3.2.2.1  Basic Concept

The overpotential is the difference between the applied electrode potential, E_applied, and E⁰ (products/substrates), at a given current density, reflecting the degree of deviation of the electrode potential from its thermodynamic equilibrium value from the Nernst equation. In other words, an overpotential is the extra amount of nonreversible energy required to overcome the energy barrier which allows the reaction to occur. It shows how far away the electrode is from its equilibrium.

The electrode overpotential mainly consists of three components: concentration overpotential, η_con; activation overpotential, η_act; and ohmic overpotential, η_ohmic, expressed as follows:

where η is the overall overpotential. Other additional overpotentials may exist, such as the chemical reaction overpotential. The cell overpotential cannot be eliminated but can be minimized by proper material modification and good cell design. Furthermore, cell working atmosphere, such as pressure and temperature, can also influence the overpotential.

3.2.2.2  Butler–Volmer Equation

As discussed in Section 3.2.1, for an electrochemical reaction with multiple steps involving one electron without any other chemical step, the activation overpotential can be correlated to current by the Butler–Volmer equation (3.13).

The transfer coefficients are considered as the fraction of the change in polarization, which leads to a change in the reaction rate constant. The transfer coefficients can be given by the Bockris model as [28]

where α_a and α_c are the anodic and cathodic transfer coefficients, respectively. γ_c is the number of charges transferred ahead of the r.d.s., υ is the total number of times of repeating the r.d.s. reaction to complete the reaction, r is the number of charges transferred in the r.d.s., and β is the symmetric coefficient (usually 0.5). Apparently, the transfer coefficients have the following relationship:

The exchange current density is a function of reaction rate constants, the concentration of the anodic and cathodic reactants, etc. It is related to the balanced forward and backward electrode reaction rates at equilibrium. A high exchange current density leads to a fast electrochemical reaction, thus good cell performance is anticipated.

A typical Butler–Volmer equation at large activation overpotential is plotted in Figure 3.6 when the exchange current density is 0.001 mA/cm² and the anodic and cathodic transfer coefficients are unity. The plot shows a high linearity, indicating the Tafel behavior. The plot is symmetric due to the same anodic and cathodic transfer coefficients. Otherwise, it should be asymmetric.

3.2.2.3  Overpotential of CO2 Electroreduction

Direct ERC on most electrode surfaces requires impractically high overpotential which is necessary to drive the process and consequently lowers the conversion efficiency [29].

The standard potential of the reduction of CO₂ to different products as shown in Table 3.1, calculated from the thermodynamic data, is much more positive than that the reduction actually takes place at as shown in Table 3.2. Such high overpotential is presumed from involvement of intermediate species, $\cdot {\text{CO}}_2^{-}$

anion radical, which requires highly negative potential for the formation. The high overpotential depends on electrode metal, crystal orientation, and reduction product alongside proton availability. CO formation takes place with relatively lower overpotentials than HCOO⁻ formation. Stabilization of $\cdot {\text{CO}}_2^{-}$ adsorption on electrodes may lead to reduction of the overpotential. Among all the electrodes, Au electrode reduces CO₂ to CO at remarkably low cathodic potential, −1.14 V at 5 mA cm⁻² (Table 3.2). It is suggested that the formed intermediate species $\cdot {\text{CO}}_2^{-}$ on the Au electrode was greatly stabilized by adsorption, leading to decrease of overpotential [30]. Coordination of the CO₂ to M can activate $\cdot {\text{CO}}_2^{-}$ and promote its reductive disproportionation to form $\text{C}{\text{O}}_{\text{3}}^{2-}$

Clearly, in order to minimize overpotentials, catalysts need to be developed that have formal potentials E⁰(Cat^n+/0) well matched to E⁰(products/CO₂), and appreciable rate constants for the electroreduction of CO₂ to appropriate products at this potential. In addition, the heterogeneous rate constant for reduction of the electrocatalyst at the electrode must be high for V_applied near E⁰(Cat^n+/0).

