1. Introduction
Fuel cells, batteries, and supercapacitors are known as the most efficient and promising technologies for electrochemical energy conversion and storage. The main constraints for their production for practical use are primarily due to the high cost of materials, as follows: e.g., noble metals such as platinum, as well as rather poor operational stability in different climate environments. The long-lasting challenge is to provide an advanced low-cost catalyst with performance as efficient as noble metals or better. Therefore, the rational design of novel multicomponent materials with improved catalytic properties can lead to alternatives that can compete with existing technologies.
Alternative energy devices are one of the most efficient and clean energy-producing sources available [
1]. Still, the sluggish kinetics of the electrochemical oxygen reduction reaction (ORR) is detected as the main limitation. Noble metals (e.g., Pt, Au) are used and considered the most efficient catalysts for their performance, but at the same time, due to their high cost and scarcity, commercialization of these energy devices is impeded [
2,
3,
4]. In the last decades, the main idea has been to reduce the loading of noble metals in the electrode material but to keep the same activity that their bulk electrode provides for the desired reactions.
In particular, micro/nanomaterials with a suitable combination of the components were investigated as a promising strategy toward high-performance electrode materials. For example, transition metal oxides (TMO) such as MnO
2, Co
2O
3, etc. have been studied, due to their rich oxidation states, high specific capacitances, low cost, and environmental friendliness [
5,
6]. Among TMOs, the focus is mostly based on MnO
2, which is already indicated as a non-carbon material of potent catalytic activity improvements while being used as a noble metal carrier [
7]. It has been reported that manganese oxide promotes oxygen disproportionation, resulting in an overall four-electron reduction of O
2 on MnOx/Nafion modified electrodes [
8]. Unfortunately, the electrochemical performance of single MnO
2 electrodes is limited by poor electrical conductivity (10
−5–10
−6 S cm
−1) [
9,
10]. An effective way to improve the utilization of MnO
2 is to combine conductive materials (metals, oxides, carbon-based materials) to form composites as electrode materials with improved characteristics. For example, Xia et al. have reported the fabrication of composite materials with core-shell structures of Co
2O
3@Pt@MnO
2 as high-performance electrode materials for supercapacitors [
10]. Lahousse et al. have found an MnO
2/TiO
2-Pt electrocatalyst of high-measured catalytic activity in comparison to the noble metal catalysts [
11]. Various combinations of mixed oxide materials have been investigated so far, e.g., 3D flower-like Co
2O
3@MnO
2 [
6], nanosheets of MnO
2 [
12], CuO/MnO
2 [
13], NiO/MnO
2 [
14], etc.
Additionally, among rare earth oxides, lanthanum oxide (La
2O
3) has been extensively analyzed for its excellent electronic, photocatalytic, and electrocatalytic properties [
15,
16]. It has been discovered that the existence of the interlayer defect structure of the oxides was helpful for the active oxygen adsorption and had a positive effect on catalyzing ORR [
17]. On the other hand, despite its outstanding electronic structure, quite similar to MnO
2, the single oxide has low electroconductivity. Therefore, lots of effort has been put into the development of a hybrid composite material that would combine all the benefits of the metals. In this regard, perovskite oxides with an ABO
3 structure have attracted great attention as a replacement metal oxides. Various perovskites have been developed, such as LaMnO
3, LaCoO
3, LaNiO
3, LaCrO
3, and LaFeO
3 [
18,
19,
20]. In particular, Co and Mn-based perovskites have shown high and stable performance for ORR as well as for the oxygen evolution reaction (OER). This bifunctional property is due to its oxidation states and spin configuration [
21]. Recently, LaMnO
3 nanoparticles were found to achieve high energy and power densities, synthesized using natural lemon juice as a green surfactant [
22]. However, due to various structures, a combination of metals, and the synthesis procedure reported, the systematic design and comparison of metal oxide-based electrode materials is necessary.
The main aim of this paper was a basic analysis and comparison of the lanthanum and manganese-based hybrid materials for ORR of reduced Pt loading. In addition, the utilization of cheap and stable carrier materials (e.g., TMOs) were used, which were supposed to contribute to the activity of a reduced number of noble metals.
