Materials and reagentsThe nickel foam (NF) (0.58 g/cm2) was purchased from Kunshan Dessco Electronics Co. Ltd. (Kunshan, China). The reagents hydrochloric acid (HCl, 38.0%), ethanol (C2H5OH, 99.7%), cobalt chloride (CoCl2·6H2O, 99.5%), ammonium chloride (NH4Cl, 99.0%), Ruthenium(III) chloride (RuCl3·xH2O, 37% Ru basis), potassium nitrate (KNO3, 90%) and potassium hydroxide (KOH, 90%) were purchased from Shanghai Macklin Biochemical Co., Ltd. All reagents were of analytical purity and used without further purification.Preparation of G-RuCoThe preparation of G-RuCo on Ni foam was based on a combined method of electrodeposition, cation exchange, and electrochemical reduction. A 0.5 × 0.5 cm2 piece of Ni foam was washed sequentially in ethanol, 0.1 M HCl, and deionized water using an ultrasonic bath to remove surface oxides. Co(OH)2/Co nanosheets were first prepared via an electrodeposition process under a -3 A/cm2 current density for 120 s in a three-electrode system consisting of a graphite rod as the counter electrode, Ag/AgCl electrode (saturated KCl solution) as the reference electrode, and Ni foam as the working electrode. The electrodeposition solution was an aqueous mixture of 0.12 M cobalt chloride, 1.5 M ammonium chloride, and 100 mL deionized water. The obtained Co(OH)2/Co nanosheets were washed with deionized water, and then cation exchange was conducted by soaking in 30 mM RuCl3 solution under ambient conditions for 40 h. The resultant catalyst was first dried at 70 °C for 1 h and then annealed in an oven at 240 °C for 3 h to convert it into Ru-Co3O4/Co. Finally, an in situ electrochemical prereduction step was performed using the chronopotentiometry method at −800 mA/cm2 for 1 h to obtain the final G-RuCo catalyst.Preparation of CoThe as-prepared Co(OH)2/Co nanosheets on Ni foam described above were directly annealed in an oven at 240 °C for 3 h to convert them into Co3O4/Co nanosheets without cation exchange. The Co catalyst was finally obtained after electrochemical reduction using the chronopotentiometry method at −800 mA/cm2 for 1 h.Preparation of RuIn the electrolytic cell, an electrolyte of RuCl3 (60 mL, 2 mg/mL), a working electrode of 0.5 × 0.5 cm2 Ni foam, and a counter electrode of a Pt plate were used. Electrodeposition was performed for 15 min at −30 mA/cm2. Subsequently, calcination was carried out at 700 °C for 2 h in a tubular furnace with 10% H2/Ar mixed gas and a heating rate of 10 °C/min to further improve the crystallinity of the Ru product.Preparation of NG-RuCoIn the electrolytic cell, an electrolyte of RuCl3 (60 mL, 2 mg/mL) and CoCl2·6H2O (23 mg/mL), a working electrode of as-prepared Co on Ni foam, and a counter electrode of a graphite rod were used. Electrodeposition was performed for 15 min at −50 mA/cm2. Finally, an in situ electrochemical prereduction step was performed using the chronopotentiometry method at −800 mA/cm2 for 1 h to obtain the final NG-RuCo on Ni foam catalyst.CharacterizationsThe morphology and elemental composition of catalysts were analyzed using a scanning electron microscope (SEM, ZEISS Sigma) equipped with an energy-dispersive X-ray spectrometer at a working voltage of 15 kV. The lattice arrangement of the catalyst was characterized by transmission electron microscopy (TEM, JEM-2100F, Japan) at an operating voltage of 200 kV; the tested catalyst comes from powder samples scraped off the nickel substrate. The crystal structure of the catalyst was analyzed using an X-ray diffractometer (XRD, Rigaku, Japan) with a Cu-Kα X-ray source (λ = 1.5418 Å). TOF-SIMS (PHI Nano TOF II Time-of-Flight SIMS) also was applied to investigated gradient elements distributions of G-RuCo. The sputter etching was performed using an Ar+ beam (3 kV 100 nA) to obtain