Electrochemical and Spectroelectrochemical Studies on Oxygen Reduction Mediated by Cu(II) Complexes Containing the Alkylamine Ligand N,N-Dimethylethylendiamine

Abstract

In this study, Cu(II) complexes containing the bidentate ligand N,N-dimethylethylendiamine (dmen), i.e., [Cu^II(dmen)₂(CH₃COO)₂], [Cu^II(dmen)₂(NO₃)₂], and [Cu^II(dmen)₂Cl₂], were developed to explore molecular catalysis for the oxygen reduction reaction (ORR). Cyclic voltammetry and UV–vis spectroelectrochemical and electrochemical impedance spectroscopy experiments were performed in the absence and presence of oxygen. The UV–vis spectroscopy results suggested that the aforementioned Cu(II) complexes present an octahedral geometry in the solid state; meanwhile, they show a square pyramidal geometry in an aqueous solution. It is proposed that the chemical species [Cu^I(dmen)₂H₂O]⁺ reacts with O₂, exhibiting an outer-sphere electron transfer mechanism. The same UV–vis spectroelectrochemical response obtained with and without O₂ indicated a direct electron transfer from Cu(II) to Cu(I), with the regeneration of catalyst and the absence of other intermediate species. Among the reported compounds, [Cu(dmen)₂(NO₃)₂] exhibited the highest catalytic rate (TOF = 1.3 × 10⁴ s⁻¹). The impedance spectroscopy results showed that the resistance charge transfer (R_ct) of the redox pair Cu^II|Cu^I decreased in the presence of O₂ from 36.391 kΩ to 5.54 kΩ. For a better understanding of the effect of aliphatic amines on the ORR, a comparison with the complex [Cu(1,10-phen)₂NO₃]NO₃ is also presented.

1. Introduction

The search for renewable and alternative energy sources has encouraged the scientific community in recent years, due to the worldwide scarcity of fossil fuel [1]. One of the most promising renewable and alternative energy options are fuel cells, which are devices capable of converting chemical energy from a reaction into electrical energy, thereby providing enough energy to fulfill vehicle demands [2,3]. In these devices, the oxygen reduction reaction (ORR) is widely used for energy generation. In an acidic medium, the ORR generates 2H₂O via 4e⁻ reduction [4]. Unfortunately, the efficiency of the ORR is determined by the overpotential associated with slow kinetics that decreases the overall efficiency to 40–60% [3,5]. Pt has been used as a catalyst for the ORR and is capable of accelerating the reaction, but with a high overpotential (300 mV). In addition to the low abundance and high cost of Pt, its implementation in commercial cells has been inviable. For this reason, investigations have focused on the search of high-efficiency catalysts based on non-precious metals [6,7,8,9].

Laccase is a metalloenzyme, capable of reducing O₂ to H₂O, with a very small overpotential (approximately 20 mV), which has led to research on anchoring it in fuel cells; however, the narrow pH range and sensitive operating conditions have limited its implementation [10,11,12]. The high catalytic activity of this enzyme has inspired the scientific community to investigate Cu complexes as ORR catalysts. The kinetics, thermodynamics, and reactivity of the reaction of Cu(I) with O₂ have been widely studied, due to its presence in biological processes [13,14]. In recent years, its use as a catalyst in fuel cells has been considered. A wide variety of mono- and binuclear Cu(II) complexes have been studied for the ORR, with different ligands such as heterocyclic porphyrins, phthalocyanines [15], substituted phenanthrolines [16,17], and pyridylalkylamine ligands [18,19,20,21,22]. Among all the reported compounds, the Cu complex with the ligand (tris(2-pyridylmethyl)amine (tmpa) has demonstrated the highest activity for the ORR, with a turnover frequency (TOF) of 1.8 × 10⁻⁶ s⁻¹ [18]. However, only a few copper complexes containing alkylamine ligands, such as tris(3-aminopropyl)amine (trpn) [23] and tris(2-aminoethyl)amine (tren) [24], have been reported for the ORR. In addition, no extensive electrochemical characterization using spectroelectrochemical and electrochemical impedance spectroscopy (EIS) has been reported to obtain information about the mechanism of these copper complexes.

In this study, we assess the electrochemical, spectroelectrochemical, and EIS response of mononuclear Cu(II) complexes with the N,N-dimethylethylenediamine (dmen) ligand in the presence and absence of O₂ to explore their ability in the ORR and to contribute to the understanding of the role of molecular structure in the process. TOFs were calculated through foot of the wave analysis (FOWA) and compared with other reported Cu complexes. As a reference, the ORR using the complex [Cu^II(1,10-phen)₂NO₃]NO₃ is also presented.

