New Platinum Complexes from Salen- and Hydroxy-Substituted Salpn-Naphthalene Ligands with CO₂ Reduction Activity

Abstract

The electrocatalytic reduction of carbon dioxide (CO₂) into added-value products is a promising alternative to completing the cycle of atmospheric CO₂. We report two new platinum complexes—a salen-like naphthalene (PtL1) and a hydroxy-substituted salpn naphthalene (PtL2)—that are capable of activating CO₂ to produce carbon monoxide (CO). The predominant keto tautomer of the non-innocent ligands was determined using DFT calculations and UV-Vis spectroscopy. The PtL2 complex has a CO Faradaic efficiency >40% in the presence of water as a sacrificial proton source at −2.5 V vs. Fc/Fc⁺. The addition of the hydroxy group in combination with water as a proton source decreased the reduction potential and increased the CO formation tenfold when compared to PtL1.

1. Introduction

The lack of sustainable energy sources has resulted in a significant dependence on fossil fuels as the primary global energy source [1]. This has consequently led to an increase in the production of anthropogenic carbon dioxide (CO₂), which is associated with global warming and climate change [2]. The electrocatalytic reduction of CO₂ is an attractive alternative for the production of value-added products while decreasing the excessive CO₂ concentrations in the atmosphere [3,4,5]. The high stability and the relative inertness of CO₂ molecules present a challenge for the electrochemical CO₂ reduction reaction (CO₂RR) [6,7]. Coordination complexes such as those synthesized from salen-like ligands provide a means to overcome the high stability of CO₂ by employing careful ligand design strategies to enhance selectivity, catalytic activity, and overpotential for the CO₂RR [7,8,9].

Salen ligands are well known for the coordination of metals ions by acting as N, N, O, O tetradentate Schiff bases (SB), and are readily obtained by means of a condensation reaction between a diamine and a hydroxy-substituted aryl ketone or aldehyde. Salen and salen-like complexes have been less extensively studied for electrochemical CO₂ reduction compared to their counterparts, porphyrins and phthalocyanines [3,10,11,12,13,14,15,16,17,18,19,20]. In addition, salen-like complexes offer advantages in terms of ease of synthesis and relatively short reaction times [21,22]. Ligand substitution on salen and salpn systems can assist in CO₂ reduction activity and selectivity, as confirmed in previous studies by Nichols et al. [9] and Bose et al. [17]. In addition, articles using salen complexes as electrocatalysts for CO₂RR have focused on 3d metal centers such as Co, Ni, and Cu in order to reduce the cost of the catalysts [4,16,17,18]. Nonetheless, these provide a mixture of products and may follow reaction pathways that compete with the Hydrogen Evolution Reaction (HER). Precious metals are known to have exceptional catalytic activity in a wide range of reactions such as metal-catalyzed coupling, water oxidation, and olefin metathesis reactions, among others [21,22,23,24,25]. Even though 3d metal complexes provide low-cost materials, the synthetic procedure and multiple purification steps taken to reach the desired metal complex increase the gross cost. Usually, noble metals have greater electrocatalytic activity; therefore, we set ourselves the task of developing Pt salen-like complexes as electrocatalysts for CO₂RR.

Herein, we present a previously reported salen naphthalene ligand (L1) and a newly synthesized hydroxy-substituted salen-like ligand (L2), and their respective Pt complexes (PtL1 and PtL2) [26]. The reported L1 and L2 ligands have been confirmed to exist as an equilibrium of tautomers, wherein the keto tautomer is the main contributor to the solid-state bulk product [26]. Through UV-Vis and DFT calculations, it was confirmed that the predominant structure is the keto tautomer in both L1 and L2 in liquid state. The synthesis pathway for the ligands and complexes is shown in Figure 1. Structural analysis was performed by means of single-crystal X-ray diffraction (SCXRD) experiments, from which we observed the square planar geometry of the PtL1 and a more distorted configuration for PtL2 due to having less angle strain in the imine moieties. Finally, the compounds were electrochemically characterized by cyclic voltammetry and studied with respect to their catalytic activity towards CO₂RR. The hydroxy-group-substituted PtL2 complex has a Faradaic Efficiency (FE) above 40% for the formation of carbon monoxide (CO). In contrast, the PtL1 complex has an FE of ~5%, demonstrating the influence of the hydroxy substituent on the diamine bridge of the metal complexes. To the best of our knowledge, these are the first SB Pt complexes that have been reported to display catalytic activity towards the electroreduction of CO₂. Mechanistic studies by means of DFT calculation were performed to elucidate the mechanism of CO formation from CO₂ using PtL1 and PtL2. It was found that the PtL2-*COOH intermediate was of lower energy than in the case of PtL1; nonetheless, the CO release for PtL2 displayed a higher Gibbs free energy due to its high degree of hydrogen bonding.

