Pt nanoparticle–nanowire hybrids supported on single-walled carbon nanotubes for enhanced oxygen reduction reaction in polymer electrolyte fuel cells

Abstract

Enhancing catalyst durability is crucial for the advancement of Polymer Electrolyte Fuel Cells (PEFCs). In this study, a hybrid catalyst composed of Pt nanoparticles and nanowires supported on single-walled carbon nanotubes (Pt_NP+NW/SWCNT) is investigated. This unique nanostructure synergistically combines the high activity of nanoparticles with the enhanced electron transport and structural stability offered by one-dimensional nanowires and SWCNTs. Pt_NP+NW/SWCNT demonstrated excellent electrochemical performance, with a half-wave potential of 0.882 V, a mass activity of 380 A g_Pt⁻¹, and a specific activity of 935 μA cm⁻², owing to its one-dimensional nanowire structure that promotes active site exposure and electron transport. The catalyst also showed superior intrinsic activity and remarkable durability. Accelerated degradation tests revealed only a 22.1% decrease in maximum power density and a minimal 14.9% loss in electrochemical surface area (ECSA) after 60 000 cycles, outperforming both Pt nanoparticles coated with N-doped carbon on SWCNTs (Pt@NC/SWCNT) and commercial Pt/C catalysts. While Pt@NC/SWCNT exhibits better resistance to acid poisoning in half-cell tests due to its N-doped carbon shell, Pt_NP+NW/SWCNT is more durable under realistic operating conditions. These results highlight the importance of structural stability in long-term fuel cell operation and suggest that Pt_NP+NW/SWCNT is a promising candidate for practical PEFC applications.

---

Introduction

Polymer electrolyte fuel cells (PEFCs) are widely regarded as promising next-generation clean energy technology due to their high energy density, portability, and environmental friendliness.^1,2 However, several critical challenges must be addressed before their large-scale deployment. Among these, the cathode electrocatalyst plays a crucial role in determining the overall performance of PEFCs.^3–5 The oxygen reduction reaction (ORR) is the rate-limiting step in fuel cell operation. Although various electrocatalysts have been developed to enhance activity and cost-effectiveness, carbon-supported platinum nanoparticles (Pt/C) remain among the most effective catalysts due to their high electrochemical activity and scalability.^6,7 While carbon supports provide a highly specific surface area that enhances catalytic performance, their long-term stability and poisoning of Pt-based catalysts under real operating conditions remain significant challenges.⁸ Under harsh conditions, such as fuel starvation, prolonged operation, and frequent start-up/shut-down cycles, carbon supports are susceptible to corrosion.^9–12 To maximize the electrochemical surface area (ECSA) and mass activity (MA), commercial Pt/C catalysts typically utilize small-sized Pt nanoparticles. However, these small nanoparticles are inherently prone to rapid degradation under fuel cell operating conditions, primarily through mechanisms such as dissolution driven by the Gibbs–Thomson effect, migration and coalescence, and Ostwald ripening.^13,14 Due to their high surface curvature and elevated surface energy, smaller particles exhibit a greater tendency toward dissolution during high potential cycling. Furthermore, the increased mobility of these nanoparticles facilitates migration and aggregation, leading to a significant reduction in the number of active sites and, consequently, a marked loss in catalytic performance.¹² In PEFCs, the decomposition of the Nafion membrane generates fluoride (F⁻) and sulfate (SO₄²⁻) ions, which cause severe sulfur poisoning of Pt catalysts and significantly degrade ORR activity.^15,16 Similarly, in PEFCs employing phosphoric acid-doped polybenzimidazole membranes, phosphoric acid and its anions, particularly adsorbed dihydrogen phosphate (H₂PO₄⁻), strongly bind to Pt active sites, leading to substantial ORR degradation.^17,18

To address these stability challenges, researchers have developed a carbon shell encapsulation strategy, in which Pt nanoparticles are physically enclosed within a protective carbon shell to mitigate dissolution and migration.^19,20 To further enhance the stability of the carbon shell, N-doped carbon shells have been developed. The incorporation of nitrogen can improve the electronic conductivity and oxidation resistance of the carbon matrix, while also strengthening the interaction between Pt and the carbon shell through electronic effects.^21,22 Since fuel cells typically operate within a relatively lower potential window, this work specifically focuses on the stability of catalysts under 0.6–0.95 V cycling conditions, which are more relevant to realistic operation. Beyond surface protection strategies, approaches such as increasing particle size and designing one-dimensional Pt nanowire structures have also proven effective in enhancing catalyst durability. Compared to small nanoparticles, large-sized Pt particles and Pt nanowires possess lower surface energy and greater structural stability, thereby exhibiting superior resistance to electrochemical dissolution and aggregation under PEFC operating conditions. Moreover, the one-dimensional morphology of Pt nanowires not only provides a large surface area but also effectively suppresses particle migration and coalescence.^23,24 For instance, Meng et al.²⁵ reported one-dimensional grain-boundary-rich Pt nanowires coupled with two-dimensional MXene nanosheets, in which the unique 1D/2D hybrid nanoarchitecture provided abundant active sites, strong interfacial electronic interactions, and robust durability, outperforming conventional Pt nanoparticle/carbon systems. In another strategy, Zhang et al.²⁶ designed composition-graded PdNi nanospheres coated with Pt monolayer shells, achieving high Pt utilization and remarkable ORR activity and stability through a core–shell structure with atomically thin Pt coverage. More recently, Wei et al.²⁷ proposed Pt nanoparticle–Mn single-atom pairs, where the synergistic interaction between Pt and Mn single atoms simultaneously optimized the electronic structure, stabilized the catalyst, and delivered outstanding ORR performance with extremely low Pt loading. These advances highlight that hybrid dimensional architectures, core–shell nanostructures, and atomic pair engineering are promising directions to overcome the activity–stability–cost trade-offs of Pt-based catalysts. To further improve catalyst stability, carbon nanotubes (CNTs) have been explored as advanced catalyst supports due to their excellent electrical conductivity, mechanical strength, and chemical durability.²⁸ Recently, the enhanced direct injection pyrolysis (e-DIPS) method has been employed for the synthesis of single-walled carbon nanotubes (SWCNTs), enabling precise control over tube diameters within the range of 1–2 nm with a narrow diameter distribution. This method utilizes two carbon sources with distinct decomposition properties, allowing for precise diameter modulation of SWCNTs by adjusting the flow rate of ethylene as the secondary carbon source.²⁹ Compared with commercially available SWCNTs (e.g., HiPco, CoMoCAT, and Carbolex), e-DIPS-derived SWCNTs exhibit superior crystallinity and purity, with a G-band to D-band intensity ratio exceeding 4 (G/D ratio >160), indicating significantly improved structural quality.³⁰ This catalyst supported by SWCNTs can effectively enhance the durability of the catalyst under harsh conditions.^31,32 Building on this foundation, we designed Pt nanoparticle–nanowire hybrid structures supported on SWCNTs to exploit the advantages of both nanoparticles and nanowires, providing a large surface area, interconnected networks, and efficient electron transport channels for enhanced ORR performance in PEFCs.

