注册 English

Multiscale engineering for regulating oxygen reduction activity in direct methanol fuel cell: A review

Shiquan Guo ,  Fei Chen ,  Congju Li

downloadPDF
Citation:  Shiquan Guo, Fei Chen, Congju Li. Multiscale engineering for regulating oxygen reduction activity in direct methanol fuel cell: A review[J]. Chinese Chemical Letters, 2026, 37(9): 112408. doi: 10.1016/j.cclet.2026.112408 shu

Multiscale engineering for regulating oxygen reduction activity in direct methanol fuel cell: A review

English

  • 1. Introduction

    Energy crisis and climate change are two major challenges facing humanity. Currently, fossil fuels continue to dominate the energy landscape in the near to medium term, exacerbating issues of severe energy shortage and environmental pollution [1,2]. These challenges have raised concerns regarding energy security and environmental sustainability. Guided by the "dual carbon" strategy, establishing a clean energy system is essential for effective resource integration and sustainable energy development [3–5]. To date, multiple energy storage and production technologies based on renewable energy sources have been developed, including hydrogen energy [6], solar power [7], wind energy [8], biomass energy [9], secondary batteries [10], and fuel cell [11]. Among them, the direct methanol fuel cell (DMFC) stands out due to its high energy density (4780 Wh/L), rapid cold-start capability, high conversion efficiency, clean and environmentally friendly fuel, and superior safety features [12–15]. Regarded as a highly promising power generation technology, DMFC is positioned to play a key role in the fast-growing portable electronics market, including laptops, smartphones, and similar devices [16–18]. Liquid methanol offers notable advantages in storage and transport compared to other fuel. Moreover, as methanol can be synthesized from CO2, making it a target product of the "liquid sunshine" initiative, it not only supports economic growth by resource recycling but also helps mitigate the shortage of liquid fuels [19,20].

    Structurally, the DMFC comprises an external circuit, ion-exchange membrane, gas diffusion layers, bipolar plates, and catalyst-based electrodes. Fig. 1 schematically illustrates the principles of both acidic and alkaline DMFC systems. In an acidic medium, methanol solution is delivered to the anode through flow channels and oxygen to the cathode. Methanol oxidation reaction (MOR) at the anode produces protons, electrons, and carbon dioxide. The protons migrate through the membrane to the cathode, while electrons travel via the external circuit, where they combine with oxygen to form water (Fig. 1a). In alkaline DMFC, oxygen reduction at the cathode generates hydroxide ions, which migrate through the anion exchange membrane to the anode and react with methanol, producing carbon oxides (Fig. 1b). The circuit completes the energy loop, powering external loads. DMFC directly converts the chemical energy of methanol into electricity, circumventing the Carnot cycle limitation, thereby achieving higher efficiency [21]. The thermodynamic cell potential is 1.21 V at 25 ℃. However, the actual voltage and power output are considerably lower due to polarization loss, mixed potential, catalyst poisoning, electrode kinetics limitation, operating condition, and material degradation [22,23]. A major factor limiting performance is the sluggish oxygen reduction reaction (ORR) kinetics at the cathode, where currently available electrocatalysts remain a critical bottleneck [24,25].

    Figure 1

    Figure 1.  (a, b) Schematic diagram of DMFC in acidic and alkaline media.
    DownLoad: Full-Size Img PowerPoint

    In fuel cell, the ORR at the cathode necessitates a high overpotential to overcome kinetic energy barrier. Therefore, the development of highly active and stable ORR catalysts has become a major research focus in the field [26–29]. An ideal cathode electrocatalyst for DMFC should possess the following features: (1) High specific surface area and rich pore structure to expose abundant active sites and facilitate the adsorption and diffusion of reactants; (2) High catalytic activity to lower the activation energy and promote ORR kinetics; (3) Good conductivity to enable efficient electron transfer; (4) Strong tolerance to methanol crossover to suppress MOR; (5) High chemical stability to prevent material degradation or failure. Significant progresses have been made in catalysts design in recent decades, driven by advances in structural innovation, nanoengineering, and advanced characterization techniques [30–32]. Among various catalysts, Pt/C remains the most active ORR electrocatalyst for metal air cells and fuel cells, despite the high cost and limited stability. However, CO poisoning is prone to occur due to the methanol crossover effect in DMFC, which deactivates catalytic sites [33,34]. Studies have shown that optimizing the surface properties and atomic arrangement of Pt-based materials is crucial for improving both durability and methanol tolerance [35–37]. Surface modification has emerged as an effective strategy to enhance ORR performance. Non-precious metal-based catalysts have also attracted attention due to their inherent methanol tolerance. Nevertheless, the catalytic activity remains inferior to that of commercial Pt/C. Enhancing the performance of such materials, particularly through systematic multi-scale engineering strategies for catalyst modification, represents a critical research challenge [38,39].

    This review is organized as follows: First, the working mechanisms of DMFC in acidic and alkaline environments are briefly discussed. Then, the kinetic mechanisms of the ORR and commonly used electrochemical evaluation methods are introduced. Subsequently, recent advances in multi-scale modulation strategies for enhancing ORR activity are systematically reviewed, including methanol tolerance optimization, rational design of active sites, electronic structure tailoring, surface/interface micro-environment engineering, and additional modulation approaches (Fig. 2). A detailed discussion is provided from multiple perspectives, including core-shell structures, crystal facet control, surface modification, confinement effects, coordination environment optimization, doping strategies, alloying, heterostructure construction, pore structure design, hydrophilicity/hydrophobicity regulation, composition and size control, and defect engineering. These strategies are examined across advanced material systems such as precious metal-based catalysts, carbon-based materials, single-atom catalysts, alloy catalysts, and metal oxides. Furthermore, the structure-activity relationships are emphasized, and the corresponding catalytic mechanisms are comprehensively analyzed. Finally, the remaining challenges and future development directions for DMFC catalysts are outlined.

    Figure 2

    Figure 2.  Modification strategies for cathode electrocatalysts in DMFC.
    DownLoad: Full-Size Img PowerPoint

    2. Fundamental understanding of ORR mechanism

    ORR is considered one of the essential electrocatalytic reactions in energy storage and conversion systems, unraveling the intrinsic mechanism of ORR is crucial for the rational design and selection of electrocatalysts with tailored structures and properties [40,41]. The reaction primarily involves three steps: (1) Adsorption of O2 on the catalyst surface; (2) Electron transfer and O2 reduction; (3) Desorption of reduction products. These processes involve the adsorption and desorption of oxygen-containing intermediates, cleavage and formation of chemical bonds, as well as protons and electrons transfer [42]. The adsorption of O2 on the electrode surface is the initial and rate-determining step of the ORR, making it essential to understand the possible adsorption modes. In general, three distinct adsorption modes are recognized on the catalyst surface, mainly the Pauling model, Bridge model, and Griffiths model, as illustrated in Fig. 3a [43]. In the Pauling model, O2 is adsorbed in an "end-on" fashion, with one oxygen atom bonds to a single site, this mode weakly perturbs the O—O bond and often leads to the 2e- reduction pathway, producing H2O2. Bridge model is known as "side-on" manner, in which both oxygen atoms bond to two adjacent sites and favors the 4e- pathway toward H2O. In the Griffiths model, the oxygen atoms bound laterally to a single active site. This facilitates O—O bond activation and cleavage, similar favoring the 4e- pathway. In summary, the adsorption mode of O2 at the active site critically determines the ORR pathway selectivity.

    Figure 3

    Figure 3.  (a) Different O2 adsorption modes on the catalytic surface and (b) the ORR reaction pathways via 2e- and 4e-. Reproduced with permission [43]. Copyright 2024, Royal Society of Chemistry.
    DownLoad: Full-Size Img PowerPoint

    Typically, the ORR typically proceeds via either a 4e- pathway to H2O, which is ideal for fuel cells and metal-air batteries, or a 2e- pathway to H2O2, which is desirable for environmental remediation and other applications (Eqs. 1–6).

    (1) 4e- pathway:

    Acid medium:

    O2+4H++4e−→2H2O,E0=1.229V

    		(1)

    Alkaline medium:

    O2+2H2O+4e−→4OH−,E0=0.401V

    		(2)

    (2) 2e- pathway:

    Acid medium:

    O2+2H++2e−→H2O2,E0=0.670V

    		(3)

    H2O2+2H++2e−→2H2O,E0=1.770V

    		(4)

    Alkaline medium:

    O2+H2O+2e−→OH−+HO2−,E0=−0.065V

    		(5)

    HO2−+H2O+2e−→3OH−,E0=0.867V

    		(6)

    The detailed reduction of O2 to H2O can be divided into three distinct pathways: Associative pathway (Eqs. 7–11), dissociative pathway (Eqs. 12–14), and 2nd associative pathway (Eqs. 15–19), as shown in Fig. 3b.

    (1) Associative pathway:

    O2+*→O2*

    		(7)

    O2*+H++e−→OOH*

    		(8)

    OOH*+H++e−→O*+H2O

    		(9)

    O*+H++e−→OH*

    		(10)

    OH*+H++e−→H2O+*

    		(11)

    (2) Dissociative pathway:

    O2+2*→2O*

    		(12)

    2O*+2H++2e−→2OH*

    		(13)

    2OH*+2H++2e−→2H2O+2*

    		(14)

    (3) 2nd associative pathway:

    O2+*→O2*

    		(15)

    O2*+H++e−→OOH*

    		(16)

    OOH*+H++e−→HOOH*

    		(17)

    HOOH*→2OH*

    		(18)

    2OH*+2H++2e−→2H2O+2*

    		(19)

    where * represents an active site on the catalytic surface. The reaction kinetics and pathways of ORR are determined by the adsorption energy of intermediates on the catalyst surface. An ideal catalyst should exhibit moderate adsorption strength toward reaction intermediates. Excessively strong adsorption may occupy active sites and inhibit desorption, impairing reaction continuity, whereas overly weak binding may be insufficient to activate key species [44,45]. Therefore, rational catalyst design, which focuses on optimizing the free energy of oxygen-containing intermediates on catalytic surface sites, is essential for enhancing ORR performance.

