Universal and Energy‐Efficient Approach to Synthesize Pt‐Rare Earth Metal Alloys for Proton Exchange Membrane Fuel Cell

Abstract Traditional synthesis methods of platinum‐rare earth metal (Pt‐RE) alloys usually involve harsh conditions and high energy consumption because of the low standard reduction potentials and high oxophilicity of RE metals. In this work, a one‐step strategy is developed by rapid Joule thermal‐shock (RJTS) to synthesize Pt‐RE alloys within tens of seconds. The method can not only realize the regulation of alloy size, but also a universal method for the preparation of a family of Pt‐RE alloys (RE = Ce, La, Gd, Sm, Tb, Y). In addition, the energy consumption of the Pt‐RE alloy preparation is only 0.052 kW h, which is 2–3 orders of magnitude lower than other reported methods. This method allows individual Pt‐RE alloy to be embedded in the carbon substrate, endowing the alloy catalyst excellent durability for oxygen reduction reaction (ORR). The performance of alloy catalyst shows negligible decay after 20k accelerated durability testing (ADT) cycles. This strategy offers a new route to synthesize noble/non‐noble metal alloys with diversified applications besides ORR.

min.Putting urea in muffle furnace in air at 550 o C for 3 h is to obtain g-C3N4.First, the g-C3N4 as ligand according to the reported method, and nitrogen doped Ketjen Black EC600JD with a high specific surface area of 1164.2 m 2 /g (Figures S28-S29) was used as support to uniformly disperse the metal salts.The detailed feeding mass of the reactants are presented as Table S6 The mixtures were filled into a high-purity graphite rod, loaded into the electric arc furnace, later pyrolyzed by general thermal treatment at different current from 70-100 A for 20-120 s.With the utilization of an infrared temperature sounder, we are able to obtain real-time temperature measurements.Subsequently, we can easily regulate the temperature by controlling the current, ensuring convenient operational control.Furthermore, the timing mechanism enables us to terminate the powder supply whenever necessary.Then, the obtained materials were treated in a beaker with 1 mol L -1 H2SO4 at 60 o C for 1 h.Additionally, after filtration, the as-synthesized Pt-RE alloy is basically wrapped by thin carbon shell (Figure S30), so it needed to be removed and put in the air at 200 o C for 12 h by mild oxidation removal.And obtained a family of Pt-RE with the 8wt% loading capacity of Pt on N-KJB.We adjusted the atom ratio of Pt and RE to achieve the optimal atom ratio of 1:2.Taking Pt-Ce alloy as typical example, when the atom ratio of Pt: Ce was 2:1, the obtained XRD pattern implied that only the diffraction peak of Pt (Figure S31) appeared.When the ratio of Pt: Ce decreased to 1:4, the alloy phase was complex, containing several phases of Pt-Ce binary alloys (Figure S32).Besides, Pt-Gd treated in 80 A but different time (80 s, 100 s, and 120 s) also showed the same law compared with Pt-Ce treated in 80 A (80 s, 100 s, and 120 s), the details shown in Figures S33-S38.
Physical characterization.Powder X-ray diffraction (XRD) measurements were performed with a MiniFlex 600 X-ray diffractometer (Rigaku) using a Cu kα (λ= 1.5418 Å) radiation source with 30 kV and 15 mA.The scan speed was 1 o per minute.The Scherrer formula: was used to calculate the grain size D, the average grain size is , the crystal shape factor K values 0.9, the diffraction angle θ and semi-peak width β, respectively.Inductively coupled plasma-optical emission spectroscopy (ICP-OES) analyses were performed in Agilent ICP-OES 720.The results based on quantitative inductively coupled plasma-optical emission spectroscopy (ICP-OES) analysis (Table S7) indicated that the molar ratio of Pt to RE in the alloy were nearly similar.Based on the above results, Pt−RE alloys with the Pt loading of about 9 wt% on the N-doped carbon support were obtained following this synthesis route.Xray photoelectron spectrometry (XPS) was carried out on a Thermo Scientific K-Alpha+ X-ray photoelectron spectrometer with an Al Kα X-ray monochromator.The auger spectrum of Ce MNN is shown in Figure S39, further confirming the metallic state of Ce. 1 Transmission electron microscope (TEM) images were obtained at 200 kV with a Talos-F200X and Tecnai G 2 F20.And the high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images were obtained with a Talos-F200X at 200 kV.The inner morphology of the as-synthesized samples was characterized by transmission electron microscopy.High-resolution (HR)-TEM used to analyze the morphology of samples.More than 300 particles from 3 to 5 TEM images were counted for statistics to determine the particle size distribution of each sample.
The elemental mappings were acquired by EDS characterization.In additionally, we also investigated the element composition from different particles (Figure S40), the results show that Pt and Re are well-dispersed in each particle.The AC-TEM image was obtained by JEM-ARM300F.The specific surface area of carbon support N-KJB were measured using the Brunauer-Emmett-Teller (BET) surface area analyzer (Autosorb IQ).

