MaterialsNickel acetate (Ni(CH3COO)2·6H2O), ammonium molybdate ((NH4)6Mo7O24), urea, sodium copper chlorophyllin, Vulcan XC-72R (Cabot Co.), ultrapure water (18.20 MΩ cm), potassium hydroxide (KOH), isopropanol (C3H8O) and Nafion solution (5 wt%) were used.Synthesis of Ni@C-MoOx
A typical method for the synthesis of Ni@C-MoOx included 1 g Ni(CH3COO)2, 0.353 g (NH4)6Mo7O24, 1 g urea and 0.1 g sodium copper chlorophyllin being added in 35 ml deionized water. After stirring vigorously for 0.5 h, the solution was transferred to a 50 ml Teflon-lined stainless-steel autoclave, then heated to 180 °C and maintained for 12 h. The autoclave was then cooled to room temperature. The obtained product was washed with deionized water and ethanol to remove ionic residue and then dried in an oven at 60 °C overnight. The resulting powder was heated in a tube furnace up to 500 °C at a ramp rate of 5 °C min−1 under H2/N2 (with 1:9 mole ratio) gas flow, and the heating was maintained at 500 °C for 2 h. Afterwards the furnace was cooled to room temperature under N2 flow, the product Ni@C-MoOx was collected.Synthesis of NiMoOx
The NiMoOx was synthesized in the same way as Ni@C-MoOx except replacing sodium copper chlorophyllin with carbon black (Vulcan XC-72R).Synthesis of Ni/CuChlThe Ni/CuChl was synthesized in the same way as Ni@C-MoOx except without adding ammonium molybdate ((NH4)6Mo7O24).Synthesis of Ni@C-MoOx-N2
The Ni@C-MoOx-N2 was synthesized in the same way as Ni@C-MoOx except annealed in N2 instead of H2.Synthesis of Ni@C-MoOx-TPThe Ni@C-MoOx-TP was synthesized in the same way as Ni@C-MoOx except using tetraphenyl porphyrin (TP) instead of CuChl.Synthesis of MoOx
One g (NH4)6Mo7O24 was heated in a tube furnace up to 500 °C at a ramp rate of 5 °C min−1 under H2/N2 (with 1:9 mole ratio) gas flow, and the heating was maintained at 500 °C for 2 h. Afterwards the furnace was cooled to room temperature under N2 flow, and the product MoOx was collected.Synthesis of C-MoOx
A 0.353 g amount of (NH4)6Mo7O24 and 0.1 g sodium copper chlorophyllin were added in 35 ml deionized water; after drying at 65 °C, the powder was heated in a tube furnace up to 500 °C at a ramp rate of 5 °C min−1 under H2/N2 (with 1:9 mole ratio) gas flow, and the heating was maintained at 500 °C for 2 h. Afterwards the furnace was cooled to room temperature under N2 flow, the product C-MoOx was collected.Electrochemical measurementsAll electrochemical experiments were performed in a standard three-electrode cell at room temperature. The cell consists of a glassy carbon (GC) working electrode (a rotating disc electrode with 5 mm in diameter, PINE: AFE3T050GC), a Hg/HgO (in saturated KCl) reference electrode and a carbon rod counter electrode. All potentials in this study are given relative to a reversible hydrogen electrode (RHE). The working electrodes were prepared by applying catalyst ink onto GC disk electrodes or carbon paper (1 × 3 cm2, HCP120, HESEN Inc.) electrodes. In brief, the electrocatalyst was dispersed in ethanol with Nafion solution (5 wt% in isopropyl alcohol) and ultrasonicated for 15 min to form a uniform catalyst ink. The well-dispersed catalyst ink was then applied onto a pre-polished GC disc to the designed catalyst loading. The total catalyst loading on GC was 0.5 mg cm−2 for all Ni-based catalysts, respectively. The catalyst-modified gas diffusion electrode (GDE) was prepared by using carbon paper as electrode substrate. A 1 × 1 cm2 catalyst-coated active area was formed by pipetting the catalyst ink onto the carbon paper and drying at 40 °C. All the electrodes were pretreated by cycling the potential between 0 and 0.3 V at a sweep rate of 10 mV s−1 for 30 cycles to remove any surface contamination before the hydrogen oxidation reaction (HOR) testing. The HOR measurements were conducted in 0.1 M KOH electrolyte, which was saturated with H2 gas by continuous purging. The HOR polarization curves were collected on the catalyst-coated GC rotation disc electrode at a controlled rotation speed of 2,500 rpm or 1,600 rpm and a potential scan rate of 5 mV s−1.Calculation of exchange current densityExchange current