Electrolytic ammonia production using transition metal oxide catalyst enhanced with a transition metal dopant

Abstract

The invention relates to a process and a system for electrolytic production of ammonia. The process comprises feeding nitrogen to an electrolytic cell, where it comes in contact with a novel cathode surface, wherein said surface consists of a catalyst comprising at least one transition metal oxide doped with at least one transition metal dopant, the electrolytic cell comprising a proton donor and running a current through said electrolytic cell, whereby nitrogen reacts with protons to form ammonia.

Description

[0001] P15340PC00 ELECTROLYTIC AMMONIA PRODUCTION USING TRANSITION METAL OXIDE CATALYST ENHANCED WITH A TRANSITION METAL DOPANT FIELD OF INVENTION The disclosure is within the field of process chemistry, and specifically relates to the production of ammonia with electrolytic methods using novel doped transition metal oxide catalysts for the cathode surface. BACKGROUND Ammonia is a high value chemical owing to its use in fertilizers, with an annual production of 175 million tonnes, but its potential as an energy storage material and a carbon-free maritime fuel have recently gained attention. For more than a century, ammonia has been produced via the Haber-Bosch process, in which nitrogen and hydrogen gases react over a promoted Fe / Ru catalyst at a temperature exceeding 400°C and pressure upwards of 150 bar; N2(g) + 3H2(g) → 2NH3(g) This leads to substantial carbon emissions, owing mostly to the production of hydrogen gas via steam reforming, but also to the extreme reaction conditions rendering the process highly energy consuming. These are in stark contrast to the ambient conditions at which the enzyme nitrogenase in bacteria naturally catalyses nitrogen reduction from solvated protons, electrons and atmospheric nitrogen; N2(g) + 8H++ 8e- → 2NH3(g) + H2(g) This natural process of nitrogen fixation has inspired the search and development of catalysts that can work at (much milder) ambient conditions. However, as is common knowledge, the search for a suitable catalyst for a specific reaction is both challenging and complex, wherein a search needs to be based on factors such as availability, cost, stability, activity, and selectivity of the catalyst material under the desired operational conditions. In particular, the realisation of the electrochemical nitrogen reduction reaction (NRR) using solid catalytic electrodes in an aqueous solution, with the protons coming from water splitting at the anode, has been an active field of research, both computationally and experimentally [1]. Despite the research interest and recent efforts, little progress has been made in finding a suitable catalyst for NRR with the main culprit being the competing hydrogen evolution reaction (HER), as most of the materials that are active towards the NRR are more selective towards HER, especially at higher overpotentials. This leads to extremely low Faradaic efficiencies, i.e., most of the P15340PC00 electrical energy supplied to the system is being wasted on the HER rather than being utilized for the NRR. To emphasize the challenge of finding a suitable NRR catalyst, a few examples from the literature and from previous efforts by the applicants are discussed below. Out of the pure transition metals, namely, Sc, Y, Ti, Zr Re, Os, Co, Ru, Rh, Ir, Ni, Pd, Pt, Cu, Ag, Au, V, Nb, Ta, Cr, Mo, W, and Fe it has been predicted (with computational studies) that none of the pure transition metals is capable of catalysing the NRR because of the facile H2 formation. [2] The linear scaling relations of adsorption energies for the pure metals in the periodic table predict that they are not able to catalyse this reaction. Some of these predictions have been verified in experiments, namely for Ru, Rh and Re, with strict protocols used to avoid false positive results. [3] In fact, it is safe to say that efficient ammonia synthesis in an aqueous solution verified with strict experimental protocols remains a challenge, with the literature littered with likely false positives [3,4,5]. Recent efforts by the present applicants have focused on novel ceramics materials, i.e., using transition metal nitride (TMN), transition metal oxide (TMO) and transition metal sulphide (TMS) catalyst surfaces, as disclosed in WO2015189865, WO2019053749 and WO2020110155, respectively,. Despite these efforts, most investigated materials still catalyse HER over NRR. The identification of a promising catalyst is one of the steps needed to realise electrochemical NRR at low to moderate overpotentials. The present invention seeks to ameliorate these problems by providing novel catalyst compounds capable of catalysing the electrochemical nitrogen reduction reaction (NRR) at mild operational conditions such as in an aqueous solution, using low to moderate overpotentials / voltage. References: [1] Guo, W., Zhang, K., Liang, Z., Zou, R., and Xu, Q. (2019). Electrochemical nitrogen fixation and utilization: theories, advanced catalyst materials and system design. Chem. Soc. Rev.48, 5658-5716. [2] Skúlason, E., Bligaard, T., Gudmundsdóttir, S., Studt, F., Rossmeisl, J., Abild-Pedersen, F., Vegge, T., Jónsson, H., and Nørskov, J.K. (2012). A theoretical evaluation of possible transition metal electro-catalysts for N2 reduction. Phys. Chem. Chem. Phys.14, 1235-1245. [3] Andersen, S.Z., Colic, V., Yang, S., Schwalbe, J.A., Nielander, A.C., McEnaney, J.M., Enemark-Rasmussen, K., Baker, J.G., Singh, A.R., Rohr, B.A., et al. (2019). A rigorous electrochemical ammonia synthesis protocol with quantitative isotope measurements. Nature. 570, 504-508. P15340PC00 [4] Tang, C., and Qiao, S.-Z. (2019). How to explore ambient electrocatalytic nitrogen reduction reliably and insightfully. Chem. Soc. Rev.48, 3166-3180. [5] Suryanto, B.H.R., Du, H.