Mn-Ni-Co-O Spinel Oxides towards Oxygen Reduction Reaction in Alkaline Medium: Mn_0.5Ni_0.5Co₂O₄/C Synergism and Cooperation

Abstract

Mn-doped spinel oxides Mn_xNi_1−xCo₂O₄ (x = 0, 0.3, 0.5, 0.7, and 1) were synthesized using the citric acid-assisted sol–gel method. The Mn_0.5Ni_0.5Co₂O₄ (x = 0.5) supported on carbon nanosheets, Mn_0.5Ni_0.5Co₂O₄/C, was also prepared using the same method employing NaCl and glucose as a template and carbon source, respectively, followed by pyrolysis under an inert atmosphere. The electrocatalytic oxygen reduction reaction (ORR) activity was performed in alkaline media. Cyclic voltammetry (CV) was used to investigate the oxygen reduction performance of Mn_xNi_1−xCo₂O₄ (x = 0, 0.3, 0.5, 0.7, and 1), and Mn_0.5Ni_0.5Co₂O₄ was found to be the best-performing electrocatalyst. Upon supporting the Mn_0.5Ni_0.5Co₂O₄ on a carbon sheet, the electrocatalytic activity was significantly enhanced owing to its large surface area and the improved charge transfer brought about by the carbon support. Rotating disk electrode studies show that the ORR electrocatalytic activity of Mn_0.5Ni_0.5Co₂O₄/C proceeds via a four-electron pathway. Mn_0.5Ni_0.5Co₂O₄/C was found to possess E_1/2(V) = 0.856, a current density of 5.54 mA cm⁻², and a current loss of approximately 0.11% after 405 voltammetric scan cycles. This study suggests that the interesting electrocatalytic performance of multimetallic transition metal oxides can be further enhanced by supporting them on conductive carbon materials, which improve charge transfer and provide a more active surface area.

1. Introduction

Fuel cells are promising and efficient energy conversion systems that utilize sustainable energy sources, such as hydrogen or biofuels. Due to the inherent efficiencies and use of sustainable fuels, fuel cells can help to reduce greenhouse gas emissions and mitigate future energy supply issues (such as blackouts or load-shedding events) [1,2]. Direct alcohol fuel cells (DAFCs) produce electricity through oxidation of alcohol fuel and reduction of oxygen gas at the anode and cathode electrodes, respectively [3,4,5]. Owing to their ability to utilize variable fuel sources, ease of integration, compatibility with different systems, and milder operating conditions, DAFCs have several advantages over their competitors [5]. However, the commercialization of DAFCs is hindered by low power density and high setup costs arising from costly platinum-based electrocatalysts. Hence, current research targets electrode material improvement, operation condition optimization, and increasing the power output.

Oxygen is the best cathode candidate for all the cathodic electron acceptors due to its high reduction potential and abundance [6,7]. The performance of fuel cell devices depends on the kinetics of the oxygen reduction reaction at the cathode. The kinetics of the ORR are generally sluggish [8,9,10,11]; therefore, a need exists to develop novel cathodic electrocatalytic materials to drive electrode reactions that regulate the power density of energy devices [10,12].

Non-precious metal-based electrocatalysts for ORR have been extensively studied. Among them, spinel metal oxides show excellent performance in an alkaline medium towards ORR. Spinel oxides are compounds that have the general structure AB₂O₄, with A and B having tetrahedral (T_d) and octahedral (O_h) occupancies, respectively [13,14,15,16]. Spinel oxides are possible substitutes for precious metal ORR electrocatalysts because they have variable valences, flexible structure, and remarkable redox stability. They are also abundant, cheap, and easy to prepare [17,18,19]. The literature has well documented an improvement in the electrochemical and physicochemical properties of transition metal oxides brought about by a stoichiometric combination of Ni and Co [17,20]. The electrocatalytic properties of spinel oxides are based on the spinel lattice’s electronic stability and geometric orientation. The cations within the spinel lattice can change valence states, leading to the formation of “electro-conductivity chains,” facilitating electron movement. Therefore, these electro-conductivity chains contribute to the electrocatalytic activity towards the ORR of spinel oxides [21].

In alkaline conditions, the cobaltite-based spinel oxides, MCo₂O₄ (where M = Mn, Ni, Cu, Fe), are very promising ORR electrocatalysts [22,23,24]. Merabet et al. [25] used the hydrothermal method to synthesize the crystalline MnCo₂O₄ spinel with a catalytic activity superior to Co₃O₄. Hu et al. [26] independently demonstrated that nanocrystalline Mn₃O₄@Co_xMn_3-xO₄ prepared by a rapid room temperature procedure possesses remarkable catalytic activity towards ORR.

Ternary spinel oxides, specifically the Mn–Ni–Co systems, possess good physicochemical characteristics, modified by intrinsic morphological nano-structural control. The properties of ternary systems are determined by the different synthetic methods that have been adopted, such as the hydrothermal method [13,27], wet methods [28], and combustion [29]. The availability of a range of methods gives room for tailored stoichiometric metallic compositions of the spinel lattice [29,30]. A ternary spinel system has a pronounced metallic synergy effect, which improves the spinel-based electrocatalysts’ catalytic and stability properties [13,31,32,33].

Of the facile synthetic methods available for preparing a ternary spinel, the citric acid-based sol–gel route has the advantage of stoichiometric control through metallic mixing at the atomic level. Moreover, if two metals have similar atomic radii, a ternary spinel cobaltite system of the type C_xA_1−xB₂O₄ is possible. For example, Dolla et al. [34] synthesized Mn_xZn_1−xCo₂O₄ by a facile co-precipitation method resulting in Mn-Zn-Co oxide microspheres. In the same vein, Nestola et al. [35] have prepared (A_xB_1−x)Fe₂O₄ using the same process. They studied the electrocatalytic activity of Mn_xCu_1−xCo₂O₄ in an alkaline medium, and it was concluded that copper substitution could enhance the ORR electrocatalytic properties [36].

Herein, we report an investigation of the effect of doping of Mn into the NiCo₂O₄ parent spinel towards the electrochemical properties for ORR. A series of Mn_xNi_1−xCo₂O₄ spinel oxide-based electrocatalysts were as synthesized using a modified citrate-assisted sol–gel method to improve ORR activity. The electrocatalytic activity of Mn_xNi_1−xCo₂O₄ oxide towards ORR was optimized, and the Mn_0.5Ni_0.5Co₂O₄ spinel oxide electrocatalyst proved to be the best ORR electrocatalyst after evaluation using cyclic voltammetry, chronoamperometry, and cyclic stability tests. The best-performing electrocatalyst, Mn_0.5Ni_0.5Co₂O₄, was then supported on a carbon sheet (Mn_0.5Ni_0.5Co₂O₄/C) to show the performance enhancement of carbon support. After this, physicochemical and electrochemical comparisons between Mn_xNi_1−xCo₂O₄ and Mn_0.5Ni_0.5Co₂O₄/C studies were made.

2. Results and Discussions

2.1. Spinel Synthesis Chemistry

A series of Mn_xNi_1−xCo₂O₄ spinel electrocatalysts were synthesized using the citric acid-assisted sol–gel method. As shown in Equation (1), acetates of Mn, Ni, and Co were mixed in water to form a metal acetate homogeneous complex.

$$\mathrm{p}\mathrm{M}+\mathrm{yH}+\mathrm{z}\left(\mathrm{cit}\right)\iff {\mathrm{M}}_{\mathrm{p}}{\mathrm{H}}_{\mathrm{q}}{\left(\mathrm{Cit}\right)}_{\mathrm{r}}$$

(1)

where ${\mathrm{M}}_{\mathrm{p}}{\mathrm{H}}_{\mathrm{y}}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}={\mathrm{xMn}}^{2+}/{\mathrm{HCit}}^{3-},\left(1-\mathrm{x}\right){\mathrm{Ni}}^{2+}/{\mathrm{HCit}}^{3-},{\mathrm{Co}}^{2+}/{\mathrm{HCit}}^{3-}$.

