Transition Metal-Based Polyoxometalates for Oxygen Electrode Bifunctional Electrocatalysis

Abstract

Polyoxometalates (POMs) with transition metals (Co, Cu, Fe, Mn, Ni) of Keggin structure and lamellar-stacked multi-layer morphology were synthesized. They were subsequently explored as bifunctional electrocatalysts for oxygen electrodes, i.e., oxygen reduction (ORR) and evolution (OER) reaction, for aqueous rechargeable metal-air batteries in alkaline media. The lowest Tafel slope (85 mV dec⁻¹) value and the highest OER current density of 93.8 mA cm⁻² were obtained for the Fe-POM electrocatalyst. Similar OER electrochemical catalytic activity was noticed for the Co-POM electrocatalyst. This behavior was confirmed by electrochemical impedance spectroscopy, where Fe-POM gave the lowest charge transfer resistance of 3.35 Ω, followed by Co-POM with R_ct of 15.04 Ω, during the OER. Additionally, Tafel slope values of 85 and 109 mV dec⁻¹ were calculated for Fe-POM and Co-POM, respectively, during the ORR. The ORR at Fe-POM proceeded by mixed two- and four-electron pathways, while ORR at Co-POM proceeded exclusively by the four-electron pathway. Finally, capacitance studies were conducted on the synthesized POMs.

1. Introduction

Ongoing energy crises and rising CO₂ emissions have highlighted the importance and urgency of transitioning from a fossil-fuels-based economy to one powered by cleaner, low- or no-carbon sources. However, the intermittent nature of renewable power sources (wind or solar energy) causes energy production to fluctuate over time, limiting their applicability and requiring eco-friendly energy storage devices.

Rechargeable aqueous metal–air batteries are sustainable and low-cost energy storage devices, manifesting high energy density and good ionic conductivity [1,2]. Nevertheless, their limitation lies in the sluggish kinetics and high overpotentials of the air electrode during discharge (oxygen reduction reaction, ORR) and charging (oxygen evolution reaction, OER) processes [2,3,4,5,6]. Being multistep electron transfer reactions, ORR and OER highly depend on the electrode material, with Pt/C being the most active electrocatalyst for ORR [3,4] and IrO₂ and RuO₂ for OER [5,6]. However, the limited stability, scarcity, and high price of these materials and the fact that they cannot efficiently catalyze both charging and discharge reactions encourage the development of new electrode materials for rechargeable metal-air batteries.

Polyoxometalates (POMs) are polyatomic transition metal oxyanions with a tridimensional framework, emerging as promising electrocatalysts owing to their high thermal stability, tunable structure, and mechanical and redox properties [7,8]. Thus, they have found applications as electrocatalysts for ORR [9,10,11,12] and OER [11,13,14,15] in fuel cells, batteries [9,11,13,16], and electrolytic cells [14]. Adjusting the electrochemical and mechanical properties of POMs can be accomplished by varying the composition of their framework by introducing different metal redox centers [17,18] or incorporating the whole POM structure into porous materials or conductive substrates [7]. Moreover, the effect of intrinsic structure adjustment in terms of anion structure regulation on POM electrocatalytic activity has been explored [19]. Anionic structure regulation led not only to the increased number of catalytically active sites, but further enhanced the catalytic site activity due to the enhancement of TM–oxygen covalence. Thus, the choice of an appropriate method for POM synthesis governs the physicochemical properties of a material (composition, surface area, porosity, etc.) and, consequently, electrochemical behavior in surface reactions at the electrolyte/electrode interface [19,20]. Adjustment of the porosity improves the contact between active sites and reactants, further enhancing POM electrocatalytic performance.

This work assessed the catalytic properties of Keggin-structure POMs with transition metal centers (Fe, Co, Cu, Mn, and Ni) toward ORR and OER (Scheme 1). The aim was to identify the most efficient lower-cost bifunctional catalysts for these reactions and understand the influence of the metal center on the structure-function properties of this POM family. These will enable the development of a rational design for POM-based electrocatalytic materials. Prepared TM-POMs were characterized by X-ray diffraction (XRD) analysis, scanning electron microscopy with integrated energy-dispersive X-ray spectroscopy (SEM-EDS) and transmission electron microscopy (TEM), and, finally, nitrogen sorption analysis. Electrochemical performance was run using voltammetry and electrochemical impedance spectroscopy (EIS).