2. Results and Discussion

2.1. Spectroscopic Properties

The UV-Vis spectra of the reported complexes L1, L2, PtL1, and PtL2 were recorded in CH₂Cl₂ at a concentration of 0.3 mM (Figures S6–S9). The first ligand, L1, had three peaks located at 325 nm, 416 nm, and 434 nm. The first absorption peak at 325 nm of L1 is attributable to a π-π* intra-ligand transition of the naphtholate chromophore [1,2,3,4]. The remaining two absorption peaks correspond to π-π* transitions of the lone pair electrons in the azomethine bonds [3]. Likewise, L2 presents absorption peaks at similar wavelengths to those of L1, at 332 nm, 408 nm, and 432 nm. The hypochromic shifts observed in the L2 azomethine bands were caused by the inductive effect due to the presence of the hydroxyl group, which is electron-donating through sigma bonds [27]. In addition, the L1 and L2 spectra showed shoulders at 385 nm and 389 nm, respectively, which also corresponds to C=N chromophores. Moreover, the coordination complexes PtL1 and PtL2 have absorption peaks analogous to those of the free ligands. The platinum complex with L1 had peaks located at 323 nm, 434 nm, and 460 nm. The lower energy transitions due to the azomethine groups were lower in intensity as well, as a result of the short Pt-N bond length [28]. In contrast, the PtL2 complex exhibited three absorption peaks at 302 nm, 427 nm, and 442 nm, respectively. The first absorption IIh is attributable to naphtholate chromophore shifting to a lower wavelength due to coordination with the metal center. By applying the Tauc plot method to the UV-vis data, the experimental band gaps (Eg) for L1, PtL1, L2, and PtL2 were calculated to be 2.73 eV, 2.67 eV, 2.76 eV, and 2.66 eV, respectively (

Table 1. Experimental HOMO, LUMO, and optical band gap (Eg) values of the studied compounds.

Compound	Experimental Eg (eV) a
L1	2.73
L2	2.76
PtL1	2.67
PtL2	2.66

^a The Eg was estimated from the UV-Vis spectra using the Tauc plot method.

) [21,29,30,31,32].

2.2. Structural Characterization

Two new crystal structures are reported herein for the PtL1 and PtL2 complexes. The PtL1 complex crystallizes under the monoclinic P2_1/c space group with a square planar geometry for the metal center. The torsion angles on the O1-N1-N2-O2 and O1-N1-N2-Pd bonds are −0.28° and −179.4°, respectively. Additionally, the square planar geometry of the PtL1 complex was confirmed using the tau parameters (τ₄ and τ′₄), which have values very close to zero (Table S3). Meanwhile, PtL2 crystallizes in the orthorhombic space group P2₁2₁2₁; the geometry of this complex differs from that of PtL1 due to the longer carbon chain in the imine bridge of L2. The complex presents a distorted geometry compared with that of PtL1, as shown in Figure 2. The determined O1-N1-N2-O2 and O1-N1-N2-M torsion angles and tau parameter of PtL2 shown in Table S4 confirm the deviation from the square planar geometry.

