The cell performance of DMFC is usually lower than that of PEMFC using hydrogen as fuel due to the sluggish MOR and ORR reaction kinetics in both the anodes and cathodes, and the Pt loading in the catalyst layer is also significantly higher. In recent years, a variety of Pt-based alloy systems with various compositions have been studied, such as Pt-Pd, Pt-Au, Pt-Ag, etc. (Alcaide et al., 2011; Wang et al., 2011b; Arikan et al., 2013; Cao et al., 2015; Feng et al., 2016; Gowthaman et al., 2018) These binary and ternary alloy catalysts have improved ORR activity and relatively large surface areas. The unique structural properties and composition increase the catalytic activity and further reduce the Pt loading. The design of binary and ternary nanostructures of Pt is a good strategy to develop low-Pt ORR catalysts, which even exhibit better performance than monometallic Pt.

Stamenkovic et al. (2006) investigated the polycrystalline alloys of Pt₃M (M = Ni, Co, Fe, or Ti) to study the effect of 3d transition metals on the ORR performance of Pt alloys. It was discovered that the activity of this alloy is determined by the surface properties of the 3d metal. They also revealed the correlations between the electrocatalytic on the surface of Pt₃M, the d-band center, and ORR activity. This correlation exhibits a “volcanic” shape where the highest activity is related to the nature of the 3d metal (Stamenkovic et al., 2007b). To boost the removal of oxygen-containing intermediates, an effective ORR electrocatalyst should have a lower binding energy compared to Pt. They revealed that an optimal ORR catalyst exhibits a weaker binding energy of about 0.2 eV for oxygen-containing species compared to pure Pt (Stamenkovic et al., 2006; Tripković et al., 2010). In this regard, Ni, Co, and Fe are the most effective alloying elements. This trend can help researchers choose the most appropriate alloy components to control the surface chemisorption characteristics of Pt. It was also found that single-crystal Pt₃Ni (111) exhibited enhanced catalytic activity for ORR, which was 10 times higher than that of commercial Pt (111). The significant increase in catalyst activity is attributed to the presence of Ni, which weakens the adsorption of OH⁻ on the catalyst surface and thus increases the adsorption sites of O₂ (Stamenkovic et al., 2007a).

As for effective Pt-Ni alloy catalysts, Antolini et al. (2005) prepared carbon-supported Pt-Ni alloys with sodium borohydride at room temperature and tested their performance for DMFCs for the first time. The obtained catalysts worked as both cathodes and anodes, and the performance of Pt-Ni as a cathode is higher than that of Pt-Ni as an anode. The catalysts were prepared with a Pt to Ni ration of 90:10 or 70:30. As a cathode material, Pt₉₀Ni₁₀ is superior to Pt in both mass activity and specific activity. The enhanced performance of Pt-Ni as cathode is attributed to the electronic effect of ORR and better methanol resistance related to the alloy. Todoroki et al. (2015) investigated Ni-Pt alloy nanoparticle-stacking thin films (NPSTF) as a novel ORR electrocatalyst. The Ni-Pt was deposited via the arc plasma deposition method. Arc plasma is created by a gas arc discharge and can generate nanosized particles with the high temperature discharge (Gerdeman and Hecht, 2012). The mass activity of the as-synthesized Pt−Ni NPSTF is 10 times higher than that of commercial Pt/C catalysts. Pt−Ni NPSTF exhibits a significant ORR activity enhancement, which is attributed to the underlying Ni atoms’ modification of the surface Pt-rich layer and the enhanced active surface area caused by the stacking of PtNi nanoparticles, as shown in Figure 5.

Zou et al. (2015) synthesized an ordered PtNi (O-PtNi/C) intermetallic compound supported by carbon from the conversion of disordered PtNi nanoparticles. The nanoparticles were converted by a heat treatment method at 550–600°C under a H₂/N₂ atmosphere. The intermetallic nanoparticles were characterized and showed increased mass and specific activity. At 0.85 V, the specific activity of the PtNi nanoparticles is about 6 times that of commercial Pt/C and 3 times that of disordered PtNi alloy. The stability of the as-prepared catalyst may be related to the structural changes during the heat treatment.

