Subsize Pt-based intermetallic compound enables long-term cyclic mass activity for fuel-cell oxygen reduction

Significance

Pt-based alloys are regarded as the oxygen-reduction electrocatalysts with the most potential to replace commercial Pt/C in proton exchange membrane fuel cells as long as their durability reaches the required standard. We demonstrated a series of Pt-based intermetallic compounds with ultrathin Pt skin and an average particle size down to about 2.3 nm that bring sustainable cyclic mass activity in fuel cells. Moreover, the subsize scale of Pt-based intermetallic compounds benefits the design of the confined structure for membrane electrode assembly, which further enhanced the stability under humid fuel-cell operations. Our work highlights advanced electrocatalysts for fuel cells and offers insight into the correlation between the structure and performance for membrane cathode electrode structures.

Abstract

Pt-based alloy catalysts may promise considerable mass activity (MA) for oxygen reduction but are generally unsustainable over long-term cycles, particularly in practical proton exchange membrane fuel cells (PEMFCs). Herein, we report a series of Pt-based intermetallic compounds (Pt₃Co, PtCo, and Pt₃Ti) enclosed by ultrathin Pt skin with an average particle size down to about 2.3 nm, which deliver outstanding cyclic MA and durability for oxygen reduction. By breaking size limitation during ordered atomic transformation in Pt alloy systems, the MA and durability of subsize Pt-based intermetallic compounds can be simultaneously optimized. The subsize scale was also found to enhance the stability of the membrane electrode through preventing the poisoning of catalysts by ionomers in humid fuel-cell conditions. We anticipate that subsize Pt-based intermetallic compounds set a good example for the rational design of high-performance oxygen reduction electrocatalysts for PEMFCs. Furthermore, the prevention of ionomer poisoning was identified as the critical parameter for assembling robust commercial membrane electrodes in PEMFCs.

The large-scale employment of precious platinum (∼0.4 mg/cm² on cathode) for oxygen reduction reaction (ORR) seriously increases the cost and hampers commercialization of proton exchange membrane fuel cells (PEMFCs) (1, 2). Great progress has been made to diminish the usage of platinum by alloying it with transition metals, such as Fe, Co, Ni, Cu, and Mn (3–6). These Pt-M alloy catalysts can improve the mass activity (MA) to a considerable level in comparison to the pure Pt catalysts. However, the rapid leaching of transition-metal atoms in the Pt-M alloy catalysts under acid PEMFC corrosion leads to the collapse of the catalysts’ skeleton, particularly for the catalysts with nonuniform particle sizes, which seriously impairs their MA and, ultimately, longevity (7–9).

Plenty of efforts have been devoted to enhancing the durability of Pt-M alloy systems, among which the deliberate design of ordered atomic structure is regarded as the most effective (10–12). For instance, Pt skin–covered ordered Pt-M intermetallic compounds with the transition-metal atoms occupying specific sites have been reported with high intrinsic stability in the ORR-based half cell (13–15). However, the formation of ordered structure necessitates high temperature annealing (>500 °C), which generally excludes the precise control of particle size and morphology (16, 17). This will lead to an inhomogeneous element distribution and unsatisfied particle sizes (>5 nm), causing Pt utilization efficiency and MA to decrease dramatically (18). Several studies use MgO, Fe₃O₄ or polymer coating on the surface of nanoparticles to avoid aggregation, but these methods are not very effective (19). Moreover, this covering shell usually causes residue concern on the Pt surface and only allows partial synthesis of intermetallic structures due to the limited atom mobility (20). In the last 2 y, plenty of meso-pore or micropore materials such as MOFs (metal organic frameworks) and zeolites have been selected as confined templates, but the ordered intermetallic–Pt skin core–shell structure in size control was still hardly realized (21). Therefore, the MA and durability over long-term cycling of currently available Pt alloy electrocatalysts are still inadequate to meet the requirements for practical applications (22, 23). Furthermore, due to the unsatisfied size of electrocatalysts, the related structure design of membrane electrode assembly (MEA) remains blank in fuel-cell environments (24, 25). In this context, breaking the size limitation in ordered structure transformation can maximize the electrocatalytic performance and make Pt intermetallic compounds promising candidates as next-generation commercial catalysts in PEMFCs.

