Polyoxometalates as Electrocatalysts for Electrochemical Energy Conversion and Storage

Abstract

Polyoxometalates (POMs) are polyatomic ions with closed three-dimensional frameworks. Their unique structure contains a large number of redox active sites, making them promising electrocatalysts for electrochemical energy conversion and storage applications. Thus, this paper presents an overview of the use of POMs as electrocatalysts for electrochemical energy conversion and storage devices, such as batteries, supercapacitors, fuel cells, or water electrolyzers. A discussion of the viability of these materials as alternatives to noble metal-based electrocatalysts is made. The current status of these materials to respond to the challenges of converting modern energy systems into more sustainable ones is also envisaged.

1. Introduction

Today’s society faces one of its biggest challenges in the form of climate change. This makes the pursuit of energy storage and production options that do not emit greenhouse gases mandatory for all developed countries [1,2]. This endeavor takes several forms: the replacement of fossil-fueled vehicles with electrical ones, the production of electricity via renewable sources for all forms of electrical consumption, and finally, efficient energy storage. This last facet is a mandatory part of the equation when it comes to renewable energy production since it tends to fluctuate greatly during the day and throughout the year in terms of energy produced.

In the endeavor to solve these issues, a potentially viable option has been found. The concept of the “Hydrogen Economy”, based on using hydrogen as an energy carrier, has now become a reality. However, most hydrogen comes from hydrocarbon reforming, a source of global pollution [3,4], which is certainly not sustainable. Using water electrolysis cells powered by renewable energy sources produces green hydrogen gas, a clean and efficient form of energy storage. Coupled with the fact that hydrogen can be fed to fuel cells to generate electricity, this makes the concept very interesting from a circular economy perspective [5].

Of course, this technology does not come without its limitations. Core reactions in water electrolyzers and fuel cells, oxygen evolution (OER) and oxygen reduction (ORR), as well as hydrogen evolution reaction (HER), have slow kinetics, requiring high overpotentials to obtain reasonable currents. Electrocatalysts are used to reduce the energy consumption in these reactions, with ruthenium oxide and platinum being the most efficient for OER and ORR/HER, respectively, but also high-cost materials. The ongoing search for low-cost, high-performance catalysts has proven to be difficult; most of these only operate in current ranges below 100 mA cm⁻², which is not satisfactory for production on an industrial scale. Another problem in testing these materials is the difference between temperature, pressure, and electrolyte concentration values of small-scale lab cells compared with industry-scale ones, which will lead to significant divergence in electrolyte conductivity, ion migratory flux, and catalyst structural stability [6]. For this reason, the present review paper focuses on the use of polyoxometalates (POMs) to overcome the aforementioned problems. POMs have shown interesting properties for OER/ORR and HER electrocatalysis, with good activity and low cost. Furthermore, they were shown to lead to promising performance of different types of batteries and supercapacitors when employed as electrode materials.

POMs are polyatomic ions composed of three or more transition metal (e.g., Mo, V, W) oxyanions linked together by shared oxygen atoms to form closed three-dimensional frameworks, Figure 1. The unique structure of POMs, which leads to a large number of redox active sites, makes them promising for energy conversion and storage applications. The study of POMs has been very in-depth, and applications have been found in the fields of material sciences, medical applications, and catalysis, which this review will cover more closely. Concerning this last field, important properties have been observed in these materials, such as high thermal stability, high sensitivity to electricity, and resistance to oxidative decomposition, making them prime candidates for electrocatalytic study.

POMs demonstrate distinctive redox characteristics and the capability of reversibly uptaking as much as 24 electrons per cluster unit [PMo₁₂O₄₀]³⁻ in the solid state [7,8,9,10,11]. This clearly indicates POMs as materials applicable for multi-electron transfer processes. The electrochemical behavior of POMs in terms of redox potential and number of electrons stored can be tailored additionally by, for example, the incorporation of redox-active metal centers through a chemical modification; such tailored-made POMs might be suitable for various electrochemical and electroanalytical applications [9].

POMs are normally insulators with only a few POMs demonstrating (semi-)conducting behavior. However, high electronic conductivity, often along with ionic conductivity, mainly protonic conductivity, is essential for the use of POMs as electrocatalysts in electrochemical energy conversion and storage devices. Thus, to use POMs in these devices, it is crucial to ensure adequate charge flow. Anchoring POMs through covalent or non-covalent means on nanostructured carbon materials, including carbon nanotubes (CNTs) and graphene, as conductive and high-surface area substrates, enables the mentioned charge flow [12]. “Wiring” of POMs to conductive organic polymers has also been suggested as a way to overcome the issue of their low conductivity [9].

2. POMs as Electrocatalysts for the Oxygen Evolution Reaction (OER)

One of the reactions involved in water electrolysis is the oxygen evolution reaction (OER), in which water is oxidized to O₂. This reaction is further essential for the operation of metal-air batteries and unitized regenerative fuel cells. It has different mechanisms depending on the pH of the electrolyte solution, as shown in Figure 2. In any case, OER encompasses all the problems that plague the production of hydrogen via electrolysis cells, as shown in Figure 3. It has high thermodynamic demands, and its kinetic obstacles lead to high activation energies. This translates into high overpotentials required for the reaction to start, making it evident that developing an efficient catalyst is mandatory when pursuing research in this field.

Good performance when using POMs as catalysts for the OER has already been reported [15]. Removing metal oxygen entities creates empty spaces in the POM structure, creating so-called lacunary POMs. These are good inorganic ligands that stabilize multi-metal oxide clusters, creating a very robust structure that can be used to assemble catalytically active transitional metal-oxo clusters. So, different POMs can be studied for the desired catalytic effect, with various metals serving as the active sites.