Besides Pt alloying, designing structures at the nanoscale is also an effective way to disperse Pt on the surface to the greatest extent and to reduce the overall Pt loading. Pt alloying can be combined with appropriate core-shell structures or nanoframes to maximize the performance of the catalysts (Wu et al., 2019). Pt-based core-shell nanostructures consist of a core made with metals, metal oxides or metal nitrides and a thin layer of Pt grown on the surface (Zheng et al., 2018). They have improved Pt utilization, catalytic activity, and stability. Yang et al. (2016) used density functional theory (DFT) calculations to screen binary and ternary Pt-shell electrocatalysts made with non-precious elements for ORR. By calculating the adsorption intensity of O on nanoalloy clusters using DFT, the trend of ORR activity of Pt-shell electrocatalysts may be determined. The equilibrium geometry of a three-layer Pt₄₂M₁₂N₁ (M, N = Fe, Co, Ni, and Cu) cluster model has an icosahedral (Ico) structure. To maximize ORR activity, all atoms in the surface layer are Pt atoms, and atoms in the subsurface layers are transition metals to lower the amount of Pt utilized. If M is the same as N, the systems are binary Pt₄₂Fe₁₃, Pt₄₂Co₁₃, Pt₄₂Ni₁₃ or Pt₄₂Cu₁₃ clusters. When M is not the same as N, the systems are ternary Pt-shell clusters, such as Pt₄₂Cu₁₂Ni₁ and Pt₄₂Fe₁Co₁. The surface of the Ico structure includes 20 facets, so the adsorption on these highly symmetric nanoclusters is worth studying. For each triangular face, six different adsorption sites were found. In binary Pt-M (M = Fe, Co, Ni, and Cu) electrocatalysts, Pt₄₂Fe₁₃ has the strongest ORR activity. While Pt₄₂Fe₁₂Cu₁ was the most effective catalyst in ternary Pt-M-N systems. Yang et al., 2021b reported a bimetallic Pt-metal core-shell structure with good ORR activity. The Pt₃Co@Pt core-shell structures had a mass activity of 0.71 mA mg_Pt ⁻¹ and specific activity of 2.75 mA mg_Pt ⁻¹ attributed to the well-defined core-shell structure.

Kuttiyiel et al. (2012) reported an electrocatalyst with low-Pt monolayer and a bimetallic IrNi core on carbon support (Pt_ML/IrNi/C) using the potential shift of a Cu layer synthesized at a negative potential. The Pt mass activity of the as-prepared electrocatalysts in the scaled-up synthesis is three times higher than that of the commercially available Pt/C electrocatalysts. The electronic and geometric effects of the IrNi substrate led to enhance activities. The experimental results are supported by DFT using a spherical model, demonstrating an effective approach to improve the activity of cathodic oxygen reduction.

Glüsen et al. (2019) fabricated a PtNi core-shell structure for reduced Pt content in DMFCs. The core-shell catalyst initially exhibited similar performance to commercial Pt. But after operation, a Pt-rich shell forms with a 0.7 nm thickness, which corresponds to 3 Pt atom layers. This Pt shell led to increased DMFC performance, which was comparable to a cathode consisting of five times the Pt amount. Figure 6A shows the difference in the Pt shell before and after the electrochemical tests, which could be attributed to leaching of the Ni atoms from the core. Figure 6B shows the stability test over a long time, and the performance showed no decay even after 1,000 h of operation due to the robust durability of the core−shell catalyst. In the future, it would be beneficial to expedite such a structure alteration so that the catalyst can achieve the optimized performance in a shorter time.

Nanoframes with 3D surfaces can achieve higher catalytic activity with lower Pt content because the atoms on the inner and outer surfaces can contribute to catalysis at the same time, resulting in a much higher utilization of active atoms. Structural control at the atomic level can precisely and efficiently tune the catalytic performance of the material, thus improving activity and durability. Chen et al. (2014) used a unique air etching method to synthesize polymetallic Pt₃Ni nanoframes as an ORR catalyst, and they conducted an in-depth study on the structure and formation process of the materials. Characterization results by elemental mapping, X-ray photoelectron spectroscopy (XPS) and transmission electron microscopy (TEM) demonstrated that a 3D nanoframe structure was formed with Pt-atom-rich surfaces due to the limited etching effect of O₂ on Ni. The obtained Pt₃Ni catalyst has a very high catalytic activity for the ORR, with a mass activity 22 times higher than that of commercial Pt/C. The method is practical for the synthesis of polymetallic nanoframes and can be used to prepare frame structures such as PtRuNi, PtPdNi, etc. The fabrication of Pt-based nanoframes has good prospects for scale-up applications.