Herein, we report a series of Pt-based intermetallic compounds (Pt₃Co, PtCo, and Pt₃Ti) with ultrathin Pt skin and an average particle size down to about 2.3 nm that bring outstanding cyclic MA in humid conditions of PEMFCs. By breaking size limitations in ordered atomic transformation, subsize Pt₃Co intermetallic compounds possess a high Pt atom utilization. The core–shell structure also introduces a 4% contraction strain for the surface Pt catalytic shell, which is believed to be beneficial for oxygen reduction. Consequently, sixfold MA was achieved on sub-Pt₃Co-MC compared with commercial Pt/C. Due to the improved catalytic stability, sub-Pt₃Co-MC could even deliver 15-fold MA compared with commercial Pt/C after 30,000 cycles. Furthermore, in fuel-cell tests, sub-Pt₃Co-MC displayed a considerably high MA retention of 81.5% compared to that of commercial Pt/C (33%), which is the highest level among the reported Pt alloys. Moreover, the subsize scale benefits the MEA to enhance its durability by preventing ionomer poisoning. Our work offers avenues to rationally design advanced Pt intermetallic electrocatalysts for oxygen reduction in fuel cells.

Results

Structural Characterizations of PtM Intermetallic–Pt Skin.

A special strategy with mesopore carbon confined was employed to obtain subsize Pt-based intermetallic (Pt₃Co, PtCo, and Pt₃Ti)–Pt skin core–shell catalysts (Fig. 1A and SI Appendix, Fig. S1). All the samples’ characteristic peaks in XRD (X-ray diffraction) patterns (SI Appendix, Figs. S2–S4) match well with corresponding Pt-based intermetallic compounds. Taking Pt₃Co intermetallic compound as an example, the transmission electron microscopy (TEM) and high-resolution TEM images (Fig. 1B and SI Appendix, Fig. S5) reveal that the Pt₃Co intermetallic compound is distributed homogeneously in the mesopore of carbon with a uniform size of 2.3 nm (c.f. size distribution in SI Appendix, Fig. S6). Lattice fringes indicate d values of 0.193 and 0.222 nm, which correspond to (200) and (111) crystal plane, respectively. The EDS (energy dispersive spectrometer) mapping in SI Appendix, Figs. S7 and S8 confirms the core–shell structure of Pt-based intermetallic–Pt skin. The Brunauer–Emmett–Teller analysis in SI Appendix, Fig. S9 indicates Pt₃Co nanoparticles are located in the mesopore of carbon support, and the relative simulation in SI Appendix, Fig. S10 exhibits the pore volume limited the growth of particle size. The X-ray Absorption Fine Structure (XAFS) measurements of Co K-edge and Pt L₃-edge were also performed to investigate the local structure of the mesopore-confined subsize Pt₃Co intermetallic compound. As shown in Fig. 1 C and D, the feature shape of Co K-edge in white-line peak changes indicates the evolution of crystal arrangement from hcp (hexagonal close packed) order in pure cobalt to alloying in Pt₃Co. The Co atoms were located at the vertex of a single-unit cell, as Pt atoms were located at a face-centered Co cube in the Pt₃Co intermetallic compound. The white-line peak of Pt L₃-edge gives information about electron vacancy with a decrease of integrated intensity. This result indicated an increasing 5d-electron density and the charge transfer from cobalt to platinum during the formation of intermetallic bonds. The corresponding R space curve-fitting results are provided in SI Appendix, Fig. S11 to exhibit the coordination numbers and the bond lengths of Pt–Pt, Pt–Co, and Co–Co. It can be found that the bond lengths of Pt–Co and Pt–Pt were 2.56 and 2.73 Å, respectively. To further verify the mesopore-confined intermetallic structure, the X-ray photoelectron spectra (XPS) before and after Ar ion etching are exhibited in Fig. 1E and SI Appendix, Fig. S12. The peak density of Pt 4f_5/2 and 4f_7/2 could hardly be observed on the surface of our sample, which is attributed to the confined structure of mesoporous carbon. However, after 1 keV Ar etching, the characteristic peak of Pt appears. The 4f_7/2 peak of Pt (0) located at 71.81 eV and Pt (II) was at 72.74 eV, respectively, which corresponds to the pure Pt skin of Pt₃Co intermetallic–Pt. After 3 keV Ar etching, the Pt 4f_7/2 peak shifted to a low binding energy (BE) region at 71.61 eV, which resulted from the charge transfer from cobalt to Pt in the Pt₃Co intermetallic core. The mass content evolution of the Pt, Co, and C elements is also exhibited in Fig. 1F. On the surface of the sub-Pt₃Co-MC sample, the carbon content occupied a large proportion at 95.5%. With the increasing depth of etching, the mass content of Pt first increased to 18.2% as Pt skin was detected. Then, after 3keV etching, the Pt content dropped and cobalt increased, which resulted from the signal of the intermetallic Pt₃Co core. All these results clearly confirm that the mesopore-confined subsize Pt-based intermetallic compound was successfully obtained.