2.3. DFT Calculations

Geometry optimizations were performed using the Gaussian 09 software at the B3LYP/6-31+G(d,p)/def2-TZVP theory level to corroborate the crystal structure parameters and the experimental optical band gap values of the studied compounds. Comparison of selected bond lengths of the optimized geometries with the crystallographic data demonstrated good accordance (Tables S4–S7). Furthermore, the theoretical band gap values obtained from the computational calculation followed the same trend as our experimental HOMO/LUMO gaps (Table S2). The theoretical band gap values of the enol form of the ligands differed greatly from the experimental data. As expected, the keto form resembled the experimental band gap value, validating the results of our previous study [26]. The enol tautomers of L1 and L2 phenolate oxygens had interatomic distances of 5.256 Å and 3.710 Å, respectively (Figure 3). In contrast, the L1 and L2 keto tautomer phenolate O1–O2 interatomic distances were 6.933 Å and 6.724 Å, respectively (Figure 3). The larger O1–O2 interatomic distance observed in the keto form of the ligands minimizes repulsion, thereby making the keto tautomer the more predominant.

To understand the higher performance in terms of the catalytic activity of PtL2 for the CO₂RR, we performed DFT computations to examine the Gibbs free energy changes along the pathway for CO production on both catalysts. We focused on the formation of the *COOH intermediate, where the C atom is adsorbed in an η1©-CO₂− (C-down) configuration, leading to the production of CO, as shown in Figure 4a. Our results showed that the adsorption of CO₂ on both catalysts was endothermic, with the Pt site serving as the active site. The rate-determining step (RDS) was the *COOH process. Interestingly, we observed hydrogen bond interactions between the *COOH intermediate and the naphtholate oxygen on PtL1 and PtL2. Notably, the *COOH intermediate with PtL2 exhibited an additional hydrogen bond interaction between the C=O group and hydrogen on the diamine bridge. This extra hydrogen bond may explain the larger Gibbs free energy obtained for CO release with PtL2 compared to PtL1 (see Figure 4b). The Hirschfeld charge analysis revealed that the Pt site in PtL2 lost electrons (from 0.243 au to 0.323 au) upon *COOH adsorption, while the one in PtL1 gained more (0.233 au to 0.318 au). This charge transfer trend in PtL1 resulted in additional higher potential and presented a higher voltage demand compared to the general direction of charge transfer in the C-down configuration, in which the C plays the role of a Lewis acid. We also noted that the distances between the *CO and Pt sites were greater than 3 Å in both PtL1 and PtL2, indicating classical physisorption. The final step of the proposed catalytic cycle was the release of CO from the Pt complex, leading to the recuperation of the neutral complex for subsequent CO₂RR.

2.4. Electrochemical Studies

Cyclic voltammetry (CV) was used to study the electrochemical behavior of compounds using a three-electrode cell with a 0.1 M TBAPF₆ in N,N-dimethylformamide (DMF) as the electrolyte solution. The L1 voltammogram in an N₂-saturated environment presented a single oxidation peak at E_a = 0.47 V vs. Fc/Fc⁺ and a reduction peak at E_c = −2.49 V vs. Fc/Fc⁺ (Figure 5a, red line) [26]. The observed oxidation was attributed to the formation of the [L1]⁺ species by Gutierrez et al. [26] A similar reduction peak was observed in all complexes; therefore, we hypothesize that this reduction is caused by the formation of the [L1]⁻ species. The voltammogram for L2 (Figure 5b, red line) under an N₂ atmosphere displayed a similar irreversible oxidation and reduction peak as that of L1 at E_a = 0.35 V vs. Fc/Fc⁺ and E_c = −2.38 V vs. Fc/Fc⁺. Similar to L1, the initial oxidation was a result of the formation of the charged species [L2]⁺, and the reduction was attributed to the [L2]⁻ species.

The blue line in Figure 5a corresponds to the CV of PtL1, the ligand-based oxidation and reduction peaks are shifted cathodically to E_a = 0.38 V vs. Fc/Fc⁺ and E_c = −2.55 V vs. Fc/Fc⁺, respectively. The PtL1 complex has an absent pre-peak in the voltammogram, suggesting greater stability following the initial oxidation than the free ligand. Furthermore, the PtL1 complex shows an additional irreversible reduction at negative potentials of E_c = −2.90 V vs. Fc/Fc⁺. In the reverse sweep, two new oxidation peaks were observed at E_a = −2.40 V and −1.69 V vs. Fc/Fc⁺, which were related to the two above-mentioned reduction processes. The reduction observed at −2.55 V vs. Fc/Fc⁺ corresponds to the formation of [PtL1]⁻ species; this process occurs at the metal center. The cyclic voltammogram of PtL2 (Figure 5b, blue line) under an inert atmosphere showed that the characteristic irreversible ligand oxidation shifted to E_a = 0.43 V vs. Fc/Fc⁺. At negative potentials, multiple indistinguishable reduction processes can be seen at 100 mV/s, but they are not distinguishable from each other. Reduction of the scan rate to 50 mV/s enabled a clear visualization of three irreversible reductions at E_c = −2.21, −2.53, and −2.86 V vs. Fc/Fc⁺ (Figure S12).