2. Results and Discussion

2.1. Complexes’ Characterization

2.1.1. Solid-State Characterization

The IR spectra of complexes present characteristic vibrations between 3000 and 2800 cm⁻¹ corresponding to the ν(C-H) vibration of the N-CH₃, -CH₃, and -CH₂ groups of dmen ligand; see Figure S1. The stretching vibration ν(Cu-N) is detected between 500 and 450 cm⁻¹, which is in agreement with the reported value for Cu(II) coordination with ethylenediamine [25,26,27]. The complex [Cu(dmen)₂(CH₃COO)₂] presents two typical stretching vibrations ν(COO⁻)_s,as at 1410 and 1572 cm⁻¹, with a Δν(COO-) = 162 cm⁻¹ associated with the monodentate coordination mode M-O-CO [28]. Furthermore, the complex [Cu(dmen)₂(NO₃)₂] presents signals at 1313 and 1414 cm⁻¹ corresponding to asymmetrical and symmetrical (NO₂) vibrations, which could be related with a monodentate coordination mode for NO₃⁻. The mass spectrum data of complexes showed m/z signals at 273, 301, and 239, corresponding to [Cu(dmen)₂(CH₃COO)]⁺, [Cu(dmen)₂(NO₃)]⁺ and [Cu(dmen)₂Cl]⁺, respectively; see Figures S2–S4. The electronic spectra of the solid state of complexes [Cu(dmen)₂(CH₃COO)₂], [Cu(dmen)₂(NO₃)₂], and [Cu(dmen)₂Cl₂] showed broad and asymmetrical absorption bands at 557 nm, 545 nm, and 585 nm, respectively (Figure 1), associated with a distorted octahedral geometry around the metal center [29]. The crystal structure of complex [Cu(dmen)₂(NO₃)₂] confirms the proposed structure, which is similar to that reported earlier by Narayanan and Bhadbhade [30]; see Figure 2 and Tables S1 and S2.

2.1.2. Characterization in Solution

The electronic spectra of all complexes in the aqueous solution showed one absorption band at 570 nm, typical of a d⟶d transition, associated with a square pyramidal geometry; see Figure S5. Furthermore, in the UV region, two absorption bands at 230 nm and 270 nm were detected, typical of the Ligand–Metal Charge-Transfer (LMCT) transitions n → σ* and σ → σ*, respectively. The same spectrum obtained for all complexes suggests an exchange between the counterion of the complexes with a water molecule of the medium, generating a common chemical species [Cu^II(dmen)₂H₂O]²⁺.

2.2. Electrochemical Behavior of Cu-dmen Complexes

2.2.1. Cyclic Voltammetry for Bis(dmen) Complexes

The electrochemical response of bis(dmen)Cu complexes was explored through cyclic voltammetry from open-circuit potential (E_oc) in the negative direction, using a glassy carbon electrode as the working electrode in the presence of 0.1 M KNO₃ in an aqueous solution as the supporting electrolyte. Figure 3a shows a typical cyclic voltammogram at 0.1 V s⁻¹ of the complex [Cu^II(dmen)₂(NO₃)₂], where one reduction signal (I_c) at an E_pc = −0.390 V Ag/AgCl and two oxidation signals (I_a and II_a) at −0.241 and 0.146 V vs. Ag|AgCl can be observed. Furthermore, a pre-wave signal I^* is also detected at −0.023 V vs. Ag|AgCl. A study of inversion of potential E_-λ in the negative direction indicates that all oxidation signals are dependent on the reduction signal I_c. Considering the dependence of signals I_c with I_a, and experiments of cyclic voltammetry at different scan rates, a difference in peak potentials ΔE_p close to 150 mV was calculated. Cyclic voltammograms in a normalized representation (i/ν^1/2), as seen in Figure 3b, exhibited an increase in normalized current for signal I_a and a shift toward more negative potential for signal I_c when the scan rate was increased. This behavior can be related to quasi-reversible electron transfer followed by a coupled chemical reaction. The peak potential values for signal II_a also presented a shift to positive values. This fact also suggests the presence of a coupled chemical reaction with the electron transfer process. The formation of Cu(0) over the electrode must be neglected, as no metallic electrodeposit was detected in cyclic voltammograms after bulk electrolysis experiments. The responses of bis(dmen)Cu complexes display similar behavior among themselves; see Figures S6 and S7; only the response of the complex [Cu^II(dmen)₂(NO₃)₂] is shown in this article. Spectroelectrochemical experiments were performed to establish the location of the electron transfer in the process I_c.

2.2.2. Spectroelectrochemical Characterization

UV–vis spectroelectrochemical experiments for a solution of 1.5 × 10⁻² M of complex [Cu^II(dmen)₂(NO₃)₂] in 0.1 M KNO₃ were performed, imposing a pulse of potential in which the process I_c is limited by diffusion (E₁ = −0.85 V vs. Ag|AgCl) by 5 min, with the simultaneous acquisition of spectra performed every 3 s (see Figure 4a). Immediately, a second potential pulse (E₂ = 0.25 V vs. Ag|AgCl) was imposed for 5 min (see Figure 4b). For the forward pulse, a decrease in the absorbance values for the signal at 570 nm, typical of a d⟶d electronic transition associated with a distorted square pyramidal geometry around the Cu(II) metal center, was recorded. For the backward pulse, the absorbance values for the electronic transition at 570 nm increased until a constant value was reached (Figure 4b). Hence, the process I_c can be attributed to the reduction Cu(II) + 1e⁻ → Cu(I); meanwhile, I_a can be associated with the oxidation Cu(I) → Cu(II) + 1e⁻; for the [Cu^II(dmen)₂(NO₃)₂] complex at 0.1 Vs.⁻¹, E_1/2 was calculated to be −0.32 V vs. Ag|AgCl.

