal. Chem., 1986, 205, 219–232.
9. C. Finn, S. Schnittger, L. J. Yellowlees and J. B. Love, Chem. Commun., 2012, 48, 1392–1399.
10. E. B. Cole and A. B. Bocarsly, in Carbon Dioxide as Chemical Feedstock, ed. M. Aresta, Wiley-VCH Verlag GmbH & Co., Weinheim, Germany, 2010, ch. 11, pp. 291–316.
11. M. N. Mahmood, D. Masheder and C. J. Harty, J. Appl. Electrochem., 1987, 17, 1223–1227.
12. M. Hammouche, D. Lexa, M. Momenteau and J. M. Saveant, J. Am. Chem. Soc., 1991, 113, 8455–8466.
13. I. Bhugun, D. Lexa and J. M. Saveant, J. Am. Chem. Soc., 1996, 118, 1769–1776.
14. A. M. Appel, J. E. Bercaw, A. B. Bocarsly, H. Doddek, D. L. DuBois, M. Dupuis, J. G. Ferry, E. Fujita, R. Hille, P. J. A. Kenis, C. A. Kerfeld, R. H. Morris, C. H. F. Peden, A. R. Portis, S. W. Ragsdale, T. B. Rauchfuss, J. N. H. Reek, L. C. Seefeldt, R. K. Thauer and G. L. Waldrop, Chem. Rev., 2013, 113, 6621–6658.
15. J. L. Inglis, B. J. MacLean, M. T. Pryce and J. G. Vos, Coord. Chem. Rev., 2012, 256, 2571–2600.
16. H. Ishida, K. Tanaka and T. Tanaka, Organometallics, 1987, 6, 181–186.
17. D. S. Laitar, P. Muller and J. P. Sadighi, J. Am. Chem. Soc., 2005, 127, 17196–17197.
18. M. Garcia, M. J. Aguirre, G. Canzi, C. P. Kubiak, M. Ohlbaum and M. Isaacs, Electrochim. Acta, 2014, 115, 146–154.
19. N. Furuya and K. Matsui, J. Electroanal. Chem. Interfacial Electrochem., 1989, 271, 181–191.
20. T. V. Magdesieva, I. V. Zhukov, D. N. Kravchuk, O. A. Semenikhin, L. G. Tomilova and K. P. Butin, Russ. Chem. Bull., Int. Ed., 2002, 51, 805–812.
21. P. G. Cozzi, Chem. Soc. Rev., 2004, 33, 410–421.
22. S. Skljarevski, A. A. Peverly and D. G. Peters, J. Electroanal. Chem., 2011, 661, 39–43.
23. D. J. Darensbourg, R. M. Mackiewicz, A. L. Phelps and D. R. Billodeaux, Acc. Chem. Res., 2004, 37, 836–844.
24. C. T. Cohen and G. W. Coates, J. Polym. Sci., Part A: Polym. Chem., 2006, 44, 5182–5191.
25. S. Gambarotta, F. Arena, C. Floriani and F. Zanazzi, J. Am. Chem. Soc., 1982, 104, 5082–5092.
26. Y. Hori, K. Kikuchi, A. Murata and S. Suzuki, Chem. Lett., 1986, 897–898.
27. Y. Hori, A. Murata and R. Takahashi, J. Chem. Soc., Faraday Trans. 1, 1989, 85, 2309–2326.
28. Y. Hori, H. Wakabe, T. Tsukamoto and O. Koga, Electrochim. Acta, 1994, 39, 1833–1839.
29. L. M. Aeshala, R. G. Uppaluri and A. Verma, Journal of CO₂ Utilization, 2013, 3–4, 49–55.
30. L. Barbora, S. Acharya and A. Verma, Macromol. Symp., 2009, 277, 177–189.
31. L. M. Aeshala, S. U. Rahman and A. Verma, Proceedings of the ASME 2011 5th International Conference on Energy Sustainability, Washington, D.C., USA, 2011.
32. L. M. Aeshala, S. U. Rahman and A. Verma, Sep. Purif. Technol., 2012, 94, 131–137.
33. Y. Hori, in Handbook of Fuel Cells, ed. W. Vielstich, John Wiley & Sons, England, 2003, vol. 2, pp. 720–732.
34. S. Singh, C. Mukherjee and A. Verma, Clean Technol. Environ. Policy, 2014 DOI:10.1007/s10098-014-0796-6.
35. B. P. Sullivan, Electrochemical and electrocatalytic reactions of carbon dioxide, Elsevier, Amsterdam, 1993.
36. J. Costamagna, G. Ferraudi, J. Canales and J. Vargas, Coord. Chem. Rev., 1996, 148, 221–248.
37. K. J. P. Schouten, Y. Kwon, C. J. M. Ham, Z. Qin and M. T. M. Koper, Chem. Sci., 2011, 2, 1902–1909.
38. L. M. Aeshala, R. G. Uppaluri and A. Verma, Phys. Chem. Chem. Phys., 2014, 16, 17588–17594.

---