Transition metal electrodes have been reported to dramatically reduce the overpotential of reduction to at least 500 mV anodic of E⁰ for CO₂ → $\cdot {\text{CO}}_2^{-}$

. Bruce et al. [32] find E⁰ = +0.1 V vs. SHE for CO₂ + 2H⁺ + 2e⁻ → HCOOH on copper, which is much more positive than that calculated from Gibbs energy of formation.

Fundamental studies will contribute to enhancement of our understanding of the overpotential of the electrodes. A wide range of Tafel slopes for CO₂ reduction from 120 mV decade⁻¹ to 350 mV decade⁻¹ has been reported [5,16,33,34]. These slopes indicate a rate-determining initial electron transfer to CO₂ forming a surface adsorbed $\cdot {\text{CO}}_2^{-}$

intermediate. The calculated exchange current density for the formation of $\cdot {\text{CO}}_2^{-}$ radical at 100 kPa and 298 K in CO₂-saturated aqueous solution (~0.036 M) is 4.8 × 10⁻⁷ kA m⁻² [35]. The exchange current densities for ERC to formate at 293 K in 0.95 M KCl + 0.05 M NaHCO₃ on In, Hg, and Sn are reported, respectively, as 1 × 10⁻⁷, 1 × 10⁻¹⁰, 1 × 10⁻⁸ kA m⁻² [6].

Gold electrodes yield CO in CO₂ reduction at −0.8 V vs. SHE with a low overpotential in 0.5 M KHCO₃ [30]. The obtained relationship between the potential and logarithm of the partial current of CO formation (i_c) (Figure 3.7) showed that the Tafel slope is approximately 130 mV decade⁻¹ in the lower current region, where the transport process of CO₂ does not interfere with the supply of CO₂ to the electrode, the Tafel slope corresponds to the transfer coefficient 0.46. i_c at a constant potential is linear to the CO₂ partial pressure. These data suggest that the CO₂ reduction proceeds in the first order with respect to CO₂, and the r.d.s. of the reaction is the first electron transfer to CO₂. These facts support the reaction scheme discussed in the following. In phosphate buffer solutions of pH 2–6.8 at an Au electrode, i_c is proportional to the pressure of CO₂ [36]. The Tafel slope is ca. 120 mV decade⁻¹. The potential partial current relation obtained with electrolytes at pH below 4.3 agreed well with that obtained with 0.5 M KHCO₃ at pH 7.5 after the double-layer correction due to the difference of the electrolyte concentration [30]. This fact shows that CO formation at Au electrode does not depend on pH of the electrolyte, and that the proton donor is not H⁺ but H₂O molecules.

The intrinsic electrocatalytical activity of nanoporous silver (np-Ag) for CO₂ reduction to CO was investigated through Tafel analysis in a CO₂-saturated 0.5 M KHCO₃, compared with polycrystalline Ag catalyst, as shown in Figure 3.8 [37] As observed, the np-Ag was more preferred for CO₂ reduction with a Tafel slope of 58 mV decade⁻¹, much smaller than that of 132 mV decade⁻¹ for polycrystalline Ag. At a overpotential higher than 0.25 V, the slope increased sharply, indicating another rate-limiting step most probably relating to the issues of the diffusion of reactants and products out of/into the catalyst nanopores. Mechanistically, there are two steps in the two-electron reduction of CO₂ to CO on an Ag surface [38]. First, one electron is transferred to a CO₂ molecule, forming a anion intermediate absorbed on the catalyst surface. Subsequently, the CO₂⁻ anion takes another electron and two protons, forming a molecule of CO and a molecule of H₂O. As reported [39], the first step at a much more negative potential compared with that of the following step is rate determining for the whole process with calculated Tafel slope of 120 mV decade⁻¹. This is very similar to that of 132 mV decade⁻¹ observed in polycrystalline Ag. For np-Ag, the sharp decrease of Tafel slope at 58 mV decade⁻¹ presents a fast step of initial electron transfer before a r.d.s. of later nonelectron transfer [39–41], indicating that the np-Ag with possibly higher index facets supported by the highly curved surface [42,[43] are able to stabilize the anion $\cdot {\text{CO}}_2^{-}$

intermediate much better than a flat one.

