Pt-based stable catalyst with ultrafine transition-metal oxide content

Abstract

A nanocatalyst includes platinum nanoparticles and base transition metal oxide clusters, wherein each base transition metal oxide cluster is enclosed by a platinum shell.

Description

Atty. Dkt. No.: 102352-1142 PT-BASED STABLE CATALYST WITH ULTRAFINE TRANSITION- METAL OXIDE CONTENT CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application is an International Patent Application that claims priority to and the benefit of U.S. Provisional Application No.63 / 580,271, filed on September 1, 2023, the contents of which are incorporated herein by reference in their entirety. STATEMENT OF GOVERNMENT SPONSORED RESEARCH

[0002] This invention was made with government support under Grant Number N0001418-1-2155, awarded by the U.S. Navy, Office of Naval Research. The government has certain rights in the invention. TECHNICAL FIELD

[0003] The present embodiments relate generally to energy storage and more particularly to a platinum (Pt)-based stable catalyst with transition-metal oxide content. SUMMARY

[0004] In one aspect, a nanocatalyst is provided that includes platinum nanoparticles and base transition metal oxide clusters, wherein each base transition metal oxide cluster is enclosed by a platinum shell. In some embodiments, the base transition metal is selected from the group consisting of Co, Fe, Ni, Mn, Cu, and Zn. In some embodiments, the base transition metal is selected from the group consisting of Co, Fe, and Ni. In some embodiments, the base transition metal is Co. In some embodiments, the platinum nanoparticles are on a carbon support. In some embodiments, the carbon support is carbon black, carbon nanotube, carbon fiber, or graphene. In some embodiments, the carbon support is carbon black. In some embodiments, the nanocatalyst has an average particle size of about 1 nm to about 5 nm. In some embodiments, the nanocatalyst has an average particle size of about 1.5 nm. In some embodiments, the base 1 4856-0223-4333.3Atty. Dkt. No.: 102352-1142 transition metal oxide clusters are sparsely embedded in the platinum nanoparticles. In some embodiments, the nanocatalyst has a Co:Pt atomic ratio of about 4.7:95.3 to about 4.9:95.1.

[0005] In another aspect, a fuel cell is provided that includes any of the nanocatalysts described herein. In some embodiments, the fuel cell is a proton exchange membrane fuel cell.

[0006] In a further aspect, a method is provided preparing the nanocatalyst described herein. The method includes mixing platinum-based catalyst precursor with base transition metal catalyst precursor on a carbon support in an organic solvent; evaporating the solvent to form a dry powder mixture; annealing the dry powder mixture to form an initial annealed mixture; washing the initial annealed mixture under acidic conditions to form an acid-treated mixture; and annealing the acid-treated mixture to provide the nanocatalyst.

[0007] In some embodiments, the platinum-based catalyst precursor is platinum(II) acetylacetonate. In some embodiments, the base transition metal catalyst precursor is cobalt(II) acetylacetonate. In some embodiments, the solvent is acetone. In some embodiments, the dry powder mixture is annealed at a temperature of at least 200⁰C under inert atmosphere. In some embodiments, the dry powder mixture is annealed at a temperature of 250⁰C under argon atmosphere. In some embodiments, the dry powder mixture is annealed for at least 6 hours. In some embodiments, the acidic conditions comprise heating with aqueous sulfuric acid. In some embodiments, the acidic conditions comprise heating at 80⁰C with aqueous sulfuric acid for at least 10 hours. In some embodiments, the acid-treated mixture is annealed at a temperature of at least 150⁰C. In some embodiments, the acid-treated mixture is annealed at a temperature of 180⁰C under hydrogen / argon atmosphere. In some embodiments, the acid-treated mixture is annealed for at least 60 minutes.

[0008] Some embodiments relate to a method to synthesize uniformly distributed, Pt- based, nanoparticles having an ultralow base transition metal content. The method includes the impregnation and adsorption of hydrophobic organic metal precursors on a hydrophobic carbon support. Then, the mixture of carbon and metal precursors is annealed in conditions that control the ultrafine nanoparticle size around 1-5 nm while ensuring sufficient utilization of the platinum precursor. After annealing, the obtained ultrafine catalyst is processed with dilute acid. AfterAtty. Dkt. No.: 102352-1142 acid treatment, a further annealing process is conducted to produce the final product. The method may be applied to binary or ternary Pt-M-N catalysts while M is a base transition metal (M, N = Mn, Fe, Co, Ni, Cu, Zn, etc.). The method can be applied to different types of carbon supports (carbon black, carbon nanotube, carbon fiber, graphene, etc.).

[0009] The performance of the developed catalysts may be evaluated using a membrane electrode assembly (MEA) test. The developed catalysts showed significant improvement in fuel cell activity and durability compared to typical commercial Pt-based catalysts, including commercial Pt / C and / or commercial Pt alloys (c-PtCo / C). Also, the developed catalysts showed excellent fuel cell performances surpassing all catalyst targets set by the U.S. Department of Energy (DOE). Notably, the durability of the developed PtCo / C represents the state-of-art stability in terms of rated power retention and voltage loss at various current ranges. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] These and other aspects and features of the present embodiments will become apparent to those ordinarily skilled in the art upon review of the following description of specific embodiments in conjunction with the accompanying figures, wherein:

[0011] FIGs.1A, 1 B, and 1C illustrate characterization of obtained and compared catalysts: (FIG.1A) is an X-ray powder diffraction (XRD) spectra of commercial Pt / C (c-Pt / C), commercial PtCo / C (c-PtCo / C), CoOx@Pt / C before acid treatment (AW), and final CoOx@Pt / C catalyst; FIG.1B is the mean diameter of c-PtCo / C and CoOx@Pt / C with standard deviation (SD) based on TEM images; and FIG.1 C shows the atomic metal ratio of c-PtCo / C and CoOx@Pt / C catalysts based on inductively coupled plasma atomic emission spectroscopy (ICP- AES) and X-ray photoelectron spectroscopy (XPS), according to the examples.

[0012] FIGs.2A, 2B, 2C, and 2D illustrate X-ray absorption spectroscopy (XAS) characterization of developed PtCo catalysts, including XAS results of CoOx@Pt / C before acid treatment (AW), and final CoOx@Pt / C catalyst (after AW+AN): FIG.2A is for Pt L3-edge X- ray absorption near edge structure (XANES); FIG.2B is for Co K-edge XANES; FIG.2C is for Pt Extended X-ray absorption fine structure (EXAFS) in R-space; and FIG.2D is for Co EXAFS in R-space, according to the examples.Atty. Dkt. No.: 102352-1142

[0013] FIGs.3A, 3B, 3C, and 3D illustrate membrane electrode assembly (MEA) tests of catalysts (Pt / C, c-PtCo / C, and CoOx@Pt / C): FIG.3A is the mass activity in H2 / O2 test; FIG.3B is the rated power density from H2 / Air test at 94°C, 250 kPaabs; FIG.3C is the voltage loss after accelerated-durability test (ADT) at different current densities at 94°C, 250 kPaabs(for each set of three bars: (left) CoOx@Pt / C, (middle) c-PtCo / C, (right) c-Pt / C); and FIG.3D is the rated power loss after ADT, according to the examples.

[0014] FIG.4 illustrates size distribution and average size of the measured particle diameter, where the size includes the mean ± standard deviation (SD), according to the examples.

[0015] FIGs.5A and 5B illustrate schematic illustration of fuel cell architecture and the catalyst design: FIG.5A includes components in a single fuel cell, where the proton exchange membrane and the catalyst layers form the MEA; and FIG.5B includes designs and advantages of the fine CoOx@Pt core–shell nanocatalyst supported on carbon, according to various embodiments.

[0016] FIGs.6A-6F illustrate MEA performance of CoOx@Pt / C, c-PtCo / C and c-Pt / C catalysts: FIG.6A is a comparison of MA obtained in H2 / O2 tests at the beginning of life (BOL; before accelerated stress test (AST)) and end of life (EOL; after AST); FIGs.6B and 6C are comparison of the rated power density (6B) and Pt utilization (6C) obtained in H2 / air tests at the BOL and EOL (DOE targets are indicated by the horizontal dashed (BOL) and dotted (EOL) lines); FIGs.6D and 6E are polarization plots (left axes) and power density plots (right axes) of c-PtCo / C (6D) and CoOx@Pt / C (6E) obtained in H2 / air tests at the BOL and EOL, highlighting the excellent power performance and stability of the MEA with CoOx@Pt / C; and FIG.6F is the voltage loss at current density values of 0.8, 1.0 and 1.25 mA cm−2(for each set of three bars: (left) CoOx@Pt / C, (middle) c-PtCo / C, (right) c-Pt / C), according to the examples. In FIGs.6A, 6B, 6C, and 6F, the data are shown as the mean ± SD, where the mean values are derived from n independent measurements (with data points shown as circles; CoOx@Pt / C: n = 5 and 4 for BOL and EOL, respectively; c-PtCo / C: n = 5 and 2 for BOL and EOL, respectively; c-Pt / C: n = 2 for BOL and EOL).Atty. Dkt. No.: 102352-1142

[0017] FIGs.7A-7F include illustrations of MEA performance comparison of CoOx@Pt / C and the reference alloy samples with higher Co content (Pt79Co21 / C): FIG.7A is a comparison of mass activity (MA) obtained in H2 / O2 flow at the beginning of life (BOL; before ADT) and end of life (EOL; after ADT). Comparison of (FIG.7B) rated power density obtained in H2 / air flow at BOL and EOL and its loss during ADT. Polarization curves of Pt79Co21 / C and CoOx@Pt / C obtained in H2 / air tests at (FIG.7C) BOL (solid lines) and (FIG.7D) EOL (dot lines). FIG.7E is the voltage loss at different current densities recorded in H2 / air flow, where the dotted line indicates DOE target. FIG.7F is the change of Co atomic composition at BOL and EOL. The error bars in (FIG.7A) and (FIG.7B) represent the standard deviation with the center representing the mean value, derived from n times of independent measurements (data points shown as circles, n=5 and 4 for CoOx@Pt / C BOL and EOL, n=3 for c-Pt29Co21 / C BOL and EOL). Data are presented as mean values ± SD.