    To evaluate the ORR catalytic performance, a rotating disk electrode (RDE) or rotating ring-disk electrode (RRDE) is typically employed in combination with an electrochemical workstation. Common electrochemical techniques include cyclic voltammetry (CV), linear sweep voltammetry (LSV), chronoamperometry (CA), and electrochemical impedance spectroscopy (EIS). Key evaluation metrics for catalytic activity comprise the onset potential (Eonset), half-wave potential (E1/2), electron transfer number, stability, Tafel slope, electrochemically active surface area (ECSA), kinetic current density (Jk), and limiting current density (JL).

    3. Optimization of ORR performance in DMFC by various regulation strategies

    The ORR process is fundamentally determined by the electronic structure and physicochemical properties of the catalyst interface, which dictate the adsorption and desorption behavior of oxygen-containing intermediates. To enhance electrocatalytic performance, various strategies, such as heterointerface engineering, defect engineering, composite design, and surface functionalization, have been extensively explored [46–49]. Specific methods like heteroatom doping, size confinement, surface reconstruction, and pore structure control are employed to tailor the electronic environment, morphological features, active sites, and porosity of catalysts, thereby improving reaction kinetics. Ideal catalysts should exhibit high specific surface area, abundant porosity, excellent conductivity, and well-exposed active sites, along with stability in acidic/alkaline environments, resistance to methanol/CO poisoning, and avoidance of mixed potentials [50,51]. In addition to experimental and advanced characterization techniques, theoretical studies play a key role in high-throughput catalyst screening and mechanistic elucidation [52,53]. Establishing a clear structure-performance relationship and understanding how various modification strategies regulate ORR activity are essential for the rational design of efficient DMFC catalysts. Tables S1 and S2 (Supporting information) summarize the recent advances in catalytic materials, highlighting their key electrochemical performance metrics across different media.

    3.1 Methanol tolerance optimization

    Precious metal catalysts (such as Pt, Pd) typically exhibit high electrocatalytic activity and are widely used as benchmarks. However, their application in DMFC is limited by methanol crossover poisoning, where methanol permeates from the anode to the cathode through the membrane, severely degrading catalyst performance [23,54]. The development of precious metal-based catalysts aims to simultaneously increase active site density and improve methanol tolerance. Nanostructure engineering and surface modification have been demonstrated to significantly enhance both the catalytic activity and methanol resistance of Pt-based catalysts. Furthermore, the improvement in ORR performance has been attributed to strain effects and multi-component synergy [32,55]. Meanwhile, non-precious metal catalysts typically exhibit intrinsic methanol tolerance, though the catalytic activity still requires further improvement.

    DMFC commonly use low-concentration methanol to suppress crossover and sustain performance [23]. However, high methanol concentrations are essential for competitive energy density. To overcome this limitation, Feng et al. [56] designed selective electrocatalysts that facilitate operation under high methanol concentrations. The resulting Au@Pd catalyst demonstrated enhanced ORR activity due to Au-Pd synergy, including electronegativity-driven electronic coupling and lattice mismatch-induced strain effects. The assembled DMFC achieved a peak power density of 89.7 mW/cm2 operating at 80 ℃ with 10 mol/L methanol aqueous solution, while maintaining high performance even at 15 mol/L. Tetrahedral Pd nanocrystals often show enhanced ORR activity in alkaline media owing to the high surface-to-volume ratio and exposed (111) facets [57]. For instance, Wang et al. [58] synthesized uniform truncated Pd tetrahedrons (T-Pd-Ths) enclosed by (111) facets, which feature low-coordination sites and lattice distortion that introduce localized strain. The T-Pd-Ths/C catalyst exhibited specific and mass activities 10.7 and 14.1 times higher than those of commercial Pt/C, respectively, along with high CO tolerance and only 7.69% activity loss after 50,000 cycles. Theoretical simulations confirmed that low coordination sites enhanced ORR performance. A DMFC using T-Pd-Ths/C achieved an open-circuit voltage of 0.927 V and a peak power density of 163.7 mW/cm2 across tested temperatures. Ultrathin carbon layers can effectively anchor Pt nanoparticles onto carbon supports, suppressing their migration, aggregation, detachment, and dissolution, while also inhibiting methanol penetration [59]. For example, Tong et al. [60] developed an in-situ polymer nanoencapsulation-graphitization strategy to prepare carbon-coated Pt/CNTs@C catalysts. The carbon layer thickness, adjustable from 0.5 nm to several nanometers via controlled polymer growth, was found to critically influence ORR activity. The optimized Pt/CNTs@C exhibited high catalytic activity, durability, and methanol tolerance. In a related approach, Lee et al. [61] encapsulated individual Pt nanoparticles within graphitic carbon shells (Pt@CS/CNF). As shown in Figs. 4a and b, transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) images revealed uniformly dispersed Pt nanoparticles (~3.8 nm) within the nanofibers, with the surrounding carbon shell effectively preventing aggregation. The carbon shell significantly blocked methanol access to the Pt core, an effect enhanced by higher graphitization, while still permitting O2 permeation to facilitate the ORR process (Fig. 4c).

    Figure 4

    Figure 4.  (a) TEM and (b) HRTEM images of Pt@CS/CNF. (c) Mechanisms of methanol tolerance and enhanced oxygen reduction. Reproduced with permission [61]. Copyright 2020, American Chemical Society. (d) Schematic of electrodes composed of Pt/C and Fe-N-C. (e) Mechanism schematic of DMFC based on composite cathode catalytic layer. Reproduced with permission [64]. Copyright 2024, Elsevier. (f) Adsorption configuration of methanol on N—C and FeN4 sites obtained by DFT calculations. (g) The adsorption energy of methanol on N—C, FeN4, CoN4, and MnN4 active sites calculated at different potentials. Reproduced with permission [66]. Copyright 2020, Royal Society of Chemistry.
    DownLoad: Full-Size Img PowerPoint

    To address the trade-off between oxygen catalytic activity and poisoning tolerance, Zhang et al. [62] successfully synthesized a class of Pd-Te hexagonal nanoplates (Pd-Te HPs) featuring a Pd20Te7 phase. Atomic force microscopy (AFM) image showed that the Pd atoms deviated from the hexagonal close-packed arrangement, resulting in exposed bridge sites that serve as catalytic centers. The Pd-Te HPs/C catalyst demonstrated superior ORR activity with E1/2 of 0.90 V, long-term stability (retaining 65% activity after 50,000 s), and exceptional methanol tolerance. Density functional theory (DFT) calculations revealed that the enhanced activity originates from the breakthrough in the linear relationship between OOH* and OH* adsorption, while the MOR remained highly endothermic. Among strategies to mitigate methanol crossover, structural modifications of the membrane electrode assembly (MEA) have shown promising results. For example, Kim et al. [63] designed a hierarchically structured cathode with an inner Pt-black layer to shield active sites from methanol and an outer Pt/C layer to enhance the electrochemical active surface area. This graded structure mitigates methanol crossover by enhancing ORR activity. Inspired by the Fenton activity of Fe-N-C, Hou et al. [64] systematically optimized electrochemical performance using composite cathode architectures (Fig. 4d). The best cell configuration composed of a layered cathode with Pt/C outside and Fe-N-C inside achieved a peak power density of 24.55 mW/cm2 with 3 mol/L methanol solution at 30 ℃. As illustrated in Fig. 4e, the Fe-N-C layer produced H2O2 via the 2e- pathway, which was then decomposed by both Fe-N-C (Fenton reaction) and Pt/C, generating reactive oxygen species (ROS). These ROS may chemically oxidize methanol that crosses over from the anode, suppressing its electrochemical oxidation on Pt and thus preserving cell performance.

    To address the issues of high methanol permeability and low conductivity in polymer electrolyte membranes, Tan et al. [65] proposed a novel strategy employing methanol-tolerant NiCo2O4 cathode catalysts combined with a permeable polymer fiber membrane to improve DMFC performance. LSV and CA tests indicated that the E1/2 of NiCo2O4 was comparable to commercial Pt/C catalysts, alongside significantly enhanced methanol tolerance. The permeable polymer fiber membrane-based DMFC achieved a maximum power density of 126.2 mW/cm2 at 60 ℃. Shi et al. [66] systematically investigated the adsorption behavior of methanol on various active sites at U = 0 and 0.8 V vs. reversible hydrogen electrode (RHE) using stable adsorption modes (Fig. 4f). Results showed that methanol adsorption strength on MN4 sites (M = Fe, Co, Mn) correlates strongly with electrode potential, especially at higher potentials (Fig. 4g). Importantly, methanol adsorption neither disrupts the structure of FeN4 sites nor interferes with the 4e- ORR pathway. Leveraging this intrinsic methanol tolerance, an atomically dispersed Fe/Co-N-C catalyst through a Co-doping and Fe-adsorption strategy was synthesized. Co was found to modulate pore structure and generate defects for anchoring FeN4 sites. The resulting DMFC, operating in a methanol-air system, achieved a peak power density of 135 mW/cm2 using 1 mol/L methanol solution. This work offers a new approach to designing dual-site catalysts with high activity and superior methanol tolerance.

    Currently, although modification strategies for precious metal catalysts, such as core-shell structures, high-index facet exposure, surface modification, selective catalytic layers, and functionally graded electrode design, can partially mitigate the methanol crossover effect, it remains challenging to achieve both high ORR activity and ideal methanol tolerance simultaneously. While non-precious metal catalysts exhibit inherent methanol tolerance, the ORR activity still significantly lags behind that of precious metal benchmarks. Consequently, the development of novel catalyst systems that combine high activity and strong tolerance continues to pose a persistent challenge [67,68].

    3.2 Rational design of active sites

    Rational modulation at the atomic/molecular level enables the precise construction of active centers with tailored geometric configurations, electronic structures, and micro-environments, thereby facilitating directed optimization of catalytic activity, selectivity, and stability. In electrocatalysis, the evolution of active sites from the nanoscale to the single-atom level induces significant changes in geometric and electronic structures, leading to distinct catalytic behaviors. Atomically dispersed catalysts, typically consisting of metal sites anchored atomically on carbon supports, have attracted extensive research interest due to the 100% atom utilization efficiency [69–71]. Among them, M-N-C materials (M = Fe, Co, Ni, etc.), formed by embedding transition metal centers into N-doped carbon matrices, have been widely demonstrated to derive the high ORR activity primarily from the metal sites atomically dispersed and coordinated with N atoms [72,73]. Thus, M-N-C materials are emerging as a mainstream direction in the design and fabrication of ORR catalysts.