2.1. Ruthenium POMs

One of the transition metals that can act as an active site of POMs is ruthenium (Ru). Even though Ru is a noble metal and, consequently, expensive, it should be mentioned that it has been demonstrated to be a powerful oxidant in a near-neutral pH. For example, [Ru₄O₄(OH)₂(H₂O)₄(SiW₁₀O₃₆)₂]¹⁰⁻, or simply (Ru₄Si₂), has been reported as being highly active for OER [14,15], having a ratio of product formed per molecule of catalyst (turn-over number, TON) of up to 180 and initial d(TON)/dt (turn-over frequency, TOF) of up to 288 h⁻¹ [16,17]. [Ru₄(O)₅(OH)(H₂O)₄(PW₁₀O₃₆)₂]⁹⁻, or (Ru₄P₂) was also shown to be able to catalyze the OER, but at a slightly lower efficiency [18].

Furthermore, studies involving Ru₄Si₂ combined with conductive materials have also been done. For example, anchoring the material to a conductive bed of multi-walled carbon nanotubes (MWCNTs) leads to higher electrocatalytic activity when compared to simply Ru₄Si₂-functionalized amorphous carbon [19]; this is most likely due to the enhanced electron transfer in MWCNTs. With these materials, a TOF of 300 h⁻¹ was reached at an overpotential of 0.6 V [20].

Composite materials based on electrostatic immobilization of Ru₄Si₂ onto graphene followed by electrochemical deposition on glassy carbon have also been studied by Guo et al. [21]. This material showed good catalytic activity and stability under near neutral pH, with currents comparable to that of IrO₂ at an overpotential of 0.35 V.

According to Quintana et al. [22], another approach to increase the electrocatalytic performance of Ru₄Si₂ was to use graphene, covalently functionalized with organic, hydrogen-bonding, cations. This hybrid material shows an overpotential of 0.3 V to reach a current density of 0.150 mA cm⁻², and a negligible loss of performance even after 4 h of testing, all at neutral pH, performing at a higher efficiency than both isolated Ru₄Si₂ and its nanotube analogue.

2.2. Cobalt POMs

Cobalt (Co) is somewhat better than Ru in terms of higher abundance, but not particularly more sustainable. Namely, most of its reserves are located in countries with fewer regulations. which further leads to less emphasis on controlling the pollutants from the extraction of this metal [23]. Thus, Co may be considered a midway between the more sustainable transition metal-based POMs and the noble metal-based POMs.

Limani et al. [24] studied the viability of four cobalt phosphotungstate materials (MWCNT_N8_Co4, GF_N8_Co4, GF_ND8_Co4, and GF_NS8_Co4) as electrocatalysts. In alkaline media, with a pH of 13, all composites showed good performance, particularly GF_N8_Co4, which had an onset potential of 0.34 V vs. RHE and a maximum current of 70 mA cm⁻² at 2 V vs. RHE, while also maintaining around 73–82% of its current after about 5.5 h. It is important to notice that these data mean that this electrocatalyst outperforms state-of-the-art IrO₂.

2.3. Manganese POMs

As mentioned, although Co is more abundant and consequently cheaper than Ru, its use comes with some drawbacks. So, better alternatives are yet to be found, for example, manganese (Mn). However, only a small number of Mn-POMs have been studied. The lack of Mn examples is not exclusive to POMs, because Mn oxide is generally much less active than the corresponding Co or Ni analogs when considering heterogeneous catalysis [25]. A good place to start the search for Mn-POMs that exhibit OER activity seems to be the Mn-analogue of the well-studied Co₄P₂, Mn₄(H₂O)₂(PW₉O₃₄)₂]¹⁰⁻, or simply (Mn₄) [26]. Even though electrochemical water oxidation experiments show initial activity comparable to the Co counterpart, the current density decreased very rapidly, becoming negligible in 30 min with the formation of an inactive MnO_x layer on the electrode.

Wu et al. [27] also studied another six POMs comprising different Mn-O clusters and Mn in different oxidation states: (Mn₂-POM), (Mn₄-POM), (Mn₆-POM-1), (Mn₆-POM-4), (Mn₁₄-POM), and (Mn₁₉-POM), both with these materials in solution, and then deposited on the surface of indium tin oxide (ITO) electrodes to form the composite films. It was found that Mn₁₄-POM displays the highest electrocatalytic performance toward oxygen evolution. It was also concluded that the oxidation state of Mn and the cubic structure of Mn-O cluster are important factors impacting the POMs electrocatalytic performance for OER.

2.4. Nickel POMs

Extensive studies of Ni-based electrocatalysts for electrochemical energy-related applications showed that their performance often overcomes that of the benchmark electrocatalysts [28]. Nanostructured Ni-based materials have especially shown promising activity for oxygen electrode reactions, as well as HER.

When it comes to POMS, nickel POMs are some of the most recent ones; [Ni₅(OH)₆(OH₂)₃(SiW₉O₃₃)₂]¹²⁻ or (Ni₅Si₂) was only first reported in 2012 by Zhu et al. [29]. The research into this type of material continued with Singh et al. [30] who in 2018 reported a hybrid POM-supported Ni^II coordination complex, [(Ni^II(bpy)₂(H₂O))(HCo^IIW^VI1₂O₄₀)]₂³⁻, with Ni^II metal ion acting as the active center. This catalyst was highly stable and robust for OER at pH 7 having a high TOF of 18.49 (mol of O₂)(mol of Ni(II))⁻¹ s⁻¹ and a Tafel slope of 168.41 mV dec⁻¹.

A simple one-step hydrothermal deposition of microcrystals of a Dexter–Silverton POM, [Co_6.8Ni_1.2W₁₂O₄₂(OH)₄(H₂O)₈], on a commercial Ni foam, resulted in an industrially viable/applicable composite electrode for OER in alkaline electrolytes [31]. Namely, the overpotential to reach a current density of 10 mA cm⁻² under OER polarization conditions was observed to be as low as 360 mV. The material’s good activity towards OER was further reflected in the Tafel slope value of 126 mV dec⁻¹, accompanied by a faradaic efficiency of over 90%. Furthermore, the material demonstrated high mechanical and chemical stability and no detachment of the POM microcrystals from the metal foam support.

Similarly, (C₅H₇N₂)₆[NiW₁₂O₄₄] showed an overpotential of 347 mV to reach 10 mA cm⁻² complemented by long-term stability (up to 96 h) under OER polarization conditions in alkaline media [32]. NiO and WO_x (x = 1 or 2), with activity for OER, were generated in situ, further contributing to the electrocatalysis of OER.

