Store or catalyze? The M–O bond decides

Abstract

The nature of the metal–oxygen (M–O) bond is a key electronic factor that decides if an oxide/hydroxide serves as a charge-storage medium or an electrocatalyst. When the M–O bond is highly ionic and electronically stable, oxygen remains inactive. In this case, the material stores charge through fast, reversible metal-centered redox, which is typical of battery-like and pseudocapacitive behavior. On the other hand, increasing M–O covalency changes the electronic environment. Antibonding states become accessible, oxygen becomes chemically active, and the M–O unit can perform the bond-making and bond-breaking steps needed for catalytic activity. Small, intentional changes in M–O covalency—such as doping, defect engineering, oxidation-state tuning, or lattice strain—can switch materials between these two modes. These changes decrease the potential barriers and allow for the formation of oxygen-evolving intermediates. This view brings these findings together in a clear framework where M–O covalency acts as a switch between “store” and “react.”

---

1. Introduction

Have we ever wondered why some materials have strong pseudocapacitive behavior while others show high OER activity, even within the same potential region? For instance, cyclic voltammetry of two Ni-based electrodes can display nearly identical redox peaks related to the Ni²⁺/Ni³⁺ (or Ni³⁺/Ni⁴⁺) transition (Scheme 1). However, in one case, the material stores charge reversibly with little oxygen evolution, while in the other, it quickly produces O₂ at the same applied potential. These observations indicate that the nominal redox transition alone does not explain the functional differences. This leads us to the question: what factors determine if the material acts as a pseudocapacitor or an OER catalyst?

Scheme 1 Conceptual illustration showing how materials with similar redox behavior can work either as pseudocapacitors or as OER catalysts under similar electrochemical conditions. The cartoon in the middle of the illustration was created using ChatGPT.

Modern energy technologies rely more on materials that can store charge effectively and speed up chemical reactions. These two functions are essential for the operation of devices such as batteries, supercapacitors, and water-splitting electrolyzer systems. Charge-storage materials allow for quick ion uptake and release, ensuring reliability and stabilizing fluctuations in renewable energy grids.^1–3 Catalysts speed up important chemical changes, such as hydrogen evolution and oxygen evolution from the electrolysis of water. This acceleration affects the overall efficiency and sustainability of electrochemical processes. The interplay between charge storage and catalysis is crucial for next-generation clean-energy systems, which need both high power delivery and precise reaction control.⁴

While strong catalytic activity is important for achieving low overpotentials, high selectivity, and fast reaction rates, it is also critical that the material does not build up excessive capacitive charge during operation. A large capacitive current, resulting from double-layer charging or pseudocapacitive faradaic steps, can obscure the material's intrinsic catalytic performance, inflate apparent current densities, and complicate accurate evaluations of turnover frequencies or product selectivity. This issue is especially pronounced in transition-metal oxides, hydroxides, and layered double hydroxides, which show significant pseudocapacitive behavior along with promising catalytic properties. In these materials, capacitive contributions can hide true faradaic processes, leading to overestimated catalytic activity and misunderstood electrochemical responses.^5–9

It is crucial to understand this relationship because too much pseudocapacitive storage can hinder catalytic turnover by trapping oxygen-containing species. This can block active sites and delay forming high-valence states needed for the oxygen evolution reaction (OER). Conversely, a material that is excessively catalytically active may lack the stability or structural reversibility for long-term energy storage.^10,11

An increasing amount of evidence shows that these competing behaviors—charge storage and catalysis—are not separate; they are fundamentally influenced by the material's electronic structure, particularly the nature of the metal–oxygen (M–O) bond (Fig. 1a). The strength, polarity, and covalency of the M–O bond determine if the surface stabilizes intermediates, promotes electron transfer, or stores the charge through surface redox. Highly ionic M–O bonds usually support fast and reversible metal-centered redox, allowing for pseudocapacitive energy storage. On the other hand, stronger M–O covalency reduces the energy of antibonding states and enables oxygen to take part in bond rearrangements. These conditions are crucial for the formation and conversion of key catalytic intermediates such as *OH, *O, and *OOH. Therefore, the same M–O unit can either accumulate charge or activate reaction intermediates, depending on its electronic configuration.^12–16

Fig. 1 (a) Schematic illustration highlighting the fundamental kinetic distinction between charge-storage materials (pseudocapacitive) and catalytic materials (OER-active); (b) comparison of current decay behavior for catalytic currents versus capacitive currents, emphasizing their different time-dependent responses; (c) cyclic voltammograms of pure Ni(OH)₂ and Fe-incorporated Ni(OH)₂ showing the reduction in potential lag and earlier onset of catalytic current upon Fe doping; (d) conceptual schematic summarizing key electrochemical differences between capacitive Ni(OH)₂ and catalytically active Fe–Ni(OH)₂, including redox behavior and onset characteristics; and (e) electrochemical response of carbonate- and sulfate-intercalated NiFe-LDH, illustrating how high-basicity CO²⁻₃(electron-rich) enhances OER activation, whereas low-basicity SO²⁻₄ (electron-poor) favors pseudocapacitive charge storage. Center CV image is reproduced from ref. 26 with permission from Royal Society of Chemistry, copyright 2024.

When looking at long-term operational stability, the current decay mechanism shows that the capacitive current, which includes both non-faradaic double-layer charge and pseudocapacitance, decreases much faster than the faradaic catalytic current. This difference in speed comes from how each of these currents works. The capacitive component, which represents electric double layer (EDL) charging, exhibits a rapid exponential decay that is determined by the RC time constant, usually ranging from µs and ms. On the other hand, the faradaic current, which is limited by mass transport or diffusion, decays more slowly in an algebraic manner. This behavior is explained using the Cottrell equation's dependence on the t^−1/2 (Fig. 1b). As a result, a high capacitive contribution can both deceptively lower the observed catalytic performance and cause reduced current stability during long electrochemical operations.

