Ideal design of air electrode—A step closer toward robust rechargeable Zn–air battery

Although intensive research on electrically rechargeable Zn–air batteries (ZABs) began in early 1960’s, a commercial breakthrough has not been accomplished yet due to the inferior overall energy efficiency and poor cyclability, which is correlated with the difficulties present in optimizing the cathode and anode components.^1–4 Despite the slow progress, ZABs have always been considered as a highly promising battery system for electric vehicles and portable applications, since their theoretical gravimetric and volumetric energy densities (⁠$500 W h kgcell−1$ and $1400 W h lcell−1$⁠) are higher than those of the commercialized lithium-ion battery (LIB) (350 W h $kgcell−1$ and 810 W h $lcell−1$⁠), as shown in Fig. 1(a).^5,6 Additionally, Zn metal is earth-abundant and highly resistant to corrosion in alkaline solutions, giving ZABs outstanding advantages such as low capital cost, safe fabrication process, flat discharge voltage, and long shelf-life.^7,8 ZABs acquire a unique half-open system with cathode utilizing oxygen from ambient air as the active material, which decreases the total mass and the volume of the cathode.^9–11 As shown in Fig. 2(a), a typical Zn–air battery (ZAB) is composed of four main components: a Zn anode (Zn metal, Zn paste, or Zn powder), an air-breathing cathode, an alkaline electrolyte, and a separator. Potassium hydroxide (KOH), typically about 6M, is the most used alkaline electrolyte for ZABs because of its high conductivity and activity for both Zn and air electrodes. To suppress the dendritic growth and shape change of the Zn anode, zincate solubility-reducing agents such as ZnCl₂, Zn(ac)₂, and ZnSO₄ are usually added in the alkaline electrolyte.¹² A microporous separator is often placed between the anode and the cathode to prevent the physical contact of electrodes and to suppress the passivation and dendrite propagation of the Zn electrode. It also serves as the electrolyte reservoir to assist the ionic transport. Note that the separator does not involve directly in any cell reactions, but its structure and properties play a significant role in determining the battery performance, including cycle life, safety, energy density, and power density, by influencing the cell kinetics.¹³ More importantly, the air electrode is the most significant and complicated part of a Zn–air battery, which allows the diffusion of O₂ from the air into the electrode and converts it to OH⁻.^14,15 It is usually made up of three parts: a catalyst layer facing the liquid electrolyte, a gas diffusion layer (GDL) exposed to the external environment, and a conductive current collector, which is either located in the middle of the other two layers or combined together with the GDL.^7,16,17 Detailed discussion about the components of the air electrode is presented in Sec. II.

More importantly, the conversion of O₂ to OH⁻ in air electrode includes the oxygen reduction reaction (ORR) and the oxygen evolution reaction (OER) during the charge and discharge processes, respectively; both reactions require high overpotentials due to sluggish kinetics related to four-electron processes and complicated mechanisms. While the overall energy efficiency and power output of ZABs are still inferior to commercialized LIBs, one critical issue that limits the practical application of Zn–air batteries is the large voltage gap (ΔE) between the charge and discharge processes, which is mainly attributed to the high overpotentials of oxygen reactions at the air electrode.^9,14,18,19 Thus, the cathode reactions (ORR and OER) are the rate-limiting steps for ZABs reactions. To overcome this problem, significant efforts have been dedicated to the development of advanced bifunctional catalysts, which can efficiently catalyze both ORR and OER.^20–25 The bottleneck for the development of these catalysts lies in that ORR and OER have different requirements for the catalytically active sites. Furthermore, even though several ways have been proposed to efficiently combine together the best available ORR and OER catalysts with exposed active sites for both oxygen reactions, most of them cannot withstand the alternating reductive–oxidative environments during discharge–charge cycles.^26–28 Thus, improving the design of bifunctional O₂ electrocatalysts is still the main focus of many researchers working on the innovation of rechargeable ZABs.^29–32 

Herein, we strongly emphasize that all the components in the air electrode play unique roles that are equally important. Therefore, to develop a robust structural design that could provide accessible active sites to oxygen, electrolyte, and electron, it is rational to modify the overall structure of the air electrode, which includes the O₂ catalyst, GDL, and conductive substrate. This minireview intends to present the recent developments in Zn–air batteries, in view of the two important design factors: (1) overall structural features of the air electrode and (2) the electronic structure of the bifunctional O₂ electrocatalyst. The first part will carefully discuss the individual components of the air electrode and the related oxygen electrochemical reactions within the battery system. We hope to emphasize the structural features required to satisfy the design of an ideal cathode compartment of ZABs, which includes but not limited to the hierarchical pore structure, wetting properties, and electrical conductivity. Moreover, because it is not easy to identify the catalytic center that drives the ORR/OER performance, we will present the associated factors of oxygen reactions based on the four major modification strategies of air catalysts, namely, (1) defect engineering, (2) cation/anion regulation in multi-component metal compounds, (3) single or multi-heteroatom doping in carbon materials, and (4) fabrication of binder-free air electrodes [Fig. 1(b)]. Thus, the second part of this review will describe the relationship between the intrinsic activity of bifunctional O₂ catalysts and the collective strategies employed to modify the electronic structure of ORR/OER catalysts. In the last part, we will review the present status of the rechargeable Zn–air battery, which utilizes the recent bifunctional O₂ catalysts in an integrated air electrode structure. The performance of thus far reported O₂ electrocatalysts will be evaluated by summarizing the bifunctional catalytic activity with respect to OER and ORR and the Zn–air battery performance.

In general, an air electrode with a structural design that could provide proficient mass and charge transport is exceedingly important since it is the dominating factor in developing a rechargeable Zn–air battery with high roundtrip efficiency and high-power output. The main role of an air electrode is to provide reaction sites for both ORR and OER, wherein the oxygen reduction reaction takes place at the gas–liquid–solid triple-phase region during the discharge process. Therefore, the air electrode is required to possess a highly porous structure to facilitate the diffusion of air and liquid and an optimal wettability property with well-balanced hydrophobicity and hydrophilicity, which can help slow down the evaporation loss of the electrolyte and avoid the adverse effect of the flooding of the electrode; both properties are important for improving the rate capability and specific capacity of ZABs.^7,11,23

An air electrode is usually made up of three parts: a catalyst layer loaded with oxygen electrocatalysts, a gas diffusion layer (GDL), and a conductive current collector. It is fabricated in such a way that the catalyst layer and GDL are facing the internal liquid electrolytes and external environment, respectively, while the highly conductive current collector either sits in the middle of the two components or is integrated together with GDL.^17,28 It is vital to emphasize that every component in an air electrode plays individual roles, which are equally important in the overall improvement of the ZAB performance. For instance, in view of the mass and charge transport, the catalyst layer should own a hierarchical pore structure with abundant macropores and mesopores functioning as the gas transport channel and reaction regions for the gas–liquid–solid triple-phase oxygen reactions, respectively. In addition, the loaded O₂ electrocatalyst is expected to deliver high electrical conductivity to expedite the electron transfer inside the catalyst, and the catalyst surface is preferable to be hydrophilic so that the active sites on its surface will be well-exposed to the reactants. More importantly, the O₂ electrocatalyst should have dual coupled active sites that are capable of catalyzing both ORR and OER or a single bifunctional active site that can facilitate both of the oxygen reactions. Figure 2(b) shows the typical illustration of the required structural properties of the O₂ electrocatalyst.^10,20

