Molecular Vanadium Oxides for Energy Conversion and Energy Storage: Current Trends and Emerging Opportunities

Abstract Molecular vanadium oxides, or polyoxovanadates (POVs), have recently emerged as a new class of molecular energy conversion/storage materials, which combine diverse, chemically tunable redox behavior and reversible multielectron storage capabilities. This Review explores current challenges, major breakthroughs, and future opportunities in the use of POVs for energy conversion and storage. The reactivity, advantages, and limitations of POVs are explored, with a focus on their use in lithium and post‐lithium‐ion batteries, redox‐flow batteries, and light‐driven energy conversion. Finally, emerging themes and new research directions are critically assessed to provide inspiration for how this promising materials class can advance research in sustainable energy technologies.


Introduction
Our way of harvesting and storing energy is beginning to change on ag lobal scale.T he transition from traditional fossil-fuel-based systems to carbon-neutral and more sustainable schemes is underway. [1] With this transition comes the need for new directions in energy materials research to access advanced compounds for energy conversion, transfer,a nd storage.I na ddition, long-term stability,e conomic viability, and sustainability will become central design criteria. [2,3] In particular,r edox-active materials are essential for electrochemical or photochemical energy conversion and storage, including solar fuels,p hotovoltaics,e lectrocatalysis,a nd batteries.M etal oxides have become established as ac hemically suitable and economically viable materials class for energy technologies such as batteries, [4] water electrolysis, [5] fuel cells, [6] and small-molecule (e.g. H 2 O, CO 2 ,N 2 )a ctivation. [7] Over the last decade,m olecular metal oxides (polyoxometalates,P OMs) have become af ocal point in energy materials research. POMs combine the fundamental chemical properties of solid-state metal oxides with the structural precision and chemical tunability inherent to molecular systems. [8][9][10] As such, they have shown outstanding performance in challenging energy-related applications,i ncluding water oxidation, [11,12] hydrogen evolution, [13,14] and batteries. [15,16] Traditionally,most POM research was focused on Wand Mo-based systems,t hat is,p olyoxotungstates and polyoxomolybdates,w ith af ocus on derivatives of the Keggin [XM 12 O 40 ] nÀ and Wells-Dawson [X 2 M 18 O 62 ] mÀ anions (X = P, Si, etc. and M = Mo,W). [17][18][19][20] Although tremendous progress was made on tungstate and molybdate chemistry in the 1990s and early 2000s, molecular vanadium oxides,o rp olyoxovanadates (POVs; Figure 1), [21][22][23] have only received widespread attention in the area of energy research over the last decade.P OVsf eature multiple accessible oxidation states (mainly based on the V V / V IV and more rarely V IV /V III redox couples), resulting in rich redox behavior. [24] In addition, they have asignificantly lower atomic weight than other POM families (V: 50.94 gmol À1 ; Mo:9 5.95 gmol À1 ;W :1 83.84 gmol À1 ), which allows higher gravimetric energy densities and makes POVs relevant for battery applications. [25] Furthermore,vanadium is obtained as ab y-product of steel manufacturing and is,t herefore, available on an industrial scale. [26] In addition, POVs are Molecular vanadium oxides,o rpolyoxovanadates (POVs), have recently emerged as an ew class of molecular energy conversion/ storage materials,w hichcombine diverse,c hemically tunable redox behavior and reversible multielectron storage capabilities.This Review explores current challenges,major breakthroughs,a nd future opportunities in the use of POVs for energy conversion and storage.T he reactivity,a dvantages,a nd limitations of POVs are explored, with afocus on their use in lithium and post-lithium-ion batteries,redoxflowb atteries,a nd light-driven energy conversion. Finally,e merging themes and new researchd irections are critically assessed to provide inspiration for how this promising materials class can advance researchins ustainable energy technologies.  [21,23,27] Thef unctionalization of molybdates and tungstates with redox-active heterometals has been widely used to tune their redox behavior and resulting reactivity for applications such as,w ater oxidation, [11,12] hydrogen evolution, [16] photooxidation chemistry, [28] and battery applications. [29] However,u ntil recently,t he functionalization of POVs with metals was mainly used for structural stabilization, [27,[30][31][32] precursor synthesis, [33,34] or to access new cluster topologies. [21,23,27] Over recent years,h owever, POVc hemistry has moved from exploring structures to functions,w hich has enabled ground-breaking research in many areas of energy conversion and storage.I nt his Review,w ew ill showcase the recent developments in the redox tuning of POVs.W ehighlight the unique advantages of POVs that can be exploited in energy research and identify areas which-from the authors point of view-offer new possibilities for fundamental and application-driven research over the coming years.