2.5. CO₂ Reduction

The CO₂ reduction activity of the reported compounds was tested by saturating the cell with CO₂; CVs were recorded to identify the peaks associated with the CO₂RR. The CV of ligands L1 and L2 showed no catalytic activity towards CO₂RR under a CO₂ saturation atmosphere, as observed in Figure S13a,b. The current densities of ligands (L1 and L2) were reduced by factors (J_CO2/J_N2) of 0.74 and 0.40, respectively. This suggests catalyst poisoning in the presence of CO₂ and a less reactive CO₂ adduct [33]. Nevertheless, the voltammograms of the PtL1 and PtL2 complexes showed increased current densities after complexation, which is in accordance with the results observed from DFT mechanistic calculations (Figure S13a,b).

The CV of PtL1 in the presence of CO₂ shows a sharp reduction process increment at E_c = −2.68 V vs. Fc/Fc⁺ with a current density of J_c = −0.90 mA/cm² under aprotic conditions (Figure 6a, red line). The current density increased by a factor of 3.46 compared to the voltammogram in the inert atmosphere. Adding water to the PtL1 complex under CO₂ saturation had a negative effect by inhibiting the current density response, as can be observed in Figure 6a (J_H2O/J_N2 = 0.67), which is indicative of the complex poor stability in the presence of water [34]. Increasing the water concentration under CO₂ saturation shifted the CO₂RR peak and onset potential to positive values (Figure S14b). The current density response of the PtL2 complex under CO₂ saturation was lower than that observed for PtL1, yet the peak associated with the CO₂ reduction occurred at a lower potential, E_c = −2.53 V vs. Fc/Fc⁺. This result validates the hypothesis that the addition of the hydroxy group enhances CO₂ reduction.

The PtL2 complex forms an adduct that facilitates the reduction of CO₂ by lowering the onset potential and increasing the interaction with CO₂ [33]. Furthermore, a pre-catalytic wave that is characteristic of the formation of a PtL2-COOH adducts was observed [15]. Finally, the activity of the PtL2 complex CO₂RR in the presence of water and CO₂ did not lower the catalytic activity of the complex as it did in PtL1. The addition of water as a proton source to the PtL2 enhanced its current density and resulted in an anodic shift of onset potential and a CO₂RR peak (−2.53 V to −2.51 V vs. Fc/Fc⁺). At 3 M of water, the current density reached its peak, and further addition of water inhibited the CO₂RR peak (Figure S14b). A linear plot was obtained from the Log–Log plot of the current density response versus the water concentration added to the electrochemical cell, indicating first-order kinetics with increasing water concentration (Figure S15). The negative slope corresponding to PtL1, shown in Figure S15, suggests that the complex was not stable with the increase concentration of water present during CO₂RR [34].

The quantification of gaseous products using a gas chromatograph enabled the accurate determination of the FE of the studied electrocatalysts. The conditions selected for the bulk electrolysis experiments were based on the CV experiments performed to determine the optimal product formation. The PtL1 complex had a calculated CO FE (FE_CO) of 5.5% under aprotic conditions at a reduction potential of −2.7 V vs. Fc/Fc⁺ after one hour of electrolysis (

Table 2. Product quantification for CO₂RR under aprotic and protic conditions.

Metal Complex	Atmosphere	Voltage (V vs. Fc/Fc+)	H2O (M)	CO FE (%)
PtL1	CO2	−2.7	0	5.5
	CO2	−2.7	3	5.4
	CO2	−2.5	3	0.8
	CO2	−2.9	3	0.9
	N2	−2.7	0	N.D.
	N2	−2.7	3	N.D.
PtL2	CO2	−2.5	0	4.4
	CO2	−2.5	3	42.0
	CO2	−2.3	3	4.15
	CO2	−2.7	3	0.8
	N2	−2.5	0	N.D.
	N2	−2.5	3	N.D.