In this study, Pt catalysts were successfully synthesized by depositing Pt nanoparticles onto SWCNTs followed by the formation of N-doped carbon shells. Subsequent hydrogen etching led to the partial decomposition of the carbon shells and the simultaneous formation of a mixed structure consisting of Pt nanoparticles and Pt nanowires. TEM revealed that most of the carbon shells were significantly degraded, exposing more Pt active surfaces. XPS confirmed the presence of residual nitrogen species, suggesting that nitrogen atoms remained incorporated in the carbon support and form Pt–N bonds. XAS further proved the results of HRTEM and XPS. The structural evolution resulted in enhanced ORR performance, which can be attributed to the increased exposure of Pt active sites, the formation of highly conductive and structurally stable Pt nanowires, and the improved interaction between Pt and the support.

Experimental section

Materials

H₂PtCl₆·6H₂O (98.5%), aniline (99.0%), ethanol (>99.5%), acetone (>99.5%), and methanol (>99.8%) were procured from Fujifilm Wako Pure Chemical Co., along with 20% Nafion™ dispersion solution DE2020 CS type and 5% Nafion™ dispersion solution DE521 CS type, all utilized in their original form. e-DIPS SWCNTs (carbon purity >99%) with an approximate diameter of 2 nm were supplied by Meijo Nano Carbon Co., Ltd. Perchloric acid (HClO₄, 70%) and 2-propanol (IPA, >99.7%) were obtained from Kanto Chemical Co. and used without further treatment. A hydrophilic nylon membrane with a 0.45 μm pore size was acquired from Merck. The anode catalyst used was a commercial Pt/C (47 wt%, 2 nm) from TANAKA Precious Metals, while a commercial Pt/C (20 wt%, <5 nm) from Sigma-Aldrich served as the reference catalyst for the cathode. Sodium dodecyl benzene sulfonate (SDBS), also purchased from Sigma-Aldrich, was employed as a dispersant for the SWCNT support. Nafion NRE-211 membranes (25 μm thickness) were sourced from Moubic and used as received.

Synthesis of Pt@NC/SWCNT and Pt_NP+NW/SWCNT

The Pt–aniline complex was synthesized by dissolving 120 mg of H₂PtCl₆·6H₂O in 35 mL of distilled aniline under N₂ protection based on a previously reported procedure, with slight adjustments.³³ The resulting solution was stirred at room temperature for 5 h to ensure thorough interaction. Subsequently, unreacted aniline was removed by washing the product with 0.2 M HCl. The final Pt–aniline complex was obtained as a dark purple powder. After that, 10 mg of SWCNTs and 10 mg of Pt–aniline complex were dissolved in 100 mL of ethanol to yield a homogeneous mixture following a 15-minute ultrasonic treatment. Then, a mixture of SWCNTs and Pt–aniline complex was obtained by removing the ethanol solvent using a rotary evaporator. Additionally, a one-hour annealing process in N₂ at 600 °C was conducted to produce SWCNTs with Pt nanoparticles encapsulated in an N-doped carbon shell (Pt@NC/SWCNT). Subsequently, the samples were annealed under 5% H₂/Ar at 700 °C for 9 h to obtain Pt_NP+NW/SWCNT.

Structural characterization

The crystalline structure of the synthesized catalysts was characterized by X-ray diffraction (XRD) using a Rigaku SmartLab diffractometer (Rigaku Corporation, Japan). XRD analysis utilized Cu-Kα radiation with a wavelength (λ) of 0.15456 nm. X-ray photoelectron spectroscopy (XPS, ESCALAB 250, VG Scientific) with Al Kα radiation was employed to analyze the chemical composition of the catalysts. Transmission electron microscopy (TEM) analysis coupled with energy-dispersive spectroscopy (EDS) was performed utilizing a JEM-2100F/HK instrument by JEOL, Japan operated at 200 kV. Ultra-high resolution STEM images were recorded on a JEM ARM200F microscope with a spherical aberration corrector operating at 200 kV. The average particle sizes of the Pt@NC/SWCNT were determined based on a statistically significant dataset comprising over 100 samples. X-ray absorption spectroscopy (XAS) measurements were carried out at the BL5S1 beamline of the Aichi Synchrotron Radiation Center (AichiSR), operated by the Aichi Science & Technology Foundation in Seto, Aichi, Japan. The beamline is equipped with a 5 T superconducting magnet as the light source, a flat-bent Rh-coated mirror for collimation, a Si (111) double-crystal monochromator, and a cylindrically bent Rh-coated mirror for focusing and suppression of higher-order harmonics. The photon flux at the sample position was approximately 3 × 10¹⁰ photons per second, with a beam spot size smaller than 0.5 mm × 0.5 mm. Photon energy calibration was performed using the characteristic peak (8980.3 eV) of the Cu K-edge X-ray absorption near-edge structure (XANES) spectrum of a Cu foil. Pt L₃-edge XANES spectra were recorded in transmission mode using the Si (111) monochromator. The mass composition of the catalysts was determined using thermal gravimetric analysis (TGA, TGA&DTG-60AH, Shimadzu Co., Japan), where the catalysts were subjected to annealing up to 1100 °C at a heating rate of 10 °C min⁻¹ under an air atmosphere.

Electrochemical measurements

The measurements were conducted using a standard three-electrode system, consisting of a glassy carbon (GC) rotating-disk electrode (RDE) with a geometric area of 0.1256 cm² (EC-Frontier Co., Ltd) employed as the working electrode, a Pt coil utilized as the counter electrode, and a reversible hydrogen electrode (RHE) employed as the reference electrode. The catalyst ink was formulated by dispersing 1 mg of sample in 3 mL of IPA and 10 μL of Nafion (5%) solution while commercial Pt/C (20 wt%) inks were prepared, similarly. After ultrasonication for 20 minutes, a uniform catalyst ink was obtained and dropped onto the GC electrode. The Pt loading of commercial Pt/C, Pt@NC/SWCNT, and Pt_NP+NW/SWCNT on the GC were 14.3 μg cm⁻², 10.2 μg cm⁻², and 11.2 μg cm⁻², respectively. ORR performance was assessed using cyclic voltammetry (CV) and linear sweep voltammetry (LSV), performed with a potentiostat/galvanostat (EC-Frontier Co., Ltd). Electrochemical measurements were conducted in 0.1 M HClO₄ electrolyte at 25 °C. The working electrode was initially activated by potential cycling between 0.05 and 1.2 V at 50 mV s⁻¹ under N₂-saturated conditions until a steady-state voltammogram was achieved. Subsequently, CV measurements of ECSA were performed at a scanning rate of 50 mV s⁻¹ and LSV measurements were performed at a scanning rate of 10 mV s⁻¹ and 1600 rpm under O₂-saturation. The ECSA was calculated assuming a hydrogen monolayer adsorption charge of 210 μC cm⁻².