Recent developments in DFT modeling, operando spectroscopy, and X-ray/IR/Raman techniques have started to uncover how electron density, orbital hybridization, and local bonding geometry change at applied potential. These tools show how active sites form, change, and deactivate, and how M–O covalency shifts during redox cycling. Such insights help minimize unwanted capacitive effects, improve catalytic turnover, and ensure stability in both structure and electrochemistry.^17–20

In the end, the M–O bond serves as the electronic pivot that determines whether a material acts as a high-capacity charge-storage medium or an effective catalyst. Moreover, instead of being a competing term, M–O covalency offers a foundation for understanding why established descriptors such as e_g occupancy, lattice oxygen participation, and conductivity are linked to catalytic activity.^21–24 Therefore, by carefully adjusting its covalency, bond strength, and orbital arrangement, it is possible to develop materials that integrate both functions, providing high energy storage, enduring long-term cycling, and excellent catalytic performance.

In this discussion, we emphasize how the electronic structure of the M–O bond shapes the trade-off between storage and catalysis. We summarize recent progress in adjusting M–O interactions and propose design principles for next-generation multifunctional materials geared toward sustainable energy technologies.

2. How pseudocapacitance competes with the OER

Fig. 1a–c illustrates the key difference between charge-storage behavior and catalytic activation in nickel hydroxides. In pure Ni(OH)₂, the main electrochemical process is the reversible Ni²⁺/Ni³⁺ redox transition, seen as strong anodic and cathodic peaks.²⁵ This process stores charge efficiently, typical of pseudocapacitive materials. However, it maintains surface oxygen in a stable, non-reactive state. As a result, the onset of catalytic current shows a noticeable potential delay after the redox peak. This potential lag directly indicates that, while charge accumulates easily, it is not immediately used to drive the oxygen evolution reaction (OER). Adding Fe (2 ppm aq. solution of Fe(NO₃)₃·9H₂O) changes this behavior significantly. Fe does not have its own distinct redox peak, but its presence alters the electrochemical response of Ni(OH)₂ in two main ways: (1) the catalytic current starts closer to the Ni oxidation potential. This shows that the system moves more easily from the redox state to the OER pathway (Fig. 1d); and (2) the gap between Ni oxidation and the rise of catalytic current shrinks. This means that less charge is held up in pseudocapacitive processes and more goes directly into catalytic turnover.

These features indicate that adding Fe improves the connection between the Ni²⁺/Ni³⁺ redox transition and the steps needed to start the OER. Instead of mainly storing electrons within the redox-active Ni framework, the charge is more effectively transferred toward initiating the reaction. This results in a quicker rise in current at lower potentials. This behavior also supports broader insights from studies on transition-metal hydroxides. Materials with strong pseudocapacitive peaks often show delayed catalytic activation. In contrast, materials with smoother redox transitions and small potential gaps typically enable faster oxygen evolution. In other words, the balance between charge storage and catalysis is directly reflected in the shape and timing of the features in the cyclic voltammogram (Fig. 1c and e).

A similar mechanistic picture explains the Fe-induced activation in Ni(OH)₂ when comparing carbonate- and sulfate-intercalated NiFe-Layered Double Hydroxide (LDH).²⁶ This reveals a common principle linking pseudocapacitive behavior with catalytic willingness. In pristine or weakly activated materials, such as Ni(OH)₂ or sulfate-NiFe-LDH, the redox transition (Ni²⁺ → Ni³⁺) occurs smoothly and reversibly. This process allows significant charge to build up in the surface or interlayer region. The stored charge results in a strong pseudocapacitive response. It also creates a noticeable potential lag between the end of Ni oxidation and the start of measurable OER current. The oxidized sites focus on charge storage rather than quickly converting into OER-relevant intermediates, such as M–OOH.

When Fe is added to Ni(OH)₂, or when a more strongly interacting interlayer ion, such as CO²⁻₃, replaces SO²⁻₄ in NiFe-LDH, this balance changes.²⁷ The redox transition becomes less isolated, the charge-storage capacity decreases, and the catalyst more easily channels the oxidized state into chemical turnover (Fig. 1e). In CO²⁻₃-intercalated NiFe-LDH, the closer electronic coupling between the layers and the metal centers limits excessive pseudocapacitive buildup and shortens the potential lag. This leads to an earlier OER onset, mirroring how Fe doping allows for a faster transition from Ni³⁺ formation to generating reactive OER intermediates. In contrast, the SO²⁻₄ variant, like undoped Ni(OH)₂, maintains a smoother and more reversible redox feature. Therefore, materials that promote large pseudocapacitive storage have delayed OER kinetics. Meanwhile, materials that direct charge away from storage and toward intermediate formation display a faster catalytic response. This parallel underscore a broader principle: modifying the local environment around Ni—whether through cation substitution (Fe doping) or anion exchange (CO²⁻₃vs. SO²⁻₄)—shifts the balance between pseudocapacitance and catalysis. This change affects how effectively the material moves from storing charge to producing oxygen.

Overall, it is notable that materials with strong pseudocapacitive peaks often have delayed catalytic activation. In contrast, materials with smoother redox transitions and small potential gaps usually allow for faster oxygen evolution. The balance between charge storage and catalysis shows up in the shape and timing of the features in the cyclic voltammogram. Since strategies such as doping, making heterostructures, creating defects, or engineering interlayers are often used to adjust either pseudocapacitive or catalytic behavior, the CV-based diagnostic parameters in Table 1 could be especially useful. These parameters offer a straightforward and quantitative way to evaluate whether a certain activation strategy enhances the material's charge storage or improves its catalytic turnover.

