SINGLE-ATOM CATALYST FOR USE IN A WATER SPLITTING PROCESS AND A METHOD FOR PREPARING THE SAME

Abstract

A single-atom catalyst for use in a water splitting process includes at least one support material and at least one metal catalyst deposited on the surface of the at least one support material. The at least one support material is made of tungsten carbide obtained from a tungstate-metal-aryl compound precursor, and the at least one metal catalyst is selected from a group including Fe, Ni, Mn, Co, Cu, Zn, V, Ru, Ir, Ca, Pd, Pt or combinations thereof.

Claims

1. A single-atom catalyst for use in a water splitting process comprising: at least one support material; and at least one metal catalyst deposited on the surface of the at least one support material, wherein: the at least one support material is made of tungsten carbide obtained from a tungstate-metal-aryl compound precursor, and the at least one metal catalyst is selected from the group consisting of Fe, Ni, Mn, Co, Cu, Zn, V, Ru, Ir, Ca, Pd, and Pt or combinations thereof.

2. The catalyst according to claim 1, wherein the metal catalyst is one of Fe, Ni or FeNi.

3. The catalyst according to claim 2, for use as a catalyst for oxygen evolution reaction in a water splitting process.

4. The catalyst according to claim 2, having an overpotential η (10 mA/cm.sup.2, alkaline solution) of less than 300 mV.

5. The catalyst according to claim 2, having a turnover frequency (η = 300 mV) of at least 0.04 s.sup.-1.

6. The catalyst according to claim 1, wherein the metal catalyst is one of Ru, Ni, RuNi, RuMn, RuV, and RuCa.

7. The catalyst according to claim 6, for use as a catalyst for hydrogen evolution reaction in a water splitting process.

8. The catalyst according to claim 6, having an overpotential η (10 mA/cm.sup.2, alkaline solution) of less than 50 mV.

9. A method of preparing the metal-tungsten-carbide catalyst according to claim 1, comprising: providing an aqueous solution of at least one aryl compound, adding a solution of at least one metal salt, wherein the metal is selected from the group consisting of Fe, Ni, Mn, Co, Cu, Zn, V, Ru, Ir, Ca, Pd, and Pt to the aryl compound solution, adding a sodium tungstate solution to the aryl compound-metal solution, whereby a tungstate-metal-aryl compound precursor is precipitated, collecting the tungstate-metal-aryl compound precursor, and calcinating the tungstate-metal-aryl compound precursor at a temperature between 700 and 1100° C. to obtain the metal-tungsten-carbide catalyst.

10. The method according to claim 9, wherein the at least one aryl compound is dopamine.

11. The method according to claim 9, wherein the total concentration of metal salt in the metal salt solution is in a range between 0.5 mmol and 1.5 mmol.

12. The method according to claim 9, wherein, when two different metal salts are used, the feeding amounts of the metal salts are adjusted to obtain different amounts and ratios of the two metals in the tungstate-metal-aryl compound precursor.

13. An electrode comprising the catalyst according to claim 1.

14. (canceled)

15. (canceled)

16. The catalyst according to claim 4, having an overpotential η (10 mA/cm.sup.2, alkaline solution) of less than 230 mV.

17. The catalyst according to claim 5, having a turnover frequency (η = 300 mV) of at least 2 s.sup.-1.

18. The catalyst according to claim 8, having an overpotential η (10 mA/cm.sup.2, alkaline solution) of less than 20 mV.

19. The method according to claim 9, wherein calcinating the tungstate-metal-aryl compound precursor is done at a temperature between 800 and 900° C.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The solution is explained in the following in more detail by means of the examples with reference to the Figures.

[0058] FIG. 1a shows a schematic illustration of the synthetic procedures: a, formation of the metal-organic complex; b, further reaction, and growth of microspherical tungsten-based precursor; c, carbonization process.

[0059] FIGS. 1b-c shows DFT module of a WCx-FeNi structure.