Fundamentals of Polyoxovanadate Chemistry
Theprincipal reactivity of POVs has been described in the literature. [21][22][23]27,35,36] Therefore,only abrief overview of their most important features is given here:POVsself-assemble in aqueous or organic solvents, [21,22] sometimes involving internal templates such as transition metals, [27,37] anions (halides, pseudohalides,o xoanions), [21,23,27] or, very rarely,n eutral organic compounds. [38] Thef unctionalization of POVs with metal cations, [23,27] semimetals, [36] or organic groups,f or example,alkoxides, [39] 6 ]c oordination modes. [21,27] However,t his coordinative versatility can also facilitate structural rearrangements,w hich needs to be considered when deploying POVs in energy applications. [27,[40][41][42][43] In addition, POVs are often described as being "less stable" than tungstates and molybdates,w hich is most likely due to ac ombination of their structural flexibility, complex protonation chemistry,and high redox activity. [21,23,27] These challenges also offer opportunities,particularly for energy materials research:k ey features which render POVs interesting for energy conversion and storage are their capability to reversibly store and transfer multiple electrons, based on the accessibility of the V V /V IV and-more rarelythe V IV /V III redox couples. [24] Despite its importance for batteries,s upercapacitors,a nd electrocatalysis,m ultielectron storage in POVs is still underexplored. [44,45] In addition, the related area of proton-coupled electron transfer in POVs requires urgent attention, [46] as it could facilitate complex multielectron transfers (e.g. water oxidation, hydrogen evolution, CO 2 reduction) to overcome activation barriers or tune redox potentials. [47] Another unique feature of POVs is their smaller HOMO-LUMO gap compared with those of Mo-or W-based POMs,s ot hat POVs can act as efficient visible-light-driven photoredox catalysts. [48] In the following sections,w ew ill discuss recent developments in the use of POVs in lithium-ion and post-lithium-ion batteries,r edoxflow batteries,a nd photochemistry.W ew ill describe unique features as well as current limitations of POVs and provide an outlook on future scenarios for POV-based energy research.

POVsi nBatteries
Rechargeable batteries are key electrochemical energy storage technologies where stored chemical energy is converted into electricity. [49] Currently,l ithium-ion batteries (LIBs) are amongst the most successful electrochemical energy storage technologies,a st hey offer high voltages and,  thus,h igh energy densities.H owever,t he currently used electrode materials appear not to be sustainable in the medium and long term because of the limited availability of several of the elements used (e.g.L i, Co). [50] Furthermore, LIBs typically rely on so-called intercalation compounds (often metal oxides), where the cycling stability is limited by the mechanical degradation of the structure during the lithium intercalation/extraction ( Figure 2). [51] To create more-stable battery electrodes,anew class of battery electrode materials is required which combines high energy density and stability with chemical tunability and economic viability.
