Effect of the Cu2+/1+ Redox Potential of Non-Macrocyclic Cu Complexes on Electrochemical CO2 Reduction

Cu2+/1+ complexes facilitate the reduction of CO2 to valuable chemicals. The catalytic conversion likely involves the binding of CO2 and/or reduction intermediates to Cu2+/1+, which in turn could be influenced by the electron density on the Cu2+/1+ ion. Herein we investigated whether modulating the redox potential of Cu2+/1+ complexes by changing their ligand structures influenced their CO2 reduction performance significantly. We synthesised new heteroleptic Cu2/1+ complexes, and for the first time, studied a (Cu-bis(8-quinolinolato) complex, covering a Cu2+/1+ redox potential range of 1.3 V. We have found that the redox potential influenced the Faradaic efficiency of CO2 reduction to CO. However, no correlation between the redox potential and the Faradaic efficiency for methane was found. The lack of correlation could be attributed to the presence of a Cu-complex-derived catalyst deposited on the electrodes leading to a heterogeneous catalytic mechanism, which is controlled by the structure of the in situ deposited catalyst and not the redox potential of the pre-cursor Cu2+/1+ complexes.


Introduction
The rational design of metal complexes for CO 2 reduction relies on developing a clear understanding of the correlations between their molecular structure and catalytic performance [1]. The structure of the complexes can be altered in several ways, including changing the coordination number of the metal centre [2,3], the number of ligand binding sites (denticity) [4,5], and the electronic charges on the ligands [6,7]. Furthermore, the strength of the coordination bond, i.e., the coupling of the frontier orbitals of the metal centre and the ligand binding site, can be altered by placing electron-donating or -withdrawing groups on the ligands [8,9]. Steric groups on the ligands can affect the length of the coordination bond, which modifies the electronic coupling between the ligand and the metal ion. A prime example is the approximately 0.3 V difference in the redox potential of structurally similar ortho-and para-methyl-substituted tetragonal Cu-bipyridine complexes [10]. Due to the steric hinderance of the methyl groups at the ortho-position, the Cu 2+/1+ -N bond is weaken, and the electron density on the Cu 2+/1+ is lowered as compared to that of the para-substituted bipyridyl ligand. Such Cu 2+/1+ complexes have been extensively used in both homogeneous and heterogeneous electrocatalytic CO 2 reduction experiments [1]. Cu complexes are one of the rare metal complexes that can reduce CO 2 to methane (CH 4 ) as well as to C-C products of ethylene (C 2 H 4 ) and ethanol (C 2 H 5 OH) with high selectivity. CO 2 reduction experiments using Cu complexes suggested that the catalytic mechanism of CO 2 reduction involved the binding of the CO 2 or its intermediates to the Cu 2+/1+ metal centre [11][12][13]. However, to date, no clear correlations between the redox potential of tetragonal Cu complexes and their CO 2 reduction performance have been established [1].
On the other hand, correlations between the molecular structure of non-Cu-based complexes and their CO 2 reduction performance have been studied in homogeneous vs. Fc/Fc + under a CO2 atmosphere while the gas products were detected using gas chromatography (GC) at regular time intervals. After chronoamperometric measurements, the carbon paper electrodes with Cu-complex-derived catalyst layers deposited were rinsed and tested in a fresh electrolyte using the same procedure, but without the Cu complex added. Finally, X-ray diffraction (XRD) and scanning electron microscopy (SEM) were employed to characterize the Cu-complex-derived catalysts deposited on the carbon paper electrode.