[0060] FIG. 1d shows k.sup.3-weighted FT spectra from the XAFS analysis.

[0061] FIGS. 2a-g show Morphology and structure of the materials: (a) Schematic illustration of WC.sub.x-FeNi catalyst, consisting of Fe and Ni atoms stabilized on WC.sub.x nanocrystallites (majority component) surrounded by carbon sheets. (b) HAADF-STEM image of a WC.sub.x nanocrystallite. (c and d) HAADF-STEM images of the atom arrays on the WC.sub.x and surface loaded single atoms. A schematic atomic structure and a simulation HAADF-STEM image is superposed on the image. (e) The corresponding intensity profiles along (1̅21̅̅) plane over the dimmer atomic columns (marked by the arrows 1-5). (f) Atomically resolved STEM spectrum imaging the edge of the nanocrystallite to show the distribution of W, Fe, Ni elements. (g) EDX spectra of a single-pixel at different positions (1-6).

[0062] FIG. 3 shows OER performances of WC.sub.x-FeNi catalysts obtained under different temperatures. The results indicate the lowest overpotential of WC.sub.x-FeNi obtained at 900° C.

[0063] FIGS. 4a-f show performances of WC.sub.x-FeNi catalysts and control samples in a three-electrode configuration in 1 M KOH at room temperature: (a and b) The CV and OER polarization curves of different catalysts with real-time iR correction (4.4 ohms). (c) The OER data analysis of overpotentials (obtained from OER polarization curves at the current density of 10 mA cm.sup.-2) and Tafel slopes. (d) TOFs of the measured catalysts calculated based on the loaded Fe, Ni, and FeNi atoms at different overpotentials. (e) Mass activities of the catalysts based on Fe, Ni, and Fe-Ni atoms. (f) Overpotentials of WC.sub.x-FeNi with different FeNi ratios (Fixed Ni (or Fe) at 0.5 at.% and change Fe (or Ni) amount), and the corresponding activity trend. Error bars represent the average values (mean ± SD, n = 5).

[0064] FIG. 5 shows OER performances of WC.sub.x-FeNi with different catalysts loading amounts: a) activity based on the geometric area of the electrode; b) activity based on the loading mass of the catalysts. The required overpotential to reach 10 mA cm.sup.-2 can be further decreased from 237 mV to 201 mV when the loading amount of WC.sub.x-FeNi on the glassy carbon electrode increased from 0.38 mg/cm.sup.-2 (contains ~3.12 wt.% of Fe (1.39 wt.%) and Ni (1.73 wt.%) in total, indicates a loading of Fe and Ni is 11.85 .Math.g/cm.sup.-2) to 1.14 mg/cm.sup.-2 (Fe and Ni: 35.57 .Math.g/cm.sup.-2), the required overpotential to reach 10 mA cm.sup.-2 is decreased from 237 mV to 201 mV.

[0065] FIG. 6 shows OER performances of WC.sub.x-FeNi with different Fe and Ni ratios in the catalysts.

[0066] FIG. 7 shows HER activity in 1 M KOH of WCx-Ru2 obtained at different temperatures.

[0067] FIG. 8 shows Investigation of Ni atom effects on the performance of Ru catalysts.

[0068] FIG. 9 shows Activity and Mass activity compare with the Pt/C and WCx-Ru; and

[0069] FIG. 10 shows WC-RuX bi-atomic system for H.sub.2 evolution reaction (X=Cu, Co, Mn, Ni, V, Ca) .