    Currently, Fe-N-C is evolving into one of the most representative M-N-C materials in DMFC due to its exceptional catalytic efficiency and intrinsic tolerance to methanol/CO poisoning [74–77]. Recent studies show that the confinement engineering based on metal-organic framework (MOF) can stabilize single atoms via spatial site isolation to prevent particle agglomeration. Moreover, such confinement engineering readily modulates the coordination environment and electronic structure of active sites via chemical bonding, thereby conferring enhanced catalytic activity [78,79]. Xu et al. [80] reported an atomically dispersed ZIF/MIL-10–900 catalyst with FeN4 configuration derived from a MOF mixture of institute lavoisier-101 (MIL-101) and zeolite imidazolate frame-8 (ZIF-8), which enabled the DMFC to achieve an enhanced peak power density of 83 mW/cm2 operating with 3 mol/L methanol solution at 80 ℃. The discrete distribution of active sites necessitates multiscale pore structures to ensure accessibility and efficiency. Shi et al. [81] constructed a hierarchically porous FeNC-DT catalyst with smooth morphology via a dual-template strategy. The as-prepared catalyst demonstrated a remarkable E1/2 of 0.954 V and low Tafel slope of 49.3 mV/dec in alkaline media. To tackle the challenges of insufficient active sites and restricted mass transport at triple-phase boundaries, Shu et al. [82] innovatively designed a core-shell structured Fe-mPDA@HPCS catalyst, which consists of hierarchically porous carbon spheres (HPCS) as the core and single-Fe-atom-modified m-phenylenediamine (Fe-mPDA) as the shell. By precisely engineering the pore architecture and strategically incorporating atomic Fe sites (Fig. 5a), the catalyst synergistically enhanced both ORR activity and mass transfer capability.

    Figure 5

    Figure 5.  (a) Schematic of Fe-mPDA@HPCS cathode structure with high-efficiency mass transfer. Reproduced with permission [82]. Copyright 2021, Wiley. (b) Proposed the possible structure of Fe/NS-C. (c) ORR polarization curves. Reproduced with permission [85]. Copyright 2023, Elsevier. (d, e) HAAFF-STEM images. (f, g) Acquired HAADF-STEM intensity line profiles. (h, i) Simulated Mg-Co HAADF-STEM image. (j) Proposed Mg-N-Co moiety in Mg-Co. Reproduced with permission [86]. Copyright 2023, Elsevier. (k) Linear correlation between -ICOHP and Bader valence electron number (N) in different single atom particle catalyst systems (left) and the dissolved Fe content for Fesac, Au-Fesac, and Pt-Fesac (right). (l) Long-term stability test at a constant current density of 0.1 A/cm2. Reproduced with permission [87]. Copyright 2023, American Chemical Society.
    DownLoad: Full-Size Img PowerPoint

    Previous studies indicate that the suboptimal adsorption energy of oxygen intermediates, induced by the four symmetric electronegative N atoms surrounding Fe centers, fundamentally restricts the catalytic activity of Fe-N-C materials. Notably, partial substitution of N with heteroatoms through secondary coordination engineering has proven effective in modulating electronic structures and enhancing reaction kinetics [83,84]. Cui et al. [85] utilized trithiocyanuric acid (TCA)-coated ZIF-8 as precursor to synthesize a hollow porous Fe, N, S-codoped carbon catalyst (Fe/NS-C). In this system, TCA simultaneously enhances N content and introduces S as a secondary dopant, thereby not only providing abundant anchoring sites for Fe single atoms but also modulating the electronic structure. The catalyst features active sites with FeN4 and FeN3S coordination configurations (Fig. 5b). Owing to the high intrinsic catalytic activity of FeN3S sites and the fully exposed active surfaces, the Fe/NS-C material delivered exceptional ORR performance with a E1/2 of 0.893 V and Jk of 41.18 mA/cm2, superior to commercial Pt/C (Fig. 5c).

    The suboptimal binding strength of reaction intermediates significantly impedes the ORR kinetic. Although dual-site catalytic mechanism can partially alleviate this limitation by synergistically facilitating oxygen dissociation and O—O bond cleavage, the precise control over the geometric and electronic structures of these dual sites that determine catalytic efficiency remains a fundamental challenge. Wang et al. [86] developed a Mg-Co dual-atom catalyst via a pyrolysis-impregnation-pyrolysis method, featuring densely populated active sites that significantly enhanced both ORR activity and DMFC performance in acidic media. As shown in Fig. 5d, HAADF-STEM image revealed Mg-Co dual sites, with brighter spots corresponding to Co atoms due to the higher atomic number compared to Mg. Atomic-resolution images (Fig. 5e) and intensity line profiles (Figs. 5f and g) further confirmed the atomic dispersion of metal species without long-range ordering. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) simulations suggested the possible presence of Mg-N2–Co configurations in the dual-site structure (Figs. 5h and i). Simulation and experimental results deduced the possible a Mg-N2–Co moiety in Mg-Co atomic pairs (Fig. 5j). The Mg-Co-based DMFC achieved a peak power density of 200 mW/cm2 and exhibited negligible performance decay after 12 h of operation in a 10 mol/L methanol solution at 80 ℃. Precise engineering of local coordination environments enables the rational design of high-performance catalytic systems.

    Research on M-N-C catalysts has predominantly emphasized activity and selectivity, with comparatively less attention given to the electrochemical dissolution of metal sites. This is particularly concerning since metal leaching not only contaminates the electrolyte but also degrades active sites, ultimately leading to performance degradation. Although interactions between metal nanoparticle and single atom have been shown to enhance the stability of single-atom catalysts and mitigate metal leaching, the underlying stabilization mechanisms remain poorly understood. Herein, Gao et al. [87] proposed and demonstrated a universal mechanism by which metal nanoparticles suppress the dissolution of Fe single atom sites. As illustrated in Fig. 5k, the projected crystal orbital Hamilton population (-pCOHP) was employed to quantitatively evaluate the strength of Fe-N bonds based on its negative values. A positive linear correlation was observed between the integrated negative of the -pCOHP (-ICOHP) and N values, a higher N value corresponds to greater electron density at Fe sites and lower Fe oxidation states, while a higher -ICOHP value indicates stronger Fe-N bonds, thereby inhibiting demetalation. This correlation confirms that metal nanoparticles act as electron donors, enhancing Fe-N bond strength by increasing electron density at FeN4 centers and thus mitigating Fe dissolution. Electrochemical results revealed a 78% reduction in Fe dissolution with the nanoparticle-decorated Fe single-atom catalyst, enabling continuous DMFC operation for 430 h at a constant current density of 0.1 A/cm2 (Fig. 5l).

    Through strategies such as confinement engineering, optimization of the coordination micro-environment, increased active site density, and suppression of demetalation, significant progress has been made in the intrinsic activity of catalysts. However, the inherently high surface energy of monodisperse site catalysts still readily leads to migration and aggregation phenomena. Coupled with the generally low metal loading in such catalysts, these issues severely restrict the accessibility and effective exposure of active sites. Furthermore, the dynamic changes in active site structure during catalytic reactions and the difficulties in characterization add to the complexity of understanding structure-activity relationships. Therefore, achieving simultaneous improvements in both stability and metal loading while maintaining high activity remains a fundamental challenge in the development of single-atom catalysts.

    3.3 Electronic structure tailoring

    The electrocatalytic activity of materials is primarily governed by the surface electronic structure. Typically, the electronic structure of catalysts can be effectively optimized by regulating the composition, defects, and heterostructures. Such electronic structure modulation strategies enable precise adjustment of the electron states, band structure, and surface adsorption capacity of catalysts. For instance, approaches including heteroatom doping, strain engineering, alloying, and heterointerface construction have become critical pathways for optimizing electronic structures and enhancing catalytic performance [88–91].

    Carbon-based materials are widely studied as ORR electrocatalysts due to the high surface area, low cost, and tunability. However, the intrinsic activity is limited and often requires modification. Heteroatom doping, particularly with N, effectively tunes the morphology, electronic structure, and chemical properties of carbon catalysts. The electronegativity difference may induce charge redistribution, break the electrical neutrality of the carbon matrix and creating charged active sites that promote O2 adsorption and reduction [92]. Beyond the catalytic role, carbon materials also function as excellent catalyst supports. Graphene is a highly conductive carbon allotrope ideal for framework materials with excellent thermal and electrical conductivity. Farzaneh et al. [93] hydrothermally synthesized a novel three-dimensional (3D) mesoporous N-doped reduced graphene oxides (NRGOs) with tunable N content (1.0%−4.7%) by controlling melamine/graphene oxide ratios. It was found that N-doping activates π-electrons to enhance low-potential ORR, while pyridinic-N boosts high-potential ORR via lone electron pairs. Integrating synergistic heteroatoms can also adjust spin density and charge polarization to enhance catalytic performance. Liu et al. [94] reported a 3D porous carbon nanosphere doped with O/N atoms, with diameters below 200 nm and specific surface area exceeding 659 m2/g, which displayed similar Eonset and E1/2 compared to commercial 40 wt% Pt/C.

    Alloying has attracted significant research interest due to its ability to alter the local electronic environment of metals and induce electronic structure reconstruction. Varying alloy compositions lead to distinct electronic and physicochemical properties, which in turn yield diverse catalytic performances. Current design of alloy catalysts focuses on composition optimization, size control, morphology engineering, and structural design [95–97]. Guided by these strategies, researchers have performed systematic and in-depth studies on alloy catalysts. Introducing transition metals into binary or multicomponent alloys can effectively modulate electronic and ligand effects. Gavidia et al. [98] synthesized various carbon-supported Pd-based alloys (PdFe, PdIr, and PdFeIr) via borohydride method with precisely controlled compositions. The results demonstrated that the charge transfer from Fe/Ir to Pd played a pivotal role in performance enhancement. Li et al. [99] employed an "activity-assisted" strategy to synthesize branched one-dimensional PdBiAu alloy nanochains (size < 6.5 nm) with abundant interfaces. The d-band center of Pd significantly downshifted due to the electronic modulation by adjacent Bi/Au species, weakening the adsorption strength of oxygen intermediates. Benefiting from the synergistic effects, including rich surface defects, lattice strain, optimized d-band center, enhanced methanol tolerance and charge transfer capability, the DFMCs modified with Pd0.60Bi0.35Au0.05/C catalyst achieved remarkable power densities of 140.1 and 112.4 mW/cm2 under 6 mol/L methanol solution when operating in pure O2 and air atmospheres, respectively. To overcome the kinetic limitations of Pd, Sathyaseelan et al. [100] incorporated oxyphilic Ni atom into Pd, successfully synthesizing a sphere-like PdNi alloy (sPdNiA). The incorporation of Ni into the Pd lattice facilitates electron transfer from Ni to Pd, leading to a downshift of the d-band center due to the alloying effect, which confirms the strain effect induced by doping. Fig. 6a illustrated the schematic configuration of a DMFC employing sPdNiA as a bifunctional catalytic layer. As the operating temperature increases, the reaction kinetics accelerate significantly, resulting in a notable improvement in open-circuit voltage and peak power density, reaching 0.68 V and 34 mW/cm2 at an operating temperature of 60 ℃, respectively (Fig. 6b).