The syntheses of hybrid electrocatalytic materials with detailed electrochemical characterization were conducted. The electrode materials based on La2O3/MnO2 (LMO) have been synthesized by one-pot synthesis, and the influence of the addition of Pt low loading has been investigated. Transition metal oxides were chosen as base support due to their double-role potential, one of which was to serve as stabilizing catalyst carriers while the other one had interactive participation in the enhancement of the catalytic properties of noble metals. Electrochemical characterizations have been carried out by means of cyclic voltammetry (CV), linear sweep voltammetry (LSV), and electrochemical impedance spectroscopy (EIS) in 0.1 M KOH. ORRs were investigated in the 3-electrode set-up using a rotating disk (RDE) as a working electrode to provide controlled mass transport conditions.
2. Results and Discussion
Stable cyclic voltammograms (CVs) of La
2O
3, MnO
2, LMO, and LMO–Pt thin-layer electrodes are shown in
Figure 1.
There are no significant differences in the La2O3, LMO, and LMO-Pt CV responses. MnO2 exhibits slightly higher cathodic currents in the range from 1.0 to −0.2 V, which meets the value for LMO–Pt material at E = −0.2 V. For an LMO–Pt sample, the onset of the hydrogen evolution reaction is clearly visible around 0.0 V due to the presence of Pt. Although being quite weak, the broad redox pair peak around 0.6 V can be distinguished for La2O3, MnO2, and LMO-Pt. The only anodic shoulder-like peak that can be distinguished for LMO.
MnO2 cyclic voltammogram in the potential range of −0.2 V to 1.6 V in 0.10 M KOH electrolyte represents a typical voltammogram with a wide reduction peak related to the reduction reactions of MnO2 to Mn2O3, Mn3O4, and finally to Mn(OH)2. The anodic counterpart appears as the increase in currents at the potentials above 1.1 V. The broad peaks of other materials, La2O3, LMO, and LMO-Pt, belong to mutual redox reactions of rare earth metals and Mn oxides. These voltammograms show forms that are typical for the materials that could be associated with the pseudocapacitive response of these faradaic processes. The absence of ideal linearity and symmetry between the charge-discharge processes can be observed as wider discharge counterparts.
The prepared electrocatalytic materials were tested for their ORR activity by means of LSV in an O
2-saturated 0.10 M KOH electrolyte.
Figure 2 presents the ORR electrochemical performances of La
2O
3, MnO
2, LMO, and LMO-Pt thin powder layers at the electrode rotating speed of 1000 rpm.
Noticeable reduction currents can be observed as a result of ORR. The ORR onset potential for MnO
2, LMO, and LMO-Pt electrodes was approx. 0.7 V, which is comparable to previous work [
23], while the onset potential for the La
2O
3 electrode was around 0.8 V. The MnO
2, LMO, and LMO-Pt electrode materials show similar curve shapes and activity. For all of the materials, the curve appears stepped, with a transition around 0.4 V. This could also be an indication of different ORR mechanisms, with La
2O
3 electrode material being quite a different mechanism with respect to MnO
2, LMO, and LMO-Pt electrode materials. It can be noticed that LMO electrode material has a more similar behavior to La
2O
3 electrode material in the potential range from 0.3 V to −0.2 V, while the former exhibits the highest activity below ca. 0.0 V. MnO
2 and LMO-Pt have quite similar curve shapes in the mentioned potential range. The MnO
2 electrode material has the smallest ORR activity amongst three named electrodes with a current plateau of around −1.9 mA cm
−2, followed by LMO (−2.0 mA cm
−2) and, finally, LMO with 5 mass% of Pt (−2.3 mA cm
−2). The La
2O
3 electrode material has a different shape of polarization curve. This indicates that different ORR mechanisms could be taking place on the La
2O
3 thin layer; the ORR starts at more positive potentials and, consequently, the overpotential is lower at the potentials positive to ca. 0.65 V). This feature is similar to the ORR activity of commercial 20 mass% Pt@XC-72 catalysts [
23,
24]. However, the ORR current plateau is considerably lower (−1.3 mA cm
−2) with respect to other samples.