2. Materials and Methods

2.1. Synthesis and Characterization of Five POMs Electrocatalysts

K₆Na₄[X₄(H₂O)₂(PW₉O₃₄)₂]·24H₂O POMs, where X = Ni, Mn, Cu, Co, and Fe were synthesized using a procedure similar to the one proposed by Clemente-Juan et al. [21]. Firstly, five 22 mM solutions were prepared by dissolving MnSO₄·H₂O, NiSO₄·6H₂O, (NH₄)₂Fe(SO₄)₂·6H₂O, CoSO₄, and CuSO₄·5H₂O in 5 mL of deionized water each. Secondly, 50 mL of a solution containing 0.1 M Na₂WO₄·2H₂O and 11 mM Na₂HPO₄ was vigorously stirred with a magnetic stirrer for 5 min, while the pH of the obtained solution was adjusted to 7 with the concentrated acetic acid. Finally, the prepared solutions of transition metal ions were added to 10 mL of tungstate solution each, and the obtained mixtures were refluxed for 2 h and then hot-filtered. A 0.4 g quantity of KCl was added to the hot filtrate, and the solution was left to crystalize overnight. The crystals were subsequently separated from the solution by filtration, redispersed in deionized water, and heated until dissolved. The following day, the recrystallized products were obtained by decanting excess solution and air-drying at room temperature.

Their structures were determined by XRD analysis using a Rigaku Ultima IV diffractometer in Bragg–Brentano geometry with Ni-filtered CuKα radiation (λ = 1.54178 Å). Moreover, Fourier transform infrared (FTIR) spectroscopy analysis was performed using a Perkin Elmer GX1 spectrometer. Scanning electron microscope Phenom™ ProX Desktop SEM (ThermoFisher Scientific™, Waltham, MA, USA) with integrated EDS detector and TEM performed on a HITACHI H-8100 microscope (Microscopy and analytical imaging research resource core laboratory, Lawrence, Kansas)were used for a complete examination of surface morphology, microstructure, and atomic composition of five TM-POMs. The sorption parameters were determined by analyzing nitrogen sorption isotherms at a temperature of 77 K. The measurements were conducted using a Micromeritics ASAP 2060 Surface Area Analyzer (MicrotracBEL, Osaka, Japan), following prior degassing of the samples at 150 °C for 36 h.

2.2. Electrocatalysis with Transition Metal POMs

The activity of the prepared POMs for OER and ORR electrocatalysis was assessed using a three-electrode cell with a saturated calomel electrode (SCE) as the reference, a graphite rod as the counter electrode, and the TM-POM deposited on a glassy carbon rotating disk electrode (RDE) as working electrode. All potentials are converted and presented versus the reversible hydrogen reference electrode (RHE). All electrodes were placed in 1 M KOH (80 mL) as the supporting electrolyte. The measurements were taken using an ALS/DY2325 bipotentiostat (ALS Co., Osaka, Japan) and an Ivium V01107 Potentiostat/Galvanostat (Ivium Technologies, Eindhoven, The Netherlands).

For the catalytic ink preparation for OER/ORR and double-layer capacitance studies, a mixture of powdered TM-POM (5 mg), distilled water (600 μL), ethanol (400 μL, 96%), and Nafion™ (25 μL, 0.5 wt.%) was sonicated in an Emmi^®—08ST sonicator, EMAG Technologies (Ann Arbor, MI, USA). The inks were pipetted onto the RDE (0.071 cm²), leading to a catalyst loading of 0.137 mg cm⁻². For the capacitance study, a mixture was composed of TM-POM (5 mg), activated carbon (1.35 mg), ethanol (27.5 μL), and Nafion™ (7.3 μL, 5 wt.%).

For capacitance measurements and one of the ORR studies, the electrolyte was saturated with N₂ gas for 20 min before the experiments. The remaining ORR experiments were run in an O₂-saturated electrolyte. The rotation rate of the disc electrode was varied: 300, 600, 900, 1200, 1800, 2400, and 3600 rpm at a scan rate of 5 mV s⁻¹. Lastly, electrochemical impedance spectroscopy experiments were conducted in the frequency range from 100 kHz to 0.1 Hz at 1.868 V.