    Figure 6

    Figure 6.  (a) Schematic representation of the DMFC assembled with the sPdNiA as bifunctional electrodes. (b) Polarization curves of the DMFC at various operating temperatures. Reproduced with permission [100]. Copyright 2023, American Chemical Society. (c) Electron transfer model on RuO2/Pd(111) interface. (d) Reaction free energy diagram of ORR at different catalytic sites. Reproduced with permission [101]. Copyright 2021, Wiley. (e) Schematic of ORR mechanism for the enhanced electrocatalytic activity of the Mn2O3@Co1.2Ni1.8O4. Reproduced with permission [106]. Copyright 2018, American Chemical Society.
    DownLoad: Full-Size Img PowerPoint

    Enhanced catalytic performance through interface engineering mainly results from strong metal-support interactions. Downsizing metal components from nano- to atomic-scale allows precise tuning of metal-oxide interfacial effects. Lyu et al. [101] reported a two-dimensional (2D) RuOx/Pd nanosheet featuring atomically dispersed interfaces that significantly enhance metal-oxide interactions. As shown in Fig. 6c, electron transfer from the Pd(111) surface to RuOx/Pd interface resulted in electron enrichment at the interface, leading to the formation of a dual-end adsorption configuration for O2. Free energy theoretical calculations further revealed that the optimal O2 adsorption configuration transformed into one oxygen atom adsorbed at the interfacial Ru site while the other bound to an adjacent Pd site after the formation of RuO2/Pd(111) interface (Fig. 6d). The dense RuOx/Pd interface effectively promoted O2 activation, O-O bond cleavage, and intermediate conversion processes. CeOx exhibits abundant surface oxygen species, excellent oxygen storage capacity, and variable valence states, which collectively contribute to its enhanced catalytic performance. As demonstrated by Wang et al. [102], the improved activity of PtMo-CeOx composite primarily stems from the ligand effect and synergistic interactions between Mo and CeOx nanoparticles. Leveraging the electronic modulation effect of low-electronegativity Ce on Pt and the catalytic effect of CeO2, Wan et al. [103] introduced Ce as a secondary component to tailor the electronic structure of Pt while constructing PtCe-CeO2 heterointerfaces to enhance reaction kinetics. The heterogeneous interaction between PtCe and CeO2 facilitated the spontaneous electron donation from Ce to Pt, thereby enhancing the ORR kinetics.

    Wang et al. [104] synthesized a Fe-urea-formaldehyde resins carbon/MnO2 composite (Fe-UFC/MnO2) via an in-situ redox approach. The resulting Fe-Nx moieties function as effective electrocatalytic active sites. Moreover, Fe incorporation narrows the band gap, promotes the conversion of Mn3+ to Mn4+, and directly supplies free electrons to oxygen-adsorbed species, significantly enhancing ORR efficiency. Liu et al. [105] prepared bimetallic NiCo2O4 with 3D dandelion-like (DL) and flower-like (FL) morphologies via a facile hydrothermal method. Electrochemical tests indicated electron transfer numbers of 3.97 and 3.91 under alkaline conditions, respectively. DMFC performance improved considerably with increasing operating temperature, reaching maximum power densities of 151 and 114 mW/cm2 at an operating temperature of 65 ℃ when assembled with NiCo2O4-DL and NiCo2O5-FL, respectively. Mechanistic studies revealed that Co3+ acts as an electron donor, directly supplying excited electrons to O2 for reduction, while Co4+ serves as an electron acceptor from the external circuit and is reduced to Co3+, establishing an efficient electron exchange process. In addition, Liu et al. [106] developed a novel Mn2O3@Co1.2Ni1.8O4 hybrid by decorating ultrathin and highly conductive Co1.2Ni1.8O4 nanosheets onto Mn2O3 surfaces. The hybrid exhibited exceptional methanol tolerance even in high-concentration methanol solution and delivered a remarkable power density of 70.5 mW/cm2 at near room temperature. As shown in Fig. 6e, the enhanced electrocatalytic performance primarily stems from the narrow bandgap of hybrid material, which facilitates electron transition from the valence band to conduction band.

    Optimizing the electronic structure of catalysts through strategies such as doping, alloying, and heterostructure engineering serves as a core approach to enhance catalytic performance by modulating the charge distribution and band structure of active sites. This method offers significant advantages, including effectively reducing reaction energy barriers, optimizing the adsorption/desorption behavior of intermediates, and enhancing intrinsic activity and selectivity. Nevertheless, this strategy still faces multiple challenges, such as the complexity of electronic modulation mechanisms, difficulties in precisely controlling doping sites and concentrations, susceptibility to phase separation and component segregation, as well as insufficient interfacial stability and charge transfer efficiency.

    3.4 Surface and interface micro-environment engineering

    Traditional catalyst design has primarily focused on electronic structure optimization, geometric structure modulation and component engineering. However, as a typical electrode-electrolyte interfacial reaction, the catalytic performance of ORR is profoundly influenced by the macro- and micro-environment at the interface, including factors such as pore structure, surface geometry, hydrophilicity/hydrophobicity, and the degree of active site exposure. These strategies collectively enhance reaction kinetics and structural stability by improving the effective specific surface area, electrical conductivity, mass transport efficiency, electronic structure, interfacial free energy, as well as the density and accessibility of active sites.

    Porous carbon offers a high surface area and abundant pore structures, which facilitate mass transfer and increase the accessibility of active sites. However, conventional synthesis methods typically yield mesoporous carbon with limited pore size, large particles (up to micrometers), and a tendency to aggregate, all of which restrict mass transport in fuel cell applications. As shown in Fig. 7a, Shu et al. [107] utilized N-rich m-phenylenediamine to modulate N active sites and pore structure during the synthesis of mesoporous N-doped carbon spheres (MNDCS). The incorporation of m-phenylenediamine effectively modulated precursors nucleation, with the surface structure directly dependent on loading amount. The surface/interface of carbon spheres exhibited abundant functional N species (~6.5 wt%), predominantly pyrrolic N and pyridinic N species. The optimized micro/mesoporous ratio combined with enhanced N-containing species significantly improved the ORR activity. The MNDCS-0.3 sample (30 wt% m-phenylenediamine) achieved a high E1/2 of 0.81 V (Fig. 7b). Due to electrode structural constraints and prolonged operation, cathode water flooding progressively worsens, significantly impairing catalyst activity and overall fuel cell performance. This issue underscores the need to construct robust oxygen transport channels tailored for ORR. Shu et al. [108] synthesized N-doped yolk-shell carbon spheres (N-YS-CS) with a distinctive core-shell structure, featuring precisely controlled mesoporous surfaces and well-defined core-shell boundaries. Fig. 7c illustrated the membrane electrode assembly in which the PTFE-treated N-YS-CS exhibited high hydrophobicity and flexibility, making it an ideal hydrophobic cathode material. The mesoporous framework of N-YS-CS effectively accommodated hydrophobic agents and formed robust hydrophobic channels within the structure, resulting in a remarkable power output of 142 mW/cm2 at an operating temperature of 60 ℃.

    Figure 7

    Figure 7.  (a) HAADF-STEM of MNDCS-0.3. (b) ORR polarization curves. Reproduced with permission [107]. Copyright 2020, Wiley. (c) Schematic of the triple-phase interface with N-YS-CS hydrophobic cathode structure. Reproduced with permission [108]. Copyright 2017, Elsevier. (d) Scheme of the microstructure for Fe/N/C cathodes. Reproduced with permission [110]. Copyright 2017, American Chemical Society. (e) Schematic of structural advantages. (f) Power density curves of DMFCs. (g) Short-term stability tests of DMFCs at 100 mA/cm2. Reproduced with permission [111]. Copyright 2021, American Chemical Society.
    DownLoad: Full-Size Img PowerPoint

    Xu et al. [109] developed a microwave-assisted method to rapidly synthesize Fe-doped ZIF-8 precursor, which was further etched with NH3 under a protective atmosphere to obtain the M15-FeNC-NH3 catalyst with a hierarchical micro/mesoporous structure. NH3 etching treatment not only generated new micropores but also expanded some existing micropores into mesopores, while simultaneously increasing the specific surface areas of micropores and mesopores to 749.9 and 554.6 m2/g, respectively. The interconnected porous network facilitated efficient mass transport of reactants, enabling the resulting DMFC to achieve a peak power density of 61 mW/cm2 and maintain 95.6% voltage retention after 24 h with a 0.5 mol/L methanol solution. Although catalytic active sites are typically located within micropores, these pores are highly susceptible to water flooding, which obstructs oxygen diffusion to the active sites and disrupts the three-phase interface. Wang et al. [110] introduced hydrophobic dimethyl silicone oil (DMS) as an additive into Fe/N/C catalyst to create a microporous environment that balances hydrophobicity and mass transfer. When the molecular weight of DMS reached 14 kDa, its distribution within the micropores was optimally regulated, effectively mitigating water flooding (Fig. 7d). Polarization curves revealed that DMFC performance initially improved and then declined with increasing DMS molecular weight, peaking at 14 kDa. Under this optimized condition, the DMFC achieved a peak power density of 102 mW/cm2 using a 3 mol/L methanol solution, matching the performance of Pt/C-based DMFC.