DETAILED DESCRIPTION

A) Oxygen Evolution Reaction (OER)

[0070] Synthesis of WC.sub.x-FeNi, WC.sub.x-Ni, WC.sub.x-Fe, and WC.sub.x catalysts:

[0071] First, 5 mmol (0.948 g) dopamine hydrochloride (DA) was dissolved in 50 mL water, resulting in a 0.1 M dopamine in water solution. The pH value of the dopamine solution was adjusted to ~2 by adding 2 mL of 1 M HCl. After that, 0.25 mmol FeCl.sub.3.Math.6H.sub.2O and 0.25 mmol NiCl.sub.2.Math.6H.sub.2O were added together into the DA solution and well mixed by stirring. Following, 50 mL sodium tungstate solution (containing 5 mmol Na.sub.2WO.sub.4.Math.2H.sub.2O) was dropped slowly (about 10 min) into the DA-FeNi solution. The reaction happens immediately, then the precipitation starts, and the color of the suspension changes from dark brown to bright yellow during the addition of sodium tungstate. Finally, a greenish-yellow precipitate formed. The reaction was further stirred for 1 h, and then the product was collected by centrifugation and washed with DI water and ethanol for 3 times. The products were dried in an oven at 60° C. overnight to obtain the DA-W-FeNi precursor.

[0072] Then the DA-W-FeNi precursor was carbonized in an Argon furnace at different temperatures (800, 900, and 1000° C.) for 2 h, with a ramp of 2° C./min. The final black powder was designated as WC.sub.x-FeNi-800, WC.sub.x-FeNi (indicating the materials carbonized at 900° C., unless specified), WC.sub.x-FeNi-1000, and collected in a glass bottle for further use. To further confirm the influence of the temperature on the OER performance and crystals structures of the synthesized catalysts, the DA-W-Ni precursor was also thermally treated at different temperatures (700, 800, 900, and 1000° C.) for 2 h.

[0073] In order to get different contents and different ratios of Fe/Ni in the final WC.sub.x-FeNi catalysts, the feeding amounts of Fe and Ni source in the DA-W-FeNi precursor were adjusted. All the synthesized catalysts with different FeNi amounts synthesized so far are summarized in Table 1. Furthermore, for comparison, the same synthetic approach as for WC.sub.x-FeNi has been carried out to synthesize WC.sub.x.

TABLE-US-00001 The feeding amount of Fe and Ni source of all the synthesized catalysts with different compounds. Sampl es WC.sub.x WC.sub.x- Fe.sub.0.5Ni.sub.0 WC.sub.x- Fe.sub.0.5Ni.sub.0.2 .sub.5 WC.sub.x- Fe.sub.0.25Ni.sub.0. .sub.5 WC.sub.x- Fe.sub.0.5Ni.sub.0.5 WC.sub.x- Fe.sub.0.5N i.sub.1 WC.sub.x- Fe.sub.1Ni.sub.0.5 WC.sub.x- Fe.sub.0Ni.sub.0.5 Fe 0 0.50 mmol 0.50 mmol 0.25 mmol 0.50 mmol 0.50 mmol 1.00 mmol 0 Ni 0 0 0.25 mmol 0.50 mmol 0.50 mmol 1.00 mmol 0.50 mmol 0.50 mmol

[0074] The synthetic process is schematically illustrated in FIG. 1a. Dopamine molecules firstly coordinate with Fe and Ni ions, which then assemble with tungstate ions to form metal-organic compounds as yellow-green powders. SEM micrograph reveal a microspherical morphology, formed by aggregated nanosheets or flakes. Thermogravimetric analysis (TGA) under N.sub.2 flow shows that weight loss occurs at around 400° C. and 800° C., while from 900° C. no further weight loss can be observed. Furthermore, the powder X-Ray diffraction (XRD) depicts nearly no structural changes from 900° C. to 1000° C.

[0075] FIGS. 1b-c illustrates the structure module of the WCx supported FeNi single atoms. FIG. 1d shows detailed structural information in the k.sup.3-weighted FT spectra in R-space at the Ni K-edge (left) and the Fe K-edge (right) from the XAFS analysis.