Fig. 1.

Structure of Pt–M intermetallic–Pt skin with mespore confined. (A) Illustration of synthesis methods for sub-Pt intermetallic-MC_. (B) HRTEM images of sub-Pt₃Co-MC. (C) The XAFS measurements of sub-Pt₃Co-MC for Co K-edge. (D) The XAFS measurements of sub-Pt₃Co-MC for Pt L₃-edge. (E) X-ray photoelectron with Ar etching of Pt 4f spectra in sub-Pt₃Co-MC. (F) X-ray photoelectron with Ar etching of element content in sub-Pt₃Co-MC.

Electrochemical Performance.

Electrochemical oxygen reduction experiments were carried out in an acidic medium to evaluate the electrocatalytic activity of as-prepared Pt-based intermetallic samples. The Pt loading of commercial Pt/C (Hispec 3000, Vulcan XC-72), sub-PtCo-MC, sub-Pt₃Ti-MC, and sub-Pt₃Co-MC is 20.2, 19.7, 20.3, and 19.8 μg_Pt/cm² (determined by ICP-OES, inductively coupled plasma-optical emission spectrometer, SI Appendix, Fig. S13), respectively. As shown in Fig. 2A, the half-wave potential of sub-Pt₃Co-MC is 0.95 V in O₂-saturated 0.1 M HClO₄ solution, which is more positive than other electrocatalysts (an upshift by 63 mV relative to commercial Pt/C). The sub-Pt₃Co-MC catalysts were further employed as cathodic catalyst in the MEA of fuel cells (Fig. 2B). The experimental tests strictly observed DOE (Department of Energy)-Fuel Cell Technical Team protocol (26–28). As shown in Fig. 2C, at the fuel-cell voltage of 0.8 V in the polarization curves, our sub-Pt₃Co-MC–based fuel cell exhibits a current density of 313 mA/cm² (201 mA/cm² for commercial Pt/C), which was 112 mA/cm² better for sub-Pt₃Co-MC. It can also be observed that the highest power density can reach 1.77 W/cm² for our sub-Pt₃Co-MC, which was 530 mW/cm² better than commercial Pt/C. These results indicated the outstanding fuel-cell performance based on our sub-Pt₃Co-MC electrocatalysts. The ECSA (electrochemical active surface area) was measured to be 49.7 m²/g by fuel-cell H_upd in SI Appendix, Fig. S14, and specific activity was calculated to be 1.74 mA/cm² (ECSA was also verified by CO stripping in SI Appendix, Fig. S15). Cycling MA is considered a more important index, which needs to reach the high standard for long-term fuel-cell conditions. In the beginning of life, sub-Pt₃Co-MC exhibits an MA of 0.92 ± 0.02 A/mg Pt at a cell voltage of 0.9 V, which is six times that of commercial Pt/C (0.15 ± 0.01 A/mg Pt) (SI Appendix, Fig. S18). After 10,000 and 30,000 cycles in accelerated durability test (ADT) (Fig. 2D), the MA of sub-Pt₃Co-MC can remain at 0.84 A/mg Pt (8.7% loss) and 0.75 A/mg (18.5% loss), respectively, which is almost 15-fold that of commercial Pt/C (0.05 A/mg), reaching the top level of cycling MA among reported Pt alloys (Fig. 2F and SI Appendix, Table S1) (14, 29–36). At a constant-potential operation test at a high working current, the current density loss of sub-Pt₃Co-MC was negligible after 500 h (Fig. 2E). H₂–air power performance is also essential for evaluation of the practical fuel-cell performance. As shown in SI Appendix, Fig. S19, the H₂ (s = 2) and air (s = 2) are at 100% RH and a back pressure of 150 kPa in the polarization curve measurements. It can still be observed that the performance of sub-Pt₃Co-MC is better than that of commercial Pt/C, sub-PtCo-MC, and sub-Pt₃Ti-MC. The maximum peak power density for sub-Pt₃Co-MC was 0.8 W/cm², which was 100 mW/cm² higher than that of commercial Pt/C. Moreover, the stability tests for cycles and constant potential were also conducted, which indicated that relative high stability could be maintained in H₂–air fuel-cell stability tests. All the above results suggest that the sub-Pt₃Co-MC possesses outstanding activity and stability for ORR in both half-cell RDE (rotating disk electrode) and practical fuel-cell tests.