Ni-POMs have also been shown to have a high photocatalytic activity to drive efficient water oxidation under visible light. Thus, Na₂₄[Ni₁₂(OH)₉(CO₃)₃(PO₄)(SiW₉O₃₄)₃]·56H₂O, Na₂₅[Ni₁₃(H₂O)₃(OH)₉(PO₄)₄(SiW₉O₃₄)₃]·50H₂O, and Na₅₀[Ni₂₅(H₂O)₂OH)₁₈(CO₃)₂(PO₄)₆(SiW₉O₃₄)₆]·85H₂O demonstrated high O₂ evolution TON of 128.2, 147.6, and 204.5, respectively [33]. The studied compounds’ photocatalytic activity was attributed to the band gap structures, numerous active sites, and the favorable structural design of POMs.

2.5. Copper and Iron POMs

Yu et al. [34,35] studied two Cu-POMs clusters, where [Cu₃(H₂O)₃(SbW₉O₃₃)₂]¹²⁻ was shown to catalyze OER at neutral pH without decomposition under homogeneous electrochemical conditions. Interestingly, [Cu₅(OH)₄(H₂O)₂(SiW₉O₃₃)₂]¹⁰⁻, the POM that could photochemically catalyze the same reaction, showed negligible electrochemical activity under the same conditions.

Azmani et al. reported the activity of Fe-POMs in relation to Co-POMs in 2021 [36], comparing [Fe^III₄(H₂O)₂(B-α-PW₉O₃₄)₂]⁶⁻ (Fe₄-WS) to its cobalt analog [Co^II₄(H₂O)₂(B-α-PW₉O₃₄)₂]¹⁰⁻ (Co₄-WS). Still, it was found that Fe^III derivatives display lower OER activity than Co^II-POMs. These results are in agreement with the theoretical considerations, as the lower Tafel slope exhibited by Co₄-WS pointed to faster OER kinetics than in the case of Fe₄-WS. This owes to the lower activation barrier in the case of Co₄-WS. These activation barriers were determined using the computation of the transition state of one water molecule forming a hydrogen bond with a bridging oxygen [36].

Han et al. synthesized sub-nanometric heterometallic CoW and FeCoW clusters by a molecule-to-cluster approach starting from several different POMs (i.e., [{Co₄(OH)₃PO₄}₄(SiW₉O₃₄)₄]³²⁻, [{Fe₂Co₂(OH)₃PO₄}₄(SiW₉O₃₄)₄]²⁴⁻, and [{FeCo₃(OH)₃PO₄}₄(SiW₉O₃₄)₄]²⁸⁻) as precursors [37]. The amount of Fe in the FeCoW clusters could be controlled by using POM precursors comprising a different number of Fe atoms. Outstanding OER activity with overpotential to reach a current density of 10 mA cm⁻², η₁₀, as low as 192 mV, and a Tafel slope as low as 36 mV dec⁻¹ was recorded for the most promising material.

A direct comparison of the performance of different transition metal-POM-based electrocatalysts is not simple/straightforward due to the different experimental conditions used in various studies reported in the literature,

Table 1. OER parameters of different POMs and their composites reported in the literature.

Catalyst	Electrolyte	η10 (mV)	Tafel Slope (mV dec−1)	Source
MWCNT_N8_Co4	0.1 M KOH	400	55	[24]
GF_N8_Co4	0.1 M KOH	340	67	[24]
GF_ND8_Co4	0.1 M KOH	490	68	[24]
GF_NS8_Co4	0.1 M KOH	460	62	[24]
Mono(aqua)nickel(II)	0.1 M phosphate buffer	-	168	[30]
Co6.8Ni1.2W12O42(OH)4(H2O)8/Ni foam	0.1 M KOH	360	126	[32]
(C5H7N2)6[NiW12O44]	1 M KOH	347	-	[32]
Ru4POM	80 mM sodium phosphate	-	120	[39]
Co4POM	0.1 M KNO3	-	80	[39]
NiP4Mo6	0.1 M KOH	-	73	[40]
Fe2Ni2@MWCNT_N6	0.1 M KOH	580	45	[41]
Fe4@MWCNT_N6	0.1 M KOH	460	102	[41]
Ni4@MWCNT_N6	0.1 M KOH	360	54	[41]
ZIF-8@ZIF-67@POM	1 M KOH	490	88	[42]
Ba[Fe4-WS]/CP	50 mM potassium phosphate buffer with 1 M KNO3	-	99	[36]
Ba[Co4-WS]/CP	50 mM potassium phosphate buffer with 1 M KNO3	-	73	[36]
PBA@POM	1 M KOH	440	235	[43]
1-CoW	1 M KOH	240	53	[37]
2-CoFeW	1 M KOH	224	43	[37]
3-CoFeW	1 M KOH	205	38	[37]

Co₄—[Co₄(H₂O)₂(PW₉O₃₄)₂]; Mono(aqua)nickel(II)—[Ni^II(2,2−bpy)₃]₃[Ni^II(2,2−bpy)₂(H₂O)HCo^IIW^VI₁₂O₄₀]₂·H₂O; Ru₄POM—[Ru₄O₄(OH)₂(H₂O)₄(γ−SiW₁₀O₃₆)₂]¹⁰⁻; Co₄POM—[Co₄(H₂O)₂(PW₉O₃₄)₂]¹⁰⁻; NiP₄Mo_6—Ni₃[Mo₆O₁₂(OH)₃(HPO₄)₃(PO₄)]₂·4bpe·10H₂O, bpe = 4,4-vinylene-dipyridine; Fe₂Ni₂—Na₁₄[(FeOH₂)₂N_i2(As₂W₁₅O₅₆)₂]·55H₂O; Fe₄—(Na₁₂[(FeOH₂)₂Fe₂(As₂W₁₅O₅₆)₂]·54H₂O; Ni₄—Na₁₂[(NiOH₂)₂Ni₂(As₂W₁₅O₅₆)₂]·54H₂O; ZIF-8@ZIF-67@POM—zeolitic imidazolate frameworks@H₃[PW₁₂O₄₀]·nH₂O; Ba[Fe₄-WS]—[Fe^III₄(H₂O)₂(B-α-PW₉O₃₄)₂]⁶⁻ and Ba[Co₄-WS]/CP—[Co^II₄(H₂O)₂(B-α-PW₉O₃₄)₂]¹⁰⁻, both with carbon paste (CP); POM in PBA@POM—H₃PMo₁₂O₄₀.

