PHOTOCATALYST

Abstract

A photocatalyst is described that is suitable for converting molecular nitrogen into ammonia. The photocatalyst comprises a layered base material comprising 1 to 100 layers, the layered base material being selected from the group consisting of molybdenum disulfide, tungsten disulfide, molybdenum telluride, tungsten telluride, molybdenum selenide and tungsten selenide, a layered base material comprising 1 to 100 layers, the layered base material being selected from the group consisting of molybdenum disulfide, tungsten disulfide, molybdenum telluride, tungsten telluride, molybdenum selenide and tungsten selenide, and 0.1-10.0% by weight, relative to the weight of the base material, of one or more Group VI, VII, VIII, IX or X transition metals. T he photocatalyst can further comprise 0.1-50.0% by weight, relative to the weight of the base material, of one or more semiconductor materials having an average particle size of 0.5-50.0 nm. The photocatalyst exhibits high catalytic efficiency without the need for high temperature and pressure. Also described is a process for the preparation of the photocatalyst, as well as uses of the photocatalyst for converting molecular nitrogen into ammonia.

Claims

1. A photocatalyst comprising: a layered base material comprising 1 to 100 layers, the layered base material being selected from the group consisting of molybdenum disulfide, tungsten disulfide, molybdenum telluride, tungsten telluride, molybdenum selenide and tungsten selenide; and 0.1-10.0% by weight, relative to the weight of the base material, of one or more Group VI, VII, VIII, IX or X transition metals.

2. The photocatalyst of claim 1, wherein the layered base material comprises 1 and 10 layers.

3. The photocatalyst of claim 1 or 2, wherein the layered base material is molybdenum disulfide.

4. The photocatalyst of claim 1, 2 or 3, wherein the photocatalyst comprises 1.0-3.0% by weight, relative to the weight of the base material, of one or more Group VI, VII, VIII, IX or X transition metals.

5. The photocatalyst of any preceding claim, wherein the one or more Group VI, VII, VIII, IX or X transition metals is selected from the group consisting of Fe, Mn, Co, Mo, Ni, Ru, Rh, Pd and Pt.

6. The photocatalyst of any preceding claim, wherein the one or more Group VI, VII, VIII, IX or X transition metals is Fe.

7. The photocatalyst of any preceding claim, wherein the size of the one or more Group VI, VII, VIII, IX or X transition metals ranges from single atoms of the transition metals to atomic clusters of the transition metals having a maximum diameter of 4.0 nm.

8. The photocatalyst of any preceding claim, wherein the photocatalyst further comprises 0.1-50.0% by weight, relative to the weight of the base material, of one or more semiconductor materials having an average particle size of 0.5-50.0 nm.

9. The photocatalyst of claim 8, wherein the one or more semiconductor materials has an average particle size of 0.5-15.0 nm.

10. The photocatalyst of claim 8 or 9, wherein the one or more semiconductor materials has the compositional formula AB.sub.xC.sub.1-x, wherein A is selected from the group consisting of Cd, Pb and In; B and C are selected from the group consisting of S, Se, Te, As and P; and x is a number ranging from 0.01 to 1.

11. The photocatalyst of claim 8, 9 or 10, wherein the one or more semiconductor materials is cadmium sulfide.

12. The photocatalyst of any preceding claim, wherein the photocatalyst has an average particle size of 0.05-100 μm.

13. A process for preparing a photocatalyst as claimed in any preceding claim, the process comprising the steps of: a) providing a dispersion of a layered base material comprising 1 to 100 layers, the layered base material being selected from the group consisting of molybdenum disulfide, tungsten disulfide, molybdenum telluride, tungsten telluride, molybdenum selenide and tungsten selenide; and b) contacting the dispersion of the layered base material with a solution of one or more Group VI, VII, VIII, IX or X transition metals.

14. The process of claim 13, wherein the layered base material having between 1 and 100 layers is prepared by exfoliating the base material in its bulk form.

15. The process of claim 14, wherein the base material in its bulk form is exfoliated by: (i) contacting an aqueous mixture of the base material in its bulk form with an intercalant; (ii) sonicating the mixture resulting from step (i); and (iii) isolating the layered base material having between 1 and 100 layers resulting from step (ii).

16. The process of claim 13, 14 or 15, wherein the solution of one or more Group VI, VII, VIII, IX or X transition metals is prepared by dissolving one or more Group VI, VII, VIII, IX or X transition metal precursor compounds in a solvent.

17. The process of any one of claims 13 to 16, wherein step b) is conducted at a temperature of 10-325° C., optionally under hydrothermal conditions.

18. The process of claim 17, wherein step b) is conducted at a temperature of 130-190° C., under hydrothermal conditions.

19. The process of any one of claims 13 to 18, wherein the photocatalyst resulting from step b) is contacted with an aqueous solution of one or more semiconductor materials having an average particle size of 0.5-50.0 nm.

20. A photocatalytic process for the conversion of molecular nitrogen to ammonia, the process comprising the step of: a) contacting molecular nitrogen with a photocatalyst as claimed in any one of claims 1 to 12 in the presence of water; wherein step a) is performed under the application of electromagnetic radiation having a wavelength of 270-1000 nm.

21. The process of claim 20, wherein the electromagnetic radiation is supplied to the mixture of step a) using a solar concentrator.

22. The process of claim 20 or 21, wherein step a) is conducted at a temperature of 5-270° C.

23. The process of clam 20, 21 or 22, wherein step a) is conducted at a temperature of 10-50° C.

24. The process of any one of claims 20 to 23, wherein the photocatalyst is provided as: A) a fixed bed; B) a suspension; or C) a thin film.

25. The process of any one of claims 20 to 24, wherein step a) is performed as: A) a batch process; or B) a continuous process.

Description

EXAMPLES

[0151] One or more examples of the invention will now be described, for the purpose of illustration only, with reference to the accompanying figures, in which:

[0152] FIG. 1 shows a reaction scheme for N.sub.2 fixation to NH.sub.3 catalysed by nitrogenase (A) and the mimicking self-ensemble Fe-sMoS.sub.2 hybrids (B), respectively. Fe protein containing [4Fe-4S] is shaded in brown colour, MoFe protein is shaded in green and the FeMoco cofactor is shaded in yellow with a ball and stick model of FeMo cluster [8Fe-9S—Mo—C]. Mo, S, Fe, C atoms are shown as balls in blue, yellow, brown, black colour, respectively. (A) shows that Fe protein consumes ATP to produce electrons (e.sup.−) and transfers them to FeMoco via outer sphere mechanism. The transferred electrons are thought to migrate along the Fe/Mo—S molecular relay centres for to electron-rich metal centre for N.sub.2 reduction. (B) shows that solar energy is supplied to generate excited hole (h.sup.+) and e.sup.−. The excited electrons are rapidly separated and transferred by the similar Fe—S—Mo units as the ‘electronic relays’ to electron-rich Fe atoms for N.sub.2 reduction by protons from water to ammonia while the excited h.sup.+ is relaxed with OH.sup.− from water to produce molecular O.sub.2.

[0153] FIG. 2 shows the exfoliation of bulk MoS.sub.2 and the doping of single transition metal on the obtained single layered MoS.sub.2.