Results of studies on the kinetics with a copper foil electrode at 273 and 295 K using two separate electrode pretreatments to remove the oxide layer, that is, 10 wt% HCl and 10 wt% HNO₃ [44], is listed in Table 3.3.

The kinetic information/data on ERC is sparse and difficult to summarize because of the variety of approaches and experimental conditions. It has to be noted that the reaction kinetics depend on many interacting factors so results obtained by one researcher under his conditions might not apply under other conditions.

3.4.4.1  Metal Electrodes

Although there is no general consensus on the mechanism for hydrogenation at Cu/ZnO catalysts, Cu(I) sites are thought to promote catalytic activity and selectivity toward CH₃OH. Although Cu–Zn alloys are considered active sites for CO₂ reduction, Cu(I) sites are considered key species for CO adsorption in hydrogenation reactions [68]. Further, Cu(I) sites are believed to stabilize reaction intermediates such as carbonates (CO₃²⁻-), formates (HCOO⁻-), and methoxy adsorbates (H₃CO⁻-) due to their higher heats of adsorption [69]. Sheffer et al. showed that alkali metals help stabilize Cu(I) active sites and increase Cu(I) concentrations which significantly improve CH₃OH yield [70]. Oxidation studies at single crystal Cu(I) shows that H₃CO adsorbed at (111) surfaces with Cu(I) atoms at the second atomic layer allows coordinately unsaturated oxygen anions to act as hydrogen abstraction sites for dehydrogenation [71]. In this case, it is possible that the unsaturated oxygen atoms at the (111) surfaces of the cuprous oxide film act as hydrogen donors sites in the reduction reaction.

A theoretical report by Peterson et al. describes a pathway for the electrochemical reduction of CO₂ to CH₄ at Cu electrodes based on density functional theory (DFT) and computational hydrogen electrode (CHE) models applied to Hori’s experimental data [72]. In that work, the authors indicate that the carbon atom of CO adsorbates may be hydrogenated via proton transfer to form HCO at −0.74 V (RHE). Hydrogenation of an adsorbed CO species is proposed to occur directly via proton addition from solution as their availability is significantly greater than hydrogenation from adsorbed hydrogen. Accordingly, once the HCO species is formed, the carbon atom continues proton and electron transfer reactions to form H₃CO adsorbates. Following this pathway, the last proton transfer to the H₃CO species (on Cu(211) surfaces) favors CH₄ formation by 0.27 eV. In the case of cuprous oxide electrodes as described here, the reduction reaction may benefit from both improved intermediate stability and the ability of H⁺ species coordinated with surface-bound oxygen. This surface would allow hydrogen addition to the oxygen atom of the H₃CO adsorbate rather than the carbon atom as shown in Figure 3.12 [73].

The formation of HCO⁻ intermediate has been reported as the r.d.s. in CO₂ reduction to CH₄, and H₂CO⁻ intermediate is detected in both hydrogenation and photoelectrocatalytic process for CH₃OH formation [72,[74]. Two possible mechanism pathways (a and b) for the direct electrochemical reduction of CO₂ to CH₃OH at Cu/ZnO catalyst are proposed in Figure 3.13 [73]. Although mechanism (a) proceeds through CO pathway, mechanism (b) undergoes formate (HCOO⁻) intermediate. The last four steps in both (a) and (b) mechanism pathways are the same with HCO adsorbate formation, H₂CO adsorbate formation, H₃CO adsorbate formation, and hydrogenation of H₃CO adsorbate to CH₃OH.