Durability measurement of the PEFC

For the single-cell tests, an electrode area of 1 cm² was used for the membrane electrode assembly (MEA). The cathode ink was prepared by mixing and dispersing Pt_NP+NW/SWCNT, SDBS, Nafion (5%), and 15 mL of ethanol. To compare the performance with Pt@NC/SWCNT, the catalyst inks of Pt@NC/SWCNT were prepared using the same method. The Pt loading of Pt_NP+NW/SWCNT and Pt@NC/SWCNT was 98 and 69 μg cm⁻². The anode catalyst ink was prepared using the doctor blade method by mixing Pt/C (47 wt%), IPA, ultrapure water and Nafion™ (20%), achieving a Pt loading of 200 μg cm⁻². The Pt/C (20 wt%) cathode with Pt loading of 110 μg cm⁻² was prepared using the same method as that for the anode catalyst. The appropriate amount of catalyst ink was filtered through a nylon membrane (0.45 μm thickness), washed with acetone, methanol, and ultra-pure water to remove excess SDBS, and dried in an oven at 80 °C for 8 h. After the drying process, the membrane was cut into 1 cm² pieces. The prepared cathode and anode were pressed onto the Nafion membrane (NRE-211, thickness: 25 μm) at 140 °C and 1.5 MPa for 3 minutes.

The assembled MEA was configured into a single fuel cell (Chemix Ltd, FC-010-01H) and evaluated using a fuel cell test system (TOYO Corp.). Carbon paper (GDL-28-AA) was used as the gas diffusion layer, a carbon plate served as the bipolar plate, and a silicon gasket was employed. Prior to all MEA measurements, a constant voltage of 0.2 V was applied at a cell temperature of 80 °C for 2 h (anode: H₂, 1 NL min⁻¹, relative humidity [RH] 81%; cathode: air, 2 NL min⁻¹, RH 81%). The I–V measurements were conducted at a cell temperature of 80 °C (anode: H₂, 1 NL min⁻¹, RH 100%; cathode: air, 2 NL min⁻¹, RH 100%) and were repeated several times until a stable maximum current density was attained. Before conducting the CV measurement, the cell was supplied with H₂ (70 NmL min⁻¹) to the anode and N₂ (166 NmL min⁻¹) to the cathode for 20 minutes to discharge the O₂ in the MEA.

Following cessation of the N₂ supply to the cathode, CV measurements were conducted at 80 °C using a scan rate of 50 mV s⁻¹ in the potential range of 0.05–0.9 V (anode: H₂, 70 NmL min⁻¹, RH 100%), repeated for eleven cycles. The ECSA was calculated assuming a hydrogen monolayer adsorption charge of 210 μC cm⁻². Durability tests were performed by applying a square-wave potential between 0.6 V (3 s) and 0.95 V (3 s) for up to 60 000 cycles under H₂ (70 mL min⁻¹) at the anode and N₂ (166 NmL min⁻¹) at the cathode.

Results and discussion

Material preparation and characterization

The Pt–aniline complex was prepared according to a previously reported procedure, with minor modifications.³³ The complex was then mixed with SWCNTs in ethanol and subjected to ultrasonic treatment for 30 minutes, followed by stirring overnight. Subsequently, the ethanol solvent was evaporated to obtain a dry mixture. The mixture was calcined in an N₂ atmosphere at 600 °C for 1 h, during which Pt ions were in situ reduced by the decomposition and carbonization of the aniline complex, while the concurrent carbonization process induced the formation of a nitrogen-doped carbon matrix.³³ The N-doped carbon shell-coated Pt nanoparticle catalyst (Pt@NC/SWCNT) was obtained. In addition, carbon etching (H₂ + C–CH₄)³⁴ was carried out slowly by calcination at 700 °C for 9 h under 5% H₂/Ar. During this process, some Pt nanoparticles gradually aggregated into nanowires. As a result, a hybrid catalyst consisting of Pt nanoparticles and nanowires supported on SWCNTs (Pt_NP+NW/SWCNT) was obtained as shown in Fig. 1. The Pt loading of Pt@NC/SWCNT and Pt_NP+NW/SWCNT obtained by TGA were 16 wt% and 20 wt%, respectively, as shown in Fig. S2.

Fig. 1 Schematic diagram of the synthesis from Pt@NC/SWCNT to Pt_NP+NW/SWCNT.

As shown in Fig. 2a, the Pt@NC/SWCNT catalyst exhibited a relatively uniform size distribution, with an average particle diameter of 2.3 ± 0.9 nm, whereas Fig. S1 indicated that Pt_NP+NW/SWCNT had a larger average size of 7.5 ± 0.9 nm and a broader distribution, with several nanoparticles exceeding 10 nm. In Fig. 2b, the carbon shells encapsulating the Pt nanoparticles within the Pt@NC/SWCNT, which exhibited an average thickness of approximately 1 nm, can be clearly observed, demonstrating a uniform and well-defined core–shell structure that serves to stabilize the nanoparticles. Moreover, the distinct weight loss in Pt@NC/SWCNT observed in the temperature range of 200–400 °C as shown in Fig. S2, which was attributed to the thermal decomposition of the carbon shells that surrounded the Pt nanoparticles, provided additional evidence confirming the successful formation and presence of carbon shells.^32,33 Furthermore, Fig. S3a presents the HAADF-STEM image of Pt@NC/SWCNT along with the corresponding elemental mapping, in which the spatial distribution of Pt was shown to closely coincide with that of nitrogen. Based on the findings reported by Karuppannan et al.³³ and Deng et al.,³⁵ it can be inferred that the nitrogen species, mainly in the form of graphitic N derived from the decomposition of aniline, were primarily incorporated into the carbon shell encapsulating the Pt nanoparticles, rather than being directly coordinated to the Pt matrix. As shown in Fig. 2c, after slow annealing of Pt@NC/SWCNT under a H₂ atmosphere at 700 °C, the carbon shells were partially decomposed due to hydrogen etching. The Pt nanoparticles grew along the axis of the carbon nanotubes to form nanowires, while a small fraction of Pt nanoparticles still aggregated into larger particles. In the STEM mapping shown in Fig. S3b, the one-dimensional morphology of the Pt nanowires is clearly observed, and the nitrogen signal is found to be uniformly distributed across the SWCNT surface. This suggested that although hydrogen etching partially decomposed the carbon shell, residual nitrogen-containing species remained adsorbed on the SWCNT surface. Furthermore, as shown in Fig. 2d, the diameter of the Pt nanowires in Pt_NP+NW/SWCNT was less than 2 nm, and the measured lattice spacing was 2.26 Å. In contrast, the lattice spacing of Pt in Pt@NC/SWCNT was 2.29 Å (Fig. S4a), which was consistent with the standard Pt (111) plane. This suggested that the Pt nanowires in Pt_NP+NW/SWCNT experienced a lattice compression of approximately 1.3%. Furthermore, interplanar spacings of the (111) planes were measured to be 2.25 Å for Pt_NP+NW/SWCNT and 2.29 Å for Pt@NC/SWCNT based on HRTEM images, confirming lattice compression (Fig. S4b and c). Fig. 2e shows the XRD patterns of commercial Pt/C, Pt@NC/SWCNT and Pt_NP+NW/SWCNT. The wide peak observed near 25° is attributed to the carbon support. In the XRD pattern of Pt@NC/SWCNT, the diffraction peaks observed at 39.9°, 46.3°, 67.9° and 81.5° are indexed to the (111), (200), (220) and (311) crystal planes of metal Pt, respectively. The diffraction peaks of Pt_NP+NW/SWCNT are sharp and free of impurity signals, indicating that it possesses high crystallinity. Compared with Pt/C, the (111) diffraction peak of the Pt_NP+NW/SWCNT catalyst is more intense, suggesting a preferential (111) orientation that is thermodynamically favored due to its lower surface energy.³⁶ As shown in Fig. 2f, the Pt(111) peak shifts to higher 2θ values relative to Pt/C and Pt@NC/SWCNT, indicating compressive strain likely caused by surface shrinkage.³⁷ The SWCNT-supported nanowire structure may further enhance this preferential orientation and strain, highlighting the synergistic effect of the support on the Pt nanostructure.