[0018] FIGs.8A-8F include illustrations of MEA performance of CoOx@Pt / C, NiOx@Pt / C, FeOx@Pt / C: FIG.8A is a comparison of mass activity (MA) obtained in H2 / O2tests at the beginning of life (BOL; before ADT) and end of life (EOL; after ADT); FIGs.8B and *C are comparisons of rated power density (8B) and Pt utilization obtained in H2 / air tests at BOL and EOL (8C), where DOE targets are represented by horizontal dash (BOL) and dot (EOL) lines; FIGs.8D and 8E are polarization plots (left axis) and power density plots (right axis) of FeOx@Pt / C (8D) and NiOx@Pt / C (8E) obtained in H2 / air tests at BOL and EOL; and FIG.8F is the voltage loss at current density of 0.8, 1.0 and 1.25 mA / cm2, where for each set of three bars: (left) CoOx@Pt / C, (middle) NiOx@Pt / C, (right) FeOx@Pt / C, according to the examples. The error bars in (FIG.8F) represent the standard deviation with the center representing the mean value, derived from 5 times of independent measurements (data points shown 10 as circles). Data are presented as mean values ± SD.

[0019] FIGs.9A-9D include illustrations comparing key performance metrics for CoOx@Pt / C catalysts at active cell areas of 5 cm2, 25 cm2and performance after compensating the additional voltage loss from wiring resistance. FIG.9A is a comparison of MA obtained in H2 / O2tests under high or low flow rate (FR) at the beginning of life (BOL; before AST) and end of life (EOL; after AST). A comparison of rated power density obtained in H2 / air tests at BOLAtty. Dkt. No.: 102352-1142 and EOL is shown in FIG.9B, with DOE targets represented by horizontal dash (BOL) and dot (EOL) lines. FIG.9C is the polarization plots (left axis) and power density plots (right axis) obtained in H2 / air tests at BOL and EOL, highlighting the extraordinary power performance and stability of the 25 cm2MEA with CoOx@Pt / C. FIG.9D is the polarization plots of 5 cm2and 25 cm2almost overlapped after compensating voltage loss from wiring resistance at each current density. The error bars in FIGs.9A and 9B represent the standard deviation with the center representing the mean value, derived from n times of independent measurements (data points shown as circles, n=5 and 4 for 5cm2cell at BOL and EOL). Data are presented as mean values ± SD.

[0020] FIG.10 illustrates change in Co atomic composition for BOL and EOL in the MEA, according to the examples.

[0021] FIG.11 illustrates a comparison of CO stripping experiments in the MEAs at 80 °C, showing the change in the oxidation peaks at BOL (solid lines) and EOL (dotted lines) for the commercial catalysts and CoOx@Pt / C, according to the examples.

[0022] FIG.12 illustrates a comparison of key fuel cell metrics with the state-of-the-art catalysts. Notably, the MEA with CoOx@Pt / C catalyst exceeds all DOE performance metrics and shows the highest EOL performances, highlighting its extraordinary activity and durability. Data are presented as mean values ± SD. *: value extracted from performance plots, whereaindicates the total loading of 0.10 mg / cm2is for H2-O2 test and 0.04 mg / cm2is for H2-Air test, andbis the MA retention value is recorded from a different MEA with a thicker membrane. DETAILED DESCRIPTION

[0023] The present embodiments will now be described in detail with reference to the drawings, which are provided as illustrative examples of the embodiments so as to enable those skilled in the art to practice the embodiments and alternatives apparent to those skilled in the art. Notably, the figures and examples below are not meant to limit the scope of the present embodiments to a single embodiment, but other embodiments are possible by way of interchange of some or all of the described or illustrated elements. Moreover, where certain elements of the present embodiments can be partially or fully implemented using known components, only thoseAtty. Dkt. No.: 102352-1142 portions of such known components that are necessary for an understanding of the present embodiments will be described, and detailed descriptions of other portions of such known components will be omitted so as not to obscure the present embodiments. In the present specification, an embodiment showing a singular component should not be considered limiting; rather, the present disclosure is intended to encompass other embodiments including a plurality of the same component, and vice-versa, unless explicitly stated otherwise herein. Moreover, applicants do not intend for any term in the specification or claims to be ascribed an uncommon or special meaning unless explicitly set forth as such. Further, the present embodiments encompass present and future known equivalents to the known components referred to herein by way of illustration.

[0024] Proton exchange membrane fuel cells (PEMFCs) that directly convert the chemical energy in hydrogen (H2) to electricity represent an attractive zero-emission power generation technology. The implementation of PEMFCs requires costly platinum group metal (PGM) catalysts to facilitate the sluggish cathodic oxygen reduction reaction (ORR), which accounts for up to 50% of the total system cost and represents a critical sustainability roadblock that prevents the widespread adoption of PEMFCs. To ensure the competitiveness of PEMFCs in the transportation sector, the U.S. Department of Energy (DOE) has set a series of technical targets for PEMFC catalysts (see Fuel Cell Technologies Office Multi-Year Research, Development, and Demonstration Plan (U.S. Department of Energy, 2017)). Generally, fuel cell performance should be evaluated under both low current density (LCD) and high current density (HCD) regimes for different operation scenarios. The kinetics-dominated LCD region strongly affects the fuel efficiency in vehicle applications, as the cells operate predominantly (90% of the time) in a low-power mode during cruising, whereas the performance under the HCD regime determines the maximum power output available for engine startup or aggressive acceleration, and generally dictates the performance retention and the PEMFC lifetime. Performance retention is especially critical in the growing application of heavy-duty vehicles, the use of which requires travelling for six times further than with typical light-duty applications and thus demands a much longer lifetime.Atty. Dkt. No.: 102352-1142

[0025] However, the key performance metrics of PEMFEC catalysts (for example, activity versus lifetime durability) often display an intrinsic trade-off relationship and are difficult to achieve simultaneously. These challenges have motivated immense efforts to improving the catalyst activity and durability. In particular, various nanostructured (nanowires, nanoframes, nanocages, core–shell entities, among others) Pt-based alloy catalysts (Pt–M, M = Co, Ni and so on) with tailored strain and ligand effects have been explored to modulate the oxygen binding energy and the ORR catalytic activity (see, e.g., (1) Liu et al., J. Am. Chem. Soc., (2020), 142:17812–17827; and (2). Zhao et al., Adv. Mater., (2019), 31:1808115). However, the application of these elaborate designs in practical fuel cells is commonly challenged by their limited stability and / or difficulties in mass production. For example, the Ostwald ripening of typical nanostructured catalysts can lead to rapid loss of the electrochemically active surface area (ECSA) and overall mass activity (MA). The Pt–M alloy catalysts often suffer from additional durability penalties associated with the dissolution of non-precious metal cations that can occupy the proton sites in the ionomers and lead to cation poisoning effects that substantially impede mass transport and compromise the practical fuel cell performance. Although the formation of PtCo / Pt core–shell catalysts can partly mitigate this problem, such alloy catalysts are of a relatively large size (>4 nm) that limits their ECSA (~40 m2gPt−1), and, accordingly, their mass- specific performance. To this end, it has been shown that structurally ordered intermetallic Pt–M alloys can improve the stability and deliver a satisfactory MA retention (>60%) in the LCD region. However, metal leaching and the associated cation poisoning effects can still compromise the mass transport, which remains a persistent challenge, particularly in the HCD region where mass transport dominates the device performance, as highlighted by an unsatisfactory rated power loss (>10%).

[0026] Considerable efforts have been devoted to mitigating the metal leaching and cation poisoning effects in Pt–M alloy catalysts. For example, it has been shown that oxide- supported Pt catalysts can exhibit improved activity and stability due to the strong metal–oxide interaction. Although the role of the oxide support in modifying the electronic structure and stabilizing the Pt has been well corroborated, the typically poor conductivity and limited stability of the oxide support itself under acidic conditions prevent its practical implementation in PEMFCs.Atty. Dkt. No.: 102352-1142

[0027] To address these issues, provided herein is a nanocatalyst containing Pt nanoparticles with sparsely embedded cobalt oxide (CoOx) clusters, in which each CoOx cluster is fully protected by a Pt shell, to form a core-shell nanostructure as a durable ORR electrocatalyst in PEMFCs. The embedded-oxide design retains the full benefit of the strong Pt– oxide interaction to modify the electronic structure of Pt, simultaneously suppressing dissolution of the Pt atoms and enhancing the catalytic activity. Meanwhile, with the unstable and less conductive oxide species fully enclosed by the metallic Pt shells, the design also mitigates the dissolution of CoOxand consequent cation poisoning effects. Moreover, it bypasses the charge- transport limitations of a typical oxide support. These combined merits deliver a high PEMFC performance and durability. With the embedded CoOx retarding Pt leaching and Oswald ripening, the nanocatalysts exhibit an extraordinary stability at a fine size (~1.5 nm), and they display a compressive surface strain to enable a superior performance compared with commercial Pt (c-Pt / C) or PtCo alloy (c-PtCo / C) catalysts.

[0028] Accordingly, in one aspect, provided herein is a nanocatalyst comprising, consisting essentially of, or consisting of platinum nanoparticles and base transition metal oxide clusters, wherein each base transition metal oxide cluster is enclosed by a platinum shell.