Table 1 CV-derived descriptors distinguishing pseudocapacitive and catalytic behavior. Oxidation-peak shifts, potential lag (ΔE), and redox reversibility help assess how easily a material stores charge or activates for the OER. The kinetic parameter b in the equation i = aν^b (where i is the current, ν is the scan rate, and ‘a’ is a constant) further distinguishes surface-controlled pseudocapacitance (b ≈ 1) from diffusion/PCET-controlled catalytic processes (b ≈ 0.5)

S. no.	CV features	Pseudocapacitive behavior	Catalytic (OER) behavior
1	Oxidation peak position	Appears at lower potential (easy M^n→M^n+1 redox)	Shift to higher potential
2	Potential lag ΔE (difference between M^n/M^n+1 peak and OER onset)	Large ΔE → redox occurs but not trigger OER	Small ΔE → fast coupling of M–O oxidation with OER onset
3	Scan rate dependence (i = aν^b)	b ≈ 1.0 →surface-controlled, pseudocapacitive	b ≈ 0.5 →catalytic, diffusion/PCET controlled
4	Reversibility	Strongly reversible (essential prerequisite)	Poor reversibility

---

In conclusion it is important to note that these electrochemical features reveal what changes during Fe incorporation and provide a deeper understanding of why this transition occurs. This necessitates us to examine the underlying electronic structure. In the following section, we discuss how variations in the M–O bond character, including its strength, polarity, and orbital interactions, govern the shift between pseudocapacitive storage and catalytic oxygen activation.

3. The role of OH⁻ binding strength and M–O bonding

In alkaline media, transition-metal oxides/hydroxides interact with OH⁻ through formation of surface M–OH and M–O species, and the strength and nature of these bonds are controlled by the electronic structure of the metal center. The initial OH⁻ adsorption induces the metal oxidation to form M–OH, deprotonation then generates M–O, and further oxidation may lead to reactive intermediates such as M–OOH (in OER) or a stable M–O redox couple (in pseudocapacitance). Whether surface intermediates such as M–OH, or M–O remain stable or become reactive (by forming M–OOH followed by O₂ removal) depends on how the metal cation redistributes charge with oxygen, a process dictated by d-orbital filling, metal–oxygen orbital overlap, and the covalent/ionic character of the M–O bond. In transition-metal oxides and hydroxides, the metal d-orbitals hybridize with the oxygen 2p orbitals to generate bonding (O-dominated) and antibonding (M-dominated) states.²⁸

In octahedral symmetry, the hybridization of transition metal (TM)-nd and O-2p leads to formation of a lower energy bonding molecular orbital (MO) band and a higher energy antibonding band, thereby formed an extended M–O lattice, as shown in Fig. 2a The lower energy bonding MOs (t_2g & e_g) are largely bearing the low-lying O character whereas the antibonding MOs ( and ) are enriched with metal character (Fig. 2b). In the hybridized state, the polarity of the M–O bond plays a critical role in determining its reactivity and associated redox behavior.²⁹ The degree of polarity depends on the electronegativity (χ) difference between oxygen and the transition metal: a large χ difference yields a more ionic M–O bond, while a smaller difference enhances covalency, as clearly shown in Fig. 2c. This trend can be rationalized using a fundamental chemical principle—the acidity of hydrogen halides (H–X, where X = F⁻, Cl⁻, Br⁻, I⁻). As we move from HF to HI, the electronegativity (χ) difference between H and X decreases, reducing bond polarity, a concept familiar from undergraduate chemistry (Fig. 2d). Similarly, in M–O units, the bonding electrons in the molecular orbitals reside largely on the oxygen atom due to its higher electronegativity.³⁰ A large χ difference between the metal and oxygen stabilizes these electrons, reinforcing the ionic character of the M–O bond. Greater electron stability on oxygen makes the oxide less reactive, shifting the electrochemical activity to the metal center rather than the M–O bond itself. Under such conditions, the metal can undergo fast and reversible redox transitions (M ↔ Mⁿ⁺ + ne⁻) without breaking the M–O framework. This behavior is characteristic of pseudocapacitance, where charge storage occurs through rapid surface redox processes while maintaining the structural integrity of the oxide.³¹

Fig. 2 (a) Schematic molecular orbital representation showing octahedral splitting of transition-metal nd levels and their interaction with O-2p orbitals; (b) simplified band diagram illustrating metal–oxygen hybridization and the relative contributions of metal nd and oxygen 2p states; (c) conceptual depiction of how electronegativity differences tune the ionic versus covalent character of the M–O bond; and (d) classical example using hydrogen halides (HX) to demonstrate how electronegativity mismatch between heteroatoms governs bond polarity and the resulting ionic/covalent nature.

This stable M–O bond under a large χ difference resists further activation from M–O to M–OOH and thereby minimizes catalytic activity. This trend is supported by the fact that the left side congener in the 3d TM series possesses very low electrocatalytic activity towards the OER, yet exhibits profound capacitor properties (Table 2). This behaviour is consistent with the experimentally established oxophilicity trend among 3d transition-metal hydr(oxy)oxides (Mn > Fe > Co > Ni), where Mn and Fe bind OH* too strongly, trapping oxygenated intermediates on the surface. As shown by Subbaraman et al., the strength of the OH_ad–M interaction dictates not only OH adsorption but also the reaction kinetics for the OER.³² When the binding is excessively strong (high χ difference and ionic M–O bond character), the surface becomes “poisoned,” preventing the M–O unit from further activation to M–OOH, which is essential for the OER sequence. Importantly, the HER, CO oxidation, and OER activities of 3d-M hydr(oxy)oxides all follow a single monotonic trend—Ni > Co > Fe > Mn—revealing that the same descriptor (OH_ad–M bond strength) controls reactivity across the entire potential window. Mn and Fe, being highly oxophilic, stabilize adsorbed OH and O to such an extent that further transformation toward reaction intermediates becomes energetically unfavourable. This intrinsically slow conversion results in negligible catalytic activity but enables fast, reversible redox of the metal center (M ↔ Mⁿ⁺ + ne⁻) without breaking the stabilized M–O framework. Such behaviour manifests as pseudocapacitive energy storage rather than catalytic turnover.