). Under protic conditions, the FE_CO (5.4%) of PtL1 did not show any major improvement, as expected from the cyclic voltammograms. The benign effect of the FE_CO after the addition of water to the PtL1 solution can be attributed to not possessing any hydroxy groups capable of activating CO₂. The PtL2 complex displayed a similar performance under aprotic conditions to PtL1. Nonetheless, the addition of 3 M H₂O promoted CO₂RR, as was observed from CV, by lowering the onset and reduction potential of CO₂RR. The bulk electrolysis experiments using PtL2 with 3 M water as the proton source yielded the best performance, increasing the selectivity for CO formation to a total of 42.0% FE. After 3 h had elapsed, the FE_CO dropped to 17.1% (Table S10). The decay of the FE_CO is characteristic of degradation and/or poisoning of the electrocatalyst over the course of the bulk electrolysis [34]. Moreover, applying different potentials under protic conditions demonstrated that the optimal potential for PtL1 and PtL2 are −2.5 and −2.7 V vs. Fc/Fc⁺, respectively (Table 2 and Figure S18).

Additionally, sealed cell experiments under an inert atmosphere (N₂) and in the presence of metal complexes were performed to confirm that the detected CO was due to CO₂ electrolysis and not degradation of the metal complexes or the DMF (Table 2). Ex situ UV-Vis spectroscopy following the bulk electrolysis experiments was performed to evaluate any possible degradation of the material. The UV-Vis spectra of both metal complexes after electrolysis showed a significant decrease in the characteristic band intensities, as can be observed in Figure S19. The results presented confirm that metal complexes indeed degrade upon reduction of the CO₂RR; however, the degradation mechanism does not contribute to the CO observed from the CO₂RR. A reaction mechanism for PtL2 as a CO₂RR catalyst is proposed in Figure 7. We hypothesize that hydrogen bonding interactions arising from the hydroxy group, CO₂, and water cause an increment in the CO₂ molecule residence time, hindering the release of the formed CO (Figure 7, step 4.5). The proposed mechanism, presented in Figure 6, follows a five-step process involving an overall two-electron transfer to achieve the desired CO product. The water in the system acts as a protonating agent of the intermediates formed during CO₂ reduction. Step 4.5 is thought to be crucial due to the high degree of hydrogen bonding, causing the CO₂ to be retained longer, shortening the life of the electrocatalyst material.

3. Experimental Section

3.1. Materials and Methods

All reactions were performed under an N₂ atmosphere using a Schlenk line, and the solvents were dried using standard techniques. Reagents and solvents were used as received unless otherwise specified. Reagents and solvents were obtained from CCSI, Toa Baja, Puerto Rico, Thermo Fisher Scientific, Cayey, Puerto Rico, and VWR, Manatí, Puerto Rico. FTIR, UV-Vis, NMR, and single-crystal X-ray crystallography measurements were recorded using a Nicolet iS50 (Rochester, NY, USA), Shimadzu UV-1800 (Kyoto, Japan), Bruker Ascend Aeon 500 MHz (Faellanden, Switzerland), and Rigaku SuperNova equipped with a HyPix3000 X-ray detector, a Cu-Kα (λ = 1.5417 Å) radiation source (The Woodlands, TX, USA), respectively. Cyclic voltammetry (CV) was used to characterize the electrochemical properties of the compounds with an Autolab PGSTAT30 (Riverview, FL, USA) potentiostat. Bulk electrolysis experiments were performed using a BioLogic SP-150 potentiostat (Knoxville, TN, USA).