Fig. 2 The overview HRTEM images of (a and b) Pt@NC/SWCNT with a histogram of the Pt particle sizes and (c) Pt_NP+NW/SWCNT. (d) HAADF-STEM image of Pt_NP+NW/SWCNT (inset: the atomic line profile). (e and f) XRD patterns of Pt/C, Pt@NC/SWCNT and Pt_NP+NW/SWCNT.

Fig. 3a shows the XPS Pt 4f core-level spectra of Pt@NC/SWCNT, Pt_NP+NW/SWCNT, and commercial Pt/C. The two significant peaks near 71.4 and 75.1 eV correspond to Pt 4f_7/2 and Pt 4f_5/2, which are divided into metallic Pt (Pt⁰) and oxidized Pt (Pt²⁺ and Pt⁴⁺), respectively.^19,32 As shown in Fig. 3b, Pt_NP+NW/SWCNT exhibited the highest ratio of metallic to oxidized Pt (68 : 32), compared to Pt@NC/SWCNT (60 : 40) and commercial Pt/C (54 : 46). This result suggested that the one-dimensional nanowire structure favored the preservation of metallic Pt species, likely due to its unique electronic environment and reduced surface oxidation tendency.³⁸Fig. 3c shows the N 1s XPS spectrum of Pt_NP+NW/SWCNT, where two distinct peaks appear at 398.7 eV and 401.0 eV, corresponding to Pt–N bonds and graphitic N, respectively.^39,40 In contrast, for Pt@NC/SWCNT, only graphitic N was detected, as shown in Fig. S5, and the C–N bond was detected as shown in Fig. S6a, indicating that nitrogen was mainly doped into the carbon shell rather than into the Pt matrix. The appearance of both Pt–N and C–N bonds in Fig. S6b for Pt_NP+NW/SWCNT suggested that high-temperature hydrogen treatment partially etched the carbon shell, exposing nitrogen species capable of coordinating with Pt atoms.⁴¹ This implied that a fraction of nitrogen-doped carbon shells remained on the nanowires, further strengthening their interaction with Pt. Additionally, a 0.4 eV positive shift in the Pt 4f binding energy for Pt_NP+NW/SWCNT compared to Pt/C (Fig. 3a) indicated reduced electron density around Pt, likely due to Pt–N coordination. The electron-withdrawing nature of nitrogen drew electrons from Pt, leading to an overall electron-deficient Pt surface. This electronic modulation enhanced the metallic characteristics of Pt, contributing to improved ORR catalytic activity and durability.⁴²

Fig. 3 (a) XPS Pt 4f core-level spectra of Pt_NP+NW/SWCNT, Pt@NC/SWCNT and commercial Pt/C. (b) Relative peak area ratio for each catalyst. (c) N 1s spectra of Pt_NP+NW/SWCNT. (d) XANES of the Pt L₃-edge for Pt_NP+NW/SWCNT, Pt@NC/SWCNT and commercial Pt/C. (e) Corresponding FT-EXFAS for Pt_NP+NW/SWCNT, Pt@NC/SWCNT, commercial Pt/C and references.

To corroborate the XPS findings, XANES measurements at the Pt L₃-edge were performed (Fig. 3d). Pt/C exhibited the highest white line intensity compared to Pt_NP+NW/SWCNT and Pt@NC/SWCNT, suggesting a higher proportion of oxidized Pt species. In contrast, Pt_NP+NW/SWCNT showed the lowest white line intensity, approaching that of metallic Pt, indicating that the Pt is predominantly in a reduced state. This observation was consistent with the results obtained from XPS analysis. FT-EXAFS analysis of the Pt L₃-edge (Fig. 3e) revealed a clear Pt–N coordination signal in Pt_NP+NW/SWCNT, absent in Pt@NC/SWCNT and Pt/C. This confirmed that N-doping significantly altered the local coordination environment of Pt. In addition, the observed Pt–Pt bond contraction compared to metallic Pt suggested lattice compression, consistent with the HRTEM results.^43,44 A residual fraction of N-doped carbon shells remained on the nanowires, serving as anchoring sites that stabilized Pt–N coordination and promoted the preferential growth of Pt along the (111) facet into one-dimensional nanowires. Furthermore, surface electronic modulation due to Pt–N coordination helped stabilize metallic Pt. Similar effects have been reported in sulfur-doped graphene systems, where heteroatom doping enhanced metal–support interactions and nanowire formation.^45,46 A control experiment further confirmed the essential role of the N-doped carbon shell. Annealing a physical mixture of SWCNTs and H₂PtCl₆·6H₂O under the same conditions (5% H₂/Ar, 700 °C, 9 h) yielded no nanowires (Fig. S7). Instead, irregular Pt particles over 50 nm in size were observed. This stark contrast underscored the importance of the N-doped carbon shell in guiding Pt nucleation, suppressing agglomeration, and promoting nanowire formation. The presence of Pt–N coordination likely contributes to lattice compression by modifying the local bonding and electronic environment.³²