Pathway (a) proceeds via CO intermediate, similar to Peterson and Gattrell’s mechanisms for CO₂ reduction at Cu surfaces [72,[75]. The first electron and proton transfer may be associated with the formation of dioxymethylene (HOCO⁻). When another electron and proton are added to the OH group of the HOCO⁻ adsorbate, the C−OH₂ bond is broken to desorb a H₂O molecule and leave CO adsorbate on the surface. The reaction continues with the formation of HCO species by H addition to the C atom. Once the HCO species are formed, the carbon atom continues proton and electron transfer reactions to form H₂CO and H₃CO adsorbates. Although it is not clear whether these adsorbates would attach to the Cu, Zn, or O bond, ZnO (10–10) surface would promote CH₃OH formation by drawing OH adsorption and allowing the H atom to attach to the O atom of the H₃CO species.

Pathway (b) proceeds via HCOO⁻ intermediate [74]. The H attaches to the C atom instead of the O atom on CO₂ adsorbate, hence, forming HCOO⁻ rather than HOCO⁻. The mechanism proceeds with H additions to the O atom of the formate (HCOO) species first to form HCO after an H₂O molecule dissociates from the surface. Similar to mechanism (a), once HCO species are formed, the reaction continues electron and proton transfer until CH₃OH is desorbed from the surface.

In Hg, Tl, Pb, In, Sn, and Cd at aqueous solution, the formation of HCOO⁻ may be initiated by one electron transfer to CO₂ at the potential negative of −1.6 V vs. SHE, forming $\cdot {\text{CO}}_2^{-}$

present mostly freely in the solution close to the electrode as depicted in Figure 3.14 (1). The hydrogen bond by water molecules [76] and the high dielectric constant of water molecule contribute to the stabilization of $\cdot {\text{CO}}_2^{-}$ in solution. The concentration of $\cdot {\text{CO}}_2^{-}$ would be ca. 10⁻⁵ mol dm⁻³, if one takes into account the standard potential −1.85 V or −1.90 V and the Nernst relation 59 mV decade⁻¹ at 25°C. Free $\cdot {\text{CO}}_2^{-}$ will take a proton at the nucleophilic carbon atom from a H₂O molecule acting as a Lewis acid, forming HCO₂. H⁺ will not be bonded to the O atom of because pKa value of the acid–base couple ($\cdot {\text{CO}}_2^{-}$ /CO₂H) is low at 1.4 [77]. HCO₂·; is subsequently reduced to HCOO⁻ at the electrode in aqueous media. The electrode potentials of Pb, Cd, and Tl are −1.6 V or more negative at 5.0 mA cm⁻² [31]. The potential of Hg is −1.51 V at 0.5 mA cm⁻², which would be equivalent to −1.7 V at 5.0 mA cm⁻² with an assumption of the transfer coefficient for HCOO⁻ formation 0.25 [5,9,65]. In nonaqueous media, H₂O molecule is not available and plenty of CO₂ is present. A CO₂ molecule will play a role of a Lewis acid with the nucleophilic C of $\cdot {\text{CO}}_2^{-}$ . The coupling of $\cdot {\text{CO}}_2^{-}$ and CO₂ will lead to the formation of an adduct intermediate ·(CO₂)₂⁻, thus producing oxalate [76], which is more probable than Savéant’s coupling mechanism [78]. By contrast, CO is produced by a mechanism different from HCOO⁻ formation which may proceed with free $\cdot {\text{CO}}_2^{-}$ intervening. CO is favorably produced from the electrode metals which stabilize $\cdot {\text{CO}}_2^{-}$ effectively.

On the metal electrode surface of Au, Ag, Zn, Pd, Ga, or Cu, giving CO as the major product, $\cdot {\text{CO}}_2^{-}$