Results and Discussion
The redox potential of the five Cu complexes was investigated by cyclic voltammetry using 1 mM Cu complexes dissolved in 0.1 M TBAP and DMF solution with 0.1 M deionized (DI) water in argon. Figure 2 shows the third cycle (see three consecutive cycles in Figure S1) of current density versus potential curves. The four Cu complexes (dmp, hp, hbpy, htpy) showed a quasi-reversible redox wave within the −2.4 to 0.52 V vs. Fc/Fc + potential range. Based on the known electrochemical behaviour of these complexes, the redox reaction was assigned to the Cu 2+ to Cu + redox reaction, modulated by the different coordination of the ligands. The redox potential of the heteroleptic Cu complexes (hp, hbpy, htpy) shifted to a more negative potential as compared to that of the homoleptic Cu complex (dmp) in Table 1. In simple terms, the negative shift can be rationalized by the stronger coordination of the HETPHEN ligand as compared to the dmp ligand. Among the heteroleptic complexes, hbpy showed a 70 mV negative shift as compared to hp, consistent with the redox potential difference between the homoleptic analogues of bpy versus dmp complexes [38]. The Cu 2+/1+ redox potential of htpy showed a larger negative shift of 450 mV as compared to the hp, which is attributed to the enhanced electron density on the Cu 2+ due to the coordination of the third N of the terpy ligand. The redox potential of the Cu 2+/1+ of cq is even more negative at −1.27 V, as previously reported [37]. The five complexes provide a redox potential range of almost 1.3 V. In the following sections, the effect of this large redox potential variation on CO2 reduction performance is investigated. To achieve the goals, five Cu complexes were designed and synthesised (dimethylphenanthroline Cu complex (dmp), Cu complexes of the 6,6 -dimesityl-2,2 -bipyridine ligand and a series of second ligands, phenanthroline (hp), bipyridine (hbpy), and terpyridine (htpy) and Cu-bis(8-quinolinolato) (cq)), as shown in Figure 1a. From this study we have excluded macrocyclic Cu complexes such as porphyrins as their redox processes in the negative regime is typically dominated by the ligand reduction and not the Cu 2+/1+ centre. The homoleptic Cu complex (dimethylphenanthroline Cu complex (dmp)) is well-studied in the literature [29]. We have recently shown highly selective CO 2 reduction to methane in an organic electrolyte in the presence of organic cations TBAP [28]. In DMF solvent, a Cu-complex-derived catalyst formed on the surface of the electrode during chronoamperometric testing under CO 2 atmosphere. The optimum applied voltage was −2.17 V versus Fc/Fc + , where high FE for methane production was observed without significant decomposition of the organic electrolyte [28]. XRD measurements showed the presence of a crystalline Cu phase, while XPS measurements suggested the presence of a N-containing organic phase. We noted that the FE of methane production using the dmp complex varied from 80% to 40%, even though the deposition was performed under nominally identical conditions. In addition, we have studied the in situ growth of the Cu-complex-derived catalyst during chronoamperometric testing under a CO 2 atmosphere [30]. Evidenced by a gradually increasing capacitance, the appearance and growth of new redox active peaks in the CVs, as well as switching the main CO 2 reduction product from CO (first 5 to 20 min) to CH 4 (>40 min), the CO 2 reduction mechanism was explained by switching from homogeneous to a heterogeneous mechanism. Some differences in the morphology of the deposited catalyst were observed, but the origin of the