density (j0) can be deduced from the Butler–Volmer equation:$${j}_{k}={j}_{0}\left[\mathrm{e}^{\displaystyle\frac{{\rm{\alpha }}{{F}}}{{RT}}\eta }-\mathrm{e}^{\displaystyle\frac{(1-{\rm{\alpha }}){{F}}}{{RT}}\eta }\right]$$where α is the charge transfer coefficient, R is the universal gas constant (8.314 J mol−1 K−1), T is the operating temperature (303 K in this work), F is Faraday’s constant (96,485 C mol−1) and η is the overpotential, respectively. In a small potential window of the micro-polarization region near the equilibrium potential (±10 mV), jk approximately equals to j. In this case, the Butler–Volmer equation can be expanded by Taylor’s formula and simplified as:$${j}_{k}=\frac{F}{{RT}}\frac{{j}_{0}}{\eta }$$By linearly fitting the polarization curve in the micro-polarization region, the j0 for Ni-based catalysts can be obtained.Acidic etching experimentsCyclic voltammetry with a potential scan rate of 5 mV s−1 at the potential range of 0–0.3 VRHE in 0.1 M HClO4 was employed for several cycles until the cyclic voltammetry curve was stable, indicating all the metallic Ni in the sample had been dissolved.Electrical resistance measurementsThe electrical resistances were determined by using a homemade button cell47. The sample was inserted between two polished smooth steel discs. A Solartron SI 1287 electrochemical interface equipped with a Solartron SI 1260 impedance/gain-phase analyser coupling system was used. The operating frequency range was between 0.1 Hz and 10 kHz, the d.c. potential was 0 V compared to an open circuit, and the a.c. amplitude was 10 mV. In this case, the phase angle between the voltage applied and the current induced is zero; the impedance of the sample as a function of frequency is present as a horizontal line. The value of resistance of sample is equal to the impedance; and the resistance can be directly read from the |Z |-axis in the Bode.Membrane electrode assembly measurementsThe catalyst ink was prepared by adding the catalyst, Vulcan XC-72 carbon (as conductive agent) and ionomer to isopropanol as solvent, followed by sonication for 0.5 h. The weight ratio of Ni@C-MoOx catalyst and the Vulcan XC-72 carbon was 5:4, and the weight ratio of the QPCBP-10 ionomer was 0.20 and 0.22 of the total mass for the anode and cathode, respectively. Next, the ink was sprayed onto the polymer (carbazolyl aryl piperidinium)48 membrane by airbrush to produce a gas diffusion electrode of 1 cm2 for the anode. The final catalyst loading was 1.0 mgNi cm−2 for anode and 0.3 mgPt cm−2 for cathode. After drying at 40 °C, the membrane electrode assembly (MEA) was immersed in H2-saturated 1 M KOH aqueous solution for 12 h for ion exchange. The MEA was assembled into the fuel cell with a fluorinated ethylene propylene gasket, two pieces of carbon paper as the gas diffusion layer, a graphite bipolar plate with 1 cm2 flow field and a gold-coated current collector on each side to complete the anion-exchange-membrane fuel cell. A fuel cell test station (Scribner 850e) with back-pressure regulators was used to measure the polarization curves and stability under H2/O2 or H2/air feeding conditions.CharacterizationsX-ray diffraction patterns were collected with a Rigaku D/MaXIIIA for Cu Kα (λ = 1.540598 Å) radiation at room temperature with a scanning speed of 10° min−1. The in situ Fourier transform infrared absorption spectra were recorded on IRTracer100 with Autolab302N. The UV–vis spectra were collected on UV-3600. The morphology and microstructures of all the catalysts were characterized by field-emission SEM (JEOL JEM-2100). TEM and HRTEM images were recorded on a JEOL JEM-2100F electron microscope equipped with a high-brightness field-emission gun and an energy-dispersive X-ray spectroscopy analyser. The ADF-STEM and EELS were recorded on Spectra 300 electron microscope. XPS spectra were recorded on a Thermo Scientific K-Alpha spectrometer equipped with a monochromatic Al X-ray source (Al KR, 1.4866 keV), charge calibration for all high-resolution spectra were carried out with the standard C 1s peak of 284.80 eV. In situ XPS measurements were conducted under similar