[0154] FIG. 3 shows atomic force microscopy (AFM) image analyses for the chemically exfoliated sMoS.sub.2. (A) AFM image of spin-coated sMoS.sub.2 with a scan line, and (B) a model of 2-H MoS.sub.2 with 3-layer structure perpendicular to c axis. (C) It can be seen that the step heights of individual layers of 0.6-0.7 nm. This value is comparable to ca. 0.65 nm of a single layer of the S—Mo—S building block as shown in B. Statistical analysis of 100 flakes produced by the lithium exfoliation method revealed that 56% of the flakes to be monolayer, 28% of two layers and 13% of three layers and so on. The average topographic height is around 1.04 nm, which agrees with typical height of a sMoS.sub.2 with the presence of water molecules (between 0.6 and 1.0 nm). Sample was prepared by spin-coating sMoS.sub.2 onto a surface of Si/SiO.sub.2 substrate. The lateral dimension of this sMoS.sub.2 nanosheet is approximately 20-40 nm.

[0155] FIG. 4 shows the characterization of the basal plane of Fe-sMoS.sub.2 and identification of the Fe atoms. (A) HAADF-STEM of the Fe-sMoS.sub.2 with a model of single layer 2H—MoS.sub.2. Blue and yellow balls represent Mo and S atoms. (B) An atomic resolution HAADF-STEM of the basal plane of Fe-sMoS.sub.2; Scan 1 (green line) contains Fe atom on Mo atop site; Scan 2(red line) contains S vacancy substituted by Fe atom; normal sites (blue line) (C) atomic EELS acquired for spots (1),(2) in (A): an EEL edge at 708 eV corresponding to L.sub.3 edge of Fe atom. (D) An atomic model based on optimized DFT for a single Fe atom on Mo atop site. (E) An atomic model based on optimized DFT for a single Fe atom substituting a S site. (F and G) ADF intensity line profiles taken along the correspondingly numbered lines (1) and (2), respectively indicated in (B).

[0156] FIG. 5 shows DFT optimized geometries of Fe binding configurations and their binding energies. Top, side and perspective views (left to right) of the DFT calculated geometries for Fe on the Mo atop site (A) and S substitution (B). The values are the calculated bond lengths with the unit of Å.

[0157] FIG. 6 shows structural analysis of Fe-sMoS.sub.2 and its comparative catalytic performance with related systems for photocatalytic ammonia synthesis. (A) Fourier transformed magnitudes of the experimental Fe k-edge EXAFS spectra of samples. (B) Wavelet transformation for the k3-weighted Fe k-edge EXAFS signals of Fe-sMoS.sub.2 based on Morlet wavelets with optimum resolutions at the first and higher coordination shells. The intensity reflects the content of scattering signal. (C) An atomic model FeMo—S cluster in FeMo cofactor of FeMoco. The four-membered ring motif of [Fe—S.sub.2-Mo] in FeMoco is shaded with green line. (D) An atomic model from optimized DFT of a single Fe on Mo atop site. The same four-membered ring motif of [Fe—S.sub.2-Mo] in Fe-sMoS.sub.2 is shaded with green line. (E) Catalytic performance of Fe-sMoS.sub.2 as compared with other MoS.sub.2 systems. Blank represents a blank experiment using CdS quantum dots instead of pre-mixing with Fe-sMoS.sub.2 (CdS-Fe-sMoS.sub.2) under the same conditions. Reaction conditions: r.t., 10 mL/min of N.sub.2, 1 h. Activity was evaluated by averaging at least 3 repeated measurements under the same conditions to generate the measurement errors. Inset figure is the ammonia production of CdS:Fe-sMoS.sub.2 as a function of reaction time.

[0158] FIG. 7 shows the k3 EXAFS curves at Fe K edge in k for Fe-sMoS.sub.2.

[0159] FIG. 8 shows wavelet transforms for the k3-weighted Fe k-edge EXAFS signals of Fe foil based on Morlet wavelets with optimum resolutions at the first and higher coordination shells. The intensity reflects the content of scattering signal.

[0160] FIG. 9 shows HAADF-STEM images of Co-sMoS.sub.2. (A) Co at the Mo-atop site model, (B) EELS acquired along the line in (A). (C) Ni at the Mo-atop site model and (D) EELS acquired along the line in (C).

[0161] FIG. 10 shows Fe k edge XANES spectra for atomically dispersed Fe-sMoS.sub.2. Fe foil, FeCl.sub.2, and FeCl.sub.3 are used as references.

[0162] FIG. 11 shows (A) An atomic model FeMo cluster in FeMo cofactor. The four-membered ring of [Fe—S.sub.2-Mo] unit in FeMoco is showed with green bond. (B) An atomic model from geometry optimized DFT of a single Fe on Mo atop site with values from EXAFS. The value represents the bond length of FeMoco and Fe-sMoS.sub.2 with the unit of Å.

[0163] FIG. 12 shows a schematic diagram of energy band potentials of the conduction band (CB) and valence band (VB) of the as-synthesized (bulk) bMoS.sub.2, (few layer) fMoS.sub.2, ultra-thin MoS.sub.2, and CdS at PH=7. The CB and VB values are compared with the potentials of nitrogen reduction and water oxidation.sup.14,15.

[0164] FIG. 13 shows mass spectroscopy of indophenols after reaction with the solution pre-trapped with gaseous product from (A) .sup.14N.sub.2/H.sub.2O and (B) .sup.15N.sub.2/H.sub.2O over Fe-sMoS.sub.2 upon visible light illumination. The aliquot solution (A) shows the formation of condensed indophenol based complex at m/z 198. In contrast, the aliquot solution (B) gives the characteristic .sup.15N-labelled indophenol based complex at m/z 199 with relative intensity significantly higher than the natural abundance ratio of .sup.14N: .sup.15N nuclei. This result clearly confirms that gaseous N.sub.2 is fixed into NH.sub.3 by the catalyst under light illumination.

[0165] FIG. 14 shows catalytic performance and mechanism study on how the Fe—S.sub.2-Mo unit promotes N.sub.2 reduction. (A) Time-resolved ATR-FTIR spectra at a flow of N.sub.2 and H.sub.2O. Time zero was set when light illumination started. (B) Photocatalytic activity for ammonia (and oxygen) production with trace H.sub.2 on first-row transition metal doped MoS.sub.2 and their electron decay time obtained from time-resolved photoluminescence; (C) Calculated HOMO-LUMO states of Fe-sMoS.sub.2. Green net represents positive electron density and brown is negative. (D) Quantum efficiency (Q.E.) of Fe-sMoS.sub.2 at 436 nm, 575 nm, 650 nm, and 750 nm, respectively of the incident light.

[0166] FIG. 15 shows In-situ cell for time-resolved ATR-FTIR.

[0167] FIG. 16 shows the UV-vis spectrum of the filtered solution with chromogenic agent para-(dimethylamino) benzaldehyde acidic solution).

[0168] FIG. 17 shows time-resolved photoluminescence of Mn, Fe, Ni-doped sMoS.sub.2 and sMoS.sub.2, measured using a pulsed Ti:sapphire laser (˜150 fs).

[0169] FIG. 18 shows HOMO (down)-LUMO (up) states of sMoS.sub.2, Mn, Co, and Ni-sMoS.sub.2. Green net mesh is positive electron density and brown is negative.