morphology change was not known. For the purposes of establishing correlations between the redox potential and the CO 2 reduction performance, we included data sets for both high FE (dmp-A) and lower FE (dmp-B) samples. In addition, we have synthesised three new heteroleptic Cu complexes using the HETPHEN strategy. The HETPHEN strategy (HETeroleptic PHENanthroline) was developed to modulate the electronic nature of tetragonal Cu complexes [31,32] by employing bulky ligands, i.e., by adding phenyl rings at the 2-and 9-position to the phenanthroline or bipyridine [33,34]. The HETPHEN ligands hinder the formation the homoleptic HETPHEN Cu complex, leading to the isolated heteroleptic Cu complex. This strategy has been used to fully characterize supramolecular Cu complexes [35,36]. The key benefit of the HETPHEN strategy for this work is the ability to tune the redox potential of Cu 2+/1+ using the isolated heteroleptic Cu complexes. The addition of a second, less bulky ligand, can potentially create a lower steric effect at a certain position of the Cu site to facilitate the bonding of CO 2 , leading to selective CO 2 reduction. The steric effect generated by heteroleptic Cu complexes would also influence the chemical nature of the Cu-complex-derived catalyst during electrolysis, such as the Cu-Cu coordination number, related to the hydrogeneration of *CO to CH 4 . HETPHEN Cu complexes are therefore used for the first time for electrochemical CO 2 reduction. We used a 6,6 -dimesityl-2,2bipyridine ligand and as the second ligand, and we chose phenanthroline (hp), bipyridine (hbpy), and terpyridine (htpy). Unlike bipyridine, terpyridine is tridentate, so the third N atom of the ligand may block the 5th coordination site of the Cu 2+ complexes. This 5th coordination site may otherwise be occupied by solvent molecules or coordinating counter ions such as Cl − [29], affecting the redox potential significantly. The counter ion for the complexes were bis(trifluoromethane)sulfonimide (TFSI) for hp and tetrafluoroborate (BF 4 ) for hbpy and htpy. The synthesis and characterization of new Cu complexes can be found in Section 3, and UV-vis spectra is provided in Figure S8. Lastly, a (Cu-bis(8-quinolinolato)) (cq) complex was also tested. In this complex, the Cu 2+ ion is coordinated with two N atoms and two O atoms. Unlike the polypyridyl complexes, the (Cu 2+ -bis(8-quinolinolato)) complex is overall neutral as the ligand is negatively charged after deprotonation. Its redox potential has been reported to be more negative than the Cu bipyridyl complexes although their geometry is similar. Pape et al. reported that the Cu 2+ /Cu 1+ redox potential in (Cu-bis(8-quinolinolato) has shifted significantly to −0.564 V vs. NHE (−1.24 V vs. Fc/Fc + ), a more than 1 V shift as compared to that of the Cu 2+ salt, measured in 0.01 M TBAClO 4 /DMSO [37].
First, cyclic voltammetry (CV) was performed to consistently determine the redox potentials of the five Cu complexes dissolved in 0.1 M TBAP/DMF electrolyte with 0.1 M water in an argon atmosphere. Then, chronoamperometry (CA) was performed at −2.17 V vs. Fc/Fc + under a CO 2 atmosphere while the gas products were detected using gas chromatography (GC) at regular time intervals. After chronoamperometric measurements, the carbon paper electrodes with Cu-complex-derived catalyst layers deposited were rinsed and tested in a fresh electrolyte using the same procedure, but without the Cu complex added. Finally, X-ray diffraction (XRD) and scanning electron microscopy (SEM) were employed to characterize the Cu-complex-derived catalysts deposited on the carbon paper electrode.