    In practical applications of Fe-N-C catalyst, a significant portion of inactive FeNx sites is embedded deep within the carbon matrix, increasing mass transport resistance and limiting active sites utilization. As shown in Fig. 7e, Xu et al. [111] adopted a size-confinement strategy to anchor EDTA-Fe groups on the external surface of ZIF-8, followed by pyrolysis to obtain ZIF@EDTAFe-1%−950. The resulting material features exposed FeN4 sites within an epitaxial growth layer. Electrochemical analysis indicated an exceptional active site utilization efficiency of 59.6%. The highly accessible Fe-N-C configuration, combined with optimized mass transport, enabled the DMFC to reach a peak power density of 71 mW/cm2 at 0.5 mol/L methanol solution (Fig. 7f). Additionally, PVDF modification of the catalyst layer formed robust hydrophobic channels around the active sites, effectively mitigating flooding. This configuration demonstrated outstanding durability, maintaining 83.9% voltage retention after 24 h of operation at 100 mA/cm2 (Fig. 7g).

    Surface and interface micro-environment engineering effectively enhances catalytic performance by modulating the local physicochemical environment around active sites. The advantages are primarily demonstrated through the construction of hierarchical pore channels, optimization of interfacial electronic structures, and regulation of hydrophilicity/hydrophobicity, which collectively improve reaction mass transfer, increase the accessibility of active sites, and enhance reaction selectivity, thereby significantly boosting catalytic efficiency and stability. However, this strategy still exhibits certain limitations. For instance, precise control of the micro-environment often relies on complex multi-step synthesis, posing challenges for scalable production and reproducibility. Moreover, the dynamic evolution mechanisms of the micro-environment under practical operating conditions remain unclear, and current characterization techniques are inadequate for real-time resolution of multiphase interfacial processes, hindering in-depth understanding of structure-activity relationships and rational design.

    3.5 Additional modulation approaches

    The activity of catalysts can be effectively modulated through multi-scale strategies ranging from macro to micro levels, encompassing various dimensions such as morphology, pore structure, particle size, phase composition, surface and interface modification, active site micro-environments, and electronic structure. The mechanisms of strain effects, defect engineering, phase transformation, and ensemble effects on electrocatalytic activity are systematically summarized, with their practical applications discussed in this section.

    Tremendous works have demonstrated that electronic and geometric structures are the key influencing factors that determines catalytic activity [42,112–115]. Meanwhile, the use of highly conductive substrates helps improve the catalytic performance of precious metals by reducing resistance-induced polarization losses. For instance, further surface modification of carbon nanotube (CNT) can strengthen interaction with precious metals, thereby preventing the detachment, aggregation, and dissolution of active species [116]. Although defect formation on CNT promotes the growth of Pt nanoparticles, effective methods for precisely controlling such defects remain scarce. To address this, Hu et al. [117] employed plasma treatment on N-doped CNT, which generated abundant surface defects that facilitated the uniform dispersion and stabilization of Pt nanoparticles (p-Pt@p-NCNT). Meanwhile, plasma reduction resulted in Pt nanoparticles with a lower proportion of surface atoms in high oxidation states. As illustrated in Fig. 8a, p-Pt@p-NCNT consisted of Pt nanoparticles supported on bamboo-shaped NCNT featuring numerous defects. Pt 4f X-ray photoelectron spectroscopy (XPS) spectrum (Fig. 8b) and DFT calculations further revealed that plasma-treated surface provided favorable sites for Pt nanoparticles attachment, inducing strong electronic coupling between Pt and NCNT support. This modification optimized the surface oxidation state of Pt, thereby lowering the energy barrier of rate determining step in ORR. Consequently, the p-Pt@p-NCNT-based DMFC achieved a high peak power density of 81.9 mW/cm2 at an operating temperature of 80 ℃. Liu et al. [118] introduced trace amounts of Ru into Co nanoclusters, resulting in atomically dispersed Ru on Co. Benefiting from the ligand and strain effects, the catalyst demonstrated significantly enhanced ORR activity, achieving a E1/2 of 0.908 V and high selectivity toward the 4e- pathway. When tested in fuel cell, the catalyst demonstrated a 60% improvement in performance compared to commercial Pt/C, highlighting the strong potential for practical fuel cell applications.

    Figure 8

    Figure 8.  (a) Aberration-corrected HRTEM. (b) high-resolution XPS spectra of Pt 4f. Reproduced with permission [117]. Copyright 2023, Wiley. (c) Schematic of the CoFe@NCNF/BCNT synthesis. (d) The peak power densities obtained at different operating temperatures. Reproduced with permission [121]. Copyright 2023, Elsevier. (e) Structure diagram of flexible DMFC. (f) The electrochemical performance of DMFC under consecutive bending from flat to 180°. Reproduced with permission [12]. Copyright 2021, American Chemical Society.
    DownLoad: Full-Size Img PowerPoint

    Currently, systematic studies on hollow-structured PtM (M = transition metal) alloy catalysts remain scarce, especially in terms of the controllable synthesis of diverse alloy phases and architectures, as well as the understanding of how compositional and morphological synergies govern electrocatalytic performance. Through precise adjustment of precursor ratios and pyrolysis temperature, Fan et al. [119] realized gradient composition control and ordered atomic arrangement in PtNi alloy nanotube. Lattice contraction effect induced by Ni-doping (wall thickness: 19 nm) resulted in optimal electrocatalytic performance at a Ni/Pt atomic ratio of 2.2. The catalyst retained 91.2% of its initial mass activity after 10,000 accelerated durability test cycles. Zou et al. [120] systematically explored the structural evolution from disordered PtNi alloys to ordered PtNi intermetallic compounds and its impact on catalytic behavior. The ordered PtNi intermetallic nanoparticles demonstrated markedly improved ORR activity, durability, and methanol tolerance compared to the disordered counterparts.

    The metal ion ratio in precursors determines the composition and electronic structure of the resulting alloy nanoparticles. As shown in Fig. 8c, our group developed a nanoengineering strategy to synthesize a series of CoFe alloy catalysts confined in N-doped carbon nanofibers (NCNF) and bamboo-like carbon nanotubes (BCNT) by pyrolyzing precursors with precisely controlled Co/Fe molar ratios (1:0, 3:1, 1:1, 1:3, 0:1) [121]. Owing to optimized N coordination, precisely tuned alloy composition, and strong electronic coupling effects, the DMFCs assembled with Co1Fe1@NCNF/BCNT and Co1Fe3@NCNF/BCNT demonstrated high power densities of 29.10 and 31.11 mW/cm2 when operating at an operating temperature of 80 ℃, respectively (Fig. 8d). Ultra-thin 2D materials have garnered growing research attention owing to their distinctive performance advantages. Nevertheless, synthetic strategies for ultrathin 2D Pd-based nanomaterials with multicomponent compositions are still scarcely reported. In an innovative approach, Zhao et al. [122] developed a visible-light-induced templating method to fabricate quaternary PdAuBiTe alloy nanosheets (PdAuBiTe ANSs) with a thickness of approximately 5 nm. The prepared catalyst demonstrated exceptional electrocatalytic performance, which can be attributed to its abundant edge and surface defects, lattice strain effects, optimized d-band center, and ensemble effects. These advantages ultimately enabled the DMFC to achieve remarkable maximum power densities of 235.7 mW/cm2 (in O2) and 173.5 mW/cm2 (in air) at 10 mol/L methanol solution. Conventional DMFC are often hindered by bulky volume, high cost, and severe methanol crossover, which present major obstacles to the development of flexible DMFC devices. As shown in Fig. 8e, Li et al. [12] innovatively proposed a flexible DMFC design strategy employing a highly active Fe(Fc)-N/S-C cathode material integrated with a mechanically robust potassium polyacrylate hydrogel (PAK gel) electrolyte. Deformation-discharge tests demonstrated that the assembled flexible DMFC maintained a stable output voltage of approximately 0.458 V even after returning to its flat state (Fig. 8f). This design offers new insights for developing flexible DMFC that combine both low cost and high energy conversion efficiency.

    In the modification of electrocatalysts, multiple strategies such as strain effects, defect engineering, phase transformation, and ensemble effects work synergistically to regulate catalytic performance across various dimensions, including electronic structure, coordination environment, crystal phase, and active-site microstructure. Specifically, lattice strain can be introduced to modulate surface electronic structure; unsaturated coordination sites can be constructed to enhance local charge density and reaction activity; structural reconstruction can be utilized to expose highly active crystal facets; and synergistic mechanisms can be employed to promote efficient multi-step reaction processes. However, these strategies still face significant challenges, such as difficulties in precisely controlling defect concentration and distribution, the irreversibility of phase transformation processes, the reliance on precise assembly for complex structures, and the unclear synergistic mechanisms among multiple effects, which represent critical bottlenecks for further development.

    4. Conclusions and perspectives

    Advances in nanotechnology have opened new opportunities for the design and optimization of electrocatalysts. This review begins with an overview of the fundamental working principles of DMFC and the mechanisms of the ORR, then systematically summarizes recent advances in modification strategies for ORR catalysts. These include methanol tolerance optimization, precise design of active sites, electronic structure modulation, surface/interface micro-environment engineering, and other synergistic methods. Multi-scale strategies, such as core-shell structure, nanoengineering, heteroatom doping, alloying, metal-support interactions, defect engineering, and interface engineering are emphasized, with the mechanisms for improving catalytic activity and methanol tolerance discussed in detail. However, current ORR catalysts still require considerable improvements in structural design and overall performance. The following key challenges remain to be addressed:

    (1) The aforementioned activity modulation strategies primarily cover precious metal catalysts, carbon-based materials, single-atom catalysts, alloy catalysts, and metal oxides. For precious metal catalysts, it is necessary to reduce precious metal loading while maintaining catalytic activity and improving methanol tolerance. Carbon-based materials suffer from low activity and poor long-term stability, with the catalytic mechanism of heteroatom doping remaining unclear. Single-atom catalysts require enhanced stability and loading capacity, the ORR activity strongly depends on the M-N-C coordination environment and metal centers, making precise structural control critical. Alloy catalysts encounter issues such as poor synthesis controllability, component segregation, and high precious metal dependence, urgently demanding advanced synthesis techniques and stability enhancement strategies. Additionally, metal oxides exhibit low conductivity, necessitating surface defect/doping engineering, conductive network construction, crystal structure modulation, and interface engineering to synergistically improve both conductivity and ORR activity. Catalyst design aims to achieve high activity, high stability, and low cost simultaneously, yet balancing all three remains challenging. An ideal catalyst should combine high site density to achieve high activity while maintaining a low dissolution rate to ensure long-term stability.