Two different electrolytes were used for capacitance measurements, namely, 3 M KOH as alkaline media and 1 M H₂SO₄ as acidic media. The voltammograms were run from 0.20 to 0.62 V for alkaline media and 0.25 to 0.95 V for acidic media, using increasing scan rates of 5, 10, 20, 50, 100, and 200 mV s⁻¹. Following these measurements, potential cycling was performed for the best material by running 1000 cycles in 3 M KOH and 1 M H₂SO₄ at 0.5 A g⁻¹ and 0.25 A g⁻¹, respectively.

Finally, EIS was performed in acidic media for all TM-POMs at open circuit potential (OCP) to determine electrolyte and charge transfer resistance.

3. Results

3.1. Characterization of TM-POMs

SEM images, together with the EDS spectra of the five POM samples, are presented in Figure 1. Two segregated phases with nanowires and large agglomerate parts were noticed in SEM images of Ni-POM (Figure 1A), Mn-POM (Figure 1B), Fe-POM (Figure 1C), and Co-POM (Figure 1E) samples, while Cu-POM (Figure 1D) also showed the presence of spherical particles. EDS spectra of five POM samples confirmed the presence of elements of POMs combined with Ni, Mn, Fe, Co, and Cu in the corresponding POM (Figure 1F illustrates the case of Fe-POM).

TEM images of POM samples showed a mixed multi-layered structure of stacked lamellae obtained for Cu-POM, Co-POM, and Ni-POM (Figure 2A,C,E), and nanobelts for Mn-POM and Fe-POM (Figure 2B,D). Q. Liu et al. investigated cluster-assembled materials based on POMs and showed five different morphologies of Keggin POM-POSS Janus clusters in different solvent mixtures, confirmed by TEM analysis [22]. The morphology of obtained POMs [22] is similar to that of the POM materials synthesized herein.

Figure 3A shows typical XRD patterns for the Keggin structure of five TM-POM samples [23,24,25,26,27]. The four characteristic peaks at 2θ of 25.02, 26.60, 37.5, and 42.01° were observed for all TM-POMs samples [24]. Lower peak intensities in the cases of Fe-POM and Ni-POM indicate lower crystallinity of their structures. Such structural disorder and the higher number of defects in nanomaterials of low crystallinity can give rise to higher electrochemical performance compared to their highly crystalline counterparts, as previously observed in the case of an electrocatalyst based on Fe₂O₃ [28] 10.3390/molecules28010434.

The FTIR spectra of the TM-POMs display bands characteristic of POMs in the 1100–740 cm⁻¹ region (Figure 3B), thus confirming the POM structure. Namely, the bands at 1040, 945, 877, and 806 cm⁻¹ correspond to the frequency of P-O in the central PO₄ and W=O in the exterior WO₆, W−O_b−W, and W−O_c−W bridges, respectively [29]. A characteristic shoulder in the P–O position results in a decreased symmetry of PO₄ tetrahedron. The bands at ~3490 and 1580 cm⁻¹ are attributed to the stretching and banding vibration of the O–H bond within free H₂O molecules, respectively.

The surface area of POMs was investigated by conducting N₂ sorption measurements.

The isotherms exhibited by all POMs were a combination of type I and IV, accompanied by H3 hysteresis loops. The resulting BET surface areas were approximately 3.2 m² g⁻¹ for Co-POM, 1.9 m² g⁻¹ for Cu-POM, 2.0 m² g⁻¹ for Fe-POM, 1.0 m² g⁻¹ for Mn-POM, and 3.0 m² g⁻¹ for Ni-POM (Figure S1). To determine the pore size distribution (PSD), the nonlocal density functional theory (NLDFT) method was employed. Analysis of the PSD revealed distinct peaks for Co-, Cu-, Mn-, and Ni-POMs, indicating a significant concentration of mesopores within the size range of approximately 9–47 nm (Figure 4). However, in the case of Fe-POM, the PSD analysis exhibited peaks in the range of ca. 32–83 nm, indicating the presence of both mesopores and macropores (Figure 4C).