Meanwhile, in order to maximize the oxygen gas permeability, air electrode is either spray-coated or directly deposited with a GDL. The primary functions of the GDL are to ensure efficient oxygen diffusion between the electrolyte and ambient air and to prevent any electrolyte flooding and leakage.^26,31 Thus, the wettability and pore structure of GDL are needed to be controlled in a way that the side in contact with the electrolyte is slightly hydrophobic with smaller pore sizes, while the side exposed to the external environment is highly hydrophobic with larger pore sizes to allow gas to diffuse efficiently into the system. The hydrophobicity of GDL can be realized by impregnating it with superhydrophobic agents such as polytetrafluoroethylene (PTFE) and high calcination temperature.³³ Note that the PTFE-membrane-like coating not only functions as a wet-proofing agent but also controls the permeation of CO₂ into the electrolyte. Alkaline electrolytes such as KOH could produce carbonate species when it reacts with CO₂, which may potentially block the GDL pores, thus restricting the supply of O₂ in the air electrode.³¹ Moreover, the GDL is typically thin but with highly effective surface area and high porosity, which is beneficial for the gas diffusion. In addition, it should be not only electrically conductive to assist charge transfer but also electrochemically durable against rapid oxidation and strong alkaline electrolytes to protect the electrode against corrosion. Notably, the lifespan of the air electrode mainly depends on the preservation of the desired wettability properties and porosity of GDL; however, controlling the layer thickness and surface porosity while maintaining the suitable wettability properties is the common difficulty in developing efficient GDLs.^14,23 On the other hand, current collectors such as nickel foam, copper foam, stainless-steel (SS) mesh, Ti mesh, and Ni mesh, which typically own high electrical conductivities and electrochemical oxidation resistance, do not only function as physical and conductive support for the catalyst layer but also provide a highway for fast and efficient oxygen diffusion during the reactions. Unlike with carbon-based current collectors that are prone to fast oxidation and deterioration, metallic substrates exhibit excellent mechanical and chemical stability, remarkable electrical conductivity, and 3D porous structure, which may lead to better performance of the catalyst layer and GDL during a long-term cycling test.^7,26

Recently, strong interactions between the catalyst particles and the conductive substrate to reduce the interface resistance have been recognized as a major target in improving the performance of ZAB.^11,20,27 The fabrication of the binder-free air electrode through direct growth of metal precursors onto the surface of the conductive collector has been considered favorably strategical because it could not only eliminate the use of any polymer binder that hinders some of the electroactive sites and improve the mechanical stability but also reduce the final weight of the electrode for about 10%–40%.^34–36 Thus, various methods have been utilized to prepare a binder-free air electrode comprising a bifunctional O₂ catalyst, a GDL and a conductive substrate, which will be further discussed in Sec. IV.

Furthermore, the typical parameters to evaluate the performance of ZABs include the round-trip efficiency, voltage gap, durability, and energy/power density. The round-trip efficiency represents the energy-utilizing efficiency for the ZABs and can be derived by dividing the voltage of the discharge and recharge process at a constant current density. The low roundtrip efficiency is one of the major challenges in ZAB; hence, it can be boosted by reducing the overpotentials of both OER and ORR. Moreover, the voltage gap could be obtained by measuring the difference between the charging and discharging curves at different current densities, usually at 5 mA cm⁻² or 10 mA cm⁻². While the changes in voltage during the cycles could be used to evaluate the stability of the ZAB, the lower voltage gap value is directly proportional to the overpotential of the oxygen reactions for ZABs. On the other hand, to characterize the bifunctional catalytic activity of the electrocatalyst, the potential difference (ΔE) between the potential for achieving a current density of 10 mA cm⁻² during the OER (E₁₀) and the half-wave potential (E_1/2) during the ORR could be calculated, which is illustrated in Fig. 2(c).^18,19

Non-precious transition metal compounds (spinel and perovskite oxides, layered hydroxides, nitrides, sulfides, and phosphides) probably represent the most interesting class of O₂ catalysts.^38,39 They possess advantages such as earth-abundant, low-cost, easy preparation, and environmental friendliness. Transition metal compounds and their composites have attracted so much attention as highly potential candidates to develop competent bifunctional ORR and OER electrocatalysts toward high-performance metal–air batteries such as the rechargeable ZAB.

Transition metal oxides (TMOs) can be categorized as single-metal oxides, spinel oxides, and perovskites-type oxides. Single manganese oxides (MnO, Mn₃O₄, Mn₂O₃, and MnO₂) have exhibited remarkable electrochemical performance due to its various properties, including a wide variety of atomic structures, phases, morphology, and porosity. Similarly, noble-free metal oxides such as Fe₂O₃, Fe₃O₄, Co₃O₄, and NiO have been receiving great research interest and are extensively investigated due to their incompletely filled d-orbitals, which can provide numerous possible oxidation states and lead to the high intrinsic electrochemical activity, including superior rate capability, high energy density, long-term life cycle, and high safety.^26,37 On the other hand, TMOs with spinel structure are represented by the formula A_xB_3−xO₄, where divalent A²⁺ (i.e., Co, Ni, Fe, and Mn) and trivalent B³⁺ (i.e., Al, Fe, and Co) cations could occupy part or all of the tetrahedral and octahedral sites, respectively. They have shown exceptional performances in oxygen evolution and reduction in alkaline solutions owing to their remarkable electrical conductivity. The electron transfer in this structure takes place with quite low activation energies between the cations of different valences via hopping processes.¹⁵ Perovskite-type oxides have been widely investigated as the bifunctional ORR/OER catalyst in alkaline electrolytes due to their electrocatalytic activities and corrosion resistance. They own a chemical structure of AA′BB′O3, where A is a rare-earth metal ion (La, Ce, Pr, and Hf) and A′ is an alkaline-earth metal (i.e., Ca, Sr, and Ba), which induce the capability of oxygen adsorption, and B and B′ are transition metal ions (Co, Mn, Fe, Ni, and Cr), which influence the activity of the adsorbed oxygen.^28,40 Through partial substitution of A or B-site cations, the properties of perovskite oxides can be easily adjusted, contributing to different structures, defects, and catalytic properties.^41–43 Overall, the catalytic activity of metal oxide O₂ catalysts can be associated with the ability of the cations to adopt different valence states, especially when they form redox couples at the potential of ORR/OER.⁴⁴ 

Layered double hydroxides (LDHs), also called anionic or hydrotalcite-like clays, are made up of layers of divalent and trivalent metal cations coordinated to hydroxide anions and incorporated with guest anions (typically $CO32−$⁠) and water between the large interlayer spaces, [$M1−x2+Mx3+$(OH)₂][Aⁿ⁻]_x/n− zH₂O. The composition of LDHs can be easily modified to enhance catalytic activity; however, the full potential of this material as bifunctional O₂ catalysts is yet to be realized since most of the LDH-based materials were investigated for OER only.^17,43,45 Meanwhile, metal sulfides, nitrides, and phosphides such as CoS_x and NiS_x, Co₄N, and CoP, respectively, have demonstrated better electrocatalytic activities for both ORR and OER when compared to their counterpart metal oxides due to their richer redox sites and higher electrical conductivity, which are beneficial for charge transfer during electrochemical reactions.^46–50 

Introduction of dopants such as nitrogen, sulfur, phosphorus, and/or boron into graphitic carbon materials can change the sp² hybrid structure of carbon atoms because the size and electronegativity of these elements are different. This method can modify the electron distribution and create surface defects that will gravely affect the catalytic activity of carbon materials. By doping carbon with more electronegative heteroatoms, a net positive charge is created on surrounding carbon atoms to improve the chemisorption of oxygen and electron transfer. Thus, the integration of heteroatoms into the carbon structure constitutes a significant molecular functionalization advance, which would control the physicochemical and electronic properties for specific applications such as ORR and OER.⁵¹ Among the various dopants, nitrogen is the most explored element due to its similar size with the carbon atom. There are two main methods to prepare nitrogen-doped carbon materials: (1) via post-synthesis treatment of un-doped porous carbon materials under the N-containing atmosphere and (2) direct synthesis of the N-containing carbon material by using N-containing precursors.¹⁰ These methods are widely used to introduce catalytic active N-doping sites to carbon materials and to create pores with a well-defined size and shape, which is a very crucial requirement of the bifunctional O₂ catalyst. For the first method, pore structures and composition of N-functionalities may vary depending on the type of reagents and treatment conditions applied, which might induce significant changes in the electronic state of the carbon surface and catalytic performance. The most commonly used reagents include ammonia and a series of related compounds such as urea, melamine, dicyanodiamide, dimethylformamide (DMF), and ethylenediamine.⁵² The latter method offers one-step synthesis to prepare heteroatom-doped carbon materials, which includes direct carbonization of N-containing carbon precursors such as molecules containing cross-linkable groups (—CN and —NH₂), N-rich polymers (polypyrrole, polyaniline, and polyacrylonitrile), ionic liquids, or biomass derivatives. This method ensures the uniform incorporation of N-functionalities throughout the carbon structures, which results in more stable and homogeneous properties of N-doped carbon materials that are critical to the catalytic reactions.⁵³ 