Fig. 2.

Electrochemical tests. (A) The CVs of sub-Pt intermetallic-MC and commercial Pt/C (recorded on rotating disk electrode at 1,600 rpm). (B) Scheme of sub-Pt₃Co-MC–based fuel cells. (C) The H₂–O₂ fuel-cell polarization curves of sub-Pt₃Co-MC and commercial Pt/C. (D) The MA of sub-Pt₃Co-MC and commercial Pt/C at the beginning and after 10,000 cycles and 30,000 cycles. (E) H₂–O₂ fuel-cell measurements of sub-Pt₃Co-MC in a constant-current operation of 3 A × cm⁻². (F) Comparison of cyclic MA and stability retention in fuel cells.

Discussion

Mechanism Understanding.

A subsize Pt₃Co intermetallic compound with 2∼3 atomic thickness pure Pt skin was tested as an ideal electrocatalyst, as the modified Pt skin cannot only realize better intrinsic activity but also prevent transition metal from leaching. DFT (density functional theory) calculation was employed to understand the outstanding activity and stability of sub-Pt₃Co-MC (Fig. 3A). The Pt (111) distance in Pt skin of Pt₃Co–Pt is 2.16 Å, which is 0.09 Å shorter than that of pure Pt. This indicates a high contraction strain of 4.4% for the Pt skin due to the formation of ordered atomic arrangements in the intermetallic–Pt skin core–shell structure, which further induced a stronger overlap of the 5d electron cloud on surface Pt sites. Based on d-band theory (37), the downward shift of the antibonding states between O–Pt exhibits a weaker BE between the oxide intermediate OH* and Pt sites, which promotes the intrinsic ORR activity of Pt (Fig. 3B) (38). The contraction strain was also verified by high-angle annular dark-field scanning TEM (HAADF) and in situ XAFS analysis. As shown in Fig. 3C, on account of regular patterns as dark Co atom surrounded by light Pt columns, the subsize Pt₃Co intermetallic compound with 2∼3 atomic thickness of Pt skin could be clearly observed along the [100] zone axis (39, 40). To observe the interplanar (111) spacing of Pt skin, the Pt₃Co intermetallic–Pt skin was also viewed along [110] zone axis in SI Appendix, Fig. S20. It should be noted that the (111) spacing of the Pt layer (0.217 nm) is lower than cubic Pt (0.226 nm), which indicated a 4% contraction strain in Pt skin as well. The in situ XAFS characterizations were performed under operation conditions with sub-Pt₃Co-MC at open circuit: 0.9 and 0.8 V for operando catalytic conditions (Fig. 3E). The sub-Pt₃Co-MC exhibits a lower white-line peak intensity than commercial Pt/C at the potential of 0.8 V (SI Appendix, Fig. S21), corresponding to the contraction stain with more electrons filling in 5d orbits. The Pt–Pt bond was also fitted by the IFEFFIT code in Fig. 3D for direct observation. The peak at 2.73 Å (phase correction) could be ascribed to Pt–Pt bonds in sub-Pt₃Co-MC, which is almost 0.05 Å shorter than Pt foil samples, attributed to the results of high contraction strain. The Pt L₃-edge after 10,000, 20,000, and 30,000 cycles was also recorded to verify the outstanding cycling MA of sub-Pt₃Co-MC. As shown in Fig. 3F, the Pt–Pt bonds at all cycles remain about 0.05 Å shorter than Pt foil without significant change, ascribed to the contraction strain in Pt skin maintained during cycling. A series of postcharacterizations of sub-Pt₃Co-MC after 30,000 cycles (SI Appendix, Figs. S22 and S23) consistently indicated the high stability of sub-Pt₃Co-MC during ORR.