[0170] FIG. 19 shows density of status for Fe, Co, Ni doped-sMoS.sub.2 and undoped sMoS.sub.2.

[0171] FIG. 20 shows the energy plot of N.sub.2 adsorption on Fe-sMoS.sub.2 from DFT calculations with the reference to the energies of Fe-sMoS.sub.2 and free N.sub.2 molecule.

[0172] FIG. 21 shows mechanism on how the Fe—S.sub.2-Mo unit promotes N.sub.2 reduction. Energies of intermediates states in the mechanism of N.sub.2 reduction at the Fe-sMoS.sub.2 from DFT calculation.

[0173] FIG. 22 shows TRPL decay time dependent on nature of gas exposure over transition Fe-doped and undoped MoS.sub.2. Adsorption of N.sub.2 from air appears to influence the TRPL over the Fe—S.sub.2-Mo unit which promotes N.sub.2 reduction as discussed in relation to FIG. 21.

1. MATERIALS

[0174] Reagents used for synthesis were: MoS.sub.2 (Sigma-Aldrich); iron acetate (reagent grade, Alfa Aesar); FeCl.sub.3.6H.sub.2O (reagent grade, Alfa Aesar); FeCl.sub.2 (reagent grade, Sigma-Aldrich); n-butyllithium/hexane (reagent grade, Sigma-Aldrich); Polyvinylpyrrolidone (PVP, reagent grade, Sigma-Aldrich); Potassium acetate (reagent grade, Sigma-Aldrich); Cd acetate (reagent grade, Sigma-Aldrich); Sodium sulfuride (reagent grade, Sigma-Aldrich); Thioglycolic acid (TGA, anhydrous, ≥99.9%, Sigma-Aldrich); KBr (reagent grade, Sigma-Aldrich); hydrazine (puriss. p.a., absolute ≥99.8% (GC), Sigma-Aldrich); isopropanol (99.9%, Sigma-Aldrich); para-(dimethylamino) benzaldehyde (reagent grade, Sigma-Aldrich); H.sub.2SO.sub.4 (≥98%, Sigma-Aldrich); .sup.15N.sub.2 (98%, CK Isotopes).

2. METHODS

2.1. Synthesis of Few-Layered MoS.SUB.2 .(fMoS.SUB.2.) and Single-Layered MoS.SUB.2 .(sMoS.SUB.2.)

[0175] Few-Layered MoS.sub.2. 6 g of bulk MoS.sub.2 powder was dispersed in 400 mL of Water/Isopropanol (1:3, v/v). 4 mL of hydrazine monohydrate was then added. The solution mixture was placed into the sonication bath for 12 hours for exfoliation, followed by centrifugation at 2000 rpm for 60 minutes. The supernatant collected was filtered using vacuum filtration, followed by washing with water. The exfoliated product was dried under vacuum for 24 hours.

[0176] Single-Layered MoS.sub.2. 0.5 g of bulk MoS.sub.2 powder was soaked in 4 mL of 1.6 M n-butyllithium/hexane under nitrogen atmosphere for 48 hours. Solid Li.sub.xMoS.sub.2 was then isolated by vacuum filtration, followed by washing with hexane to remove excess n-butyllithium. It was then dried under vacuum for 24 hours. The dried product was then immersed into 250 mL of water. The solution was placed into the sonication bath for 12 hours and then centrifuged at 5000 rpm for 15 minutes. The supernatant collected was filtered using vacuum filtration, followed by washing with water. The exfoliated product was dried under vacuum for 24 hours.

2.2. Synthesis of Single Fe Atom Doped bMoS.SUB.2./fMoS.SUB.2./sMoS.SUB.2

[0177] Fe precursor solution was prepared by dissolving 0.2 mM metal ions into 1 mL of 0.5 mM thiourea solution and left for overnight to form a metal complex. The metal complex solution was mixed with 30 mL of colloid solution, which was made by dispersing 30 mg of sMoS.sub.2 (b MoS.sub.2 or fMoS.sub.2) in 30 mL of water/isopropanol (1:3, v/v) and 30 mg of PVP (stabiliser). The solution mixture was then transferred to an autoclave and then placed into an oven at 160 ° C. for 24 hours. Afterwards, the precipitate was washed with deionized water and dried under vacuum for 12 hours to obtain the solid product.

2.3. Synthesis of CdS:Fe-sMoS.SUB.2

[0178] CdS quantum dots were synthesized according to previous reports with slight modifications.sup.4. Briefly, 250 uL of TGA was added into 50 mL of Cd acetate (10 mM) aqueous solution, and N.sub.2 was bubbled throughout the solution to remove O.sub.2 at 110° C. During this period, 1.0 M NaOH aqueous solution was slowly added with adjustment to raise the pH to 11 gradually. Following this step, 5.5 mL of 0.1 M Na.sub.2S aqueous solution was injected into the CdS quantum dots. The reaction mixture was refluxed under N.sub.2 atmosphere for 4 h. Finally, the desired CdS quantum dots were obtained and stored in a refrigerator at 4° C. for further use. To load the CdS quantum dots onto the basal plane of sMoS.sub.2, 10 mg of Fe-sMoS.sub.2 was dipped into 4 mL of CdS quantum dots aqueous solution (0.25 mg/mL) for 1 h.

2.4. Characterisation

[0179] High-angle annular dark field scanning transition electron microscopy (HAADF-STEM). The finely ground samples were placed onto the holey carbon coated Cu-TEM grid for analysis. The analysis was performed by JEOL-JEM2100 Aberration-Corrected Transmission Electron Microscope in Birmingham. A voltage of 60 kV to avoid beam excitation and damage was applied for the imaging. An off-axis annular detector imaging was used for Dark-field (Z-contrast) imaging and atomic-resolution imaging. Compositional analysis by X-ray emission detection was also conducted. For the EDX detector, Bruker 5030 SDD detector with a window area of 30 mm.sup.2 was used. All results were then processed with Esprit 2.0 software.

[0180] Inductively coupled plasma (ICP). The finely ground samples were dissolved and diluted with 5 wt. % HCl for ICP analysis. The analysis was performed by ICP optical emission spectroscopy (Optima2100DV, PerkinElmer). The doped-metal content was controlled at around 3 wt. % with error ±0.5 (Fe 3.3 wt. %, Co 3.0 wt. %, Ni 3.5 wt. %).

[0181] Extended X-ray absorption fine structure (EXAFS). Fe K-edge and Mo K-edge X-ray absorption spectra was conducted in fluorescence mode at the BLO7A XAS beamline at NSRRC, Taiwan. To examine the local chemical environment around Fe and Mo atoms, EXAFS data were extracted from XAS spectra. The Demeter ATHENA program was used for XAFS data analysis, where the data were background subtracted, normalised and Fourier transformed. The Demeter ARTEMIS program was used to perform the least-squares curve fitting analysis of the EXAFS .sub.X(k) data. The EXAFS Wavelet analysis was performed following the protocol and calculations developed by Marina Chukalina and Harald Funke, where the backscatter atoms are distinguished within the same atomic shell.sup.16. To confirm the reproducibility of the experimental data, at least 2 scan sets were collected and compared for each sample. The spectra were calibrated with Fe and Mo foil as reference. The amplitude reduction factor was obtained from analysis of the Fe and Mo foil, which was used as a fixed input parameter to allow refinement in the coordination number and bond distance of the absorption element.