Citric acid was used as a chelating agent. After the homogeneous metal acetate solution was formed, ammonia was added to increase the pH to 7 (these metal ions reached complete complexation at this optimal pH) according to Equation (2):

$${\mathrm{M}}_{\mathrm{p}}{\mathrm{H}}_{\mathrm{y}}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}+{\mathrm{NH}}_3\leftrightarrow {\mathrm{NH}}_3{\mathrm{H}}_{\mathrm{y}}^{+}+{\mathrm{M}}_{\mathrm{p}}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}$$

(2)

where ${\mathrm{M}}_{\mathrm{p}}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}={\mathrm{xMn}}^{2+}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}+\left(1-\mathrm{x}\right){\mathrm{Ni}}^{2+}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}+{\mathrm{Co}}^{2+}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}$.

To the metal acetate solution, citric acid was added to form citrate anions coordinated to metallic cations. To ${\mathrm{M}}_{\mathrm{p}}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}$ ethylene glycol was added to assist with the gelation process. The reaction is illustrated in the following equation.

$${\mathrm{xMn}}^{2+}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}+\left(1-\mathrm{x}\right){\mathrm{Ni}}^{2+}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}+{\mathrm{Co}}^{2+}{\left(\mathrm{Cit}\right)}_{\mathrm{z}}\stackrel{\mathrm{a},\mathrm{b},\mathrm{c}}{\to }[{\mathrm{xMn}}^{2+}·\left(1-\mathrm{x}\right){\mathrm{Ni}}^{2+}·{\mathrm{Co}}^{2+}]\mathrm{gel}$$

(3)

a = ethylene glycol, b = 80–90 °C, c = 6 h

The resulting $[{\mathrm{xMn}}^{2+}·\left(1-\mathrm{x}\right){\mathrm{Ni}}^{2+}·{\mathrm{Co}}^{2+}]$-polymeric gel was dried overnight at 100 °C to evaporate the water formed during the citrate–ethylene glycol polymerization reaction. After that, the dried gel was calcined at 450 °C, and the metal oxide spinel system was developed. During calcination, carbon dioxide gas was produced, thus leading to a material with increased porosity. The following equation shows this two-stage process:

$$({\mathrm{xMn}}^{2+}·\left(1-\mathrm{x}\right){\mathrm{Ni}}^{2+}·{\mathrm{Co}}^{2+}) \mathrm{gel}\stackrel{\mathrm{a},\mathrm{b}}{\to }{\mathrm{Mn}}_{\mathrm{x}}{\mathrm{Ni}}_{1-\mathrm{x}}\mathrm{Co}\stackrel{\mathrm{c},\mathrm{d},\mathrm{e}}{\to }\begin{array}{c}{\mathrm{Mn}}_{\mathrm{x}}{\mathrm{Ni}}_{1-\mathrm{x}}{\mathrm{Co}}_2{\mathrm{O}}_4\\ \mathrm{Black} \mathrm{powder}\end{array}$$

(4)

a = dry overnight, b = 100 °C, c = 450 °C, d = 5 h, e = calcination

As shown in reaction Scheme S2, Mn_0.5Ni_0.5Co₂O₄/C was synthesized by mixing stoichiometric amounts of metallic precursors to form $[0.5{\mathrm{Mn}}^{2+}·\left(0.5\right){\mathrm{Ni}}^{2+}·{\mathrm{Co}}^{2+}]-{\mathrm{C}}_6{\mathrm{H}}_{12}{\mathrm{O}}_6-\mathrm{cubic} \mathrm{NaCl}$. Sodium chloride was utilized as a template because it is less soluble than metal precursors and glucose. $[0.5{\mathrm{Mn}}^{2+}·\left(0.5\right){\mathrm{Ni}}^{2+}·{\mathrm{Co}}^{2+}]-{\mathrm{C}}_6{\mathrm{H}}_{12}{\mathrm{O}}_6-\mathrm{cubic} \mathrm{NaCl}$ was oven treated to form $[0.5{\mathrm{Mn}}^{2+}·\left(1-\mathrm{x}\right){\mathrm{Ni}}^{2+}·{\mathrm{Co}}^{2+}]/\mathrm{C}-\mathrm{coated} \mathrm{cubic} \mathrm{NaCl}$. The resultant composite was then pyrolyzed at 750 °C in the presence of PVP, and the release of CO₂ led to the formation of a material with a porous structure (see SEM in the SI section). The $[0.5{\mathrm{Mn}}^{2+}·\left(0.5\right){\mathrm{Ni}}^{2+}·{\mathrm{Co}}^{2+}]/\mathrm{C}-\mathrm{coated} \mathrm{cubic} \mathrm{NaCl}$ was then calcined in air at 450 °C to form porous crystalline Mn_0.5Ni_0.5Co₂O₄/C−coater cublc NaCl. The NaCl template was washed off with H₂O/CH₃CH₂OH (0.5:1, v/v).

2.2. Physicochemical Characterization

The phase purity and crystal structures of the Mn_xNi_1−xCo₂O₄ were investigated using powder X-Ray Diffraction (XRD) and are shown in Figure 1a. The XRD pattern confirmed that all the spinel oxides exhibited a cubic spinel structure (Fd$\bar{3}$m) and no impurity was observed. The diffraction peaks are observed at 2θ values of 18.88°, 31.18°, 36.7°, 44.64°, 55.67°, 59.05°, and 64.89°, which were, respectively, indexed to the (111), (220), (311), (400), (422), (511), and (440) crystal planes of a cubic NiCo₂O₄ spinel (JCPDS No. 73-1702) [37,38]. All the diffraction patterns of the Mn-doped samples were found to be similar to the NiCo₂O₄, thus providing evidence for the formation of a homogeneous solid solution and the adoption of the spinel cubic structure. The average crystal sizes were calculated using the Scherrer equation on peak widths at 2θ = ~36° (331) and are presented in

Table 1. The structural and electrochemical characteristics of the synthesized electrocatalysts.

Material	Physicochemical Characteristics					Electrochemical Activity
	Crystal Size (nm)	Lattice Parameter (Å)	BET Surface Area (m2g−1)	Pore Volume (cm3g−1)	Average Pore Size (nm)	Current Density (mAcm−2)	* Current Retention (%)
x = 0	19.5	8.1084	63.19	0.1263	7.6	−3.80	56.46
x = 0.3	14.6	8.1357	55.17	0.1028	7.0	−3.49	41.30
x = 0.5	15.1	8.1480	83.42	0.2819	13.3	−3.93	64.98
x = 0.7	9.7	8.1402	44.49	0.2288	19.5	−2.51	34.91
x = 1	15.5	8.1281	36.63	0.2315	19.7	−1.88	54.55
Pt/C	-	-	-	0.2668	4.85	−4.42	-
Mn0.5Ni0.5Co2O4/C	-	-	209.52	0.1263	7.6	−5.54	86.20

$* \mathrm{Retained} \mathrm{current} (\%)=\left[1 - \frac{\mathrm{Initial} \mathrm{current} \left(\mu \mathrm{A}\right)-\mathrm{Final} \mathrm{current} \left(\mu \mathrm{A}\right)}{\mathrm{Initial} \mathrm{current} \left(\mu \mathrm{A}\right)}\right] \text{×} 100$%.