Results and Discussion
The redox potential of the five Cu complexes was investigated by cyclic voltammetry using 1 mM Cu complexes dissolved in 0.1 M TBAP and DMF solution with 0.1 M deionized (DI) water in argon. Figure 2 shows the third cycle (see three consecutive cycles in Figure S1) of current density versus potential curves. The four Cu complexes (dmp, hp, hbpy, htpy) showed a quasi-reversible redox wave within the −2.4 to 0.52 V vs. Fc/Fc + potential range. Based on the known electrochemical behaviour of these complexes, the redox reaction was assigned to the Cu 2+ to Cu + redox reaction, modulated by the different coordination of the ligands. The redox potential of the heteroleptic Cu complexes (hp, hbpy, htpy) shifted to a more negative potential as compared to that of the homoleptic Cu complex (dmp) in Table 1. In simple terms, the negative shift can be rationalized by the stronger coordination of the HETPHEN ligand as compared to the dmp ligand. Among the heteroleptic complexes, hbpy showed a 70 mV negative shift as compared to hp, consistent with the redox potential difference between the homoleptic analogues of bpy versus dmp complexes [38]. The Cu 2+/1+ redox potential of htpy showed a larger negative shift of 450 mV as compared to the hp, which is attributed to the enhanced electron density on the Cu 2+ due to the coordination of the third N of the terpy ligand. The redox potential of the Cu 2+/1+ of cq is even more negative at −1.27 V, as previously reported [37]. The five complexes provide a redox potential range of almost 1.3 V. In the following sections, the effect of this large redox potential variation on CO 2 reduction performance is investigated.  In Figure S2, the CVs of five Cu complexes recorded under argon and CO2 atmospheres in the −2.37 to −0.672 V vs. Fc/Fc + potential range are shown. The CVs measured in CO2 show sharply rising current densities with no discernible redox peaks, while the CVs under argon atmosphere show a diverging behaviour with some samples showing a redox peak at highly negative potentials indicative of ligand reduction (dmp-A and dmp-B) and Cu 2+ reduction (cq) with a possible contribution from the deposition of the Cu-complexderived catalyst and/or electrolyte decomposition. This potentials range is more negative than the Cu 2+/1+ redox peaks identified under an argon atmosphere in Figure S2 except for the cq complex. Within this range, water reduction to H2, CO2 reduction to various products, as well as ligand reduction and deposition of a Cu-complex-derived catalyst are all possible reactions contributing to the current [28]. Next, chronoamperometric measurements were performed at −2.17 V vs. Fc/Fc + under a CO2 atmosphere and the products of CO, CH4, and H2 were collected using the GC after 20, 40, 60, and 80 min, as shown in Figure S3. The catalytic conversion of CO2 to various products could be a superposition of a homogeneous and heterogeneous catalytic mechanism, as explained above, i.e., it may involve the molecular catalyst of Cu 2+/1+ complexes dissolved in electrolytes or the catalytic activity of a Cu-complex-derived catalyst deposited during the course of the chronoamperometric measurements. It is expected that towards the end of the testing, heterogeneous catalysis may start to dominate, as the coverage of the carbon paper electrode with the catalyst deposited in situ is more compact. The ratio of FE values determined for CO,  In Figure S2, the CVs of five Cu complexes recorded under argon and CO 2 atmospheres in the −2.37 to −0.672 V vs. Fc/Fc + potential range are shown. The CVs measured in CO 2 show sharply rising current densities with no discernible redox peaks, while the CVs under argon atmosphere show a diverging behaviour with some samples showing a redox peak at highly negative potentials indicative of ligand reduction (dmp-A and dmp-B) and Cu 2+ reduction (cq) with a possible contribution from the deposition of the Cu-complex-derived catalyst and/or electrolyte decomposition. This potentials range is more negative than the Cu 2+/1+ redox peaks identified under an argon atmosphere in Figure S2 except for the cq complex. Within this range, water reduction to H 2 , CO 2 reduction to various products, as well as ligand reduction and deposition of a Cu-complex-derived