[0029] In some embodiments, the base transition metal may be Co, Fe, Ni, Mn, Cu, Zn, or a mixture of any two or more thereof. In some embodiments, the base transition metal is Co. In some embodiments, the base transition metal is Fe. In some embodiments, the base transition metal is Ni. In some embodiments, the base transition metal is Mn. In some embodiments, the base transition metal is Cu. In some embodiments, the base transition metal is Zn. In some embodiments, the base transition metal is Co, Fe, Ni, or a mixture of any two or more thereof.

[0030] In some embodiments, the platinum nanoparticles are on a carbon support. Non- limiting examples of the carbon supports include carbon black (e.g., Ketjenblack EC-300J), carbon nanotube, carbon fiber, or graphene. In some embodiments, the carbon support is carbon black.

[0031] In another aspect, provided is a nanocatalyst comprising, consisting essentially of, or consisting of platinum nanoparticles with embedded cobalt oxide clusters, wherein each cobaltAtty. Dkt. No.: 102352-1142 oxide cluster is enclosed by a platinum shell. Such platinum-containing nanocatalysts are referred herein as CoOx@Pt nanocatalysts, or more particularly CoOx@Pt / C nanocatalysts when the nanocatalysts are on a carbon support.

[0032] In some embodiments, the nanocatalyst has an average particle size of about 1 nm to about 5 nm. This includes an average particle size of about 1 nm, about 1.5 nm, about 2 nm, about 2.5 nm, about 3 nm, about 3.5 nm, about 4 nm, about 4.5nm, and about 5 nm, including any value therebetween. In some embodiments, the nanocatalyst has an average particle size of about 1.5 nm.

[0033] In some embodiments, the nanocatalyst has an average particle size of 1 nm to 5 nm. This includes an average particle size of 1 nm, 1.5 nm, 2 nm, 2.5 nm, 3 nm, 3.5 nm, 4 nm, 4.5nm, and 5 nm, including any value therebetween. In some embodiments, the nanocatalyst has an average particle size of 1.5 nm.

[0034] In some embodiments, the base transition metal oxide clusters are sparsely embedded in the platinum nanoparticles. In some embodiments, the cobalt oxide clusters are sparsely embedded in the platinum nanoparticles.

[0035] In some embodiments, the nanocatalyst has a Co:Pt atomic ratio of about 4.7:95.3 to about 4.9:95.1. In some embodiments, the nanocatalyst has a Co:Pt atomic ratio of 4.7:95.3 to 4.9:95.1. In some embodiments, the nanocatalyst has a Co:Pt atomic ratio of 4.7:95.3. In some embodiments, the nanocatalyst has a Co:Pt atomic ratio of 4.9:95.1.

[0036] In another aspect, a fuel cell is provided that includes any of the above nanocatalysts. In some embodiments, the fuel cell is a proton exchange membrane fuel cell.

[0037] In another aspect, a method of preparing the nanocatalysts described herein is provided. The methods include mixing platinum-based catalyst precursor with base transition metal catalyst precursor on a carbon support in an organic solvent; evaporating the solvent to form a dry powder mixture; annealing the dry powder mixture to form an initial annealed mixture; washing the initial annealed mixture under acidic conditions to form an acid-treated mixture; and annealing the acid-treated mixture to provide the nanocatalyst.Atty. Dkt. No.: 102352-1142

[0038] In some embodiments, the platinum-based catalyst precursor is platinum(II) acetylacetonate. In some embodiments, the base transition metal catalyst precursor is cobalt(II) acetylacetonate.

[0039] In some embodiments, the solvent is a volatile organic solvent. In some embodiments, the solvent is acetone, methanol, ethanol, or propanol, or a combination of two or more thereof. In some embodiments, the solvent is acetone.

[0040] In some embodiments, the dry powder mixture is annealed at a temperature of at least 200⁰C under inert atmosphere (e.g., argon atmosphere). In some embodiments, the dry powder mixture is annealed at a temperature of 200⁰C, 210⁰C, 220⁰C, 230⁰C, 240⁰C, 250⁰C, 260⁰C, 270⁰C, 280⁰C, 290⁰C, or 300⁰C, or more, or any value therebetween, under inert atmosphere. In some embodiments, the dry powder mixture is annealed at a temperature of 250⁰C under an argon atmosphere. In some embodiments, the dry powder mixture is annealed for at least 6 hours. This includes 6, 7, 8, 9, 10, 11, or 12 hours, or more, including any value therebetween. In some embodiments, the dry powder mixture is annealed for 8 hours.

[0041] In some embodiments, the acidic conditions include heating with aqueous sulfuric acid. In some embodiments, the acidic conditions include heating at 80⁰C with aqueous sulfuric acid for at least 10 hours. This includes 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, or 24 hours, or more, including any value therebetween. In some embodiments, the acidic conditions include heating at 80-90⁰C with aqueous sulfuric acid for 8-24 hours. In some embodiments, the acidic conditions include heating at 80⁰C with aqueous sulfuric acid for 12-16 hours. In some embodiments, the aqueous sulfuric acid has a concentration of 0.05 M to 0.5 M. This includes 0.05 M, 0.1 M, 0.15 M, 0.20 M, 0.25 M, 0.3 M, 0.35 M, 0.40 M, 0.45 M, or 0.5 M, or any value therebetween.

[0042] In some embodiments, the acid-treated mixture is annealed at a temperature of at least 150⁰C. This includes 150⁰C, 155⁰C, 160⁰C, 165⁰C, 170⁰C, 175⁰C, 180⁰C, 185⁰C, 190⁰C, 195⁰C, or 200⁰C, or more, including any value therebetween. In some embodiments, the acid- treated mixture is annealed at a temperature of 150-200⁰C under hydrogen / argon atmosphere. In some embodiments, the acid-treated mixture is annealed at a temperature of 180⁰C underAtty. Dkt. No.: 102352-1142 hydrogen / argon atmosphere. In some embodiments, the acid-treated mixture is annealed for at least 60 minutes. This includes 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, or 150 minutes, or more, including any value therebetween. In some embodiments, the acid-treated mixture is annealed for 90 minutes. In some embodiments, the acid-treated mixture is annealed for 90-120 minutes.

[0043] In another aspect, provided herein is a strategy to synthesize uniformly distributed Pt-based nanoparticles with ultralow base transition metal content, comprising: the impregnation and adsorption of hydrophobic organic metal precursors on a hydrophobic carbon support; and annealing the mixture of carbon and metal precursors in conditions that control the ultrafine nanoparticle size around 1-5 nm while ensuring sufficient utilization of the platinum precursor. In some embodiments, the optimized annealing conditions include the atmosphere, heating rate, temperature, and duration. In some embodiments, the strategy further comprises, after annealing, processing the obtained ultrafine catalysts by diluted acid. In some embodiments, the strategy further comprises, after acid treatment, conducting a further annealing process to produce the final product. In some embodiments, the strategy is applied to binary or ternary Pt- M-N catalysts while M is a base transition metal (M, N = Mn, Fe, Co, Ni, Cu, Zn, etc.). In some embodiments, the strategy is a type of the carbon support is one of carbon black, carbon nanotube, carbon fiber, graphene, etc. EXAMPLES

[0044] Example 1: ultrafine Pt catalyst with Co oxide cluster core (CoOx@Pt / C). A typical synthesis includes the mixing of 150-180 mg carbon black, 150-170 mg platinum (II) acetylacetonate [Pt(acac)2], 100-160 mg cobalt (II) acetylacetonate [Co(acac)2], and 6-10 mL acetone in a 35ml glass vial. The mixture then undergoes ultrasonication for 10-30 min, followed by the evaporation of the acetone solvent. The evaporation process can be done at ambient temperature or pressure lower or equal to 1 atm. After the complete evaporation of the solvent, the remaining powder is collected and finely ground.

[0045] A fine powder of metal precursor adsorbed carbon black (0.5-1.5g) obtained from the above process is then loaded in a quartz boat and annealed. The annealing conditions areAtty. Dkt. No.: 102352-1142 optimized to ensure ultrafine particle size (e.g. approximately 2 nm) while maximizing utilization of the platinum precursor. First, the oxygen is removed from the annealing system by repeated vacuum pumping and Ar purging. Then, a H2 / Ar mixture gas (H2 volume ratio is 0.2- 2.0%) is purged when the sample is heated from room temperature to 200-300°C and maintained at the final temperature for 6 to 12 hr. After that, the annealed sample is cooled down to room temperature slowly under the purging of Ar.

[0046] After the annealing step, an acid wash process (AW) is conducted. A typical treatment involves the dispersion of 120-200 mg obtained catalyst in 15-20 mL of N2saturated 0.1-0.5 M H2SO4 in a 35 ml capped vial. The vial was kept at 80-90 ºC for 8-24 hr. This acid treatment process can also be conducted in a larger container with multiple additions of the catalyst powder and diluted acid mentioned above. After acid wash, the catalyst was rinsed with ultrapure water till the pH was neutral. Then the catalyst is collected via centrifugation and dried in a vacuum oven. The dry catalyst is then annealed (AN) under the following conditions: 0.5- 1.5g catalyst powder is loaded in a quartz boat, and the oxygen is removed from the annealing system by repeated vacuum pumping and Ar purging; After the removal of oxygen, a H2 / Ar mixture gas (H2 volume ratio is 0.2-1.0%) is purged when the sample is slowly heated from room temperature to 150-200°C and maintained at the final temperature for 1.5-2 hr. The annealed sample is cooled down to room temperature slowly under the purging of Ar.

[0047] Characterizations and fuel cell performance evaluations.