Table 2 Pseudocapacitive behavior in 3d transition-metal materials as a function of M–O bond ionicity and redox reversibility

Electrode material	Sp. capacitance (F g^−1)	Reference
Scandium (Sc) doped into NiCo-LDHs	1695	33
PAN-PPh./TiO2	484	34
Aluminum vanadium oxide	761	35
Cr2O3/graphene nanocomposite	462.2	36
MnCo2O4@NiMo-LDH	1009	37
Iron oxide	191.4	38
Co(OH)2/rGO	2688	39
NiS	1066	40
Cu(OH)2	2777	41

---

The difference in electronegativity (χ) between the metal and oxygen determines the initial polarity of the M–O bond. This difference indicates whether the M–O interaction is mostly ionic or covalent. A large χ difference results in a highly polarized, ionic M–O bond. This stabilizes O²⁻ and keeps the electrons localized on oxygen. These conditions support fast metal-centered redox processes and allow for pseudocapacitive charge storage without breaking the M–O bond. However, the polarity driven by χ is not the only factor that affects M–O reactivity. The oxidation state of the transition metal influences the number of antibonding states and the position of the d-band center. At lower oxidation states, the metal has more d-electrons. This keeps the d-band center closer to the Fermi level. The larger electronegativity difference between metal and oxygen creates a more ionic M–O bond. This stabilizes lattice O²⁻ and allows for fast, reversible M ↔ Mⁿ⁺ redox without breaking M–O linkages. This behavior is typical of pseudocapacitance. At higher oxidation states, the count of d-electrons decreases and the d-band center moves further below the Fermi level. This increases M–O covalency through stronger hybridization between metal and oxygen orbitals. The lowered d-band makes antibonding states easier to access, reduces the stabilization of lattice oxygen, and allows oxygen atoms to take part in chemical reactions. These conditions are necessary for the OER, where M–O bonds must form and break.

This electronic picture is strongly supported by experimental observations involving Fe³⁺ incorporation into Ni(OH)₂ and Co(OH)₂ during alkaline OER. Even small amounts of Fe impurities in commercial KOH significantly improve the electrocatalytic activity of Ni- and Co-based hydroxides. In the study by Boettcher et al., adding Fe³⁺ ions to the electrolyte (from Fe-free to 5–25% Fe) results in a gradual increase in OER activity along with a rise in electrical conductivity.⁴² While the authors link this improvement to increased conductivity and Fe-induced partial charge transfer activation, the CV profiles provide a deeper understanding consistent with our electronic framework.

In Fe-free Ni(OH)₂ (Fig. 3a–b), the anodic sweep displays a sharp and intense Ni²⁺/Ni³⁺ redox peak that corresponds to the Ni(OH)₂ + OH⁻ ↔ NiOOH + e⁻ + H₂O process. After this transformation, a noticeable potential gap (lag) appears between the redox activation and the start of the OER. This behavior occurs when the M–O bond is mainly ionic, with the Ni d-band closer to the Fermi level and the antibonding states mostly filled. Under these conditions, the M–O framework stays stable, the redox cycling is quick and reversible, and the system acts as pseudocapacitive system—storing charge without transforming M–O (or M–OH) into M–OOH. As the amount of Fe³⁺ incorporation increases, the redox peak shifts, and the potential lag between Ni oxidation and catalytic current gradually disappears. Fe³⁺ has a higher effective electronegativity and a higher oxidation state, pulling electron density from neighboring Ni centers through M–O–Fe bonding. This electron withdrawal lowers the Ni d-band center compared to the Fermi level and reduces the occupation of M–O antibonding states. Consequently, the M–O bond becomes more covalent, oxygen becomes chemically active, and the barrier for the formation OER decreases. Thus, Fe incorporation electronically drives Ni(OH)₂ from stable ionic M–O pseudocapacitance to reactive covalent M–O OER catalysis.

Fig. 3 Effect of Fe³⁺ incorporation on Ni(OH)₂ and Co(OH)₂ activation toward the OER. (a and b) Increasing Fe³⁺ content reduces the potential gap between the Ni²⁺/Ni³⁺ redox peak and OER onset. It also increases conductivity, which indicates faster activation of NiOOH for catalysis; reproduced from ref. 42 with permission from American Chemical Society, copyright 2014. (c and d) A similar pattern occurs with Co(OH)₂. Fe incorporation shifts the redox peak and reduces the activation delay. This shows increased M–O covalency and better OER kinetics; and (e and f) TOF plots display a volcano trend with Fe content. This shows that an optimal level of Fe maximizes intrinsic activity by balancing OH⁻ adsorption and oxygen release; reproduced from ref. 43 with permission from American Chemical Society, copyright 2015.

A similar trend is seen in Co(OH)₂ (Fig. 3c and d). Increasing Fe content shifts the Co²⁺/Co³⁺ redox peak to a higher potential and narrows the activation-kinetic current separation.⁴³ This shows the same electron-withdrawal effect: the decrease in d-electron population due to Fe increases M–O covalency and helps initiate the OER. An intrinsic TOF comparison (Fig. 3e and f) further confirms that Fe incorporation boosts catalytic turnover efficiency.

Pseudocapacitance and the OER both start with OH⁻ adsorption, forming M–OH and M–O species. However, their energetic needs differ. Pseudocapacitance requires strong, stable M–O bonds that can store charge without changing the structure. In contrast, the OER needs weakened and covalent M–O bonds that can activate oxygen and enable catalytic turnover. Therefore, tuning the d-band center based on oxidation state acts as an electronic switch between ionic M–O bonds for energy storage and covalent M–O activation for catalysis.

