Intermetallic Platinum Alloy Nanoparticles and Single Atoms Hybrid Electrocatalysts for Proton Exchange Membrane Fuel Cells

Proton exchange membrane fuel cell converts hydrogen and oxygen into electricity with zero emission 1 . The high cost and low durability of Pt-based electrocatalysts for oxygen reduction reaction hinder its wide applications 2,3 . The development of non-precious metal electrocatalysts also reaches the bottleneck because of the low activity and durability 4,5 . Here we rationally design a hybrid electrocatalyst consisting of atomically dispersed Pt and Fe single atoms and intermetallic PtFe alloy nanoparticles. The Pt mass activity of the hybrid catalyst is 3.5 times higher than that of commercial Pt/C in a fuel cell. More importantly, the fuel cell with an ultra-low Pt loading in the cathode (0.015 mg Pt cm -2 ) shows unprecedented durability, with 93.6% activity retention after 100,000 cycles and no noticeable current drop at 0.6 V for at least 206 h. These results highlight the importance of the synergistic effects among active sites in hybrid electrocatalysts and provide an alternative way to design more active and durable low-Pt electrocatalysts for electrochemical devices.


Introduction
Proton exchange membrane fuel cells (PEMFCs) as a promising clean energy conversion technology have gained considerable attention. However, the high cost and low durability of Pt-based nanocatalysts for the cathodic oxygen reduction reaction (ORR) hinder the wide adoption of this technology 2,6 .
According to the ultimate cost target of $30 kW − 1 for the fuel-cell stack, the Pt loading in the catalyst layers must be below 0.125 mg cm − 2 . However, as the Pt loading decreases, the oxygen transfer resistances increase because of limited accessible active sites, resulting in more deteriorated durability 7 . Thus, the ambition of developing low-Pt-loading cathodes poses great challenges in the areas of Pt utilization and intrinsic durability of Pt-based electrocatalysts. Despite great efforts on the development of advanced Pt-based catalysts to improve the Pt utilization and mass activity (MA) toward ORR 1,8 , their high activities/durability measured in liquid cells have rarely been realized in fuel cells. On the other hand, carbon-based Pt group metal (PGM)-free ORR electrocatalysts consisting of highly dispersed transitionmetal single atoms in nitrogen-coordinated carbon surfaces (Me-N-C) are promising candidates to replace Pt 5 . Unfortunately, the poor durability of Me-N-C has limited their practical applications 9 . Recently, Liu et al. reported a hybrid catalyst consisting of Pt-Co alloy nanoparticles supported on Co-N-C with an unprecedented ORR activity (1.77 A mg Pt −1 at 0.9 V) 10 . Despite the excellent Pt MA of this hybrid ORR catalyst, it still suffered signi cant activity losses during potential cycling (83% after 30,000 cycles between 0.6 and 0.95 V) and potential hold (45% after 22 h at 0.75 V) 10 .
Herein, we report a hybrid electrocatalyst (denoted as Pt-Fe-N-C) consisting of multiple types of active sites that shows not only a tenfold higher Pt MA, but also unprecedented durability. The performance loss is negligible even after 100,000 potential cycles, and no current drop is observed for at least 206 h at 0.6 V in a fuel cell test with an ultra-low Pt loading (0.015 mg Pt cm − 2 ) in the cathode. To the best of our knowledge, no comparable durability for a cathode with an ultra-low Pt loading has previously been reported in fuel cells.