POMs have inspired ground-breaking research on battery materials as they can bridge the gap between molecular designer materials and technologically important solid-state metal oxides.Whereas pioneering studies of POMs as battery components were focused on polyoxomolybdates, [16,52,53] recent studies have reported breakthroughs in the use of POVs as active components for battery electrodes.POVsare the lightest class of POMs available and they,therefore,offer higher gravimetric energy densities,which is important for emobility applications.I na ddition, POVs can, in principle, undergo multiple redox processes per Vc enter (POVs containing V V ,V IV ,a nd more rarely V III have been reported), [24] which would further increase their electronstorage capacity. [54] These benefits combined with the economic viability of vanadium has led to fast-paced progress in the exploration of POV-based batteries. [55] To date,m ost studies have been focused on the POV prototype decavanadate,which is typically synthesized under aqueous conditions and isolated as ah ydrated salt, that is, M 6 [V 10 O 28 ]·x H 2 O(M= Li or Na; x = 9-16). As aconsequence of this synthetic approach, the crystal lattice typically contains water, which needs to be thermally removed, as otherwise this would lead to "gassing" (O 2 and H 2 evolution) under typical LIB operation conditions. [56] Forthis reason, previous studies used thermally dehydrated Li 6 [V 10 O 28 ]a saL IB cathode material, which led to initial discharge capacities of about 130 mAh g À1 at ac urrent density of 0.2 mA cm À2 and potentials between 2.0 and 4.2 Vv s. Li + /Li. [57] Subsequent studies investigated the performance of the component after thermal treatment at higher temperatures (450 8 8C). Thes amples showed high initial specific capacities (ca. 400 mAh g À1 )b ut low cycling stability. [58] In contrast, samples treated at 600 8 8C showed lower initial specific capacities (ca. 200 mAh g À1 )but high cycling stability. [58] Subsequent studies used the thermally dehydrated decavanadate salt Na 6 [V 10 O 28 ]a st he cathode active material in LIBs.B yu sing in situ X-ray absorption near-edge spectroscopic (XANES) studies,t he authors showed that all ten V 5+ ions can be reversibly reduced to V 4+ in ap otential range of 4-1.75 Vv s. Li + /Li. [59] The decavanadate was also used as as tarting material for the fabrication of POV-based electrodes for sodium-ion batteries (NIBs), and Na 6 [V 10 O 28 ]w as reported as an anode material with ar eversible capacity of about 280 mAh g À1 ,a na verage discharge potential of 0.4 Vv s. Na + /Na, and high cycling stability. [55] In most of these studies,t he removal of lattice water involved thermal treatment at high temperatures. However,P OVsa re known to easily convert into solid-state vanadium oxides, [60,61] and the decavanadate cluster is particularly susceptible to undergo thermally induced structural rearrangements;t he dehydration of lithium decavanadate Li 6 [V 10 O 28 ]·16 H 2 Ol eads to the formation of two solid-state oxides,LiVO 3 and LiV 3 O 8 ,even at moderate temperatures of about 120 8 8C. [62] Consequently,most studies of decavanadates as molecular electrode components reported to date were in fact analyzing the performance of nano-or microstructured solid-state lithium vanadium oxides. [62] In addition to studies using the decavanadate prototype, other common POVs such as K 5 [67,68] have also been used as NIB and LIB cathode materials and showed promising performance.I n addition, Dong, Cronin, and co-workers designed symmetric LIBs that employed Li 7 [V 15 O 36 (CO 3 )] as the active anode and cathode material. Thes ystem combined battery-like energy density (125 Wh kg À1 )a nd supercapacitor-like power density (51.5 kW kg À1 at 100 Ag À1 )a nd could serve as am odel to bridge these two technologies. [68] However,a sd iscussed for the decavanadate,questions remain on the actual structure of the active materials and their chemical evolution under battery cycling conditions. [61] To explore the performance of truly molecular POV battery electrodes,A njass,S treb,a nd co-workers proposed the use of supramolecular crystal engineering for POV stabilization. [63] To this end, dimethylammonium cations were used for electrostatic and hydrogen-bonding stabilization of the decavanadate cluster in the crystal lattice.T his prevented thermal degradation of the POVi nto solid-state oxides and led to the retention of the molecular structure of decavanadate in the resulting LIB electrodes (Figure 3). Initial tests of these cathodes in LIBs showed specific capacities up to 290 mA hg À1 in av oltage range between 1.2  . [69] Thegroup incorporated the material in LIB cathodes and reported ad ischarge capacity of 156 mA hg À1 in av oltage range between 1.5 and 3.8 Vvs. Li + /Li. Further work by Liu and co-workers showed that the combination of Mg 2+ and NH 4 + cations in Mg 2 -(NH 4 ) 2 [V 10 O 28 ]also results in stable crystal lattices and gives access to LIB cathodes with discharge capacities of about 200 mA hg À1 in the potential range of 1.0 and 3.8 Vv s. Li + / Li. [70] These reports highlight the urgent need for fundamental studies on the reactivity of POVs,with afocus on thermally or chemically harsh conditions as used in many energy technologies.Based on this understanding,new concepts for stabilizing POVs (and POMs in general) could open new approaches for the design of materials for the field. In addition, fundamental questions also remain with respect to interfacial reactivity and the role of the nano-and microstructuring of POM-based electrodes, [16] particularly for applications where contact with electrolytes is essential. Pioneering works which address these questions will be discussed as part of the Outlook (Section 6).