Fig. 3.

Strain analysis. (A) DFT calculation for Pt(111) distance and charge densities distribution in Pt skin of Pt₃Co intermetallic compound and pure Pt. (B) The d-band theory analysis of sub-Pt₃Co-MC. (C) The HAADF images of as-prepared sub-Pt₃Co-MC. The red ball represents the cobalt atom and the blue one represents the platinum atom. (D) XAFS measurements fitted by the IFEFFIT code of sub-Pt₃Co-MC without phase correction. (E) The in situ EXAFS of Pt L₃-edge for sub-Pt₃Co-MC for different potentials. (F) The in situ XANES (X-ray absorption near edge structure) of Pt L₃-edge for sub-Pt₃Co-MC for different cycles without phase correction.

Besides the long-time maintained contraction strain and ordered structure of the subsize Pt₃Co intermetallic compound in favor of electrocatalysis, the specific mesopore-confined subsize structure could also benefit MEA to improve its durability in humid fuel-cell environments. With ideal size control, the subsize Pt₃Co intermetallic compound confined in ∼5-nm mesopores can prevent poisoning by ionomer and maintains oxygen transfer in the fuel-cell oxygen reduction process. The CO stripping in half-cell RDE was performed with different dosages of ionomers in Fig. 4A and SI Appendix, Fig. S25. As the content of Nafion increased to 25% and 50%, sub-Pt₃Co-MC exhibited only 23% and 47.7% loss of surface area. However, for commercial Pt/C (Hispec 3000, Vulcan XC-72), it was decreased by 44.6% and 71.6% (Fig. 4A, left axis). These results clearly indicated that the space-confined structure can prevent Pt sites from being poisoned by ionomers. To evaluate the mass transfer capacity in the space-confined structure, the oxygen permeability was also investigated. It indicates the oxygen permeability is 41.6 × 10⁹ cm³/m² for sub-Pt₃Co-MC but 9.4 × 10⁹ cm³/m² for commercial Pt/C (Fig. 4A, right axis). We conclude, with the space-confined structure, that the Nafion ionomers cannot directly cover the Pt₃Co particle. As for commercial Pt/C, Pt is exposed on the surface of carbon and can be directly covered by ionomers, leading to suppressed oxygen transfer and reduced active-site exposure (41). Because the nanoparticle is not in direct contact with ionomers in sub-Pt₃Co-MC, the proton transfers rely on the water film on the inner wall of the channel and ionomers. As shown in Fig. 4C and SI Appendix, Fig. S26, the model for enhanced performance in humid fuel-cell conditions was proposed. Moreover, the seepage mechanics analysis method was further used to prove the remarkable advantage of the space-confined structure in sub-Pt₃Co-MC (42–44). Molecular dynamics (MD) simulations were adapted to reveal the transport behavior of oxygen and protons in mesopore-confined subsize Pt₃Co (45, 46). As shown in Fig. 4B and SI Appendix, Fig. S27, under the circumstances of ionomer poison, the oxygen diffusion within different pore sizes was recorded as a function of time. It was found that a pore size of 5 nm exhibited the best performance for oxygen diffusion, preventing ionomer poisoning. The rate of oxygen diffusion increases with pore size rising, and the larger pore size further limits proton transfer due to the limited diffusion coefficient (SI Appendix, Fig. S28). In our case, sub-Pt₃Co-MC was confined in 5-nm mesopores, which could maintain efficient gas transfer without interference of ionomer (Fig. 4E). However, for commercial Pt/C, ionomers seriously hinder the contact between surface Pt and oxygen. Based on the above analysis, the subsize structure of Pt₃Co also benefits MEA design in fuel-cell environments, in which the confined structure can prevent ionomer poisoning electrocatalysis for better durability in humid fuel-cell conditions.

Fig. 4.

Triple-phase boundary for MEA. (A) Surface loss by CO stripping and oxygen permeability measurements. (B) The proton and gas diffusion capacity with pore-size change. (C) The antipoison effect with three-phase interface in sub-Pt₃Co-MC. (D) The poison effect by Nafion ionomer in commercial Pt/C. (E) The simulation of three-phase interface in sub-Pt₃Co-MC.