Tafel analysis is generally utilized to compare the electrocatalytic activity of the electrodes and to discriminate the reaction mechanisms of the reactions with mass transfer limitations at low to moderate overpotentials.
Ohmic-drop-compensated pоtentiodуnаmic Tafel сurvеs for the ORR at La
2O
3, MnO
2, LMO, and LMO-Pt materials in O
2-saturated 0.10 M KOH are shown in
Figure 3 (the overall current is examined since well-developed diffusion current density plateaus have not been registered to apply the typical Koutecky–Levich analysis).
It can be seen in
Figure 3 that all the electrode materials exhibit Tafel slope at low overpotentials (up to 100 mV) between 53 and 60 mV dec
−1 (lines). Tafel analysis leads to the following two important physical parameters: the Tafel slope and the exchange current density. Based on the theory of ORR mechanisms, the Tafel slope of 60-mV dec
−1 determines the overall reaction rate when the surface adsorption of the species MOO and MOO− is most likely the rate-determining step [
25,
26]. One can observe that, in the measured overpotential range, La
2O
3 material has the highest currents. At lower overpotentials (0.8–0.9 V potential region), MnO
2 has one of the highest currents among Mn-containing samples; at higher overpotentials (at potentials negative to ca. 0.75 V), the highest ORR currents are registered for Pt-free Mn oxide-based materials. In this high-overpotential range, LMO-Pt’s activity appears to approach the activities of MnO
2 and LMO. Samples are of considerably lower ORR activity than bulk Pt (the Tafel slope of 52 mV dec
−1 was registered (not shown)). The lower activity could be attributed to the fact that there is only 5 mass% of Pt in the LMO-Pt electrolytic material, which is quite low content to induce the ORR activities close to bulk Pt. Moreover, from
Figure 2, it can be seen that La
2O
3 electrolytic material loses considerably its highest low-overpotential activity and reaches the lowest limiting currents.
The current densities per mol of a metal could be a useful indication of the specific activity of the components in the composites [
23]. This parameter was calculated at the potential in the region of calculated slopes of Tafel dependencies, and the results are shown in
Table 1.
It can be concluded that La improves the activity of Mn at lower overpotentials but not at higher ones.
To further analyze the ORR activities and their correlation to sample structures, potentiostatic electrochemical impedance spectroscopy (PEIS) measurements were performed at open circuit potential for all the samples in an N2-saturated 0.1 M KOH solution. Galvanostatic electrochemical impedance spectroscopy (GEIS) measurements were performed at OCP and the current from a low- to moderate-overpotential region of −0.765 mA cm−2 (with an amplitude of 0.510 mA cm−2) for all the samples in an O2-saturated 0.1 M KOH solution in order to evaluate the details of ORR activity and the electrochemical behavior of the tested materials.
The PEIS results, shown as the capacitance complex plane plots (CCPPs) of the synthesized materials, are given in
Figure 4a. CCPPs show differences in capacitive features between the tested electrodes. The mutual characteristic is a well-resolved capacitive loop in the high-frequency domain followed by straight-line behavior in the low-frequency domain. The capacitive loop for the Pt-modified LMO electrode is of a higher diameter than LMO itself, and the highest among all of the samples, meaning the capacitance is considerably increased by Pt–modification. Samples involving single oxides–MnO
2 and La
2O
3 exhibit similar EIS features of two indicatively overlapped capacitive loops of similar diameters. The region consisting of a loop-like response at high frequencies corresponds to pseudocapacitive redox transitions of a metal ion from the oxide. The straight-line behavior in the low-frequency region is attributed to the ionic exchange in the electrolyte/material interphase.
The impedance features from
Figure 4a of the investigated electrodes have been simulated by the equivalent electrical circuit (EECs). The impedance characteristics of EECs are indicated by lines in the capacitance complex plane plots of
Figure 4a. Samples’ PEIS responses are described by different circuits and are shown in
Figure 4b–d. The framework for fitting was to quantify the EIS behavior through the simplest suitable EEC with the lowest possible fitting error. The EECs correspond to those applied in the literature [
27,
28] and are typical for porous structures. EECs are presented as resistors and constant phase elements (CPEs) through the branches. The Chi-squared calculated was below 0.004, while the relative error of parameter values did not exceed 34%. The equivalent electric circuits used to fit the data gave similar fitting quality for the electrode based on LMO and MnO
2.