Structure And Composition Of Pt-Fe-N-c Figure 1a shows a typical transmission electron microscopy (TEM) image of the as-synthesized Pt-Fe-N-C catalyst, clearly revealing nanoparticles with a main size distribution of 2-3 nm ( Supplementary Fig. 1) dispersed on a carbon substrate with a Brunauer-Emmett-Teller surface area of 750 m 2 g − 1 ( Supplementary Fig. 2). High-angle annular dark-eld scanning transmission electron microscopy (HAADF-STEM) images of Pt-Fe-N-C ( Fig. 1b and Supplementary Fig. 3) show a high density of isolated atoms anchored on the carbon substrate in addition to nanoparticles. A representative atomically resolved STEM image of a single particle (Fig. 1c) shows the atomic arrangement view along the [100] direction. The characterized atom contrast as well as the lattice distance of 3.78 Å (001) reveals an ordered structure, which is consistent with the atomic model and simulated STEM image of a facecentered-tetragonal PtFe structure. It is worth noting that the ordered structure can be clearly identi ed in almost every nanoparticle ( Supplementary Fig. 3). Two sets of bright spots, distinguished by their Zcontrast difference 11 , could be detected in Fig. 1d and Supplementary Fig. 4. As shown in Fig. 1e, the local electron energy loss spectroscopy (EELS) pro le of the nanoparticle shows strong Fe and O signals, while the Pt signal could not be detected owing to the high energy loss of the Pt M-edge 12 . The pro le of the single atom in a blue dashed circle shows weak Fe and N signals, suggesting the weaker spot to be an Fe-N moiety. In contrast, the pro le of the atom in a red dashed circle only shows a weak N signal, indicating the brighter spot to be a Pt-N moiety. These results are in good agreement with their contrast differences in STEM images.
According to the inductively coupled plasma mass spectrometry results, the metal loadings in Pt-Fe-N-C are 2.0 wt% Fe and 1.7 wt% Pt . The oxidation states of metals and types of nitrogen dopant are characterized by X-ray photoelectron spectroscopy in Supplementary Fig. 5. The structure of Pt-Fe-N-C was further characterized by X-ray absorption spectroscopy (XAS). Indicating from X-ray absorption nearedge structure (XANES) result, there was no obvious difference in Pt L 3 -edge (Fig. 1f) in the pre-edge region in comparison with a metallic Pt foil. The stronger intensity of the white line for Pt-Fe-N-C, resulting from the electron transfer from Pt 5d to Fe 3d orbitals, suggested that Pt is in the oxidized form in the Pt-Fe-N-C 13 . For the Fe K-edge in Fig. 1g, the intensity of the pre-edge peak is the highest for the Pt-Fe-N-C sample, while the white-line intensity has decreased dramatically compared with Fe-N-C and Fe oxides. The formation of Pt-Fe alloys and coordination information of Pt and Fe single atoms were con rmed by Fourier transforms of extended X-ray absorption ne structure data (EXAFS, Supplementary Fig. 6) and relative tting results (Supplementary Fig. 7 and Supplementary Table 1).

Performance Evaluation Of Pt-Fe-N-C
A cyclic voltammogram (CV) curve of Pt-Fe-N-C did not present profound hydrogen adsorption/desorption peaks characteristic of Pt because of the ultra-low Pt loading ( Supplementary Fig. 8). Pt-Fe-N-C showed a signi cantly improved ORR activity over Fe-N-C, with the half-wave potential (HWP) shifting from 0.790 to 0.909 V ( Supplementary Fig. 9). The corresponding Pt MA of Pt-Fe-N-C at 0.9 V reached 1.74 A mg Pt −1 , which was more than ten times that of Pt/C (0.18 A mg Pt −1 ) and much higher than most of the Pt-based electrocatalysts (Supplementary Table 2 Table 2).
The cells assembled with Pt-Fe-N-C, Pt/C and Fe-N-C cathodes were further subjected to accelerated durability testing under repeated square-wave cycles at 0.6 and 0.95 V by holding at each potential for 3 s, following the DOE testing protocol. As shown in Fig. 2b, the fuel cell polarization curves with the Pt-Fe-N-C cathode showed no change after 30,000 cycles, and a negligible change after 100,000 cycles. The cell retained the MA of 0.88 A mg Pt −1 at 0.9 V and the power density of 0.90 W cm − 2 at 2.0 A cm − 2 even after 100,000 cycles. These results surpassed DOE's durability goal of less than 40% MA loss after 30,000 cycles. In comparison, Pt/C and Fe-N-C cathodes showed substantial performance loss after 30,000 cycles ( Supplementary Fig. 13).