3.2. Catalysis of OER Using TM-POMs

TM-POMs were studied under OER polarization conditions in 1 M KOH electrolyte solution (Figure 5A). TM-POMs possess unique redox-active metal oxide clusters that can facilitate electron transfer processes, making them attractive candidates for electrocatalytic applications. Herein, Fe-POM showed the highest current density during OER (Equations (1)–(5)) of 93.8 mA cm⁻², followed by Co-POM with 43.1 mA cm⁻² and Ni-POM with 11.8 mA cm⁻², while Cu-POM and Mn-POM were found to have a negligible current density at 2 V. Onset potential (E_onset), i.e., the potential to reach a current density of 1 mA cm⁻², was found to increase in the order Ni-POM (1.453 V) > Fe-POM (1.588 V) > Co-POM (1.683 V) > Cu-POM (1.948 V) > Mn-POM (1.966 V),

Table 1. Kinetic parameters for OER at TM-POMs compared with similar literature reports.

Catalyst	Eonset/V	η10/mV	j400/mA cm−2	b/mV dec−1	Source
Fe-POM	1.588	434	4.78	87	This work
Co-POM	1.683	551	0.14	99	This work
Ni-POM	1.453	752	4.13	241	This work
Cu-POM	1.948	–	0.04	598	This work
Mn-POM	1.966	–	0.03	672	This work
([PVIM][V-Co4])	–	430	–	–	[30]
PNiW11@amZIF	–	375	–	69	[31]
PW12@amZIF	–	423	–	86	[31]
PW12@FeOOH-P	–	230	–	79	[32]
PMo12@FeOOH-P	–	235	–	123	[32]
SiW12@FeOOH-P	–	260	–	126	[32]
FeOOH-P	–	290	–	202	[32]
CNCP-0	–	310	–	105.8	[33]
CNCP-1	–	293	–	88.2	[33]
CNCP-12	–	269	–	79.5	[33]
IrO2	–	–	–	90.2	[33]
Fe4@MWCNT_N6	–	580	–	102	[34]
NiP4Mo6	–	250	–	73	[9]

. Fe-POM showed the lowest overpotential to reach a current density of 10 mA cm⁻² (η₁₀) among the studied materials, with a value of 434 mV, i.e., 117 mV and 318 mV lower than that of Co-POM (551 mV) and Ni-POM (752 mV), respectively, Table 1. Previously studied POM combined with poly(vinyl butyl imidazolium) cations and cobalt ([PVIM][V-Co₄]) showed an η₁₀ of 430 mV in alkaline media, which is similar to the herein-obtained value for Fe-POM [30]. Ni-modified Keggin-type POM ([PNi(H₂O)W₁₁O₃₉]⁵⁻) prepared with the zeolitic imidazolate framework ZIF 67 (PNiW₁₁@amZIF) and PW₁₂@amZIF examined for OER in 1 M KOH showed an η₁₀ of 375 mV and 423 mV, respectively, which is somewhat lower/similar to the value obtained herein [31]. Cu-POM and Mn-POM did not reach a current density of 10 mA cm⁻² within the studied potential range. Furthermore, the current density at an overpotential of 400 mV (j₄₀₀) was found to be the highest for Fe-POM and to decrease in the order Fe-POM (4.78 mA cm⁻²) > Ni-POM (4.13 mA cm⁻²) > Co-POM (0.14 mA cm⁻²) > Cu-POM (0.04 mA cm⁻²) > Mn-POM/rGO (0.03 mA cm⁻²), Table 1.

Tafel analysis was then performed to gain deeper insight into the electrocatalytic performance of the studied TM-POMs for OER, Figure 4B. The lowest Tafel slope (b) value of 87 mV dec⁻¹ was determined for OER at Fe-POM. Co-POM (99 mV dec⁻¹) showed a somewhat higher Tafel slope value, while Ni-POM (241 mV dec⁻¹), Cu-POM (598 mV dec⁻¹), and Mn-POM (672 mV dec⁻¹) showed notably higher values, Table 1. Shang et al. [32] prepared different POM electrocatalysts, which were substituted with transition metal and mixed with β-FeOOH to obtain PW₁₂@FeOOH-P, PMo₁₂@FeOOH-P, SiW₁₂@FeOOH-P, and FeOOH-P. The lowest Tafel slope of 79 mV dec⁻¹ was obtained for PW₁₂@FeOOH-P, and then somewhat higher values of 123 and 126 mV dec⁻¹ were obtained for PMo₁₂@FeOOH-P and SiW₁₂@FeOOH-P, respectively. The highest Tafel value was calculated to be 202 mV dec⁻¹ for the FeOOH-P electrocatalyst during OER. These results are comparable to those obtained for Fe-POM and Co-POM (Table 1). Different CNCP electrocatalyst nanoframes were prepared with a dodecahedral POM nanostructure and [Co(CN)₆]³⁻ nanochannels and examined for OER in alkaline media [33]. CNCP-0, CNCP-1, CNCP-12, and IrO₂ showed Tafel slopes of 105.8, 88.2, 79.5, and 90.2 mV dec⁻¹, respectively. These values are comparable with the herein-determined values. Additionally, the best-performing material within this study, Fe-POM (87 mV dec⁻¹), has a Tafel slope comparable to or lower than those reported in the literature. For instance, Fe₄@MWCNT_N6 (102 mV dec⁻¹) electrocatalyst prepared with Fe-POMs compound and multi-walled carbon nanotubes [34] as well as NiP₄Mo₆ (73 mV dec⁻¹) synthesized with POM and appropriate Ni and Mo salts [9].