conditions; the catalysts are directly reduced under the H2 atmosphere in the chamber of XPS without contacting any air or oxygen. XPS data were fitted using Avantage software, peak fit parameters: peak background is smart, maximum iterations is 100, convergence is 0.0001, fitting algorithm is Powell, Gauss–Lorents mix is product.UPS measurements were performed in a Thermo ESCALAB 250XI PHI5000 Versa Probe III with a He Iα UV source (21.20 eV), applying a −5.00 V bias to the sample. Work function (Φ) was calculated as the energetic difference between the Fermi level and the vacuum level, as defined by the secondary-electron edge, which is based on the Einstein’s equations for photoelectricity:$$\Phi ={\rm{h}}\upnu -{{{W}}}_{0}$$The hν is the energy of He1α, the W0 is the energy of the cut-off edge and Φ is the calculated work function. The ionization energy is estimated as the onset of the density of states at the top of the valence band. The W0 for Ni@C-MoOx and NiMoOx are 16.27 eV and 17.38 eV, so that the calculated Φ are 4.93 eV and 3.82 eV, respectively49.The X-ray absorption spectra of Ni, Mo and C were collected at the Singapore Synchrotron Light Source centre in the transmission mode, where a pair of channel-cut Si (111) crystals were used in the monochromator. The X-ray absorption fine structure (XAFS) spectroscopy in the stability test for Ni was carried out using the RapidXAFS 2 M (Anhui Absorption Spectroscopy Analysis Instrument Co.) by transmission mode at 20 kV and 30 mA, and the Si (551) spherically bent crystal analyser with a radius of curvature of 500 mm was used for Ni. The first inflection point (the first peak in the first derivative of XANES) of its XANES was defined to be E0 as the reference value (20.00 keV for Mo and 8.33 keV for Ni). XAFS data were analysed with ATHENA and ARTEMIS, linear function was subtracted from the pre-edge region, then the edge jump was normalized using Athena software. The absorption curves were normalized to 1, and the EXAFS signals χ(k) were obtained after the removal of the pre-edge and post-edge background. The k3 weighted χ(k) data were Fourier transformed after applying a Hanning window function in the range of 3–12 Å−1. The models of MoO3 and Mo2C were used to calculate the scattering paths of Mo; the model of Ni was used to calculate the scattering path of Ni in Ni@C-MoOx. First, feeff paths were obtained by calculating the feeff input files through FEEFF package in ARTEMIS; secondly, the EXAFS fitting results were got by setting different structure parameter until the fitting data highly consistent with experimental data50. Hama Fortran.exe software was used to calculate the wavelet transform of EXAFS spectra; the Morlet wavelet with finite length is used as the fundamental wave51.$$\varphi \left(t\right)=\frac{1}{\sqrt{2\pi }\sigma }\left[\exp \left({ikt}\right)-\exp \left(-\frac{{k}^{2}}{2}\right)\right]\exp \left(-\frac{{t}^{2}}{{2\sigma }^{2}}\right)$$k and σ are the programme parameters kappaMorlet and sigmaMorlet.The model function is a sum of two sinuses modulated by Gaussian.$$\mathrm{signal}=\mathrm{part}\,1+\mathrm{part}\,2$$$$\begin{array}{l}\mathrm{part}\,1=\mathrm{ampli}\,1\times\frac{1}{\sqrt{2\pi {\sigma}1}}{\exp }^{-\displaystyle\frac{{(k-{\mathrm{centum}}\,1)}^{2}}{{{\sigma}1}^{2}}}\\\qquad\qquad\times\sin (\mathrm{frequency}\,1\times[k-\mathrm{phase}\,1])\end{array}$$$$\begin{array}{l}\mathrm{part}\,1=\mathrm{ampli}\,2\times\frac{1}{\sqrt{2\pi {\sigma}2}}{\exp }^{-\displaystyle\frac{{\left(k-\mathrm{centum}2\right)}^{2}}{{{\sigma}2}^{2}}}\\\qquad\qquad\times\sin (\mathrm{frequency}\,2\times[k-\mathrm{phase}\,2])\end{array}$$The input parameters: Mother wavelet function-Morlet function, Kappa Morlet = 1, Sigma Morlet = 10.The equation used to model and interpret EXAFS is:$$\chi (k)=\sum\limits_{j}\frac{{N}_{j}{S}_{0}^{2}\;{f}_{j}(k){\mathrm{e}}^{-2{R}_{j}/\lambda (k)}}{k{R}_{j}^{2}}\sin \left[2k{R}_{j}+{\delta }_{j}(k)\right]$$Where the sum could be over shells of atoms or over scattering paths for the photoelectron, the f(k) and δ(k) are photoelectron scattering properties of the neighbouring atom (and