. For instance, theoretical considerations of various possible transition metal and heteroatom combinations for Weakley POMs led to the conclusion that the activity of metal sites for OER decreases in the order Ru > Mn~Co > Fe > Ni for this particular type of POMs [38]. Moreover, Cr Weakley POM with Al as the heteroatom is expected to show an overpotential as low as 310 mV, i.e., lower than that of a commercial IrO₂ electrocatalyst [38]. Still, when it comes to transition metal-based electrocatalysts, the use of a combination of Ni and Mo, Fe, or Co was demonstrated to lead to the highest number of active sites and thus the highest electrocatalytic performance for water electrolysis and other electrochemical energy conversion reactions [28].

3. POMs as Electrocatalysts for the Hydrogen Evolution Reaction (HER)

Although the OER is deliberated as an efficiency-limiting process in electrolytic water splitting, the hydrogen evolution reaction (HER), which involves several steps (Figure 4), also poses several challenges that must be tackled. As stated before, hydrogen production is a possible solution to the world’s current energy problems, but existing platinum-based electrocatalysis solutions need to be replaced by lower-cost alternatives. POMs are potential candidates as electrocatalysts for green hydrogen production. They have been shown to have a close enough performance to commercial carbon-supported platinum (Pt/C), the most used electrocatalyst for acidic and alkaline media, while being considerably cheaper.

Singh et al. [44] studied the copper-based POM [{Cu^II(2,2′-bpy)(H2O)₂}] [{Co^IIW^VI₁₂O₄₀}{Cu^II(2,2′-bpy)(H2O)}{Cu^II(2,2′-bpy)}]·2H₂O, close to neutral pH. In this, POM {Cu^II(2,2′- bpy)(H2O)₂}²⁺ acts as the active center, catalyzing the HER. The overpotential to reach 1 mA cm⁻² was determined to be 520 mV and TOF to be 6329 mol of H₂ evolved per mole of Cu atom per hour at the current density of 1 mA cm⁻². This was a huge increase when compared with previously reported Cu clusters, whose TOF was 457 mol of H₂ per mol of catalyst per hour at 817 mV [45]. These values are also more favorable than those of other complexes reported in the literature; namely, Ni- [46] and Co-complexes [47] were found to have TOF values of only 970.45 and 871.17 mol of H₂ mol catalyst⁻¹ h⁻¹, respectively, both measured at an overpotential of 837 mV.

Several Dexter–Silverton POM/Ni foam composites (NiM-POM/Ni; M=Co, Zn, Mn) were shown to be efficient electrocatalysts for HER in alkaline media (1 M KOH). NiCo-POM/Ni exhibited the highest activity towards HER in terms of the lowest overpotential (64 mV) to reach 10 mA cm⁻² accompanied by a low Tafel slope of 75 mV dec⁻¹ and faradaic efficiency as high as ca. 97%. Moreover, NiCo-POM/Ni was observed to be both mechanically and chemically stable [48].

Ni-based polyoxovanadate K₂[Ni(H₂O)₆]₂[V₁₀O₂₈]·4H₂O was found to exhibit homogeneous electrocatalytic activity for HER in acidic media with a low overpotential of 127 mV to reach half of the maximum current and TOF of 2.1 s⁻¹ at pH 2.3 [49]. It is worth noting that [Ni(H₂O)₆]²⁺ on its own does not exhibit any activity for HER, only when present in a POM matrix. Additionally, high stability, i.e., no decomposition, of K₂[Ni(H₂O)₆]₂[V₁₀O₂₈]·4H₂O under HER polarization conditions was detected.

In 2020, Wang et al. [50] prepared and investigated the electrocatalytic properties of another three copper-based POMs, {Cu₂(3-bptzp)₃(H₂O)₄[SiW₁₂O₄₀]}·H₂O, {Cu₂(3-bptzpe)₂(H₂O)₈[SiW₁₂O₄₀]}·4H₂O, and {Cu₂(3-bptzh)₃(H₂O)₆[SiW₁₂O₄₀]}, where (3-bptzp = 1,4-bis (5-(3-pyridyl)- tetrazolyl)- butane; 3-bptzpe = 1,4-bis(5-(3-pyridyl)tetrazolyl)-pentane, and 3-bptzh = 1,4-bis(5-(3-pyridyl)tetrazolyl)-hexane). All these POM-based metal-organic complexes exhibited good electrocatalytic activity under alkaline and acidic media. In particular, {Cu₂(3-bptzp)₃(H₂O)₄[SiW₁₂O₄₀]}·H₂O achieved a current density of 10 mA cm⁻² at an overpotential of 59.4 mV vs. RHE, leading to a Tafel slope of 74.2 mV dec⁻¹ in a 0.1 M KOH solution. It was further determined that even though the materials were stable under the full pH range, acidic media led to higher overpotentials and Tafel slopes.

Fernandes et al. [51] studied three further nanocomposites based on POMs on reduced graphene oxide (rGO) for HER in acidic media, and they were all shown to have overpotentials to reach 10 mA cm⁻²—very close to the state-of-the-art commercial Pt/C. P2W18@rGF_ox, P5W30@rGF_ox, and P8W48@rGF_ox were prepared by an easy and cost-effective one-step electroreduction of POM@rGF_ox. P5W30@rGF_ox showed the best activity with a Tafel slope of 33 mV dec⁻¹, but even the worst one, P8W48@rGF_ox, had a Tafel slope as low as 41 mV dec⁻¹, both relatively close to that of commercial Pt/C, 30 mV dec⁻¹.