In this work, we report a mesopore-confined strategy to synthesize a series of Pt-based intermetallic compounds (Pt₃Co, PtCo, and Pt₃Ti) enclosed by ultrathin Pt skin with an average particle size down to about 2.3 nm, bringing ultralong-term cyclic MA in fuel cells. In H₂-O₂ fuel-cell tests, the as-prepared sub-Pt₃Co-MC exhibits a high MA of 0.92 ± 0.02 A/mg, which is about six times that of commercial Pt/C. As for cycling ADT, the MA decayed only by 18.5% after 30,000 cycles, which represents the top cycling MA among reported Pt alloy ORR catalysts. According to theoretical and experimental analysis, the 2.3-nm Pt₃Co intermetallic compound was verified to exhibit contraction strain, which improved the intrinsic activity and stability. Moreover, the interaction between subsize Pt₃Co and confined support in MEA was investigated by MD simulations. Our work paves the way toward the rational structural design of Pt alloy electrocatalysts for fuel cells.

Materials and Methods

Synthesis of Subsize Pt₃Co Intermetallic Compounds with Mesopore Confine (Denoted as Sub-Pt₃Co-MC).

Typically, 23.4 mg H₂PtCl₆ ⋅ 6H₂O, 5 mg CoCl₂ ⋅ 6H₂O and 14 mg tetrabutylammonium chloride were dissolved in 2 mL ethanol. The mixture was stirred at room temperature for 1 h, which is then added to 71 mg Ketjenblack-600. Then, the mixed solution was stirred for 6 h at room temperature until almost all ethanol was volatilized. After that, the precursor was annealed in H₂/Ar (5%) atmosphere at 800 °C for 2 h to obtain the mesopore-confined subsize Pt₃Co intermetallic compound. Afterward, the Pt₃Co intermetallic compound was acid etched at 80 °C for 8 h in a 30% nitric acid solution and annealed at 800 °C for 1 h under H₂/Ar (5%) gas atmosphere to obtain the mesopore-confined subsized Pt₃Co–Pt core–shell structure.

Synthesis of Subsize PtCo Intermetallic Compound with Mesopore Confine (Denoted as Sub-PtCo-MC).

The synthesis procedure is similar with the subsize Pt₃Co intermetallic compound with a mesopore confine. The only difference is the amount of CoCl₂ ⋅ 6H₂O increasing to 15 mg.

Synthesis of Subsize Pt₃Ti Intermetallic Compound with Mesopore Confine (Denoted as Sub-Pt₃Ti-MC).

The synthesis procedure is similar with the subsize Pt₃Co intermetallic compound with a mesopore confine. The only difference is Pt₂(dba)₃ (bis dibenzylidene acetone) was used as the Pt source and TiCl₄(THF)₂ was used as the Ti source. Then, the annealing temperature was raised to 950 °C.

Material Characterization.

The crystal structure of as-prepared products was examined by an X-ray powder diffraction (Philips X’ Pert Pro Super diffractometer) instrument with Cu K_a radiation (λ = 1.54178 Å). XPS spectra were recorded on an ESCALAB MK II X-ray photoelectron spectrometer with Mg Kα as the excitation source. The BE scale calibration was conducted using the standard Au 4f_7/2. XPS software was used for analyzing the spectra. All spectra were charge corrected according to the C1s main peak (BE of 284.8 eV). The TEM images were operated by JEM-2100F field-emission electron. The sample was dispersed by ethanol and dropped on the copper grids. The sample was mixed with graphite by grind for XAFS analysis, and the source beamline was BL14W1 in Shanghai Synchrotron Radiation Facility. The operation protocol for HAADF-STEM (high-angle annular dark field scanning transmission electron microscope) is similar with the transmission electron microscope. The characterization instrument is JEOL JEM-ARF200F (28).

Electrochemical Tests.