3.2. Synthesis and Characterization

Synthesis and characterization of L1 (2′-((1′,1′E)-(propane-1,2-diylbis(azaneylylidene))bis(ethan-1-yl-1-ylidene))bis(naphthalen-1-ol)) and L2 (2′-((1′,1′E)-((2-hydroxypropane-1,3-diyl)bis(azaneylylidene))bis(ethan-1-yl-1-ylidene))bis(naphthalen-1-ol)): The synthesis of L1 was performed as reported by Gutierrez et al. [26]: 1.50 g (8.1 mmol) of 1-hydroxy-2-acetonaphthone and 0.37 g (5.0 mmol) of 1-2-diaminopropane. L1 yield: 1.40 g, 85% based on 1-hydroxy-2-acetonaphthone. λ_max (DCM)/(nm): 325 (ε/dm³ mol⁻¹ cm⁻¹ 2990), 416 (5639), 434 (5259). FTIR (V_max/cm⁻¹): 3368 (OH, br), 1584 (conj. cycl. C=N, s), 1532 (C=C aromatic ring, s), 1271 (CO, s). ¹H-NMR (500 MHz, CDCl₃): ppm 1.58–1.60 (3H, d, -CH₃, J = 6.45 Hz), 2.41 (3H, s, N=C-CH₃), 2.46 (3H, s, N=C-CH₃), 3.83–3.94 (2H, m, -CH₂), 4.35–4.41 (1H, sept, -CH, J = 6.48 Hz), 6.84–6.87 (2H, dd, Ar-H, J₁ = 4.3 Hz, J₂ = 9.1), 7.27–7.28 (1H, d, Ar-H, J = 5.30 Hz), 7.41–7.44 (2H, t, Ar-H, J = 15.1 Hz), 7.51–7.54 (2H, t, Ar-H, J = 14.7 Hz), 7.60–7.62 (2H, d, Ar-H, J = 7.75 Hz), 8.46–8.47 (2H, d, Ar-H, J = 8.15 Hz). For the synthesis of L2, 1.00 g (5.4 mmol) of 1-hydroxy-2-acetonaphthone was added to a round-bottom flask along with MeOH until dissolved. In a separate container, 0.27 g (3.0 mmol) of 1,3-diamino-2-propanol was weighed and dissolved in MeOH and added to the round-bottom flask. The reaction mixture was refluxed until a yellow solid appeared. The obtained solid was isolated by vacuum filtration and washed several times with EtOH and Et₂O. L2 yield: 0.93g, 84% based on 1-hydroxy-2-acetonaphthone. λ_max (DCM)/(nm): 332 (ε/dm³ mol⁻¹ cm⁻¹ 7995), 385 (3134), 408 (4383), and 432 (4901). FTIR (cm⁻¹): 3374 (OH, br), 1594 (conj. cycl. C=N, s), 1546 (C=C aromatic ring, s), 1271 (CO, s). ¹H-NMR (500 MHz, DMSO-d⁶): ppm 2.54 (6H, s, N=C-CH₃), 3.71–3.76 (2H, ddd, H-CH, J₁ = 5.28 Hz, J₂ = 7.00 Hz, J₃ = 12.5 Hz) 3.87–3.83 (2H, dt, H-CH, J₁ = 4.6 Hz, J₂ = 14.0 Hz), 4.14 (1H, m, C-H), 5.85 (1H, s br, OH), 6.75–6.76 (2H, d, Ar-H, J = 9.20), 7.34–7.38 (2H, td, Ar-H, J₁ = 1.21 Hz, J₂ = 6.95 Hz), 7.46–7.48 (2H, d, Ar-H, J = 9.22), 7.50–7.53 (2H, td, Ar-H, J₁ = 1.33 Hz, J₂ = 6.93 Hz), 7.61–7.63 (2H, d, Ar-H, J = 7.82), 8.35–8.37 (2H, dt, Ar-H, J₁ = 1.25 Hz, J₂ = 8.13 Hz).