[0182] Time-resolved photoluminescence (TRPL) spectroscopy. Photoluminescence spectra and corresponding lifetime of excitons were obtained from a bespoke micro-photoluminescence setup, in which a Ti-Sapphire laser (λ=266 nm, pulse duration=150 fs, repetition rate=76 MHz) was directed onto the sample. Time-resolved measurements were performed using the spectrometer as a monochromator before passing the selected signal to a photomultiplier tube (PMT) detector with an instrument response function width of ˜150 ps connected to a time-correlated single-photon counting module. Parameters describing the photoluminescence were obtained by fitting the background-corrected PL spectra with a monoexponential decay function of the form y=A.sub.1exp(−x/t.sub.1)+y.sub.0 for sMoS.sub.2. A double-exponential model using equation of l(t)=A.sub.1exp(−t/T.sub.1)+A.sub.2 exp(−t/T.sub.2) when d orbital metal (Mn, Fe, Co, and Ni) was introduced.sup.17,18.

[0183] Attenuated total reflection fourier transform infrared (ATR-FTIR) spectroscopy. In situ ATR-FTIR spectra were collected using a multiple-reflection ATR accessory (PIKE Technologies, custom-modified GladiATR) in a Varian 680-IR spectrometer, controlled by Resolutions Pro software. A trapezoidal Si internal reflection element (IRE, Crystal Gmbh, 8.39×5×1 mm.sup.3) with a face angle of 39° was sealed into a polyether ether ketone (PEEK) baseplate using silicone sealant, and a custom cell sealed on top.sup.19. A layer of water molecules, which were necessary to provide protons, was first pre-adsorbed on the surface from a drop of water onto the catalyst. Subsequently, 50 mL/min of N.sub.2 saturated with H.sub.2O was passed over the catalyst while the visible light source was turned on and the IR absorption monitored with an MCT detector over the course of the reaction.

[0184] Ultraviolet-visible (UV-vis) absorption spectroscopy. UV-vis absorption spectrum was collected using a Varian 100 Bio UV-Visible Spectrometer in absorbance mode with a step interval of 1 nm. The solution after reaction overnight was filtered. The obtained 5 mL was mixed with 5 mL of 0.14 M para-(dimethylamino) benzaldehyde and 1 M H.sub.2SO.sub.4 solution, finally transferred into an optical glass cuvette for hydrazine measurement. The concentration of ammonia solution is also detected using UV-vis spectrum with Nessler's agent.

[0185] DFT Theoretical Calculation. All calculations were performed using the first-principles density of functional theory (DFT) as implemented in Vienna ab initio simulation packages (VASP).sup.20, the exchange-correlation energy functional described by generalized gradient approximation using Perdew-Burke-Ernzerhof (PBE) functional.sup.21, and the ion-electron interaction was treated using the projector-augmented wave (PAW) method.sup.22 with a plane-wave cutoff energy of 400 eV. A (3×3) supercell of 2H-MoS.sub.2 was selected to simulate single-layered MoS.sub.2 (sMoS.sub.2), periodic boundary conditions were employed and 15 Å of vacuum in the z-direction was set to separate neighboring single-layered MoS.sub.2. The Brillouin zone has been sampled using a 2×2×1 and an 8×8×4 Monkhorst-Pack.sup.23 grid of k-points for geometry optimizations and orbital analysis calculations, respectively. Both lattice constants and atomic positions were relaxed until the forces on atoms were less than 0.02 eV {circumflex over (Å)}.sup.−1 and the total energy change was less than 1.0×10.sup.5 eV. To rationalize the different performance of sMoS.sub.2 and transition metal doped MoS.sub.2 in catalytic ammonia photosynthesis, density of states and frontier orbitals topology analysis were performed at the PBE/PAW level of theory.

2.5. Ammonia Synthesis Measurement

[0186] All photocatalytic activity experiments were conducted at ambient temperature using a 70 W tungsten lamp (Glamox Professional 2000) with UV light cut-off to simulate visible light, respectively. For the fixation of molecular nitrogen, 4 mg of photocatalyst was added into 100 mL of double distilled water in a reactor. The reactor was equipped with water circulation in the outer jacket in order to maintain at room temperature of 25° C. The mixture was continuously stirred in the dark and under visible light with high-purity N.sub.2 (99.99%) bubbled at a flow rate of 10 mL/min. Five milliliters of the solution was taken out each 30 min and after filtering to remove the photocatalyst, and the concentration was monitored by colorimetry with the UV-vis spectrometer. For the measurement of ammonia yields, a specialized highly sensitive ammonia detector was used (Thermo Sicentifc™ Orion™ Ammonia Gas Sensing ISE Electrode). Quantum efficiency measurements were carried out under a 300 W Xenon lamp through quartz windows using bandpass filters of 437±10 nm, 575±25 nm, 650±20 nm, and 750±20 nm.

[0187] Isotopic N.sub.2 was used to prove that the obtained ammonia derives from N.sub.2 gas rather than some other sources. Indophenol assays were prepared by adding 0.5 mL of aliquot solution after 1-h reaction to 0.1 mL of 1% phenolic solution in 95% ethanol/water. Stepwise, 0.375 mL of 1% NaClO in alkaline sodium citrate solution and 0.5 mL of 0.5% Na[Fe(CN).sub.5NO] solution were added. The assayed aliquots were aged overnight before analyzing on a Xevo LCMS-ESI system.

3. RESULTS AND DISCUSSION

3.1. Synthesis

[0188] The 2-D single molecular layer MoS.sub.2 (termed sMoS.sub.2) consisting of three-sub-layers of S—Mo—S in a trigonal prismatic 2-H structure was first synthesized via exfoliation of bulk MoS.sub.2 using n-butyllithium. Subsequently Fe atoms were attached to sMoS.sub.2 using a hydrothermal method for in-situ formed sulphide species (FIGS. 2 and 3), followed by H.sub.2 reduction to afford the molecular Fe-sMoS.sub.2 in nanosize. In this synthesis, Fe atoms were atomically dispersed and assembled onto the basal plane of sMoS.sub.2.

[0189] Similar transition metal-doped catalysts were prepared in the same manner using Mn, Co or Ni instead of Fe.

3.2. Characterisation

[0190] FIG. 4A shows a typical high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image of Fe-sMoS.sub.2 with corresponding 2-H characteristic pattern. As revealed from the distinctive brighter spots than the surrounding Mo and S.sub.2 sites in 2-H arrangement in FIG. 4A (red circles) and the enlarged image FIG. 4B show that individual Fe atoms are uniformly dispersed and overlap the position of Mo and S sites in the structural motif of 2H-sMoS.sub.2. To further confirm the nature of this adsorbed atom, atomic resolved electron energy loss spectroscopy (EELS) were performed on these brighter spots. The EEL spectra (FIG. 4C) at the corresponding positions demonstrate the presence of Fe atoms with the characteristic signature L.sub.3 edge at 708 eV.sup.24. The images for HAADF-STEM and EELS analysis show that isolated Fe atoms are deposited at the two favoured positions in the basal plane of sMoS.sub.2. A typical intensity profile analysis of the HAADF shown in FIG. 4F demonstrates that a Fe atom commonly takes residence on atop site of Mo of the basal plane. Occasionally, as shown in FIG. 4G, an Fe atom can be found in the position of S site as substitution. The corresponding structural models shown in FIGS. 4D and 4E based on DFT calculations also confirm the existence of these two energetic favoured atomic positions (FIG. 3).