is stabilized by adsorption both in aqueous and nonaqueous electrolytes. Electrophilic reagents, H₂O in aqueous solution, react with the O atom of adsorbed $\cdot {\text{CO}}_2^{-}$ , forming CO(ad) and OH⁻ as depicted in Figure 3.14 (2). H⁺ will not take part in the CO formation as the partial current of CO formation is independent of pH. CO(ad) is readily desorbed from the electrode as a gaseous molecule. The sequence of CO selectivity roughly agrees with that of the electrode potentials shown in Table 3.2. In nonaqueous media, a CO₂ molecule reacts as a Lewis acid with adsorbed $\cdot {\text{CO}}_2^{-}$ , and allows a C−O bond of the $\cdot {\text{CO}}_2^{-}$ to be broken. This process forms CO(ad) and $\cdot {\text{CO}}_2^{-}$ as postulated by Hammouche and his coworkers [79] for electrochemical reduction of CO₂ catalyzed by iron porphyrins. CO(ad) thus formed is easily desorbed, as shown in Figure 3.14 (2). Cd, Sn, and In, of medium CO selectivity, do not strongly adsorb $\cdot {\text{CO}}_2^{-}$ .$\cdot {\text{CO}}_2^{-}$ will be most freely present in aqueous electrolyte owing to the stabilization by water molecules over hydrogen bond [76] and the high dielectric constant.$\cdot {\text{CO}}_2^{-}$ stabilized in the electrolyte will be further reduced to HCOO⁻. However, $\cdot {\text{CO}}_2^{-}$ is not sufficiently stabilized in nonaqueous electrolyte due to lack of hydrogen bond formation and low dielectric constant of the solvents. Thus, $\cdot {\text{CO}}_2^{-}$ adsorbed on Cd, Sn, and In may be relatively more stable than $\cdot {\text{CO}}_2^{-}$ dissolved in the electrolyte. These metals yield CO in nonaqueous media in the same manner as Au, Ag, and Zn in Figure 3.14 (2). The CO selectivity mentioned above will be closely connected to the stability of adsorbed $\cdot {\text{CO}}_2^{-}$ on the electrode.

In H₂O–DMF solutions, the electrochemical reduction of CO₂ at Pb and Hg give mainly oxalate or formate, depending on the concentration of H₂O, with scarce formation of CO at relatively low yield [80–82]. It is proposed that an electron transfer to CO₂ molecule initiates the process, forming $\cdot {\text{CO}}_2^{-}$

, which is not adsorbed on Pb, Hg, and Tl, and is freely present in nonaqueous electrolyte solution as well. And the interactions between the electrode and the reactants, intermediates, and products are negligible as the CO₂ reduction at Hg and Pb electrodes proceeds at potentials close to the standard potential of$\cdot {\text{CO}}_2^{-}$ /CO₂ couple. Several competing homogeneous reactions sequentially take place in parallel in the electrolyte solution (Figure 3.15). The product distribution at Pb and Hg electrodes is determined by the current density and the concentration of CO₂ and H₂O [80]. Oxalate is formed by coupling of $\cdot {\text{CO}}_2^{-}$ , Figure 3.15 (1), formate is formed in a way similar to that in aqueous electrolyte, Figure 3.15 (5)(5′), and CO is formed by sequential reactions (2) and (3) or (3′) in Figure 3.15.

3.4.4.2  Molecular Electrocatalysts

In the proposed mechanism of molecular catalyst of [Ni^IIcyclam]²⁺ complex (structure 1, Figure 3.16), the [Ni^Icyclam]⁺ complex is assumed to play a crucial role as shown in Figure 3.17, and in particular, a [Ni^Icyclam(CO)]⁺ intermediate species has been detected in the course of selective electroreduction of CO₂ to CO in aqueous solution. Fourteen-membered cyclam framework is a crucial prerequisite for the encircled nickel center to act as catalyst in the CO₂ electroreduction. The planar ligand geometry of the Ni(cyclam), allowing access to the metal center, has been suggested to be of significance with respect to the catalytic ability exhibited. The system favors the formation of the Ni^I state [83,[84]. [Ni(II)-cyclam] is reportedly a very selective electrocatalyst for CO₂ reduction to CO and was found to catalyze carbon dioxide reduction on mercury electrodes at ca. 0.5 V below the calculated thermodynamic value (E⁰ = −0.41 V vs. NHE at pH 5) at a “remarkable” velocity of ca. 32 h⁻¹ with 95% CO formation CE. The catalyst was reported to cycle more than 1000 times with no notable deactivation. The system was found to be pH dependent, anion specific, prone to deactivation through CO buildup, resulting in Ni(cyclam)(CO) precipitation if the solution is not stirred and reported to require an Hg surface for appreciable rates and turnover [85,86].