On the other hand, dual or multiple heteroatom doping of carbon materials could modulate the electronic properties and surface polarities and eventually increase the electrocatalytic activity of nanostructured carbon materials. Recent Density Functional Theory (DFT) studies revealed that due to possible synergistic effects, reactant adsorption could be facilitated due to more active sites than single doped samples, and thus, multiple heteroatom doped-carbon materials attain much enhanced catalytic performance.^54,55 Doping N-doped carbon nanomaterials with secondary heteroatoms with lower electronegativity such as B, S, and P is an effective way to further optimize the electronic and chemical features of the carbon materials. In addition, this method can reduce the tendency of the carbon catalyst to oxidize under a high anodic potential during OER, which is important in achieving high electrocatalytic bifunctionality.^56–58 

Various carbon-based metal-free materials, including graphene, carbon nanotubes (CNT), and porous carbon, have been synthesized and utilized as efficient bifunctional oxygen catalysts in metal–air batteries because of their advantages such as high electrical conductivity, diverse structure, rich defect chemistry, high surface area, and low cost.^17,20 These materials have been proven to facilitate the electron transfer and mass diffusion for ORR/OER. Their electronic structures can be readily regulated with heteroatom-doping, rational defect engineering, or binding with other species, leading to improved catalytic performance.⁵ However, it is noteworthy that carbon-supported catalysts may suffer from relatively low lifetimes due to electrochemical corrosion (oxidation) of the carbon, particularly at highly oxidative potentials of OER catalysis.¹⁰ Carbon nanostructures can be further categorized into three groups: one-dimensional (1D) nanotubes and nanofibers, 2D graphite and graphene nanosheets, and 3D nanoporous architectures.¹⁵ 

The unique hollow geometry consisting of the conjugated all-carbon structure, along with important properties such as high surface area, remarkable electrical, mechanical, and chemical stability, has made CNTs as ideal electrocatalysts toward various electrochemical applications.⁵⁹ The introduction of nitrogen in the structure of CNT could induce changes in physicochemical properties, including an increase in surface defects, active sites, surface area, electrical conductivity, and electrochemical activity.

Intrinsic properties of graphene-based materials such as large theoretical specific surface area (∼2630 m² g⁻¹), high electrical conductivity, excellent mechanical flexibility, remarkable thermal conductivity, and porous structure make them a highly suited candidate as O₂ catalysts for fuel cells and metal–air batteries applications. By the introduction of heteroatom dopants such as boron, nitrogen, phosphorous, sulfur, halogen, or metal atoms into the structure of 2D-ultrathin graphene sheets, the electrochemical stability of graphene could be further enhanced to exhibit improved catalytic activity toward ORR and OER.^29,60

Nitrogen-doped porous carbon (NPC) materials have been reported to exhibit electrocatalytic activities and good durability in both acidic and alkaline electrolytes. Nitrogen doping of porous carbon has several advantages, including (1) changing the electronic structures while minimizing the lattice mismatch in porous carbon, (2) enhancing the hydrophilicity of the surface, which improves catalyst dispersion in aqueous media, and (3) improving the interaction between carbon surfaces and reaction intermediates. While the adjacent carbon atoms modulated by the nitrogen atoms are believed to be the active sites in N-doped porous carbon materials, both pyridinic and graphitic nitrogen atoms were identified as the catalytic center of this bifunctional O₂ catalyst.^17,20,61,62

The configurations of heteroatom-doped carbon vary depending on the intrinsic features of each dopant, which could lead to different catalytic properties. For example, the P atom, being larger than C and N, always moves out of the planar surface in the graphitic framework and occupies the edge sites. Doping carbon with P can improve the electron-donor properties of carbons along with the enhanced catalytic activity. However, the chemical bonding state of P in carbon materials remains unclear since P is easily oxidized into P—O bond, which is usually unavoidable during the heteroatom doping process, and hence becomes inactive toward ORR/OER.^63–65 On the other hand, boron has a lower electronegativity than carbon so that the positively charged B atoms could attract the negatively charged O₂ atoms, leading to chemisorption. Boron doping provides boron sites that could act as electron donors and improve the graphitization level and durability of carbon materials. Meanwhile, S-doping only leads to negligible polarization due to the similar electronegativity values of the S and C atoms. The mismatch of the outermost orbitals of the S and C atoms could induce a non-uniform distribution of spin density, endowing graphene with catalytic abilities.

In general, transition metal compounds are accountable for OER activity, while heteroatom doped carbon materials are highly efficient for ORR. While the major drawback of metal compounds as O₂ electrocatalysts is their poor electric conductivity, carbon materials suffer from low lifetimes due to electrochemical oxidation. Thus, it is strategically valid to combine these two materials to form a hybrid catalyst that will have the ability to synergistically enhance both ORR activity and OER activity.^51,66,67 One of the most studied non-noble O₂ electrocatalysts is the composite composed of a transition metal and nitrogen co-doped carbon materials, denoted as MNC (M = transition metal, N = nitrogen, and C = carbon) materials. In this hybrid form, the synergy between N-doped carbon and transition metal is often observed, resulting in higher ORR and OER activities compared with the activities of their individual components. Such synergetic effects observed with this hybrid catalyst are attributed to the facts that (1) the electronic interaction between carbon and metal species could have created rapid electron transfer paths and that (2) the active carbon species initially facilitate the reduction of O₂ to HO²⁻ and, then, the metal species subsequently catalyze HO²⁻ to OH⁻, which eventually results in the overall four-electron reduction process.²⁹ 

Until now, Pt-based and Ir/Ru-based compounds have been the benchmark catalysts for ORR and OER, respectively. However, due to the high cost, low stability, and scarcity of these materials, many researchers have devoted tremendous efforts to find alternative materials suitable for ORR, OER, or ORR/OER catalysts. Most explored materials that showed potential bifunctional catalytic activity could be categorized into two main groups, namely, (1) transition metal compounds (spinel and perovskite oxides, layered hydroxides, nitrides, sulfides, and phosphides) and (2) carbon-based materials (carbon nanotube, graphene, metal-free or metal-based heteroatom doped-carbon, and g-C₃N₄). Various strategies have been employed to improve the electrocatalytic properties and bifunctional performance of these ORR/OER electrocatalysts.^25,68,69 However, controversial results concerning the origin of catalytic activity that drives the ORR/OER performance still exist, which inhibits the progress of the catalyst developments toward the Zn–air battery.

This section will discuss the most recent bifunctional O₂ electrocatalysts that showed promising performances toward ORR/OER and extended their function as air catalyst for the Zn–air battery. Subsections IV A–IV C will present various bifunctional electrocatalysts based on major modification strategies of O₂ catalysts, namely, (1) defect engineering, (2) cation/anion regulation in multi-components transition metal compounds, and (3) single or multi-heteroatom doping in carbon materials (metal-free and metal-based materials). The last part of this section will discuss the recent developments in binder-free air electrodes utilized for rechargeable ZAB, where highly active bifunctional O₂ catalysts (possessing gas diffusion layer properties) are integrated with a conductive substrate.