[0191] As shown in k3-weighted Fourier transformed spectra in the extended X-ray absorption fine structure (EXAFS) for Fe post-k edge analysis of Fe-sMoS.sub.2 (FIG. 6A), there are clearly Fe-S contributions at a distance of 1.7Å for Fe-sMoS.sub.2, which is distinctively different from the Fe-Fe contributions at a distance of 2.2 Å calibrated by Fe foil. The distance is close to the Fe-Cl contributions in FeCl.sub.3, indicating that Fe is atomically isolated as revealed by HAADF-STEM images shown in FIGS. 4A and 4B. There is a small but new peak at 2.1 Å attributed to Fe-Mo interaction, which is agreeable to the envisaged bonding environment of Fe atop to the Mo site on 2-H sMoS.sub.2 and the theoretical model from the DFT calculations (FIG. 6D). Similarly, the small peak at 2.7 Å can be attributed to the long Fe-S bonds where Fe substitutes into the S site (FIG. 6E, see model in FIG. 5). The EXAFS curve fit matches with the expected coordination number of the nearest sulfur atoms around the isolated Fe atom of 2.9±0.3 with the distance of 2.1 Å, and the nearest Mo atoms around the isolated Fe atom is 1±0.2 according to the atop model (FIG. 7 and Table 1).

TABLE-US-00001 TABLE 1 The fitted coordination environment of Fe in sMoS.sub.2 Scattering Path Enot (eV) R (Å) CN D-W factor Fe—S −5.7 2.27 ± 0.02 2.9 ± 0.3 0.010 Fe—Mo 1.1 2.54 ± 0.03 1.0 ± 0.2 0.003 Fe—S′ −5.7 3.08 ± 0.03 3.7 ± 0.4 0.010 “CN” is coordination number; “D-W” is Debye Waller (thermal atomic uncertainty); R is the bond length; Enot is the energy difference between theoretical and calculated scattering path.
R factor=1.3%; Kwt=1,2,3; k range 3-10; No Fe-Fe bonds can be fitted in the first 2 shells, indicating the Fe species are in form of single atoms.

[0192] FIG. 6B shows the WT-EXAFS wavelet transformed analysis based on Morlet wavelets, which can be used to differentiate closely-related spatial interactions.sup.25. For Fe-sMoS.sub.2, the hot spot of the WT maximum at ˜3 Å.sup.−1 is well-resolved at the first coordination shell, which can clearly be related to the Fe—S bond at atop site. In contrast, the WT intensity hot spot at ˜8 Å.sup.−1 region corresponding to the Fe—Fe bond was not detected in Fe-sMoS.sub.2, which indicates the sole dispersion of individual Fe atoms in Fe-sMoS.sub.2 (FIG. 8). In addition, there is an associated weak asymmetry WT intensity area ranging from 2-3 Å for Fe-sMoS.sub.2, which is attributed to the mixed contributions of Fe-Mo bonds and longer Fe—S bonds at the S substitution sites (reference to FIG. 5). It is noted that the WT intensity for this Mo atop site is much stronger than that of the S substitution site, demonstrating that this is the principle site for the Fe, as reflected by the HAADF-STEM analysis.

[0193] Similar structure was obtained for Co and Ni-doped sMoS.sub.2 at comparable doping levels (FIG. 9).

[0194] X-ray absorption near-edge structure (XANES) analysis was also carried out to better understand the single-atom Fe-sMoS.sub.2 catalyst. As shown in FIG. 10, the Fe k-edge XANES spectrum of quenched Fe-sMoS.sub.2 from photocatalysis is drastically different from that of metallic Fe foil (Fe.sup.0), but its absorption edge in the right-shift position between Fe.sup.IICl.sub.2 and Fe.sup.IIICl.sub.3 implying that the average working oxidation state is between them for the anchored Fe.

[0195] The molecular models of FeMoco and Fe-sMoS.sub.2 shown in FIG. 6C and FIG. 6D, respectively illustrate their similar structural motifs of the four-membered [Fe—S.sub.2-Mo] rings. Interestingly, the derived Fe-S, Mo-S and Fe-Mo bonding lengths of the [Fe—S.sub.2-Mo] in this single layer molecular Fe-sMoS.sub.2 catalyst are extremely close to that of the reported molecular [Fe—S.sub.2-Mo] unit in FeMoco from single crystal data within the average deviations of 10%.sup.28 (see Table S2(2), FIGS. 5 and 11).

TABLE-US-00002 TABLE 2 The bond lengths of [Fe—S.sub.2—Mo] in FeMoco and Fe—sMoS.sub.2 on basis of EXAFS experiments in comparison with optimised DFT calculations. Bond length (Å) Bond species FeMoco Experiment Calculation Fe—S1 2.248 2.27 ± 0.02 2.137 Fe—Mo 2.666 2.54 ± 0.03 2.539 Fe—S2 2.213 2.27 ± 0.02 2.136 Mo—S1 2.356 2.40 ± 0.01 2.542 Mo—S2 2.336 2.40 ± 0.01 2.542

3.3. Conversion of N.SUB.2 .to NH.SUB.3

[0196] FIG. 6E confirms that Fe-sMoS.sub.2 with [Fe—S.sub.2-Mo] is active to convert N.sub.2 to NH.sub.3 via photo-activation to provide excited electrons for the N.sub.2 fixation in H.sub.2O at ambient conditions. Bulk MoS.sub.2 is shown to be inert for N.sub.2 reduction presumably because of its conduction band (CB) is more positive than that of the N.sub.2/NH.sub.3 redox couple.sup.27 (see FIG. 12). By reducing the layers of MoS.sub.2, the activity for nitrogen fixation to ammonia is gradually enhanced. It has been proven that the band gap of MoS.sub.2 could be enlarged with a more negative CB edge striding over the N.sub.2/NH.sub.3 redox couple.sup.27,28. In addition, indirect band excitation over few-layer MoS.sub.2 can be switched to a more efficient direct band excitation for single layer MoS.sub.2.sup.29. While the single layered MoS.sub.2 materials display a negligible activity in H.sub.2/O.sub.2 splitting from water by visible light, a substantial higher photocatalytic ammonia production rate is recorded. Notably, the introduction of the [Fe—S.sub.2-Mo] motifs into the basal planes of single layered MoS.sub.2 displays a far more superior activity for ammonia and (stoichiometric oxygen) production from N.sub.2 and H.sub.2O reaction with trace hydrogen gas formation than most recent reported photocatalysts in visible light regime (Table 3), thereby mimicking the FeMoco.