Energy consumption calculation.
We use electric arc furnace (Panasonic, YD -400SS) to synthesize.As it shown in the machine's instruction, "60 A / 22.4 V ~ 400 A / 36 V", there is a linear relationship between voltage and current.Taking Pt-Ce-80A-100s as an example, when the current is 80 A, voltage values 23.2 V. We used to calculate the energy consumption by eq 1 and 2.

𝑊 = 𝑃 • 𝑡 (eq 1)
Where P is the power, t is the synthesis time.

𝑃 = 𝑈 • 𝐼 (eq 2)
Where U is the output voltage, I is the current.So, during the 100 s, energy consumption W is about 0.052 kW•h.
In the calculation process, the processing time refers to the sum of the heating time and holding time.When calculating the energy consumption required for the process reported in the references, the working parameters of the OTF-1200X1200 Celsius opening tube furnace from Hefei Kejing Company were used for calculations, and the power is 3 kW.
The thermal consumption  is calculated by eq R3 and the energy efficiency η is calculated by eq R4.
Where M is the mass of graphite rod, C is the specific heat capacity and ΔT is the temperature difference before and after reaction.
Where   is output power,   is input power,  is the total energy consumption, and  is thermal consumption.Take Pt-Ce-80A-100s as an example.The average mass of graphite rod is 2.5×10 -3 kg, C is 710 J/(kg•K), and ΔT is nearly 1000 K, thus Q values 1775 J, equals 4.93×10 - 4 kW•h.While the total energy consumption is about 0.052 kW•h, thus η values nearly 99.1%.

Electrochemical measurements.
In order to assess the ORR performance, a typical three-electrode configuration with rotating disk electrode (RDE) setup was applied.A Pt foil was applied as the counter electrode, and a commercial Ag/AgCl was used as the reference electrode.5 mg of the catalysts were added to 1 mL of solution consisted of Nafion solution (5 wt%) and iso-propanol.The Pt loading of Pt-RE and the commercial Pt/C (20 wt %, Johnson-Matthey Co.) on the glassy carbon electrode (0.196 cm 2 ) was kept at 8.9 μgPt/cm 2 .All electrochemical measurements were performed in 0.1 M HClO4 at 25 o C. Cyclic voltammetry (CV) measurement was conducted in O2-saturated 0.1 M HClO4 for at least 60 cycles.Secondly, the linear sweep voltammetry (LSV) curves were collected at 10 mV/s to evaluate the ORR activity, and the curves were calibrated with the background subtraction and solution resistance.According to the Koutecky-Levich equation (eq 5), the kinetic currents could be calculated.The mass activity (MA) could be calculated according to eq 6.And as for eq 7, the specific activity (SA) could be calculated.
where i is the measured current and iL is the diffusion-limited current.
where mPt is the Pt loading of the catalyst used in the working electrode.
where ECSA is the electrochemical surface area of the catalyst, which can be estimated by a hydrogen underpotential deposition (Hupd) method.Besides, to assess the ECSA, the Hupd method (assuming factor = 210 μC cm -2 ) was applied.The Hupd method was conducted in 0.1 M HClO4 saturated with N2 (100 mV/s).In the potential range of around 0.011 V and 0.406 V after the deduction of the capacitive current, which contain the Hupd adsorption/desorption peak, and the charge (QH) was obtained.The baseline used for integration of the QH belonging to the Hupd adsorption/desorption peak was the current value at around 0.406 V.According to the eq 8, the ECSA obtained via the Hupd method was estimated.
where mPt is the Pt loading of the as-prepared catalyst used in the working electrode.
The average charge of the H adsorption and desorption peak was applied to estimate the ECSA (denoted as H ̅ s) for comparison.At last, the stability of the catalysts was assessed by using an accelerated durability test (ADT) for 10k and 20k cycles at 0.6-1.0V (vs RHE, 100 mV/s) in the O2 atmosphere.
PEMFC measurement.The catalyst powder, Nafion (5wt%, Dupont) and isopropanol were ultrasonically mixed in a mass ratio of 1:X:100 to prepare the catalyst ink (X=12.5In Figure S25a, we constructed two Pt-Ce models that includes both intermetallic compound and solid solution.In Figure S25b, at potential of U=0 V, the potential value of the rate-determining step in intermetallic Pt5Ce compound is about 0.62 eV, which is larger than that in solid solution (0.41 eV), indicating the lower adsorption energies of ORR intermediates.This result reveals that the ordered coordination effect of Ce on Pt and the electron transfer from Ce to Pt generated by intermetallic compounds are beneficial for improving ORR performance.Therefore, the ORR performance of Pt5Ce alloys in intermetallic compound state is better than that in solid solution state.
for Pt-Ce, X=5 for 20 wt% JM Pt/C).The JM Pt/C cathode and Pt-Ce cathode are prepared by spraying the correspond catalyst ink on the gas diffusion layer (GDL, Sigracet 25 BC), the Pt loading is controlled within 0.1 mgPt cm -2 , respectively.All the anode of MEAs were prepared by spraying the JM Pt/C on the GDL (Sigracet 25 BC) with a Pt loading of 0.1 mgPt cm -2 .The MEAs with an active area of 25 cm 2 were fabricated by hot-pressing the prepared anode and cathode onto the opposite sides of a Nafion-117 membrane at 80 o C for 3 min under a pressure of 3 MPa.The MEA performance of the catalysts as the cathode was tested at a fuel cell testing system (Scribner 850e).The backpressure was set at 150 kPa for both the anode and cathode at H2-O2 and H2-Air.For all fuel cell tests, the cell temperature was kept at 80 o C. Fully humidified hydrogen (flow rates: 600 sccm) and air (flow rates: 1500 sccm) at 80 o C were supplied to the anode and cathode, respectively.A chronoamperometric test at a constant voltage of 0.6 V and at 80 o C in H2(anode)-O2(cathode) was adopted to evaluate the stability of Pt-Ce alloy catalyst.We also applied Nafion 211 to PEMFC devices and conducted performance tests on Pt-Ce and commercial Pt/C using the same test process.The results are shown in the Figure S41.Density functional theory (DFT) calculations.All the DFT computations were conducted employing the Vienna ab initio simulation package (VASP).The exchangecorrelation energy was assessed utilizing the generalized gradient approximation (GGA)with the Perdew-Burke-Ernzerhof (PBE) functional, and the projector augmented wave (PAW) method was employed to model electron-ion interactions.The energy cutoff for plane wave expansions was set at 450 eV, with a convergence threshold established at 10 −5 eV in energy and 0.03 eV Å −1 in force, respectively.For slab model calculations, a vacuum space of 15 Å was introduced to prevent interactions between periodic images.Furthermore, a dipolar correction was applied to slab models with symmetrization switched off.K-POINT was set to 2 × 2 × 1.For intermetallic compounds, we constructed the structure of Pt5Ce (1 1 1) crystal facet, comprising 4 layers.For solid solution structure, we employed the Pt (1 1 1) crystal facet, randomly arrange Pt and Ce atoms, also containing 4 layers.Both structures encompassed 16 Ce atoms and 80 Pt atoms, with the lower three layers firmly held during the computational process.To evaluate the ORR performance for slab models, we adopted the typical adsorbate evolution mechanism as proposed by Nørskov et al.2,3