a depth profile. The surface valence state and chemical composition of the catalyst were studied via X-ray photoelectron spectroscopy (XPS; Thermo ESCALAB 250) using monochromatic Al-Kα radiation (1486.6 eV). All XPS spectra were calibrated by shifting the detected carbon C 1 s peak to 284.8 eV. XAFS experiments were performed at the 1W1B beamline of the Shanghai Synchrotron Radiation Facility. The XAFS spectra were analyzed with the Athena software package. The k-weighting was set to 2 for the Fourier transforms.Electrochemical measurementsAll electrochemical tests were performed under environmental conditions using a three-electrode system, and the results were recorded by an electrochemical workstation (CS310, Wuhan Kesite) in a customized H-type cell with an anion exchange membrane (separated by a Nafion 117 membrane; magnetic stirring at 1500 rpm). A figure of the experimental set-up was provided in the Supplementary Fig. 45. Unless otherwise specified, G-RuCo on Ni foam (0.5 × 0.5 cm2, the loading of the catalyst is about 8 mg/cm2) catalyst was typically employed as the working electrode, with platinum wire and a Hg/HgO electrode (filled with 1.0 M KOH solution) serving as the counter electrode and reference electrode, respectively. And the hydrogen reversible reaction was used to calibrate the reference electrode. In addition, a solution of 2000 ppm NO3− in 1.0 M KOH was used as the electrolyte, with the initial electrolyte volume set at 30 mL for the H-cell measurements. The electrolyte solution was bubbled with Ar gas for 10 min before the experiment to remove O2 and N2. Before testing, all catalysts were first electrochemically reduced at −0.2 V vs. RHE for 600 s in a 1.0 M KOH solution to eliminate surface oxidation. Electrochemical NRA measurements were performed using linear sweep voltammetry polarization curves via the potential dynamic method at a scanning rate of 1 mV/s in 1.0 M KOH electrolyte (the pH value was 13.7). All potentials were calibrated to the RHE by the equation:$${{\rm{E}}}({{\rm{V}}} \, {vs}. \, {{\rm{RHE}}})={{\rm{E}}} \, ({{\rm{V}}} \, {vs}.\,{{\rm{Hg}}}/{{\rm{HgO}}}) \,+\, 0.0591\times {{\rm{pH}}}+0.098$$
(4)
All measured potentials were 50% iR-compensated by the solution resistance, unless otherwise specified. Electrochemical impedance spectroscopy (EIS) was performed in the frequency range of 0.1 Hz–200 kHz with the amplitude of 10 mV at the overpotentials of −60 mV vs. RHE under the NRA operating condition. Rs is related to the solution resistance. Rct denotes the charge transfer resistance. Long-term stability was examined through chronopotentiometry tests at −300 and −1000 mA/cm2 in a flow-system H-cell with a 30 mL/min electrolyte flow rate.ECSA analysisFor the electrochemical active surface area (ECSA), we used the double-layer capacitance method in an electrolyte of 1.0 M KOH in the non-Faradaic potential range with different scanning rates of 10, 20, 30, 40, 50 and 60 mV/s. The ECSA of the working electrodes was calculated according to the following equations:$${I}_{C}=\upsilon {C}_{{dl}}$$
(5)
$${{\rm{ECSA}}}=\frac{{C}_{{dl}}}{{C}_{s}}$$
(6)
where Ic represents the charging current at different scan rates, ν is the scan rate, Cdl is the double-layer capacitance, and Cs is the specific capacitance for a flat metallic surface, which is generally in the range of 20–60 μF/cm2 (we assume a value of 40 μF/cm2 here)37,38.FE, yield rate and current density determinationThe FE of NRA for NH3 and NO2− was calculated according to:$${{\rm{FE}}}=(n\times V\times c\times F)/Q$$
(7)