The Ru–Co₃O₄ system shows how the electronic structure works in practice.⁴⁴ As the content of Ru increases from 0 to 20%, the area of the redox peak increases. This indirectly shows a higher TOF and explains how the variation of Ru affects the volcano plot shown in Fig. 4a and b. This change indicates a higher buildup of positive charge on the electrode surface because of increased metal activation through M ↔ Mⁿ⁺ redox. More charge accumulation leads to more OH⁻ adsorption, which initially raises the OER current density (Fig. 4c). However, when the surface charge becomes too high, the M–O bond gets very stable. In this situation, the catalyst surface locks into the *O intermediate state and cannot progress to the formation of *OOH. This behavior is similar to pseudocapacitive storage. Charge keeps building up through reversible M–O redox, but the M–O bond is too stable to break or form O–O. On the other hand, when charge accumulation is low, OH⁻ adsorption is not enough, and catalytic processes slow down. Only with an intermediate level of surface charge (Ru–Co₃O₄-15), where the M–O bond strength is moderate, is there an optimal balance between adsorption and desorption, leading to the highest OER current density. A similar trend is seen in the Ru-decorated CoFe-LDH system (Ru@CoFe-LDH) reported by Karmakar et al.⁴⁵ In this study, Ru nanoparticles were placed on the CoFe-LDH surface. The accumulated surface charge during the oxygen evolution reaction (OER) was measured from the redox area under the pre-oxidation peaks. Increasing the Ru loading raised the accumulated charge on the LDH surface. This change showed a higher positive charge density and stronger metal oxidation during the activation step. Generally, a higher accumulated charge encourages stronger OH⁻ adsorption. Therefore, one might expect that the highest Ru loading would improve OER kinetics. However, the catalytic activity did not follow a straightforward trend. Instead, Ru@CoFe-LDH(3%) showed better OER activity than both lower (0–1%) and higher (5%) Ru loadings. When plotting accumulated charge against current density, a clear volcano relationship appeared (Fig. 4d). The catalyst with intermediate Ru content (3%) sat at the peak. The resulting volcano-shaped relationship between OER activity and accumulated charge directly proves the Sabatier principle.

Fig. 4 Effect of Ru incorporation on OER activity and charge-storage behavior: (a) TOF increases with Ru loading and peaks at Ru–Co₃O₄-15, indicating optimal catalytic turnover; (b) volcanic relationship between OER current density and accumulated charge, showing that intermediate Ru content provides the best balance of adsorption/desorption (Sabatier principle); reproduced from ref. 44 with permission from Royal Society of Chemistry, copyright 2023; (c) reduction peak area increases with Ru loading, indicating enhanced pseudocapacitive charge storage at higher Ru levels; and (d) similar volcano trend observed for Ru@CoFe-LDH, confirming that intermediate Ru loading optimizes surface charge and maximizes OER activity. Reproduced from ref. 45 with permission from American Chemical Society, copyright 2023.

This distinction becomes clearer when we consider the Sabatier principle. It states that the adsorption of reaction intermediates must be neither too strong nor too weak to achieve the best catalytic turnover. In alkaline OER, the catalytic cycle progresses through the sequential adsorption and conversion of surface intermediates (M − OH* → M − O* → M − OOH*). If the M–O bond is too strong, as seen in materials with a high d-band center and mostly ionic M–O character, the surface gets stuck in the *O state. This stops it from progressing to *OOH; the catalyst becomes electronically locked and behaves more like a pseudocapacitive charge-storage material. On the other hand, if the M–O bond is too weak, intermediates can desorb too easily. This prevents the formation of O–O bonds needed for the OER. Only when the d-band center is adjusted to an intermediate level does the metal bind *OH/*O/*OOH with just the right strength to allow for adsorption, transformation, and desorption, enabling efficient OER. Thus, the Sabatier principle offers a thermodynamic framework that adds to the discussion of electronic structure; pseudocapacitors operate where strong OH⁻ binding and stable M–O bonds exist. Efficient OER catalysts work in a zone where M–O bonding is weakened and covalent enough to activate lattice oxygen, but not so weak that it loses control over adsorption. The best OER catalyst balances adsorption strength, M–O covalency, and d-band positioning to meet the Sabatier requirement.

Therefore, for bifunctional materials that can store charge and catalyze reactions, the best M–O bond should not be purely ionic or purely covalent. Instead, it should exist in a middle ground that can change. A bond with some ionic character allows for quick and reversible metal-centered redox reactions and effective pseudocapacitive charge storage. At the same time, a bond with some covalent character enables partial charge sharing with oxygen, stabilizes high-valence states, and allows for the creation of oxygenated intermediates important for catalysis. Therefore, the perfect bifunctional M–O bond acts like an electronic switch, changing from a state favorable for storage to one favorable for catalysis when a bias is applied. To achieve this behavior, careful control over the material's composition, coordination, defects, and surface chemistry will likely be necessary to keep it close to this electronic transition point.

4. From ionic to covalent: periodic trends governing energy storage and electrocatalysis

A clear picture develops when we look at the pseudocapacitive (Table 1) and OER catalytic (Table 2) datasets through the lens of metal–oxygen (M–O) bond covalency, while also considering structural factors. Generally, materials with a large electronegativity gap between the metal and oxygen, such as Sc, Ti, V, Mn, and Fe, form relatively ionic M–O bonds. These bonds localize electron density around oxygen. This localization stabilizes low-valence redox couples (M²⁺/M³⁺) and promotes fast, reversible ion–electron exchange. As a result, we see strong pseudocapacitive responses in Sc-LDHs, MnCo₂O₄@NiMo-LDH, and Fe-based LDHs listed in Table 2. In contrast, Table 3 shows that highly active OER catalysts typically have more covalent M–O bonds, such as Co³⁺–O and Ni³⁺/Ni⁴⁺–O. Here, strong 3d–2p mixing stabilizes high-valence intermediates (M–OOH*) and reduces barriers for O–O bond formation. This places these M–O bonds in the optimal bonding window for catalysis, consistent with the Sabatier principle.

Table 3 Catalytic behavior of 3d transition-metal materials depends on M–O bond covalency and intermediate binding strength

Catalyst	Overpotential	Ref.
Sc@CoTe/CNT	1.59	46
Metal doped TiO2	1.45	47
Vanadium pentoxide (V2O5)	330 mV	48
CoCr2O4	370 mV	49
Mn/Ru oxide	351	50
Fe50Ni30P13C	327	51
Co3O4	230 mV	52
NiOOH	130 mV	53
CuO	490	54

---

A notable exception to this covalency pattern is Cu-based hydroxides, which appear in the high-capacitance list even though Cu and O have a relatively small electronegativity difference, indicating greater covalency. This exception occurs because Cu²⁺ compounds experience strong Jahn–Teller distortion, which weakens certain Cu–O bonds and creates highly anisotropic, labile coordination environments. These weakly bonded axial Cu–O sites allow for rapid redox cycling (Cu²⁺/Cu⁺) and easy OH⁻ insertion/extraction. These behaviors are typical of pseudocapacitive storage rather than catalytic turnover. Additionally, Cu(OH)₂ nanowire and nanosheet structures enhance surface-controlled charge storage due to abundant defects and high surface area.