[0048] The developed ultrafine PtCo supported on carbon black with ultralow Co content (CoOx@Pt / C), commercial Pt / C (c-Pt / C), and commercial PtCo / C (c-PtCo / C) were characterized and evaluated in a 5 cm2fuel cell stack. X-ray powder diffraction (XRD) spectra revealed that the nanoparticles in all catalysts are face center cubic (fcc) packing (Figure 1a). After acid treatment, the (111) peak shifted towards lower degree while the width increased, indicating the leaching of Co content and crystalline size growth (Figure 1a). The size distributions of c- PtCo / C and CoOx@Pt / C were analyzed by transmission electron microscopy (TEM), which revealed the size of nanoparticles to be 4.7 ± 1.4 nm and 1.5 ± 0.2 nm, respectively (Figure 1b). The compositions of the catalysts were studied by inductively coupled plasma atomic emission spectroscopy (ICP-AES) and X-ray photoelectron spectroscopy (XPS) (Figure 1c). From theAtty. Dkt. No.: 102352-1142 ICP-AES result, the developed CoOx@Pt / C exhibited an ultralow Co content of 5 wt.% compared to that of the c-PtCo / C. In addition, the compositions extracted from XPS showed a higher Pt ratio than those based on ICP-AES for PtCo / C, indicating a Pt shell structure with Co embedded inside the CoOx@Pt / C nanoparticle (Figure 1c).

[0049] The detailed structures of the developed CoOx@Pt / C catalysts were further demonstrated by X-ray absorption spectroscopy (XAS). According to the X-ray absorption near edge structure (XANES) results, both Pt and Co were in the oxidation state before acid wash (Figure 2a, b). After acid wash and annealing, Pt was reduced to metallic form while the Co oxidation state remained. From Extended X-ray absorption fine structure (EXAFS) results, Pt and Co were atomically mixed in a “single-atom-oxide composite” structure before acid wash (Figure 2c, d). After acid wash and annealing, most Co oxide was removed and Pt-Pt interaction emerged, while the residue Co was still mainly in the single atom Co2+O6 structure.

[0050] The fuel cell performances of the catalysts are evaluated in a 5 cm2membrane electrode assembly (MEA) test. The durability of the fuel cell catalysts was evaluated by applying an accelerated-durability-test (ADT) of 30000 square-wave cycles strictly following the U.S. Department of Energy (DOE) protocol. In H2 / O2 test, the developed CoOx@Pt / C showed significant higher oxygen reduction reaction (ORR) mass activity (MA) of 1.10 A / mgPt and MA retention of 88.2% compare to commercial c-Pt / C (MA of 0.34 A / mgPt, MA retention of 22.5%) and c-PtCo / C (MA of 0.57 A / mgPt, MA retention of 36.8%) (Figure 3a). The performance surpassed the MA targets set by US DOE at both beginning-of-life (BOL) and end-of-life (EOL). The H2 / Air tests were conducted for power performance evaluation, which showed the developed CoOx@Pt / C exhibited the highest rated power performance of 1.04 W / cm2compared to commercial c-Pt / C (0.88 W / cm2) and c-PtCo / C (0.92 W / cm2), and exceeded the rated power target of 1 W / cm2set by US DOE (Figure 3b). Notably, the developed PtCo / C demonstrated the lowest voltage loss (Figure 3c) and rated power loss of only 7.5% (Figure 3d) after the aggressive ADT, exhibiting tremendous advantages over both commercial c-Pt / C (63.6% loss in rated power) and c-PtCo / C (33.7% loss in rated power). Thus, the developed ultrafine CoOx@Pt / C with embedded Co oxide cluster core with promising fuel cell performance andAtty. Dkt. No.: 102352-1142 durability holds great potential to resolve the major challenges of cost and durability for practical fuel cell applications.

[0051] Example 2. Synthesis and structural characterization.

[0052] Chemicals and materials. Pt(acac)2and Aquivion D83-06A ionomer dispersion were purchased from Sigma Aldrich. Co(acac)2, 10 wt% Pt / C, and 40 wt% Pt / C were purchased from Alfa Aesar. Commercial PtCo / C catalyst (c-PtCo / C) catalyst was purchased from Premetek. Acetone and isopropanol were purchased from Fisher Scientific. Freudenberg H14C7 gas diffusion layers and polytetrafluoroethylene (PTFE) gaskets were purchased from Fuel Cell Store. The carbon support (Ketjenblack EC-300J) was purchased from Fitz Chem. The water for the experiments (18.2 MΩ cm) was obtained using a Millipore Ultrapure water system.

[0053] Catalyst preparation. A carbon support (Ketjen Black EC-300J) was activated in a H2 / Ar mixture before use as noted Zhao et al. (Matter (2020), 3:1774–1790). In a typical synthesis, the carbon support (150–180 mg), platinum(II) acetylacetonate (Pt(acac)2) (150–170 mg), cobalt(II) acetylacetonate (Co(acac)2) (150–160 mg), and acetone (15 ml) were mixed under ultrasonication for 20 min. After evaporation of the acetone, the resultant metal-precursor- adsorbed carbon support was collected as a dried powder mixture and then annealed in a quartz tube with Ar gas flow, whereby the temperature was heated from room temperature (20 °C) to 250 °C and maintained at 250 °C for 8 h. The sample was collected after it had cooled to room temperature. Then the obtained powder was washed in N2-saturated 0.05 M H2SO4 solution at 80 °C for 12–16 h. After acid washing, the catalyst was rinsed with ultrapure water until the pH was neutral, after which it was collected via centrifugation. After thoroughly drying the catalyst under vacuum, it was further annealed in a H2 / Ar mixture at 180 °C for 1.5 h. After cooling to room temperature, the sample collected was the CoOx@Pt / C catalyst, which was ready for testing.

[0054] Structure and composition characterization. TEM images were obtained using a JEOL JEM 2800 transmission electron microscope operated at 200 kV. Atomic-resolution high- angle annular dark-field images, as well as energy-dispersive X-ray spectroscopy (EDS) maps, were obtained using a JEOL Grand ARM300CF scanning / transmission electron microscopeAtty. Dkt. No.: 102352-1142 operated at 300 kV. The TEM / STEM sample grids were prepared by dispersing the sample in a mixture of water and ethanol, then drop-casting the dispersion on to lacey carbon film-coated copper grids (for high-resolution STEM and EDS analyses) using a pipette and drying under ambient conditions. For size analysis of the catalysts from the TEM / STEM images, measurements were taken from multiple regions and averaged from more than 400 particles in each sample. The membrane electrode assembly (MEA) cross-section samples at a thickness of ~150 nm were prepared using an ultramicrotome, and they were analyzed using the JEOL JEM 2800 transmission electron microscope operated at 200 kV. The fraction of Pt mass loss in the Pt band was calculated by integrating the Pt L edge from the Pt band region divided by the total Pt signal from the entire region (membrane plus cathode catalyst layer), and the results were averaged from around ten different regions in each sample; the Pt band region was carefully selected by inspecting the location before and after the acquisition. The concentration of the catalysts was determined using inductively coupled plasma atomic emission spectroscopy (ICP- AES) (Shimadzu ICPE-9000) and EDS coupled via the JEOL Grand ARM300CF instrument— note that EDS may not provide an accurate absolute composition due to background interference, although it can reflect the relative composition change. Powder X-ray diffraction patterns were collected using a Panalytical X’Pert Pro X-ray powder diffractometer with Cu-Kα radiation. X- ray photoelectron spectroscopy (XPS) tests were carried out using a Kratos AXIS Ultra DLD spectrometer. The XPS depth profiling was conducted using the instrument’s Ar+source operated at 4 kV, 50 μA, and rastered over a 3 × 3 mm area for 4 min. The atomic composition was determined on the basis of the photoelectron peak areas and the relative sensitivity factors provided in the CasaXPS software (see, e.g., Fairley et al., Appl. Surf. Sci., (2021) 5:100112).

[0055] Scanning transmission electron microscopy (STEM) and high-resolution TEM images showed that the resulting CoOx@Pt / C product was uniformly dispersed on the carbon support with an average particle diameter of 1.5 ± 0.2 nm (FIG.4). Powder X-ray diffraction studies showed a similar diffraction pattern to those of the c-Pt / C and c-PtCo / C catalysts, with a face-centered cubic structure. A notably broader peak in the X-ray diffraction pattern was observed, on the basis of which an average particle size of 1.5 nm is estimated, consistent with microscopic size analysis.Atty. Dkt. No.: 102352-1142

[0056] Composition analysis via ICP-AES shows that CoOx@Pt / C has a Co:Pt atomic ratio of 4.7:95.3, whereas analysis with EDS shows a consistent Co:Pt ratio of 4.9:95.1. Electron energy-loss spectroscopy (EELS) mapping and the corresponding spectra show that the Co content is sparsely dispersed inside the nanoparticle. Interestingly, XPS analysis of CoOx@Pt / C shows no Co 2p peak, indicating that the majority of Co is embedded inside the catalyst. Indeed, the Co 2p peak (consistent with Co2+) emerged after removing the surface atomic layer via in situ Ar+etching.