[0076] FIG. 2a illustrates the structure of the catalysts as concluded from the conducted measurements (vide infra). First, the SEM images show that all carbonized materials maintain the microspherical morphology of the precursor. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images depict that the WC.sub.x catalysts consist of WC.sub.x nanocrystallites surrounded by carbon sheets. As single-atom doped carbon show insufficient catalytic activity and durability for the OER, the interest regarding possible active sites for OER was mainly focused on the well-crystallized WC.sub.x nanocrystallites (FIG. 2b). The atomic-resolution STEM images of WC.sub.x nanocrystallites observed along [110] and [111] crystallographic directions (FIGS. 2c, 2d) shows a bright W atom array with some dark atomic columns, which can be distinguished as Fe/Ni atoms on the surface. The line-profiles for the HAADF image (FIG. 2e) are taken across a dimmer atomic column along the (1̅21̅) plane (FIG. 2d right, arrow 1 to 5), which demonstrates the significant variation in atomic column intensity, proving the presence of Fe/Ni atoms with a random distribution on or near the surface of WC.sub.x.

[0077] The electrocatalytic OER activities of WCx-Ni and WC.sub.x-FeNi prepared at different temperatures was compared by linear sweep voltammetry (LSV) scan using a rotating disk electrode (RDE) at a scan rate of 10 mV s.sup.-1 with simultaneous iR correction (FIG. 3) confirming that the sample calcined at 900° C. shows the lowest overpotential.

[0078] Cycle voltammetry (CV) displayed in FIG. 4a shows the Ni.sup.0 to Ni.sup.2+, and Ni.sup.2+ to Ni.sup.3+/4+ redox features of WC.sub.x-Ni, confirming the presence of Ni.sup.0 single atoms in the catalyst. The addition of Fe changes the redox features of Ni, as the oxidation potential shifts to a higher voltage (from around 1.34 V vs. RHE to 1.41 V vs. RHE) in WC.sub.x-FeNi, indicating the synergetic effects between Fe and Ni atoms in electrochemical conditions. At the reduction cycle, the reduction peak is as well observed at higher potentials in WC.sub.x-FeNi than in WC.sub.x-Ni.

[0079] The OER polarization curves are compared in FIG. 4b, showing that WC.sub.x-Ni requires an overpotential of 275 mV at 10 mA cm.sup.-2 (see also Table 2), which is 136 mV lower than that of WC.sub.x-Fe. However, when Fe and Ni are introduced simultaneously, the WC.sub.x-FeNi requires an overpotential of only 237 mV at 10 mA cm.sup.-2, which is much lower than for WC.sub.x-Fe, WC.sub.x-Ni, and bare WC.sub.x, demonstrating that the high OER activity results from a synergistic effect of Fe and Ni sites. The Tafel slope to overpotential (FIG. 4c) and electrochemical impedance (not shown) give consistent results, in particular, the WC.sub.x-FeNi requires the lowest overpotential (237 mV) and Tafel slope (44 mV/dec) and shows the highest mass and electron transfer speeds. Besides, the required overpotential to reach 10 mA cm.sup.-2 can be further decreased from 237 mV to 201 mV, when the loading amount of WC.sub.x-FeNi increases from 0.38 mg/cm.sup.-2 to 1.14 mg/cm.sup.-2 (FIG. 5).

[0080] To investigate the intrinsic oxygen-evolving catalytic activity of WC.sub.x-supported single atoms, the turnover frequencies (TOFs) (FIG. 4d) and mass activities (FIG. 4e) are calculated. The WC.sub.x-FeNi exhibits an extremely high TOFs value of 2.18 s.sup.-1 per ½ FeNi atoms and mass activities of 14.5 A mg.sup.-1 at the overpotential of 300 mV. When increasing the applied overpotential to 320 mV, the value of TOFs and mass activity increased to 3.63 s.sup.-1 and 24.5 A mg.sup.-1, which are more than 5 times of the values for WC.sub.x-Ni and even 70 times higher than for WC.sub.x-Fe. WCx-FeNi also compares well to other state-of-the-art non-noble metal-based OER catalysts based on recent reports (Table 2).