. The lattice parameters were calculated and observed to be consistent with the previous report [13]. The lattice parameters increase with manganese concentrations from x = 0 to 0.5 and decrease onwards to x = 1. Based on the inverse spinel-type NiCo₂O₄ crystal structure, ${\mathrm{Co}}_{\mathrm{x}}^{2+}{\mathrm{Co}}_{1-\mathrm{x}}^{3+}\left[{\mathrm{Co}}^{3+}{\mathrm{Ni}}_{1-\mathrm{x}}^{2+}{\mathrm{Ni}}_{\mathrm{x}}^{3+}\right]{\mathrm{O}}_4$ [39], it can be explained that manganese substitution for nickel takes place in the octahedral position up to x = 0.5, thus increasing the lattice parameter [40]. Beyond x = 0.5, either the manganese or nickel or both cations can move from O_h (A) to T_d (B) sites, leading to the Mn²⁺/Mn³⁺ or Ni²⁺/Ni³⁺ O_h (A) and T_d (B) partial residence. In Figure 1b, the peak at 2θ = 26.1° (marked with * asterisks) is attributed to carbon coexisting in the obtained product showing the carbonization process’s effectiveness. Thus, Mn_0.5Ni_0.5Co₂O₄/C has a graphitic carbon structure and is well developed inside the composite.

To further understand the structures of Mn_xNi_1−xCo_2−xO₄ and the carbon support formed on Mn_0.5Ni_0.5Co₂O₄/C), Raman spectroscopy analysis was carried out, and the results are shown in Figure 2a,b, respectively. The characteristic peaks for the M–O bond stretching vibration were observed at 435 cm⁻¹ to 735 cm⁻¹ for the Mn_xNi_1−xCo_2−xO₄ series corresponding to the Mn–O, Ni–O, and Co–O bonds [13,41,42]. A slight shift of peaks was observed across the synthesized spinel oxide series, and these were attributed to the defects and strains arising from the doping of Mn in NiCo₂O₄. In the Raman spectrum of Mn_0.5Ni_0.5Co₂O₄/C, two characteristic peaks appearing at 1333 cm⁻¹ and 1564 cm⁻¹ were observed, and these were indexed to the D and G bands of the graphitic carbon, respectively. The integrated peak intensity ratio of the D and G bands (I_D/I_G) provides a base assessment for the degree of graphitization of carbon material. The I_D/I_G ratio can be used to quantify the defects within carbonaceous materials. The D band is due to the disorderliness or the defects within the graphitized lattice plane, whereas the G band is attributed to the extent of the sp²-hybridization of the graphitized structure [43]. For the Mn_0.5Ni_0.5Co₂O₄/C hybrid, the calculated I_D/I_G is 1.17. A large value of I_D/I_G shows a high degree of graphitic defects, which is advantageous in electrocatalysis because the defects contribute towards the anchoring of Mn_0.5Ni_0.5Co₂O₄ in the carbon matrix, thus allowing easy electrolyte movement and exposure of active catalytic sites that facilitate electrocatalytic activity.

The FTIR of the series of Mn_xNi_1−xCo₂O₄ spinel oxides (Figure 3) was used to identify the bending and stretching vibrations and give more insight into the spinel formation. Two intense peaks were observed below 1000 cm^–1. Specifically, the peaks appearing in the range 500–688 cm^–1 (oval dotted line in Figure 3) were ascribed to the stretching vibrations of tetrahedral (T_d) and octahedral (O_h) occupancy of M–O (M = Mn, Ni or Co). M–O intensities were much pronounced for Mn_0.3Ni_0.7Co₂O₄, Mn_0.5Ni_0.5Co₂O₄, Mn_0.7Ni_0.3Co₂O_4, and Mn_0.5Ni_0.5Co₂O₄/C as compared to NiCo₂O₄ and MnCo₂O₄. The high intensities are probably due to the three compounded M–O moieties, namely Mn–O, Ni–O, and Co–O. A shift in the peak position of the doped samples (Mn_0.3Ni_0.7Co₂O₄, Mn_0.5Ni_0.5Co₂O₄, Mn_0.7Ni_0.3Co₂O₄, and Mn_0.5Ni_0.5Co₂O₄/C) relative to the undoped samples (NiCo₂O₄ and MnCo₂O₄) was observed. This shift can be attributed to the T_d and O_h occupancy stretches arising from the Mn doping of the spinel matrix. We believe that the presence of a split OH peak appearing between 3160 and 3421 cm^–1 might be due to the different OH adsorbed on spinel and carbon sheets. This, in turn, confirms the presence of a carbon sheet in the synthesized sample due to OH peak overlap. The FTIR also showed absorption peaks between 2297 and 2441 cm^–1, due to the CO₂ sorption molecule at the surface. The stretching vibrations of the CH₂ group (shown by a big purple arrow) occurred between 2822 and 2972 cm^–1 [14]. Carboxyl (–C=O) symmetric and asymmetric stretching bands were observed at 1547–1711 cm^–1. These bands emanate from the trace organic precursors of the support material [44], which were also confirmed by the presence of C 1s core level of the XPS spectrum shown in Figure 4i.

X-ray photoelectron spectroscopy (XPS) of the Mn_xNi_1−xCo₂O₄ series and Mn_0.5Ni_0.5Co₂O₄/C was used to investigate the oxidation state and to understand the electronic structure and composition of the elements on the surface. Survey scans for the Mn_xNi_1−xCo₂O₄ and Mn_0.5Ni_0.5Co₂O₄/C samples (Figure 4a) show the presence of Mn, Ni, and Co (the LMM Auger lines of Mn, Ni, and Co), O (the O KVV Auger lines), and C (C 1s) at 285 eV binding energy (BE), which is due to the presence of absorbates on the surface of the Mn_xNi_1−xCo₂O₄ (ex situ samples).

The Co 2p spectra for Mn_xNi_1−xCo₂O₄ and Mn_0.5Ni_0.5Co₂O₄/C are displayed in Figure 4b,c. This BE region is divided into two main regions, that is, ~777–792 eV for Co 2p_3/2 and ~792–807 eV for Co 2p_1/2 [45,46], and this is consistent with the line shape of the mixed-valent Co₃O₄ where Co ions have both Co³⁺ and Co²⁺ oxidation states [47,48]. For example, the fit to the x = 1 sample is shown at the bottom of Figure 4b for the entire series. The spectrum for Mn_0.5Ni_0.5Co₂O₄/C is fitted separately (Figure 4c). The best fit to the data for the x = 1 sample in Figure 4b shows two spin–orbit doublets (separated by a spin–orbit energy splitting of 15.2 eV) for the main peaks and one doublet with a broad line shape for the satellite. These two components can be ascribed to the Co³⁺ (BE 2p_3/2 = 780.35 eV), which accounts for roughly two-thirds of the spectral weight (taking into account the potential error associated with the choice of background for the fit) and Co²⁺ oxidation states (BE 2p_3/2 = 782.27 eV), which accounts for the other one-third of the spectral weight. Similarly, for Mn_0.5Ni_0.5Co₂O₄/C, the BE for Co³⁺ 2p_3/2 (66% of the spectral weight) is 780.02 eV, while that of Co²⁺ 2p_3/2 (34%) is 781.94 eV, which is consistent with the relevant literature on Co₃O₄-type compounds [47,48]. The Co 2p core-level line shape shows that Co is very robust towards the Mn-to-Ni substitution.

Figure 4d shows the Mn 2p BE region for the x = 0.3, 0.5, 0.7, 1 series, and Mn_0.5Ni_0.5Co₂O₄/C. This energy region comprises two main BE regions, namely ~638 eV to ~649 eV (2p_3/2) and ~649 eV to ~660 eV (2p_1/2). The Mn 2p BE region overlaps with the Ni LMM Auger emission, whose line shape was measured for the Mn-free sample of the series and was appended as a solid black line underneath the Mn 2p spectra plotted in Figure 4d. Before fitting the data, the line shape of the Ni Auger peak (normalized to the stoichiometric amount of Ni in each sample) was subtracted from the row data in order to obtain the Mn 2p contribution to the XPS signal in this BE region. Mn is present mainly in the Mn²⁺ oxidation state [13,34]. The BEs of the 2p_3/2 peaks of the fitted components are 641.95 eV (Mn²⁺—80% of the spectral weight) and 644.39 eV (Mn³⁺). The line shape of Mn 2p for Mn_0.5Ni_0.5Co₂O₄/C overlaps very well with that of the non-supported equivalent sample x = 0.5 of the spinel oxide series, indicating that the presence of carbon support does not affect the line shape of this core level.