Another notable exception can be observed in the case of Co(OH)₂/rGO, which gives a high-capacitance value (Table 2) despite cobalt oxides/hydroxides usually having high covalency and strong OER activity (Table 3). This isn't a contradiction. Instead, it shows how structural and interfacial engineering takes precedence over intrinsic covalency. The rGO scaffold improves conductivity and charge delocalization, while Co(OH)₂ allows for quick Co²⁺/Co³⁺ redox transitions. Together, they create outstanding pseudocapacitance despite cobalt's natural catalytic tendencies. Similar design-based differences can be seen in V₂O₅, where the layered structure aids ion intercalation despite the long V–O bonds. An apparent counter example to the covalency–activity relationship is seen in early transition-metal oxides such as TiO₂. Where despite strong Ti–O covalency it shows poor intrinsic OER activity. The empty d-orbital and energetically low-lying d-band limit the redox participation of the metal center for the progress of the OER.

Apart from the above mentioned exemption, taken together, these tables support a straightforward design rule for electrochemical materials:

• More ionic M–O bonds lead to electron localization, resulting in pseudocapacitance dominance.

• More covalent M–O bonds lead to electron delocalization, resulting in high OER activity.

However, structural distortions, morphology, and conductive frameworks, such as those in Cu(OH)₂ or Co(OH)₂/rGO, can override intrinsic covalency trends. These insights illustrate that while electronic covalency sets a baseline, local geometry and defect chemistry determine whether a material acts as a charge-storage electrode or an OER catalyst. Additionally, the framework proposed here should be seen as identifying covalency as an enabling factor that must work alongside suitable electronic filling and redox accessibility for high catalytic activity (Fig. 5).

Fig. 5 Conceptual map connecting M–O bond character with storage and catalytic performance, showing how material tuning shifts behavior between these regimes.

Additionally it is important to differentiate between inherent and potential-induced M–O covalency. Inherent covalency is determined by composition, coordination geometry, and orbital alignment. It affects the thermodynamic accessibility and stability of high-valence metal states, as shown in recent experimental and computational studies.^22,23,55 On the other hand, applied anodic potential changes the electronic structure by oxidizing the metal center and causing surface reconstruction. This process increases metal–oxygen hybridization and changes the M–O bond character in real time, as seen in operando XAS and spectroscopic studies.^56,57 Therefore, inherent covalency influences how easily a material can reach high-valence, catalytically active states, while potential cycling actively drives and stabilizes these states through dynamic changes in the M–O bond.

5. Future perspective

The discussion here shows that the metal–oxygen (M–O) bond is more than just a structural connection; it is the key factor that determines whether a material can store charge, catalyze reactions, or do both effectively. The M–O bond controls redox reversibility, intermediate stability, electron-transfer pathways, and active-site evolution. While considering the scope and main constraints of the M–O covalency descriptor, it is indeed necessary to state that although M–O bond covalency provides a strong electronic framework for explaining trends in charge storage and catalytic activity, it does not account for all factors affecting electrochemical performance. Structural features such as surface area, porosity, defect density, amorphization, hydration, and electrolyte-specific interactions significantly influence kinetics, mass transport, and stability. These factors must be considered alongside electronic descriptors. Additionally, M–O covalency represents an average electronic property and may not fully show spatial differences or dynamic changes during operation. Therefore, M–O covalency should be seen not as a single predictor of activity but as a common electronic view that helps interpret the effects of structural and chemical changes.

The degree of metal–oxygen covalency can be measured experimentally using O K-edge and metal L-edge X-ray absorption spectroscopy (XAS) (pre-edge intensity, edge shifts), XPS core-level shifts, Auger parameters, and vibrational spectroscopy (M–O stretching frequencies). For computational evaluation, covalency can be assessed through the overlap of metal d and oxygen 2p projected density of states, Bader or Hirshfeld charge analysis, crystal orbital Hamilton population (COHP) or integrated COHP (ICOHP) analysis, bond orders, and the positions of the metal d-band and oxygen 2p band centers. These metrics allow for a direct connection between electronic structure, adsorption energies, and electrochemical performance.

For bifunctional materials that can store charge and catalyze reactions, the best M–O bond should not be purely ionic or purely covalent. Instead, it should exist in a middle ground that can change. A bond with some ionic character allows for quick and reversible metal-centered redox reactions and effective pseudocapacitive charge storage. At the same time, a bond with some covalent character enables partial charge sharing with oxygen, stabilizes high-valence states, and allows for the creation of oxygenated intermediates important for catalysis. Therefore, the perfect bifunctional M–O bond acts like an electronic switch, changing from a state favorable for storage to one favorable for catalysis when a bias is applied. To achieve this behavior, careful control over the material's composition, coordination, defects, and surface chemistry will likely be necessary to keep it close to this electronic transition point.

Looking to the future, this field is important for two main reasons. First, as clean energy technologies grow, such as electrolyzers, CO₂ conversion devices, metal–air batteries, and hybrid supercapacitors, we need to control storage and catalytic functions in the same material. Future devices will require more than just catalysts or capacitors. They will need smart materials that can dynamically manage charge, stay stable under extreme conditions, and activate reaction intermediates only when necessary. Second, ongoing issues with capacitive interference in catalytic measurements point out a major practical challenge. Future progress should focus on: (a) advanced spectroscopies and in situ microscopy must be developed to capture how M–O bonding changes in real time under working conditions, identifying where, when, and how active species form; (b) predictive electronic-structure modeling: beyond static DFT, we need time-dependent and potential-dependent simulations to connect redox states, pseudocapacitance, and catalytic kinetics; (c) techniques such as interlayer-ion engineering, creating defects, substituting heteroatoms, tuning strain, and controlling amorphization provide effective ways to adjust charge-accumulation pathways and catalytic activation thresholds; (d) To accurately determine turnover frequency, active-site density, and kinetic metrics, we need analytical methods that distinguish faradaic responses from non-faradaic ones; (e) next-generation materials should intentionally use the tunability of the M–O bond to achieve both high storage capacity and catalytic reactivity, rather than viewing these functions as separate.