TABLE-US-00002 Comparison of catalytic parameters of WC.sub.x supported single-atom catalysts and literature reported catalysts. All current densities are based on geometric area. Samples Overpotential at 10 mA/cm.sup.2 (mV) Tafel slop (mV/dec) TOF at overpotential of 300 mV (s.sup.-1) Catalysts loading (mg/cm.sup.2) Ref. no WCx 573 235 -- 0.38 This work WCx-Fe 411 98 0.04 0.38 This work WCx-Ni 275 59 0.36 0.38 This work WCx-FeNi 237 44 2.18 0.38 This work WCx-FeNi 201 56 1.10 1.14 This work G-FeCoW 223 0.46 0.21 Zhang, B. et al., Science 352, 333-337, (2016) A-FeCoW 301 0.17 0.21 Zhang, B. et al., Science 352, 333-337, (2016) LDH FeNi ~300 ~38 0.05 0.07 Song, F. & Hu, X., Nature Comm. 5, 4477, (2014) LDH NiCo ~330 ~40 0.01 0.07 Song, F. & Hu, X., Nature Comm 5, 4477, (2014) IrO.sub.2 260 0.01 0.21 Song, F. & Hu, X., Nature Comm 5, 4477, (2014) Ni-N.sub.4-C.sub.4 331 63 0.72 0.28 Fei, H. et al. Nature Catalysis 1, 63-72, (2018) W-Ni(OH).sub.2 237 33 0.74 0.20 Yan, J. Q. et al., Nature Comm 10, (2019) NiCo-UMOFNs 250 42 0.86 0.20 Zhao, S. et al. Nat Energy 1, 16184, (2016)

[0081] To reveal the kinetic activities of Fe and Ni, the feeding amount of Fe atoms were kept at 0.5 at.% and changed the Ni ratio from 0 to 1.0 at.% in relation to W (FIG. 4f, FIG. 5). The overpotential at 10 mA/cm.sup.2 decreases dramatically from 411 mV (WC.sub.x-Fe.sub.0.5Ni.sub.0) to 248 mV (WC.sub.x-Fe.sub.0.5Ni.sub.0.25). With a further increase of Ni in the system, the lowest overpotential reaches 237 mV (WC.sub.x-Fe.sub.0.5Ni.sub.0.5) and 236 mV (WC.sub.x-Fe.sub.0.5Ni.sub.1). If the Ni amount is fixed to 0.5 at.%, the overpotential decreases along with increased Fe ratio and reaches to the lowest overpotential of 237 mV at 0.5 at.% Fe ratio (WC.sub.x-Fe.sub.0.5Ni.sub.0.5), which then increases to 247 mV, when further increasing the Fe amount to 1.0 at.% (WC.sub.x-Fe.sub.1Ni.sub.0.5). This activity trend study allows us to optimize the catalytic performance of WC.sub.x-FeNi catalysts. Considering the better kinetics of WC.sub.x-Fe.sub.0.5Ni.sub.0.5 than WC.sub.x-Fe.sub.0.5Ni.sub.1, the following studies on durability, structure, and theoretical calculation will mainly focus on the WC.sub.x-Fe.sub.0.5Ni.sub.0.5 catalyst.

[0082] The operating durability is essential to assess the device application potential of OER catalysts, thus water oxidation tests were performed utilizing the optimized WC.sub.x-FeNi catalyst deposited on Ni foam under constant current of 10 mA cm.sup.-2, 20 mA cm.sup.-2, 50 mA cm.sup.-2, and 10 mA cm.sup.-2 continuously for 130 hours. No appreciable increase in overpotential was observed in this time interval (not shown). To gain insight into the thermal durability of WC.sub.x-FeNi, the material was subjected to a post thermal treatment in air (200° C.) or under Argon (900° C.). However, after these thermal treatments, the overpotentials show hardly any change at 10 mA cm.sup.-2, indicating the thermal durability of these oxygen-evolving catalytic single-atom sites on the WC.sub.x surface.