A dual-atom catalyst comprising O-bridged W-Mo atoms onto N-doped graphene vacancies (with W-O-Mo-O-C configuration) was developed by controlled self-assembly of POMs and subsequent nitridation [52]. The performance of the prepared W-Mo heterodimer electrocatalyst overpasses the performance of homodimer Mo-Mo and W-W and it is comparable to that of Pt/C, along with remarkable stability under HER polarization conditions in a wide pH range.

Novel heterostructure comprising zinc iron oxide (ZnFe₂O₄) and POM nanoplates (POM–ZnFe₂O₄) synthesized via a hydrothermal method exhibited excellent performance for both HER and OER [53]. Thus, the current densities of 20 and 50 mA cm⁻² were reached at low overpotentials of 268 and 356 mV, respectively, under HER polarization conditions, and at 220 and 290 mV, respectively, under OER polarization conditions. Moreover, an electrolytic cell with POM–ZnFe₂O₄ serving both as the cathode and as anode delivered a current density of 10 mA cm⁻² at an operating voltage of 1.53 V. The electrochemical activity of POM–ZnFe₂O₄ overpassed that of commercial Pt/C and RuO₂ catalysts for HER and OER, respectively, and some catalysts recently reported in the literature. The acceleration of HER and OER by POM–ZnFe₂O₄ most likely originates in the modulation of its electronic and chemical characteristics along with the creation of heterojunction of ZnFe₂O₄ and POM, as well as its high electrochemically active surface area (ECSA).

All these studies on POMs for HER show excellent potential for this class of compounds. They also encourage further work with other materials to be combined with the POMs, creating highly conductive nanocomposites. These mainly include nanostructured carbon materials, such as the already mentioned graphene and reduced graphene oxide, carbon nanotubes, as well as conductive polymers.

4. POMs as Electrocatalysts for the Oxygen Reduction Reaction (ORR)

The presence of transition metal ions in the POM framework enables fine-tuning of their redox potentials. Incorporating a POM-based catalyst can increase the rate of the electroreduction process and lower the overpotential. The POM is said to “act as a powerful electron reservoir” and can usually provide electrons to other species [53].

The oxygen reduction reaction (ORR) is an integral process in fuel cells and metal-air batteries. Its catalysis is crucial because it presents sluggish kinetics, requiring high overpotentials to start. Generally, catalysts for ORR applications are selected to preferentially have a four-electron transfer mechanism, where O₂ is reduced to water (Equations (1) and (4) for alkaline and acidic media, respectively), in detriment to the two-electron pathway, which produces hydrogen peroxide (H₂O₂) (Equations (2), (3), (5), and (6) for alkaline and acidic media, respectively) [54]. Still, according to Zheng et al. [40], the two-electron transfer pathway can sometimes be desirable, as it is a safer and cheaper way to produce H₂O₂ than current technology, and POMs could also serve to catalyze this process.

Alkaline electrolyte

(a)

Direct 4-electron pathway

$${\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O}+4{\mathrm{e}}^{-}\to 4{\mathrm{OH}}^{-} { \mathrm{E}}^{°}=0.401 \mathrm{V}$$

(1)

(b)

Peroxide (2-electron) pathway

$${\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}+2{\mathrm{e}}^{-}\to {\mathrm{HO}}_2^{-}+{\mathrm{OH}}^{-}{ \mathrm{E}}^{°}=+0.065 \mathrm{V}$$

(2)

$${\mathrm{HO}}_2^{-}+{\mathrm{H}}_2\mathrm{O}+2{\mathrm{e}}^{-}\to 3{\mathrm{OH}}^{-}{ \mathrm{E}}^{°}=+0.867 \mathrm{V}$$

(3)

Acidic electrolyte

(a)

Direct 4-electron pathway

$${\mathrm{O}}_2+4{\mathrm{H}}^{+}+4{\mathrm{e}}^{-}\to 2{\mathrm{H}}_2{\mathrm{O} \mathrm{E}}^{°}=+1.229 \mathrm{V}$$

(4)

(b)

Peroxide (2-electron) pathway

$${\mathrm{O}}_2+2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to {\mathrm{H}}_2{\mathrm{O}}_2{ \mathrm{E}}^{°}=+0.67 \mathrm{V}$$

(5)

$${\mathrm{H}}_2{\mathrm{O}}_2+2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to 2{\mathrm{H}}_2{\mathrm{O} \mathrm{E}}^{°}=+1.77 \mathrm{V}$$

(6)

Another critical factor for the optimal running of rechargeable metal-air batteries or unitized regenerative fuel cells is bifunctionality—the ideal electrocatalysts should be active for both OER and ORR. For fuel cell applications, it is also important to find catalysts selective for ORR, i.e., “resistant” to fuel (for instance, methanol) crossover.

Zhang et al. [55] reported the efficiency of various POM-based materials for ORR, involving different supports for Ni and Co POM core. The supports were thermalized triazine-based frameworks (TTFs), fluorine-doped TTF (TTF-F), and rGO. All compared favorably to commercial Pt/C, and particularly {(PW₉)₂Ni₇}/{Cu(ethylenediamine)₂}/TTF-F as the first example of POM-based noble metal-free ORR catalyst that possessed comparable ORR activity to Pt/C while having much better stability in neutral medium. At this pH, Cu₆Ni₇/TTF-F, Cu₆Ni₇/rGO, and Cu₆Ni₇/C were the first examples of POM-based electrocatalysts that could support the one-step reduction of oxygen to water.

Liu et al. recently reported a novel top-down POM-based single-atom catalyst design approach. Immobilization of POM-single site catalysts on high surface-area mesoporous carbon at high loading was demonstrated [56]. Keggin-polyoxomolybdate [PMo₁₂O₄₀]³⁻ was used as a precursor for anchoring POM-like molybdenum (VI)-oxo subnanometer clusters ([Mo-oxo]_n, n = 1–20) onto carbon material. Such prepared material exhibited outstanding electrocatalytic activity for ORR, comparable to or even superior to that of commercial Pt/C catalysts.