[0057] XAS data collection and analysis. Electrode inks for the X-ray absorption spectroscopy (XAS) electrodes were composed of 60% ethanol, 40% water, 5 wt% Nafion solution, and the catalyst powder. The ink was spray-coated on to a carbon paper with a total catalyst loading about 12.5 mg cm−2(with a Co loading of 0.05 mg cm−2and a Pt loading of 2.7 mg cm−2). The XAS experiments were conducted at room temperature using an in situ spectroelectrochemical half-cell circulated with a continuously O2-purged 0.1 M HClO4 aqueous solution (see, e.g., Arruda et al., J. Phys. Chem. C, (2008) 112:18087–18097), using beamlines ISS 6-BM and 8-ID at National Synchrotron Light Source II (Brookhaven National Laboratory, New York). Before being transferred into the cell, the electrodes were conditioned in 0.1 M HclO4under vacuum for 40 min without applying the potential to remove surface contaminants. Potentiostatic control was maintained using an Autolab PGSTAT30 potentiostat (Metrohm, formerly Brinkman Instruments). Data were collected on the same electrode in transmission mode at the Pt L3edge, and in fluorescence mode at the Co K edge, with a Pt / Co reference foil positioned between I2 and I3 as a reference. The voltage cycling limits were from 0.05 to 1.00 V versus the reversible hydrogen electrode (RHE). Data collection was performed at the chosen potentials held during anodic sweeps. The electrode was fully cycled following each potential hold to clean the electrode surfaces. Typical experimental procedures were utilized with details provided previously (Jia et al., ACS Nano, (2015) 9:387–400). The data were processed and fitted using the IFEFFIT-based Athena (see Newville, J. Synchrotron Radiat., (2001) 8:322–324) and Artemis (Ravel et al., J. Synchrotron Radiat., (2005) 12:537–541) programs. Scans were calibrated, aligned and normalized with the background removed using the IFEFFIT suite (see Newville et al., Phys. Rev. B, (1993) 47:14126–14131). The χI data were modelled using single scattering paths calculated using FEFF6 (Ankudinov et al., Phys. Rev. B, (1998) 58:7565–7576).Atty. Dkt. No.: 102352-1142

[0058] In situ XAS studies in 0.1 M HClO4aqueous solution were conducted to understand the local atomistic and electronic structures. The Pt L3-edge X-ray absorption near- edge structure spectra show that the majority of Pt is in the metallic phase. The results from extended X-ray absorption fine-structure (EXAFS) fitting give a Pt–Pt bond length of 2.74 Å, which is 1.4% smaller than that in bulk Pt (2.78 Å in Pt foil), suggesting a compressive strain that is beneficial for the ORR (see, e.g., (1) Escudero-Escribano et al., Science, (2016) 352:73– 76; (2) Luo et al., Nat. Rev. Mater., (2017) 2:17059). The Co K-edge X-ray absorption near- edge structure results confirm the Co(II) oxidation state.

[0059] Importantly, the Co EXAFS and fitting results show a single Co–O peak with a coordination number of 6 ± 1, with no obvious Co–Co and Co–O–Co interactions, indicating sparsely dispersed Co(II)O6clusters. Moreover, the Co K edge is found to be independent of the applied potential, suggesting that the Co(II)O6 clusters are well-protected inside the nanoparticles and do not directly interact with the electrolyte. Meanwhile, the intensity of the Pt–O interaction reduced with decreasing potential from 0.90 V to 0.54 V, which is consistent with the desorption of oxygenated species on the Pt surface in this potential regime. Interestingly, a clear Pt–O interaction remains at 0.54 V, where the electrode surface is presumably free of oxygenated adsorbates (Jia et al., ACS Nano, (2015) 9:387–400). Such a remaining Pt–O interaction is attributable to the interaction between Pt atoms and the embedded Co(II)O6species. These combined STEM-EELS / EDS, XPS and in situ XAS characterizations robustly demonstrate that the sparsely dispersed Co(II)O6clusters are well-protected by the Pt shell, which is not surprising as any surface CoOxspecies will be removed during the acid-washing process.

[0060] Fuel cell performance. The practical fuel cell performance of the CoOx@Pt / C nanocatalyst was evaluated in a fuel cell fixture that consists of a single cell with an active area of 5 cm2(FIG.5A ). The membrane electrode assembly (MEA) was prepared by coating the catalyst ink directly on to the proton exchange membrane via an ultrasonic spray with a cathode loading of 0.09 mgPGMcm−2and an anode loading of 0.01 mgPGMcm−2(see below for more details). The performance was benchmarked against commercial c-Pt / C and c-PtCo / C under identical conditions. Notably, CoOx@Pt / C exhibits an initial, beginning-of-life (BOL) MA of 1.10 A mgPGM−1(FIG.6A), which is considerably higher than those of c-Pt / C (0.34 A mgPGM−1),Atty. Dkt. No.: 102352-1142 c-PtCo / C (0.57 A mgPGM−1) and the DOE target (0.44 A mgPGM−1). Furthermore, CoOx@Pt / C retains 88.2% of its initial MA after the aggressive square-wave AST (FIG.6A), which is much higher than the 22.5% and 36.8% MA retention values observed in c-Pt / C and c-PtCo / C, respectively. This MA retention achieved via the demanding square-wave AST compares favorably with other reported fuel cell catalysts (Stariha et al., J. Electrochem. Soc., (2018) 165: F492–F501). Notably, the end-of-life (EOL) MA (0.97 A mgPGM−1) is 3.7 times that of the DOE target (0.26 A mgPGM−1), 4.6 times the c-PtCo / C catalyst (0.21 A mgPGM−1) and 13.1 times the c- Pt / C catalyst (0.07 A mgPGM−1) (see, e.g., (1) Han et al., Sci., (2015) 8:258–266;(2) .Ko et al., J. Electrochem. Soc., (2021) 168:024512; (3) Energy Environ. Sci., (2021) 14:4948–4960; (4) Duan et al., Adv. Energy Mater., (2022) 12:2103144; (5) Li et al., Joule, (2019) 3:124–135; (6) Zhao et al., Matter, (2020) 3:1774–1790; (7) Chong et al., Science, (2018) 362:1276–1281; (8) Spendelow, Advanced Electro-Catalysts through Crystallographic Enhancement (U.S. Department of Energy, 2019).

[0061] Whereas the MA test under pure oxygen (O2) evaluates the intrinsic activity of the catalyst with little impact from mass-transport issues, the rated power density measured during the H2 / air test directly reflects the practical performance. The H2 / air fuel cell with the CoOx@Pt / C catalyst delivers a rated power density of 1.04 W cm−2, outperforming those with c- Pt / C (0.88 W cm−2) and c-PtCo / C (0.92 W cm−2) (FIG.6B), and exceeding the DOE target of 1.00 W cm−2. In addition, a useful metric for evaluating the PGM utilization in fuel cells is to normalize the power by the total PGM loading (measures in watts per milligram of PGM). In this regard, the CoOx@Pt / C MEA delivers a high Pt utilization of 10.4 W mgPGM−1(FIG.6C), which is well above the DOE target (8.0 W mgPGM−1). Furthermore, CoOx@Pt / C demonstrated a notably better performance than c-Pt / C or c-PtCo / C throughout the entire current density regime (Figures 6d and 6e). These power performances of CoOx@Pt / C are among the top reported performances for Pt-based catalysts (FIG.12). More importantly, CoOx@Pt / C exhibits a high EOL rated power density (Pt utilization) of 0.96 W cm−2(9.6 W mgPGM−1) (FIG.6B), much higher than those of c-Pt / C (0.32 W cm−2, 3.2 W mgPGM−1) and c-PtCo / C (0.61 W cm−2, 6.1 W mgPGM−1).Atty. Dkt. No.: 102352-1142

[0062] The stability can be evaluated further via the voltage loss at a given current density (for example, 0.8 A cm−2). The DOE 2020 target suggested a voltage loss of <30 mV at 0.8 A cm−2. When comparing the BOL and EOL polarization curves, CoOx@Pt / C exhibits superior durability, greatly outperforming c-Pt / C or c-PtCo / C, with only a 13.3 mV loss at 0.8 A cm−2(FIG.6F), which is substantially smaller than that of Pt / C (162.5 mV) or c-PtCo / C (61.3 mV). Moreover, there was only a slight drop throughout the entire curve after the AST for CoOx@Pt / C (FIG.6F), even in the more challenging HCD region (for example, an 11.1 mV loss at 1.0 A cm−2and a 12.0 mV loss at 1.25 A cm−2), clearly emphasizing the high stability of the CoOx@Pt / C catalyst. With the high MA retention of 88.2% and the remarkably low rated power loss of 7.5%, the CoOx@Pt / C nanocatalyst satisfies both the MA and the more challenging power durability targets. It is interesting to note that such a small voltage loss in the HCD region can be translated into an extraordinary projected lifetime of around 15,000 hours, which is three times that of the DOE 2020 target (5,000 hours) for light-duty transportation applications (see Fuel Cell Technologies Office Multi-Year Research, Development, and Demonstration Plan (U.S. Department of Energy, 2017)). To exclude other factors (for example, the carbon support) that may contribute to the improved durability, the in-house PtCo alloy catalyst (Pt79Co21 / C) on the same carbon support was also prepared, which shows considerably more Co leaching and worse durability for the EOL performance, confirming the critical role of the embedded CoOxin the CoOx@Pt / C catalyst (FIG.7).

[0063] In addition, a similar strategy was used to prepare NiOx@Pt / C and FeOx@Pt / C catalysts with sparsely dispersed metal oxide clusters, which also showed an impressive performance and durability (FIG.8). These studies suggest that the strategy described herein is transferable to other catalyst systems, providing insight into the catalyst design beyond conventional alloy systems.

[0064] The BOL and EOL performance metrics of the CoOx@Pt / C catalyst (BOL Mass Activity: 1.10 ± 0.046 A / mgPt; EOL Mass Activity: 0.97 ± 0.048 A / mgPt) simultaneously exceed the DOE targets (BOL Mass Activity: >0.44 A / mgPt; EOL Mass Activity: >0.264 A / mgPt). It should be highlighted that although a number of advanced catalyst designs have led to impressive BOL performances previously, the durability has been a persistent challenge and theAtty. Dkt. No.: 102352-1142 EOL performance is generally far from satisfactory (see, e.g., (1) Chong et al., Science, (2018) 362:1276–1281; (2) Zhao et al., Nat. Nanotechnol., (2022) 17:968–975).