[0083] To gain insight into the electronic structure of WC.sub.x-FeNi, X-ray photoelectron spectroscopy (XPS) was applied, revealing the surface oxidation state. The high-resolution W 4f spectra show a decreased intensity of surface oxygenation peaks for WC.sub.x-Ni, WC.sub.x-Fe, and WC.sub.x-FeNi, when compared to pure WC.sub.x (not shown). The O content at the surface of pure WC.sub.x is 10.48 at.%, while it decreases to 4.93 at.% for WC.sub.x-Ni, 8.76 at.% for WC.sub.x-Fe, and 7.97 at.% for WC.sub.x-FeNi. This variation can be attributed to the different oxidation properties of Fe and Ni species. Ni 2p peaks of WC.sub.x-Ni show a distribution of both ionic Ni.sup.2+ (in the state of Ni-OH, and Ni-O) and metallic Ni.sup.0, with a calculated Ni.sup.2+: Ni.sup.0 peak area ratio of 0.8. After adding Fe, the Ni.sup.2+: Ni.sup.0 peak area ratio increases to 2.9, thus the presence of Fe yields a higher amount of oxidized Ni-species. The Fe 2p spectra of WC.sub.x-Fe and WC.sub.x-FeNi both show a mixture of Fe.sup.0, Fe.sup.2+, and Fe.sup.3+ valences.

[0084] X-ray absorption near-edge structure (XANES) spectra show a reduction of the white-line intensity (gray arrow) at the W L.sub.3 edge compared with commercial WC powder, thus corroborating the lower valence state of W in WC.sub.x-FeNi, which can be explained by the formation of a W.sub.2C and WC mixture. Fourier transforms (FTs) of the extended X-ray absorption fine structure (EXAFS) spectra of W reveal a similar WC structure with the commercial WC powder, however with lower W-W and W-C coordination numbers (.sup.aN), which confirms the high amount of surface defects and C atom vacancy on WC.sub.x.

[0085] XANES of WC.sub.x-FeNi at the Ni K-edge reveals an average valence between Ni.sup.0 and Ni.sup.2+, and the Fe K-edge reveals an average valence between Fe.sup.2+ and Fe.sup.3+. The FTs and wavelet-transform images from the EXAFS spectra in WC.sub.x-FeNi depict that the Fe coordination and valence environment are contributed partly by the Fe.sup.0 in Fe-W, and Fe.sup.2+/3+ ions in Fe-O/C, while Ni is contributed partly by Ni.sup.0 in Ni-W, and Ni.sup.2+ in Ni-O/C. The EXAFS fitting parameters at the Ni and Fe K-edge of WC.sub.x-FeNi show that the bond distances of Ni-W (3.43 and 3.60 Å) and Fe-W (3.52 Å) are much longer than the W-W bond (2.88 Å). Combined with the high mobility of Ni and Fe atoms under the electron beam, it is suggested that the atomically dispersed FeNi metal sites are weakly bond on the surface of the WC.sub.x nanocrystallite instead of inserted in the crystal or replacing W/C atoms.

B) Hydrogen Evolution Reaction (HER)

[0086] Synthesis of HER catalysts WC.sub.x-Rui, WC.sub.x-RuNi catalysts was conducted in analogy to the OER catalysts as described above.

[0087] FIG. 7 illustrates the HER activity of WCx-Ru2 in 1 M KOH obtained at different temperatures, whereby the best performance obtained at 800° C. The atomic content of Ru in the WCx-Ru2 catalyst is 0.49 at.%, which is weight percentage of 2.28 wt.%

[0088] Noble metals, such as Pt, Ru, and Pd, have considered to be ideal HER electrocatalysts. However, their activity in alkaline condition is much lower than that in acidic condition. To improve the activity of Ru catalyst for HER in alkaline condition, WC was used as a support, Ni atoms are investigated as mediator to adjust the H* adsorption, and the OH- desorption steps, therefore enhance the HER activity (FIG. 8).