Figure 4e,f displays the Ni 2p BE region for samples (d) x = 0, 0.3, 0.5, 0.7 and Mn_0.5Ni_0.5Co₂O₄/C. This energy region comprises two main BE regions, namely from ~850 eV to ~870 eV (Ni 2p_3/2) and ~870 eV to ~890 eV (2p_1/2) (Table S1). A spin–orbit energy splitting of 17.37 eV was used. The component at lower BE was assigned to the Ni²⁺ oxidation state, and the one at higher BE to the 3+ oxidation state [13,49,50]. Interestingly, there is a change in the Ni²⁺/Ni³⁺ ratio across the series. Samples x = 0 and x = 0.3 have the Ni³⁺ oxidation state in the majority. With regard to the Ni 2p spectrum for the x = 0.7 sample, the vertical arrow in the figure indicates the Co Auger broad peak, which overlaps the Ni 2p core level in this BE region; the Co Auger line becomes more relevant in this sample as it is the one with the least amount of Ni in the series. This Auger line’s presence severely affects the goodness of fit for this sample, as the fit parameters were kept consistent with those for the other samples. However, the oxidation states are inferred from the relative intensity of the main peaks, so the inconsistency of the fitting for the x = 0.7 sample in the satellite region should not affect the final result. By contrast, the other two samples have an almost even distribution of the Ni²⁺ and Ni³⁺ oxidation states. This is due to the electrocatalytic activity of ratio x = 0.5 and the disappearance of the pseudo-bifunctionality in x = 0.7 (Figure S6) and x = 0 (Figure S7). Furthermore, Mn_0.5Ni_0.5Co₂O₄/C shows the least amount of Ni³⁺ compared to the Mn_xNi_1−xCo₂O₄ series. This means that, during the electrocatalytic process, the redox couple Ni²⁺/Ni³⁺ can provide electrons easily for the electro-reduction process.

Figure 4g,h shows the O 1s spectra for the Mn_xNi_1−xCo₂O₄ series and Mn_0.5Ni_0.5Co₂O₄/C, respectively. The three components are labeled O1, O2, and O3. O1 is ascribed to the stoichiometric oxygen in the main matrix of the perovskites [13,51] and O2 corresponds to oxygen vacancies or defects [13]. As shown in

Table 2. BEs (in eV), FWHM (in eV), and relative percentage areas (in %) of the three components of the O 1s core level for the MnNiCoO series and Mn_0.5Ni_0.5Co₂O₄/C.

Sample	O1 (eV)	O2 (eV)	O3 (eV)	FWHM (O1, eV)	FWHM (O2, eV)	FWHM (O3, eV)	O1 (%)	O2 (%)	O3 (%)
x = 0	529.49	531.14	533.24	1.23	2	2	39.1	51	9.9
x = 0.3	529.62	531.27	533.27	1.29	2	2	40.4	42.8	16.8
x = 0.5	529.8	531.36	533.42	1.29	2.1	2.1	44.3	41.6	14.1
x = 0.7	530.06	531.56	533.59	1.41	1.8	1.8	70.2	25.6	4.2
x = 1	530.2	531.8	533.4	1.35	1.57	1.9	71.4	17.6	11
Mn0.5Ni0.5Co2O4/C	529.79	531.23	533.36	1.25	2.2	2.2	41	45.6	13.4

, Mn_0.5Ni_0.5Co₂O₄/C has the highest percentage of component O2 (high oxygen vacancy/defects corresponding to high catalytic activity). Lastly, O3 is attributed to surface contaminants, that is, chemisorbed oxygen [41]. Interestingly, while the energy difference between each component remains consistent from sample to sample, the overall BE shifts to higher values from x = 0 to x = 1, suggesting that the addition of Mn makes the system slightly more insulating. This highlights the need for supporting Mn_0.5Ni_0.5Co₂O₄ on a carbon sheet to enhance conductivity.

Lastly, Figure 4i shows the fitted C 1s spectrum for Mn_0.5Ni_0.5Co₂O₄/C and relative percentages in Table S2. While C1 was attributed to C–C/C=C chemical bonds, C2 was attributed to C–OH bonds; by the same token, C3 to C=O and C4 to O–C=O are various oxygen moieties [52]. These correspond with the results of the FTIR spectra shown in Figure 3. C1 has the highest percentage of C–C/C=C chemical bonds, thus confirming the graphitic and carbonaceous nature of the support material and the carbon source’s graphitization (glucose). This, therefore, confirms the presence of carbon sheets.

The morphology of the prepared Mn_xNi_1−xCo₂O₄ and the Mn_0.5Ni_0.5Co₂O₄ supported on carbon was investigated using transmission electron microscopy (TEM). Figure 5a shows TEM images of the Mn_0.5Ni_0.5Co₂O₄, which are agglomerated. Figure 5b,c shows the TEM images of Mn_0.5Ni_0.5Co₂O₄/C at different magnifications. Figure 5b shows the overview of Mn_0.5Ni_0.5Co₂O₄ (black dots) on the carbon sheet. Figure 5c shows the magnified image of Mn_0.5Ni_0.5Co₂O₄/C, showing the dispersion of the Mn_0.5Ni_0.5Co₂O₄ nanoparticles in the carbon matrix. Thus, Figure 5b,c shows that employing carbon support helps in reducing the agglomeration, concluding that Mn_0.5Ni_0.5Co₂O₄ nanoparticles are spread well over the carbon nanosheet. The SAED (selected area electron diffraction) pattern in Figure 5d shows that Mn_0.5Ni_0.5Co₂O₄/C is polycrystalline. Figure 5e shows the EDS spectrum of the Mn_0.5Ni_0.5Co₂O₄ sample, confirming the presence of Mn, Ni, Co, and O in the sample. In Figure S3, the compositions with x = 0.3, 0.5, and 0.7 show cubic type morphology (red squares), while the shapes of x = 0 and 1 are not clearly defined due to agglomeration. The doping of Mn had affected the “shape” morphology as shown when x = 0 to 1 (Figure S3). On the one hand, Figure S4a,b shows the respective SEM and TEM images of Mn_0.5Ni_0.5Co₂O₄. The TEM images show aggregates of small nanoparticles (as dark spots), which could be due to the interaction between the nanoparticles resulting from the high surface energy of the nanoparticles.