3.3. Catalysis of ORR Using TM-POMs

The performance of TM-POMs was next evaluated under ORR polarization conditions in 1 M KOH. A comparison of the current densities recorded in N₂- and O₂-saturated solutions shows an apparent increase in the presence of O₂ in the case of all TM-POMs, corresponding to oxygen reduction [35]. Figure 6A illustrates the case for Fe-POM (Figure S2 presents data for other TM-POMs). The most positive value of ORR peak potential (E_p) of 0.632 V was noticed for Fe-POM (Figure 6A), followed by 0.622, 0.558, 0.529, and 0.494 V recorded in the case of Mn-POM, Co-POM, Cu-POM, and Ni-POM, respectively. Herein, obtained peak potential values are in agreement with the literature reports [36,37,38]. All TM-POMs showed ORR peak current density of the same order of magnitude.

RDE voltammograms of five TM-POMs collected at 1800 rpm are compared in Figure 6C. Mn-POM gave the highest ORR current density, followed by Fe-POM and Cu-POM, with somewhat lower ORR current densities. On the other hand, Ni-POM and Co-POM gave two times lower ORR current densities. Fe-POM, Cu-POM, and Mn-POM show slightly lower onset potential than Ni-POM and Co-POM (

Table 2. Kinetic parameters of ORR for TM-POMs compared with similar literature reports.

Electrocatalysts	Eonset/V	E1/2/V	Epp/V	b/mV dec−1	n	Source
Fe-POM	0.695	0.580	0.632	85	2.8–2.9	This work
Co-POM	0.655	0.552	0.558	109	3.8–3.9	This work
Ni-POM	0.639	0.533	0.494	144	2.3	This work
Cu-POM	0.695	0.580	0.529	94	2.4–2.5	This work
Mn-POM	0.694	0.556	0.622	107	2.6–2.9	This work
NiP4Mo6	–	–	–	106	3.80–3.85	[9]
S-NiP4Mo6	–	–	–	98	2.08–2.15	[9]
Co4(PW9)2@N-CNT	0.90	–	–	89	2.73	[39]
[Mo-oxo]n clusters	0.75	–	0.62	109	~1.5	[40]
Fe4@MWCNT_N6	0.81	–	–	35.4	2.9	[34]
Ni4@MWCNT_N6	0.80	–	–	34.7	2.7	[34]
Fe2Ni2@MWCNT_N6	0.80	–	–	37.9	3.2	[34]
[PW11MO39]5@Ru-rGO (M = Co, Ni, Cu)	–	–	–	–	3.9–4.2	[41]

). Fe-POM and Cu-POM gave the most positive values of half-wave potential (E_1/2) of 0.580 V, whereas Co-POM and Mn-POM gave values lower for 28 and 24 mV, respectively. The lowest value of E_1/2 during ORR was found to be for the Ni-POM.

Figure 6C shows obtained Tafel plots and slopes (b) for five TM-POMs, where Fe-POM and Cu-POM offered the lowest values of 85 and 94 mV dec⁻¹ (Table 2), Mn-POM and Co-POM gave a slightly higher value of 107 and 109 mV dec⁻¹, and finally Ni-POM showed the highest value of 144 mV dec⁻¹. Similar POM electrocatalysts, NiP₄Mo₆ and S-NiP₄Mo₆, presented Tafel slopes of 106 and 98 mV dec⁻¹, respectively [9]. Nanocomposite synthesized of phosphotungstate with carbon nanotubes doped with nitrogen (Co₄(PW₉)₂@N-CNT) led to an ORR Tafel slope 89 mV dec⁻¹, comparable with the herein-tested TM-POM electrocatalysts [39]. POM-like metal oxo clusters with molybdenum ([Mo-oxo]_n clusters) synthesized on carbon showed a higher Tafel slope of 109 mV dec⁻¹ during ORR [40].