The performance of the electrocatalysts is greatly affected by the adsorbate binding energy of reaction intermediates.^70–72 In this context, regulation of the surface adsorbate binding energy of oxygen intermediates during oxygen catalysis is very important in achieving highly active and stable OER/ORR bifunctional catalysts.^73,74 In particular, defect structures in the electrocatalysts present multiple advantages in oxygen electrocatalysis, including (1) optimization of the adsorption energy of reaction intermediates, (2) improvement in charge transfer ability, and (3) an increase in the number of low-coordinated sites, which could play as active sites.^27,75 Moreover, for oxygen electrocatalysis, oxygen defects in the catalyst surface play a vital role in improving the performance of both OER and ORR.⁷⁶ Therefore, many studies that focused on the fabrication of highly active air electrode catalysts have adopted the strategy to generate defect structures, such as atomic vacancies, hetero-interfaces, grain boundaries, and adatom.^29,64,77

The non-metal counterparts in transition metal compounds, such as sulfur and selenium, could assist the generation and modulation of oxygen defects in the catalyst.^78–82 For example, Yin et al. reported the synthesis of hetero-interfaced NiS₂/CoS₂ porous nanowires with dominant oxygen vacancies (NiS₂/CoS₂–O NWs) for O₂ electrocatalysts for ZAB.⁸³ Initially, NiCo₂O₄ nanowires (NWs) were calcined with sulfur powder to prepare NiS₂/CoS₂ NWs. The as-prepared NWs were then submitted to electrochemical oxidation in 1.0M KOH electrolyte at 1.6 V vs RHE to induce the surface oxide on the NiS₂/CoS₂ and to obtain NiS₂/CoS₂—O NWs catalysts. The domination of oxygen vacancies in the NiS₂/CoS₂—O NWs was confirmed by the XPS and ESR spectra of NiS₂/CoS₂—O NWs, wherein results revealed the significant sulfur loss on the surface and the formation of abundant oxygen vacancies during the electrochemical oxidation after the electrochemical treatment. Benefited by abundant oxygen vacancies, the NiS₂/CoS₂—O NWs exhibited an overpotential of 235 mV to drive a current density of 10 mA cm⁻² smaller than NiS₂/CoS₂ NWs (320 mV), NiCo₂O₄ NWs (360 mV), and commercial Ir/C (300 mV). Based on the DFT calculation results, the active sites for OER were revealed to be oxygen vacancies on NiS₂/CoS₂—O. Moreover, the ORR half-wave potential of the NiS₂/CoS₂—O NWs catalysts exhibited 0.70 V, which is close to that of commercial Pt/C. The rechargeable Zn–air battery using the NiS₂/CoS₂—O NWs as an air electrode exhibited an open-circuit voltage of 1.49 V for more than 10 h. In addition, the recharging of the fabricated ZAB showed a stable performance of more than 30 h at the current density of 5 mA cm⁻². Likewise, Zheng et al. reported the activity boost of bifunctional OER/ORR catalysts by engineering oxygen vacancies in nickel–cobalt selenide catalysts.⁷⁸ The surface of the nickel–cobalt selenide nanocrystals (Ni_0.6Co_0.4Se₂), prepared by the polyol method, was transformed to amorphous oxides by in situ electrochemical oxidation. The formation of oxygen vacancies was assisted by the dissolution of Se elements, as confirmed by soft XAS and XPS measurements (Ni_0.6Co_0.4Se₂—O). Benefited by the oxygen vacancies, the Ni_0.6Co_0.4Se₂—O catalysts showed a much smaller OER overpotential (285 mV) to achieve 10 mA cm⁻² than those of Ni_0.6Co_0.4Se₂ (363 mV), NiSe₂ (418 mV), CoSe₂ (462 mV), and commercial Ir/C (384 mV) catalysts. This excellent OER performance largely originated from the increased number of active sites and the intrinsic metallic conductivity from oxygen vacancies in the Ni_0.6Co_0.4Se₂—O catalyst. Moreover, the Ni_0.6Co_0.4Se₂—O catalyst showed better ORR performances than other catalysts, whose onset potential was only 40 mV larger than that of the commercial Pt/C catalyst. Motivated by the excellent bifunctional activity of Ni_0.6Co_0.4Se₂—O for OER and ORR, the catalyst was applied as the O₂ electrode catalyst for flexible ZAB to show the voltage gap of 0.82 V with corresponding energy density 944 W h kg Zn⁻¹ at 10 mA cm⁻² current density and the capacity to support a timer for more than 30 h.

Moreover, anion vacancies in the catalyst can also lead to excellent performances of the air electrode catalyst. For example, Li et al. reported CuCo₂S₄ nanosheets (NSs) with sulfur vacancies as an OER/ORR bifunctional electrocatalysts, whose thickness is 4–6 atomic layers.⁷⁹ The CuCo₂S₄ NSs were prepared by exfoliating bulk CuCo₂S₄, which was synthesized by the hydrothermal method, through ultrasonication. Contrary to bulk CuCo₂S₄, CuCo₂S₄ NSs showed disordered surface and coordination deficiency of Cu, revealed by EXAFS spectra, which originated from S vacancies in CuCo₂S₄ NSs. The CuCo₂S₄ NSs exhibited an overpotential of 287 mV to drive an OER current density of 10 mA cm⁻², which is lower than those of bulk CuCo₂S₄ (327 mV) and commercial Ir/C (310 mV). This improved OER activity is correlated with the optimized binding energy of surface oxygen intermediates, assisted by in situ formed oxygen vacancies inherited from the sulfur vacancies of the CuCo₂S₄ NSs. Additionally, the ORR onset potential of CuCo₂S₄ NSs showed 0.90 V, which is comparable to that of commercial Pt/C (0.94 V) and much superior to those of bulk CuCo₂S₄ (0.86 V) and commercial Ir/C (0.68 V). By Koutecky–Levich (K–L) plots of the catalysts, CuCo₂S₄ NSs exhibited a higher electron transfer number of 3.9 compared to bulk CuCo₂S₄ (2.9), which originated from the metallic character of CuCo₂S₄ NSs. With excellent bifunctional performance for OER and ORR, the CuCo₂S₄ NSs were applied as air electrode catalysts for flexible, all-solid-state ZAB. The fabricated flexible ZAB exhibited an open-circuit potential of 1.20 V and the energy density of 424 W h kg⁻¹ at 1.0 mA cm⁻², along with the rechargeable capacity for more than 18 h.

On the other hand, the oxygen defects in the transition metal oxide catalysts could also be generated during the formation of carbonaceous materials by pyrolysis. For instance, Jiang et al. reported the beneficial effect of defect-rich interfaces between metallic Co and Co₃O₄ in porous graphitized shells (PGS) toward air electrode catalysts for ZAB.⁸⁴ A hybrid catalyst composed of Janus Co/Co₃O₄ nanoparticles in PGS (Co/Co₃O₄@PGS) was prepared by pyrolysis of Co²⁺-exchanged Zn-based metal–organic frameworks (MOFs). The Co 2p and O 1s XPS analysis of the Co/Co₃O₄@PGS catalysts, as shown in Figs. 3(a) and 3(b), indicated the higher number of oxygen defects in the catalyst than Co₃O₄@PGS, showing that the oxygen defects are largely produced at Co/Co₃O₄ interfaces in the catalyst. Due to the oxygen defects, the electrical conductivity of Co/Co₃O₄@PGS (8.41 S cm⁻¹) was much higher than those of Co@PGS (7.2 S cm⁻¹) and Co₃O₄@PGS (3.05 S cm⁻¹), suggesting the fast and efficient charge transfer in the Co/Co₃O₄@PGS catalyst.⁸⁵ The Co/Co₃O₄@PGS catalyst exhibited 350 mV overpotential to drive an OER current density of 10 mA cm⁻², which is smaller than that of commercial Ir/C (410 mV). Moreover, Co/Co₃O₄@PGS showed excellent ORR activity, whose half-wave potential (0.89 V) was 15 mV higher than that of commercial Pt/C. With the smallest OER/ORR voltage gap (0.69 V), the Co/Co₃O₄@PGS catalyst was applied to the air electrode of ZAB. Co/Co₃O₄@PGS demonstrated a higher power density of 118.27 mW cm⁻² than the reference (90.60 mW cm⁻²) based on the state-of-the-art Pt/C and Ir/C mixture. The Co/Co₃O₄@PGS catalyst also exhibited excellent stability with little voltage fading (0.91–0.96 V) over 800 h at a current density of 10 mA cm⁻². The comparison of electrochemical performances of the modified catalyst against others are presented in Figs. 3(c) and 3(d). Chen et al. also reported the bifunctional oxygen catalysis of hollow CoO with abundant oxygen vacancies embedded in N, S-co-doped porous carbon.⁵² The hollow CoO nanoparticles in N and S co-doped mesoporous carbon (CoO-NSC-900) catalysts were synthesized by the pyrolysis of PEI/Co²⁺/lignosulfonate composites under the N₂ atmosphere at 900 °C. The oxygen deficiency in the cobalt oxide phase in CoO-NSC-900 was confirmed by O 1s and Co 2p XPS spectra, which was produced by the reducing ability of carbonaceous materials.⁸⁶ Benefited from the oxygen vacancies in the catalyst, the CoO-NSC-900 catalyst exhibited lower OER overpotential (470 mV) to drive a current density of 10 mA cm⁻² than RuO₂ (570 mV) catalyst. The CoO-NSC-900 catalyst also exhibited a ORR half-wave potential of 0.83 V, which is comparable to that of commercial Pt/C (0.83 V). With a low OER/ORR voltage gap of 0.86 V, the CoO-NSC-900 catalysts were applied to the air electrode of ZAB, which exhibited energy densities of 871 W h kg Zn⁻¹ at 10 mA cm⁻², along with the high stability for more than 60 h.