Synthesis and characterization of PtL1 and PtL2: To a round-bottom flask, 150 mg (0.4 mmol) of L1 was added and dissolved in 5 mL of DMF; the solution was stirred at 75 °C until dissolved. Afterward, 2 equivalents of NaOAc were suspended in the L1 solution. Then, a 2 mL solution of DMSO with 150 mg of K₂PtCl₄ (0.4 mmol) was added to the mixture. The reaction was constantly stirred and heated continuously at 75 °C for 24 h. Subsequently, the crude was purified by column chromatography using silica gel as the stationary phase and a mobile phase consisting of CH₂Cl₂:Ethyl acetate. X-ray-quality single crystals were obtained from vapor diffusion of Et₂O into a CH₂Cl₂ solution. PtL1 yield: 58.5 mg, 28%. λ_max (DCM)/(nm): 323 (ε/dm³ mol⁻¹ cm⁻¹ 8898), 434 (2209), and 460 (1751). FTIR (cm⁻¹): 1585 (C=C aromatic ring, s), 1452 (conj. cycl. C=N, s) 1273 (C-O, s). ¹H-NMR (500 MHz, DMSO-d⁶): ppm 1.34–1.35 (3H, d, CH₃, J = 6.35 Hz), 2.41 (3H, s, N=C-CH₃), 2.44 (3H, s, N=C-CH₃), 3.88–3.91 (2H, dd, H-CH, J₁ = 5.38 Hz, J₂ = 13.0 Hz), 4.37–4.42 (1H, m, C-H), 7.05–7.07 (2H, d, Ar-H, J = 9.15 Hz), 7.51–7.55 (2H, t, Ar-H, J = 15.1 Hz), 7.6118–7.6412 (2H, t, Ar-H, J = 14.7 Hz), 7.73–7.78 (5H, m, Ar-H), 8.70–8.72 (2H, d, Ar-H, J = 8.3 Hz). PtL2 was synthesized following a similar method to that used for PtL1, but using L2. L2 150 mg (0.4 mmol) was added to K₂PtCl₄ 150 mg (0.4 mmol). The resulting solution was purified by column chromatography using silica gel and EtOAC:CH₂Cl₂ as the eluent. X-ray-quality single crystals were obtained from vapor diffusion of Et₂O into a CH₂Cl₂ solution of the complex. PtL2 yield: 42.1 mg, 20%. λ_max (DCM)/(nm): 302 (ε/dm³ mol⁻¹ cm⁻¹ 12,270), 427 (3908), and 442 (3720). FTIR (cm⁻¹): 1577 (conj. cycl. C=N, s), 1552 (C=C aromatic ring, s), 1252 (C-O, s). ¹H-NMR (500 MHz, DMSO-d⁶): ppm 2.41 (6H, s, N=C-CH), 3.57–3.62 (2H, m, CH-H), 4.13–4.16 (2H, dd, CH-H, J₁ = 3.58 Hz, J₂ = 14.4 Hz), 4.61–4.66 (1H, m, C-H), 5.68 (1H, d, OH, J = 4.76 Hz), 6.95–6.97 (2H, d, Ar-H, J = 9.04 Hz), 7.43–7.46 (2H, t, Ar-H, J = 15.1 Hz), 7.51–7.57 (4H, Ar-H, m), 7.65–7.67 (2H, d, Ar-H, J = 8.00 Hz), 8.53–8.54 (2H, d, Ar-H, J = 8.24 Hz).

3.3. DFT Calculations

Geometry optimizations, single-point energy, and frequency calculations were performed using the Gaussian 09 software package using the B3LYP functional with the 6-31+G(d,p) basis set for H, C, N, and O atoms, and def2-TZVP (with effective core potentials) for Pt [35]. Frequency calculations performed on each structure confirmed that the optimized geometry had reached the minimum values on the basis of the absence of negative frequencies. Reaction pathway computations were performed by DFT with the DMol3 package [36,37]. The Perdew, Burke, and Ernzerhof (PBE) functional within the generalized gradient approximation (GGA) was used to describe exchange–correlation interactions. Dispersion corrections were treated by the Tkatchenko and Scheffler (TS) scheme [38]. The double numerical plus polarization (DNP) basis sets were adopted with a global orbital cutoff of 4.5 Å. The energy and force convergences were set as 10–5 Ha and 0.002 Ha/Å, respectively. To simulate the solvent environment, a conductor-like screening model (COSMO) was used with a dielectric constant of 78.54. The standard hydrogen electrode (SHE) [38] was used, but then converted to the Fc/Fc⁺ reference to compare with experiment data. The Gibbs free energy difference between steps was determined as ∆G = ∆E + ∆ZPE − T∆S, where ∆E, ∆ZPE, ∆S, and T are the electronic energy difference, zero-point energy difference, entropy difference, and temperature, respectively.