POVsi nRedox-Flow Batteries
Redox-flow batteries (RFBs) charge and discharge (highly concentrated) solutions of redox-active species.T he charged solutions can be stored in external tanks,w hich can be varied in size to allow inexpensive and simple scaling of the battery capacity without altering the electrode area. This results in decoupled power and capacity,which is not possible with other battery technologies and is an advantage for stationary energy-storage applications. [3] RFBs based on vanadate salts dissolved in aqueous acid have been successfully commercialized (Figure 4). However,t heir volumetric energy density is limited to around 50 Wh L À1 by the solubility of the vanadate salts. [71] Forcomparison, Panasonic LIBs used in the 2013-2017 Te sla electric vehicles had energy densities of 670-683 Wh L À1 . [72] Intensive research is underway to develop next-generation RFB technologies by creating new redox-active materials to achieve higher energy densities and compete with LIB technology.T his would enable RFBs to store more energy and operate on longer timescales,which is crucial for switching to renewable power sources.I mproved specific and volumetric energy density would also allow RFBs to compete with LIBs in terms of applications in electric vehicles.
When designing new redox-active species for RFBs,t he two main chemical considerations are 1) how much charge can be stored per liter (volumetric capacity) and 2) how much energy can be stored per charge (the difference between the redox potentials of the redox couples in the separate electrolytes). Since RFBs are mainly aimed at (large-scale) stationary energy storage,l ow materials cost is an important third consideration. Molecular and macromolecular species with multielectron storage capability (such as POMs or redoxactive polymers) can fulfil all three design criteria. [73][74][75][76][77][78][79] In one ground-breaking study,aRFB based on ap olyoxotungstate anolyte and abromide/bromine-based catholyte was reported to reach 200 Wh L À1 , [74] which is the energy density of the LIBs used in the 2013 Honda Fitelectric vehicle. [72] Although most of the POM-RFB studies have thus far been focused on tungstate POMs,P OVso ffer some features which make them ideal charge-storage materials in RFBs. POVs can often be accessed as mixed-valence species,sothat they can be both oxidized and reduced. [23] This allows them to be used in symmetric RFBs,where the catholyte and anolyte use the same redox-active species. [73] POVs also have significantly lower molecular weights compared with molybdate and tungstate species.I nt he future,t his should allow    [76] POMs/POVs generally exhibit electrochemically reversible (fast) electron transfer, while electron transfer in conventional all-vanadium RFBs is typically slow.T his limits the achievable current density and power output of typical RFBs.Stimming and co-workers used the POM/POVs ystem to achieve ap ower output that was 50 mW cm À2 greater than an all-vanadium RFB. [76] O 40 ]anolyte was reduced by 2e lectrons per cluster. Therefore,t he clusters were used in a2 :1 molar ratio (tungstate/vanadate) to ensure that all the redox-active material could be charged. [80] This also illustrates the advantages of using two independent electrolytes in asymmetric RFBs,a st he anolyte and catholyte can be adjusted individually to achieve an optimized whole-cell performance.I n addition, Stimming and co-workers observed by 51 VN MR spectroscopy that, upon degradation, the [PV 14 O 42 ] 9À cluster reassembled under operational conditions.T his behavior could lead to significantly longer operational lifetimes for POM-based RFBs compared to other RFB systems. [76,81] The POM/POVsystem has also recently been successfully scaledup and operated in ap ilot-scale reactor. [81] These studies demonstrate the scalability and advantages of aqueous POM/ POVRFBs.