Electrocatalytic properties for an efficient ORR catalyst

The electrochemical measurement results in the three-electrode system were evaluated at room temperature, and the results are shown in Fig. 4. CVs of commercial Pt/C (Pt loading: 20 wt%), Pt@NC/SWCNT and Pt_NP+NW/SWCNT in 0.1 M HClO₄ saturated with N₂ and O₂ are shown in Fig. 4a–c. The CV curves reveal the adsorption/desorption peaks of hydrogen and reduction–oxidation peaks of Pt within the potential range of 0.05–0.4 V and 0.6–1.2 V, respectively. If the oxidation of monolayer hydrogen requires 210 μC cm⁻², the ECSA value is calculated by integrating the charge collected in the hydrogen desorption region. The ECSA values of Pt@NC/SWCNT (43.3 m² g_Pt⁻¹) and Pt_NP+NW/SWCNT (40.6 m² g_Pt⁻¹) were slightly lower than that of commercial Pt/C (47.4 m² g_Pt⁻¹). The lower ECSA value of Pt_NP+NW/SWCNT was due to the presence of larger Pt nanoparticles, as shown in the HRTEM (Fig. S1). Under N₂-saturated conditions, the Pt_NP+NW/SWCNT exhibited a higher ORR peak potential (787 mV) compared to Pt/C (735 mV) and Pt@NC/SWCNT (722 mV), indicating more favorable oxygen adsorption/desorption characteristics and a well-preserved electrochemical surface. Upon switching to O₂-saturated conditions, the ORR peak current of Pt_NP+NW/SWCNT increased significantly, and the ORR peak potential further shifted positively to 831 mV, outperforming Pt/C (792 mV) and Pt@NC/SWCNT (803 mV). These results suggested that Pt_NP+NW/SWCNT possessed superior ORR kinetics and more efficient utilization of active sites, attributed to its unique nanoparticle–nanowire hybrid architecture supported on SWCNTs.

Fig. 4 CVs of (a) Pt/C, (b) Pt@NC/SWCNT and (c) Pt_NP+NW/SWCNT measured under both N₂- and O₂-saturated 0.1 M HClO₄ solution. (d) LSV curves recorded under O₂-saturated conditions at 1600 rpm, 25 °C (scan rate: 10 mV s⁻¹). (e) The half-wave potential of the catalysts (f) MA and SA of the catalysts.

As shown in Fig. 4d and e, the half-wave potentials (E_1/2) of Pt@NC/SWCNT, Pt_NP+NW/SWCNT, and commercial Pt/C are 0.846 V, 0.882 V, and 0.850 V (vs. RHE), respectively, indicating the superior ORR activity of the Pt_NP+NW/SWCNT catalyst. This improvement was attributed to the large specific surface area and the unique one-dimensional nanowire structure, which not only offered abundant accessible active sites but also promoted efficient electron transport.^47,48 As shown in Fig. 4f, the MA values for Pt@NC/SWCNT and Pt_NP+NW/SWCNT were found to be 160 A g_Pt⁻¹ and 380 A g_Pt⁻¹, respectively, both exceeding that of commercial Pt/C (154 A g_Pt⁻¹). Similarly, the specific activity (SA) for Pt@NC/SWCNT, Pt_NP+NW/SWCNT and Pt/C was 369 μA cm⁻², 935 μA cm⁻², and 324 μA cm⁻², respectively. Although the ECSA of Pt_NP+NW/SWCNT was relatively low, its MA and SA were significantly higher than those of Pt@NC/SWCNT and commercial Pt/C, with the SA reaching more than twice the value observed. The significant enhancement of SA in Pt_NP+NW/SWCNT may be attributed to the excellent geometric structure of Pt nanowires, which may facilitate the preferred exposure of Pt atomic sites on the surface. These sites are usually considered to have higher intrinsic ORR activity.^47,49 Meanwhile, the presence of Pt–N coordination bonds is believed to modulate the electronic structure of surface Pt atoms, thereby optimizing the adsorption energies of oxygen intermediates and improving catalytic efficiency.^32,50

Evaluation of anion adsorption effects on ORR performance

To further evaluate the resistance to negative ion poisoning, we conducted ORR tests on electrolytes containing H₃PO₄ and H₂SO₄. As shown in Fig. 5a and b, adding different amounts of H₃PO₄ (1 mmol, 10 mmol) to 0.1 M HClO₄ electrolyte solution causes the E_1/2 of all catalysts to gradually shift negatively, indicative of phosphate-induced site blockage and deactivation. Among them, Pt@NC/SWCNT exhibited the smallest drop in E_1/2 and ECSA, highlighting its superior resistance to phosphate poisoning, likely attributed to the protective effect of the N-doped carbon shell.¹⁵ In comparison, Pt/C and Pt_NP+NW/SWCNT experienced more pronounced performance losses at higher phosphate concentrations. Fig. 5c and d show that H₂SO₄ contamination also negatively affected both E_1/2 and ECSA. Notably, Pt_NP+NW/SWCNT retained the highest E_1/2 under sulfuric acid exposure, suggesting structural benefits in maintaining ORR kinetics despite a significant decrease in ECSA. The corresponding CV curves and LSV curves are shown in Fig. S8 and S9. Overall, Pt@NC/SWCNT demonstrated the best comprehensive resistance to acid-induced deactivation, retaining both ORR activity and active surface area more effectively than the other catalysts, while Pt/C showed poorer tolerance under both contamination conditions.

Fig. 5 (a) E_1/2 of Pt/C, Pt@NC/SWCNT, and Pt_NP+NW/SWCNT catalysts before and after H₃PO₄ contamination in O₂-saturated 0.1 M HClO₄ solution (rotation rate: 1600 rpm; scan rate: 10 mV s⁻¹; 25 °C). (b) ECSA of Pt/C, Pt@NC/SWCNT, and Pt_NP+NW/SWCNT catalysts before and after H₃PO₄ contamination. (c) E_1/2 of Pt/C, Pt@NC/SWCNT, and Pt_NP+NW/SWCNT catalysts before and after H₂SO₄ contamination in O₂-saturated 0.1 M HClO₄ solution (rotation rate: 1600 rpm; scan rate: 10 mV s⁻¹; 25 °C). (d) ECSA of Pt/C, Pt@NC/SWCNT and Pt_NP+NW/SWCNT catalysts before and after H₂SO₄ contamination.