[0065] In particular, there has been no report of a catalyst system that can deliver a rated power retention of 90% and an MA retention of 60% while satisfying all of the BOL performance metrics set in the DOE targets (see Fuel Cell Technologies Office Multi-Year Research, Development, and Demonstration Plan (U.S. Department of Energy, 2017)). As fuel cells must meet the performance requirements throughout their service lifetime, the EOL instead of the BOL performance ultimately determines the amount of Pt needed (see Cullen et al. Nat. Energy, (2021) 6:462–474). In this regard, the high EOL performance achieved in the nanocatalysts disclosed herein is particularly important for producing PEMFCs with a substantially reduced lifetime-adjusted cost in practical applications. More importantly, the performance and the durability of the developed CoOx@Pt / C catalyst are validated using the fuel cell with the larger active area of 25 cm2(FIG.9), further demonstrating the industrial significance of the catalyst design described herein.

[0066] The size distribution of the nanocatalyst in the MEA before and after the AST was analyzed. Overall, the size of c-Pt / C increased dramatically from 3.5 ± 0.8 nm at BOL to 7.2 ± 3.6 nm at EOL; a similar size growth was also observed in c-PtCo / C (from 4.7 ± 1.4 nm at BOL to 8.1 ± 3.5 nm at EOL). Such a substantial size increase can be attributed to nanoparticle detachment, movement and coalescence, as well as oxidative dissolution, diffusion and the Ostwald ripening process (see Borup et al., Curr. Opin. Electrochem., (2020) 21:192–200). By contrast, the average size of CoOx@Pt / C increased only moderately from 1.5 ± 0.2 nm at BOL to 2.5 ± 1.2 nm at EOL. Similarly, from carbon monoxide (CO) stripping experiments in the MEAs at 80 °C, the CoOx@Pt catalyst exhibited a much higher EOL ECSA (35.9 ± 0.9 m2gPGM−1) compared with the c-Pt / C (8.9 ± 1.8 m2gPGM−1) and c-PtCo / C (12.3 ± 1.0 m2gPGM−1) catalysts. These analyses are consistent with the observed EOL performance trend and clearly highlight the benefit of the catalyst design disclosed herein for effectively restraining the dissolution of Pt and thus limiting the size growth. In particular, the embedded CoOxstrengthens the interactions with the surface Pt atoms due to the strong metal–oxide interaction, greatly slowing down theAtty. Dkt. No.: 102352-1142 oxidation and dissolution of surface Pt atoms and retarding the Ostwald ripening process (FIG, 5B).

[0067] In addition, the MEA cross-section after the AST was characterized, and the Pt loss to the Pt band as a result of redeposition of the dissolved Pt at the membrane interface was evaluated (see Padgett et al., J. Electrochem. Soc., (2019) 166:F198–F207). Analyses show that the MEA with CoOx@Pt / C shows only a 12.8 ± 1.9% Pt mass loss into the Pt band, which is much smaller than that of the MEA with c-PtCo / C (25.4 ± 5.3% Pt mass loss). Considering the much smaller size of CoOx@Pt / C (~1.5 nm) than c-PtCo / C (~4.7 nm), the surface-area- normalized dissolution rate of the CoOx@Pt / C catalyst is only ~16–25% of that of the much larger-sized c-PtCo / C sample, further highlighting the high stability of the fine CoOx@Pt / C nanoparticles in spite of their small size.

[0068] Furthermore, EDS composition analyses and dissolution experiments confirmed that the Co content of the CoOx@Pt nanoparticles was retained well during the AST, with a minor decrease from 4.9 at% at BOL to 4.5 at% at EOL (FIG.10), and with the Co content remaining sparsely embedded inside the nanoparticle. By contrast, there was a drastic composition change for the c-PtCo nanoparticles from 26.4 at% (BOL) to 9.8 at% (EOL) (FIG. 10), indicating substantial leaching of Co cations that may cause a considerable drop in intrinsic activity and cation poisoning effects on the mass transport and overall device performance (see, e.g., (1) Braaten et al., J. Electrochem. Soc., (2019) 166:F1337–F1343; (2) Dumont et al., ECS Trans., (2017) 80:861–867).

[0069] Moreover, CO stripping experiments were used to probe the change in the chemisorption behavior of different catalysts. Previous studies have shown that the lower the peak potential of CO oxidation, the weaker the CO binding strength, which indicates a lower Pt d-band center and weaker oxygen adsorption energy that is favorable for ORR kinetics (see Nilsson et al., Catal. Lett., (2005) 100:111–114). Interestingly, after the AST, both c-Pt / C and c- PtCo / C showed a notable shift to a higher binding energy, whereas such a shift was not observed in CoOx@Pt / C (FIG.11), indicating that the favorable weak oxygen adsorption characteristics of CoOx@ Pt / C were retained, which is essential for an excellent EOL performance.Atty. Dkt. No.: 102352-1142

[0070] To further understand the fundamental origin of the high stability of the CoOx@Pt / C catalyst, the atomistic configurations, energetics and electronic structure of the fine Pt nanoparticles with embedded CoOx clusters were computationally predicted to compare their thermodynamic and kinetic stability with pure Pt and PtCo alloy nanoparticles. No rigorous information exists in the literature on the atomistic configurations of Pt-based nanoparticles in this size regime, not even for pure Pt, which must be predicted ab initio. Extensive density- functional global optimization (DF-GO) (see Aprà et al., Phys. Rev. B, (2006) 73:205414) on nanoparticles with different atom numbers and stoichiometries were carried out to determine their global minimum (GM) configurations, the corresponding energetics, electronic structure and interaction with oxygen adatoms. A simplified but representative model of a single CoOx inside the Pt shell (around 80 atoms), at a size (of ~1 nm) was used, which is at the smaller end of the experimental nanoparticle that is accessible for both accurate and affordable computations. Calculations indicated that the preferred stoichiometry of the CoOx unit was CoO6, in agreement with experimental EXAFS studies.

[0071] Interestingly, the DF-GO studies showed that the most stable structure in this size range (for example, Pt78, Pt77Co and Pt77CoO6) typically exhibits a barrel-like structural motif, which is energetically more stable by several electronvolts with respect to other motifs, such as the truncated octahedron Pt79with a closed-shell crystalline-like Wulff shape. It is noteworthy to observe that simulated TEM images for the Pt77CoO6 GM configuration show apparently similar features to those observed in experimental STEM studies. Moreover, Bader analysis shows a Bader charge of 1.34 on the Co atoms, which compares well with the Bader charge of 1.36 for Co in bulk CoO, consistent with the Co2+status observed in XAS studies. It was also noted that the predicted O and Pt coordination numbers of Pt (evaluated with a cut-off distance of 3.00 Å) are 0.21 and 6.23, in agreement with the experimental values of 0.4 ± 0.3 and 6 ± 1, respectively. The predicted structure also exhibits a significant compressive strain, in agreement with XAS experiments.

[0072] More importantly, the calculations show that the free energy of the Pt nanoparticle with embedded CoO6(Pt77CoO6) is 1.85 eV lower than that of pure Pt (Pt78) and 1.55 eV lower than that of the PtCo alloy (Pt77Co) structures that have a similar size. The workAtty. Dkt. No.: 102352-1142 function of these structures was further calculated, and the order Pt78(4.94 eV) < Pt77Co (4.97 eV) < Pt77CoO6 (5.04 eV) was observed, which correlates well with the experimentally observed order of stability under bias. Indeed, the nanocatalysts need to adjust their work function via oxygen adsorption and / or adding positive charge so that their work function equalizes the target electrode potential under the operating conditions (for example, 0.90–0.95 V) (see Sundararaman et al., J. Chem. Phys., (2017) 146:114104). A larger work function implies a lower positive charge density and makes the luster less liable to oxidation, which contributes to improved stability against oxidative dissolution.

[0073] Next, the relative kinetic stability of these clusters under ORR conditions was explored. Previous studies have shown that Pt dissolution during fuel cell operation mainly occurs via an oxidative / reductive dissolution mechanism (see Kodama et al., Nat. Nanotechnol., (2021) 16:140–147). Therefore, the oxygen adsorption energies on the surface sites of the Pt78, Pt77Co and Pt77CoO6 nanocatalysts was calculated to probe the kinetics of Pt dissolution. It is interesting to note that both Pt78and the Pt77Co alloy have a considerably higher population of surface sites with stronger oxygen adsorption energy values, which are considered to be more vulnerable sites, that is, more prone to oxidation and oxidative dissolution (see, e.g., (1) Zhu et al. Adv. Funct. Mater., (2022) 32:2203883; (2) Lopes et al., ACS Catal. (2016) 6:2536–2544; (3) Huang et al., Nat. Catal., (2022) 5:513–523).

[0074] By contrast, the surface Pt sites on Pt77CoO6 generally show weaker oxygen adsorption energy values and thus have a higher kinetic barrier for oxidative dissolution. Overall, these calculations explain the high stability of CoOx@Pt / C against the oxidative dissolution of surface Pt atoms and the Oswald ripening process, which further prevents the embedded CoOx from leaching out, as observed via experiment. Although these beneficial effects of embedded CoOxcan be reasonably demonstrated using the simplified Pt77CoO6model, a larger predictive structural model (Pt124Co6O18) with a size of approximately 1.5 nm in diameter was also generated, and a composition was reached that matched closely with the experimental results. Preliminary simulation results confirmed that the Pt124Co6O18model is also energetically stable, thus suggesting a general behavior for CoOx@Pt particles in the investigated size and composition ranges.Atty. Dkt. No.: 102352-1142

[0075] The MEA made with the CoOx@Pt nanocatalyst demonstrates a high initial MA of 1.10 A mgPGM−1and a high rated power density of 1.04 W cm−2(10.4 W mgPGM−1) at a total PGM loading of just 0.10 mg cm−2. With minimized metal dissolution, the MEA delivers a notably high durability, achieving an MA retention of 88.2%, a voltage loss of only 13.3 mV at 0.8 A cm−2and a small rated power loss of 7.5% after an accelerated stress test (AST), greatly exceeding the relevant DOE performance targets.