Subsequently, voltammograms were recorded using different electrode rotation rates, Figure 6B and Figure S3. Koutecký–Levich analysis was performed using background-corrected data. The number of electrons (n) transferred during ORR was determined from the slope of Koutecký–Levich plots (j⁻¹ vs. ω^−1/2). ORR n values for Fe-POM were found to range from 2.8 to 2.9, Figure 6D, suggesting that the reduction of oxygen proceeds by both the two-electron pathway, where HO₂⁻ is formed as an intermediate (Equations (6)–(8)), and four-electron pathway, where O₂ is directly reduced to OH⁻ (Equation (9)). n values for Cu-POM, Mn-POM, and Ni-POM were lower, suggesting a two-electron pathway. Similar results were reported for other POM electrodes [9,34,39,40]. The only exception was Co-POM with n values of 3.8–3.9 (Figure 6D), indicating that the process takes place exclusively through a four-electron pathway. [PW₁₁MO₃₉]^5-@Ru-rGO (M: Co, Ni, and Cu) electrocatalysts gave n between 3.9 and 4.2, which is comparable with the n value obtained for the herein-examined Co-POM electrocatalyst [41].

The results obtained for OER/ORR catalysis by Fe-POM illustrate the importance of the electrocatalyst’s electronic, structural, and morphological properties on its electrochemical performance. Fe’s ability to undergo reversible redox reactions and the combination of catalytically active sites in high oxidation states with their easy accessibility due to a hierarchical porous structure and low crystallinity give rise to the high performance of Fe-POM for oxygen electrode reactions, mainly OER. Fe provides active adsorption sites capable of effectively binding oxygen species, thus modifying the Gibbs energy associated with oxygen adsorption as an important step in ORR.

3.4. Capacitance and Impedance Study

The electrocatalytic performance of materials for oxygen electrode reactions relies significantly on their pseudocapacitive characteristics, particularly oxygen adsorption (underpotential deposition) on the electrocatalyst’s surface, which serves as a prerequisite for oxygen electrode processes. Enhanced pseudocapacitive behavior correlates with improved electrocatalytic performance, indicating a greater abundance of active sites available for participation in oxygen electrode processes. Characterization via cyclic voltammetry at various scan rates elucidates the rate capability of materials, reflecting their efficacy in adsorbing substantial amounts of oxygen and subsequently desorbing with high coulombic efficiency. Optimal pseudocapacitive behavior, manifested by the rectangular shape of CV curves, directly correlates with heightened electrocatalytic activity, as evidenced by lower overpotential values.

Double-layer capacitance (C_dl) values of studied TM-POMs (Figure S1) were found to be of the same order of magnitude, though higher in the case of Co-POM (400 µF cm⁻²) compared to the other four TM-POMs (100–200 µF cm⁻²). These values further suggest a similar electrochemically active surface area with active sites at the solid/electrolyte interface [35]. Thus, the difference in the TM-POMs’ electrochemical response suggests different reactivity of the available catalytically active sites for the studied oxygen electrode reactions among five TM-POMs. The oxidation state and accessibility of active sites, along with other factors like charge transfer resistance, might play a role.

The charge storage behavior of five materials was further scrutinized in both alkaline (3 M KOH) and acidic (1 M H₂SO₄) media; Figure 7A,B illustrate the case of Co-POM, and Figure S4A,B illustrate a comparison of five TM-POMs’ behavior at a common potential scan rate of 20 mV s⁻¹. The shape of the voltammograms clearly indicates the pseudocapacitance behavior of the studied TM-POMs in both media. Namely, both acidity and alkalinity in the electrolyte solution clearly activate faradaic processes in TM-POM surface atoms (at 0.75 V in 1 M H₂SO₄ and −1.12 V in 3 M KOH) that contribute to the charge storage capability of the materials’ surface [42].