Study on the durability of the MEA

Durability tests of Pt@NC/SWCNT, Pt_NP+NW/SWCNT and commercial Pt/C were conducted using MEA in a single-cell configuration within a rectangular potential oscillation ranging from 0.6 V (3 s) to 0.95 V (3 s), as shown in Fig. 6. This process simulated the potential fluctuations experienced by fuel cells during actual operation to accelerate the degradation of Pt over long periods of time. I–V curves were recorded at 80 °C, where the anode and cathode were supplied with fully humidified H₂ and air, respectively. As shown in Fig. 6a–c, at the initial stage, the current density of Pt_NP+NW/SWCNT at 0.6 V was markedly higher than that of Pt@NC/SWCNT and commercial Pt/C. With increasing cycling numbers, a rapid decline in current density was observed for Pt/C. After 60 000 accelerated degradation test (ADT) cycles, the current density of Pt/C at 0.6 V decreased by 276 mA cm⁻², corresponding to a loss of 58.1%. In comparison, Pt@NC/SWCNT exhibited a more pronounced degradation, with a current density loss of 320 mA cm⁻² (75.1%). In contrast, Pt_NP+NW/SWCNT demonstrated superior durability, with a reduction of only 239 mA cm⁻², corresponding to a relatively lower loss of 34.1%. In Fig. 6d, after 60 000 cycles, the maximum power density of Pt/C decreased from 663 mW cm⁻² to 330 mW cm⁻², a reduction of 50.2%. The maximum power density of Pt@NC/SWCNT dropped from 395 mW cm⁻² to 222 mW cm⁻², a decrease of 43.8%, while that of Pt_NP+NW/SWCNT dropped from 506 mW cm⁻² to 394 mW cm⁻², a decrease of only 22.1%. In Fig. 6e, the maximum power density per unit Pt load of Pt/C rapidly decreases from 6.0 mW cm⁻² μg_pt⁻¹ to 3.0 mW cm⁻² μg_pt⁻¹. The maximum power density per unit Pt load of Pt@NC/SWCNT decreased from 5.7 mW cm⁻² μg_pt⁻¹ to 3.2 mW cm⁻² μg_pt⁻¹. The maximum power density per unit Pt load of Pt_NP+NW/SWCNT decreased from 5.1 mW cm⁻² μg_pt⁻¹ to 3.95 mW cm⁻² μg_pt⁻¹. Even after 60 000 cycles, Pt_NP+NW/SWCNT still exhibited good activity and stability. The ECSA of all catalysts exhibited a gradual decline with increasing cycling numbers during the ADTs, as shown in Fig. S10. In Fig. 6f, the ECSA of Pt/C reveals severe degradation after 60 000 cycles, decreasing from 55.6 to 38.1 m² g_Pt⁻¹, representing a loss of 31.5%. Pt@NC/SWCNT also suffered a significant reduction, with its ECSA decreasing from 59.0 to 43.1 m² g_Pt⁻¹, corresponding to a loss of 26.9%. In contrast, Pt_NP+NW/SWCNT retained a relatively higher ECSA of 45.1 m² g_Pt⁻¹ after 60 000 cycles, only dropping from 53.7 m² g_Pt⁻¹, indicating a smaller loss of 14.9%. These results suggested that the Pt_NP+NW/SWCNT catalyst possesses superior structural and electrochemical stability under prolonged operation.

Fig. 6 Durability tests in a single-cell configuration by using a fuel-cell test system. Potential cycling was performed by applying a rectangular potential wave between 0.6 V and 0.95 V with a period of 6 s per cycle. I–V measurements of MEAs with (a) Pt/C, (b) Pt@NC/SWCNT and (c) Pt_NP+NW/SWCNT. (d) Variation of maximum power density, (e) maximum power density per Pt loading amount at various potential cycles, and (f) ECSA calculated from the CV curves of MEA during the durability tests.

Interestingly, these results contrasted with the findings from the half-cell poisoning experiments (Fig. 5), where Pt@NC/SWCNT showed the highest resistance to H₃PO₄ and H₂SO₄ contamination, while Pt_NP+NW/SWCNT was more vulnerable. This discrepancy highlighted the distinct degradation pathways in different testing environments. In RDE measurements, catalyst deactivation was dominated by surface adsorption of anionic species, where the N-doped carbon shell effectively mitigates phosphate and sulfate ion adsorption. However, in MEA tests, long-term performance was strongly influenced by the structural stability of the catalyst. The one-dimensional nanowire framework of Pt_NP+NW/SWCNT provided enhanced resistance against Pt dissolution and agglomeration, as well as better retention of ECSA (only 14.9% loss versus 26.9% and 31.5% for Pt@NC/SWCNT and Pt/C, respectively), thereby contributing to its superior performance and durability. Overall, while Pt@NC/SWCNT offered better short-term tolerance to acid poisoning, Pt_NP+NW/SWCNT proved to be more robust under realistic fuel cell operating conditions, making it a promising candidate for practical applications.

TEM images of Pt@NC/SWCNT and Pt_NP+NW/SWCNT after 60 000 potential cycles between 0.6 and 0.95 V revealed significant morphological changes. For Pt@NC/SWCNT, the average particle size of Pt nanoparticles increased from 2.3 ± 0.9 nm to 2.7 ± 0.9 nm, as shown in Fig. 7a and the absence of part of the carbon shell can be seen in Fig. S11, suggesting partial collapse of the N-doped carbon shell and the loss of structural confinement. In contrast, Pt_NP+NW/SWCNT (Fig. 7b) maintained a well-dispersed hybrid nanostructure composed of Pt nanoparticles and nanowires, which remained tightly anchored to the SWCNT structure. The one-dimensional nanowires exhibited minimal structural damage, suggesting a potential enhancement in resistance to Pt dissolution and Ostwald ripening during the durability test. XPS analysis of the Pt 4f spectra further provided evidence of the structural and chemical stability of the Pt_NP+NW/SWCNT catalyst. As shown in Fig. 7c and d, both catalysts exhibited a slight shift in the Pt 4f peak after the ADT test, indicative of partial surface oxidation. The extent of this shift was comparable between Pt@NC/SWCNT (Fig. 7c) and Pt_NP+NW/SWCNT (Fig. 7d), although Pt_NP+NW/SWCNT retained a marginally higher proportion of metallic Pt⁰. This suggests a slightly better preservation of metallic Pt and implies a subtle improvement in durability under fuel cell operating conditions.

Fig. 7 TEM images of (a) Pt@NC/SWCNT and (b) Pt_NP+NW/SWCNT after 60 000 cycles with the potential waves from 0.6 to 0.95 V. Pt 4f XPS spectra of (c) Pt@NC/SWCNT and (d) Pt_NP+NW/SWCNT catalysts in their initial condition before the durability test and after 60 000 potential cycles (0.6–0.95 V) in the durability test.

Generally, the degradation of PEFC performance during long-term operation is primarily attributed to several interrelated factors, including the increased oxidation state of Pt, electrochemical dissolution and redeposition (Ostwald ripening), carbon support corrosion, and the gradual reduction of ECSA due to poisoning or detachment of Pt species.⁵¹ These degradation pathways cause substantial losses in catalyst activity and structural integrity, leading to poor durability under practical operating conditions. In this study, the Pt_NP+NW/SWCNT catalyst demonstrates remarkable structural and electrochemical stability compared to Pt@NC/SWCNT and commercial Pt/C. The incorporation of one-dimensional Pt nanowires not only suppresses particle coalescence and dissolution due to their lower surface energy and extended geometry, but also improves anchoring strength with the SWCNT network, thereby reducing Pt migration and detachment.