Transition-metal nitrides (TMN), oxides, and carbides have ORR activities and can be used in Pt-free catalysts, as will be discussed below, but they can also be combined with Pt-based materials to improve the overall activity and durability of the composite catalyst (Yuan et al., 2022). Carbide-based Pt alloy composites are fabricated in various forms, such as core–shell structure, nanoparticles, and single atoms. With the excellent durability and large specific surface area of carbides, these compounds have great potential in ORR catalysis (Li et al., 2019a). They not only reduce noble metal usage but have also been shown to enhance the intrinsic activity and durability of Pt in ORR (Park et al., 2014; Zhang et al., 2014; Zhao et al., 2015; Lin et al., 2019). Transition metal oxide-supported electrocatalysts are regarded as one of the most promising electrode catalyst alternatives due to their great stability than carbon under various operating conditions. The introduction of metal oxides can provide anchor points for the nucleation and growth of Pt but also increases the exposure of active interfaces for efficient catalytic reactions (Liu et al., 2017; Song et al., 2018; Chen et al., 2021). In addition, the strong interfacial active band of the metal oxide itself and the interaction between Pt and metal oxide can not only turn Pt into an electron-rich state but also increase the surface energy and stabilize the growth size of the material. The goal of speeding up the chemical reaction is met while the amount of Pt in the electrocatalyst is decreased (Bertram et al., 2020; Xu et al., 2020; Yang et al., 2021a; Chen et al., 2021; Valizadeh et al., 2022).

In addition, transition metal nitrides with high electrical conductivity and thermal stability are suitable ORR electrocatalysts (Ham and Lee, 2009; Cao et al., 2013; Kumar et al., 2013; Youn et al., 2013). The formation of nitrides alters the electronic structure of the catalyst, but the interactions within TMN can also enhance the adhesion of the supported noble metals, thus enhancing the electron transfer of the catalytic process (Zhong et al., 2006; Xia et al., 2008; Cui et al., 2013; Yang et al., 2013; Cui et al., 2014). Tian et al. (2016) synthesized an effective TMN catalyst with TiNiN as the core coated with a layer of Pt . The TiNiN@Pt catalyst showed great ORR catalytic activity, with a more than 4-times increase in mass activity compared to commercial Pt/C. It also possessed long-term durability within 10,000 potential cycles due to the strong binding between Pt nanoparticles and TiNiN. Figures 7A,B shows the TEM images of the particles, and Figures 7C,D shows the CV curves and ORR LSV curve of the as-synthesized catalyst. This catalyst avoided the corrosion problem of conventional carbon supports by using highly stable transition metal nitride and successfully enhanced the durability of the Pt/TMN composites.

As a new type of Pt-free oxygen reduction catalyst, transition metal sulfide compounds have attracted research attention because of their abundant sources and good stability in acidic media. The most studied transition metal sulfide compounds are mainly Ru-based materials serving as alternative Pt-free catalysts. The transition metal sulfide compounds can be divided into two types: binary compounds and ternary compounds. Brouzgou et al. (2012) summarized the methanol resistance of the transition metal chalcogenides. The activity of the Co-Se carbon-loaded ORR catalyst remained constant in the presence of methanol (Nekooi et al., 2010). Pd-Se catalysts also showed great activity and good ORR tolerance. According to Jeng et al. (2011), the performance of DMFC using RuSe/CNT as the cathode catalyst was better than that of Pt/C, depending on the type of carbon support.

Ru-based and Ir-based chalcogenides are also promising ORR catalysts with high methanol tolerance in acidic media. Lee et al. reported the iridium-selenium (Ir-Se) system with high ORR catalytic activity, which may be due to the bimetallic interaction effects (Lee et al., 2007b). Alonso-Vante et al. (2002) used in situ extended X-ray absorption fine structure (EXAFS) to study Ru_xX_y (X = S, Se, and Te) compounds. It was found that the catalyst had a Ru atomic nucleus with triangular coordination and direct metal-metal bonding. The catalytic activity depends on Ru atomic cluster size and O₂ interaction. Among the as-synthesized chalcogenides, Ru_xSe_y binary compounds are the most catalytic. To boost ORR catalytic activity and methanol tolerance, transition metal sulfide compounds need improvements in compositional and structural properties. Recently, composite materials with cobalt and sulfur supported by carbon have also made progress. Zhang et al. synthesized CoS_0.197-C composite as efficient ORR catalyst from Co nitrilotriacetic acid (Co-NTA) polymer precursor (Zhang et al., 2019c). The unique 1D porous structure provides more active sites and contributes to the ORR catalytic performance; it also enhances the electron transport between the catalyst and the carbon support.