Specific capacitance (C_s) values of the same order of magnitude were determined for the five TM-POMs (Table S1) (in accordance with the similar specific surface areas), though they were higher in the case of Co-POM (47.6–34.2 F g⁻¹) versus other materials—Mn-POM (37.0–22.4 F g⁻¹), Cu-POM (32.8–25.0 F g⁻¹), Ni-POM (27.3–20.9 F g⁻¹), and Fe-POM (25.9–16.6 F g⁻¹)—in acidic media. Furthermore, the capacitance values of the same order of magnitude were determined for two media, demonstrating that electrolyte solution composition does not notably affect the charge storage properties of TM-POM (Figure 7C, Table S1). The case of Co-POM that generally showed the highest values illustrates somewhat higher values of Co-POM in alkaline media compared to the acidic media at lower scan rates: capacitance of 51.6 and 47.2 F g⁻¹ in 3 M KOH versus 47.6 and 45.6 F g⁻¹ in 1 M H₂SO₄ at 5 and 10 mV s⁻¹, respectively. But at scan rate values of 20 mV s⁻¹ and higher, the opposite becomes true, and capacitance values in H₂SO₄ become higher than in KOH with values ranging from 44.0 to 34.2 F g⁻¹ versus 40.8 to 20.2 F g⁻¹.

Following this, continuous cycles were run in both media, with Figure 6D presenting specific capacitance values of Co-POM. Stable charge storage behavior could be observed with a decrease of only 2.95% over 1000 cycles in 1 M H₂SO₄. Moreover, overall higher capacitance values were determined in the KOH solution, increasing over time to 115.5% of the initial values. This phenomenon can be attributed to the increase in electrolyte–electrode contact area, where running the measurement over a longer time allows for electrolyte ions to enter the material’s pores [43].

EIS data of the five TM-POMs recorded at 1.868 V, i.e., in OER conditions, is presented in Figure 8A, which were fitted by the equivalent circuit given in the inset. Fe-POM gave the lowest charge transfer resistance (R_ct) of 3.35 Ω, followed by Co-POM, with a R_ct of 15.04 Ω (

Table 3. EIS parameters of five TM-POMs in 1 M KOH at 1.868 V.

Electrocatalyst	Rs/Ω	Re/Ω	Rct/Ω	Qe/mF	Qdl/mF
Fe-POM	5.27	0.15	3.35	3.20 × 10−3	7.40 × 10−4
Co-POM	50.49	1.36	15.04	3.25 × 10−4	1.93 × 10−2
Ni-POM	184.80	44.78	137.00	8.08 × 10−4	1.94 × 10−4
Cu-POM	31.49	22.08	67.28	4.72 × 10−4	6.57 × 10−4
Mn-POM	9.25	2.68	239.00	8.22 × 10−2	1.65 × 10−4

R_s—electrolyte resistance; R_e—electronic resistance of the material and the associated capacitance; R_ct—charge transfer resistance; Q_e—constant phase element; Q_dl—ideal double-layer capacitor.

). Significantly higher R_ct values of 67.28, 137.0, and 239.0 Ω were observed for Cu-POM, Ni-POM, and Mn-POM electrocatalysts, respectively. These parameters agree with the OER kinetics parameters given in Table 1. These confirm the higher catalytic activity of Fe-POM and Co-POM towards OER compared to the rest of the TM-POMs.

EIS was also run in 1 M H₂SO₄ at the OCP to evaluate the TM-POMs’ resistance in acidic media. Figure 8B evidences a high impedance for the five TM-POMs. Frequency values, in inverse proportion with the relaxation time during which charge storage and transport take place (i.e., lower the frequency, longer the time for electrochemical phenomena to take place [43]), were found to be similar for the studied TM-POMs.

4. Discussion

Considering the kinetic parameters analyzed above (Table 1 and Table 2), the optimum catalyst for OER can be determined. Fe-POM presented the lowest potential to reach the benchmark current density of 10 mA cm⁻², along with the lowest Tafel slope value and the highest OER current density of 93.8 mA cm⁻². Correspondingly, the current density at an overpotential of 400 mV delivered by Fe-POM was much higher than the other four composites, leading to the conclusion that this material exhibits the optimum performance for OER among the five studied TM-POMs. This material performance is also comparable to other POMs reported in the literature, Table 1 [8].