2.2.3. Chronoamperometry

The diffusion coefficient (D_o) was calculated using simple pulse chronoamperometry by applying a potential value of −0.5 V vs. Ag|AgCl, with a perturbation time of 60 s; see Figure S8a. The linear relationship of i vs. t^−1/2 indicated that the process obeys the Cottrell law (Equation (1)), Figure S8b, with the equation i(t) = −0.01004t^−1/2 − 2.941 × 10⁻⁴, where r = 0.99836. From the slope, the diffusion coefficient D_o was calculated to be 5.36 × 10⁻⁶ cm²s⁻¹.

$$\mathrm{i}\left(\mathrm{t}\right)={\mathrm{i}}_{\mathrm{d}}\left(\mathrm{t}\right)={\displaystyle \frac{\mathrm{nFA}{\mathrm{D}}_{\mathrm{o}}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}{\mathrm{C}}_0^{\ast }}{{\mathrm{\pi }}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}{ \mathrm{t}}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}}}$$

(1)

2.2.4. Proposed Mechanism

As demonstrated above, spectroelectrochemical experiments and cyclic voltammetry signals I_c and I_a are related to the redox couple Cu(II)|Cu(I), with a coupled chemical reaction. Meanwhile, the oxidation process II_a and the pre-wave signal for adsorbed spices are related to the oxidation of Cu(I) species generated in the redox process I_c. Considering the previously mentioned evidence and the similarity of voltametric profile obtained in this study with that reported for the ECE-DISP mechanism [31,32,33], the following mechanism is proposed.

[Cu^II(dmen)₂H₂O]²⁺ + e⁻ ↔ [Cu^I(dmen)₂H₂O] ⁺  I     E

[Cu^I(dmen)₂ H₂O] ⁺ ⟶ [Cu^I(dmen)₂]⁺ + H₂O      C

[Cu^I(dmen)₂]⁺ ↔ [Cu^II(dmen)₂]²⁺ + e⁻    I_a    E

[Cu^I(dmen)₂ H₂O]⁺ + [Cu^II(dmen)₂]²⁺ ⟶[Cu^II(dmen)₂ H₂O]²⁺ + [Cu^I(dmen)₂]⁺  DISP

[Cu^I(dmen)₂ H₂O]⁺ _(ads) ⟶[Cu^II(dmen)₂ H₂O]²⁺ + e⁻  I*

2.3. Electrochemical Characterization of Cu-1,10-Phen Complex

Cyclic Voltammetry for bis(1,10-phen)Cu complex

To understand the effect of aliphatic amines on the electrochemical reduction of O₂, a comparison with the complex [Cu^II(1,10-phen)₂NO₃]NO₃ is presented. Cyclic voltammetry at 0.1 V seg⁻¹ initiated from open-circuit potential in the negative direction of this complex is presented in Figure S9a, where one reduction signal I_c with an E_pc = 0.007 V vs. Ag|AgCl and one oxidation signal I_a with an E_pa = 0.123 V vs. Ag|AgCl can be observed, with a difference between peak potentials (ΔE_p) of 116 mV. The fine shape of the wave and the value of E_p/2 = 14.7 mV for I_c indicate that the process has a contribution from adsorbed species at the electrode surface. Cyclic voltammograms at different scan rates in normalized representation (i/ν^1/2) voltammetry, as seen in Figure S9b, show that the normalized peak current and peak potential values for I_c for I_a do not converge, confirming the contribution of adsorbed species in both processes. According to the literature and considering the mentioned results, the signals I_c and I_a can be associated with the redox process [Cu^II(1,10-phen)₂NO₃]⁺_(ads) + e⁻ ↔ [Cu^I(1,10-phen)₂NO₃] _(ads) [34,35].

2.4. Electrochemical Studies in Presence of O₂

2.4.1. Cyclic Voltammetry in Presence of O₂

Figure 5a shows the electrochemical response of O₂ in the presence of 0.1 M KNO₃ (black line), where one irreversible reduction signal III_c with an E_pc = −0.655 V vs. Ag|AgCl and an E_p/2 = 184 mV, associated with the redox process O₂ + 4H⁺+ 4e⁻ ⟶ 2H₂O, was observed. In the same figure, the electrochemical response of the complex [Cu^II(dmen)₂(NO₃)₂] in the absence (red line) and in the presence of molecular oxygen (blue line) is presented. It is observed that, when oxygen is present in a solution (blue line), the signal II_c is not detected, and the signal I_c shows an increase in its current value, with an E_pc = −0.390 V vs. Ag|AgCl, related to the electrochemical reduction in Cu(II) to Cu(I), which is shifted by 10 mV toward a positive value in comparison to when oxygen is absent. These changes are associated with a typical molecular catalysis that occurs during O₂ reduction, where the electrogenerated Cu(I) species acts as a redox mediator, which reduces the O₂ molecule with a homogeneous electron transfer mechanism. When the experiments were carried out at a variable scan rate (Figure 5b), the signal I_a did not show an increase in its normalized current value, and the potential values of signal II_a were shifted toward more positive values when the scan rate was increased, which is typically observed in a coupled chemical reaction. Moreover, in the presence of oxygen, the signal I^* is not detected at any scan rate. The inversion potential analysis confirms the dependencies of signals I_c and II_a; see Figure S10. For all the other bis(dmen)Cu complexes, similar responses were recorded; see Figure S11.