3.4. Electrochemical Setup

The electrochemical setup consisted of three electrodes and a sealed electrochemical cell. The cell was cleaned with aqua regia before each experiment; it was composed of a glassy carbon (GC) working electrode (⌀ = 3 mm), a nonaqueous Ag/AgNO₃ reference electrode, and a graphite counter electrode. The GC electrode was physically cleaned by polishing the GC surface in an eight-form pattern with different granular pore sizes (1.0, 0.5, and 0.03 µm) of alumina micro-polish (Buehler) and pure water. The solution of all the compounds consisted of 1 mM in 0.1 M TBAPF₆ in DMF as a supporting electrolyte. All the voltammogram electrochemical potentials were corrected with respect to the Fc/Fc⁺ redox couple. The N₂ and CO₂ saturation atmospheres were obtained by bubbling the DMF electrolyte solution for 20 min.

Bulk electrolysis experiments were carried out in a five-port EuroCell. For each experiment, 30 mL of 0.1 M TBAPF₆ in DMF solution was added to the cell. The working electrode was a 100 mm graphite rod (Ø = 6.7 mm), the counter electrode was graphite, and a nonaqueous Ag/AgNO₃ reference electrode was used. The counter electrode was isolated from the main solution by a glass frit separator. Enough catalyst was added to create a 1 mM solution, and if necessary, water was added to the 1 mM catalyst solution as a proton source. Inlet and outlet gas concentrations from electrolysis experiments were measured using an online gas chromatograph (Shimadzu GC-2014) equipped with flame ionization detector (FID), thermal conductivity detector (TCD), and methanizer assembly for the detection of low levels of CO₂ and CO through conversion to methane by FID.

4. Conclusions

In conclusion, we synthesized two new platinum salen-like naphthalene complexes (PtL1 and PtL2). All ligands and metal complexes were structurally characterized by single-crystal X-ray diffraction methods. The use of UV-Vis was crucial to determining the spectroscopic properties as well as the optical band gap, which was used as a descriptor when performing DFT calculation to define the predominant tautomer of the ligands. DFT calculations confirmed that the keto form was predominant over the enol form in solution, as well as the crystal structure of L1 and L2 [26]. Geometry optimizations using DFT corroborated the reported crystal structure bond lengths and angles. Additional DFT calculations were performed to elucidate the CO₂RR mechanism, and it was observed that PtL2 was better able to stabilize the *COOH intermediate than PtL1. However, since multiple hydrogen bonding sites were available for PtL2, hindering the release of CO, it displayed a higher Gibbs free energy than PtL1 (Figure 4b). In terms of the electrochemical properties of the reported compounds, the ligands were classified as non-innocent due to characteristic redox processes after metal coordination present in their respective CV (Figure 4). For the PtL1 complex under a CO₂-saturated environment in aprotic and protic conditions, the CO₂RR performance was substandard with a FE_CO of 5.5% and 5.4%, respectively. In addition, the optimal potential for the CO₂RR using PtL1 was determined to be −2.7 V vs. Fc/Fc⁺. Meanwhile, PtL2 showed a lower optimum reduction potential for CO₂ than its PtL1 counterpart, at −2.5 V vs. Fc/Fc⁺. The CO₂RR peak and onset potential of PtL2 shifted to positive potentials when water was added, with an FE_CO of 42.0% being achieved. Conversely, when water was not present, the FE_CO was 4.4%. The electrocatalyst was found to be stable for 1 h; afterwards, a significant decrease in FE_CO was observed due to the degradation of the electrocatalyst. This degradation did not result in the production of CO, as it was not detected during the bulk electrolysis experiments under N₂ atmosphere. Moreover, the electrocatalysis kinetics were of the first order with respect to water. In the case of PtL1, the negative slope of the Log[J(mA/cm²)] vs. Log[H₂O] plot is indicative of the complex’s lack of stability in the presence of water, unlike in the case of PtL2, where the synergy between the ligand’s hydroxy group and water as the sacrificial proton source enhanced the catalytic activity of CO₂RR. Although deactivation of the reported electrocatalysts is an issue that needs to be addressed, the high selectivity at the optimal potential of PtL2 and the facile synthesis of PtL1 and PtL2 demonstrate the possibility of using them as CO₂RR electrocatalysts. Future studies include optimization of the ligand backbone by adding different substituents, enhancing ligands’ rigidity and the study of the CO₂RR catalytic activity of these ligands’ coordination complexes with 3d metal ions.