catalyst are all possible reactions contributing to the current [28]. Next, chronoamperometric measurements were performed at −2.17 V vs. Fc/Fc + under a CO 2 atmosphere and the products of CO, CH 4 , and H 2 were collected using the GC after 20, 40, 60, and 80 min, as shown in Figure S3. The catalytic conversion of CO 2 to various products could be a superposition of a homogeneous and heterogeneous catalytic mechanism, as explained above, i.e., it may involve the molecular catalyst of Cu 2+/1+ complexes dissolved in electrolytes or the catalytic activity of a Cu-complex-derived catalyst deposited during the course of the chronoamperometric measurements. It is expected that towards the end of the testing, heterogeneous catalysis may start to dominate, as the coverage of the carbon paper electrode with the catalyst deposited in situ is more compact. The ratio of FE values determined for CO, CH 4 , and H 2 after the first 20 min of chronoamperometry as well as after 80 min in Table 1 were correlated with the redox potential determined by CV, as shown in Figure 3a-f. The correlation in Figure 3a shows a distorted bell-shaped curve, i.e., as the redox potential decreases, the FE ratio for CO at 20 min increases from 14.6% to as high as 57.7% in the order of dmp-A, dmp-B, hp, and hbpy (see line drawn to guide the eye). However, the Faradaic efficiency ratio for CO decreases for htpy and cq complexes. The distorted bell-shape curve can be explained by a homogeneous CO 2 to CO conversion mechanism involving the dmp, hp, and hbpy complexes. In these complexes, increasing the electron density on the Cu 2+/1+ (easier to oxidize) facilitates that conversion of CO 2 to CO, possibly through enhanced CO 2 binding to Cu ions. The drop in performance for the htpy complex can be explained by the weakened coordination of CO 2 due to the more crowded Cu 2+ coordination sphere occupied by the extra N atom at the fifth coordination site of the terpy ligand. The homogeneous cq complex catalyzes water reduction to H 2 over CO 2 reduction to CO, as shown in Figure 3b, which could be attributed to its ability to bind water molecules to the Cu 2+ site [39]. Note that about 30 times more H 2 was collected in the presence of cq molecular catalyst as compared to that of using carbon paper alone at the same applied potential (Table S1). Two H 2 O molecules can bind to the Cu site, enhancing water reduction. A distorted bell-shaped curve excluding cq complex is therefore clearer (Figure S9), as the other Cu complexes favour catalyzing CO 2 reduction. In contrast, Figure 3c shows an inverse bell-shaped curve, i.e., as the redox potential increases, the FE for CH 4 at 20 min increases in the order of hbpy, hp, dmp-A, and dmp-B complexes. Similarly to CO in Figure 3a, the FE for CH 4 is lower for the htpy and cq complexes. In a previous work using the dmp complex, we found that the main product of CO 2 reduction switched from CO to CH 4 during chronoamperometry. CH 4 production was attributed to a heterogeneous catalytic mechanism driven by the deposition of a Cu-complex-derived catalyst [30]. The inverse relationship between the redox potential and CH 4 evolution could therefore be attributed to the rate of forming a heterogeneous catalyst on the working electrode surface as well as its catalytic activity. The more positively shifted the redox potential of the Cu 2+/1+ is, the less stable the complex is, and therefore, the faster the formation of the Cu-complexderived catalyst is. We studied the correlations between the redox potential and the specific capacitance, the current density in CV, and the total charge during CA in Figure S11a-c. The specific capacitance (F/g) was calculated at the same potential (0.2 V) in Table S2 of the Cu-complex-derived catalysts using the CV data in Figure S10. We could not calculate the specific capacitance of the sample derived from the hp complex due to the broad redox peaks at 0.2 V. We could not see any significant correlations in Figure S11a-c, which implies that other factors, i.e., the solubility of the ligands or the reduced Cu complex, also contribute to the rate of catalyst decomposition. dmp-B and dmp-A complexes showed the highest FE ratio for CH 4 as 72.4 and 69.6% at 20 min, while the FE ratio for CH 4 using the other complexes