[0089] As illustrated in FIG. 9 the mass activity of WCx-Ru catalyst is more than 50 times higher than Pt/C at the overpotential at 100 mV and the mass activity of WCx-Ru2Ni2-800 catalyst is more than 50 times higher than Pt/C at the overpotential at 95 mV.

[0090] Besides Ni further metals such as Cu, Co, Mn, Ni, V, Ca were tested in a WC-RuX bi-atomic system for H.sub.2 evolution reaction. In particular WC-Ru system comprising Mn, V, Ca showed good results (see FIG. 10).

Chemical Reagents and Materials

[0091] Na.sub.2WO.sub.4.Math.2H.sub.2O, dopamine hydrochloride, HCl, FeCl.sub.3.Math.6H.sub.2O, NiCl.sub.2.Math.6H.sub.2O, and KOH were obtained from Alfa Aesar. Nafion 117 solution was obtained from sigma aldrich. Unless otherwise stated, all the reagents were of analytical grade and were used as received. All aqueous solutions were prepared with DI water produced from Millipore purification system.

Characterizations and Methods

Scanning Electron Microscopy (SEM)

[0092] The morphology of the prepared precursors and final materials were observed by high-resolution FE-SEM (Hitachi S-4000 and S-4800). All the carbonized materials were observed directly without gold coating. For the non-conductive precursors, the gold was deposited with a layer of about 1 nm.

Scanning Transmission Electron Microscopy (STEM)

[0093] For STEM characterization, the NPs were drop-casted on Lacey carbon-coated Cu grids. STEM investigations were performed using a JEOL JEM-ARM 200F scanning transmission electron microscope equipped with a cold field emission electron source, a DCOR probe corrector (CEOS GmbH), a 100 mm.sup.2 JEOL Centurio EDX detector, and a Gatan GIF Quantum ERS electron energy-loss spectrometer. The microscope was operated at 200 kV, a semi-convergence angle of 20.4 mrad, resulting in a probe size of 0.8 Å (1 Å for the EELS and EDXS analyses). 83 - 205 mrad collection angles were used to obtain the HAADF images. A collection semi-angle of 83 mrad was used for EELS measurements. The pixel dwell time for the spectrum imaging was set to 0.01 seconds; EDX and EELS signals were recorded simultaneously. The HAADF-STEM image simulations were performed using the QSTEM image simulation software [C. T. Koch, ProQuest Dissertations and Theses, Arizona State University 2002].

X-Ray Photoelectron Spectra (XPS)

[0094] XPS was measured on K-Alpha™ + X-ray Photoelectron Spectrometer System (Thermo Scientific) with Hemispheric 180° dual-focus analyzer with 128-channel detector. X-ray monochromator is Micro focused Al-Kα radiation. For the measurement, the prepared powder samples were pressed and loaded on carbon taps, then pasted onto the sample holder for measurement. The data was collected with X-ray spot size of 400 .Math.m, 20 scans for the survey, and 50 scans for the specific regions.

X-Ray Diffraction (XRD)

[0095] Powder XRD of all the prepared samples were measured with the same condition on a Bruker D8 Advance instrument with Cu Kα radiation (λ=1.54 Å) at a generator voltage of 40 kV and a generator current of 50 mA.

Thermogravimetric Analysis (TGA)

[0096] TGA was conducted in N.sub.2 conditions from room temperature to 1000° C., with a rapid of 10° C./min.

X-Ray Absorption Spectra (XAS)

[0097] The Tungsten L.sub.III-edges XAFS spectra of the standards and samples were collected at the beamline BL1W1B of the Beijing Synchrotron Radiation Facility (BSRF). The typical energy of the storage ring was 1.5 GeV, and the electron current was 180 mA in the top-up mode. The white light was monochromatized by a Si (111) double-crystal monochromator and calibrated with a W foil (L.sub.III edge 10207 eV). Samples were pressed into thin slices and positioned at 90° to the incident beam in the sample-holder. The XAFS spectra were recorded in transmission mode with two ion chambers.