where n is the electron-transfer number (8 for 1 mol NH3, 2 for 1 mol NO2−), V is the volume of the catholyte of the cathode chamber (30 mL), c represents the concentration of the outlet products (M), F is the Faraday constant (96,485 C/mol), and Q represents the applied overall coulomb quantity (C).The yield rate and current density of NH3 were calculated according to the following equations:$${{{\rm{Y}}}}_{{{{\rm{NH}}}}_{3}}=(c\times V) \, / \, (S\times t)$$
(8)
$${{{\rm{j}}}}_{{{{\rm{NH}}}}_{3}}=(Q\times {{{\rm{FE}}}}_{{{{\rm{NH}}}}_{3}}) \, / \, (S\times t)$$
(9)
where S is the area of the geometrical cathode and t is the electrolysis time.Ammonia calculationThe concentration of NH3 was spectrophotometrically determined using the indophenol blue method56,57. First, 2 mL of the diluted electrolyte solution was mixed with 2 mL of chromogenic agent (a mixture of 1.0 M KOH solution, 0.36 M salicylic acid, and 0.18 M sodium citrate). Then, 100 μL of 0.05 M NaClO solution (containing 4.00–4.99% effective chlorine) was added, followed by 0.2 mL of 0.034 M (1 wt.%) sodium nitrite ferrocyanide solution (stored at 4 °C) to initiate the color reaction. After allowing the mixture to stand at room temperature for 1 h, the absorbance spectrum was measured using a UV-vis spectrophotometer, and the formation of indophenol blue was determined at a wavelength of 655 nm. A standard concentration–absorbance calibration curve was prepared in advance using a range of NH4Cl (≥99.5%) solutions; the concentration of the NH3 product was then calculated based on the measured absorbance and standard curve.Nitrite detectionThe nitrite concentration was detected using UV‒vis spectrophotometry10. Initially, the collected electrolyte was diluted to the detection range. Next, 1 mL of 1.0 M HCl was added to 5 ml of diluted electrolyte, followed by adding 0.1 mL of a NO2−-specific color developing agent (a mixed solution of 0.2 g N-(1-naphthyl) ethylenediamine hydrochloride, 4.0 g sulfanilamide and 10 mL phosphoric acid (85 wt.% in H2O) in 50 ml deionized water). The absorbance intensity at a 540 nm wavelength was tested using UV‒vis spectrophotometry after allowing the resultant solution to react for 20 min at room temperature. A concentration–absorbance calibration curve was obtained by linear fitting of a series of standard potassium nitrite solutions, and the nitrite concentration was calculated based on the measured absorbance and standard curve.Nitrate detectionThe nitrate concentration was detected using UV‒vis spectrophotometry58. First, 4 ml of diluted electrolyte was mixed with 1 mL of 1.0 M HCl and 0.1 mL of sulfamic acid (0.8 wt.%) to form a mixed solution. Following a 20-minute reaction at room temperature, the absorption intensities at 220 nm and 275 nm wavelengths were recorded using UV‒vis spectrophotometry. The final absorbance (A) was calculated with the following equation: A = A220nm−A275nm. A concentration–absorbance calibration curve was established by linear fitting of a series of standard potassium nitrate solutions, and the nitrate concentration was calculated based on the measured absorbance and standard calibration curve.Determination of ammonia by 1H NMRTo detect the FE of 14NH4+ after 1 h of electrolysis at −0.1 V (vs. RHE) in 2000 ppm K14NO3, a calibration curve of 1H NMR (400 MHz) measurements was constructed using a series of 14NH4Cl standard solutions with specified concentrations (0, 10, 20, 30, and 40 mM). Subsequently, 0.5 mL of electrolyte, mixed with 15 mM maleic acid, 50 μl of 4 M H2SO4, and 50 μL DMSO-d6, was sealed into an NMR tube for 1H NMR. Next, 2000 ppm K15NO3 was used to qualitatively determine the source of NH3. Electrolysis is performed for 1 h at −0.1 V (vs. RHE), and 15NH4+ in the electrolyte is detected with 1H NMR59.H2 detectionFor gaseous products (HER, OER), the FE has been monitored by measuring the volume of gas collected from the working electrode in an inverted