In the case of the complex [Cu^II(1,10-phen)₂NO₃]NO₃, a signal for O₂ reduction is not observed at −0.62 V vs. Ag|AgCl (IIIc) (Figure S12a). For the process I_c, a low increase in current is recorded, which is associated with the low molecular catalysis of oxygen reduction at the surface electrode, due to the already mentioned adsorption process (Figure 6a). A new signal is recorded at −0.26 V vs. Ag|AgCl for the desorption processes of Cu(II) complexes promoted by oxygen and can be attributed to the molecular catalysis of O₂ reduction in the solution (Figure 6a).

Experiments at a variable scan rate, as seen in Figure S12b, show changes in potential peaks of signals I_c and II_c to negative values and a low increase in normalized current (i/v^1/2) when the scan rate is increased and recorded. For the signal I_c, a shift toward positive values is observed when the scan rate is increased. A comparison of electrochemical responses of [Cu^II(dmen)₂(NO₃)₂] and [Cu^II(1,10-phen)₂NO₃]NO₃, as seen in Figure 6a,b, indicates a higher performance for oxygen reduction for the complex [Cu^II(dmen)₂(NO₃)₂], due to the presence of 1,10-phenanthroline in [Cu^II(1,10-phen)₂NO₃]NO₃ that promotes the adsorption of the complex at the electrode surface, decreasing its ability to carry out the molecular catalysis of oxygen reduction.

2.4.2. Spectroelectrochemical in Presence of O₂

Due to the adsorption of the [Cu^II(1,10-phen)₂(NO₃)]NO₃ complex on the electrode surface, only electronic absorption behavior during the ORR for the bis(dmen)Cu complex was explored with UV–vis spectroelectrochemical characterization, using a 1.5 × 10⁻² M solution of the complex [Cu^II(dmen)₂(NO₃)₂] in 0.1 KNO₃ in the presence of O₂. Imposing a first potential E₁ = −0.85 V vs. Ag|AgCl in the first 5 min and, immediately, a second pulse E₂ = 0.25 V vs. Ag|AgCl, also for 5 min, spectra were acquired simultaneously every 3 seg (Figure S13). In the absence of O₂, before imposing any pulse potential, Cu complexes exhibit a well-defined absorbance band at 570 nm associated with a d→d electronic transition, corresponding to Cu⁺²; in the presence of O₂, the absorbance band remains at 570 nm, indicating a higher cathodic potential of E_pc; no intermediated species are formed. Figure 7 displays Abs vs. t at a fixed wavelength of 570 nm, both in the absence and presence of O₂ at E₁ = −0.85 V vs. Ag|AgCl; see only the redox reaction involving the consumption of Cu²⁺ to Cu⁺. The same UV–vis spectroelectrochemical response obtained with and without O₂ indicated a direct electron transfer from Cu(II) to Cu(I) and the regeneration of catalyst without any intermediate species.

2.4.3. Proposed Mechanism for O₂ Reduction

For bis(dmen)Cu complexes, the following mechanism is proposed, which is similar to that presented in the absence of oxygen, except the adsorption process and the molecular catalysis EC_i’.

[Cu^II(dmen)₂H₂O]²⁺ + e⁻ ↔ [Cu^I(dmen)₂ H₂O]⁺        E

O₂ + 4H⁺ + 4[Cu^I(dmen)₂ H₂O]⁺⟶ 4[Cu^II(dmen)₂ H₂O]²⁺ + 2H₂O    C_i^’

Meanwhile, for the bis(1,10-phen)Cu complex, the following mechanism is proposed.

[Cu^II(1,10-phen)₂NO₃]⁺_(ads) + e⁻ ↔ [Cu^I(1,10-phen)₂NO₃] _(ads) I_c

O₂ + 4H⁺ + [Cu^I(1,10-phen)₂NO₃] _(ads) ⟶ 4[Cu^II(1,10-phen)₂NO₃] _(ads) + 2H₂O     C_i^’

[Cu^II(1,10-phen)₂NO₃]⁺ + e⁻ ↔[Cu^I(1,10-phen)₂NO₃]      II_c

O₂ + 4H⁺ + 4[Cu^II(1,10-phen)₂NO₃] ⟶ 4[Cu^I(1,10-phen)₂NO₃]⁺ + H₂O    C_i^’’

Based on the above discussion, it can be proposed that the presence of aromatic system in the 1,10-phenanthroline ligand promotes the adsorption of molecular catalysts, decreasing the electrocatalytic activity.

2.4.4. Foot of the Wave Analysis (FOWA) for O₂ Reduction

Considering a 4e⁻ mechanism for O₂ reduction using the CV response of the complex [Cu(dmen)₂(NO₃)₂] in the presence of molecular oxygen, a FOWA was carried out to calculate the homogenous electron transfer rate constant k_obs, using the following equation [36]:

$${\displaystyle \frac{{\mathrm{i}}_{\mathrm{c}}}{{\mathrm{i}}_{\mathrm{p}}}}={\displaystyle \frac{2.24\mathrm{n}\sqrt{\frac{\mathrm{R}\mathrm{T}{\mathrm{k}}_{\mathrm{o}\mathrm{b}\mathrm{s}}}{\mathrm{F}\mathrm{\nu }}}}{1+{\mathrm{\epsilon }}_1}}$$