To further evaluate the performance of the catalyst, additional back-pressures of 100 kPa and 150 kPa were applied during the MEA test process in accordance with the standards set by the New Energy and Industrial Technology Development Organization (NEDO)⁵² and the U.S. Department of Energy (DOE),⁵³ respectively. These tests were designed to be compared with the initial ADT measurements, which were performed under atmospheric pressure without any additional back-pressure applied. As shown in Fig. S12, all catalyst systems exhibited significantly enhanced current densities and maximum power densities with increasing pressure, primarily due to improved mass transport and water management under higher operating pressure. Among the tested samples, the Pt_NP+NW/SWCNT exhibited the highest current density of 1157 mA cm⁻² and maximum power density of 759 mW cm⁻² at 150 kPa, outperforming both Pt/C (1098 mA cm⁻², 831 mW cm⁻²) and Pt@NC/SWCNT (781 mA cm⁻², 565 mW cm⁻²), while Pt/C achieved the highest power density at 150 kPa (831 mW cm⁻²). These results highlighted the superior pressure adaptability of Pt_NP+NW/SWCNT, which combined high activity under ambient pressure with excellent scalability under elevated back-pressure, making it a promising candidate for practical PEFC applications under NEDO/DOE-relevant operating conditions.

Conclusions

In summary, a novel Pt-based hybrid catalyst composed of Pt nanoparticles and nanowires supported on SWCNTs was developed, exhibiting outstanding ORR activity in acidic media. Electrochemical characterization revealed that this catalyst possesses significant E_1/2 (0.882 V vs. RHE), MA (380 A g_Pt⁻¹), and SA (935 μA cm⁻²) values compared to those of commercial Pt/C and Pt@NC/SWCNT. The enhanced activity was attributed to the unique one-dimensional structure that enabled efficient electron transfer, improved reactant accessibility, and favored exposure of highly active Pt surface sites. The incorporation of nanowires not only increased structural rigidity but also optimized the electronic properties through Pt–N coordination. In MEA tests, Pt_NP+NW/SWCNT showed exceptional practical performance and stability, with only a 14.9% reduction in ECSA. It maintained 65.9% of its initial current density and retained 77.9% of its initial maximum power density after 60 000 potential cycles, significantly outperforming Pt@NC/SWCNT and commercial Pt/C in terms of degradation resistance. TEM and XPS analyses after long-term cycling confirmed that the Pt_NP+NW/SWCNT catalyst retained its structural integrity and chemical composition. This work highlighted the effectiveness of combining Pt nanowires and nanoparticles on a conductive and highly crystalline carbon nanotube support for developing next-generation PEFC cathode catalysts with both high ORR activity and excellent durability, crucial for demanding applications like heavy-duty vehicles.

Author contributions

Qiao Chen—original draft—preparation and creation of the published work, specifically writing the initial draft. Chu-Yang Yu—conducting the research process, performing data collection. Takashi Watanabe—conducting XAFS measurements. Masaya Kawasumi—experimental design, review, and editing—critical review. Miftakhul Huda—experimental design, review, and editing—critical review, commentary or revision, including pre- and post-publication stages. Yutaka Matsuo—supervision, conceptualization, ideas, review, funding acquisition, formulation and evolution of overarching research goals and aims. All authors have made valuable contributions to thie research and the interpretation of the results.

Conflicts of interest

There are no conflicts to declare.

Data availability

All data supporting the findings of this study are available from the corresponding authors upon reasonable request. Supplementary information, including details of ECSA calculation, synthesis and characterization of Pt@NC/SWCNT and Pt_NP+NW/SWCNT (HRTEM, HAADF-STEM, STEM-EDS, TGA, N 1s and C 1s spectra, Pt (111) distance), electrochemical measurements (CV, LSV, durability tests in RDE and single-cell configuration), and MEA performance under different back pressures, is available via the Royal Society of Chemistry website at DOI: https://doi.org/10.1039/d5nr02497h.

Acknowledgements

The authors would like to thank Mr Takeshi Hashimoto (Meijo Nano Carbon) for the carbon nanotubes and support with valuable data and discussion, and Mr Kimitaka Higuchi (Nagoya University) for technical assistance in ultra-high resolution STEM imaging. This study was supported by the “Knowledge Hub Aichi”, the Priority Research Project V from Aichi Prefectural Government.