TABLE-US-00003 TABLE 3 Comparison of photocatalytic ammonia synthesis over Fe—sMoS.sub.2 and CdS—Fe—sMoS.sub.2 along with others from recent literature Activity Light Reactant/ Published Entry Sample (μmol/g/h) source reagent Reference year 1 Fe—sMoS.sub.2 249 70 W UV cut-off N.sub.2, H.sub.2O This work — tungsten lamp 2 CdS—Fe—sMoS.sub.2 459 70 W UV cut-off N.sub.2, H.sub.2O This work — tungsten lamp 3 FeS.sub.2/CNT ~67 Xenon lamp N.sub.2, H.sub.2O (30) 2018 4 JRC-TIO-6 0.7 300 W Hg lamp N.sub.2, H.sub.2O (37) 2017 (TiO.sub.2) (280-420 nm) 5 Fe—TiO.sub.2 10 360 W Hg lamp N.sub.2, H.sub.2O/ (32) 1977 HCl 6 BiOCl 68.9 500 W N.sub.2, H.sub.2O/ (33) 2015 Xenon lamp CH.sub.3OH 7 5% Ru@n- 120 290-380 nm N.sub.2, H.sub.2 (34) 2017 GaN NWs UV irradiation 8 W.sub.18O.sub.49 195.5 300 W N.sub.2, H.sub.2O/ (35) 2018 Xenon lamp Na.sub.2SO.sub.3

[0197] Light driven nitrogen fixation over solid catalysts in aqueous medium has been intensively studied with continual interests. TiO.sub.2 has been receiving considerable attention due to its outstanding photochemical properties but the wide band gap of 3.2 eV denies the direct ammonia production by visible light activation. As a result, modified TiO.sub.2 materials to reduce the band gap or include promotors to capture visible light have been commonly applied. Despite these attempts, low activities for photo ammonia production using visible light are generally obtained. The low levels of ammonia can be seen from the typical modified TiO.sub.2 such as entries 4 and 5 (Table 3) where a significant quantity of ammonia is produced due the use of unfiltered light source with UV component. Other semi-conductive oxide-based materials such as BiOCl and W.sub.18O.sub.49 (entries 6 and 8 in Table 3) are also known to capture visible light for ammonia production but they show low activities, presumably due to poor charge separation (short lifetime for exciton recombination) from lack of rapid charge separation component for these structures. It should be particularly noted from entry 7 (Table 3) that 5% Ru@n-GaN NWs is a promising material, which exhibits higher ammonia production activity by using N.sub.2/H.sub.2 at room temperature. In contrast with the recent literature, Fe-sMoS.sub.2 with [Fe—S.sub.2-Mo] motifs on 2-D single layered MoS.sub.2 show the highest ammonia production activity from direct visible light activation without using sacrificial reagent (see entries 1 and 2 in Table 3). This indicates its unique structure for the efficient charge separation and activation of N.sub.2 in visible light and water for the ammonia production.

[0198] Although only a small quantity of photocatalyst was made, it is sufficient to cover more than 6 m.sup.2 of farmland per gram of catalyst assuming the leaching rate of 100 kg N ha.sup.−1 for a selected crop is used. This value for decentralised photocatalytic ammonia fertilizer production was estimated as follows: [0199] Assuming the highest leaching rate of 100 kg N ha.sup.−1 (1 ha=10,000 m.sup.2) per year of a selected crop (see K. Sieling, et al. Journal of Agricultural Science, Cambridge (1997), 128, 79-86). [0200] Thus, the leaching N rate=0.714 mole N/m.sup.2/yr. [0201] The photo-catalyst produces 500μ mol N/g/h=500×24×365=4.38 mole N/g/yr. [0202] The depleted-N area replenished per gram of catalyst=4.38/0.714=6.134 m.sup.2/g.

[0203] Isotope labelled .sup.15N.sub.2 was used to track the nitrogen source of ammonia, which confirmed that gaseous .sup.15N.sub.2 was fixed by this FeMoco-like Fe-sMoS.sub.2 (FIG. 13). Interestingly, further activity promotion by incorporation of light-captured CdS quantum dots to Fe-sMoS.sub.2 can be achieved. The rate for N.sub.2 reduction to NH.sub.3 maintains for at least 120 min under constant illumination without showing any obvious attenuation (FIG. 6E inset). It is expected that the CdS quantum dots can contribute additional electron-hole pairs from visible light illumination, the significant activity enhancement reflects their efficient charge separation by the [Fe—S.sub.2-Mo] motifs in Fe-sMoS.sub.2 at the materials interface.

[0204] The dynamic N.sub.2 reduction to NH.sub.3 over Fe-sMoS.sub.2 was also studied using in-situ ATR-FTIR with light illumination (FIGS. 14A and 15). The IR absorption bands of 3303 and 1634 cm.sup.−1 shown in FIG. 14A can be attributed to the O—H stretching and H—O—H bending of adsorbed water molecules, respectively on the catalyst structure. Their decreasing signals from background as a result of the consumption of the adsorbed water molecules upon the light illumination in N.sub.2. Simultaneously, four bands at 1431, 1278, 1106, and 956 cm.sup.−1 were arisen, which can be attributed to the H—N—H bending, —NH.sub.2 wagging, —NH.sub.2 twisting, and N—N stretching of adsorbed N.sub.2Hy (2≤y≤4) species, respectively.sup.36. Notably, the latter species suggest that the N.sub.2 reduction on the Fe-sMoS.sub.2 may follow the association pathway under the light illumination. The content of hydrazine was analysed using para-(dimethylamino) benzaldehyde acidic solution, which gave a small but detectable peak at around 450 nm in UV-vis spectroscopy, as shown in FIG. 16. This indicates that the formation of N.sub.2H.sub.4 from N.sub.2 reduction, which forms a complex with the benzaldehyde compound.sup.37. Thus, the Fe-sMoS.sub.2 appears to undertake the same association pathway for N.sub.2 fixation as that of nitrogenase with both structures containing the common motifs of four membered [Fe—S.sub.2-Mo] rings.

[0205] To prove the unique feature of [Fe—S.sub.2-Mo] in photocatalytic ammonia production, the activities of some selected first-row transition metal analogues were compared and are shown in FIG. 14B. The same volcano activity relationship for typical ammonia production rate from N.sub.2 reduction with respect to d orbital filling and position at the optimal value of Fe is presented. Time resolved photoluminescence (TRPL) spectra of sMoS.sub.2, Mn, Fe, Co, and Ni-doped sMoS.sub.2 are also shown in FIGS. 14B and 17. As seen from the TRPL spectra, the instantly generated excited electrons and holes in sMoS.sub.2 annihilate rapidly within a few nanoseconds. Doping single transition metal atoms onto this structure apparently increases their recombination time, suggesting that the metal exerts an enhanced degree of charge separation by accepting excited electrons. Mn and Ni-doped sMoS.sub.2 show a similar exciton lifetime, followed by Co-sMoS.sub.2. Interestingly, Fe-doped sMoS.sub.2 with optimal d-band filling and position also gives the longest excitons lifetime with the slowest PL decay curve. Notably, the rank of their lifetimes (Table 4) shows a strong inverse relationship with photocatalytic activity for ammonia production (FIG. 14B). It is anticipated that the prolonged excitons lifetime is critical to allow chemical reactions of the excitons to occur before they recombine for relaxation, leading to photocatalytic N.sub.2 fixation. Thus, the Fe-doped sMoS.sub.2 with the Fe—S.sub.2-Mo motifs displays the best combination of metal site and ‘electron relay’ components for charge separation analogously to that in the biological system.