    (2) With the advancement of catalytic science, traditional methods have become inadequate in deciphering the complex and non-linear structure-activity relationships between high-dimensional features of catalysts (e.g., chemical composition, morphology, size distribution, coordination environment) and the performance metrics including activity, selectivity, and stability. Recent breakthroughs in characterization techniques, including aberration-corrected transmission electron microscopy (AC-TEM), in situ Raman spectroscopy, and in situ X-ray absorption fine structure (XAFS) spectroscopy, now allow real-time tracking the coordination environment and valence changes of active sites, providing direct evidence for mechanisms behind ORR activity. DFT calculations complement experiments by revealing adsorption free energy changes of intermediates and electronic properties. Meanwhile, machine learning identifies patterns in data to build predictive models and guide inverse catalyst design, uncovering new mechanisms and shifting research from trial-and-error to precision approaches. Together, these integrated techniques deepen understanding of catalytic processes and establish a robust foundation for designing next-generation high-performance catalysts.

    (3) The ORR catalysts developed in laboratories often demonstrate superior intrinsic activity compared to commercial Pt/C in three-electrode system, primarily due to the idealized testing environment (including high O2 solubility, ultrathin catalyst layers, and efficient mass transport conditions), which maximizes active site utilization and optimizes ion/electron transport pathways. However, in practical DMFC operation, complex gas-liquid-solid interfaces, mass transport resistance caused by thicker catalyst layers, electron/proton conduction imbalance, and material degradation during long-term operation (e.g., support corrosion and active metal leaching) significantly limit the practical performance of electrocatalysts. In contrast, precious metal catalysts still dominate the fuel cell field due to the mature electrode engineering technology. This performance gap highlights the critical technical challenges in transitioning from intrinsic material properties to device-level engineering applications. Three-electrode measurements are essential in basic research for rapid material screening and mechanistic studies. In applied research and development, integrated testing of flexible/wearable DMFC is crucial to evaluating device performance, system compatibility, and practical engineering challenges.

    CRediT authorship contribution statement

    Shiquan Guo: Writing – original draft, Writing – review & editing, Investigation, Conceptualization. Fei Chen: Formal analysis, Methodology. Congju Li: Writing – review & editing, Supervision, Funding acquisition.

    Declaration of competing interest

    The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

    Acknowledgments

    This work is supported by the Beijing Municipal Natural Science Foundation (No. 2252043), the National Natural Science Foundation of China (No. 52170019), the Fundamental Research Funds for the Central Universities (No. 06500100), and the “Ten Thousand Plan”-National High-level personnel of special support program.

    Supplementary materials

    Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2026.112408.


    1. [1]

      F. Creutzig, Nature 606 (2022) 460–462. doi: 10.1038/d41586-022-01616-z

    2. [2]

      M. Murunga, C. Macleod, G. Pecl, Nat. Clim. Change 14 (2024) 126–133. doi: 10.1038/s41558-023-01904-0

    3. [3]

      K. Yuan, T.T. Zhang, X.M. Xie, et al., J. Cleaner Prod. 414 (2023) 137590. doi: 10.1016/j.jclepro.2023.137590

    4. [4]

      Y.Q. Yang, S. Bremner, C. Menictas, M. Kay, Renew. Sustain. Energy Rev. 167 (2022) 112671. doi: 10.1016/j.rser.2022.112671

    5. [5]

      F. Belaïd, A. Al-Sarihi, R. Al-Mestneer, Renew. Energy 205 (2023) 534–542. doi: 10.1016/j.renene.2023.01.083

    6. [6]

      M.L. Yue, H. Lambert, E. Pahon, et al., Renew. Sustain. Energy Rev. 146 (2021) 111180. doi: 10.1016/j.rser.2021.111180

    7. [7]

      Z.T. Li, X.T. Xu, X.R. Sheng, et al., ACS Nano 15 (2021) 12535–12566. doi: 10.1021/acsnano.1c01590

    8. [8]

      Y.D. Lei, Z.L. Wang, D.Y. Wang, et al., Nat. Clim. Change 13 (2023) 693–700. doi: 10.1038/s41558-023-01692-7

    9. [9]

      X. Zhao, H. Zhou, V.S. Sikarwar, et al., Energy Environ. Sci. 10 (2017) 1885–1910. doi: 10.1039/C6EE03718F

    10. [10]

      O.W. Sheng, C.B. Jin, X.F. Ding, et al., Adv. Funct. Mater. 31 (2021) 2100891. doi: 10.1002/adfm.202100891

    11. [11]

      Y. Wang, Y.H. Pang, H. Xu, A. Martinez, K.S. Chen, Energy Environ. Sci. 15 (2022) 2288–2328. doi: 10.1039/d2ee00790h

    12. [12]

      X.H. Li, T.T. Chao, Y.E. Duan, et al., ACS Sustainable Chem. Eng. 9 (2021) 15138–15146. doi: 10.1021/acssuschemeng.1c03047

    13. [13]

      D.F. Chen, P.C. Pei, Y.H. Li, et al., Energy Convers. Manage. 261 (2022) 115651. doi: 10.1016/j.enconman.2022.115651

    14. [14]

      D.X. Xie, S.J. Hu, D.H. Teng, et al., Chem. Eng. J. 457 (2023) 141310. doi: 10.1016/j.cej.2023.141310

    15. [15]

      H. Chen, J. Song, X. Long, et al., Appl. Catal. B 361 (2025) 124643. doi: 10.1016/j.apcatb.2024.124643

    16. [16]

      M.F.R. Hanifah, J. Jaafar, M.H.D. Othman, et al., Electrochim. Acta 403 (2022) 139608. doi: 10.1016/j.electacta.2021.139608

    17. [17]

      Ş. Neaţu, F. Neaţu, I.M. Chirica, et al., J. Mater. Chem. A 9 (2021) 17065–17128. doi: 10.1039/d1ta03644k

    18. [18]

      K.L. Wang, P.C. Liu, M.Z. Wang, et al., Chin. Chem. Lett. 36 (2025) 110532. doi: 10.1016/j.cclet.2024.110532

    19. [19]

      Z.L. Wu, Q.H. Yang, Y. Liu, et al., ACS Catal. 12 (2022) 3063–3082. doi: 10.1021/acscatal.1c06034

    20. [20]

      J. Liu, J. Zhao, H.Q. Wei, Q. Zhu, Y. Li, Energy Convers. Manage. 302 (2024) 118128. doi: 10.1016/j.enconman.2024.118128

    21. [21]

      F.S. Nanadegani, B. Sunden, Int. J. Hydrogen Energy 48 (2023) 32875–32942. doi: 10.1016/j.ijhydene.2023.05.052

    22. [22]

      D. Gerteisen, J. Power Sources 195 (2010) 6719–6731. doi: 10.1016/j.jpowsour.2010.04.004

    23. [23]

      Y.H. Zuo, W.C. Sheng, W.Q. Tao, Z. Li, J. Mater. Sci. Technol. 114 (2022) 29–41. doi: 10.1016/j.jmst.2021.10.031

    24. [24]

      S. Zaman, L. Huang, A.I. Douka, et al., Angew. Chem., Int. Ed. 133 (2021) 17976–17996. doi: 10.1002/ange.202016977

    25. [25]

      Y. Wang, D.S. Wang, Y.D. Li, Adv. Mater. 33 (2021) 2008151. doi: 10.1002/adma.202008151

    26. [26]

      Y.H. He, Q. Tan, L.L. Lu, J. Sokolowski, G. Wu, Electrochem. Energy Rev. 2 (2019) 231–251. doi: 10.1007/s41918-019-00031-9

    27. [27]

      L.M. Dai, Y.H. Xue, L.T. Qu, H.J. Choi, J.B. Baek, Chem. Rev. 115 (2015) 4823–4892. doi: 10.1021/cr5003563

    28. [28]

      C.Z. Wan, X.F. Duan, Y. Huang, Adv. Energy Mater. 10 (2020) 1903815. doi: 10.1002/aenm.201903815

    29. [29]

      Y.X. Sun, H.Y. Li, S.Q. Guo, C.J. Li, Chin. Chem. Lett. 35 (2024) 109418. doi: 10.1016/j.cclet.2023.109418

    30. [30]

      M.L. Liu, Z.P. Zhao, X.F. Duan, Y. Huang, Adv. Mater. 31 (2019) 1802234. doi: 10.1002/adma.201802234

    31. [31]

      X.Q. Wang, Z.J. Li, Y.T. Qu, et al., Chem 5 (2019) 1486–1511. doi: 10.1016/j.chempr.2019.03.002

    32. [32]

      X.L. Tian, X.F. Lu, B.Y. Xia, X.W. Lou, Joule 4 (2020) 45–68. doi: 10.1016/j.joule.2019.12.014

    33. [33]

      G. Alemany-Molina, C. Lo Vecchio, V. Baglio, et al., J. Colloid Interface Sci. 691 (2025) 137426. doi: 10.1016/j.jcis.2025.137426

    34. [34]

      L.L. Zhang, P.P. Lu, M.M. Yin, et al., Rare Met 44 (2024) 1767–1776. doi: 10.3390/en17071767

    35. [35]

      K. Singh, E.B. Tetteh, H.Y. Lee, T.H. Kang, J.S. Yu, ACS Catal. 9 (2019) 8622–8645. doi: 10.1021/acscatal.9b01420

    36. [36]

      H. Xu, H.Y. Shang, C. Wang, Y.K. Du, Adv. Funct. Mater. 30 (2020) 2000793. doi: 10.1002/adfm.202000793

    37. [37]

      C. Li, N. Clament Sagaya Selvam, J.Y. Fang, Nano-Micro Lett. 15 (2023) 83. doi: 10.1007/s40820-023-01060-2

    38. [38]

      S.S.A. Shah, T. Najam, M.S. Bashir, et al., Small 18 (2022) 2106279. doi: 10.1002/smll.202106279

    39. [39]

      T. Najam, S.S.A. Shah, S. Ibraheem, et al., Energy Storage Mater. 45 (2022) 504–540. doi: 10.1016/j.ensm.2021.11.050

    40. [40]

      J.Y. Zhao, J. Lian, Z.X. Zhao, X.M. Wang, J.J. Zhang, Nano-Micro Lett. 15 (2022) 19.