burette or graduated cylinder60. The H2-FE measurement in the work was based on the water drainage method at different potentials for 1 h for different catalysts. Firstly, connect the 50 ml inverted burette to the electrolyte (2000 ppm NO3− and 1.0 M KOH) containing electrolysis chamber on one side of the working electrode through a gas guide tube. Then, apply an external potential and conduct an electrolysis reaction at a constant potential. The generated hydrogen gas enters the top of the inverted burette through the gas guide tube. Finally, after 1 h of reaction, record the volume of the hydrogen product generated.In situ Raman spectroscopyIn situ Raman measurements were carried out using a Raman microscopy system and an electrochemical workstation. Raman spectroscopy was conducted with a Lab-RAM HR Raman microscopy system (Horiba Jobin Yvon, HR550) equipped with a 532 nm laser as the excitation source, a water immersion objective (Olympus LUMFL, 50×), a monochromator (1800 grooves/mm grating), and a Synapse charge-coupled device (CCD) detector. The electrolytic cell was made of polytetrafluoroethylene, and the working electrode was immersed into the electrolyte through the cell wall, with its plane remaining perpendicular to the incident laser. Platinum wire and Ag/AgCl electrodes were used as the counter and reference electrodes, respectively. Electrochemical intermittent in situ Raman spectroscopy was performed with a Renishaw InVia Qontor Raman system at 0.1 V intervals over a potential range of +0.5 to −0.2 V vs. RHE. After obtaining the first Raman spectrum, we added 0.1 mL of 2000 ppm KNO3 solution to the electrolyte. Each spectrum is an average of five continuously acquired spectra, with a collection time of 50 s for each collection. The cycle test was repeated four times.EPR experiments5,5-Dimethyl-1-pyrroline N-oxide (DMPO) was used to capture unstable hydrogen radicals by forming DMPO-H adducts, and the resulting EPR spectra were analyzed to detect the hydrogen radical signals produced by the catalyst during the reaction process44. In an H-type cell, the catalyst served as the working electrode, and constant electrolysis was performed at −0.1 V vs. RHE in a solution of 1.0 M KOH and 2000 ppm KNO3 for 5 min. After the reaction, 5 mL of the electrolyte solution was collected, and 10 μL of DMPO capturing agent was added, followed by Ar2 degassing. EPR measurements were carried out using a Bruker EMX-10/12 spectrometer under a frequency of ~9.5 GHz, a sweep width of 200 G, and a power of 20 mW.In situ FTIR spectroscopyElectrochemical in situ FTIR spectroscopy measurements were collected using Nicolet Nexus 8700 FTIR spectrometer equipped with a liquid N2-cooled system and MCT-A detector23,61. The Hg/HgO electrode and platinum foil electrode (Area~2 cm2) were used as the reference and counter electrode, respectively. The working electrode was prepared by depositing our developed catalysts as an active material over the glassy carbon electrode. The uniform thin layer (~10 μm) on the working electrode was obtained by vertically pressing it on the CaF2 window. The working electrode surface was set perpendicular to incoming infrared beam for obtaining the maximum response signals during electrochemical NO3− reduction. The catalyst’s in situ IR spectra (Rs) were obtained in the potential range of 0.5 V to −0.5 V at a scan rate of 100 mV/s. All the spectrums were reported after using the relation: ΔR/R = (RS−RRef)/ RRef, where the spectrum obtained at 0.5 V were considered as reference RRef. 62.GIXRDIn the GIXRD configuration, the incident X-ray beam with a wavelength of 0.6877 Å and an energy of 18 keV is kept at a small angle α concerning the sample surface, and the distance of samples to the