Table 1 lists the textual properties of Mn_xNi_1−xCo₂O_4, and Mn_0.5Ni_0.5Co₂O₄/C. Figure 6 shows the nitrogen adsorption–desorption isotherms and pore size distribution of a series of Mn_xNi_1−xCo₂O₄. As per IUPAC guidelines, the nitrogen adsorption–desorption isotherms for the Mn_xNi_1−xCo₂O₄ series are type IV. The exception is observed with the Mn_0.5Ni_0.5Co₂O₄ sample, classified as a type V isotherm (elongated S-type). The change in the sizes of the nanoparticles probably accounts for changes in the agglomeration and hence changes in the voids between the nanoparticles, which could account for the change in the isotherm shapes. The elongated S-type means that as more O₂ molecules move towards the Mn_0.5Ni_0.5Co₂O₄-modified GCE during hydrodynamic analysis, it becomes easier for more O₂ to be fixed. This phenomenon occurs in a cooperative adsorption manner via side-by-side adsorption association between O₂-adsorbed molecules, proving the cooperative nature of the electrocatalysts [53]. This elongated S-type shape supports the activity of Mn_0.5Ni_0.5Co₂O₄ since the O₂ molecule is (a) monofunctional, (b) exhibits moderate intermolecular attraction, and (c) meets strong competition due to the hydrodynamic ORR process. All the requirements listed in a, b, and c should be met for the elongated S-shape to be realized. At P/Po = 0.7–0.8, the hysteresis loops show distinctive H3 hysteresis loops for the Mn_xNi_1−xCo₂O₄ series, indicating that the agglomerated nanoparticles’ mesoporous nature makes up the various spinel samples. However, the Mn_0.5Ni_0.5Co₂O₄/C sample exhibits a H1 hysteresis loop, meaning that Mn_0.5Ni_0.5Co₂O₄/C may have a more uniform porous structure. The Mn_0.5Ni_0.5Co₂O₄/C showed a different hysteresis loop relative to Mn_0.5Ni_0.5Co₂O₄ because of the pore overlap between the Mn_0.5Ni_0.5Co₂O₄ and the carbon sheet. This overlap might have resulted in a narrow range of uniform mesopores [54,55,56].

2.3. Electrochemical Evaluation of ORR Activity

The electrocatalytic activities of Mn_xNi_1−xCo₂O₄, 0 ≤ x ≤ 1 nanocomposites were elucidated using cyclic voltammetry within the 0.55 to 1.35 (V vs. RHE) range (Figures S3–S8). This necessitated checking the electrocatalytic behavior of the synthesized electrocatalysts in an alkaline medium and selecting the best electrocatalyst within the Mn_xNi_1−xCo₂O₄, 0 ≤ x ≤ 1 series. The best-performing electrocatalyst was then supported on a carbon sheet to improve the electrocatalytic activity, durability, and stability. The stability and durability were evaluated using chronoamperometry and cyclic voltammetry studies.

The dependence of the electrochemical behavior of Mn_xNi_1−xCo₂O₄ on the scan rate (10–100 mV/s) was studied (Figures S3–S8). It was possible to conduct the scan rate studies in the whole range because the faradaic current was more dominant than the capacitive current. A linear correlation of scan rate to cathodic peak current was observed (Figures S3–S8b), indicating that the electroactive O₂ is firmly entrapped on the modified GCE cathode surface. This means that the species are confined in the bulk of the electrode surface, except for Mn_0.7Ni_0.3Co₂O₄, which does not have a well-defined pattern. This anomaly can be explained using the XPS result (Table S1), which shows that Mn_0.7Ni_0.3Co₂O₄ has approximately the same Ni²⁺/Ni³⁺, leading to a reduction in the redox couple behavior, as evidenced in Figure S6. A negative shift in potential was observed when the scan rate was increased, indicating greater propensity for oxygen reduction. Figures S3–S8b show a plot of E_pc vs. log v, which gave linear regressions with R² greater than 0.9600. This indicates a strong correlation between the cathodic peak potential and the logarithmic value of potential. An anomaly was observed for Mn_0.7Ni_0.3Co₂O₄, which has an R² of 0.7416 (Figure S6b). From the plots of E_pc vs. log v, a Tafel slope (b) was calculated using Equation (5) [57]:

$$E_{pc}=\frac{b}{2}log v+constant$$

(5)

where $\frac{b}{2}$ is the plot slope and b is the Tafel slope.

Evaluating $\frac{\partial Epc}{\partial \left(log v\right)}$ gives V dec⁻¹ and no distinct change in the slope of the plots was observed, indicating the exact mechanism in the O₂ electro-reduction. The $\frac{\partial Epc}{\partial \left(log v\right)}$ for Mn_0.5Ni_0.5Co₂O₄ was found to be 0.400 V dec⁻¹ (the highest Tafel slope of all ratios) and $\frac{b}{2}=0.2002$, thus giving $b=2\times 0.2002=0.400 \mathrm{V} \mathrm{dec}$⁻¹. The other Tafel values are shown in Table S3. The higher Tafel value obtained for Mn_0.5Ni_0.5Co₂O₄ indicates the porous nature of the synthesized materials. The small Tafel (0.0623 V dec⁻¹) for Mn_0.3Ni_0.7Co₂O₄ means a probable 2-electron oxygen electro-reduction pathway, confirmed by the Koutecky–Levich plots. The Tafel value relates to a fast over-potential increase with the current intensity, which indicates the good catalytic activity of the electrocatalysts and is therefore commendable for electrocatalytic reduction of O₂.

In Figures S3–S8c, the plot of I_cp vs. $\sqrt{scan rate }$ emanating from the voltammetry studies gave a linear correlation between the two, indicating a diffusion-controlled process. Thus, no external force is required to push the reacting species to the surface of the electrodes. The surfaces and intrinsic properties of the electrocatalysts are essential in the electrochemical performance of the Mn_xNi_1−xCo₂O₄ spinel oxide series. To evaluate these characteristics, cyclic voltammetry within the 0.55 to 1.35 (V vs. RHE) range at 50 mV s⁻¹ was used, and the resulting voltammograms are shown in Figure 7. As shown in Figure 7, the electrocatalysts have well-established peaks within the potential window, relative to Figure S9 where the cv cycles show no reduction peaks in N₂-saturated 0.1 M KOH. The cyclic voltammograms for NiCo₂O₄, Mn_0.3Ni_0.7Co₂O₄, and Mn_0.5Ni_0.5Co₂O₄ (Figure 7a) show that their reduction curves have shifted to the more positive values, relative to Mn_0.7Ni_0.3Co₂O₄, MnCo₂O₄, and Mn_0.5Ni_0.5Co₂O₄/C), which have reduction curves shifted to the negative values. Mn_0.5Ni_0.5Co₂O₄ of Mn_xNi_1−xCo₂O₄ has a more positive onset potential than all ratios (1.10 (V vs. RHE)), indicating its low over-potential ORR. Reaction kinetics of Mn_0.5Ni_0.5Co₂O₄ occurred at 0.78 (V vs. RHE) half-wave potential, with a 1.10 (V vs. RHE) onset potential Tafel of 0.400 V dec⁻¹. This qualifies Mn_0.5Ni_0.5Co₂O₄ as the most electroactive electrocatalyst of the series for the reduction of oxygen because it follows Equations S(2)–S(18).

Following the onset potential and current density, the order of ORR activities of the electrocatalysts is as follows; Mn_0.5Ni_0.5Co₂O₄/C > Mn_0.5Ni_0.5Co₂O₄ > NiCo₂O₄ > Mn_0.3Ni_0.7Co₂O₄ > Mn_0.7Ni_0.3Co₂O₄ > MnCo₂O₄. From these ORR activities, it can be seen that Mn_0.5Ni_0.5Co₂O₄/C (Figure 7b) has the highest current density due to the increased electroactive surface area (Table 1) resulting from high oxygen vacancy sites revealed by XPS [13], low charge transfer resistance (

Table 3. Resistance values for the cathodes calculated from the fitting results of Nyquist plots.