was less than 52%. This suggests that the HETPHEN strategy is better for CO 2 reduction to CO in a homogeneous mechanism than depositing a heterogeneous catalyst performing CH 4 production. In summary, the increased stability of the Cu complex with more negative redox potential may explain the preference of CO production by the Cu complex molecular catalyst, while the decreased stability of the Cu complex with a more positive redox potential may explain the faster deposition of the Cu-complex-derived catalyst leading to the preference for CH 4 production via a heterogeneous mechanism. The correlation between the FE ratio of gas products at 80 min of CA and the redox potential in Figure 3d-f shows similar trends as compared to that in Figure 3a-c except for cq complex. The similar trends in product FE ratios at 20 min and 80 min in Figure 4a-c suggest that even at higher heterogeneous catalyst loading at 80 min, the conversion of CO 2 to CO and H 2 evolution from water still proceed through a homogeneous mechanism, while CH 4 is produced by a heterogeneous mechanism involving the Cu-complex-derived catalyst. However, the changes in the behaviour using the cq molecular catalyst, i.e., the increased FE ratio for CO from 17.1% to 29%, indicates a switch to a heterogeneous mechanism for this catalyst. The purely heterogeneous mechanism was tested by performing CA using the working electrodes with the deposited Cu-complex-derived catalyst in a fresh 0.1 M TBAP/DMF electrolyte without the Cu complex. The amount of gas products detected using the GC are shown in Figure S4. XRD and SEM measurements show the morphology of the Cu-complex-derived catalysts in Figures S5 and S6, respectively. Figure 3g-i shows the correlation between the redox potential of the Cu 2+/1+ complex used to deposit the catalyst and the catalytic performance of the Cu-complex-derived catalysts in the purely heterogeneous mechanism. Unlike in Figure 3a,b, no clear correlation between the FE ratio for CO and the redox potential of the original Cu 2+/1+ complex is found in Figure 3g. This suggests that the mechanism switched from a homogeneous to a heterogeneous mechanism for CO in the absence of a dissolved molecular catalyst. The CO product ratio increased for cq, dmp-B, htpy, and hp, while it decreased for dmp-A and hbpy. The main difference between the highly efficient CH 4 catalyst dmp-A sample is the ability to maintain high CH 4 FE while dmp-B partially produced CO after the dmp complexes were removed from the electrolyte. The heterogeneous catalysis results suggest that CO 2 binding strength to the new Cu-complex-derived catalyst is unrelated to the redox potential of the Cu 2+/1+ ion. This is not surprising since XRD shows the presence of Cu metal, hence the binding energy should be similar provided that the CO 2 binds to the same crystal facet. The FE ratio for H 2 decreased in all samples in Figure 3h in the purely heterogeneous mechanism, confirming that H 2 evolution predominantly followed a homogeneous mechanism, shown in Figure 3b,e. The correlation between the redox potential and FE ratio of CH 4 in the purely heterogeneous mechanism still exists in Figure 3i but it seems less correlated as compared to Figure 3c,f. The weaker linear correlations in product FE ratio for CH 4 between the purely heterogeneous mechanism and at 20 min and 80 min of CA in the Cu complex dissolved electrolyte ( Figure S7) confirmed that dmp-B and hbpy are out of the linear correlations. dmp-B produced more CO than dmp-A in the purely heterogeneous mechanism, leading to the difference in FE ratio for CH 4 (20%). dmp-A and dmp-B showed the same redox potential of Cu 2+/1+ in CV; however, showing different selectivity for CO and CH 4 indicates that other factors may affect the product selectivity, i.e., by changing the morphologies or crystal facets during homogeneous catalysis. The stability of the dmp-complex-derived catalyst was tested for 5 h [28]. The current gradually decayed and the FE for CH 4 decreased over time. The decaying performance was thought to originate from the restructuring of the catalyst, evidenced by the morphological change before and after the long-term test. hbpy in the purely heterogeneous mechanism would be more active to produce CH 4 compared to hp and htpy.