TABLE-US-00004 TABLE 4 Fitted decay time from TRPL over transition metal-doped and undoped MoS.sub.2 Sample τ.sub.1 τ.sub.2 T(ns) sMoS.sub.2 0.72 — 0.72 Mn—sMoS.sub.2 0.51 3.02 2.47 Fe—sMoS.sub.2 0.61 4.31 3.79 Co—sMoS.sub.2 0.32 3.33 3.11 Ni—sMoS.sub.2 0.47 2.93 2.44 T (average), τ.sub.1, and τ.sub.2 represents the decay times for two components, which is fitted with a double-exponential model using equation of I(t) = A.sub.1exp(−t/τ.sub.1) + A.sub.2 exp(−t/τ.sub.2) when d orbital is introduced. On the contrary, sMoS.sub.2 was fitted with one exponential model.

3.4. Mechanistic Studies

[0206] For N.sub.2 activation over nitrogenase, it was suggested from theoretical calculations that N.sub.2 could linearly bind to either the molybdenum atom over the distal pathway (hydrogenation starts at terminal N), or the iron atom over the alternating pathway (hydrogenation starts at N in proximity to Fe) in the FeMoco.sup.9. The electron states of HOMO and LUMO and band structures in Mn, Fe, Co, and Ni-doped sMoS.sub.2 were then modelled (FIGS. 14C, 18, and 19).

[0207] As shown in FIG. 18, the HOMO and LUMO orbitals concentrate on the edge of sMoS.sub.2 with relatively low electron delocalization, verifying the highly active edge site of s-MoS.sub.2 as that reported in literature. Transition metal atom doped distinctly improves the degree of delocalization of the frontier orbitals, especially to their LUMO, the frontier orbitals delocalization follows the order: Fe>Co>Mn≈Ni. The higher degree of delocalization indicate the more stable population of photo-excited electrons in LUMO orbitals, thus accounting the longer lifetime for the recombination of excited photo-generated electrons and photo-generated holes. This is in good agreement with the TRPL experimental results. Among them, the LUMO orbital distribution over the Fe atom in Fe doped sMoS.sub.2 should be noted (FIG. 14D). Particularly, they demonstrate that excited electrons could be transferred from valence band to conduction band of sMoS.sub.2 via the conductive Fe—S.sub.2-Mo motifs and resided on to the Fe atom during the photo-exciting process to enter to the anti-bonding orbital of absorbed N.sub.2 molecule and thus facilitating the hydrogenation reaction of N.sub.2 for ammonia production. On the other hand, the density of state calculation indicates a smaller band gap of Fe doping sMoS.sub.2 relative to other catalyst materials, which is also favorable for electron transfer from HOMO to LUMO by photo-excitation.

[0208] Clearly, excited electrons from CB of sMoS.sub.2 after photo-excitation show a strong propensity to transfer and accommodate at Fe.sub.1 atom than other transition metals. According to further DFT calculations, it was also found that wherever N.sub.2 was placed on Fe.sub.1 atom doped sMoS.sub.2 slab, the N.sub.2 adsorption was always converged onto the Fe.sub.1 atom in [Fe—S.sub.2-Mo] as the end on mode spontaneously (FIG. 20). Once the electronegative N.sub.2 moiety is taken up by the Fe.sub.1 atom excited electrons during visible light illumination are expected to retain to further prolong the exciton lifetime for subsequent protons reduction to ammonia on the N.sub.2-Fe.sub.1 against the typical fast recombination of excitons from this layer structure, which substantially promotes the N.sub.2 to NH.sub.3 reaction over H.sub.2O photolysis without in contact with nitrogen gas. In addition, the nitrogen fixation to ammonia on the Fe.sub.1 over [Fe—S.sub.2-Mo] appeared to go through the alternating pathway (FIGS. 21 and 22), indicating the similarity in mechanism for both non-biological and biological processes in ammonia synthesis.

[0209] It is generally recognized that ammonia synthesis at nitrogenase follows an associative pathway without breaking N≡N triple bonds directly in transition state. N.sub.2 adsorption and the following first proton and electron reactions of adsorbed N.sub.2 (formation of *N.sub.2H) are two key steps in this non-dissociative reduction of N.sub.2. The energy plots in FIG. 14D and FIG. 21 then show that the first hydrogenation step by adding hydrogen atom is the most challenging step with the energy going uphill. Hydrogenating the proximal N to Fe is found to be less favourable with a higher energy state whereas, the species from hydrogenating the terminal N is relative stable. The following hydrogenation steps can be separated into two pathways: distal and alternating as shown in the FIG. 21. The intermediate species via the latter pathway is more stable compared with that via the former pathway. Based on the energy plots, the nitrogen fixation to ammonia over [Fe-S.sub.2-Mo] sites through the alternating pathway appears to be more energetically favoured despite the higher activation barrier in the first hydrogenation step.

[0210] Quantum efficiency (Q.E.) for photon to hydrogen in ammonia is the key parameter to evaluate the conversion efficiency of renewable light energy. FIG. 14D shows that the Q.E. of this nitrogenase-mimic Fe-sMoS.sub.2 can be up to 3.5% at 436 nm, which is believed to be the highest value reported in photo-ammonia synthesis.

2.5. Conclusion

[0211] In conclusion, a bio-inspired solid structure consisting of nitrogenase-like [Fe—S.sub.2-Mo] four membered rings in 2D single layer of MoS.sub.2 is for the first time synthesized. The material mimics the nitrogenase enzyme, which shows the strong ability to reduce N.sub.2 to NH.sub.3 in aqueous solution under mild conditions with visible light illumination where excited electrons from the sMoS.sub.2 slab are conducted to the redox active Fe site through the [Fe—S.sub.2-Mo] as the electron relay units. Introduction of light-sensitive CdS quantum dots can further boost the NH.sub.3 harvest. From DFT calculations and ATR-FTIR analysis, the [Fe—S.sub.2-Mo] motif is clearly shown to carry out an associative mechanism in converting N.sub.2 to NH.sub.3. N.sub.2 prefers to bind linearly on the Fe atom in the [Fe—S.sub.2-Mo], which will undergo stepwise hydrogenations to NH.sub.3 with the successive formation of hydrogen atom from H.sup.+/e.sup.−pairs. Thus, the photocatalytic method for ammonia synthesis over this type of materials although small in quantity may open up an exciting possibility for the decentralization of ammonia supply for fertilizer to local farmlands.

[0212] While specific embodiments of the invention have been described herein for the purpose of reference and illustration, various modifications will be apparent to a person skilled in the art without departing from the scope of the invention as defined by the appended claims.

REFERENCES

[0213] 1. Avenier, P. et al. Dinitrogen dissociation on an isolated surface tantalum atom. Science 317, 1056-1060 (2007).

[0214] 2. Liu, J. et al. Nitrogenase-mimic iron-containing chalcogels for photochemical reduction of dinitrogen to ammonia. Proc. National Acad. Sci. 113, 5530-5535 (2016).

[0215] 3. IFA, 2007 Annual Production and International Trade Statistics. (Series of statistical reports on the 2007 production capacity, production and international trade of key fertilizers, raw materials and intermediates see www.fertilizer.org/ifa/HomePage/STATISTICS. 2008.