Sample Name	Solution Resistance (RC) (Ω)	Charge Transfer Resistance (RCT) (Ω)
NiCo2O4	6.621	230.3
Mn0.3Ni0.7Co2O4	25.08	519.8
Mn0.5Ni0.5Co2O4	2.776	272.43
Mn0.7Ni0.3Co2O4	12.09	304.00
MnCo2O4	105.2	799.9
Mn0.5Ni0.5Co2O4/C	1.076	92.31

), and low agglomeration (Figure 5a). This highly pronounced ORR activity can be accounted for by the synergy of the carbon support and existence of multimetallic centers metallic within the spinel lattice. The carbon sheet as a support material for Mn_0.5Ni_0.5Co₂O₄ enhanced the electron movement during the electro-reduction process. For the reduction peaks shown by the Mn_xNi_1−xCo₂O₄ spinel series at $E_{\frac{1}{2}}$, the oxygen reduction pathway and mechanisms vary with catalytic composition. Thus, O₂ adsorption is relative to catalytic surface geometry, the binding energy, and the available active sites of the electrocatalysts. For the creation of hydroxyl species on the metallic oxide surface, the surface of the oxygen ligand must be protonated, which is counteracted by charge-compensation via metal cation (M) reduction (Mn³⁺, Mn⁴⁺, Ni³⁺, and so on) [13,58]. The formed metal hydroxide species (M–OH^—) will interact with O₂ end-on or side-on arrangements [59]. Thus, the ORR mechanism proposed paths on the metal oxide surfaces would follow Equations S(2)–S(18) (Supplementary Materials) [60].

From the proposed mechanism, the Mn_xNi_1−xCo₂O₄ spinel series utilizes electron hopping within the spinel lattice. There are also well-defined anodic peaks shown by NiCo₂O₄, Mn_0.3Ni_0.7Co₂O₄, and Mn_0.5Ni_0.5Co₂O₄, which can be ascribed to the oxidation of M²⁺ to M³⁺/M⁴⁺ (M = Co). Thus, in the alkaline medium, the mechanism is as follows [61]:

Step 1: OH^— adsorption

$$A^{*}+O{H}^{-}\to e^{-}+A^{*}OH$$

(6)

Step 2: O–H bond breaking

$$A^{*}OH+O{H}^{-}\to e^{-}+H_{2}O+A^{*}O$$

(7)

Step 3: formation of oxylato-oxygen/hydroperoxyl

$$A^{*}O+O{H}^{-}\to e^{-}+A^{*}OOH$$

(8)

Step 4: H₂O and adsorbed O₂ formation

$$A^{*}OOH+O{H}^{-}\to H_{2}O+e^{-}+A^{*}OO$$

(9)

Step 5: O₂ desorption

$$A^{*}OO\to O_2+A^{*}$$

(10)

where A* is the active site.

For Mn_0.5Ni_0.5Co₂O₄/C, the redox peaks at 1.06 and 1.08 (V vs. RHE) are ascribed to reversible oxidation CoOOH + OH^— $\rightleftharpoons$ CoO₂ + H₂O + e⁻, that is the redox Co²⁺/Co³⁺ peak depicted in Equation S(1), which follows from Equation S(10) under Step 4 (Figure S8).

The ORR is a multi-electron process, depending on the catalyst used in the electro-reduction process. In alkaline electrolytes, ORR follows the direct four-electron pathway that involves direct oxygen reduction to hydroxyl ions (Equation (11); most desired pathway) and the indirect two-electron pathway, which produces $H{O}_2{}^{-}$ intermediate (Equations (12)–(14); less efficient pathway). The $H{O}_2{}^{-}$ is undesirable because it causes catalyst poisoning and corrosion of the fuel cell components, reducing electricity production. Thus, electrocatalysts must be engineered to facilitate the direct four-electron pathway, enhancing maximum electricity generation in fuel cells.

$$O_2+2H_{2}O+4e^{-}\to 4O{H}^{-}$$

(11)

$$O_2+2H_{2}O+2e^{-}\to H{O}_2{}^{-}+O{H}^{-}$$

(12)

$$H{O}_2{}^{-}+H_{2}O+2e^{-}\to 3O{H}^{-}$$

(13)

$$2H{O}_2{}^{-}\to 4O{H}^{-}+O_2$$

(14)

To better understand the ORR performance and mechanism of the electrocatalysts, we performed linear sweep voltammetry (LSV) measurements on the RDE in 0.1 M KOH saturated with O₂. The RDE we used was modified with NiCo₂O₄, Mn_0.3Ni_0.7Co₂O₄, Mn_0.5Ni_0.5Co₂O₄, Mn_0.7Ni_0.3Co₂O₄, MnCo₂O₄, and the best-performing carbon sheet supported Mn_0.5Ni_0.5Co₂O₄/C. When potential is 0.6 (V vs. RHE), the resultant current density absolute values were in the sequence of Mn_0.5Ni_0.5Co₂O₄/C > Pt/C > Mn_0.5Ni_0.5Co₂O₄ > NiCo₂O₄ > Mn_0.3Ni_0.7Co₂O₄ > Mn_0.7Ni_0.3Co₂O₄ > MnCo₂O₄. It should be noted that the amplification of electrocatalytic activity as instigated by carbon support follows the same order of CV performance. Based on the effect of carbon support on electrocatalytic activity, the best-performing Mn_0.5Ni_0.5Co₂O₄ was supported on a carbon sheet to enhance its performance further.

Figure 8a shows LSVs at different rotation speeds 400–2000 rpm of the electrocatalyst used to investigate the electron transfer pathway for ORR. The LSV curves by RDE, Mn_0.3Ni_0.7Co₂O₄, MnCo₂O_4, and Mn_0.5Ni_0.5Co₂O₄ showed a characteristic two plateaued LSV curve signifying a peroxide pathway (Figure 8b). This provides evidence for an oxygen reduction mechanism via two successive two-electron processes (Equations (12)–(14)) [62]. From the LSV results, the Koutecky–Levich (K–L) plots (Figure 8c) were drawn following the K–L shown in Equation (15). The linear K–L line fitting plots suggest first-order reaction kinetics of O₂ reduction. The average number of transferred electrons was calculated at different potentials (Figure 8d). The Mn_0.5Ni_0.5Co₂O₄/C, NiCo₂O_4, and Mn_0.7Ni_0.3Co₂O₄ spinel oxides had an average electron transfer exceeding 3.5. The most active electrocatalyst, Mn_0.5Ni_0.5Co₂O₄/C, had an average electron transfer of 3.7, comparable to the n value of Pt/C commercial catalyst n value. This pronounced electron transfer might be due to the synergistic interaction between the Mn_0.5Ni_0.5Co₂O₄ and the carbon sheet forming Mn_0.5Ni_0.5Co₂O₄/C that possesses the enhanced electron transfer facilitated by metallic cooperation within the spinel lattice. The hybridization of the carbon orbitals (of the support material) and Mn_0.5Ni_0.5Co₂O₄ occurs between split 3d orbitals of Mn, Ni, and Co and the 2p carbon orbitals. This interaction will cause electron density to be filled in the Fermi level, hence the exhibited electro-reduction catalytic activity of Mn_0.5Ni_0.5Co₂O₄/C. Thus, the electrons will flow from the carbon sheet to Mn_0.5Ni_0.5Co₂O₄, which causes the supported Mn_0.5Ni_0.5Co₂O₄ to be negatively charged, leaving the carbon sheet positively charged, hence the electrostatic inductive effect. The electrostatic inductive effect enhances electrocatalytic properties, for example, the electron conductivity resulting from the interaction between the carbon sheet pi-orbitals and Co split 3d orbitals, which comes in as a source of electrons for redox couples during oxygen electro-reduction.

Li et al. [63] have also reported the same interaction in spinel MnCo₂O₄/nanocarbon hybrids. The carbon introduction amplified the electronic state of the M^3+/M²⁺ redox couple participation in ORR. The carbon sheet established covalent interfacial M–O–C interactions. The C–O electron sharing reduces the electron density around M. This shifts the redox couple to the M²⁺, which improves the ORR electrocatalytic activity [14]. With that in mind, it is worth mentioning that the nearly two-electron electrocatalysts can also be enhanced by utilizing carbon support to enhance electron and charge movement. In turn, the improved electron and charge movement might improve the n number during ORR, pointing them to a nearly four-electron pathway. Thus, it should be noted that employing a support material enhances the electrocatalytic activity that follows the same activity order as the evaluations from CV. The comparison of the as-prepared electrocatalyst with the previously reported spinel oxide-based catalysts further expounds Mn_0.5Ni_0.5Co₂O₄/C being a promising candidate for ORR electrocatalysts (Table S4).