Inspired by the high reactivity of spinel-structured Co—Mn—O nanocrystals toward oxygen reactions, due to the high surface area, numerous defects, and abundant vacancies, Du et al. anticipated that the composite material of Co₃O₄ nanoparticles uniformly coated on the surface of the MnO₂ nanotubes could lead to the improvement of electrocatalytic activity for both ORR and OER due to the expected synergistic effect and the interface effect between the two metal oxides.⁸⁷ As expected, Co₃O₄ nanoparticles provided high electrical conductivity and active sites for OER activity, while MnO₂ nanotubes improved the charge and mass transport of the solvated ions and upgraded the performance of ORR.⁸⁸ However, the modified material did not show impressive results as O₂ catalyst for the air electrode of the Zn–air battery, which lasted for only 60 cycles. To fulfill the requirements of long-cycle ZAB, Guo et al. designed a core–shell nanostructure composed of Co₃O₄ nanowire arrays as the core element and ultrathin NiFe-layered double hydroxides (NiFe LDHs) as the shell. Schematic illustration of the synthetic process of the Co₃O₄@NiFe LDH hybrid nanowire arrays on Ni foam and the carbon cloth electrode is shown in Fig. 4(a).⁸⁹ TEM analysis confirmed the successful formation of the multicomponent core–shell material, as shown in Figs. 4(b) and 4(c), which demonstrated synergistic effects and electronic structure modulation wherein electron transfer from higher valence states Ni and Fe ions to lower valence state Co species had created additional active sites and accelerated the oxygen reactions. The modulation of the electronic structure was dependent on the growth time of NiFe LDH shells to achieve balanced surface chemistry for OER and ORR. The bifunctional Co₃O₄@NiFe displayed a lower potential difference (ΔE) value of 0.78 V, which is superior to those of Pt/C (0.81 V) and Ir/C (0.79) catalysts. More importantly, the fabricated ZAB maintained a low voltage gap of 0.8 V for over 1200 cycles (>200 h) and obtained a round-trip efficiency of 60%, as displayed in Fig. 4(d). Similarly, Majee et al. designed a core–shell structured composite by encapsulating Ba_0.5Sr_0.5Co_0.79Fe_0.21O_3−δ (BSCF) nanoparticles with two dimensional NiFe LDH sheets to overcome the surface instability of the perovskite oxide-based OER catalyst and to promote the bifunctional activity of the BSCF/NiFe-LDH-based catalyst for OER and ORR. In situ growth of NiFe-LDH over BSCF effectively increased the electrochemically active surface area (ECSA), promoted crystalline boundary for electron conduction pathways, and improved the structural stability of BSCF during oxygen reactions, which is highly desirable for long-term rechargeable ZAB performance.

Furthermore, incorporation of Ni atoms into the octahedral sites of the spinel crystal structure of Co₃O₄ to form ternary NiCo₂O₄ could improve the electrochemical activity of this highly active and corrosion-resistant O₂ catalyst.^90,91 Combination of the formed ternary metal oxides with carbon-based materials such as carbon nanotubes (CNT), graphene sheet, or porous carbon could further improve the electrochemical properties of Co₃O₄. Lee et al. prepared a NiCo₂O₄–graphene hybrid material via poly(vinyl pyrrolidone) (PVP)-assisted one-pot synthesis process to demonstrate bifunctional catalytic activity toward ORR and OER.⁹² The incorporation of Ni cations into the octahedral sites and the presence of PVP during synthesis as a capping agent created additional active sites with a lower energy barrier on the surface of Co₃O₄, which boosted the electrocatalytic activity for both OER and ORR. Additionally, the graphene sheets provided high electrical conductivity, which facilitated the charge transfer during the oxygen reaction at the surface of the catalyst. Furthermore, in situ nitration of NiCo₂O₄ nanowires (NWs) on a carbon paper was conducted by Yin et al. to synthesize NiO/CoN with porous interface NW arrays.⁹³ Heat treatment of the metal oxides under the flowing NH₃ atmosphere was effective in creating oxygen vacancies, porous structure, and robust nanointerface between NiO and CoN domains, all of which were factors responsible for the enhancement of the electrocatalytic activity of OER and ORR. Considering the excellent bifunctional catalytic activity of NiO/CoN for OER and ORR, which showed a ΔE of 0.85 V, the group extended the application of the modified material as O₂ catalyst for the renewable Zn–air battery. The low voltage gap of 0.84 V with a corresponding power density (79.6 mW cm⁻² at 200 mA cm⁻²), energy density (945 W h kg⁻¹ at 10 mA cm⁻²), and the capacity to power up a clock device for more than 12 h has indicated the remarkable performance of ZAB with the air electrode composed of NiO/CoN as the O₂ catalyst.

Transition metal nitrides such as metallic Co₄N catalysts also displayed superior OER activity to that of Co-based oxides due to its enhanced electrical conductivity. However, its ORR activity still showed unsatisfactory results because of the limited catalytic activity of the Co-based material toward oxygen reduction. Therefore, it is necessary to combine Co₄N with ORR active materials to promote bifunctional activity. Meng et al. proposed a strategy for the fabrication of a bifunctional electrode composed of Co₄N, carbon fibers network (CNW), and carbon cloth (CC).⁹⁴ The polypyrrole (PPy) nanofiber network (PNW) was first electro-deposited on the surface of CC, followed by a facile chemical reaction in a solution containing 2-methylimidazole and Co(NO₃)₂ to form the ZIF-67/PNW/CC. The as-prepared material was then calcined at 700 °C for 2 h under the N₂ atmosphere to convert ZIF-67 to Co₄N. The schematic illustration of the synthesis of Co₄N/CNW/CC is shown in Fig. 4(e). The high-resolution TEM image and SEM-EDS analysis confirmed the morphology of the material and the uniform distribution of elements C, Co, and N, as shown in Figs. 4(f) and 4(g). The combination of Co₄N and CNW on the CC promoted the formulation of a large surface area catalytic material with highly active sites for both OER and ORR; the oxides on the surface of Co₄N functioned as OER active sites, while PPy-derived N-doped carbon fibers were the active sites for ORR. The modified catalyst showed excellent bifunctional activity toward oxygen reactions with potential difference as low 0.74 V, and the Co₄N/CNW/CC electrode-based cable-type ZAB displayed good performance by maintaining the voltage gap of 1.09 V for more than 136 h at a current density of 50 mA cm⁻², as shown in Fig. 4(h).