    41. [41]

      H.X. Zhong, M.C. Wang, G.B. Chen, R.H. Dong, X.L. Feng, ACS Nano 16 (2022) 1759–1780. doi: 10.1021/acsnano.1c10544

    42. [42]

      H.D. Xu, J. Yang, R.Y. Ge, et al., J. Energy Chem. 71 (2022) 234–265. doi: 10.3390/jof8030234

    43. [43]

      W.Y. Zhang, S.Y. Yi, Y.H. Yu, et al., J. Mater. Chem. A 12 (2024) 87–112. doi: 10.1039/D3TA05147A

    44. [44]

      M.Y. Chen, J.Q. Guan, Adv. Funct. Mater. 35 (2025) 2423552. doi: 10.1002/adfm.202423552

    45. [45]

      Y.F. Guo, L.L. Zhao, N. Zhang, et al., Energy Storage Mater. 70 (2024) 103556. doi: 10.1016/j.ensm.2024.103556

    46. [46]

      X. Wu, Q. Yan, H. Wang, et al., Adv. Funct. Mater. 34 (2024) 2404535. doi: 10.1002/adfm.202404535

    47. [47]

      J.J. Huo, Y.H. Dou, C. Wu, et al., Adv. Mater. 37 (2025) 2416483. doi: 10.1002/adma.202416483

    48. [48]

      Z. Zhang, P.Z. Liu, Y.H. Song, et al., Adv. Sci. 9 (2022) 2204297. doi: 10.1002/advs.202204297

    49. [49]

      R.Y. Ge, S.H. Yu, Y.W. Li, et al., Adv. Mater. 37 (2025) 2503223. doi: 10.1002/adma.202503223

    50. [50]

      Z.W. Zhang, H.H. Jin, J.W. Zhu, et al., Carbon 161 (2020) 502–509. doi: 10.1016/j.carbon.2020.01.108

    51. [51]

      H.J. Shen, T.J. Thomas, S.A. Rasaki, et al., Electrochem. Energy Rev. 2 (2019) 252–276. doi: 10.1007/s41918-019-00030-w

    52. [52]

      S.L. Li, L. Shi, Y.J. Guo, et al., Chem. Sci. 15 (2024) 11188–11228. doi: 10.1039/d4sc02853h

    53. [53]

      Z.Q. Zhu, Z.J. Li, Z.H. Liu, et al., Small Methods 10 (2026) 2500411. doi: 10.1002/smtd.202500411

    54. [54]

      M.A. Ud Din, M. Idrees, S. Jamil, et al., J. Energy Chem. 77 (2023) 499–513. doi: 10.1016/j.jechem.2022.11.023

    55. [55]

      C. Lee, C.K. Hwang, J.W. An, et al., Appl. Catal. B 358 (2024) 124371. doi: 10.1016/j.apcatb.2024.124371

    56. [56]

      Y. Feng, H. Liu, J. Yang, Sci. Adv. 3 (2017) e1700580.

    57. [57]

      H.F. Zhang, X.Y. Qiu, Y.F. Chen, et al., Small 16 (2020) 1906026. doi: 10.1002/smll.201906026

    58. [58]

      X.L. Wang, J.W. Li, X.T. Yang, et al., Nano Res. 15 (2022) 7951–7958. doi: 10.1007/s12274-022-4492-2

    59. [59]

      K. Cheng, Z.K. Kou, J. Zhang, et al., J. Mater. Chem. A 3 (2015) 14007–14014. doi: 10.1039/C5TA02386F

    60. [60]

      X. Tong, J.M. Zhang, G.X. Zhang, et al., Chem. Mater. 29 (2017) 9579–9587. doi: 10.1021/acs.chemmater.7b04221

    61. [61]

      D. Lee, S. Gok, Y. Kim, et al., ACS Appl. Mater. Interfaces 12 (2020) 44588–44596. doi: 10.1021/acsami.0c07812

    62. [62]

      Y. Zhang, B.L. Huang, G. Luo, et al., Sci. Adv. 6 (2020) eaba9731. doi: 10.1126/sciadv.aba9731

    63. [63]

      S. Kim, J.E. Park, W. Hwang, Y.H. Cho, Y.E. Sung, Appl. Catal. B 209 (2017) 91–97. doi: 10.15384/kjhp.2017.17.2.91

    64. [64]

      C.J. Hou, W.J. Yuan, Y.J. Zhang, et al., Fuel 379 (2025) 133111. doi: 10.1016/j.fuel.2024.133111

    65. [65]

      Q. Tan, T. Qu, C.Y. Shu, et al., ACS Sustainable Chem. Eng. 7 (2019) 17145–17153. doi: 10.1021/acssuschemeng.9b03497

    66. [66]

      Q.R. Shi, Y.H. He, X.W. Bai, et al., Energy Environ. Sci. 13 (2020) 3544–3555. doi: 10.1039/d0ee01968b

    67. [67]

      S.K. Singh, K. Takeyasu, J. Nakamura, Adv. Mater. 31 (2019) 1804297. doi: 10.1002/adma.201804297

    68. [68]

      J. Zhang, J.J. Zhang, F. He, et al., Nano-Micro Lett. 13 (2021) 65. doi: 10.1007/s40820-020-00579-y

    69. [69]

      W.Q. Song, C.X. Xiao, J. Ding, et al., Adv. Mater. 36 (2024) 2301477. doi: 10.1002/adma.202301477

    70. [70]

      M.M. Liu, L.L. Wang, K.N. Zhao, et al., Energy Environ. Sci. 12 (2019) 2890–2923. doi: 10.1039/c9ee01722d

    71. [71]

      Z.L. Yang, L.Q. Liu, X.B. Rao, et al., Chin. Chem. Lett. 36 (2025) 111440. doi: 10.1016/j.cclet.2025.111440

    72. [72]

      J.F. Huang, S. Ajmal, A. Kumar, et al., J. Energy Chem. 101 (2025) 132–155. doi: 10.1016/j.jechem.2024.08.068

    73. [73]

      M.J. Gong, A. Mehmood, B. Ali, K.W. Nam, A. Kucernak, ACS Catal. 13 (2023) 6661–6674. doi: 10.1021/acscatal.3c00356

    74. [74]

      I. Martinaiou, A.H.A. Monteverde Videla, N. Weidler, et al., Appl. Catal. B 262 (2020) 118217. doi: 10.1016/j.apcatb.2019.118217

    75. [75]

      L.Z. Gu, L.H. Jiang, X.N. Li, et al., Chin. J. Catal. 37 (2016) 539–548. doi: 10.1016/S1872-2067(15)61049-X

    76. [76]

      X.T. Zhao, X. Maimaitiyiming, M. Tursun, H. Lin, Fuel 364 (2024) 131089. doi: 10.1016/j.fuel.2024.131089

    77. [77]

      D. Sebastián, A. Serov, K. Artyushkova, et al., J. Power Sources 319 (2016) 235–246. doi: 10.1016/j.jpowsour.2016.04.067

    78. [78]

      X.Y. Zhang, D.P. Xue, S. Jiang, et al., InfoMat 4 (2022) e12257. doi: 10.1002/inf2.12257

    79. [79]

      H.G. Huang, K. Shen, F.F. Chen, Y.W. Li, ACS Catal. 10 (2020) 6579–6586. doi: 10.1021/acscatal.0c01459

    80. [80]

      X.L. Xu, Z.X. Xia, X.M. Zhang, et al., Appl. Catal. B 259 (2019) 118042. doi: 10.1016/j.apcatb.2019.118042

    81. [81]

      J.W. Shi, H.B. Shao, F. Yang, et al., Chem. Eng. J. 445 (2022) 136628. doi: 10.1016/j.cej.2022.136628

    82. [82]

      C.Y. Shu, Q. Tan, C.W. Deng, et al., Carbon Energy 4 (2022) 1–11. doi: 10.1002/cey2.136

    83. [83]

      A. Yu, Y. Yang, Angew. Chem., Int. Ed. 64 (2025) e202424161. doi: 10.1002/anie.202424161

    84. [84]

      Y. Wang, Y.X. Wang, J. Zhao, et al., ACS Sustainable Chem. Eng. 12 (2024) 6982–6989. doi: 10.1021/acssuschemeng.4c00036

    85. [85]

      L.F. Cui, J. Hao, Y. Zhang, et al., J. Colloid Interface Sci. 650 (2023) 603–612. doi: 10.1016/j.jcis.2023.06.153

    86. [86]

      K.W. Wang, H.J. Zou, J.Z. Meng, et al., J. Magnesium Alloys 11 (2023) 1959–1969. doi: 10.1016/j.jma.2023.04.008

    87. [87]

      X.B. Gao, Y.C. Wang, W.C. Xu, et al., J. Am. Chem. Soc. 145 (2023) 15528–15537. doi: 10.1021/jacs.3c04315

    88. [88]

      Q.J. Niu, B.L. Chen, J.X. Guo, et al., Nano-Micro Lett. 11 (2019) 8. doi: 10.1159/000496218

    89. [89]

      C.G. Hu, R. Paul, Q.B. Dai, L.M. Dai, Chem. Soc. Rev. 50 (2021) 11785–11843. doi: 10.1039/d1cs00219h

    90. [90]

      C. Kim, F. Dionigi, V. Beermann, et al., Adv. Mater. 31 (2019) 1805617. doi: 10.1002/adma.201805617

    91. [91]

      K.S. Bejigo, R.A. Raj, S.J. Kim, Sustainable Mater. Technol. 42 (2024) e01153. doi: 10.1016/j.susmat.2024.e01153

    92. [92]

      L.Z. Liu, G. Zeng, J.X. Chen, et al., Nano Energy 49 (2018) 393–402. doi: 10.5194/piahs-379-393-2018

    93. [93]

      A. Farzaneh, N. Saghatoleslami, E.K. Goharshadi, H. Gharibi, H. Ahmadzadeh, Electrochim. Acta 222 (2016) 608–618. doi: 10.1016/j.electacta.2016.11.015

    94. [94]

      Y. Liu, T. Zhang, X. Dai, et al., ACS Appl. Nano Mater. 3 (2020) 5139–5148. doi: 10.1021/acsanm.0c00438

    95. [95]