For the commercial application of the electrocatalysts, durability and lifespan studies are important. To this end, 405 scan cycles were performed within the 0.55 to 1.35 (V vs. RHE) range in O₂-saturated KOH _(aq), and the potential sweep voltammograms shown in Figure S10 were produced. After 405 scan cycles, a shift in current peaks characterized by current reduction was observed. Of all the synthesized Mn_xNi_1−xCo₂O₄ series (Figure S10a–f), Mn_0.5Ni_0.5Co₂O₄ was found to possess the least current loss (0.21%; Figure S10c) and MnCo₂O₄ the highest current loss of 17.03% (Figure S10e). At 0.11%, Mn_0.5Ni_0.5Co₂O₄/C had the smallest current loss (Figure S10f). The stability during the cyclization is ascribed to the Mn_0.5Ni_0.5Co₂O₄ and carbon sheet interaction. The current loss of the electrocatalysts followed the following order: Mn_0.5Ni_0.5Co₂O₄/C < Mn_0.5Ni_0.5Co₂O₄ < NiCo₂O₄ < Mn_0.3Ni_0.7Co₂O₄ < Mn_0.7Ni_0.3Co₂O₄ < MnCo₂O₄ (Table S3). This high durability is also attributed to the alkaline electrolyte because the presence of hydroxide ions in solution enhances metal hydroxide (M–OH) formation, stabilizing the cyclic voltammetry [64].

Apart from the electrochemical catalytic activity of the as-prepared electrocatalyst, catalytic stability is another crucial factor to consider for the practical applicability of the electrocatalysts. The long-term stability of Mn_xNi_1−xCo₂O₄ was tested using the chronoamperometry technique to evaluate the current loss percentage at a fixed potential of 1.2 (V vs. RHE) for 2000 s (Figure 9a,b) and 3600 s for Mn_0.5Ni_0.5Co₂O₄/C (most active; Figure 9c) in an O₂-saturated 0.1 M KOH solution. For the first few seconds, a steady current decay was observed for all the electrocatalysts. After 250 s, the chronoamperometry curves flattened, thus confirming stability for all the unsupported electrocatalysts, with Mn_0.5Ni_0.5Co₂O₄ showing excellent stability. From the chronoamperometry curves, the stability measured as retained current followed the order Mn_0.5Ni_0.5Co₂O₄/C > Mn_0.5Ni_0.5Co₂O₄ > NiCo₂O₄ > MnCo₂O₄ > Mn_0.3Ni_0.7Co₂O₄ > Mn_0.7Ni_0.3Co₂O₄ > MnCo₂O₄, which correspond to Table 1. The stability of Mn_0.5Ni_0.5Co₂O₄ can be explained following a series of reactions that show how ~50–50% of Mn²⁺/Mn³⁺ is essential.

Figure S10f shows that Mn_0.5Ni_0.5Co₂O₄/C has the highest current retention, suggesting a very stable cyclization behavior. Since this study focuses on alcohol fuels, there is a possibility of ethanol crossover during fuel cell operation. Hence, an ethanol poisoning test was conducted for 3600 s using chronoamperometry (Figure 9c). Ethanol addition was done at different time intervals as shown in Figure 9c. Then, 5 mL of 1 M ethanol was first added at the point where the chronoamperometry curve for Mn_0.5Ni_0.5Co₂O₄/C without ethanol started to stabilize (first addition). A sharp current decay was observed at the first addition, indicating a set of exposed active sites were easily poisoned. This poisoning might have been due to the fast formation of CH₃CH₃OH_ad, CO_ad, and CH₃CO_ad. Thus, some of the active sites have an excellent affinity for ethanol. During the second and third addition of 5 mL of 1 M ethanol solutions, the chronoamperometry curve showed a gradual current decay that reached a steady state after ~800 s.

Interestingly, when ethanol was added, anodic peaks were observed on the upper-side of the chronoamperometry curve. This is an indication that the remaining sites can oxidize ethanol and thus minimize poisoning. After the fourth addition of ethanol, the magnitude of ethanol oxidation remains the same, indicating that the electrocatalyst must have had the same set of exposed functionally active sites that resist ethanol poisoning.

To get more insight into the kinetics of electrocatalysts, we investigated them using electrochemical impedance spectroscopy (EIS). The insert shows a circuit fit, where R_CT is the charge transfer resistance, R_S the solution resistance, and CPE the constant phase element. The results presented in Figure 10a and Table 3 show that Mn_0.5Ni_0.5Co₂O₄ electrode possesses the smallest R_CT (charge transfer resistance) value of 272.43 Ω when compared with the other Mn_xNi_1−xCo₂O₄ series. The R_CT value was even smaller when Mn_0.5Ni_0.5Co₂O₄ was supported on a carbon sheet (Mn_0.5Ni_0.5Co₂O₄/C; 92.31 Ω; Figure 10b). A lower charge transfer resistance results in reduced energy loss and greater energy recovery efficiency in a catalytic reaction [65]. This translates to improved electrocatalytic activity for ORR. Thus, a faster charge transfer indicates fast electron transfer in the electrolyte, leading to good electrical conductivity and good pore accessibility on the surface of the electrode. Therefore, the smaller the semi-circle, the lower the total resistance. To this end, the charge transfer kinetics and conductivity will be high. The fast charge transfer can be attributed to differing redox processes within the spinel oxide, that is, Mn²⁺/Mn³⁺, Ni²⁺/Ni³⁺_, and Co²⁺/Co³⁺ coupled to the synergy of Ni/Co/Mn/O within the spinel oxide system. The cobalt intermediate spin state results in a half-metal characteristic, which in turn enhances conductivity. Thus, there is more favorable reaction kinetics towards electro-reduction of oxygen due to an enhanced electron transfer. The fitted R_CT values for the electrocatalysts shown in Table 3 follow the order Mn_0.5Ni_0.5Co₂O₄/C > Mn_0.5Ni_0.5Co₂O₄ > NiCo₂O₄ > Mn_0.7Ni_0.3Co₂O₄ > Mn_0.3Ni_0.7Co₂O₄ > MnCo₂O₄. Generally, Figure 10 shows a decrease in resistance when the concentration or number of defects are increased. Thus, metallic defects can act as a charge transfer enhancer along the defective channels within the spinel structure.

3. Materials and Methods

3.1. Synthesis of Mn_xNi_1−xCo₂O₄

A modified citrate sol–gel method [13] was utilized to synthesize Mn_xNi_1−xCo₂O₄ (x = 0, 0.3, 0.5, 0.7, and 1). Metal precursors Mn(Ac)₂.4H₂O (99% purity), Ni(Ac)₂.4H₂O (98% purity), and Co(Ac)₂.4H₂O (AR grade) were obtained from Sigma Aldrich, Darmstadt, Germany, and citric acid monohydrate (C₆H₈O₇.H₂O) was acquired from SAR-CHEM. In a typical synthesis, a stoichiometric measure of Mn, Ni, and Co metal precursors was dissolved in double-distilled water under ambient conditions in a 500 mL beaker, with a magnetic stirrer bar, and placed on a magnetic stirrer (model 725, LABOTEC South Africa) set at 150 rpm to form a homogeneous mixture of Mn, Ni, and Co metal acetate solution. Amid constant mixing of metal acetate (molar proportion of metal ions to citric acid: 1:1.5), there was a slow addition of citric acid. The pH of the resultant solution was kept at 7.0 by the controlled addition of a 25% solution of NH₃ acquired from ACE Chemicals, (Theta, T, Johannesburg South). Then, 40 mL of ethylene glycol was added to the mixture to assist with the gelatinization. The resulting suspension was stirred for 6 h at 85 °C to produce a gel, which was dried overnight at 100 °C. Calcination was performed at 450 °C in air and a ramping temperature of 2 °C min⁻¹ for 5 h inside a quartz tube furnace to produce a black powder.