Radhakrishnan et al. (2019) synthesized amorphous ReS₂ nanosheets as methanol-tolerant ORR catalysts. Figures 8A,B show the FESEM and TEM images of ReS₂, respectively. Figure 8C shows the illustration of the ORR process on the surface of the catalyst. As shown in Figure 8D, the ReS₂ electrode obtained increased current density within the potential range of 0.6 to −0.3 V at higher rotation rates. More importantly, the ReS₂ electrode exhibited 3 times higher current density than MoS₂ under the same rotation rates. The catalyst also outperformed commercial Pt/C in terms of long-term durability and methanol tolerance.

Studies on Ru-based and Ir-based compounds have made some progress, but as noble metals, Ru- and Ir-based catalysts are still costly. Like Pt-based catalysts, alloying would be a good solution to reduce the fabrication cost. Both Ir and Ru work well with Co as composites for enhanced ORR activity. Ir_xCo_1−x alloys exhibited good ORR activity and better methanol tolerance compared with Pt/C catalysts (Lee et al., 2007a).Liu et al. (2021) introduced Ru into Co clusters with ZnRuCo ZIF as a precursor, and synthesized Co encapsulated with nitrogen-doped carbon (α-Ru@Co/CN). Although in this work the obtained catalyst was tested in direct ethanol fuel cells, the ORR activity of the clusters was remarkable with a half wave potential of 0.908 V vs. RHE. The performance was attributed to the strain and ligand effects of the encapsulated structure and the introduction of the Ru element.

Bisen and Nanda (2021) synthesized a Ru nanoparticle-loaded carbon structure for ORR. The illustration for the synthesis is shown in Figure 9A. The Ru@NC catalyst was synthesized by one-step pyrolysis of the dicyandiamide and ruthenium phthalocyanine at 800°C. The Ru@NC exhibited great ORR activity (Figure 9B), which was due to the improved 4-electron transfer via the inner-sphere electron transfer (ISET) mechanism. Figure 9C shows the stability of the as-synthesized catalyst compared to Pt/C catalyst. This work illustrates the effects of synthesis temperature and rotation speeds on Ru@NC catalyst performance. Moreover, the Ru@NC largely suppresses intermediate HO₂ ⁻, which leads to high stability over 10,000 cycles and surpasses commercial Pt/C.

Transition metal-nitrogen-carbon (M-N-C) complexes are good non-precious metal ORR catalysts due to their high activity and resistance to methanol crossover effects. As a possible substitute for precious metal-based cathode catalysts, M-N-C materials have caught the attention of researchers and are expected to be a key component of the most advanced cathode catalysts. The study of M-N-C compounds for ORR began in 1964, when Jasinski (1964) revealed that cobalt phthalocyanine possessed ORR activity. Yeager (1984) discovered the pyrolysis of M-N₄-macrocycle precursors to produce the first M-N-C composite for oxygen reduction. Many effective ORR catalysts based on M-N-C have been created since then.

Lin et al. (2014) utilized a conjugated organic molecule containing pyridine nitrogen as a ligand to synthesize a metal coordination polymer. They ligated the polymers with Fe²⁺ ions and then pyrolyzed them to generate a self-supported catalyst (Fe-N/C) with a high nitrogen and iron doping density. The synthesis route and SEM image of this catalyst is depicted in Figures 10A,B. The catalysts calcined at 800°C possessed a relatively high density of surface-active sites with outstanding ORR activity. Although their specific surface area was not particularly high, their catalytic activity under acidic and basic conditions was comparable to that of Pt/C. What’s more, the Fe-N-C structure was more resistant to methanol than Pt/C. Figures 10C,D show the activity of the Fe-N/C-800 before and after stability tests, in alkaline and acidic media, respectively. The results indicate that the composite materials have great durability in solutions with a wide range of pH.

Recently, Bisen et al. (2021) reported a simple method to conduct pyrolysis of cobalt phthalocyanine (CoPc) and successfully synthesized Co-N-C catalysts without any chemical waste. The Co-N-C catalyst demonstrates improved ORR activity, it was also effective in alkaline media. Martinaiou studied the activity of Fe-N-C under the effect of methanol and tested their degradation process. The results showed that adding methanol has no negative effect on the durability of the catalyst, indicating that Fe-N-C is a stable ORR catalyst with good methanol tolerance (Martinaiou et al., 2020b).