The morphology and structure of Pt-Fe-N-C catalyst after 100,000 cycles in a fuel cell were further analyzed by STEM-EELS. Abundant single atoms with the preservation of N coordinated con guration were still uniformly distributed on the carbon support ( Fig. 3a, b). However, the structure and composition of nanoparticles did change during potential cycling ( Supplementary Fig. 14). If the particle size was smaller than ~ 4 nm, a solid PtFe@Pt core-shell structure was formed (Fig. 3c), indicating by the lattice spacing of 0.191 nm for PtFe (002) in the core and 0.204 nm for Pt (002) in the shell. This conclusion was further supported by the line intensity pro le (Fig. 3d). A periodic oscillation of intensity in the center and monotonicity in the shell were observed, which can be attributed to the contrast differences between Pt and Fe in an ordered lattice. Bigger particles tended to form a percolated structure (Fig. 3e) with a lattice spacing of 0.204 nm, which is close to the spacing of Pt (002). The energy-dispersive X-ray spectroscopy (EDX) line pro le of Pt (Figs. 3f, g) shows a clear concavity in the middle of the particle, indicating the formation of a pit in the nanoparticle. Relative EDX mapping (Supplementary Fig. 15) also implies the formation of a Pt-rich percolated structure. Similar phenomena have been observed in Pt-Ni alloy nanoparticles during potential cycling 17 . It is worth noting that the majority of the PtFe nanoparticles were transformed into a more stable core-shell structure, which played an important role in achieving the unmatched durability of Pt-Fe-N-C. The high durability of intermetallic nanoparticles has been reported by other groups 18 .
A chronoamperometric test at a voltage of 0.6 V under H 2 /air was also conducted. As shown in Fig. 3c, a Pt-Fe-N-C cathode showed a nearly constant current density during the 206-h measurement, while a Pt/C cathode with a much higher Pt loading suffered rapid deterioration, losing approximately 55% of the current density after only 100 h. For Fe-N-C, the current density drop was even faster, with 75% loss after 62 h, which was similar to results reported in the literature 19,20 . Iron is a concern for Na on-based membranes and ionomers because of the Fenton reaction 21 . The F − concentrations in e uent water at 0.6 V were monitored as a function of time to assess their degradation rate. As shown in Fig Table 4, Pt-N 1 C 3 exhibited the best ORR activity, with a downhill trend across all of the elementary reactions except the last *OH protonation step, and the energy barrier during the ORR was only 0.17 eV, which was much smaller than that on Fe-N 1 C 3 (0.53 eV), Pt-N 2 C 2 (0.54 eV) and Fe-Pt dual metal (0.75 eV). The Pt ML /PtFe(111) was also predicted to be active for ORR because of a low energy barrier of 0.18 eV in the nal *OH protonation step, which was close to that on Pt-N 1 C 3 (0.17 eV). Such a core-shell structure with a 1-ML Pt skin shows better ORR activity than Pt(111) because of the weaker bindings to reaction intermediates, especially for *O and *OH. Simulations were also conducted on 2Pt ML /PtFe(111) and 3Pt ML /PtFe(111). As shown in Supplementary Fig. 21, all Pt skins showed better ORR activities than pure Pt. Among them, Pt ML /PtFe(111) was the best, with the lowest barrier for the *OH-to-H 2 O step. Thus, Pt-N 1 C 3 and core-shell nanoparticles formed after leaching Fe in the surface and sub-surfaces were proposed to be the most active sites in Pt-Fe-N-C.