λ(k) is the photoelectron mean-free-path).In situ XRD testIn situ electrochemical measurements were conducted in a polytetrafluoroethylene electrochemical cell containing Ni@C-MoOx-coated carbon paper as the working electrode, carbon rod as the counter electrode and Hg/HgO electrode as the reference electrode. The electrochemical cell was assembled, and 0.1 M KOH solution was used as the electrolyte. In situ XRD experiments were performed using a laboratory X-ray powder diffractometer (EX-CalibueR, with Cu Kα1 radiation) at room temperature with a scanning speed of 2° min−1. XRD characterizations of Ni@C-MoOx were tested during step-potential test from 0.1 to 0.7 V (vs RHE) at 1 h 0.1 V−1.ICP-OES testI CAP 6300 Duo instrument of Thermo Fisher Technology Co. was used for the inductively coupled plasma-optical emission spectrocopy (ICP-OES) testing to accurately analyse the actual content of metal elements in the sample. The catalyst was placed in a clean corundum crucible and heated to 900 °C min−1 at 5 °C min−1 in Muffle furnace, maintained for 2 h. After cooling to room temperature, the crucible was placed on a 180 °C furnace and boiled in aqua regia to dissolve the metal/metal oxide obtained after high temperature calcination. Then 50-ml ultrapure water was used to prepare a clear and transparent solution with a certain concentration. Before the ICP-OES test, the standard solution with the configured concentration gradient was tested to obtain a standard curve (correlation coefficient > 0.999) and then compared the results of the sample to determine the actual content of metal elements.Computational detailsAll the periodic models of density functional theory (DFT) in this work were calculated with Vienna Ab-initio Simulation Package code52. The electron-exchange correlation can be described by the Perdew–Burke–Ernzerhof function with generalized gradient approximation53, and the electron-ion correlation energy was determined by the projector-augmented-wave pseudopotential method54. The cut-off energy of the plane wave basis set was selected at 500 eV, and (3 × 1 × 1) Monkhorst–Pack k-point was set for the k-space integration. The calculation exits after the energy and the force have converged to 10−5 eV and 0.02 eV Å−1 with dipole correction and spin polarization. The implicit solvation model VASPsol was used to simulate the effect of water solvent on species adsorption55,56.The MoOx/Ni(111) model was constructed by a stick model of a MoOx(001) nanorod supported on four layers of Ni(111) (3 × 8) unit cell, and the MoOC/Ni(111) model was loaded with MoOC(002) on four layers of Ni(111) (4 × 8) unit cell. A vacuum layer of 15 Å was placed in the z axis direction of the model to eliminate the interaction between the periodic images. The bottom two layers of Ni were fixed during DFT calculations. The adsorption-free energy of intermediate species was obtained by$$\Delta {{G}}=\Delta {{E}}+\Delta {\rm{ZPE}}{\rm{{-}}}{{T}}\Delta {{S}}$$
(1)
where ∆G, ∆E, ∆ZPE and ∆S are the free energy change, the adsorption energy, the zero-point energy change and the entropy change, respectively, and T denotes the temperature of 298.15 K. The adsorption energy was calculated by$${\Delta {{E}}}_{\ast {\rm{ads}}}={{{E}}}_{\ast {\rm{ads}}}-{{{E}}}_{\ast }-{{{E}}}_{{\rm{ads}}}$$
(2)
where E*ads, E* and Eads are the energies of the surface with adsorbates, clean surface and adsorbates. The reference state of H* was derived from isolated H2 according to EH = ½ EH2 (ref. 57).The Bader analysis used in this work is a common method that defines the electron density zero flux plane as the interface between atoms. The independent space of each atom is the atomic basin, and the Virial theorem is satisfied in each atomic basin. Therefore, the atomic charge can be obtained by:$${q}_\mathrm{A}={Z}_\mathrm{A}-{\int_{\varOmega}^{1}}\rho (r){\rm{d}}r$$where Ω is obtained by integrating the electron density in the atomic pot and calculating the difference with the nuclear charge. A stands for atomic pot58.
Quantum confinement-induced anti-electrooxidation of metallic nickel electrocatalysts for hydrogen oxidation