As being developed to date, the most efficient catalyst of modified Fe-porphyrin [87] with phenolic groups in all ortho and ortho’ positions on the phenyl groups [88] as depicted in Figure 3.18 was reported to give CO in Faradaic yields >90% through 5 × 10⁷ turnovers over 4 h at an overpotential of just 0.465 V in DMF with 0.2 M water added causing no observable degradation. The enhancement in activity was attributed to the higher local concentration of protons associated with the phenolic hydroxyl groups.

For [Pd(triphosphine)(CH₃CN)]²⁺ catalysts (structure 2, Figure 3.16), the dissociation of a weakly coordinating solvent molecule of acetonitrile favors M–O bond formation during the electrocatalytic reduction cycle of CO₂ to CO. In addition, electron-donating substituents on the triphosphine ligand result in increased catalytic rates. The rate constant is between 5 and 300 M⁻¹ s⁻¹ [89]. The detail catalytic cycle is proposed as shown in Figure 3.19. As the Pd(II) complex gained an electron to form a Pd(I) intermediate, a reaction with CO₂ occurred to form a five-coordinate CO₂ adduct. Upon transfer of the second electron, the Pd(0) intermediate dissociated the solvent. Protonation of one of the oxygen atoms of coordinated CO₂ affords a metallocarboxylic acid intermediate, Pd–COOH. It is believed that the metallocarboxylic acid is protonated again to form a “dihydroxy carbene,” and that CO is then formed by dehydration of the dihydroxy carbene. CO dissociation and solvent association regenerates the initial complex. In solutions of high acid concentration, the r.d.s. was found to be the reaction of the Pd(I) intermediate with CO₂, though, in solutions of low acid concentrations, the cleavage of the C–O bond to form carbon monoxide and water limits the rate of the catalytic cycle. These classes of Pd phosphine complexes have shown catalytic rates in the range of 10–300 M⁻¹ s⁻¹ and with >90% CEs for CO production. Overpotentials were in the range of 100–300 mV, yet turnover numbers were low (ca. 10–100) and the decomposition to Pd(I) dimers and hydrides eventually causes cessation of catalytic activity [85,90,91].

Molecular CO₂ reduction catalyst of binuclear palladium complex (structure 3, Figure 3.16), in which two or more independent Pd triphosphine units Pd(P₃) are incorporated and separated by a methylene spacer, binds CO₂ through two Pd sites, with one metal interacting with C, the other with O. This complex palladium catalysts show very high catalytic rates, increasing by three orders of magnitude, for CO₂ reduction (>104 M⁻¹ s⁻¹), but with a turnover number of ca. 10. Along with this increase in catalytic activity came an increase in the formation of Pd(I)–Pd(I) bonds, thus decreasing catalyst lifetimes. And when the bridge spacer is not methylene, the dendrimers of the Pd catalyst showed decreased activity and selectivity [91].