Marques et al. [41] recently characterized three Fe and Ni-based POMs on MWCNTs, (Na₁₂[(FeOH₂)₂Fe₂(As₂W₁₅O₅₆)₂]·54H₂O, Na₁₂[(NiOH₂)₂Ni₂(As₂W₁₅O₅₆)₂]·54H₂O, and Na₁₄[(FeOH₂)₂Ni₂(As₂W₁₅O₅₆)₂]·55H₂O), in alkaline media. They all showed good ORR performances, having onset potentials of ca. 0.80 V vs. RHE and diffusion-limited current densities between −3.19 and −3.66 mA cm⁻². Stability tests showed Fe₄@MWCNT_N6 and Fe₂Ni₂@MWCNT_N6 to be the most promising, maintaining 84 and 80% of the current, respectively, after 12 h. Lastly, it was concluded that the number of electrons transferred per molecule of O₂ was close to three, suggesting a mixed 2e–4e mechanism.

RuPOM demonstrated high bifunctional activity for OER/ORR when combined with Ketjenblack (KB) carbon [57]. Its activity for OER was superior to that of commercial Pt/C catalysts and somewhat inferior in the case of ORR. Nevertheless, when it comes to the specific activity, i.e., the current density per mass of a precious metal, RuPOM significantly overpasses Pt/C. Thus, RuPOM/KB (in 8:2 ratio) gave 25- and 11-times higher ORR specific current density than RuO₂/KB and Pt/C, respectively. A seawater battery employing RuPOM/KB as a bifunctional catalyst for both OER and ORR exhibited notably improved performance, as evidenced by a decreased potential difference between charging (OER) and discharging (ORR) from 1.30 to 0.76 V, and an increase of the output power.

Limani et al. [24] reported good ORR performance of MWCNT_N8_Co4, GF_N8_Co4, GF_ND8_Co4, and GF_NS8_Co4. These composites presented onset potentials in the range 0.83–0.85 V vs. RHE and a number of exchanged electrons of ca. 4, indicating that ORR proceeds as a direct four-electron reduction of oxygen to water. Moreover, these composites exhibited good tolerance to methanol crossover with ORR current retention of 88–90% and good stability with current retention of 73–82% after 20,000 s at 0.55 V vs. RHE.

Table 2. ORR parameters of different POMs and their composites reported in the literature.

Catalyst	Electrolyte	Eonset (V vs. RHE)	Tafel slope (mV dec−1)	Source
MWCNT_N8_Co4	0.1 M KOH	0.85	41	[24]
GF_N8_Co4	0.1 M KOH	0.83	50	[24]
GF_ND8_Co4	0.1 M KOH	0.85	90	[24]
GF_NS8_Co4	0.1 M KOH	0.84	40	[24]
NiP4Mo6	0.1 M KOH	-	106	[40]
S-NiP4Mo6	0.1 M KOH	-	98	[40]
Fe4@MWCNT_N6	0.1 M KOH	0.80	35	[41]
Ni4@MWCNT_N6	0.1 M KOH	0.80	35	[41]
Fe2Ni2@MWCNT_N6	0.1 M KOH	0.81	38	[41]
Pd8Ni2/rGO-POM	0.1 M KOH	0.90	60	[58]
Co4(PW9)2@SWCNT	0.1 M KOH	0.77	68	[59]
Co4(PW9)2@GF	0.1 M KOH	0.89	71	[59]
Co4(PW9)2@N-CNT	0.1 M KOH	0.90	92	[59]
Co4(PW9)2@N-FLG	0.1 M KOH	0.89	63	[59]
Co4(PW9)2@Pt/C	0.1 M KOH	-	89	[59]
Mo-based species #1	0.1 M KOH	0.75	109	[56]
Mo-based species #2	0.1 M KOH	0.88	68	[56]
Mo-based species #3	0.1 M KOH	0.93	44	[56]

Co₄—[Co₄(H₂O)₂(PW₉O₃₄)₂]; NiP₄Mo₆—Ni₃[Mo₆O₁₂(OH)₃(HPO₄)₃(PO₄)]₂·4bpe·10H₂O, bpe = 4,4-vinylene-dipyridine; Fe₂Ni₂—Na₁₄[(FeOH₂)₂N_i2(As₂W₁₅O₅₆)₂]·55H₂O; Fe₄—(Na₁₂[(FeOH₂)₂Fe₂(As₂W₁₅O₅₆)₂]·54H₂O; Ni₄—Na₁₂[(NiOH₂)₂Ni₂(As₂W₁₅O₅₆)₂]·54H₂O; Pd₈Ni₂—H₃[PMo₄W₈O₄₀]; Co₄(PW₉)₂@X—[(Co₄(PW₉)₂)₇H₃[Co₄(H₂O)₂(PW₉O₃₄)₂], X—single-walled carbon nanotubes (SWCNT), graphene flakes (GF), carbon nanotubes doped with nitrogen (N-CNT), or nitrogen-doped few layer graphene (N-FLG).

summarizes ORR data reported in the literature for different POMs and POM-based materials.

5. POMs for Batteries and Supercapacitors

A particularly difficult challenge for energy storage today is the development of rechargeable, high-performance, low-cost, and environmentally friendly batteries. Namely, new electrode materials that can lead to higher voltages and higher capacities need to be developed, and POMs have contributed to this development.

5.1. Lithium-Ion Batteries

Lithium-ion batteries (LIBs) (Figure 5) represent the currently most used energy storage devices due to their high voltage (>3.5 V), high energy density (<200 W h kg⁻¹), and reasonably long cycle life. Typically, lithium transition metal oxides including LiCoO₂, LiMn₂O₄, and LiFePO₄, are employed as cathodes in LIBs. Still, these materials suffer from a decrease of capacity during cycling as well as scarcity and high price, specifically cobalt-based ones. Graphite is the most frequently employed anode in commercial LIBs. Still, graphite’s low capacity limits its viability in large-scale systems. Bearing in mind the limitations of current cathode and anode materials used in LIBs, great attention has been devoted to the development of novel materials for this type of battery. POMs have been studied for use in the field of LIBs with their electrochemical performance being enhanced by the usage of nanosized particles and/or mixing them with highly electrically conductive materials. Additionally, POMs molecular clusters show multielectron redox properties as individual molecules, making their cycling stability and capacity independent of their crystalline stability. For instance, using polyoxomolybdate (NH₄)₆[NiMo₉O₃₂] nanoparticles with KB carbon as cathode material led to an initial capacity discharge of 490 mA h g⁻¹ with a capacity retention as high as 87.6% after 50 cycles at 17 mA g⁻¹ [60]. Employing Mn-Anderson-type POM on single-walled carbon nanotubes as the anode material led to an initial capacity discharge as high as 3405.4 mA h g⁻¹, although with low-capacity retention (ca 27.4% after 100 cycles at 0.5 mA cm⁻²) [61].