3.2. Synthesis of Porous Mn_0.5Ni_0.5Co₂O₄/C Nanosheets

The synthesis followed a modified template-assisted citrate sol–gel method reported by spinel MnCo₂O₄ nanoparticles cross-linked with two-dimensional porous carbon nanosheets as a high-efficiency oxygen reduction electrocatalyst. Typically, a mixture of 122.5 mg of Mn(Ac)₂∙4H₂O, 122.5 mg of Ni(Ac)₂∙6H₂O, 498 mg of Co(Ac)₂∙4H₂O, 1.6 g of C₆H₁₂O₆ (the carbon precursor), and 6 g of NaCl (as a template) was ground for 1 h 15 min using a pestle and mortar. After adding 9 mL of double-distilled water, the resultant solution was stirred for 30 min before being oven-dried for 8 h at 40 °C. The resulting product was mixed with 20 mL of 2 M CO(NH₂)₂ and 25 mg of polyvinylpyrrolidone (PVP) [66,67], and after that, transferred to a Teflon-lined stainless autoclave for 15 h at 100 °C. The autoclaved product was vacuum-dried overnight at 50 °C. After that, the sample was calcined at 450 °C in air and a ramping temperature of 2 °C min⁻¹ for 5 h inside a quartz tube furnace. After that, carbonization was performed at 750 °C for 6 h under an inert atmosphere of N₂. The resultant fluffy product was repeatedly washed with a mixture of distilled water and ethanol to remove NaCl to get the porous crystalline Mn_0.5Ni_0.5Co₂O₄ supported on carbon nanosheets (Mn_0.5Ni_0.5Co₂O₄/C). The final drying was carried out for 30 min in an oven set at a temperature of 40 °C.

3.3. Physicochemical Characterizations

Powder X-ray diffraction (XRD) measurements were conducted at room temperature on an X’Pert PRO PANalytical diffractometer (CuKα, λ = 1.5406 Å) over the 2θ range of 10–90°. FTIR measurements were carried out using a Perkin Elmer FTIR spectrophotometer. BET analysis was conducted using an ASAP Tristar II 3020 analyzer. A thermogravimetric study was conducted on a PerkinElmer Pyris 1 TGA PerkinElmer, (Doornfontein, DF, South Africa) using the Pyris Software. The thermogravimetric analysis was conducted over a temperature range of 25–910 °C ramping at 10 °C per min and an airflow rate of 90 mL min⁻¹. The microstructure images of the synthesized materials were obtained by scanning transmission electron microscopy (STEM, JEOL JEM-ARM200F, 200 kV) Auckland Park, APK, Gauteng, South Africa. X-ray photoelectron spectroscopy (XPS) was carried out to probe the electronic structure of the samples. XPS spectra were obtained at room temperature in an ultra-high vacuum (UHV) chamber with a base pressure of 2 × 10⁻¹⁰ mbar. The UHV chamber was equipped with a SPECS XR 50M monochromatized X-ray source with an Al anode (Al K_α excitation line, or hν = 1486.71 eV) SPECS PHOIBOS 150 hemispherical electron energy analyzer. The overall energy resolution of the combined analyzer + photon source system was set at 0.8 eV for the survey scans and 0.55 eV for all the other high-resolution spectra shown in this work.

3.4. Preparation and Electrochemical Characterization of Modified Electrodes

3.4.1. Preparation of Modified Electrodes

The modified electrode was prepared by dispersing dry Mn_xNi_1−xCo₂O₄ or Mn_0.5Ni_0.5Co₂O₄/C (10 mg) electrocatalyst in 1.5 mL ultrapure water containing 5 μL of 20 wt% of Nafion solution. The catalyst ink was ultrasonicated for 30 min. The homogeneous catalyst ink (10 µL) was drop-casted on the active area of the GCE and dried at room temperature. Commercial Pt/C ink was prepared following the same procedure. A three-electrode electrochemical cell encompassed by a glassy carbon electrode (GCE; standard electrode or rotating disk electrode), both as working electrodes, 3 M KCl (Ag/AgCl) as the reference, and a platinum counter-electrode.

3.4.2. Electrochemical Characterization

The electrochemical measurements were recorded using a standard three-electrode system on a Dropsens potentiostat 8000. Furthermore, a rotating disk electrode (RDE; Metroham-Autolab) was used to investigate the spinel oxide-based electrocatalysts’ ORR catalytic activity. Electrochemical impedance spectra (EIS) were recorded at 1.4 V (vs. RHE) within the frequency range 0.005 Hz to 10,000 kHz. Chronoamperometry measurements were recorded in O₂-saturated 0.1 M KOH at 1.2 (V vs. RHE). The CV measurements were performed within the 0.55 to 1.35 V (V vs. RHE) potential window at 50 mV s⁻¹ (scan direction from left to right) and the RDE measurements were conducted at varying rotation speeds between 200 and 2000 rpm in 0.1 M KOH, 10 mV s⁻¹ in the potential window 0.6 to 1.5 (V vs. RHE). The Koutecky–Levich plots (J⁻¹ vs. $\mathsf{\omega }$^−1/2) were analyzed at different potentials. The electron transfer during ORR was calculated based on the following Koutecky–Levich equation:

$$\frac{1}{\mathrm{J}}=\frac{1}{{\mathrm{J}}_{\mathrm{L}}}+\frac{1}{{\mathrm{J}}_{\mathrm{K}}}$$

(15)

${\mathrm{where} \mathrm{J}}_{\mathrm{L}}=0.2{\mathrm{nFAD}}^{2/3}{\mathsf{\omega }}^{1/2}{\mathrm{v}}^{-1/6}\mathrm{C}$,${ \mathrm{J}}_{\mathrm{K}}=\mathrm{nFKC}$, n is the number of electrons transferred in the ORR, F is the Faraday constant (96,485 mol L⁻¹), D is the diffusion coefficient of O₂ (1.9 × 10⁻⁵ cm²s⁻¹), $\mathsf{\omega }$ is the electrode rotation speed,$\mathrm{v}$ is the solution kinetic viscosity (0.01 cm²s⁻¹), and $\mathrm{C}$ is the bulk concentration (1.2 × 10³ mol L⁻¹).

4. Conclusions

This work presents a modified citric acid-assisted sol–gel method to synthesize ternary spinel oxides Mn_xNi_1−xCo₂O₄ (x = 0, 0.3, 0.5, 0.7 and 1) as effective electrocatalysts for the oxygen reduction reaction. Among the series, Mn_0.5Ni_0.5Co₂O₄ (x = 0.5) was the best-performing electrocatalyst for reducing oxygen carried out in alkaline medium. The electrocatalytic activity of Mn_0.5Ni_0.5Co₂O₄ is further enhanced after supporting it on the carbon sheet (Mn_0.5Ni_0.5Co₂O₄/C). This provides evidence for the important role carbon supports play in the enhancement of electrocatalytic activity. The catalytic activity increased in the order Mn_0.5Ni_0.5Co₂O₄/C > Mn_0.5Ni_0.5Co₂O₄ > NiCo₂O₄ > Mn_0.3Ni_0.7Co₂O₄ > Mn_0.7Ni_0.3Co₂O₄ > MnCo₂O₄. The improved electrocatalytic activity of the synthesized Mn_0.5Ni_0.5Co₂O₄/C was due to the high surface area (209.52 m²g⁻¹), high porosity of the support material, and the high oxygen vacancy sites and it has demonstrated high stability (retained current of 86.20%). Mn_0.5Ni_0.5Co₂O₄/C is shown to follow a four-electron pathway in ORR.