6. Conclusion

In conclusion, the M–O bond is the central concept that ties together energy storage, electrocatalysis, material stability, and device durability. By mastering this bond—its strength, electronic alignment, and dynamic behavior—we can create multifunctional materials that work efficiently across different electrochemical environments. The future lies in combining insights into electronic structure with careful material design, bridging the gap between molecular interactions and real-world energy technologies. This perspective highlights that the future of sustainable electrochemistry will be shaped not just by finding new materials but also by understanding and controlling the fundamental chemical bond that influences their behavior.

Conflicts of interest

There are no conflicts to declare.

Data availability

All the data in this MS are available from the authors upon request.

Acknowledgements

Asha K. Satheesan wishes to acknowledge University Grant Commission (UGC) for a Junior Research Fellowship (JRF). Subrata Kundu wishes to acknowledge the DST for CRG (Core Research Grant) funding of number # CRG/2021/001089 dated November 20th, 2021. CSIR-CECRI manuscript number: CECRI/PESVC/Pubs./2025-174.

References

1. R. Zhang, Z. Xu, Z. Hao, Z. Meng, X. Hao and H. Tian, Carbon Energy, 2025, 7, e630.
2. Y. Wang, X. Cheng, K. Zhang, G. Chen, R. Wang and J. Zhang, Mater. Adv., 2022, 3, 7384–7740.
3. Z. Li, M. Xu, Y. Xia, Z. Yan, J. Dai, B. Hu, H. Feng, S. Xu and X. Wang, Nat. Commun., 2025, 16, 3704.
4. J. Lu, R. Wan, S. Shi, J. Han, Y. Li, B. He, Z. Jin and G. Wang, J. Electroanal. Chem., 2025, 997, 119447.
5. C. Liang, M. Liu, J. Yu, Y. Hao and Q. Lu, Adv. Eng. Mater., 2025, 27, e202501642.
6. A. R. Kadam, K. R. Zakde, S. J. Dhoble and C. Hendrik, Swart Materials Science for Future Applications, 2025,  DOI:10.1201/.9781003491439.
7. D. Khalafallah, M. A. Ibrahim, H. Hou, J. Wang, C. Liu and Q. Zhang, Sustain. Mater. Technol., 2025, 43, e01286.
8. N. Kumar, P. Bhattacharya, A. Jana and S. Kumar, Inorg. Chem. Commun., 2025, 177, 114445.
9. R. Hu, Y. Luo, X. Li, Z. Yang and D. Chen, Electrochim. Acta, 2026, 548, 147917.
10. J. Zeng, J. Liu, S. S. Siwal, W. Yang, X. Fu and Q. Zhang, Appl. Surf. Sci., 2019, 491, 570–578.
11. G. John, A. Dennyson Savariraj, J. Archana, W. Xie and P. Justin Jesuraj, Int. J. Hydrogen Energy, 2025, 167, 150987.
12. Q. He, H. Sun, W. T. Bi, X. Y. Wang, B. Li, F. Li, Z. G. Guo, J. Ding and J. B. He, Chem. Eng. J., 2024, 495, 153461.
13. K. Ahmad and T. H. Oh, Fuel, 2026, 407, 137332.
14. A. S. Raja, R. Sasikumar, S. M. Chen and B. Kim, Small, 2025, 21, e2411728.
15. Y. Liu, Z. Liu, X. Zhang, D. Sun, G. Wang, Y. Fu and H. Ma, Ionics, 2025, 31, 12213–12226.
16. D. Zhang, Y. jie Hua, X. man Fu, C. yu Cheng, D. meng Kong, M. ling Zhang, B. X. Lei and Z. Q. Liu, Adv. Funct. Mater., 2025, 35, 2414686.
17. L. Kampermann, J. Klein, T. Wagner, A. Kotova, C. Placke-Yan, A. Yasar, L. Jacobse, S. Lasagna, C. Leppin, S. Schulz, J. Linnemann, A. Bergmann, B. R. Cuenya and G. Bacher, ACS Catal., 2025, 15, 18391–18403.
18. R. Khalid, M. Tahir, M. Umar, P. Fang and Y. Li, Carbon Energy, 2025, e70067.
19. R. Jena, V. Kashyap, R. Jana, T. Mandal, T. N. Das, F. A. Rahimi, S. Barman, D. Maity, R. Kumar, D. Bhattacharyya, A. Datta and T. K. Maji, Angew. Chem., Int. Ed., 2025, 64, e202510741.
20. U. Shamraiz, B. Raza and A. Badshah, Mater. Today Energy, 2025, 53, 102006.
21. G. Tian, H. Xu, X. Wang, X. Wen, T. Zeng, S. Liu, F. Fan, W. Xiang and C. Shu, Nano Energy, 2023, 117, 108863.
22. W. T. Hong, M. Risch, K. A. Stoerzinger, A. Grimaud, J. Suntivich and Y. Shao-Horn, Energy Environ. Sci., 2015, 8, 1404–1427.
23. M. S. Fujii and S. Portegies Zwart, Science, 2011, 334, 1380–1383.
24. P. Liu, C. Wang, C. Zeng, S. Wang, X. Yu, H. Xiao, Y. Huang, Y. Zhang, Y. Zeng, C. Shu and Z. Liang, SusMat, 2025, 5, e70007.
25. R. Madhu, A. Karmakar, K. Karthick, S. Sam Sankar, S. Kumaravel, K. Bera and S. Kundu, Inorg. Chem., 2021, 60, 15818–15829.