The relationship of the redox potential of the Cu 2+/1+ complexes and their product selectivity for CO 2 reduction is shown to reveal previously unknown correlations, i.e., a distorted bell-shaped curve. This finding does not follow the correlation found in previous studies between the molecular structure (electron-donating or -withdrawing groups) and the CO 2 reduction performance using non-Cu-based complexes [8,14,15]. A possible reason for the more complex behaviour is the interplay between the transformation of the Cu complex to the Cu-complex-derived catalyst and CO 2 reduction to CO, both affected by the redox potential of the Cu 2+/1+ complexes. This study is among the first attempts to find correlations between the redox potential and CO 2 reduction performance using Cu 2+/1+ complexes. We emphasize that from the point of view of designing new complexes with improved performance, not only the correlations between the redox potential and CO production using the Cu complex catalyst in homogeneous mechanisms, but also the lack of correlations with the catalytic performance in purely heterogeneous mechanisms, are also important. The cq complex catalyst performed highly efficient hydrogen evolution (FE ratio for 53.1% at 20 min of CA). The homogeneous cq complex catalyst could be further explored for hydrogen evolution at the conditions not forming Cu deposits, i.e., lower negative potential or aqueous solution. The correlations in trying to separate homogeneous and heterogeneous mechanisms will be further studied by the new techniques, i.e., in situ and in operando spectroscopic techniques (in situ FT-IR, XPS, XAS) and quantum chemical methods to investigate the catalyst structures and catalytic reaction mechanisms. Those techniques will create big data sets, i.e., crystallographic data, morphology, Cu-Cu coordination number, and Cu-ligand bonding energy. Our approach of using the correlation analysis of five molecular structures should be expanded by using machine learning algorithms on big data sets by studying hundreds of molecular structures as the next step. Such big-data research methodologies using advanced statistical tools and machine learning algorithms are expected to dominate the development of redox-active catalysts in the future. By utilizing those techniques, the origin of catalytic performance of the macrocyclic Cu complexes, which are dominated by ligand reduction and are known to form reversible Cu-complex-derived catalysts during CA, could also be also investigated. Separating homogeneous and heterogeneous mechanisms using correlation analysis will help researchers design Cu complexes specifically tailored for homogeneous or heterogeneous catalysts.    The relationship of the redox potential of the Cu 2+/1+ complexes and their product selectivity for CO2 reduction is shown to reveal previously unknown correlations, i.e., a distorted bell-shaped curve. This finding does not follow the correlation found in previous studies between the molecular structure (electron-donating or -withdrawing groups) and the CO2 reduction performance using non-Cu-based complexes [8,14,15]. A possible reason for the more complex behaviour is the interplay between the transformation of the Cu complex to the Cu-complex-derived catalyst and CO2 reduction to CO, both affected by the redox potential of the Cu 2+/1+ complexes. This study is among the first attempts to find correlations between the redox potential and CO2 reduction performance using Cu 2+/1+ complexes. We emphasize that from the point of view of designing new complexes with improved performance, not only the correlations between the redox potential and CO production using the Cu complex catalyst in homogeneous mechanisms, but also the lack of correlations with the catalytic performance in purely heterogeneous mechanisms, are also

Electrochemical Measurements
Electrochemistry was conducted in a three-electrode, two-compartment cell (Pine Instruments, Durham, NC, USA) with a potentiostat (650D, CHInstrument). A carbon paper electrode was employed as the working electrode in a compartment and Ag/AgNO 3 was used as the reference electrode in the same compartment. A platinum mesh was used as the counter electrode in another compartment. Two compartments were separated by a glass frit. Surface area of carbon paper electrode was 1.5 cm 2 . An amount of 1 mM Cu complexes were dissolved in DMF, 0.1 M TBAP and 0.1 M water. Before the measurement, the cathodic compartment was purged by argon for 30 min. After CV measurement in argon, the cathodic chamber was purged by CO 2 for 30 min and CO 2 gas was bubbled with a 15 mL/min constant flow rate during CV and CA measurement. The electrolyte was stirred with 300 rpm during electrochemical measurement. The redox potential of the Cu complex was determined by using oxidation and reduction potentials in Figure 2 using the equation E redox = (E oxi + E red )/2. hp and hbpy complexes showed two oxidation and reduction potentials. A reduction and an oxidation peak at more negative potential were due to the presence of possible homoleptic complexes [34]. The redox potential of hp and hbpy complexes was therefore calculated by the more positive potentials. All potentials applied in this work were adjusted to the potential versus Fc/Fc + added as an internal standard.

Gas Product Analysis
The gaseous products were measured through the gas outlet of electrochemical cell to a gas chromatograph (GC2030, Shimadzu, Kyoto, Japan) equipped with thermal conductivity (TCD) and flame ionization detectors (FID). The amount of gas products was calculated by calibration curves of known volumes of gas.