[0098] The XAFS data were analyzed using the software packages Demeter (Ravel and Newville, 2005, Ravel, B., Newville, M., 2005. ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray absorption spectroscopy using IFEFFIT. Journal of Synchrotron Radiation 12, 537-541). The spectra were normalized using Athena firstly, and then shell fittings were performed with Artemis. The .sub.χ(k) function was Fourier transformed (FT) using k.sup.3 weighting, and all fittings were done in R-space. The amplitude reduction factor (S.sub.0.sup.2) was estimated to be 0.827 according to the fitting results of the W foil. The coordination parameters of sorption samples were obtained by fitting the experimental peaks with theoretical amplitude.

[0099] To further investigate the first-shell backscattering atoms and detect light and heavy scatters, wavelet transform (WT) analysis was employed using the Igor pro script developed by Funke et al. (Funke et al., 2005, Funke, H., Scheinost, A.C., Chukalina, M., 2005. Wavelet analysis of extended X-ray absorption fine structure data. Physical Review B 71). This qualitative analysis was primarily focused on the nature of the backscattering atoms as well as the bond lengths owing to the fine resolution in both wavenumbers k and radial distribution function R, and complemented the limitation of FT analysis. A Morlet wavelet was chosen as basis mother wavelet, and the parameters (η = 8, σ = 1) were used for a better resolution in the wave vector k.

Electrochemical Measurements

Ink Preparation

[0100] The catalyst ink was prepared by blending the catalyst powder (15 mg) with 100 .Math.L Nafion solution (5 wt. %) and 900 .Math.L ethanol in an ultrasonic bath. 5 .Math.L of catalyst ink was then pipetted onto the GC surface, leading to a catalyst loading of 0.38 mg/cm.sup.2. The TOFs data was calculated based on the atomic content (from XPS) of the Fe and/or Ni in the catalysts.

Electrodes and Measurements

[0101] All the electrochemical measurements were carried out in a conventional three-electrode cell using the Gamry reference 600 workstations (Gamry, USA) at room temperature. A commercial RHE electrode was used as the reference electrode, and the graphite rod was used as counter electrode. The Ag/AgCl reference electrode calibrated with RHE in 1 M KOH was used as reference electrode for long-time stability measurement. A glassy carbon (GC) RDE electrode with an area of 0.196 cm.sup.2 served as the substrate for the working electrode to evaluate the OER activities of various catalysts. The electrochemical experiments were conducted in N.sub.2 saturated 1 M KOH electrolyte. The CV and OER polarization curves of different catalysts with real-time iR corrected by Gamry reference 600 potentiostats at a resistance of 4.4 ohms. The RDE measurements were conducted at a rotating speed of 1600 rpm with a sweep rate of 10 mV/s.

[0102] Electrochemical impedance spectroscopy (EIS) was carried out with a potentiostatic EIS method with a DC voltage of 1.5 vs. RHE, in an N.sub.2 saturated 1.0 M KOH electrolyte from 100 kHz to 0.1 Hz with a 10 mV AC potential at 1600 rpm. The stability tests for the catalysts were conducted using chronopotentiometry at the constant working current densities of 10 mA/cm.sup.2, 20 mA/cm.sup.2, 50 mA/cm.sup.2, and 10 mA/cm.sup.2.

[0103] To investigate the intrinsic oxygen-evolving catalytic activity of WC.sub.x supported single atoms, the turnover frequencies (TOFs) and mass activities are calculated by X-ray photoelectron spectroscopy (XPS) data. To reveal the activity of each Fe and Ni active site, the TOFs of WC.sub.x-Fe, WC.sub.x-Ni, and WC.sub.x-FeNi are calculated based on the amount of Fe, Ni, and ½ FeNi atoms in the materials. It was confirmed that the WC.sub.x support materials contribute very low activity for OER at the overpotential higher than 550 mV, which is meaningless to discuss its performance.