26. H. S. Nishad, V. Kotha, P. Sarawade, A. C. Chaskar, S. Mane, J. Lee and P. S. Walke, J. Mater. Chem. A, 2024, 12, 9494–9507.
27. B. M. Hunter, W. Hieringer, J. R. Winkler, H. B. Gray and A. M. Müller, Energy Environ. Sci., 2016, 9, 1734–1743.
28. N. Zhang and Y. Xiong, J. Phys. Chem. C, 2023, 127, 2147–2159.
29. X. Jiao, Y. Lei, Y. Liu, Z. Huang, X. Wang, Z. Liu, C. Shi, X. Jiang, C. Xing, X. Wang and A. Cabot, J. Mater. Chem. A, 2025, 13, 20974–20983.
30. G. L. Miessler, P. J. Fischer and D. A. Tarr, Inorganic Chemistry, 5th edn, 2014, ISBN-13, 978-0-321-81105-9.
31. S. Fleischmann, J. B. Mitchell, R. Wang, C. Zhan, D. E. Jiang, V. Presser and V. Augustyn, Chem. Rev., 2020, 120, 6738–6782.
32. R. Subbaraman, D. Tripkovic, K. C. Chang, D. Strmcnik, A. P. Paulikas, P. Hirunsit, M. Chan, J. Greeley, V. Stamenkovic and N. M. Markovic, Nat. Mater., 2012, 11, 550–557.
33. W. Hu, B. Hu, Z. Wu, M. Zuo, Y. Song, Q. Zheng, G. Shan and M. Du, Small Struct., 2024, 5, 2300574.
34. E. R. Kenawy, Y. I. Moharram, F. S. Abouharga and M. Elfiky, Sci. Rep., 2024, 14, 6683.
35. A. Poongan, M. Kesava, H. Zeng, S. Karthikeyan and X. Jiang, Inorg. Chem. Commun., 2025, 182, 115663.
36. M. Asim, J. Iqbal, B. Islam, J. Bashir, N. Maqsood, K. Skotnicová and A. Nawaz, J. Sci.:Adv. Mater. Devices, 2025, 10, 100939.
37. Y. Wu, X. Qing, Y. Cao, B. Geng, Y. Zou, C. Xiang, F. Xu and L. Sun, J. Energy Storage, 2025, 131, 117529.
38. K. Pandey, Y. G. Lee and H. K. Jeong, Curr. Appl. Phys., 2025, 72, 56–64.
39. C. Jagtap, V. Kadam, B. Kamble, P. E. Lokhande, A. Pakdel, D. Kumar, R. Udayabhaskar, A. Vedpathak, N. B. Chaure and H. M. Pathan, J. Energy Storage, 2024, 83, 110666.
40. S. K. Godlaveeti, A. El-marghany, R. R. Nagireddy, S. Gedi and R. Chintaparty, Ceram. Int., 2025, 51, 20620–20627.
41. J. Bai, Q. Wang, Z. Yang, H. Yu, L. Yang and J. Zhu, J. Alloys Compd., 2023, 976, 173095.
42. L. Trotochaud, S. L. Young, J. K. Ranney and S. W. Boettcher, J. Am. Chem. Soc., 2014, 136, 6744–6753.
43. M. S. Burke, M. G. Kast, L. Trotochaud, A. M. Smith and S. W. Boettcher, J. Am. Chem. Soc., 2015, 137, 3638–3648.
44. R. Madhu, A. Karmakar, P. Arunachalam, J. Muthukumar, P. Gudlur and S. Kundu, J. Mater. Chem. A, 2023, 11, 21767–21779.
45. A. Karmakar, R. Jayan, A. Das, A. Kalloorkal, M. M. Islam and S. Kundu, ACS Appl. Mater. Interfaces, 2023, 15, 26928–26938.
46. S. Tamilarasi, R. S. Kumar, A. R. Kim, H. J. Kim and D. J. Yoo, Small, 2024, 20, 2405939.
47. N. Roy, Y. Sohn, K. T. Leung and D. Pradhan, J. Phys. Chem. C, 2014, 118, 29499–29506.
48. V. Mounasamy, G. Srividhya and N. Ponpandian, Energy Adv., 2023, 2, 784–788.
49. B. He, P. Hosseini, T. Priamushko, O. Trost, E. Budiyanto, C. Bondue, J. Schulwitz, A. Kostka, H. Tüysüz, M. Muhler, S. Cherevko, K. Tschulik and T. Li, Nat. Commun., 2025, 16, 9895.
50. M. P. Browne, H. Nolan, G. S. Duesberg, P. E. Colavita and M. E. G. Lyons, ACS Catal., 2016, 6, 2408–2415.
51. Z. Jia, Y. Zhao, Q. Wang, F. Lyu, X. Tian, S. X. Liang, L. C. Zhang, J. Luan, Q. Wang, L. Sun, T. Yang and B. Shen, ACS Appl. Mater. Interfaces, 2022, 14, 10288–10297.
52. Q. Deng, H. Li, K. Pei, L. W. Wong, X. Zheng, C. S. Tsang, H. Chen, W. Shen, T. H. Ly, J. Zhao and Q. Fu, ACS Nano, 2024, 18, 33718–33728.
53. M. Khateri and M. M. Najafpour, ACS Appl. Energy Mater., 2024, 7, 5028–5037.
54. S. Pundir, S. Upadhyay, N. Kumar, N. Chandra Joshi, R. Priya, R. Ahmad Mir, I. Hossain and O. P. Pandey, Mater. Lett., 2023, 336, 133921.
55. I. Hod, P. Deria, W. Bury, J. E. Mondloch, C. W. Kung, M. So, M. D. Sampson, A. W. Peters, C. P. Kubiak, O. K. Farha and J. T. Hupp, Nat. Commun., 2015, 6, 8304.
56. M. Jiang, S. Li, Y. Zhang, X. Wang, J. Zeng, X. Wu, L. Yao, Q. Zhu, Z. Shao and X. Wu, ACS Sustain. Chem. Eng., 2024, 12, 205–215.
57. J. Zhang, G. Chen, D. Sun, Y. Tang, W. Xing, H. Sun and X. Feng, Chem. Sci., 2024, 15, 17900–17911.

---

Footnote

† These authors have contributed equally.

---