Material Characterization
X-ray diffraction (XRD) patterns of rinsed carbon paper electrodes were measured with a PANalytical Empyrean goniometer with a long focus Cu anode tube. The accelerating voltage was 45kV and the current was 40 mA in a standard Brag-Brentano reflection geometry. All scans were performed in a 2theta angle range of 20 to 90 • . Field-emission scanning electron microscopy (FE-SEM, JEOL, JSM-7500FA) was used to obtain the morphologies of deposits on carbon paper electrode.

Conclusions
In summary, we investigated whether controlling the redox potential of Cu 2+/1+ complexes influences their CO 2 reduction performance for the first time. We have tested five Cu complexes with different ligands, including new HETPHEN complexes and Cu-(bis (8quinolinolato), showing a wide range of redox potentials of Cu 2+/1+ . We have found that the redox potential significantly affected the FE ratio of CO production by the Cu complex homogeneously dissolved in electrolyte. Even through a Cu-complex-derived catalyst was formed during CA, homogeneous catalysis of CO 2 to CO by the Cu complex was still dominant except for the cq complex. The cq complex catalyst mainly performed the hydrogen evolution by binding water molecules. CH 4 was produced by the Cu complexderived catalyst formed in situ, and dmp-A and dmp-B showed the highest FE ratio for CH 4 . HETPHEN complexes performed lower FE for CH 4 but higher FE for CO than that of dmp complexes, suggesting that HETPHEN complexes are better for performing homogeneous catalysis for CO production. In purely heterogeneous mechanism, there was no correlation between the redox potential and the FE ratio for CH 4 . CH 4 production is probably controlled by the structure and morphology of the Cu-complex-derived catalyst rather than the redox potential of Cu 2+/1+ . This study opens the possibility of systematic studies between molecular structure, transformation of Cu complexes, and CO 2 reduction performance, by the benefit of machine learning algorithms, for the design of new homogeneous and heterogeneous Cu complexes.

Supplementary Materials:
The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/molecules28135179/s1. Figure S1: Three cycles of CV curves of five Cu complexes in argon; Figure S2: CVs of Cu complexes dissolved in 0.1 M TBAP/DMF electrolyte with 0.1 M water in argon (a) and CO 2 (b); Figure S3: (a) CA in the presence of dissolved complexes at −2.17 V vs. Fc/Fc + , (b) FE of gas products at 20 min of CA, and (c) FE of gas products at 80 min of CA; Figure S4: (a) CA in purely heterogeneous mechanism at −2.17 V vs. Fc/Fc + , (b) FE of gas products at 60 min of CA; Figure S5: XRD patterns of carbon paper and Cu-complex-derived catalystsafterCA; Figure S6: SEM images of Cu-complex-derived catalysts with deposits after CA; Figure S7: FE ratio for CH 4 in heterogeneous catalysis versus at 20 min (a) and at 80 min (b) of CA in the Cu complex dissolved electrolyte; Figure S8: UV-vis spectra of Cu complexes synthesized using the HETPHEN strategy; Figure S9: The Cu redox potential in argon (excluding cq complex) versus the Faradaic efficiency ratio of CO at 20 min of CA at −2.17 V vs. Fc/Fc + ; Figure S10: CVs of the Cu-complex-derived catalysts in CO 2 ; Figure S11: The Cu redox potential in argon versus specific capacitance (F/g) (a) and current density at −2.17 V in CV under CO 2 (b) and total charge collected during CA (c); Table S1: The amount of H 2 produced using carbon paper with and without cq complex in the electrolytes during CA at −2.17 V vs. Fc/Fc + ; Table S2: Specific capacitance of the Cu-complex-derived catalysts. Data Availability Statement: Data reported in the manuscript will be available upon reasonable request to the corresponding author.