The catalytic activity of Fe-POM in OER stems from its ability to undergo reversible redox reactions, providing active sites for the adsorption and activation of reactant molecules (Equations (1)–(5)) and facilitating the transfer of electrons during the reaction [44]. Fe-POM indeed showed the lowest charge transfer resistance value under OER conditions. Its mentioned lower crystallinity with accompanying structure disorder and presence of defects boost the oxygen electrode kinetics. Additionally, the presence of Fe in POM can play a crucial role in promoting specific catalytic pathways and improving overall catalytic efficiency. The herein-prepared Fe-POM has a hierarchical porous structure that combines the advantages of macro- and mesopores for improved kinetics [35]. Macropores present in Fe-POM contributed to the enhanced electrode kinetics by increasing the effective surface area accessible to the electrolyte and available for electrochemical reactions. Although the primary surface area contribution comes from mesopores (2–50 nm) and micropores (<2 nm), macropores facilitate the accessibility of these smaller pores by electrolyte ions. Furthermore, present macropores act as channels facilitating electrolyte ion movement to/from the surface active sites and thus reducing the ions’ diffusion path length. By facilitating ion transport, macropores reduced the overall ionic resistance of Fe-POM. Finally, macropore presence is especially important in reactions involving gas evolution, such as the herein-studied OER improving gas management, i.e., allowing for generated gas bubbles to escape more easily, preventing the build-up of pressure, and ensuring the electrode surface remains accessible for the reaction.

M + OH⁻ → MOH + e⁻

(1)

MOH → MO + H⁺ + e⁻

(2)

2MO → 2M + O_2(g)

(3)

MO + H₂O → MOOH + H⁺ + e⁻

(4)

MOOH → M + O_2(g) + H⁺ + e⁻

(5)

Sum reaction: 4OH⁻ → O₂ + 2H₂O + 4e⁻

(6)

Regarding the lowest onset potential and Tafel slope, Fe-POM demonstrated the best ORR performance as well, followed by Co-POM. On the other hand, Ni-POM showed the lowest ORR performance with the highest Tafel slope and the lowest ORR current density. The reactivity of individual steps in an ORR is intricately linked to several factors, notably, the energy associated with oxygen adsorption, the dissociation energy of the O-O bond, and the binding affinity of hydroxide ions (OH⁻) to the surface (Equations (7)–(10)).

O_2,ads + H₂O + 2e⁻ → HO_2(ads)⁻ + OH⁻

(7)

HO_2(ads)⁻ + H₂O + 2e⁻ → 3OH⁻

(8)

2HO_2(ads)⁻ → 2OH⁻ + O₂

(9)

O_2(ads) + 2H₂O + 4e⁻ → 4OH⁻

(10)

Consequently, the heightened activity of Fe-POM in ORR may be attributed to its comparatively lower Gibbs free energy of oxygen adsorption, influenced by both geometric and electronic effects. Fe is posited to offer active adsorption sites capable of effectively binding oxygen species, thereby enhancing the overall adsorption capacity towards oxygen. Consequently, a plentiful presence of Fe sites on the catalyst surface can mitigate the Gibbs energy associated with oxygen adsorption during the initial step of the ORR.

5. Conclusions

Transition metal (Co, Cu, Fe, Mn, and Ni) polyoxometalates of the Keggin structure were synthesized and characterized by XRD, N₂-sorption, SEM-EDS, and TEM techniques. The bifunctional activity of TM-POMs for OER and ORR was evaluated in alkaline media. Fe-POM showed the best performances for both oxygen electrode reactions, followed by the Co-POM. Comparable Tafel slope values of 87 and 99 mV dec⁻¹ were obtained for Fe-POM and Co-POM during OER, respectively. The rest of the TM-POM electrocatalysts showed substantially higher Tafel slopes in the range of 241 to 679 mV dec⁻¹. The OER current density at an overpotential of 400 mV delivered by Fe-POM was considerably higher than that of the other four POMs. Furthermore, Fe-POM showed the lowest onset potential and Tafel slope during ORR, and these values of Fe-POM were followed by the Co-POM. The ORR for Cu-POM, Mn-POM, and Ni-POM proceeded by the two-electron pathway, for Fe-POM by mixed two- and two-electron pathways, and for Co-POM by a four-electron pathway. Fe-POM’s good performance arises from its ability to undergo reversible redox reactions along with its hierarchical porous structure and low crystallinity, i.e., easily accessible active sites with appropriate Gibbs free energy of oxygen species adsorption. The demonstrated activity of TM-POMs, specifically Fe-POM and Co-POM, for OER/ORR could be promising for their potential applications in rechargeable metal-air batteries and unitized regenerative fuel cells.