Similarly, transition metal sulfides such as NiCo₂S₄ showed edges over oxide-based catalysts due to their higher conductivity, which is beneficial for charge transfer during electrochemical reactions. Because of their richer redox sites and promising bifunctional activities toward oxygen reactions, transition metal sulfides have been justified as potential cathode material for batteries and supercapacitors.^46,47,79,95 To prompt the electron transportation and generate a larger surface area, Wang et al. directly grew NiCo₂S₄ nanotubes on the surface of the 3D N-doped carbon carrier (3DNCC) and utilized the composite as a cathode material for ZAB.⁹⁶ The nanotube structure of NiCo₂S₄ with the mesoporous surface not only increased the surface area but also favored the reduction and evolution of O₂. The controlled growth of NiCo₂S₄ nanotubes on 3DNCC facilitated the permeation of the electrolyte to the inner space of the catalyst and reduced diffusion resistance, which are beneficial for easier utilization of active catalytic sites. As an air electrode for ZAB application, it demonstrated an outstanding performance with a voltage gap as low as 0.67 V and a corresponding energy density of 688 W h kg⁻¹.

Metal-free materials have also been intensively studied to replace the Pt- and Ir/Ru-based catalysts. Aside from the above-mentioned advantages of carbon materials such as high electrical conductivity, diverse structure, rich defect chemistry, and high surface area, the great interest for these materials also arises from the weight of the final cathode, which has a direct effect on the overall energy density of the device.⁹⁷ While un-functionalized carbon materials usually display poor electrocatalytic activity, metal-free heteroatom-doped carbon materials are considered as potential candidate electrodes due to their unique electronic and structural features, robustness in the various electrolyte, and remarkable bifunctional oxygen electrocatalytic activity.⁹⁸ Pendashteh et al. fabricated carbon nanotube fibers (CNTfs) via the chemical vapor deposition (CVD) spinning method and tuned its N-content and defect density by using the urea assisted hydrothermal route.⁶² The insertion of N atoms in the CNTfs promoted the formation of active sites and assisted the cleavage of O—H bonds in the water molecule and the formation of O—O bonds in oxygen molecules. Based on the XPS survey spectra for the pristine CNT fiber and treated samples at various temperatures [Figs. 5(a) and 5(b)], the ratios of sp³/sp² and O/C were increased upon the variation of temperature, which might be correlated with the charge redistribution on adjacent carbons; this charge redistribution is important in facilitating the adsorption of intermediate species during oxygen reactions. Additionally, N-doped CNTFs owned a high concentration of pyridinic N-species, which are considered to expedite the oxygen adsorption and hydroperoxide decomposition during the electrochemical process. The modifications done on the structure of the carbon nanotube resulted in excellent bifunctional oxygen reaction, and when extended its function as air electrode, the N-doped CNTFs-based ZAB attained a roundtrip efficiency of 66% with a low voltage gap of 0.97 V at a current density of 20 mA cm⁻².

Furthermore, doping N-doped carbon nanomaterials with secondary heteroatoms is an effective way to further optimize the electronic and chemical features of the carbon materials. For example, Hu et al. designed N, O dual-doped graphene nanorings–integrated boxes (denoted as NOGB) via high-temperature pyrolysis to promote multi-functional catalytic activity toward HER, OER, and ORR and employed them as an air electrode for the rechargeable Zn–air battery.⁵⁷ After the pyrolysis process, a strong acid etching procedure was done to remove the metal core/template and to introduce oxygen elements in the N-doped graphene. Based on SEM and TEM images, the synthesized catalyst obtained a hollow-nanobox morphology with numerous graphene nanorings inside, and the elements C, N, and O were uniformly distributed on the surface of the catalyst, as shown in Figs. 5(c)–5(f). The high-resolution C 1s spectra of NOGB-800 corresponded to C—C, C—N, C—O, and C=O species, which not only indicate the successful formation of N, O-dual doped carbon but also represent the generation of ketonic C=O active sites, which is highly beneficial for the OER and HER process.⁹⁹ In addition, the hierarchically porous structure of the as-prepared catalyst provided access to active sites, which prompts the mass transport kinetics, and the high graphitization degree of carbon enhanced the electrical conductivity.

In the case of N, P dual-doped nanocarbon, it is assumed that the presence of pyridinic/graphitic N species and P atoms could alter the chemical/electronic environments of adjacent carbon atoms, which would lead to the enhancement of electrocatalytic activity. Specifically, graphitic N atoms could initiate charge density redistribution and work as electron-donating species on the neighboring C atoms, while P atoms with a much larger atomic size could promote a defect-induced active surface for O₂ adsorption.¹⁰⁰ On the other hand, heat treatment has been known as a vital process to prepare heteroatom-doped carbon. Controlling its parameter has a direct impact on the electrocatalytic performance such that (1) a higher temperature could alter the graphitization level and increase the conductivity but, at the same time, lowers the concentration of dopants, which results in the loss of active sites and (2) a lower temperature could result in poor electrical conductivity because of the incomplete carbonization of precursors. In line with this, Cai et al. systematically submitted their as-prepared vertically aligned CNT on the surface of graphene foam (VACNT-GF) to a heat treatment at 800 °C under an inert gas atmosphere with 0.76/0.41 N/P ratio to obtain a hierarchically porous N, P dual-doped VACNT-GF.⁹⁸ X-ray photoelectron spectroscopy measurements confirmed that the prepared material has a dopant-rich outer layer with a high amount of N—P bonds and abundant pyridinic N species, which could control the oxygen chemisorption mode during oxygen reaction. Additionally, after a controlled annealing process, the CNT core retained its highly graphitic structure, which could act as electron highways. All of these factors helped boost the electrocatalytic activity of the N, P dual-doped VACNT-GF for OER and ORR.

In addition, multi-heteroatoms (N, F, and P) were uniformly distributed on the surface of carbon fibers to achieve a high surface area and a macroporous structure and to avail the synergistic effect of doped heteroatoms.¹⁰¹ The long carbon nanofibers were prepared via an electrospinning process of NaPF₆ and PAN dissolved in N,N-dimethylformamide (DMF) and then followed by annealing treatment under the Ar atmosphere at 1000 °C. The modified fibers displayed a 3D network structure with abundant macropores and obtained a specific surface area of 1230.1 m² g⁻¹, which is much higher than that of un-doped carbon fibers (13.7 m² g⁻¹). The synergistic effect of N, F, and P dopants in the structure of carbon, the large surface area exposed with numerous active sites, and the highly porous structure, which is advantageous for efficient mass transfer movement, were responsible for boosting the catalytic activity of ternary doped carbon fibers toward ORR and OER, which showed better results than pure carbon fibers and commercial Pt/C catalysts.

Recent studies have shown that transition metal compound (spinel and perovskite oxides, layered hydroxides, nitrides, sulfides, and phosphides) catalysts coupled with N-doped carbon materials have the potential to exhibit higher bifunctional catalytic activity toward ORR and OER. However, the metal-based nanoparticles in this type of hybrid catalyst are prone to severe agglomeration and fast detachment from the carbon matrix over the electrochemical operation, and thus, their application as an air electrode for ZAB is limited due to the challenge in long term stability. Guo et al. utilized a two-step thermal treatment to synthesize a highly stable bifunctional electrode based on core–shell structured transition metal oxide catalysts coupled with the N-doped carbon material.¹⁰² Initially, ZIF-67 was prepared by mixing Co(NO₃)₃.6H₂O and 2-methylimidazole with methanol. The as-prepared MOF material was carbonized under Ar for 2 h to derive metallic Co nanoparticles encapsulated in the N-doped carbon matrix, wherein some CNTs were also derived, and then was heated in air for another 2 h to partially oxidize the metallic Co and generate Co@Co₃O₄ embedded into N-doped carbon polyhedral. The modified hybrid material displayed not only excellent stability but also obtained high content of graphited carbon/CNTs with multiple N-containing functional groups, which assisted the overall catalytic activity of Co@Co₃O₄/N-doped carbon.¹⁰³ Additionally, the synergy between the semi-conducting Co₃O₄-shell and the conducting Co core with a typical porous structure and the presence of CoN_x species promoted catalytic properties toward OER and ORR.¹⁰⁴ The hybrid material functioned as a good air electrode for ZAB and maintained a low voltage gap of 0.66 V for over 200 h.