The results obtained for La
2O
3, MnO
2, LMO, and LMO-Pt electrolytic material in N
2-saturated 0.10 M KOH from fitting results using EECs are given in
Table 2.
From the results shown in
Figure 4 and
Table 2, one can conclude that there are two time constants. CPE1 corresponds to the capacitive response of the outer surface of the electrode materials readily available to the electrolyte, and CPE2 corresponds to the capacitive response of the inner layer surfaces that are less accessible. The R1 resistance corresponds to the solution resistance. It can be seen that R3 is present in the EEC for MnO
2 and LMO. It appears that R3 describes the feature related to MnO
2 and its redox transitions to different oxidation states. This statement is supported by the value of R3 of LMO, in which the content of Mn is lower with respect to the pure MnO
2 layer. The R3 value of LMO is almost three times higher than the R3 value of MnO
2. If R2 is taken into consideration, the most inclined conclusion could be that it represents the redox resistance at the inner surfaces rather than the pore resistance, which should be comparatively negligible for Mn-containing samples (R2 is also higher for LMO in comparison to pure MnO
2). On the other hand, R2 could be dominated by the pore resistance in La
2O
3 due to less pronounced pseudocapacitive behavior since R3 was not required in the EEC. Consequently, R2 is an order of magnitude higher for La
2O
3 in comparison to MnO
2 and LMO. In the case of LMO-Pt, Pt causes these resistances to not affect the PEIS response.
The GEIS results for La
2O
3, MnO
2, LMO, and LMO-Pt in O
2-saturated 0.10 M KOH are shown in
Figure 5.
With an amplitude of 0.510 mA cm
−2, the GEIS measurements of La
2O
3 electrode material enter the limiting current region according to
Figure 2. Two overlapped semicircles can be seen for all the materials. Taking the semicircle diameter as a measure of the charge transfer resistance, i.e., ORR activity, the activity series in the tested region is as follows: MnO
2 > LMO-Pt > LMO > La
2O
3. The ratio of the two more or less overlapped semicircles for the materials differs too. For LMO-Pt and MnO
2, the diameter of the high-frequency semicircle is a bit larger than that of the low-frequency one. For LMO, the two diameters are of similar values, whereas the diameter of the low-frequency semicircle is larger for La
2O
3. The ORR is of mixed limiting character, being electrochemically- and diffusion-controlled. High-frequency response corresponds to charge transfer, and low-frequency corresponds to mass transfer limitations. The authors are fond of believing, according to findings from PEIS measurements, that the limiting current for the tested materials is the consequence of the chemical character of the materials rather than mass transfer. Despite pronounced activity at low overpotentials, La
2O
3 has much larger mass transport limitations (the diameter of the low-frequency semicircle is considerably larger) in comparison to other samples. Apparently, this reflects the lowest position of the limiting current plateau of La
2O
3 from
Figure 2. This most pronounced ORR limiting condition also affects the diameter of the high-frequency loop, which is not straightforwardly seen for Mn-containing samples. These features indicate that ORR at La
2O
3 is not subjected only to diffusion limitations, but to limitations related to some slow chemical reaction(s).
The fitting results gained by EEC (
Figure 5b) for ORR at La
2O
3, MnO
2, LMO, and LMO-Pt are given in
Table 3.
Owing to finite diffusion of O2, the finite length Warburg short circuit terminus was introduced (Ws). This element is also known as a Generalized Finite Warburg element (GFW), which is an extension of the Finite-Length Warburg (FLW). To allow its full finite diffusion assignment, the Ws-P parameter was fixed to 0.5. The FLW represents the solution of the one-dimensional diffusion equation of diffusion species, which is completely analogous to wave transmission in a finite-length RC transmission line. Ws-T parameter equals L2/D, where L is the effective diffusion thickness, and D is the effective diffusion coefficient. The Ws-R parameter defines a finite resistivity to diffusion. At very low frequencies, Z’ approaches Ws-R while Z” tends to zero.