detector was 315 mm. The incident X-ray beam with a wavelength of 0.8266 Å and an energy of 15 keV is kept at 0.1°, 0.2°, 0.3°, 0.4°, 0.5° concerning sample surface, and the distance of samples to the detector was 1946 mm63.Computational methodsAll calculations were carried out using density functional theory with dispersion correction D3 (DFT-D3), and the projected augmented wave (PAW) scheme was implemented in the Vienna ab initio simulation software package (VASP)64,65,66. For the structural relaxation and energy calculations, the generalized gradient approximation with Perdew-Burke-Ernzerhof (PBE) parameterization was used. The cut-off energy of the plane wave function was 500 eV. For the converged unit cell models of Co (2.49 × 2.05 × 4.02 Å3), the Brillouin zone was sampled with a 15 × 15 × 6 Γ-point centered Monk horst–Pack mesh, the energy convergence criterion was within 10−5 eV, and the force tolerance was smaller than 0.01 eV Å−1 on each atom. The (100) surface of Co was used as the catalytic substrate, and the Ru-Co gradient model was constructed based on Co (100). That is because it was observed in the experiment that the Co (100) crystal plane was exposed and was the lowest energy crystal plane67. The constructed (100) surface models contained six layers; the bottom two layers were fixed, and the top four layers were fully relaxed for geometry optimization. We applied a vacuum layer of at least 20 Å in the Z-direction of the slab models to prevent interactions between the slabs in the vertical direction. The energy convergence criteria were set to 10−4 eV, and the force tolerance of each atom was smaller than 0.02 eV/Å. The Brillouin zone was sampled by a k-point mesh of 3 × 3 × 168. The calculations involving all molecules and intermediate species on the Co (100) and Ru-Co (100) substrates were conducted with spin polarization.The Gibbs free energy change for the adsorbed *NO2 on an electrode surface to nitrite in aqueous solution (forming NO2−(l)) was calculated in three steps using the thermodynamic cycle shown in equations (Supplementary Table 8)69,70. The formation energy of NO2− on G-RuCo and NG-RuCo is calculated with the according to the following formula:$$\Delta {{\rm{G}}}\left({NO}_{2}^{-}\right)=G(*)-G(*{NO}_{2})+1/2{G}_{{gas}}\left({H}_{2}\right)-{G}_{{gas}}\left({HNO}_{2}\right)$$\({G}_{{gas}}\left({H}_{2}\right)\) and \({G}_{{gas}}\left({{HNO}}_{2}\right)\) are the corresponding Gibbs free energies of H2 and HNO2 molecules in the gas phase at 300 K and 1 atm. The entropic (ΔS) and enthalpic (ΔH) contributions to the free energy of the gaseous species were obtained from the NIST database. Nørskov’s computational hydrogen electrode model is used in the calculations71.Potential industrial applicationFor the scaled-up NH3 production process, we used chronopotentiometry to demonstrate the potential industrial application. A homemade NRA-OER industrial electrolytic cell in an MEA flow reactor was assembled. Here, the Co2P catalyst on Ni foam was selected as the anode electrode due to the potential of Co2P material in hydrogen evolution and oxygen evolution72,73. Self-supported G-RuCo on Ni foam catalysts (1 cm × 1 cm) and self-supported Co2P on Ni foam catalysts (1 cm × 1 cm) were directly used as cathodes and anodes, respectively, with anion exchange membranes (Nafion AMI-7001S) as separators. The catholyte was 1.0 M KOH and 2000 ppm KNO3 mixed electrolyte, while the anolyte was 1.0 M KOH with an electrolyte flow rate of 50 mL/min. Polarization curves and chronopotentiometry were used to evaluate the prospects of industrial nitrate electroreduction for ammonia production.
Gradient-concentration RuCo electrocatalyst for efficient and stable electroreduction of nitrate into ammonia