( a )
Improvement factor was determined by MA (sample)/MA(Pt) in the literature.(b) Improvement factor was determined by SA (sample)/SA(Pt) in the literature.(c) Improvement factor was determined by after ADT SA (sample)/SA(Pt) in the literature.

Figure S1 .
Figure S1.The temperatures measured by the real-time infrared thermometer at (a) the

Figure S9 .
Figure S9.Dot line diagram of the relationship between mean particle sizes and power.

Figure S11 .Figure S12 .
Figure S11.The XRD patterns of five Pt-RE alloys.The reference patterns are Pt2Y

Figure S14 .
Figure S14.(a) The XPS spectra of Gd 4d and Pt 4f.(b) The XPS spectra of La 3d and Figure S15.The XPS Pt 4f spectrum of JM Pt/C.

Figure S16 .
Figure S16.(a) CV curves of Pt-Ce and Pt/C after 10k cycles obtained in N2-saturated

Figure S22 .
Figure S22.The TEM image of Pt/Ce after 10k ADT cycles test.

Figure S23 .
Figure S23.The HAADF-STEM elements mapping of Pt/C after 10k ADT cycles test.

Figure S25 .
Figure S25.(a) Schematic illustration of Pt5Ce in the states of intermetallic compound

Figure S26 .
Figure S26.The curves of i−V polarizations and power density of MEA operated in

Figure S27 .Figure S28 .
Figure S27.H2-O2 fuel cell i−V polarization and power density curves of MEA initial

Figure S30 .
Figure S30.(a) The TEM image of Pt-Ce with carbon shell wrapping.(b) The HAADF-

Figure S31 .
Figure S31.The XRD pattern of Pt-Ce, which atom ratio off is 2:1.

Figure S35 .
Figure S35.(a) CV curves of Pt-Ce which in different particle size obtained in N2-

Figure S36 .
Figure S36.The Tafel plots of Pt-Ce which in different particle size.

Figure S37 .
Figure S37.(a) CV curves of Pt-Gd which in different particle size obtained in N2-

Figure S38 .
Figure S38.The Tafel plots of Pt-Gd which in different particle size.

Figure S39 .
Figure S39.The auger spectrum of Ce MNN, further confirming the metallic state of

Figure S40 .
Figure S40.The HAADF-STEM elements mapping at low magnification.It is clear

Table S1 .
Different synthetic conditions of preparing Pt-RE alloys reported recent years.

Table S2 .
The synthesis time and energy consumption compared with peers.

Table S3 .
The electrochemical tests of Pt-RE reported recent years.

Table S4 .
The bulk elemental compositions of samples determined by XPS.

Table S5 .
The PEMFC performance of Pt-based alloy catalysts reported recent years.

Table S6 .
The amounts of precursors and the molar ratio of Pt particles and rare-earth metals.

Table S7 .
The bulk elemental compositions of samples determined by ICP and XPS.