References

1. A.-G. Olabi, T. Wilberforce and M.-A. Abdelkareem, Energy, 2021, 214, 118955.
2. M. K. Debe, Nature, 2012, 486, 43–51.
3. M. Uchida, Curr. Opin. Electrochem., 2020, 21, 209–218.
4. G. Charalampopoulos and M.-K. Daletou, Mater. Rep.: Energy, 2025, 100328.
5. D.-Y. Chung, J.-M. Yoo and Y.-E. Sung, Adv. Mater., 2018, 30, 1704123.
6. A. Kulkarni, S. Siahrostami, A. Patel and J. K. Nørskov, Chem. Rev., 2018, 118, 2302–2312.
7. M. Haque, T. Kawawaki and Y. Negishi, Int. J. Hydrogen Energy, 2024, 85, 30–47.
8. M. Tang, T. Yang, X. Yang, Y. Li, Z. Shi, X. Wang, C. Liu, W. Xing and J. Ge, Electrochim. Acta, 2023, 446, 142120.
9. S. Chung, K. Ham, S. Kang, H.-K. Ju and J. Lee, Electrochim. Acta, 2020, 348, 136346.
10. S. Chung, K. Ham, S. Kang, H. Ju and J. Lee, Electrochim. Acta, 2020, 348, 136346.
11. E. Antolini, Appl. Catal., B, 2009, 88, 1–24.
12. V. Nallathambi, J.-W. Lee, S.-P. Kumaraguru, G. Wu and B.-N. Popov, J. Power Sources, 2008, 183, 34–42.
13. J.-C. Meier, C. Galeano, I. Katsounaros, A.-A. Topalov, A. Kostka, F. Schüth and K.-J. Mayrhofer, ACS Catal., 2012, 2, 832–843.
14. P. Ren, P. Pei, Y. Li, Z. Wu, D. Chen and S. Huang, Prog. Energy Combust. Sci., 2020, 80, 100859.
15. T. Nakamura, Y. Takagi, S. Chaveanghong, T. Uruga, M. Tada, Y. Iwasawa and T. Yokoyama, J. Phys. Chem. C, 2020, 124, 17520–17527.
16. J.-H. Park, N. Saito and M. Kawasumi, Int. J. Hydrogen Energy, 2024, 72, 642–651.
17. M. Wu, X. Xu, X. Zhang, Z. An, H. Zhang, S. Wang and G. Sun, Energy Technol., 2023, 11, 2300042.
18. D. Xue and J.-N. Zhang, Ind. Chem. Mater., 2024, 2, 173–190.
19. J.-H. Park, K. Kim, X. Wang, M. Huda, Y. Sawada, Y. Matsuo and M. Kawasumi, J. Power Sources, 2023, 580, 233419.
20. K. Wang, H. Chen, X. Zhang, Y. Tong, S. Song, P. Tsiakaras and Y. Wang, Appl. Catal., B, 2020, 264, 118468.
21. H. Lee, Y.-E. Sung, I. Choi, T. Lim and O.-J. Kwon, J. Power Sources, 2017, 362, 228–235.
22. X. Tong, J. Zhang, G. Zhang, Q. Wei, R. Chenitz, J.-P. Claverie and S. Sun, Chem. Mater., 2017, 29, 9579–9587.
23. S. Sui, Z. Wei, K. Su, A. He, X. Wang, Y. Su and S. Du, Int. J. Hydrogen Energy, 2018, 43, 20041–20049.
24. Y. Lu, S. Du and R. Steinberger-Wilckens, Appl. Catal., B, 2016, 199, 292–314.
25. W. Meng, H.-Y. He, L. Yang, Q.-G. Jiang, B. Yuliarto, Y. Yamauchi, X.-T. Xu and H.-J. Huang, Chem. Eng. J., 2022, 450, 137932.
26. L.-X. Luo, F.-J. Zhu, R.-X. Tian, L. Li, S.-Y. Shen, X.-H. Yan and J.-L. Zhang, ACS Catal., 2017, 7, 5420–5430.
27. X.-Q. Wei, S.-J. Song, W.-W. Cai, Y.-Q. Kang, Q. Fang, L. Ling, Y.-J. Zhao, Z.-X. Wu, X.-K. Song, X.-T. Xu, S.-M. Osman, W.-Y. Song, T. Asahi, Y. Yamauchi and C.-Z. Zhu, ACS Nano, 2024, 18, 4308–4319.
28. W. Wang, F. Lv, B. Lei, S. Wan, M. Luo and S. Guo, Adv. Mater., 2016, 28, 10117–10141.
29. M.-F. De Volder, S.-H. Tawfick, R.-H. Baughman and A.-J. Hart, Science, 2013, 339, 535–539.
30. Y. Miyata, K. Mizuno and H. Kataura, J. Nanomater., 2011, 2011, 786763.
31. M. Huda, T. Kawahara, J.-H. Park, M. Kawasumi and Y. Matsuo, ACS Appl. Energy Mater., 2023, 6, 12226–12236.
32. Q. Chen, C.-Y. Yu, Y.-C. Zhai, T. Watanabe, M. Kawasumi, M. Huda and Y. Matsuo, Small Methods, 2025, 2500074.
33. M. Karuppannan, Y. Kim, S. Gok, E. Lee, J. Y. Hwang, J.-H. Jang, Y.-H. Cho, T. Lim, Y.-E. Sung and O. J. Kwon, Energy Environ. Sci., 2019, 12, 2820–2829.
34. S.-S. Chougule, A.-A. Jeffery, S. Roy Chowdhury, J. Min, Y. Kim, K. Ko, B. Sravani and N. Jung, Carbon Energy, 2023, 5, e293.
35. J. Deng, P. Ren, D. Deng and X. Bao, Angew. Chem., Int. Ed., 2015, 54, 2100–2104.
36. S. Sun, G. Zhang, D. Geng, Y. Chen, R. Li, M. Cai and X. Sun, Angew. Chem., Int. Ed., 2011, 50, 422–426.
37. L. Wang, Z. Zeng, W. Gao, T. Maxson, D. Raciti, M. Giroux, X. Pan, C. Wang and J. Greeley, Science, 2019, 363, 870–874.
38. F.-P. Kong, M.-N. Banis, L. Du, L.-J. Zhang, L. Zhang, J.-J. Li, K. Doyle-Davis, J.-N. Liang, Q.-S. Liu, X.-F. Yang, R.-Y. Li, C.-Y. Du, G.-P. Yin and X. Sun, J. Mater. Chem. A, 2019, 7, 24830–24836.
39. D. Zhang, R.-X. Ding, Y.-Z. Tang, L.-X. Ma and Y. He, ACS Appl. Nano Mater., 2022, 5, 10026–10035.
40. Y.-J. Xiong, Y.-N. Ma, L.-L. Zou, S.-B. Han, H. Chen, S. Wang, M. Gu, Y. Shen, L.-P. Zhang, Z.-H. Xia, J. Li and H. Yang, J. Catal., 2020, 382, 247–255.
41. D. Guo, R. Shibuya, C. Akiba, S. Saji, T. Kondo and J. Nakamura, Science, 2016, 351, 361–365.
42. L. Chen, Y. Peng, J. E. Lu, N. Wang, P. Hu, B. Lu and S. Chen, Int. J. Hydrogen Energy, 2017, 42, 29192–29200.
43. Z.-L. Jiang, S.-J. Song, X.-B. Zheng, X. Liang, Z.-X. Li, H.-F. Gu, Z. Li, Y. Wang, S.-H. Liu, W.-X. Chen, D.-S. Wang and Y. D. Li, J. Am. Chem. Soc., 2022, 144, 19619–19626.
44. S. H. Ye, W. D. Chen, Z. J. Ou, Q. H. Zhang, J. Zhang, Y. L. Li, X. Z. Ren, X. P. Ouyang, L. R. Zheng, X. Q. Yan, J. H. Liu and Q. L. Zhang, Angew. Chem., 2025, 137, e202414989.
45. R. Wang, D. C. Higgins, M. A. Hoque, D. Lee, F. Hassan and Z. Chen, Sci. Rep., 2013, 3, 2431.
46. Y. Zhou, K. Neyerlin, T.-S. Olson, S. Pylypenko, J. Bult, H.-N. Dinh, T. Gennett, Z.-P. Shao and R. O'Hayre, Energy Environ. Sci., 2010, 3, 1437–1446.
47. Z.-Y. Yao, Y.-L. Yuan, T. Cheng, L. Gao, T.-L. Sun, Y.-F. Lu, P. Galindo, L. Xu, H. Yang and H.-W. Huang, Nano Lett., 2021, 21, 9354–9360.
48. H. Huang, K. Li, Z. Chen, L. Luo, Y. Gu, D. Zhang, C. Ma, R. Si, J. Yang, Z. Peng and J. Zeng, J. Am. Chem. Soc., 2017, 139, 8152–8159.
49. M.-F. Li, Z.-P. Zhao, T. Cheng, A. Fortunelli, C.-Y. Chen, R. Yu, Q. H. Zhang, L. Gu, B.-V. Merinov, Z. Y. Lin, E. B. Zhu, T. Yu, Q. Y. Jia, J. H. Guo, L. Zhang, W.-A. Goddard III, Y. Huang and X. Duan, Science, 2016, 354, 1414–1419.
50. B.-X. Ni, P. Shen, G.-R. Zhang, J.-J. Zhao, H.-H. Ding, Y.-F. Ye, Z.-Y. Yue, H. Wei and K. Jiang, J. Am. Chem. Soc., 2024, 146, 11181–11192.
51. Z.-B. Wang, P.-J. Zuo, Y.-Y. Chu, Y.-Y. Shao and G.-P. Yin, Int. J. Hydrogen Energy, 2009, 34, 4387–4394.
52. New Energy and Industrial Technology Development Organi-zation. https://www.nedo.go.jp/content/100963953.pdf.
53. The US Department of Energy (DOE). Energy Efficiency and Renewable Energy (EERE). https://www.energy.gov/sites/prod/files/2017/05/f34/fcto_myrdd_fuel_cells.pdf.

---