where i_c and i_p are the catalytic current and the diffusion current, respectively; the term ${\epsilon }_1=\mathrm{e}\mathrm{x}\mathrm{p}\left[{\displaystyle \frac{R}{RT}}\left(E-E_{{\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}}\right)\right]$ contains the half-wave potential E_1/2 for the catalytic wave; and n refers to the number of transferred electrons. According to the above reaction, the k_obs = TOF value of the 4e⁻ reduction in dioxygen was calculated from the slope m = 2.24n$\sqrt{{\displaystyle \frac{RT{k}_{obs}}{F\nu }}}$ of the plot i_c/i_p,red vs. 1/(1 + ε₁) (see Figures S14–S16), where capacitance contribution does not compromise analysis [37]. Hence, linear regression in the range of 10⁻⁵ < 1/1 + ε₁ < 0.001 was used. The resulting values for the 4e⁻ reduction of O₂ to H₂O mediated by [Cu(dmen)₂(CH₃COO)₂]²⁺, [Cu(dmen)₂(NO₃)₂] ²⁺, and [Cu(dmen)₂Cl₂] ²⁺ at 0.1 V Vs.⁻¹ were 4.84 × 10³ s⁻¹, 1.36 × 10⁴ s⁻¹, and 5.39 × 10³ s⁻¹, respectively.

Table 1. The ORR activity of few Cu(II) complex catalysts.

Catalyst	Cond [pH/Electrode]	E1/2 CuII|CuI	TOF [s−1]	Reference
[Cu(dmen)2(CH3COO)2]	aqueous solution, glassy carbon	−0.326 V vs. Ag|AgCl	4.8 × 103	this work
[Cu(dmen)2(NO3)2]	aqueous solution, glassy carbon	−0.32 V vs. Ag|AgCl	1.3 × 104	this work
[Cu(dmen)2Cl2]	aqueous solution, glassy carbon	−0.324 V vs. Ag|AgCl	5.3 × 103	this work
[Cu(1,10-phen)2(NO3)2]NO3	aqueous solution, glassy carbon	N.R.	N.R.	this work
[Cu(tmpa)(L)]2+	pH 7.0 PB, glassy carbon	0.002 V vs. Ag|AgCl	1.8 × 106	[18]
Cu-phthalocyanine	Au(111) electrode	N.R.	N.R.	[15]
[Cu(1,10-phen)]	pH 7.0 PB, Au electrode	N.R.	N.R.	[16]
[Cu(tpmen)](ClO4)2	pH 7.0 Pb, glassy carbon	0.002 V vs. Ag|AgCl	6.5 × 104	[19]
[Cu(pmea)(L)]2+	pH 7.0 PB, glassy carbon	0.162 V vs. Ag|AgCl	1.4 × 103	[38]
[Cu(tren)(L)]2+	rotating electrode glassy carbon	N.R.	N.R.	[39]
[Cu(tren)(ImH)](ClO4)2	pH 7.0 PB, pyrolytic graphite	N.R.	N.R.	[24]
[Cu(trpn)(ImH)](ClO4)2	pH 6.4 Britton–Robinson, pyrolytic graphite	N.R.	N.R.	[23]

N.R.—“not reported”.

shows parameters of Cu(II) complexes related to the molecular catalysis in the ORR, including other examples taken from the literature. It can be observed that E_1/2 is an indicator of the viability of a mediator to catalyze the oxygen reduction reaction; the smaller the overpotential of E_1/2, the more feasible it is to carry out the 4e⁻ reduction of O₂ to H₂O. The bis(dmen)Cu complexes reported in this study showed the lowest value of E_1/2 for the oxygen reduction reaction (Table 1). The fact that catalytic data for some Cu complexes listed in Table 1 are not reported makes it difficult to compare with the studied systems. Furthermore, it is expected that both the ligands pyridylalkylamine and alkyl amines in the copper complexes favor a flexible pentacoordinate geometry, which facilitates the interaction between O₂ and the Cu center, thus improving the electron transfer. However, in the case of the N,N-dimethylethylenediamine complexes, the presence of ECE and disproportion reactions compete with redox mediation, decreasing the TOF value of catalysis. The dmen ligand is an inexpensive and available option for molecular catalysis in the ORR, in contrast to more sophisticated tetradentate ligands such as tmpa or tpmen, making the complexes more accessible and economically viable for large-scale catalytic applications. Although tmpa and tpmen complexes achieve higher intrinsic turnover frequencies, the bis(dmen)Cu systems reported in this study combine reasonable activity with ease of synthesis, operational robustness in simple aqueous media, and compatibility with low-cost electrodes. These features position them as attractive candidates for scalable oxygen reduction electrocatalysis.