In their efforts to increase the energy density,t he RFB community has started to explore non-aqueous RFBs.Switching to organic electrolytes increases the solubility of many redox-active organic materials,a nd extends the functional voltage window of RFBs by preventing the electrolysis of water into oxygen and hydrogen. [82,83] In apioneering series of studies,M atson and co-workers developed vanadium-based Lindqvist clusters featuring alkoxy ligands for symmetric nonaqueous RFBs.T he highly organo-soluble species [V V 2 V IV 4 O 7 (OR) 12 ]( R= Me,B u, C 2 H 4 OCH 3 ,e tc.) exhibited solubilities of up to 1.2 m in acetonitrile (containing 0.1m nBu 4 NPF 6 ). [29,39,84,85] Them ixed oxidation states of the vanadium centers in [V V 2 V IV 4 O 7 (OR) 12 ]a llows the cluster to be reduced or oxidized by up to 2electrons.T his enabled the development of an RFB,where the same POVisused as the active species in the anolyte and catholyte.T his can be advantageous because it prevents long-term performance loss arising from active species crossover. Matson and co-workers went on to demonstrate that chemical modification of their Lindqvist cluster can be used to tune the electrochemical performance.F or example,t uning the redox potentials by incorporation of heterometals into the cluster framework is possible, [29] whereas modification of the alkoxide ligands allows the electron-transfer kinetics and the solubility (and therefore charge-storage capacity) of the electrolyte to be tuned. [29,[84][85][86]

POVsi nLight-Driven Catalysis
Thea bility of POVs to absorb in the visible-light region makes them well-suited for photoredox catalysis using sunlight. [87] Although traditional photoredox research on POMs was focused on UV-light-driven conversions of organic substrate by polyoxotungstates, [88] the last decade has seen significant progress in POV-mediated visible-light-driven photoredox processes.
Early studies in visible-light-induced POVphotooxidation chemistry explored the unique structural flexibility of the compound class to allow the in situ formation of visible-lightabsorbing species.T othis end, Streb and co-workers used the thermal conversion of the UV-absorbing cluster [V 4 O 12 ] 4À ({V 4 })t og ive the visible-light-absorbing [V 5 O 14 ] 3À ({V 5 }). [42] Irradiation of this species in organic solution with visible light in the presence of methanol resulted in the two-electron/two proton oxidation of the alcohol to formaldehyde.O ver the course of the reaction, the cluster underwent ar eductioninduced "dimerization" leading to the two-electron-reduced . Thea uthors were able to close the catalytic cycle by reoxidation of the decanuclear species using molecular oxygen (slow) or hydrogen peroxide (fast;F igure 5).
Subsequent studies showed that the photooxidative reactivity of POVs can be further tuned by the incorporation of high-valent metal cations,f or example,C e III [89] or Bi III . [90] Notably,t hese systems sometimes even resulted in ceriumor bismuth-functionalized POVs with helical chirality. [91,92] Although initial studies only explored the photooxidative activity of the racemic mixtures of the respective POVs, future studies could build on this and target enantioselective photooxidation using the enantiopure cluster species. [93] Building on these studies of metal-functionalized POV photooxidation catalysts,G uldi, Streb,a nd co-workers demonstrated that modification of the internal cluster template anion can also be used to modulate the photoreactivity.T he authors explored the model compounds [X(Bi-(dmso) 3 ) 2 V 12 O 33 ] À ,w hich only differ in their internal template,w here X is either chloride or bromide.C onsequently, the two clusters show almost identical light absorption. The authors examined the photoredox activity of both clusters ). [42] under identical conditions,u sing the photooxidative degradation of the model pollutant patent blue V( PBV) as at est reaction. This comparative study showed that the bromidetemplated system exhibited significantly faster PBV degradation kinetics compared with the chloride species.B yusing ultrafast photophysical analyses together with theoretical computations,itwas proposed that, upon photoexcitation, the heavier bromide template enables am ore efficient singlettriplet transition because of the enhanced spin-orbit coupling (the so-called heavy-atom effect). [91] This results in af aster reaction with the pollutant and the observed higher reaction kinetics.T he same trend was observed for the quantum efficiencies of the dye degradation, which were significantly higher for the bromide-templated species compared with the chloride-based cluster.M any POV( and POM in general) photooxidation studies currently use molecular oxygen as ar eoxidant to close the catalytic cycle.H owever,r ecent studies have shown that interfacial mass transport of O 2 from the gas to the liquid phase can severely limit the catalytic performance of POVs. [94] This finding emphasizes that detailed understanding is required across multiple scales to understand the optimization potential for these complex reactions,sothat photophysical processes on the femtosecond can be effectively coupled to mass transport on the second timescale,a nd molecular design on the sub-nanometer scale can be integrated with chemical reactor design on the micrometer scale and beyond.