In order to meet the above-mentioned challenges from severe aggregation and rapid peeling off of active particles on the surface of carbon support, Fu et al. prepared N-doped carbon aerogels with FeCo particles strongly anchored on its framework.¹⁰⁵ A mixture of Fe and Co salt precursors with chitosan and graphene oxide was submitted to facile sol–gel polymerization, followed by freeze-drying and heat treatment, wherein chitosan played dual roles for promoting gelation and N-dopants. The 3D interconnected porous structure of as-prepared carbon aerogels not only restrained active particles from aggregation and detachment but also provided multi-dimensional charge transport pathways during the electrochemical operation. The presence of N-dopants in the scaffold of carbon offered more active sites for oxygen reactions, while the encapsulated metal alloy particles enhanced the ORR/OER catalytic activities owing to its high electrical conductivity. The modified catalyst was employed as an air electrode for ZAB and obtained a high power/energy density of 115 mW cm⁻² and 988 W h kg⁻¹.

One salient difficulty of transition metal and N-doped carbon materials is the lack of a simple preparation method to acquire a highly active bifunctional hybrid material, particularly the one with the 3D porous structure and high specific surface area. Such a structure is highly desired because it could inhibit agglomeration and detachment, expose more potential active centers to the electrolyte, and protect the metal particles from rapid corrosion during the long-term electrochemical process. To prepare Co nanoparticles enclosed in 3D N-doped porous carbon foams (CoNCF), Jiang et al. presented a natural gas-foaming strategy, which is a facile carbonization process of the carbon source, ammonia, and metal salts, under controlled temperature to manipulate the final morphology and structure.¹⁰⁶ The designed CoNCF material obtained a remarkable specific surface area of 1641 m² g⁻¹, which exposed more active sites and facilitated the rapid electron transport and oxygen diffusion. The synergistic effect coming from C—N, Co—N_x, and Co—O species and the presence of abundant structure defects promoted bifunctional activity toward OER and ORR. When applied as a bifunctional catalyst in rechargeable ZAB, the CoNCF-based air electrode demonstrated a notable roundtrip efficiency of 62.5% and maintained a low voltage gap of 0.75 V for 166 h. Based on the obtained results, the proposed gas-foaming strategy could be a facile preparation method to construct various 3D porous materials with high specific surface area and expose to different active centers.

Moreover, the introduction of multiple heteroatoms on the carbon matrix is becoming one of the major trends, since it could create a distinctive electronic structure with a synergistic coupling effect among the heteroatom dopants and further improve the overall catalytic activity and stability.¹⁰⁷ For example, Zeng et al. modified the bifunctional activity of NiCo₂O₄ nanosheets by incorporating it into the N,O-dual-doped carbon nanotubes (N-OCNT) film. The attachment of porous NiCo₂O₄ nanosheets to the carbon structure was easily executed due to the presence of hydrophilic oxygen-containing groups in the functionalized CNT. The OER activity of the final material was ascribed to the functionalization of CNT films via the anodic oxidization method wherein it promoted abundant defects and catalytic sites, while ORR activity is correlated with the presence of three different types of N species, namely, pyridinic, pyrrolic, and quaternary N. The modified material showed good bifunctional activity toward oxygen reactions by obtaining a half-wave potential of 0.86 V (vs RHE) and a low overpotential of 270 mV at 10 mA cm⁻².¹⁰⁸ Likewise, the introduction of phosphate groups into the surface of the N, P-co-doped carbon support can tune the surface hydrophilicity and improve the mass transfer at the interface between electrons and electrolytes. Wang et al. synthesized a new kind of Ni metaphosphate on the surface of N,P dual doped carbon (NPC) support [Ni(PO_xN_3−x)₂/NPC] to develop a bifunctional O₂ electrocatalysts for rechargeable Zn–air batteries.¹⁰⁹ Through post-treatment of nickel(II) acetate/phytic acid, followed by carbonization under Ar and NH₃ atmosphere at 800 °C, a honeycomb-like porous morphology with a specific surface area of 669.8 m² g⁻¹ was obtained. While the catalytic activity of Ni(PO_xN_3−x)₂/NPC was expected to be improved due to the availability of N-dopants, the combined presence of metaphosphate metal-containing parts, $PO4−$ groups, and N,P dual heteroatom-doped carbon support have further modified the surface hydrophilicity and the electronic structure, which benefited the mass transfer movement and electrical conductivity, respectively. The electrochemical performances of Ni(PO_xN_3−x)₂/NPC for OER and ORR showed favorable results with a low overpotential of 370 mV at 10 mA cm⁻² and a half-wave potential of 0.83 V, which lead to the small potential difference of 0.77 V.

Furthermore, aside from the presence of multiple oxygen vacancies in the CoO-NSC-900 catalyst (vide infra), the incorporation of CoO hollow nanoparticles to N,S-dual doped porous carbon is equally important. The N, S dual doping could improve the electrocatalytic activity of graphitic carbon by modulating the electron distribution and adjusting the adsorption of reaction intermediates for the ORR/OER process.⁵² As an air electrode for ZAB, it obtained a roundtrip efficiency of 56% and maintained a voltage gap of 0.8 V with a corresponding high energy density of 943 W h kg⁻¹ for 360 continuous cycles. Similar catalytic activity trend was observed on the CoP quantum dots (QDs) embedded in S,N-codoped graphite carbon (CoP@SNC), which was prepared by Meng et al.¹⁰⁷ The dual doping of S and N species into carbon support not only promoted electroactive sites for both OER and ORR but also influenced the conductivity and charge transfer capability of the modified catalyst. The CoP@SNC favored the four-electron pathway for reversible OER and ORR processes and obtained a low ΔE value of 0.79 V.

Some of the bifunctional O₂ catalysts discussed above were fabricated as a binder-free air electrode for the Zn–air battery. The fabrication of the binder-free air electrode through direct growth of metal precursors onto the surface of a conductive collector not only eliminates the use of any polymer binder that hinders some of the electroactive sites and improves the mechanical stability but also reduces the final weight of the electrode to about 10%–40%.^34,35 Thus, various methods have been utilized to prepare a binder-free air electrode comprised of the bifunctional O₂ catalyst, GDL, and conductive substrate.

For example, Guo and Meng et al. prepared a flexible 3D free-standing bifunctional electrode composed of Co₃O₄@NiFe LDHs and Co₄N/carbon fibers network rooted in carbon cloth, respectively, and used them directly as air electrode for the rechargeable Zn–air battery.^89,94 In both studies, the carbon cloth served as a highly conductive substrate and an effective gas transport highway during oxygen reactions. Owing to the 3D network structure with large specific surface area and abundant active sites, the Co₄N/CNW/CC electrode showed excellent bifunctional catalytic activity toward OER and ORR and even exceptional performance for both primary and rechargeable Zn–air batteries in terms of current density and stability. Impressively, the discharge performance remained almost unchanged even after bending in different angles (bent at 30°, 60°, 90°, and 120°), as shown in Figs. 6(a) and 6(b). The bendable and twistable properties of carbon cloth make it highly suited for fabricating the flexible Zn–air battery to power up a portable and wearable electronic device. Similarly, Wang et al. prepared a binder-free air electrode via self-growth of NiCo₂S₄ nanotubes on the filter paper derived 3D N-doped carbon carrier (NiCo₂S₄/3DNCC).⁹⁶ One face of as-prepared NiCo₂S₄/3DNCC was covered by the polytetrafluoroethylene (PTFE) layer and functioned as a gas diffusion layer. After it was fully covered by PTFE, the assembled NiCo₂S₄/3DNCC–PTFE was employed as air electrode, while the Zn plate served as the anode for ZAB. The stability of Zn—NiCo₂S₄ was studied in a voltage range from 1.3 V to 2.05 V at a current density of 1 mA cm⁻². After 400 continuous charge−discharge processes for 400 h, the capacity of the hybrid battery remained almost constant and its binder-free character remained stable. Its 3D mesoporous structure not only facilitates the rapid charge transfer but also favors the reduction and evolution of O₂.