Studies on the CO₂ reduction to CO by the Re(bpy-R)(CO)₃X system (R = H, Me, tBu, X = halogen, OTf, or a neutral ligand with a noncoordinating anion) [93] indicated that the complex undergoes a two-electron reduction and loses the labile X ligand to form an anionic tricarbonyl species (A), reported to be the active form of the catalyst, best described as a Re⁰(bpy–R)⁻¹ species [94]. The subsequent mechanistic steps are depicted in Figure 3.20. An intermediate of carboxylate species (B) is formed upon reaction of carbon dioxide with the anion (A). The carboxylate species is protonated to form a carboxylic acid species (C), which could then react with a second proton to liberate water and produce a tetracarbonyl cationic intermediate (D). Although the exact order of addition of protons, addition of electrons, and loss of product remain undetermined, the ligand loss and subsequent transfer of electron density to the metal center have been reported as essential steps in the process [93]. The intermediate species involved in catalysis will likely depend on reduction potentials and pKa’s of the complexes compared with the applied potentials and pKa’s of the Brönsted acids available in solution. TOFs of the bpy-tBu system greater than 250 s⁻¹ have been reported [93]. The increasing TOF, compared with that of relatively low value of 21.4 h⁻¹ for Re(bpy)(CO)₃Cl in DMF–H₂O (9:1) solution, forming CO at −1.49 V vs SCE with CEs at 98% and excellent selectivity over H₂ production [95], attributed to the inclusion of tertiary butyl groups at the 4,4′ positions of the bipyridyl ligand [93] and protic additives to the system [96]. To date, it is one of the best and most well-studied electrocatalyst system of CO₂, still underway almost 30 years on [92].

The bipyridine complexes of ruthenium, Ru(bipy)(CO)₂²⁺, were found to electrocatalytically reduce CO₂ to CO and HCOO⁻ at −1.40 V vs. SCE, when water was present [97], by a two-electron reduction process as shown in Figure 3.21 [85,[97]. Following the production of CO, a five-coordinate neutral complex is formed. In the presence of CO₂, the complex forms an η¹-CO₂ adduct of Ru(0), which can also be formed by addition of two equivalents of OH⁻ to Ru(bipy)(CO)₂²⁺. Addition of a proton forms the LRu(CO)(COOH) species which under acidic conditions (pH 6.0) gains another proton to lose water and regenerate the catalyst. Under basic conditions (pH 9.5), the catalyst may undergo a two-electron reduction with the participation of a proton to create HCOO⁻ and regenerate the five-coordinate Ru(0) complex. The insertion and two-electron attack rely on interactions between the reduced complexes and CO₂ as the electrophile. In this case, both are activated by the occupation of the π* orbitals of bipyridine. The ligand-based reduction exerts a strong influence on reactivity as electron density at the metal center and in the metal hydride bond is reported as increased making them more susceptible to CO₂ attack. The system is helpful to elucidate several of the key intermediates in the reduction of CO₂, though with low turnover numbers and low selectivity.

cis[Os(bpy)₂(CO)H][PF₆] reported by Bruce et al. [32] leads to associative activation of CO₂ in nonaqueous solution. At a platinum electrode, the reduction took place at −1.4 to −1.6 V (vs. a NaCl-saturated calomel electrode). The proposed mechanism is illustrated in Figure 3.22. The complex accepts two electrons to give the active species which reacts with CO₂, binding as an extra ligand, leading to coordination sphere expansion. There are different routes proposed where either the initial interaction with the catalyst results in the disproportionation or two of the direduced molecules may combine to give an uncharged complex and disproportionation products, CO and carbonate. In the experiments where water was present, for example, including trace water of less than 10 mM, they reported the incorporation of a proton into the intermediate direduced species, followed by combination to give formate or CO and hydroxide. The complex structure was investigated with the H found cis to the carbonyl being swapped for ligands of different sizes, H > CH₃ > CH₂–C₆H₅ > C₆H₅. It was seen that the rate constant for the reaction decreased with increasing size of this ligand. The trans structure was also trialed as a catalyst but the less sterically demanding configuration was less active. This highlighted the impact of relatively small differences in the molecular structure of the catalyst with a pronounced effect upon catalytic ability seen.

A dinuclear copper(I) complex incorporating pyridyl-like groups was reported to activate and convert CO₂ selectively into oxalate at only a very small applied potential of −0.03 V vs NHE, via a very different and effective catalytic cycle as seen in Figure 3.23. The electrocatalytic reduction of the copper(II) ion to copper(I) appears to be rate limiting, which occurs at a readily accessible potential, around 2 V less negative than that required for CO₂ reduction to $\cdot {\text{CO}}_2^{-}$

in acetonitrile on a glassy carbon electrode [98]. Two dinuclear direduced copper complexes are thought to form a tetranuclear complex with two bridging CO₂-derived oxalates which are subsequently released as Li₂C₂O₄ precipitate. The reduction of CO₂ in this system is found at a significantly less negative potential than that of dioxygen allowing the selective reduction of CO₂ from air. Unfortunately, this catalyst only completed six turnovers with 12 equivalents of oxalate formed in 7 h. So it cannot be considered suitable as a realistic catalyst, yet however is the most efficient route to oxalate in terms of overpotential reported to date [87].