Pd-based catalysts are great substitutes for Pt-based ORR catalysts (Sanij et al., 2021). To address the issue that Pt-catalysts are easily poisoned in DMFC, Zhang et al. (2020) fabricated Pd-Te hexagonal nanoplates (HPs) as an effective ORR catalyst. The catalyst consists of an ordered arrangement of Pd atoms in the Pd-Te nanoplates that supply active sites for the reactions. As a result, Pd-Te HPs/C exhibited promising ORR activity and excellent stability over more than 50,000 s of chronopotentiometry, with little activity degradation and restricted structure/composition alterations. In DMFC devices, Pd-Te HPs/C outperforms commercial Pt/C in terms of methanol tolerance and antipoisoning stability. DFT simulations confirmed that the unique surface arrangements of the obtained catalysts have an intrinsic contribution to the great performance of the catalyst. Pd-Te HPs’ high activity and superior methanol tolerance ensure their viability as possible electrocatalysts for DMFC and beyond. This work provides a novel method for constructing and developing 2D Pd-based catalysts for enhanced activity towards ORR.

Porphyrin-based transition metal macrocyclic compounds are considered as potential electrocatalysts for ORRs in DMFCs. Based on literature, the performance of these catalysts is unstable in acidic media, but it can be significantly improved by pyrolysis (Gojković et al., 1999). A cobalt tetramethoxyphenylporphyrin (CoTMPP) precursor was pyrolyzed to fabricate an ORR catalyst with high methanol tolerance (Wang et al., 2011a). The current density of the CoTMPP/C-pyrolyzed electrocatalyst in the presence of ethanol was consistent with that in the absence of methanol, indicating the great methanol tolerance of the ORR catalyst. The activity of the as-synthesized catalyst is related to the temperature of the pyrolysis and the rotation rates. A higher rotation rate at 700°C contributed to better oxygen reduction activity.

Piela et al. (2010) examined six different non-precious catalysts as cathodes for fuel cell applications, including pyrolyzed CoTMPP. They used non-precious metals as cathodes and Pt-Ru/C as anodes to test the performance of the fuel cells by rotating disk electrode (RDE). CoTMPP exhibited high ORR selectivity and performance, which presented a N₄-metal structure at 500–700°C. CoTMPP shows great potential value for mixed-reactant feed DMFCs that require high ORR selectivity. The performance of pyrolyzed porphyrin-based ORR catalysts greatly depends on the conditions of the heat treatment. It is generally believed that heat treatment can enhance both the activity and stability of these electrocatalysts. For DFMCs with porphyrin-based macrocyclic compounds as cathodes, however, the total fabrication cost is not low because the raw materials are expensive.

Combining transition metal with C and N to form metal carbides (TMC) and nitrides can change the properties of metal centers, thus improving their catalytic performance. The Pt-like behavior of tungsten carbide was observed earlier by Levy and Boudart in several catalytic reactions, which motivated the early studies on the properties of metal carbides and nitrides (Levy and Boudart, 1973; Chen, 1996). TMC or TMN are formed by doping C or N atoms into the interstitial transition metal network, respectively. Abdelkareem et al. (2021) summarized the progress of TMC and TMN and their performances in fuel cells. Since there are various preparation methods for TMC and TMN, which have a high similarity in the structures. For example, TiC nanocrystals synthesized by a simple sonication method can efficiently improve the activity of ORR compared with native TiC, as reported by Wang et al. (2020) With abundant defects, oxygen vacancies, and a large pore volume, the as-prepared catalyst also exhibited good stability and methanol tolerance. This increased activity of atomically thin TiC was related to the high specific surface area and abundant oxygen vacancies, which together promote oxygen adsorption and ORR rate.

TMN is usually more stable than TMC under extreme working conditions; therefore, it is more widely used as an ORR catalyst or catalyst carrier. Yuan et al. (2020) reported the preparation of nanoparticulate zirconium nitride (ZrN) using the urea–glass method that works great in alkaline media. The ORR performance of ZrN tested after cycling in 0.1 M KOH at 1,000 rpm had better stability than Pt/C catalysts. The high activity was attributed to the moderate temperature in the urea–glass method, which avoided the nanoparticle aggregation that occurs during the preparation of ZrN from ZrO nanoparticles treated with NH₃ at 1,200°C. In addition, the authors demonstrate that the thin layer of ZrO generated on the surface of ZrN nanoparticles due to their exposure to air acts as a protective layer and would not hinder electron transfer.