The as-prepared electrocatalyst was first placed in 1 mL 1:1 water/ethanol solution with a serum bottle. Then, the mixed ink was further prepared through ultrasonic dispersion for 60 min. Using the ink, the sample was spray coated on glassy carbon. The loading of Pt for each electrocatalyst was determined as 0.02 mg/cm², and 0.1 M HClO₄ was selected as the electrolyte. The Pine ASR instrument and CHI760D electrochemical station were used for half-cell tests. Luggin capillary–based electrochemical cells were used. The electrolyte was selected as 0.1 M HClO₄ with sufficient O₂ saturated in standard conditions. The Nernst equation was employed for the calibration of the potential between Ag/AgCl reference electrodes and reversible hydrogen electrodes (28). The counterelectrode is graphite rod. The rotating speed is 1,600 rpm with a sweep rate of 10 mV/s. The CO adsorption procedure in CO-stripping voltammogram measurements was accomplished by polarizing the electrode at 0.2 V with CO bubbling in electrolyte solution for 10 min to adsorb monolayer CO molecules. Then, the electrode was transferred to another cell filled with Ar-saturated 0.1 M HClO₄ solution (without CO). Then, cyclic voltammograms (CVs) were conducted from 0.05 to 1.2 V with a scan rate of 50 mV/s. The theoretical charge per unit area was used as Q_theo, Pt = 420 mC/cm² and corrected for capacitive contributions (the 2-s cycle).

Fuel-Cell Measurements.

MEA was prepared by a hot pressing method. Catalyst “ink” was prepared by ultrasonic dispersion of catalyst (1.4 to 1.6 mg) and 5 weight (wt)% Nafion solution in 1.0 mL deionized water for 1 h. Toray carbon paper (1.2 × 1.2 cm²) pretreated with PTFE was used as the cathode, and the ink was directly coated on it without using a microporous layer. The platinum loading of all catalysts was about 0.2 mg/cm². The anode catalyst was 60 wt% Pt/C, and the loading was 0.2 mg/cm². Then, the MEA was prepared by pressing cathode, anode, membrane NRE 211 Nafion membrane, and gasket at about 3 MPa for 180 s around 150 °C. The compression ratio was 15 to 20%. The effective area of MEA was 1.2 × 1.2 cm². The catalyst-based MEA was tested at 353.15 K (80 °C) with 100% RH using the 850e fuel-cell test system (Scribner Associates Inc.). The total outlet pressure was adjusted to 150 kPa with the stoichiometric flow rates of anode (s = 2) and cathode (s = 9.5 for O₂ and s = 2 for air). Before the polarization curves were recorded, the MEA was fully activated by holding at 0.9 V for 20 min and 0.7 V for 20 min to stabilize potential and current density. For MA measurements, the potential was also held at 0.9 V to measure the MA (holding 30 min, sample interval = 2 s). The square wave catalyst AST consisted of potential cycling between 0.6 and 0.95 V at 3 s at each potential for 30,000 cycles for a total test time of about 50 h with potential changed at 700 mV/s. This AST was run at 200 SCCM (standard cubic centimeter per minute) H₂/N₂, 100% RH and atmospheric pressure. The ECSA of the cathode was measured using a CV between 0.03 and 1.15 V (versus anode) in fuel cells. A sweep rate of 25 mV/s was used with 0.5 SLPM (standard liter per minute) of 4% H₂ (balance nitrogen) flowing over the anode (as reference as well as counterelectrode) and 0.5 SLPM of N₂ flowing over the cathode (as working electrode) at 80 °C, 150 kPa. The ECSA values were calculated by integrating the hydrogen adsorption charge in the voltammogram (0.05 to 0.35 V), 210 µC/cm² (theoretical hydrogen monolayer adsorption on Pt), and the cathode initial Pt loading (mg_Pt/cm², per milligram platinum in per square meter).

MD Simulations.

The gas-proton nano-channel transport model is composed of two walls (carbon slabs), oxygen, water, and proton particles. The interaction field of atoms consider long-range Coulombic interactions and Lennard-Jones (LJ) pairwise potential (43). The energy between intermolecular atoms of water and oxygen should concurrently account for the Coulombic and LJ energies. The Lorentz–Berthlot mixing rules were used to determine the LJ potential parameter. The simple point charge (SPC/E) model was employed for water molecules, and the other field parameters of carbon, oxygen, and proton were acquired (44, 45). The cutoff for LJ interactions was 12 Å, and the Coulomb forces were calculated by using the particle–particle particle–mesh method (accuracy: 10^-4). The temperature of the whole system was controlled by a Nose–Hoover thermostat in NVT (Number, Volume and Temperature Canonical Ensemble) ensemble. The simulation box was applied with periodic boundaries along the three directions, and the time for each calculation step was 1.0 fs. All the simulations were carried out based on the open-source software Large-scale Atomic/Molecular Massively Parallel Simulator (46).

Data Availability

All study data are included in the article and/or SI Appendix.