The Warburg element is dominant in the low-frequency region. The larger the Ws-T parameter, the greater the diffusion thickness of specific species. R2 is charge transfer resistance, which shows the activity of the electrode materials towards ORR. From the results given in
Table 2, GEIS shows that at measured current density, MnO
2, LMO, and LMO-Pt have almost the same activity towards ORR, while the activity of La
2O
3 is almost twice as low. The higher the resistance, the lower the activity. On the other hand, the W1-T time constant is in direct correlation with the thickness of the double layer.
Table 2 shows that the effective diffusion thickness of La
2O
3 and MnO
2 is double the value of LMO and LMO-Pt. Comparing the results from
Table 2 and
Table 3, it can be seen that the presence of oxygen doubles the capacitance of the double layer. The values are even higher if the material is more active towards ORR. This is reasonable since adsorbed oxygen-containing species occupy some space in a double layer.
Figure 5 shows that galvanostatic EIS gives a more reliable and accurate response in comparison to potentiostatic conditions (data not shown; there were difficulties in recording low-frequency responses). However, GEIS requires a large amplitude of currents, which can provoke a considerable impact on the nonlinear response. In order to check the influence of higher harmonics on the apparently linear response of investigated electrode materials, GEIS was recorded for La
2O
3, MnO
2, LMO, and LMO-Pt in O
2-saturated 0.10 M KOH at OCP with a current amplitude of 0.510 mA cm
−2. This should induce the changes in potential output in a much wider range (and reasonably the widest) with respect to the GEIS measurements presented in
Figure 5, and consequently a nonlinear muti-harmonic response. Since EIS theory is based on the linear response of the first harmonic, its value should be negligibly affected by the intensity of the third harmonic [
29,
30]. The modulus of the third harmonic of ORR on the investigated electrode material is analyzed and presented in
Figure 6.
At higher frequencies (from 10 Hz to 100 kHz), the intensity of the third harmonic is weakly dependent on frequency, with LMO-Pt showing the most pronounced nonlinearity, followed by La
2O
3, MnO
2, and LMO. At lower frequencies, the nonlinearity contributes gradually more due to diffusion impact. The MnO
2 electrode material is of the lowest intensity of the third harmonic in the low-frequency domain, while the intensities of La
2O
3 and LMO increase to approach the intensity of LMO-Pt. Registered spectra of the third harmonic indicate that diffusion contributes much more to the nonlinearity than charge transfer does, whereas the increased number of components in the catalyst induces the nonlinearity of the charge transfer process. Combining the data for the activity from
Table 3 (R2, activity series, as follows: LMO ≈ MnO
2 > LMO-Pt >> La
2O
3) and those for CP1-T from
Table 2 as a measure of active surface area (LMO-Pt > La
2O
3 > MnO
2 > LMO), it can be concluded that
Figure 6 clearly shows that the most ORR active samples, LMO and MnO
2, are of the least expressed charge transfer nonlinearity. This also holds to a less pronounced extent for the region of diffusion control at low frequencies.
The influence of the third harmonic on the first is illustrated in
Figure 7 for the LMO-Pt sample, which showed the highest intensities of the third harmonic. The third harmonic starts increasingly to influence the output at the frequencies below 1 Hz and reaches almost 20% of the complete output intensity at the lowest applied frequency. This indicates that the output intensity should be corrected for the contribution of the third harmonic to gain reliable GEIS data related to linear response. The impedance complex plane plots with the contribution of the third harmonic to the output excluded are shown in
Figure 7. Loop-like dependences are obtained, which appear more or less elongated, depending on the sample, in the direction of increasing the imaginary impedance. This behavior could be associated with the inductive contribution, which is usually attributed to the adsorption phenomena in the ORR mechanism expressed in the low-frequency domain [
31].
Figure 7 shows that more pronounced elongations are found for La-containing samples. This introduces La
2O
3 as the catalyst component that affects mostly the adsorption steps in the ORR mechanism and consequently facilitates the spill-over issues.