Thef ields of light-driven POVc atalysis and POVm etal functionalization have recently been combined to give access to bio-inspired POV-based water oxidation catalysts.I no ne example,S treb and co-workers reported the manganesefunctionalized POV[Mn 4 V 4 O 17 (OAc) 3 ] 3À ({Mn 4 V 4 })asafunctional model of the oxygen-evolving complex [CaMn 4 O 5 ], which enables water oxidation in the natural photosyste-mII. [95] When coupled with the photosensitizer [Ru(bpy) 3 ] 2+ (bpy = 2,2'-bipyridine) and the terminal oxidant persulfate, the catalyst shows water oxidation under irradiation with visible light. [95] Tu rnover numbers of about 12 000 and turnover frequencies of about 100 min À1 were achieved with this system. [96] This first example of aP OV water oxidation catalyst highlights the usefulness of stabilizing polynuclear, highly redox-active transition-metal-oxo clusters as electrontransfer and storage sites for (proton-coupled) multielectron transfer. Thee xample also highlights the vast opportunities offered by POMs for undertaking fundamental as well as applied water oxidation reactivity studies. [97] Recent reports, for example,b yH ayashi and co-workers show that there might be awhole family of Mn-functionalized vanadates with redox-catalytic activity waiting to be explored. [98]

Outlook and Emerging Topics
To-date,POV energy research has mainly been focused on component design for batteries and photoredox catalysis. However,the development of advanced POVs together with adeeper understanding of their reactivity and stability could lead to accelerated advances in established fields and new applications in sustainable energy research. Thef ollowing section highlights current research opportunities in the areas discussed in Sections 3-5, as well as pioneering examples of emerging trends and future directions.

Future Directions in Established POV Energy Research
One major challenge in future POVe nergy research is acomprehensive analysis of POVreactivity and stability.This requires POVe xperts to collaborate closely with materials scientists,electrochemists,and photochemists to gain adeeper understanding of POVs tructure-function relationships.T his could be used to develop new POVs with tunable reactivity and identify new POVstructures as targets of future research.  12 ]. This is as tark contrast to the fast-paced recent progress in the design and functionalization of new POVarchitectures-many of which have not been studied for their possible use in energy technologies. [21,23,27,36] Thus,n ew concepts such as POVm odification with metal cations or organic ligands need to be aligned with the requirements of modern energy technologies.G round-breaking work in this direction has recently been reported by Matson and coworkers,w ho used both organic functionalization and metal substitution in [V 6 O 7 (OR) 12 ]totune the redox reactivity and electron-storage capability. [29,54,99] Related work by Streb and co-workers showed that the incorporation of redox-active transition metals into the organo-soluble dodecavanadate [V 12 O 32 Cl] 5À is afacile route to tune redox behavior. [100] More recently,the group showed that even incorporation of redoxinactive Ca 2+ ions into the dodecavanadate cluster leads to as ignificant enhancement of the cluster redox activity,t hus allowing the fabrication of LIB cathodes with improved energy density.T he authors attributed this surprising finding to the electrostatic and structural stabilization of the reduced cluster by Ca 2+ ions. [45] Beyond the chemical tuning of the POVi tself,m ajor challenges related to the interaction and deposition of POVs onto surfaces to construct electrodes still exist. These need to be overcome to establish general procedures for the stable "wiring" of POVs to electrodes or semiconductor surfaces. This is essential for any charge-transfer application, and to prevent the leaching of POVs (e.g. in batteries or electrolysis), which is still am ajor challenge in the field. Recent initial studies have demonstrated that POVd eposition on nanostructured carbon mediated by an ionic-liquid "binder" could be one viable approach. [101] In arelated study,Sonoyama and co-workers demonstrated that coating {PV 14 } nanoparticles with the conductive polymer (CP) polypyrrole results in significant improvements of the LIB capacity and cycling stability; [102] thus,P OM-CP composites could open new research possibilities. [103] In addition, organo-functionalized POVs could be anchored to electrodes through covalent [104,105] or supramolecular interactions. [16] As an alternative approach, Walsh, Newton, Khlobystov,a nd co-workers recently demonstrated the embedding of POMs inside electroactive carbon nanotubes.T his ground-breaking concept could in future be used to design composite electrode materials where the POM Angewandte Chemie Reviews 7528 www.angewandte.org is electrically "wired" to ac arbon support and leaching