Nonmetal nitrogen-based pyridinium cation (structure 4, Figure 3.16) catalyst [99,[100] acts as one-electron shuttle to perform multiple-electron, multiple-proton reduction of CO₂ effectively in aqueous solution with Faradaic efficiencies of methanol observed ~30% at overpotentials of only −0.2 V on metal electrode. When this pyridinium ion catalyst was anchored to a p-GaP electrode, this photochemical system yielded nearly 100% Faradaic efficiency of methanol at potential 0.3 V below the thermodynamic potential of −0.52 V vs SCE for the reaction at pH 5.2 [101]. The detailed mechanism of pyridinium-catalyzed CO₂ reduction in aqueous solution was proposed [99] as given in Figure 3.24. The 6e-reduced product of methanol was through a pathway of a series of single-electron transfer rather than the typical multielectron charge transfer. At metal electrodes, formic acid and formaldehyde were observed to be intermediate products with the pyridinium radical playing a role in the reduction of both intermediate products. The r.d.s. of the aqueous system was reportedly determined as being the initial step, formation of the pyridinum–CO₂ radical.

Large NR₄⁺ cations reportedly shift CO₂ reduction to more negative potentials and decrease the associated rate constant as their size increases. This is hypothesized to be the blocking effect of the compact layer of adsorbed cations, perhaps linked to the catalysis by tetraalkylammonium cations on CdTe in DMF and acetonitrile [102,[103]. The catalysis was most greatly enhanced by the shorter alkyl chain length with NH₄⁺ optimal. The supposed mechanism through which the enhancement was obtained was given as [103]:

1. NR₄⁺ + e⁻ → NR₄^·(ads)
2. NR₄^·(ads) + CO₂(ads) → NR₄⁺ + $\cdot {\text{CO}}_2^{-}$
   (ads)
3. $\cdot {\text{CO}}_2^{-}$
   (ads) + H⁺ + e⁻ → CO(ads) + OH⁻
4. OH⁻ + H⁺ → H₂O

Upon addition of crown ether catalysts with ring structures in the solution, the tetraalkylammonium species, best able to fit within the cavity, yielded higher catalytic activity. The additional lowering of the overpotential in this case was thought to be a result of a similar mechanism as mentioned above, with the crown ether allowing the ammonium ion to get closer to the electrode surface, which is of specific adsorption in the inner Helmholtz plane, rather than outer Helmholtz, increasing the rate of electron transfer to the NR₄⁺ species [103].

3.4.4.3  Electrocatalytic Activity Degradation

The reduction of CO₂ at Cu is inefficient, occurring at high overpotentials toward most probable products of hydrocarbons, and the activity drops rapidly as well [4,104–107]. The proposed mechanisms for the activity degradation include adsorption of organic intermediate [105,[106], black carbon deposition [105,[108], poison of copper oxide patina [107], accumulation of an unknown low vapor pressure and soluble CO₂ reduction product [109], and contaminants Fe²⁺ and Zn²⁺ deposition [4]. However, the real reason behind this is still unclear. The formation mechanism [110] of hydrocarbons [111] is still of a great challenge, though a good start of simulating elements of the reaction mechanism on Cu using DFT [72,112,113]. It is worth noting that Cu electrodes containing a Cu₂O surface layer demonstrated high activity toward CH₃OH production [114,[115]. Cu(I) species was considered to play a critical role in the activity of CH₃OH formation. CH₃OH yielded qualitatively follow Cu(I) concentrations.