One of the major concerns with Me-N-C catalysts in fuel cells is the formation of a large amount of H 2 O 2 as the nal product, which is detrimental to the membrane and ionomers 22 . In our hybrid catalysts, H 2 O 2 generated at the Fe-N-C or even Pt-N-C sites may be further reduced to H 2 O on nearby Pt-Fe nanoparticles.
To validate this hypothesis, H 2 O 2 reduction on Pt ML /PtFe(111) and Pt(111) surfaces was compared ( Fig.   4b and Supplementary Pt-Fe-N-C hybrid electrocatalyst. According to our previous work, impregnation of Fe-N-C with an ultra-low Pt loading could signi cantly promote the durability of Fe-N-C, while it did not present noticeable activity improvement, especially in the kinetic region 24 . In this follow-up work, ammonia heat treatment was designed to expose more active sites, tailoring the properties of carbon support and the coordination number of single atoms. Secondary Ar pyrolysis after ammonia treatment is designed to stabilize the carbon framework. Thus, Fe-doped ZIF-8 was used as the support for Pt impregnation in this work for simplifying heat treatment protocols. 10 mg platinum (II) acetylacetonate was homogeneously dispersed in 3 mL ethanol via ultrasonication until the solvent became transparent. In the meantime, 110 mg 1,10phenanthroline monohydrate dissolved in 10 mL ethanol was added to the Pt solution to provide a su cient nitrogen source for Pt coordination. Around 400 mg of Fe-doped ZIF-8 was added to the above solution to form a uniform suspension. After drying at 60°C in a vacuum oven overnight, the solid was collected and ball milled (ZrO 2 ball, 350 rpm, 4 h) to uniformly distribute Pt and N sources on the Fedoped ZIF-8 support. The mixed precursors were rst treated in NH 3 gas at 900°C for 15 min and then transferred to an Ar atmosphere at 1000°C for 1 h to remove Zn in the precursor and stabilize the whole carbon framework. The nal catalyst was denoted as Pt-Fe-N-C. The Pt and Fe loadings were around 1.7 wt% and 2.0 wt%, respectively.
Pt-N-C electrocatalyst. As a reference electrocatalyst, Pt-N-C was prepared from a ZIF-8 support, which was formed by mixing 1mM Zn(NO 3 ) 2 ·6H 2 O and 8.21 g 2-methylimidazole in a methanol solvent and following the same collecting and drying protocols as for Fe-doped ZIF-8. Similarly, 10 mg platinum (II) acetylacetonate was homogeneously dispersed in 3 mL ethanol via ultrasonication until the solvent became transparent. In the meantime, 110 mg 1,10-phenanthroline monohydrate dissolved in 10 mL ethanol was added to the Pt solution to provide a su cient nitrogen source for Pt coordination. Around 400 mg ZIF-8 was added to the above solution to form a uniform suspension. After drying at 60°C in a vacuum oven overnight, the solid was collected and ball milled (ZrO 2 ball, 350 rpm, 4 h) to uniformly distribute Pt and N sources on the ZIF-8 support. The mixed precursors were also initially treated in NH 3 gas at 900°C for 15 min and then transferred to an Ar atmosphere at 1000°C for 1 h. The nal catalyst was Pt-N-C with a Pt loading of 2.3 wt%.

Physical characterization
The TEM data were collected with a double Cs-corrected FEI Themis G2 operating at 300 kV. HAADF-STEM, EDX, and EELS were performed to obtain detailed structure and composition information on the hybrid catalyst with atomic resolution. Bulk and surface compositions were evaluated through inductively coupled plasma mass spectrometry (Varian 820) and X-ray photoelectron spectroscopy (Axis Ultra DLD), respectively. The surface area of the catalyst was measured by the Brunauer-Emmett-Teller (AUTOSORB-1) method. XAS characterizations, including XANES spectroscopy and EXAFS spectroscopy, were performed at beamline 20-BM of the Advanced Photon Source at Argonne National Laboratory, using a Si(111) monochromator. XAS data were processed with the Athena and Artemis software package. Micro-Raman (InVia (Renishaw)) was used to probe the change in the carbon support before and after durability measurement. E uent water samples collected from the cathode gas outlet were analyzed by ion chromatography (Metrohm/881with UV and conductivity detector) to evaluate the uoride emissions.