is virtually impossible. [106] In contrast, the challenge for RFBs is how to keep POVs in different charge states in solution and prevent precipitation or decomposition. This is particularly challenging in organic electrolytes,e ven for organo-functionalized POVs.R ecent studies with tungstate POMs,h owever, have opened af acile route to overcome this issue.P OM anions can be combined with large organic (often alkylammonium) cations to form socalled POM ionic liquids (POM-ILs), which are highly soluble in most organic solvents. [107,108] We suggest that transferring this concept to POVc hemistry could lead to highly organosoluble POV-ILs for POV-based RFBs.López and co-workers have recently shown that computational methods can be used to understand and predict the behavior of POV/organocation mixtures in organic solvents,w hich could open new avenues towards designer RFB electrolytes. [109] Another RFB-related research theme requiring urgent attention is the effect of the supporting electrolyte on the POVp erformance.V ery recently,M atson and co-workers showed that replacing bulky,low-charge-density cations (e.g. nBu 4 N + )i nt he supporting electrolyte with Li + or Na + significantly altered the redox potentials of their POV. [29,110] Thes tudies highlight that tuning the electrolyte composition in addition to tuning the POVi sapromising means for controlling POVe lectrochemistry.T hat said, high-enough concentrations of soluble POVs alts or POV-ILs would remove the need for added supporting electrolyte.
In the field of light-driven POVc hemistry,amajor bottleneck is to increase the visible-light harvesting of POVs, and new design concepts are required to shift the absorption of light by POVs (and POMs in general) into the visible region by tuning of the HOMO and LUMO gap.Inaddition, the ability to adjust the HOMO and LUMO energy levels (i.e. their redox potentials) would enable the targeted design of POVs for specific reduction or oxidation reactions.I nspiration for this HOMO-LUMO tuning might come from modern semiconductor chemistry,w here modification of the analogous valence and conduction bands (e.g. by heteroelement doping) is well-established. [111,112] These concepts,h owever, have so far not been systematically transferred to POV chemistry and have only recently received attention for POMs in general. [113][114][115] In pioneering studies,Newton and coworkers have demonstrated that the organic functionalization of Dawson polyoxotungstates can be used to control the HOMO-LUMO gap and HOMO-LUMO position, thereby enabling the controlled tuning of the resulting electro-and photochemistry. [116,117] An alternative approach towards higher visible-light photoactivity is the coupling of POVs with metal complexes or organic photosensitizers, [87] or POVd eposition on lightabsorbing semiconductors. [118,119] Principal studies in this direction have been undertaken with POMs;h owever,t hese studies also identified new challenges which arise from these more complex systems. [118] In alandmark study,K raus,Bren, Matson, and co-workers have very recently expanded this concept to POVs:t he authors deposited [V 6 O 7 (OEt) 12 ]a s hole scavengers on semiconducting CdSe light absorbers and demonstrated that this system shows enhanced light-driven hydrogen evolution. Thea uthors assign this observation to the ability of the POVt oa ct as ah ole scavenger, which facilitates proton reduction on the semiconductor surface. [120] In the field of C À Hactivation, MacMillan and co-workers recently reported ag round-breaking study in which the tungstate archetype [W 10 O 32 ] 4À was used as aphotocatalyst to trigger the abstraction of aC ÀHh ydrogen atom, which enabled the arylation [121] or trifluoromethylation [122] of aliphatic C À Hb onds.T his opens new opportunities for POV chemistry,w here HOMO-LUMO tuning could be used to control reactivity and adjust selectivity as well as lead to control over product formation and the suppression of side reactions. [87]

Emerging Themes in POV Energy Research
POVb attery research is now well-established for LIBs and-to al esser degree-NIBs.I nc ontrast, other batteries, for example,M g-or Ca-based systems,h ave not yet been explored. [123,124] In addition, the use of POVs in supercapacitors has thus far not been explored in detail. Supercapacitors can store energy through ac ombination of electrical double layer capacitance and pseudocapacitance (i.e.f aradaic electrochemical reactions), [49] so the use of highly redox-active POVs in supercapacitors seems promising.G round-breaking studies by Stimming, Srinivasan, and co-workers reported the successful use of Na 6 [V 10 O 28 ]a sa ne lectrode material for as upercapacitor, [125] which gave an energy density of 73 Wh kg À1 and ap ower density of 312 Wkg À1 ,b ased on doublelayer and pseudocapacitance.T he study lays the foundations to expand POVr esearch into the supercapacitor domain.