[0216] 4. Hoffman, B. M. et al. Mechanism of nitrogen fixation by nitrogenase: the next stage. Chem. Rev. 114,4041-4062 (2014).

[0217] 5. Gray, H. B. & Ellis Jr, W. R. Electron transfer. Bioinorganic Chemistry. Sausalito, Calif.: University Science Books, 315-363. (1994).

[0218] 6. Rod, T. H., Logadottir, A. & N∅rskov, J. K. Ammonia synthesis at low temperatures. J. Chem. Phys. 112, 5343-5347 (2000).

[0219] 7. Hoffman, B. M. et al. Nitrogenase: a draft mechanism. Acc. Chem. Res. 46, 587-595 (2013).

[0220] 8. Rittle, J., McCrory, C. C. L. & Peters, J. C. A 106-fold enhancement in N.sub.2-binding affinity of an Fe.sub.2(μ-H).sub.2 core upon reduction to a mixed-valence Fe.sup.IIFe.sup.I state. J. Am. Chem. Soc. 136, 13853-13862 (2014).

[0221] 9. Rittle, J. & Peters, J. C. An Fe—N.sub.2 complex that generates hydrazine and ammonia via Fe□N═NH.sub.2: demonstrating a hybrid distal-to-alternating pathway for N.sub.2 reduction. J. Am. Chem. Soc. 138, 4243-4248 (2016).

[0222] 10. Buscagan, T. M., Oyala, P. H. & Peters, J. C. N.sub.2-to-NH.sub.3 conversion by a triphos—iron catalyst and enhanced turnover under photolysis. Angew. Chem. Int. Ed. 56, 6921-6926 (2017).

[0223] 11. Li, Y. et al. Ammonia formation by a thiolate-bridged diiron amide complex as a nitrogenase mimic. Nature Chem. 5, 320-326 (2013).

[0224] 12. Broda, H. & Tuczek, F. Catalytic ammonia synthesis in homogeneous solution—biomimetic at last? Angew. Chem. Int. Ed. 53, 632-634 (2014).

[0225] 13. Eda. G. et al. Coherent atomic and electronic heterostructures of single-layer MoS.sub.2. ACS Nano 6, 7311-7317 (2012).

[0226] 14. Reddy, D. A. et al. Hydrazine-assisted formation of ultrathin MoS.sub.2 nanosheets for enhancing their co-catalytic activity in photocatalytic hydrogen evolution. J. Mater. Chem. A 5, 6981-6991 (2017).

[0227] 15. van der Ham, C. J. M., Koper, M. T. M. & Hetterscheid, D. G. H. Challenges in reduction of dinitrogen by proton and electron transfer. Chem. Soc. Rev. 43, 5183-5191 (2014).

[0228] 16. Harald, F., Chukalina, M. & Scheinost, A. C. A new FEFF-based wavelet for EXAFS data analysis. J. Synch. Rad. 14, 426-432 (2007).

[0229] 17. Gangwar, J. et al. Time-resolved and photoluminescence spectroscopy of θ-Al.sub.2O.sub.3 nanowires for promising fast optical sensor applications. Dalton Trans. 43, 17034-17043 (2014).

[0230] 18. Li, M. M. J. et al. Importance of the structural integrity of a carbon conjugated mediator for photocatalytic hydrogen generation from water over a CdS-carbon nanotube-MoS.sub.2 composite. Chem. Commun. 52, 13596-13599 (2016).

[0231] 19. Hidalgo, R. et al. Infrared spectroscopy during electrocatalytic turnover reveals the Ni-L active site state during H.sub.2 oxidation by a NiFe hydrogenase. Angew. Chem. Int. Ed. 54, 7110-7113 (2015).

[0232] 20. Kresse, G. & Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B 47, 558-561 (1993).

[0233] 21. Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865-3868 (1996).

[0234] 22. Blochl, P. E. Projector augmented-wave method. Phys. Rev. B: Condens. Matter Mater. Phys. 50, 17953-17979 (1994).

[0235] 23. Monkhorst, H. J. & Pack, J. D. Special points for Brillouin-zone integrations. Phys. Rev. B: Solid State 13, 5188-5192 (1976).

[0236] 24. Fei, H. et al. Atomic cobalt on nitrogen-doped graphene for hydrogen generation. J. Bio. Ino. Chem. 20, 447-460 (2015).

[0237] 25. R. Bjornsson, et al., The discovery of Mo (III) in FeMoco: reuniting enzyme and model chemistry. J. Bio. Ino. Chem. 20, 447-460 (2015).

[0238] 26. van der Ham, C. J. M., Koper, M. T. M. & Hetterscheid, D. G. H. Challenges in reduction of dinitrogen by proton and electron transfer. Chem. Soc. Rev. 43, 5183-5191 (2014).

[0239] 27. Reddy, D. A. et al. Hydrazine-assisted formation of ultrathin MoS.sub.2 nanosheets for enhancing their co-catalytic activity in photocatalytic hydrogen evolution. J. Mater. Chem. A 5, 6981-6991 (2017).

[0240] 28. Shi, H. Y. et al. Exciton dynamics in suspended monolayer and few-layer MoS.sub.2 2D crystals. ACS nano 7, 1072-1080 (2013).

[0241] 29. Brown, K. A. et al. Light-driven dinitrogen reduction catalyzed by a CdS: nitrogenase MoFe protein biohybrid. Science 352, 448-450 (2016).

[0242] 30. Lashgari, M. & Zeinalkhani, P. Ammonia photosynthesis under ambient conditions using an efficient nanostructured FeS.sub.2/CNT solar-energy-material with water feedstock and nitrogen gas. Nano Energy 48, 361-368 (2018).

[0243] 31. Hirakawa, H. et al. Photocatalytic conversion of nitrogen to ammonia with water on surface oxygen vacancies of titanium dioxide. J. Am. Chem. Soc. 139, 10929-10936 (2017).

[0244] 32. Schrauzer, G. N. & Guth, T. D. Photocatalytic reactions. 1. Photolysis of water and photoreduction of nitrogen on titanium dioxide. J. Am. Chem. Soc. 99, 7189-7193 (1977).

[0245] 33. Li, H. et al. Facet-dependent solar ammonia synthesis of BiOCl nanosheets via a proton-assisted electron transfer pathway. Nanoscale 8, 1986-1993 (2016).

[0246] 34. Li, L. et al. Nitrogen photofixation over III-nitride nanowires assisted by ruthenium clusters of low atomicity. Angew. Chem. 129, 8827-8831 (2017).

[0247] 35. Zhang, N. et al. Refining defect states in W.sub.18O.sub.49 by Mo doping: a strategy for tuning N.sub.2 activation towards solar-driven nitrogen fixation. J. Am. Chem. Soc. 140,9434-9443. (2018).

[0248] 36. Watt, G. W. & Chirsp, J. D. Spectrophotometric method for determination of hydrazine. Anal. Chem. 24, 2006-2008 (1952).

[0249] 37. Medford, A. J. et al. From the Sabatier principle to a predictive theory of transition-metal heterogeneous catalysis. J. Catal. 328, 36-42 (2015).