Electrochemical measurement
To prepare the catalyst ink, 2.5 mg of catalyst was uniformly dispersed in 500 µL of mixed solvent (water and isopropanol at a 4:1 volume ratio) and 10 µL of a 5 wt% Na on™ 117 solution. The rotating-ring disk electrode (RRDE, 5.5 mm in diameter) was polished with Al 2 O 3 powder (50 nm). 10 µL of catalyst ink was dropped onto the electrode. After drying in air, the thin-lm electrode was evaluated by an electrochemical workstation (CHI 760E). A carbon rod and Ag/AgCl were used as counter electrode and reference electrode, respectively. All potentials are referred to the reversible hydrogen electrode (RHE). 20 cycles of CVs in the potential range of 0 to 1.2 V at 100 mV s -1 were applied to clean the thin lm in an Ar-saturated 0.1 M HClO 4 solution, followed by taking a stable CV curve in the same potential range at 50 mV s -1 . The The collection e ciency of the RRDE was rst determined in an Ar-saturated 10 mM K 3 Fe(CN) 6 +1 M KOH solution. The electrode was rotated at 1600 rpm (corresponding to the angular velocity applied during the measurement) and the amperometric i-t measurements were performed by setting the ring and disk voltages to1.5 V and 0.1 V, respectively.
Page 10/20 The disk (I d ) and ring (I r ) currents were recorded. The measurement was repeated once with a disconnected disk to obtain another ring current (I r0 ), which includes all currents (not from Fe(CN) 6 4-) reduced on the disk. The collection e ciency could be calculated from: The where and are the reference pressure of 1 bar. While and during the measurement are 2.5 bar and 1.5 bar, respectively. is the effective density and is assumed to be 0.4 g·cm -3 for a non-precious-metal catalyst in a porous cathode 26 . is the transfer coe cient and theoretically < 1, and is generally assumed to be 1 in the calculation 16,27 .
Theoretical calculations DFT calculations were performed to explore the ORR mechanisms of Pt-Fe-N-C hybrid electrocatalysts by using the Vienna Ab-Initio Simulation Package (VASP) code. 28,29 The projector augmented wave method 30,31 pseudopotentials with the revised Perdew-Burke-Ernzerhof generalized gradient approximation (GGA-RPBE) 32 functional, which were provided in the VASP database, were used to describe the electron-ion interactions. The plane-wave cutoff energy was set to be 400 eV. The Gaussian smearing scheme was used with a width of 0.1 eV. For structure optimizations, the total energy convergence was set to be smaller than 1 × 10 -5 eV, and the force convergence was set to be lower than 0.01 eV/Å on atoms. Dipole corrections were applied in all the slab simulations. Spin polarization was considered in our calculations. The DFT-D3 method of Grimme with zero damping 33  grids of 3 × 3 × 1 and 5 × 5 × 1 were sampled for carbon-based models and core-shell slab models, respectively. The GGA+U (U = 3.29 eV) was used to describe the localized 3d orbital electrons for Fe atoms, considering the magnetic moment of Fe correctly [35][36][37] .
The calculation of Gibbs free energy for each elementary step is based on the computational hydrogen electrode scheme proposed by Norskov and co-workers 38 , which is calculated at 298 K and 1 atmosphere according to the equation G = E total + ZPE -TS, where E total can be directly obtained from DFT calculations, and ZPE and TS are the zero-point vibrational energy correction and entropy correction, respectively. The ORR reaction pathway on various active sites was considered as follows: Therefore, at U = 0 (vs RHE) and standard condition, the free energy for elementary steps can be calculated as follows:  showing the mass content differential between Fe and Pt and further revealing the formation of a pit in the center.