The textural attributes of the air electrode, such as surface area, pore-volume, and pore size, have been identified as important properties since they could provide accessible active sites to oxygen, electrolyte, and electron during the electrochemical process. To provide efficient mass/charge transport, the selected material that could be used as a scaffold for the freestanding or binder-free electrode should possess excellent textural properties. For instance, Cai et al. prepared a metal-free and free-standing air electrode based on N, P dual-doped vertically aligned carbon nanotubes on graphene foam (NP-VACNTs-GF).⁹⁸ The 3D structure of hierarchical VACNTs on graphene foam allowed the diffusion of O₂ to reach the catalytically active sites effectively, enhanced the conductivity, and reduced the final weight of the air electrode. The assembled Zn–air battery was composed of a polished Zn plate as the anode and as-prepared NP-VACNTs-GF as the air electrode. A small amount of 0.5% PTFE solution was drop-cast onto the cathode material and functioned as GDL. Due to its high conductivity, hierarchical structure, and numerous highly active catalytic sites, the NP-VACNTs-GF-based electrode showed better performance than the similarly prepared Pt/C/IrO2-based electrode coupled Zn–air battery, as it remained stable for over 145 h continuous operation at a current density of 10 mA cm⁻², which is much longer than the latter electrode. Furthermore, Zeng et al. prepared an “all-in-one” air electrode for ZAB using a binder/substrate/GDL-free bifunctional NiCo₂O₄@N,O dual-doped CNT composite film, as shown in Fig. 6(c).¹⁰⁸ They assembled an ∼50 cm long cable-like solid-state ZAB comprising a zinc wire as the anode, free-standing NiCo₂O₄@NOCNT films as the air electrode without adding any binder, substrate or gas diffusion layer, and alkaline polyvinyl alcohol (PVA) as the gel electrolyte [Fig. 6(d)]. All-solid state, highly flexible, and rechargeable cable-like ZABs based on NiCo₂O₄@N-OCNT film cathode not only showed low overpotential, great stability, and remarkable flexibility [Figs. 6(e)–6(h)] but also opened a new way to develop next-generation wearable and portable energy-storage devices.

Overall, this review emphasized the structural features required to satisfy the design of an ideal cathode compartment of the Zn–air battery and presented the associated factors that drive the oxygen reactions in the air electrode, based on the relationship between the intrinsic activities of bifunctional O₂ catalysts and the collective strategies employed to modify the electronic structure of ORR/OER electrocatalysts. In addition to what we have discussed carefully in the previous sections, we present Table I to show the summary of the most recent reported bifunctional catalyst in view of their bifunctional catalytic activities and Zn–air battery performances. To measure the bifunctional catalytic activity of the O₂ electrocatalyst, the potential difference (ΔE) between the potential for achieving a current density of 10 mA cm⁻² during the OER (E₁₀) and the half-wave potential (E_1/2) during the ORR has been calculated. Those electrocatalysts that obtained lower potential difference are considered good and have the potential to be efficient O₂ catalysts for the air electrode of the Zn–air battery. Among the list, some of the multi-components transition metal compounds including the combination of spinel oxides and layered double hydroxides, hydroxides and sulfides, perovskites and LDH, phosphides and oxides, and their corresponding counterparts that were coupled with single or multi-heteroatom doped-carbon materials (mostly N-doped based) have shown high bifunctional catalytic activity toward ORR and OER by obtaining lower potential gap (<0.8 V); thus, these materials have displayed remarkable results when employed as an air electrode for ZAB. However, it is difficult to directly compare the reported results of the recent air electrodes since their performances are dependent on other parameters, including the concentration of electrolyte, mass loading, applied current density, and fabrication method of the electrodes. The more concerning fact is that most of the reported studies omit some of the necessary experimental results, making the assessment of the overall performance incomplete. Therefore, it is vital to set a general evaluation rule to properly gauge the air electrode performance, which may include the list of relevant data, such as mass loading of working electrode, roundtrip efficiency, voltage gap, durability (number of cycles), power density, and energy density. This would not only provide uniform standard when gauging the present performance of air electrodes but also be greatly helpful to view the actual status of the progress of the Zn–air battery.

In this review, we summarized the properties of air electrodes needed to fulfill the requirements of an electrically rechargeable Zn–air battery: (1) high specific surface area with abundant active sites exposed to catalyze both the ORR and OER efficiently, (2) hierarchical porous structures that consist of mesopores and macropores, wherein the former function as reaction sites for the triple-phase oxygen reactions, while the latter serves as the gas transport channel during electrochemical process, (3) catalyst layer with a superhydrophilic surface not only to improve the mass transfer at the interface between electrons and electrolyte but also to increase the electrical conductivity, (4) gas diffusion layer with a controlled hydrophobic surface and abundant pores to allow O₂ in the gas phase to move toward the catalyst layer while inhibiting any liquid electrolyte leakage, and (5) strong interactions between the catalyst-GDL and conductive substrate to reduce the interface resistance. These properties could address the challenges of ZAB properly; however, fabricating an individual air electrode possessing all of these characteristics is extremely difficult. For instance, to attain the ideal catalytic performance of an air electrode, most of the studies have been focused on the design and modifications of the morphology, active sites, and electronic structures of the O₂ electrocatalysts. A great improvement has been accomplished since the realization of the bifunctional ORR/OER electrocatalyst. Nevertheless, there is an impending need of advanced fabrication methods that could effectively (1) increase the specific surface area of the catalyst exposed with catalytic active sites, (2) control the porous structure, (3) enhance the electrical conductivity, and (4) prevent the severe agglomeration problem during the charge–discharge process. All of them are highly necessary, but the latter part is particularly urgent since the agglomeration of nanoparticles on the catalyst-GDL layer is likely to block the diffusion pathways, hence limiting the catalytic activities. In line with this, the fabrication of the direct growth or binder-free air electrode has been considered as a strategical move, since the active nanoparticles are directly deposited or grown on the scaffolds of the highly porous conductive substrate, thus preventing the rapid agglomeration of the nanoparticles. In addition, direct growth of nanoparticles on the surface of the current collector eliminates the use of any polymer binders that hinder some of the electroactive sites, improves the mechanical stability, and reduces the final weight of the air electrode to about 10%–40%. However, this method still needs a lot of improvement not only to ensure the strong interactions between the catalyst-GDL layer and the current collector but also to obtain a defined structure of the attached catalyst.

Furthermore, it should be emphasized that all the components in an air electrode (catalyst layer, gas diffusion layer, and current collector) play individual roles, which are equally important in the overall improvement of the ZAB performance. Unfortunately, there are not enough references that have focused on the progress of GDL and current collector nor on the wettability properties of the air electrode. The catalyst layer may be really important in promoting catalytic activities, but GDL ensures the efficient oxygen diffusion between the electrolyte and ambient air and prevents any electrolyte flooding/leakage. Current collectors do not only function as physical and conductive support for the catalyst layer but also provide a highway for fast and efficient oxygen diffusion during the reactions. More importantly, the lifespan of an air electrode mainly depends on the preservation of the desired wettability properties and porosity of GDL, and thus, the report on the hydrophilicity or hydrophobicity of this layer is mandatory for further development of ZAB technology. Finally, we hope that this review could stimulate new ideas for the future designs of the air electrode in order to meet the critical requirements of an electrically rechargeable Zn–air battery for it to enter the commercialization path.

This work was supported by the National Research Foundation of Korea (Grant Nos. NRF-2020R1A2B5B03002475 and NRF2019R1A6A1A11044070), the Hydrogen Energy Innovation Technology Development Program of the National Research Foundation of Korea (NRF) funded by the Korean government [Ministry of Science and ICT (MSIT)] (Grant No. NRF-2019M3E6A1064709), the Korea Basic Institute under the R&D program (Project No. C38530) supervised by the Ministry of Science, the National Key Research and Development Program of China (Grant No. 2016YFA0203101), and the National Natural Science Foundation of China (Grant No. 51572139).