[0076] MEA fabrication. MEA fabrication followed the protocol disclosed previously (see Li et al. Science, (2016) 354:1414–1419). The catalyst ink was made by mixing the catalysts with the ionomer solution (Aquivion D83-06A) and water–isopropanol solvent through ultrasonication. The catalyst ink was then spray-coated on to a reinforced perfluorosulfonic acid membrane (12 μm thickness) using a Sono-Tek ultrasonic spray system to form a catalyst-coated membrane (CCM).

[0077] For a fair comparison, the cell components and fabrication conditions were the same in all MEAs except for the cathode catalysts. Accordingly, the MEAs were labelled by the cathode catalysts. The anode catalyst was the commercial 10% Pt / C with a PGM loading fixed at 0.01 mgPGMcm−2. The mass ratio of anode ionomer to carbon was 0.6. The commercial 40% Pt / C (c-Pt / C) and 40% PtCo / C (c-PtCo / C) were used as a benchmark for the cathode catalyst. The ionomer-to-carbon ratio was 0.6 for c-Pt / C and c-PtCo / C, whereas it was 0.7 for CoOx@Pt / C. The cathode catalyst loading was controlled to be about 0.09 mgPGMcm−2. The PGM loading on the anode / cathode was confirmed via ICP-AES measurements. The fabricated CCM was dried under vacuum for an hour to evaporate the solvents. For the MEA fabrication, two gas diffusion layers, two PTFE gaskets and the prepared CCM were pressed together. The thickness of the PTFE gasket was 127 μm. The thickness of the gas diffusion layer, which includes a microporous layer, was 175 μm.

[0078] MEA activity tests. The MEA was loaded in a single-cell fixture with an active area of either 5 or 25 cm2with a serpentine flow field (Fuel Cell Technology) and tested in the Scribner 850e fuel cell test stand with an automatic back-pressure system. First, MEA activation was carried out by holding the voltage at 0.5 V under a H2 / air flow of 100 / 100 standard cubic centimeters per minute (sccm) for the 5 cm2cell and 500 / 500 sccm for the 25 cm2cell at 80 °C,Atty. Dkt. No.: 102352-1142 100% relative humidity (RH), 150 kPaabs(all pressure noted herein refers to the absolute pressure) until the current reached a plateau (usually within about 30 min). Then, the MEA was further activated at oversaturated conditions, where the fuel temperate was set at 80 °C for both the cathode and anode before entering the cell (which was held at 60 °C). The MA tests were performed at 80 °C, 100% RH, 150 kPaabswith a respective H2 / O2flow rate of 835 / 2,000 sccm.

[0079] The rated power density at BOL and EOL was tested using the Scribner 850e fuel cell fixture at 94 °C, 250 kPaabsand 100% RH (see Escudero-Escribano et al., Science, (2016) 352, 73–76). The H2 / air gas flow rate was 126 / 400 sccm, respectively, for the 5 cm2cell and 630 / 2,000 sccm for the 25 cm2cell, which is equivalent to a stoichiometry of 1.5 / 2.0 at 2.4 A cm−2(the endpoint of the current density plot). The DOE has set a heat-rejection limit (Q / ΔTi) of <1.45 for light-duty vehicles, assuming a 90 kW stack gross power (Pstack) required for an 80 kW net power and an ambient temperature (Tambient) of 40 °C as the baseline. Tstack is defined as the temperature of the fuel cell stack. The power density at the rated voltage (Vrated) is defined as the rated power, which is the maximum power aligned with the heat management capability. The rated voltage is defined in equation (1) (see Yarlagadda et al., ACS Energy Lett., (2018) 3:618– 621):

[0080] The DOE protocol assumes Pstack = 90 kW and Tambient = 40 °C, and if Tstack = 94 °C then Q / ΔTi= 1.445, which meets the target (Q / ΔTi≤ 1.45), and Vratedwill be around 0.67 V. Thus, the rated power will be measured at 0.67 V (see Yarlagadda et al.).

[0081] All of the MEA’s BOL and EOL performance metrics, such as MA, ECSA and H2 crossover, were recorded using the Scribner 850e fuel cell test station and a Scribner 885 potentiostat. The ECSA was determined by integrating the CO stripping peak, assuming 420 μC cmPt−2. The CO stripping test in the MEA followed the protocol from Garrick et al. (J. Electrochem. Soc., (2016) 164:F55–F59).

[0082] MEA accelerated stress testing. For the ASTs, the more aggressive 30,000-cycle square-wave protocol was adopted, with each cycle holding the MEA at the voltage of 0.60 V forAtty. Dkt. No.: 102352-1142 3 s and then at 0.95 V for 3 s. The full AST was carried out at 80 °C, ambient pressure, 100% RH with a H2 / N2 flow of 100 / 100 sccm for the anode and cathode, respectively. The AST protocol used herein strictly followed the DOE protocol for PGM-based catalysts (see Fuel Cell Technologies Office Multi-Year Research, Development, and Demonstration Plan (U.S. Department of Energy, 2017)).

[0083] Calculation of additional voltage loss in the load cables of a 25 cm2MEA cell. As the cell current in the 25 cm2cell is five times that of the 5 cm2cell, the voltage loss due to the wiring resistance will also increase. This additional voltage loss compared with the 5 cm2cell at a different operating current in the 25 cm2cell can be calculated via idifference × Rcables. The idifference is the difference in absolute current for a 25 cm2cell and a 5 cm2cell, and Rcables is the ohmic resistance of the fuel cell wiring cables. For example, at 1.5 A cm−2, idifferencewill be 30 A and the additional voltage loss will be 28.8 mV, given a total cableof 0.96 mΩ (0.48 mΩ per 60 cm, for four-gauge cell cable) (see Application note – minimum load and wiring resistance. Scribner https: / / www.scribner.com / wp-content / uploads / 2019 / 02 / Minimum-Load- Resistance.pdf (2022)).

[0084] Projection of fuel cell lifetime. As defined by the DOE, the fuel cell lifetime is equal to the operation time until a 10% loss is reached in cell voltage at the rated power (see (1) Fuel Cell Technologies Office Multi-Year Research, Development, and Demonstration Plan (U.S. Department of Energy, 2017); (2) Marcinkoski et al., Hydrogen Class 8 Long Haul Truck Targets (U.S. Department of Energy, 2019)). The current density at 1.5 A cm−2was chosen to study the voltage loss, as it is the current density where the targeted rated power would be delivered given the rated voltage of 0.67 V. The lifetime of the tested MEA is then predicted from the ratio between 10% of the initial voltage (Vinitial) at 1.5 A cm−2and the voltage loss rate at 1.5 A cm−2(equation (2)). An acceleration factor of 100 (with respect to time) over the drive cycle is used according to Stariha et al. (Cells. J. Electrochem. Soc., (2018) 165:F492–F501). Note that a full square-wave cycle takes 6 s (0.60 V for 3 s and 0.95 V for 3 s).Atty. Dkt. No.: 102352-1142

[0085] Computational studies. DF-GO simulations were performed using the basin- hopping (BH) algorithm (see (1) Aprà et al., Phys. Rev. B, (2006) 73:205414; and (2) Wales et al., J. Phys. Chem. A, (1997) 101:5111–5116) as implemented using in-house Python code. In each BH step, starting from a given locally relaxed initial structure defined via 3N atomic coordinates (where N is the number of atoms in the cluster), the GO algorithm generates a new configuration via a random perturbation of the coordinates that is then subjected to a local geometry optimization, after which the new configuration is accepted or rejected according to a Metropolis criterion, that is, with a probability min{1, exp[(ΔE / KbT)]}, where ΔE is the energy difference between the initial and the new locally relaxed configurations, Kb is the Boltzmann constant, and T is a fictitious temperature (which in the simulations varies between 3,500 and 5,000 K). The procedure starts from a sample of 3N arbitrary or selected atomic coordinates and iteratively continues with the random generation of new local minima starting from the currently accepted one, eventually providing a putative GM as the lowest-energy configuration. All of the local minima, which belong to a DFT-PBE (Perdew et al., Phys. Rev. Lett., (1996) 77:3865– 3868) potential energy hyper-surface, were generated using the OpenMX code (Ozaki, Phys. Rev. B., (2003) 67:155108), which solves the Kohn–Sham equations within the pseudopotential- LCAO (linear combination of atomic orbitals) framework using localized pseudo-atomic numerical basis sets (Ozaki et al., Phys. Rev. B, (2005) 72:045121), whereas the Hartree potential is evaluated using an FFT (fast Fourier transform) solver that relies on the use of a plane-wave basis set whose size depends on a chosen cut-off energy (Duy et al., Comput. Phys. Commun., (2014) 185:153–164). In simulations, the Kohn–Sham energies were evaluated by feeding OpenMX with ‘Quick’-type basis sets, norm-conserving pseudopotentials and an energy cut-off of 150 Ry (Ozaki et al., Phys. Rev. B, (2004) 69:195113).

[0086] The energy sequence as a function of the BH iterations was carried out for two typical GO runs: a DF-GO run on the Pt79 cluster starting from the Pt79 truncated octahedron as the initial configuration; and random sampling of the oxygen binding sites on the Pt77CoO6 GM cluster. In the latter case, oxygen binding sites were singled out using a Voronoi tessellation procedure (Barber et al., ACM Trans. Math. Softw., (1996) 22:469–483). After tessellating the surface using the Voronoi procedure, a stochastic approach was used to sample the adsorption sites on the Voronoi tessels. In detail, for each tessel (triangle), the outward axis perpendicularAtty. Dkt. No.: 102352-1142 to the tessel’s surface was calculated, and the geometry relaxation with an oxygen atom randomly located in a plane perpendicular to the tessel and containing its outward axis was initiated. All of the tools needed for analysis of the results of the simulations are encoded in Python software.