      L.Y. Gong, J. Liu, Y. Li, et al., Chem. Eng. J. 428 (2022) 131569. doi: 10.1016/j.cej.2021.131569

    96. [96]

      P. Anandha ganesh, D. Jeyakumar, Electrochim. Acta 262 (2018) 306–318. doi: 10.1016/j.electacta.2018.01.002

    97. [97]

      W.J. Wang, Z.Q. Jiang, X.N. Tian, T. Maiyalagan, Z.J. Jiang, Carbon 201 (2023) 1068–1080. doi: 10.1016/j.carbon.2022.10.022

    98. [98]

      L.M. Rivera Gavidia, G. García, D. Anaya, et al., Appl. Catal. B 184 (2016) 12–19. doi: 10.1016/j.apcatb.2015.11.021

    99. [99]

      X. Li, K.X. Yao, F.L. Zhao, et al., Nano Res. 15 (2022) 6036–6044. doi: 10.1007/s12274-022-4299-1

    100. [100]

      A. Sathyaseelan, V. Elumalai, K. Krishnamoorthy, A. Sajeev, S.J. Kim, ACS Sustainable Chem. Eng. 11 (2023) 5345–5355. doi: 10.1021/acssuschemeng.2c05220

    101. [101]

      Z.X. Lyu, X.G. Zhang, Y.C. Wang, et al., Angew. Chem., Int. Ed. 133 (2021) 16229–16236. doi: 10.1002/ange.202104013

    102. [102]

      P. Wang, H. Cui, C.X. Wang, Chem. Eng. J. 429 (2022) 132435. doi: 10.1016/j.cej.2021.132435

    103. [103]

      Z.Q. Wan, J.Q. Ye, J.Y. Zhang, et al., J. Mater. Chem. A 13 (2025) 9729–9736. doi: 10.1039/d5ta00198f

    104. [104]

      Y.H. Wang, F. Wang, Y. Fang, et al., Appl. Surf. Sci. 496 (2019) 143566. doi: 10.1016/j.apsusc.2019.143566

    105. [105]

      Y. Liu, C.Y. Shu, Y. Fang, Y.Z. Chen, Y.N. Liu, J. Power Sources 361 (2017) 160–169. doi: 10.1117/12.2283957

    106. [106]

      Y. Liu, Y.Z. Chen, S. Li, et al., ACS Appl. Mater. Interfaces 10 (2018) 9485–9494. doi: 10.1021/acsami.8b00613

    107. [107]

      C.Y. Shu, Z. Gan, Y.Y. Hou, et al., Energy Environ. Mater. 4 (2020) 81–87.

    108. [108]

      C.Y. Shu, B. Song, X.D. Wei, et al., Carbon 129 (2018) 613–620. doi: 10.1016/j.carbon.2017.12.049

    109. [109]

      X.L. Xu, X.M. Zhang, Z.X. Xia, et al., J. Energy Chem. 54 (2021) 579–586. doi: 10.1016/j.jechem.2020.06.046

    110. [110]

      Y.C. Wang, L. Huang, P. Zhang, et al., ACS Energy Lett. 2 (2017) 645–650. doi: 10.1021/acsenergylett.7b00071

    111. [111]

      X.L. Xu, X.M. Zhang, Z.X. Xia, et al., ACS Appl. Mater. Interfaces 13 (2021) 16279–16288. doi: 10.1021/acsami.0c22968

    112. [112]

      X.Q. Zheng, Z.C. Wang, Q. Zhou, et al., J. Energy Chem. 88 (2024) 242–251. doi: 10.1016/j.jechem.2023.08.051

    113. [113]

      P. Basumatary, J.H. Choi, D. Konwar, et al., Small Methods 8 (2024) 2301651.

    114. [114]

      M.X. Du, D. Li, S.Z. Liu, J.Q. Yan, Adv. Funct. Mater. 33 (2023) 2301527. doi: 10.1002/adfm.202301527

    115. [115]

      Q.Y. Guo, Y. Yang, T.T. Hu, et al., Chin. Chem. Lett. 36 (2025) 110235. doi: 10.1016/j.cclet.2024.110235

    116. [116]

      X.L. Zhao, F. Li, R.N. Wang, et al., Adv. Funct. Mater. 27 (2017) 1605717. doi: 10.1002/adfm.201605717

    117. [117]

      T.T. Hu, W.H. Chen, Y.B. Liu, et al., Small 19 (2023) 2304076. doi: 10.1002/smll.202304076

    118. [118]

      Z. Liu, S.F. Zhou, S.X. Ma, et al., Mater. Today Phys. 17 (2021) 100338. doi: 10.1016/j.mtphys.2020.100338

    119. [119]

      C.T. Fan, G.L. Wang, L.L. Zou, et al., J. Power Sources 429 (2019) 1–8. doi: 10.1117/1.jei.28.2.023004

    120. [120]

      L.L. Zou, J. Fan, Y. Zhou, et al., Nano Res. 8 (2015) 2777–2788. doi: 10.1007/s12274-015-0784-0

    121. [121]

      S.Q. Guo, S.Y. Yu, F. Chen, et al., J. Colloid Interface Sci. 652 (2023) 429–439. doi: 10.1016/j.jcis.2023.08.104

    122. [122]

      F.L. Zhao, L.R. Zheng, Q. Yuan, et al., Adv. Mater. 33 (2021) 2103383. doi: 10.1002/adma.202103383

  • Figure 1  (a, b) Schematic diagram of DMFC in acidic and alkaline media.

    下载: 全尺寸图片 幻灯片

    Figure 2  Modification strategies for cathode electrocatalysts in DMFC.

    下载: 全尺寸图片 幻灯片

    Figure 3  (a) Different O2 adsorption modes on the catalytic surface and (b) the ORR reaction pathways via 2e- and 4e-. Reproduced with permission [43]. Copyright 2024, Royal Society of Chemistry.

    下载: 全尺寸图片 幻灯片

    Figure 4  (a) TEM and (b) HRTEM images of Pt@CS/CNF. (c) Mechanisms of methanol tolerance and enhanced oxygen reduction. Reproduced with permission [61]. Copyright 2020, American Chemical Society. (d) Schematic of electrodes composed of Pt/C and Fe-N-C. (e) Mechanism schematic of DMFC based on composite cathode catalytic layer. Reproduced with permission [64]. Copyright 2024, Elsevier. (f) Adsorption configuration of methanol on N—C and FeN4 sites obtained by DFT calculations. (g) The adsorption energy of methanol on N—C, FeN4, CoN4, and MnN4 active sites calculated at different potentials. Reproduced with permission [66]. Copyright 2020, Royal Society of Chemistry.

    下载: 全尺寸图片 幻灯片

    Figure 5  (a) Schematic of Fe-mPDA@HPCS cathode structure with high-efficiency mass transfer. Reproduced with permission [82]. Copyright 2021, Wiley. (b) Proposed the possible structure of Fe/NS-C. (c) ORR polarization curves. Reproduced with permission [85]. Copyright 2023, Elsevier. (d, e) HAAFF-STEM images. (f, g) Acquired HAADF-STEM intensity line profiles. (h, i) Simulated Mg-Co HAADF-STEM image. (j) Proposed Mg-N-Co moiety in Mg-Co. Reproduced with permission [86]. Copyright 2023, Elsevier. (k) Linear correlation between -ICOHP and Bader valence electron number (N) in different single atom particle catalyst systems (left) and the dissolved Fe content for Fesac, Au-Fesac, and Pt-Fesac (right). (l) Long-term stability test at a constant current density of 0.1 A/cm2. Reproduced with permission [87]. Copyright 2023, American Chemical Society.

    下载: 全尺寸图片 幻灯片

    Figure 6  (a) Schematic representation of the DMFC assembled with the sPdNiA as bifunctional electrodes. (b) Polarization curves of the DMFC at various operating temperatures. Reproduced with permission [100]. Copyright 2023, American Chemical Society. (c) Electron transfer model on RuO2/Pd(111) interface. (d) Reaction free energy diagram of ORR at different catalytic sites. Reproduced with permission [101]. Copyright 2021, Wiley. (e) Schematic of ORR mechanism for the enhanced electrocatalytic activity of the Mn2O3@Co1.2Ni1.8O4. Reproduced with permission [106]. Copyright 2018, American Chemical Society.

    下载: 全尺寸图片 幻灯片

    Figure 7  (a) HAADF-STEM of MNDCS-0.3. (b) ORR polarization curves. Reproduced with permission [107]. Copyright 2020, Wiley. (c) Schematic of the triple-phase interface with N-YS-CS hydrophobic cathode structure. Reproduced with permission [108]. Copyright 2017, Elsevier. (d) Scheme of the microstructure for Fe/N/C cathodes. Reproduced with permission [110]. Copyright 2017, American Chemical Society. (e) Schematic of structural advantages. (f) Power density curves of DMFCs. (g) Short-term stability tests of DMFCs at 100 mA/cm2. Reproduced with permission [111]. Copyright 2021, American Chemical Society.

    下载: 全尺寸图片 幻灯片

    Figure 8  (a) Aberration-corrected HRTEM. (b) high-resolution XPS spectra of Pt 4f. Reproduced with permission [117]. Copyright 2023, Wiley. (c) Schematic of the CoFe@NCNF/BCNT synthesis. (d) The peak power densities obtained at different operating temperatures. Reproduced with permission [121]. Copyright 2023, Elsevier. (e) Structure diagram of flexible DMFC. (f) The electrochemical performance of DMFC under consecutive bending from flat to 180°. Reproduced with permission [12]. Copyright 2021, American Chemical Society.

    下载: 全尺寸图片 幻灯片
  • 加载中
WeChat 扫一扫看文章
计量
  • PDF下载量:  0
  • 文章访问数:  42
  • HTML全文浏览量:  6
文章相关
  • 发布日期:  2026-09-15
  • 收稿日期:  2025-06-16
  • 接受日期:  2026-01-12
  • 修回日期:  2025-12-13
  • 网络出版日期:  2026-01-13
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

/

返回文章

All Rights Reserved by the Chinese Chemical Society   中国化学会 版权所有 京ICP备05002797号

本系统由北京仁和汇智信息技术有限公司开发    技术支持: info@rhhz.net