The doping of other elements in carbon materials can change the structure of them, thus affecting their hydrophilicity, electrical conductivity, catalytic activity, etc. Progress has been made in recent years on the addition of heteroatoms, such as B (Yang et al., 2011), N [Gong et al., 2009; Liu et al. (2010a), Yu et al., 2010b], S Yang et al. (2012b), Se Yang et al. (2012b), P (Liu et al., 2011); Yang et al. (2012a), and F (Sun et al., 2013) to modulate the intrinsic properties of carbon materials. Besides the common N-doped materials, other doped carbon catalysts such as B, P, and S also have great ORR catalytic performances. The binary or ternary doped carbon materials such as B-N, N-P, and N-S all have good ORR catalytic activity (Wang et al., 2012; Zheng et al., 2013). Although these catalysts exhibit good catalytic activity under alkaline conditions, their activity under acidic conditions is not as good.

Among the various dopants mentioned above, nitrogen-doped carbon catalysts are the most studied. The N-bonded graphite structure has N atoms instead of carbon atoms. As a result, they have the same structure as graphite carbon atoms, but they introduce more electrons into the off-domain Π-system. Pyridine N can also exist in an oxidized form. These different N functional groups usually coexist, and their concentrations can actually be adjusted (Jin et al., 2011). Among these N-functional groups, pyridine, pyrrole, and graphitic N are usually considered to contribute to the ORR catalytic activity. Some researchers have suggested the involvement of pyridine N in ORR activity (Lee et al., 2010), while others have proposed graphitic N (Luo et al., 2011). One study suggested that graphite and pyridine N sites are interconvertible in ORR (Kim et al., 2011). In addition to the type of nitrogen sites, the number of carbon edge sites and total N content were also found to affect the ORR activities of the materials (Choi et al., 2012).

In general, N/C is prepared mainly by three different pathways. The first method is the in-situ introduction of nitrogen atoms into the carbon skeleton through a process such as chemical vapor deposition (CVD) to form graphite planes (Wei et al., 2009; Reddy et al., 2010). This method is usually not suitable for large-scale applications. The second method is post-heating graphitic carbon, including nitrogen-containing materials such as carbon nanotubes, graphene, and fullerenes (Wang et al., 2009a; Gao et al., 2013). The last method is the direct pyrolysis of nitrogen-containing structures, such as graphitic carbon nitride (Zheng et al., 2011), melamine foam (Lee et al., 2013) and polymer backbones (Zhao et al., 2012). This last method is currently the most popular one due to its simple and manageable preparation technique. However, direct pyrolysis at high temperatures often leads to significant loss of active N species and an inability to control the internal pore structure, resulting in limited ORR active site formation and poor migration properties (Nie et al., 2015).

Carbon nanotubes (CNTs) can help increase fuel cell performance. For example, Pt can be attached to the inner and outer walls of CNTs, which could lead to an improvement in the ORR catalytic performance of Pt/CNT composites (Zhao et al., 2016a; Wang et al., 2023). But sometimes the nitrogen-doped CNTs can serve as ORR catalysts themselves. Gong et al. (2009) recently found that vertically aligned nitrogen-doped carbon nanotubes (VA-NCNTs) can serve as very effective metal-free ORR electrocatalysts. Well-arranged carbon nanotubes with good electrical and mechanical properties, as well as excellent thermal stability, make them suitable electrode materials under harsh environmental conditions. Based on DFT calculations, the improved catalytic activity was attributed to the electron accepting ability of nitrogen atoms.

The theoretical specific surface area of graphene is 2,630 m² g⁻¹, which is much higher than that of carbon black and carbon nanotubes (Hanaei et al., 2016). Due to its huge surface area and improved catalyst dispersion, graphene becomes a promising metal support for ORR catalysts. Chen et al. (2019) developed a single-atom electrocatalyst supported on nanographene for enhanced ORR activity. The catalyst had single Fe atoms anchored by N-doped carbon loading on reduced graphene oxide (Fe-N-NG/RGO). Compared to traditional carbon supports in ORR materials that is generally highly graphitized, the Fe on nanographene had bigger interlayer spacing for better diffusion of O₂ molecules. The as-prepared catalyst can achieve great ORR activity by the transport of O₂ to active sites of Fe-N species. Dong et al. (2017) also utilized N-doped reduced graphene oxide to support binary transition metal nitrides for enhanced ORR activity. They found that N-RGO supported transition metal nitride was much more active than the two components alone, indicating that there was a strong synergistic effect to promote the overall activity between the two materials.