[0087] Starting from the OpenMX DF-GO database, a more accurate approach was used to predict the energies and atomic equilibrium configurations via local geometry relaxation of candidate low-energy GO structures using the plane-wave QUANTUM ESPRESSO (QE) code (Giannozzi et al., J. Phys. Condens. Matter, (2009) 21:395502). All of the energies reported in the main text refer to the QE approach. The QE calculations were conducted spin-unrestricted at the DFT / PBE level augmented with both Grimme-D3 dispersion and Hubbard-U corrections (U = 3.5 eV for Co and zero for the other atomic species). A cut-off energy of 40 Ry was used for describing the wavefunctions, and a cut-off energy of 200 Ry was used for describing the electron density and the GBRV ultrasoft pseudopotential library (Garrity et al., Comput. Mater. Sci., (2014) 81:446–452). Since the investigated systems are finite, the k-space grid sampling was limited to the Γ point only. Cubic unit cells with an edge of 22 Å were used. The threshold criteria for the energy and forces were 1.0 × 10−4a.u. and 1.0 × 10−3a.u., respectively, for terminating the relaxation runs. Work functions were evaluated according to the procedure proposed in the literature (see, e.g., (1) Dabo et al., Phys. Rev. B, (2008) 77:115139; and (2) Andreussi et al., Phys. Rev. B, (2014) 90:245101). TEM images were simulated using the QSTEM code (Koch, C. Determination of core structure periodicity and point defect density along dislocations. Dissertation, Arizona State University (2002). https: / / www.physik.hu- berlin.de / en / sem / images / koch02_phdthesis.pdf).

[0088] The herein described subject matter sometimes illustrates different components contained within, or connected with, different other components. It is to be understood that such depicted architectures are illustrative, and that in fact many other architectures can be implemented which achieve the same functionality. In a conceptual sense, any arrangement of components to achieve the same functionality is effectively “associated” such that the desired functionality is achieved. Hence, any two components herein combined to achieve a particular functionality can be seen as “associated with” each other such that the desired functionality isAtty. Dkt. No.: 102352-1142 achieved, irrespective of architectures or intermedial components. Likewise, any two components so associated can also be viewed as being “operably connected,” or “operably coupled,” to each other to achieve the desired functionality, and any two components capable of being so associated can also be viewed as being “operably coupleable,” to each other to achieve the desi”ed functionality. Specific examples of operably coupleable include but are not limited to physically mateable and / or physically interacting components and / or wirelessly interactable and / or wirelessly interacting components and / or logically interacting and / or logically interactable components.

[0089] With respect to the use of plural and / or singular terms herein, those having skill in the art can translate from the plural to the singular and / or from the singular to the plural as is appropriate to the context and / or application. The various singular / plural permutations may be expressly set forth herein for sake of clarity.

[0090] It will be understood by those within the art that, in general, terms used herein, and especially in the appended claims (e.g., bodies of the appended claims) are generally intended as “open” terms (e.g., the term “including” should be interpreted as “including but not limited to,” the term “having” should be interpreted as “having at least,” the term “includes” should be interpreted as “includes but is not limited to,” etc.).

[0091] Although the figures and description may illustrate a specific order of method steps, the order of such steps may differ from what is depicted and described, unless specified differently above. Also, two or more steps may be performed concurrently or with partial concurrence, unless specified differently above. Such variation may depend, for example, on the software and hardware systems chosen and on designer choice. All such variations are within the scope of the disclosure. Likewise, software implementations of the described methods could be accomplished with standard programming techniques with rule-based logic and other logic to accomplish the various connection steps, processing steps, comparison steps, and decision steps.

[0092] It will be further understood by those within the art that if a specific number of an introduced claim recitation is intended, such an intent will be explicitly recited in the claim, and in the absence of such recitation, no such intent is present. For example, as an aid toAtty. Dkt. No.: 102352-1142 understanding, the following appended claims may contain usage of the introductory phrases “at least one” and “one or more” to introduce claim recitations. However, the use of such phrases should not be construed to imply that the introduction of a claim recitation by the indefinite articles “a” or “an” limits any particular claim containing such introduced claim recitation to inventions containing only one such recitation, even when the same claim includes the introductory phrases “one or more” or “at least one” and indefinite articles such as “a” or “an” (e.g., “a” and / or “an” should typically be interpreted to mean “at least one” or “one or more”); the same holds true for the use of definite articles used to introduce claim recitations. In addition, even if a specific number of an introduced claim recitation is explicitly recited, those skilled in the art“w”ll recog“iz” that such recitation should typically be“interpreted ”o me“n at least ”he recited number (e.g., the bare recitation of “two recitations,” without other modifiers, typically means at least two recitations, or two or more recitations).

[0093] Furthermore, in those instances where a convention analogous to “at least one of A, B, and C, etc.” is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., “a system having at least one of A, B, and C” would include but not be limited to systems that have A alone, B alone, C alone, A and B together, A and C together, B and C together, and / or A, B, and C together, etc.). In those instances where a convention analogous to “at least one of A, B, or C, etc.” is used, in general, such a construction is intended in the sense one having skill in the art would understand the convention (e.g., “a system having at least one of A, B, or C” would include but not be limited to systems that have A alone, B alone, C alone, A and B together, A and C together, B and C together, and / or A, B, and C together, etc.). It will be further understood by those within the art that virtually any disjunctive word and / or phrase presenting two or more alternative terms, whether in the description, claims, or drawings, should be understood to contemplate the possibilities of including one of the terms, either of the terms, or both terms. For example, the phrase “A or B” will be understood to include the possibilities of “A” or “B” or “A and B.”

[0094] Further, unless otherwise noted, the use of the words “approximate,” “about,” “around,” “substantially,” etc., mean plus or minus ten percent.Atty. Dkt. No.: 102352-1142

[0095] Although the present embodiments have been particularly described with reference to preferred examples thereof, it should be readily apparent to those of ordinary skill in the art that changes and modifications in the form and details may be made without departing from the spirit and scope of the present disclosure. It is intended that the appended claims encompass such changes and modifications.

Claims

Atty. Dkt. No.: 102352-1142 WHAT IS CLAIMED IS:

1. A nanocatalyst comprising platinum nanoparticles and base transition metal oxide clusters, wherein each base transition metal oxide cluster is enclosed by a platinum shell.

2. The nanocatalyst of claim 1, wherein the base transition metal is selected from the group consisting of Co, Fe, Ni, Mn, Cu, and Zn.

3. The nanocatalyst of claim 1 or claim 2, wherein the base transition metal is selected from the group consisting of Co, Fe, and Ni.

4. The nanocatalyst of any one of claims 1-3, wherein the base transition metal is Co.

5. The nanocatalyst of any one of claims 1-4, wherein the platinum nanoparticles are on a carbon support.

6. The nanocatalyst of claim 5, wherein the carbon support comprises carbon black, carbon nanotube, carbon fiber, or graphene.

7. The nanocatalyst of claim 5 or claim 6, wherein the carbon support comprises carbon black.

8. The nanocatalyst of any one of claims 1-7, wherein the nanocatalyst has an average particle size of about 1 nm to about 5 nm.

9. The nanocatalyst of any one of claims 1-8, wherein the nanocatalyst has an average particle size of about 1.5 nm.

10. The nanocatalyst of any one of claims 1-9, wherein the base transition metal oxide clusters are sparsely embedded in the platinum nanoparticles.

11. The nanocatalyst of any one of claims 1-10, wherein the nanocatalyst has a Co:Pt atomic ratio of about 4.7:95.3 to about 4.9:95.

1.

12. A fuel cell comprising the nanocatalyst of any one of claims 1-11.

13. The fuel cell of claim 12, wherein the fuel cell is a proton exchange membrane fuel cell.Atty. Dkt. No.: 102352-1142 14. A method of preparing the nanocatalyst of any one of claims 1-11, the method comprising: mixing platinum-based catalyst precursor with base transition metal catalyst precursor on a carbon support in an organic solvent; evaporating the solvent to form a dry powder mixture; annealing the dry powder mixture to form an initial annealed mixture; washing the initial annealed mixture under acidic conditions to form an acid-treated mixture; and annealing the acid-treated mixture to provide the nanocatalyst.

15. The method of claim 14, wherein the platinum-based catalyst precursor is platinum(II) acetylacetonate.

16. The method of claim 14 or claim 15, wherein the base transition metal catalyst precursor is cobalt(II) acetylacetonate.

17. The method of any one of claims 14-16, wherein the solvent is acetone.

18. The method of any one of claims 14-17, wherein the dry powder mixture is annealed at a temperature of at least 200⁰C under inert atmosphere.

19. The method of claim 18, wherein the dry powder mixture is annealed at a temperature of 250⁰C under argon atmosphere.

20. The method of any one of claims 14-19, wherein the dry powder mixture is annealed for at least 6 hours.

21. The method of any one of claims 14-20, wherein the acidic conditions comprise heating with aqueous sulfuric acid.

22. The method of any one of claims 14-21, wherein the acidic conditions comprise heating at 80⁰C with aqueous sulfuric acid for at least 10 hours.

23. The method of any one of claims 14-22, wherein the acid-treated mixture is annealed at a temperature of at least 150⁰C.Atty. Dkt. No.: 102352-1142 24. The method of any one of claims 14-23, wherein the acid-treated mixture is annealed at a temperature of 180⁰C under hydrogen / argon atmosphere.

25. The method of any one of claims 14-24, wherein the acid-treated mixture is annealed for at least 60 minutes.

Images