5.2. Sodium-Ion Batteries

Sodium-ion batteries (SIBs) are a promising alternative to LIBs for large-scale energy storage applications, bringing the advantage of a high abundance of sodium and thus lower cost. The possibility of employing aluminum as a low-cost current collector for both anode and cathode further lowers the price of this type of battery compared to LIBs. The use of aluminum is possible since sodium and aluminum do not form an alloy, unlike lithium and aluminum. However, this low cost comes with a problem: SIBs energy density is lower than LIBs as a consequence of the large radius of the Na⁺ ions, large atomic weight compared to lithium, and low operating potential [63]. A possibility to overcome such problems is the use of new cathode materials. Therefore, this area of study has garnered tremendous attention lately. Liu et al. [64] and Chen et al. [65] studied POM-based materials for cathodes, while Hartung et al. [66] and Lin et al. [67] focused on anode materials.

Na₂H₈[MnV₁₃O₃₈] (NMV) [64] as SIB cathode demonstrated a high specific capacity of 190 mA h g⁻¹ (at 0.1 C) with high retention of, for instance, 81% at 0.2 C. The material’s high performance was attributed to its “electron/Na⁺-ion sponge” behavior with its structure providing numerous diffusion channels for Na⁺ ions. The main contributor to its high capacity is the acceptance of 11 electrons/Na⁺ per mole. The good rate capability is enabled by the material’s flexible 2D lamellar network with 1D open channels providing numerous migration pathways for Na⁺ ions. Finally, the good cycling stability is due to the formation of [MnV₁₃O₃₈]₂₀ with size expansion upon insertion of Na⁺ ions of ca. 7.5% compared to the parental [MnV₁₃O₃₈]₉⁻.

Nanosized polyoxovanadate Li₇V₁₅O₃₆(CO₃) [65] envisaged to enable the transport of large cations, including Na⁺, demonstrated a high specific capacity of 240 mA h g⁻¹ (at 50 mA g⁻¹) and led to a specific energy of 390 W h kg⁻¹ when employed as the cathode of a rechargeable SIB.

Na₆[V₁₀O₂₈].16H₂O [68] and Na₇[H₂PV₁₄O₄₂] [67] as the SIB anode attained a capacity of ca. 276 (at 20 mA g⁻¹) and 322 mA h g⁻¹ (at 25 mA g⁻¹), respectively, along with good cycling stability in case of both studied materials. For instance, capacity retention in the case of Na₇[H₂PV₁₄O₄₂] was as high as 87% upon 120 cycles.

Still, it was generally concluded that storage capacities in SIBs were much lower than in LIBs, as shown in

Table 3. Performance of different types of batteries involving POMs and their composites reported in the literature.

Material	Batteries	Electrode	Initial Discharge Capacity (mA h g−1)	Current/Current Density/Rate	Capacity Retention (%)	Cycle Number	Source
ANM/KB	LIBs	cathode	490	17 mA g−1	87.6	50	[60]
Mn-Anderson/ SWNT	LIBs	anode	3405.4	0.5 mA cm−2	27.4	100	[61]
Ni-POM	LIBs	anode	1325	100 mA g−1	54	50	[69]
HP-NENU-5/CC	LIBs	anode	~2346	200 mA g−1	73	100	[70]
HP-HKUST-1/CC	LIBs	anode	1259	200 mA g−1	74.5	100	[70]
HP-PMo12/CC	LIBs	anode	1147	200 mA g−1	88	100	[70]
POMOFs-based Pmo10V2/rGO	LIBs	anode	2367.8	50 mA g−1	45	100	[71]
PMO12/PANI/MWNTs	LIBs	anode	1572	0.5 mA cm−2	63.6	100	[72]
Composite 1/SWNTs	LIBs	anode	1469	100 mA g−1	67.3	100	[73]
Li7[V15O36(CO3)]	LIBs	cathode	170	2 A g−1	84	100	[74]
Pmo12	SIBs	anode	15.5	1.0 A g−1	5.2	100	[75]
Pmo12/MIL-88B/GO	SIBs	anode	430.2	1.0 A g−1	67	100	[75]
Pmo12/MIL-53/GO	SIBs	anode	152.4	1.0 A g−1	40	100	[75]
Pmo12/MIL-101/GO	SIBs	anode	44.5	1.0 A g−1	18.6	100	[75]
Pmo12/GO	SIBs	anode	103.4	1.0 A g−1	-	100	[75]
Li7V15O36(CO3)	SIBs	cathode	171	0.1 A g−1	85	100	[65]
Na6[V10O28]	SIBs	anode	300	20 mA g−1	96	10	[66]
Na7[H2PV14O42]	SIBs	anode	397	25 mA g−1	71	120	[67]
Li5BW12O40-Lil	RFBs	-	24.8	100 mA g−1	86.6	100	[76]
rGO-POMs HENFs	RFBs	-	-	4.0 A g−1	97	250	[77]
all-H6[CoW12O40]	RFBs	-	13.6	25 mA cm−2	99	30	[78]

. This could be explained by low kinetics resulting from a Na₂O passivating film layer formed during the first charge/discharge cycles.

5.3. Redox Flow Batteries

Redox flow batteries (RFBs) might be considered ideal for stationary applications. This type of battery uses two electrolytes containing different electroactive species, stored outside the cell and continually pumped in, as shown in Figure 5 [68]. Two electrolyte solutions are separated by a membrane preventing their direct mixing but allowing the passage of ions. Such a unique construction design and operational mechanisms allow a RFB to be sized to meet the requirements of various energy and power demands [79].