2.5. Chronoamperometry in Presence of O₂

To establish the changes in potential when oxygen reduction takes places in the redox mediation by copper complexes, current-sampled versus potential plots I(t)–E were constructed from single-step chronoamperometric experiments at different potentials. Figure 8a shows a typical I(t)–E graph for oxygen in a 0.1 KNO₃ solution, where the broad shape of the wave from 0.5 to −0.5 V vs. Ag|AgCl with a E_1/2 =−0.44 V vs. Ag|AgCl indicates slow electrode kinetics for oxygen reduction. A plateau current with a value close to 1.9 µA was detected from −0.6 V vs. Ag|AgCl, which indicates that the current is limited by the mass transport from this potential value. In the case where the compound [Cu(dmen)₂(NO₃)₂] is in a 0.1 KNO₃ solution, as seen in Figure 8b, the I(t)–E(t) profile indicates a fast electrode kinetic process with a E_1/2 = −0.32 V vs. Ag|AgCl and a current plateau of 2.0 µA, recorded from −0.4 V vs. Ag|AgCl. When [Cu(dmen)₂(NO₃)₂] and oxygen are in a solution (see Figure 8c), a reversible wave is observed with a E_1/2 = −0.24 vs. Ag|AgCl and a limited current of −5.0 µA from −0.4 V vs. Ag|AgCl. A comparison of Figure 8a,c indicates a diminution of 0.2 V and an increase of 3.0 µA in the limited current for oxygen reduction in the presence of the complex [Cu(dmen)₂(NO₃)₂], and this can be used as a criterion to assess whether this electrochemical measurement is improved in a time window larger than that used in cyclic voltammetry.

2.6. Electrochemical Impedance Spectroscopy Studies

The electrode–electrolyte interface phenomena of mononuclear Cu(II) complexes were studied using EIS in the absence and presence of O₂ to identify parameters related to the ORR. Experiments were carried out at a more cathodic potential than the E_pc for the reduction in complexes, E_DC = −0.5 V vs. Ag|AgCl, with a frequency range from 1.2 kHz to 1 Hz.

Characteristic Nyquist diagram and Bode-module and Bode-phase representations are shown in Figure 9a–c. All spectra were fitted in the equivalent circuit (EC) inset in Figure 8a. Instead of an ideal capacitance, a constant phase element (CPE, Q₁) was used to compensate surface roughness and other surface inhomogeneities, using the expression Z = [Y(jω)ⁿ]⁻¹ where n ≈ 1 [40]. For the diffusional phenomenon, a constant phase element (CPE, Q₂) with the same expression and n ≈ 0.5 was considered [41,42]. R₁ and R₂ correspond to solution resistance (R_s) and charge-transfer resistance (R_ct), respectively (see SM, Table S3).

In the Nyquist representation, in the absence of O₂, semicircles are not well-defined in comparison to those in experiments where O₂ is present (Figure 9a). In both conditions, the Bode-module representation (Figure 9b) showed a charge-transfer reaction (at frequencies from 1.2 kHz to 50 Hz) and diffuse behavior (at frequencies from 15 Hz to 1 Hz). On the other hand, the Bode-phase representation (Figure 9c), in a low-frequency system, tends to be 45°, corresponding to diffusional behavior. In the absence and presence of O₂, R_ct for [Cu(dmen)₂(NO₃)₂] decreased from 36.391 kΩ to 5.54 kΩ. This decrease is attributed to the molecular catalysis of O₂ reduction mediated by [Cu^I(dmen)₂ H₂O]⁺. The EIS results of the other Cu(II) complexes with the dmen ligand display their similar behavior, as well as those of the CV experiments. Hence, only the results of complex 2 are shown, and the rest can be found in SM, Figures S17 and S18.

Finally, the EIS response for the ORR with the [Cu(1,10-phen)₂NO₃]NO₃ complex was also studied. It was different from that of Cu-dmen complexes in the absence of O₂; the spectra of [Cu(1,10-phen)₂NO₃]NO₃ in the Nyquist representation showed a well-defined semicircle; see SM Figure S19. Spectra were fitted to the same equivalent circuit; see the inset of Figure S20. However, the low values of Q₁ in comparison to those obtained with Cu(II) complexes can be associated with the adsorption processes [43,44]. In the absence and presence of O₂, a small change in R_ct from 2.51 kΩ to 2.38 kΩ could be related to the low molecular catalysis of O₂ reduction.

3. Materials and Methods

3.1. Materials and Physical Measurements

All chemicals and solvents were purchased from the commercial source and used without further purification: N,N’-dimethylethylenediamine, dmen (95%, Sigma-Aldrich, St. Louis, MO, USA), Cu(CH₃COO)₂ H₂O (98+%, Strem Chemicals, New Buryport, MA, USA), CuCl₂ • 2 H₂O (99.2%, J. T. Baker, Radnor, PA, USA), Cu(NO₃)₂ 2.5 H₂O (99.99%, Sigma-Aldrich, St.Louis, MO, USA), KNO₃ (99.4%, Fermont, Moterrey, Mexico), KBr (>99%, Sigma-Aldrich, St.Louis, MO, USA), diethyl ether (water < 0.01%, J.T. Baker, Radnor, PA, USA), N,N’–dimethylformamide (99.95%, water < 0.05%, J. T. Baker, Radnor, PA, USA), and acetonitrile anhydrous (water < 0.001%, Sigma Aldrich, St. Louis, MO, USA).

The IR spectra were obtained using a Shimadzu FT-IR IRAffinity-1S spectrophotometer, in a range between 450 and 4000 cm⁻¹ using KBr disks. UV–vis spectra were obtained with a Thermo Scientific Evolution Array Spectrophotometer with a spectral range of 200–1100 nm in water and in 0.1 M KNO₃ aqueous solutions, using quartz cells with an optical path length of 1 cm. The solid-state UV–vis spectra were obtained using a Shimadzu 2600-UV–vis Spectrophotometer, in a range between 200 and 900 nm. Molar conductivity was measured using a Metrohm conductivity module, model 856, with a cell having a cell constant C = 0.613 cm⁻¹.