Another area which holds great potential for POVs is electrocatalysis.H ere,P OVsc ould be used as molecular species in homogeneous solution [20,97,126] or deposited on electrodes for heterogeneous applications. [16,103,118] To date, POVe lectrocatalysis is largely unexplored. Most of the existing studies reported the incorporation of POVs as as olid into carbon paste electrodes,s ot hat analyses of the reactivity and stability of the POVu nder electrochemical conditions is challenging. [62] In addition, the studies often targeted facile electrocatalytic processes,f or example,t he conversion of nitrite,bromate,iodate,orascorbic acid. [127][128][129] In contrast, little has been reported on more challenging (proton-coupled) multielectron reactions for small-molecule activation (e.g. H 2 ,O 2 ,CO 2 ,N 2 ). In one recent example,Streb and co-workers demonstrated that the molecular-water oxidation catalyst {Mn 4 V 4 } (see Section 4) also shows sustained oxygen evolution under electrochemical conditions; [95,97] however,t he study provided little insight into the mechanism for the oxidation of water, and no information is available on the cluster stability or the role of the vanadate ligand during catalysis.I na nother pioneering example,L i, Yang et al. used Cu + ions to link organo-functionalized Lindqvist POVs into metal-organic frameworks,w hich showed intriguing reactivity for the electrocatalytic oxygen reduction reaction when deposited on acarbon electrode. [130] Building on these pioneering reports,f uture studies could focus on exploring structure-reactivity-stability relationships on the molecular level to identify the nature of the electrocatalytically active species.I na ddition, focus on technologically important reactions is required so that fundamental mechanistic studies and application-driven POVd evelopment can be combined to access high-performance electrocatalysts.A gain, the stable "wiring" of POVs to electrode surfaces is still virtually unexplored and forms aprerequisite to access systems with high durability and technological relevance. [118] In the field of small-molecule activation, Kikukawa et al. have recently broken new ground by using the bowl-shaped dodecavanadate [V 12 O 32 ] 4À to bind elemental Br 2 in the central cavity,t hereby leading to polarization of the BrÀBr bond and significant selectivity changes in organic bromination reactions. [131] In ar elated example,D as and co-workers demonstrated that POVs can sequester CO 2 from air and store it as an internal CO 3 2À template,thereby resulting in the species [H 8 V IV 15 O 36 (CO 3 )] 6À . [132] If this approach can be coupled to ar eversible template release (as reported for several POMs), [133][134][135] this discovery could open new horizons for the capture and storage of molecular carbon using the unique structural disassembly and reassembly properties of polyoxometalates. [32,40,43,136,137]

Conclusion
POVs have proven to be versatile redox-active materials that can be used to overcome challenges in many sustainable energy applications,f rom batteries to light-driven catalysis. Despite this,only as mall fraction of the known POVcluster types have been investigated as components of energy conversion or storage systems.F urthermore,m ore extensive investigation of the physical properties of existing species is needed to use them effectively in energy applications,such as, determining trends in redox behavior, photophysical properties,thermal stability,and solubility.Expanding the available knowledge of POVp roperties will allow the field to move beyond the model POVs described here and allow POVs to be designed to overcome current limitations.B ya ddressing the issues we have highlighted, we believe that POVs can become key components in the development of future sustainable energy storage systems.