POMs have been studied in RFBs as a strategy for enhancing their electrochemical performance [62,80,81]. Pratt et al. [82] developed three phosphorus-based POMs (A-a-PV₃W₉O₄₀⁶⁻, B-a-PV₃W₉O₄₀⁶⁻, and P₂V₃W₁₅O₆₂⁹⁻) for use as the electroactive components of RFBs electrolytes. The electrochemical measurements showed good pairs to act as positive and negative electroactive species, such as (P^VIV₃W^VI₉O₄₀⁹⁻/P^VV₃W^VI₉O₄₀⁶⁻) and (P^VIV₃W^VI₉O₄₀⁹⁻/P^VIV₃W^V₃W^VI₆O₄₀¹²⁻). However, the studied POMs still present low coulombic efficiency and poor cycling stability, a challenge that needs to be overcome if POMs are to be used in RFBs.

As seen in many other applications, the hybridization of POMs with other materials has been proposed to increase their electrochemical activity. Dubal et al. [77] proposed the hybridization of graphene and POMs to generate electroactive nanofluids for application in rGO/phosphomolybdate-based RFBs. These devices showed encouraging electrochemical features, including a specific capacitance of 305 F g⁻¹ as well as durable coulombic efficiency (~77–79%) after 2000 cycles. Even though these results still pale in comparison to their vanadium counterparts, more effort and new approaches may lead to breakthroughs in using POMs in electroactive nanofluids of RFBs.

5.4. Supercapacitors

Supercapacitors (SCs) are praised for their high energy storage capacity and ability to deliver it at a higher power output compared to batteries. Their other advantages include durable cycle life, long-term stability, the capability of delivering a high current, and remarkable energy efficiency. Thus, SCs are perfect for applications requiring a primarily large number of quick charge/discharge cycles instead of long-lasting compact energy storage; for instance, they are used for burst-mode power delivery in different types of vehicles, elevators, and forklifts. Still, they suffer from a rather low cell voltage and their energy density could be further increased. POMs have been suggested as electrode materials suitable for pseudo-capacitors. Namely, supercapacitors can be classified in three groups: electric double layer (EDL) capacitors, pseudo-capacitors, and hybrid capacitors, having different charge storage mechanisms, as shown in Figure 6.

Electric double-layer capacitors store charge by the reversible electrostatic accumulation of ions at the electrode/electrolyte boundary. Typically, high-surface area, porous carbon materials are employed in this category of SCs with graphene being considered the benchmark material [83]. The search for the optimal materials for EDL capacitors has led to achieving EDL capacitances in the 200–300 F g⁻¹ range. [84] Extrinsic pseudocapacitive behavior is another favorable approach for increasing energy density. In this case, electroactive species react with surface functional moieties in a rapid reversible redox reaction [85]. Electrode materials used in pseudocapacitors include transition metal (such as Mn or Ru) oxides, heteroatom-doped carbon materials, and conducting polymers. POMs, with a high surface-to-bulk ratio and metal species with a number of different redox states, seem to be great contenders for pseudocapacitors.

These POMs properties have been known and studied for several decades. As early as 1997, Gómez-Romero and Lira-Cantú [86] combined POMs with different conducting polymers for SC applications. Over the years, organic-inorganic hybrids comprising polypyrrole (PPy) [87], poly(3,4-ethylenedioxythiophene) (PEDOT), or polyaniline (PANI) [88], and POMs have been reported.

Vanadate POMs and activated carbon were combined as the positive and negative electrodes, respectively, in an asymmetric cell [89]. Such a cell reached a capacitance of 354 F g⁻¹, as well as a specific energy of 73 W h kg⁻¹ at 0.1 A g⁻¹ and a specific power of 312 W kg⁻¹.

Gupta et al. [90] also delved into the research of POMs for use in supercapacitors, more specifically, into hybrids of phosphomolybdate acid-H₃PMo₁₂O₄₀ and phosphotungstic acid-H₃PW₁₂O₄₀ with rGO. The effect of this hybridization on the electrochemical properties was very positive and the specific capacitance increased from 70 F g⁻¹ measured for rGO to 350 F g⁻¹ for the hybrid material. A higher current carrying capacity was also achieved with good retention (94 %), and consequently higher specific energy and specific power density, as shown in Figure 7.

6. Conclusions

This review recapitulates the main properties of POMs that make them suitable for electrochemical energy conversion and storage applications. Numerous examples of their successful employment in the electrocatalysis of reactions operating water electrolyzers, fuel cells, different types of batteries, and supercapacitors are presented.

POMs exhibit distinctive redox characteristics and the capability of reversibly uptaking as much as 24 electrons per cluster unit. Their redox potential and the number of electrons stored can be tailored additionally by, for instance, the incorporation of redox-active metal centers through a chemical process. One of the main drawbacks of POMs as electrocatalysts is their low electrical conductivity; still, this problem can be overcome by designing composite materials comprising POMs and highly conductive materials such as carbon-based ones.

Consequently, POMs and their composites have been investigated as electrocatalysts for ORR, OER, and HER, sometimes reaching performance comparable to that of commercial Pt/C benchmark electrocatalyst.

POMs and their composites have also been explored as both cathode and anode materials for LIBs and SIBs with LIBs exhibiting especially promising performance. The lower performance of SIBs employing POMs was considered to be a consequence of the formation of a Na₂O passivating film layer during the first charge/discharge cycles. Still, promising performance can be achieved by tailoring the redox properties and structural design of POMs, resulting in the acceptability of a high number of electrons as well as various pathways of Na⁺ ion migration and their easy intercalation. An additional advantage of POMs as electrocatalysts is their recyclability due to their ease of recovery from used electrodes of batteries by simple water treatment. Studies involving RFBs generally showed limited cycling stability and excessive hydrogen production. Nevertheless, several redox-active species have been tested for RFBs throughout the years, with different levels of success.

The potential of POMs as electrode materials for supercapacitors has been recognized early; they indeed demonstrated a high contribution to specific energy and specific power density. However, some challenges regarding POMs as electrode materials for supercapacitors have been encountered, including secure grafting of POMs onto a substrate/host material and attaining a durable cycle life. These challenges are yet to be overcome.