3.2. Synthesis of Complexes

Synthesis of complex [Cu(dmen)₂(CH₃COO)₂] H₂O: A 1 mmol quantity of Cu(CH₃COO)₂ H₂O (185 mg) was dissolved in 10 mL of MeCN. Then, 2 mmol of dmen (185 mg), dissolved in 10 mL of MeCN, was added to the copper solution. The resulting reaction mixture was stirred for 30 min at room temperature (25 °C), forming a purple precipitate, which was collected by filtration, washed with MeCN, and ground to obtain a dark purple powder. Characteristic IR bands (cm⁻¹): 2872, 2968 ν(C-H)N-CH₃; 2900 ν(C-N)R-N-CH₃; 2851, 2954 ν(C-H)R-CH₂-N; 2838, 2986 ν(C-H)R-CH₂-NH₂; 3096, 3191 ν(N-H)R-CH₂-NH₂; 1656 δ(NH₂)R-CH₂-NH₂, 1464 δ(CH₂), 1410, 1572 ν(COO⁻); 2943 ν(C-H)CH₃COO; 457, 494, 579 ν(Cu-N). UV–vis peaks in the aqueous solution λ_máx/nm (ε/Lmol⁻¹ cm⁻¹): 264 (238), 329 (242), 575 (94). Conductivity measurement (⋀/Ω⁻¹cm² mol⁻¹): 170.9. Mass spectrum data FAB (+) m/z: 273, [Cu(dmen)₂(CH₃COO)]⁺, m/z: 239, [Cu(dmen)₂]⁺.

Synthesis of complex [Cu(dmen)₂(NO₃)₂] H₂O: A 2 mmol quantity of dmen (185 mg) and a 1 mmol quantity of Cu(NO₃)₂ •H₂O (187 mg) were dissolved separately in 10 mL of MeCN. The ligand solution was added to the metallic salt solution. The resulting reaction mixture was stirred for 30 min at room temperature (25 °C) to obtain a purple precipitate. Then, the powder was washed with MeCN and ground. Crystals suitable for the X-ray analysis were obtained with the diethyl ether diffusion method using a mixture of DMF/MeOH. Characteristic IR bands (cm⁻¹): 2873, 2969 ν(C-H)N-CH₃; 2900 ν(C-N)R-N-CH₃; 2857, 2988 ν(C-H)R-CH₂-N; 2839, 2954 ν(C-H)R-CH₂-NH₂; 3097, 3193 ν(N-H)R-CH₂-NH₂; 1640 δ(NH₂)R-CH₂-NH₂, 1460 δ(CH₂), 1414, 1313 ν(NO₂)NO₃; 458, 505, 572 ν(Cu-N). UV–vis peaks in the aqueous solution λ_máx/nm (ε/Lmol⁻¹ cm⁻¹): 268 (241.63), 330 (245.53), 575 (100.01). Conductivity measurement (⋀/Ω⁻¹cm² mol⁻¹): 259.7. Mass spectrum data FAB (+) m/z: 301, [Cu(dmen)₂(NO₃)]⁺, m/z: 239, [Cu(dmen)₂]⁺.

Synthesis of complex [Cu(dmen)₂Cl₂] H₂O: A 2 mmol quantity of the ligand dmen (185 mg) dissolved in MeCN was added to 1 mmol CuCl₂ • 2H₂O (170 mg) previously dissolved in 10 mL of MeCN. The reaction mixture was stirred for 30 min at room temperature (25 °C), forming a blue precipitate, which was collected by filtration, washed with MeCN, and ground into a dark blue powder. Characteristic IR bands (cm⁻¹): 2874, 2971 ν(C-H)N-CH₃; 2899 ν(C-N)R-N-CH₃; 2841, 2988 ν(C-H)R-CH₂-N; 2801, 2955 ν(C-H)R-CH₂-NH₂; 3094, 3188 ν(N-H)R-CH₂-NH₂; 1640 δ(NH₂)R-CH₂-NH₂, 1465 δ(CH₂); 403 ν(M-Cl); 458, 492, 584 ν(Cu-N). UV–vis peaks in the aqueous solution λ_máx/nm (ε/Lmol⁻¹ cm⁻¹): 269 (241.54), 330 (243.22), 575 (98.72). Conductivity measurement (⋀/Ω⁻¹cm² mol⁻¹): 271.3. FAB mass spectrum (+) m/z: 274, [Cu(dmen)₂(Cl)]⁺, m/z: 239, [Cu(dmen)₂]⁺.

Synthesis of the complexes [Cu(1,10-phen)₂NO₃]NO₃·H₂O: The synthesis of this metal complex was carried out by dissolving 0.1 mmol of Cu(NO₃)₂·2.5 H₂O in 5 mL of anhydrous ethanol. Then, this solution was added dropwise to the metallic salt solution of 0.2 mmol of the ligand 1,10-phenanthroline, previously dissolved in 5 mL of anhydrous ethanol. The ligand solution was heated at 70 ^○C for complete dissolution; a change from blue to green was observed. The reaction mixture was heated and stirred for 2 h. Solvent was removed by slow evaporation until a powder was observed. The product was filtered and washed with diethyl ether. Characteristic IR bands (cm⁻¹): 1627, 1520, ν(C=C) + ν(C=N); 3057ν(=C-H); 1605, 1586, 1493, ν(C=C)_ring; 1315 ν_s(N=O); 1330 ν_a(NO₂); 1383 NO₃⁻ ionic, ν(Cu-N) 431, ν(Cu-O) 4010. Anal. Calc. for CuC₂₄H₁₈N₆O₇ (MW = 565 g/mol) %C, 50.9; %H, 3.2; %N, 14.8. Found: %C, 50.9; %H, 3.6; %N, 15.0. Mass spectrum FAB(+) m/z = 485, [Cu(1,10-phen)₂NO₃]⁺, m/z = 423 [Cu(1,10-phen)₂]⁺.

3.3. Single-Crystal XRD Analysis

A single crystal of [Cu(dmen)₂(NO₃)₂] was mounted on a glass fiber and analyzed using an Oxford Diffraction Gemini “A” diffractometer, equipped with a charge-coupled device area detector, an X-ray tube (λ_MoKα = 0.71073 Å), and a graphite monochromator. The software packages CrysAlis PRO and CrysAlis RED version 171.44 were used for data collection and integration. The data were corrected for absorbance using an analytical numerical correction [45]. Structure solution and refinement were carried out using the software Olex2 [46]. Graphics were obtained using the software Mercury 4.0 and Olex2 [47]. Full-matrix least-squares refinement was performed by minimizing (Fo² − Fc²)². All non-hydrogen atoms were refined anisotropically. Hydrogen atoms of the water molecules (H–O) were located with a difference map and refined isotropically with Uiso(H) = 1.5 for H–O. H atoms attached to C atoms were placed in geometrically idealized positions, with C–H = 0.95 and 0.98 Å and U_iso(H) = 1.2U_eq(C) and 1.5U_eq(C) for aromatic and methyl groups, respectively. The crystal data and experimental details of the structural determinations are listed in Tables S1 and S2 of Supplementary Materials. Crystallographic data have been deposited in the Cambridge Crystallographic Data Center (CCDC) with supplementary material number CCDC 125785. Copies of the data can be obtained free of charge upon application to CCDC: 12 Union Road, Cambridge CB2 1EZ, UK (email: deposit@ccdc.cam.ac.uk).

3.4. Electrochemical Experiments

Electrochemical measurements were carried out with a Biologic SP-300 potentiostat/galvanostat using the software EC-Lab^® V11.02. A typical three-electrode array was used with a glassy carbon electrode (Ø = 3 mm), a Pt wire, and an Ag|AgCl electrode (KCl 3 M), as the working, counter, and reference electrodes, respectively. All experiments were carried out in the presence of 0.1 M KNO₃ in an aqueous solution as the supporting electrolyte. The working electrode was manually polished with an alumina suspension (0.3 μm), sonicated for 90 s, and rinsed with deionized water. Before each measurement, the solution was bubbled for 5 min and maintained under a nitrogen atmosphere during the experiment. Ohmic drop (IR) was corrected by positive feedback compensation. Cyclic voltammetry experiments were started for open-circuit potential (E_oc) in the negative direction at different scans rates from 100 to 1000 mVs.⁻¹. One-step chronoamperometric experiments were carried out imposing pre-polarization at open-circuit potential for 20 s, followed by first-step potential at different values more negative than E_oc for 60 s. Electrochemical impedance spectroscopy experiments were acquired using a direct potential value more cathodic than the E_pc for the reduction in complexes with a stabilization time of 5 min and applying a sinusoidal perturbation E_AC = 10 mV with a frequency range from 100 kHz to 1 Hz and 10 points per decade.

3.5. Spectroelectrochemical Studies

Spectro electrochemical experiments were performed with a BioLogic SP-300 potentiostat/galvanostat coupled to a diode array UV–vis spectrophotometer (Thermo-Scientific-Evolutions Array). A Pine commercial platinum honeycomb spectroelectrochemical cell with a Ag|AgCl(KCl 3M) reference electrode was used. The concentration of the complexes was 1.5 × 10⁻² M in the presence of 0.1 M KNO₃ in an aqueous solution. Spectra were acquired every 10 sec, and simultaneously, a double-potential-step chronoamperometry sequence was imposed (E₁: −0.85 V vs. Ag|AgCl and E₂: 0.25 V vs. Ag|AgCl (KCl 3 M and t₁ = t₂ = 5 min)).

4. Conclusions

Mononuclear Cu^II complexes containing the dmen ligand used to catalyze the oxygen reduction reaction are no longer studied due to the slow electronic transfer interaction between the Cu(II) center and O₂ during the reduction reaction, compared to other Cu(II) complexes with pyridylalkylamine ligands. However, we verified the viability of bis(dmen)Cu complexes as ORR catalysts, specifically [Cu(dmen)₂(NO₃)₂] with a TOF = 1.36 × 10⁴ s⁻¹; although they were not as efficient as the complex [Cu(tmpa)(L)]²⁺ with a TOF = 1.8 × 10⁶ s⁻¹, their efficiency is comparable to that of other complexes with pyridylalkylamine ligands such as [Cu(tpmen)(ClO₄)₂] with a TOF = 6.4 × 10³ s⁻¹.

More research needs to be conducted in the future on Cu(II) complexes with these types of ligands as oxygen reduction catalysts for increasing their catalytic capability and defining their limits. We also believe that this study can bring into focus a long-abandoned area of research and expand it.