-L., Wang, D., Chen, J., Simonov, A.N., and MacFarlane, D.R. (2019). Challenges and prospects in the catalysis of electroreduction of nitrogen to ammonia. Nat. Catal.2, 290-296 SUMMARY OF THE INVENTION The present invention is specified in the claims as well as in the below description. In broad terms, the invention relates to the electrochemical synthesis of ammonia from molecular nitrogen, i.e., the electrochemical nitrogen reduction reaction. In one embodiment, the process may be carried out in aqueous solution and at ambient conditions at low to moderate overpotentials. As discussed above, the competitive HER and the instability of previously mentioned promising catalyst candidates (i.e., the proneness of these materials to decompose) renders the search for viable and stable catalyst that is selective towards NRR extremely challenging. However, it is found by the applicants that by using highly specific novel transition metal oxide catalysts, wherein the transition metal oxide catalysts are doped with carefully selected transition metal dopant(s) to alter the electronic properties of the transition metal oxide material to enhance the nitrogen reduction reaction occurring on the surface of the catalyst may be enhanced, while simultaneously limiting the competing hydrogen evolution reaction (HER). Moreover, given the importance of ammonia production and the energy intensive and environmentally unfavourable conditions typically required to produce ammonia, the invention finds important applicability in a plethora of industries such as, but not limited to, the production of fertilizer. In the first aspect of the present invention a process for producing ammonia is provided, wherein the process comprises the steps of: feeding N2 to an electrolytic cell that comprises at least one source of protons, allowing the N2 to come into contact with the cathode surface in the electrolytic cell, wherein the cathode surface comprises at least one catalyst surface comprising at least one transition metal oxide doped with at least one type of transition metal and running a current through said electrolytic cell, whereby nitrogen reacts with protons to form ammonia. In an embodiment of the present invention, the at least one catalyst surface comprising at least one transition metal oxide is doped with at least one type of transition metal dopant. The at least one transition metal oxide doped with at least one type of transition metal may be P15340PC00 selected from the group consisting of: Hafnium Oxide doped with Molybdenum and / or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum and / or Tungsten and / or Hafnium and / or Niobium and / or Nickel and / or Osmium and / or Ruthenium and / or Titanium and / or Vanadium and / or Zirconium, Iridium and / or Rhenium, Niobium Oxide doped with Molybdenum and / or Titanium, and / or Tungsten, Osmium Oxide doped with Rhenium and / or Tungsten and / or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt and / or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium and / or Niobium, Titanium Oxide doped with Rhenium and / or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt and / or Molybdenum and / or Niobium and / or Tantalum. In another embodiment, the at least one catalyst surface comprising at least one transition metal oxide is doped with at least one type of transition metal dopant. The at least one transition metal doped with one type of transition metal is selected from the group consisting of: Hafnium Oxide doped with Molybdenum or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum or Tungsten or Hafnium or Niobium or Nickel or Osmium or Ruthenium or Titanium or Vanadium or Zirconium, Iridium or Rhenium, Niobium Oxide doped with Molybdenum or Titanium, or Tungsten, Osmium Oxide doped with Rhenium or Tungsten or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium or Niobium, Titanium Oxide doped with Rhenium or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt or Molybdenum or Niobium or Tantalum, or preferably from a group consisting of: Molybdenum Oxide doped with Tungsten or Osmium or Titanium or Vanadium, Osmium Oxide doped with Molybdenum, Rhenium Oxide doped with Molybdenum or even more preferably from a group consisting of: Molybdenum Oxide doped with Tungsten and Rhenium Oxide doped with Molybdenum. Therefore, in accordance with the present invention, a process for producing ammonia is provided, wherein said process comprises: feeding N2 to an electrolytic cell that comprises at least one source of protons, allowing the N2 to come into contact with a cathode surface in the electrolytic cell, wherein the cathode surface comprises at least one catalyst surface comprising at least one transition metal oxide doped with at least one type of transition metal selected from the group consisting of: Hafnium Oxide doped with Molybdenum and / or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum and / or Tungsten and / or Hafnium and / or Niobium and / or Nickel and / or Osmium and / or Ruthenium and / or Titanium and / or Vanadium and / or Zirconium, Iridium and / or Rhenium, Niobium Oxide P15340PC00 doped with Molybdenum and / or Titanium, and / or Tungsten, Osmium Oxide doped with Rhenium and / or Tungsten and / or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt and / or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium and / or Niobium, Titanium Oxide doped with Rhenium and / or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt and / or Molybdenum and / or Niobium and / or Tantalum, and, running a current through said electrolytic cell, whereby nitrogen reacts with protons to form ammonia. In one embodiment, the at least one transition metal oxide doped with at least one type of transition metal dopant is selected from the group consisting of: Hafnium Oxide doped with Molybdenum or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum or Tungsten or Hafnium or Niobium or Nickel or Osmium or Ruthenium or Titanium or Vanadium or Zirconium, Iridium or Rhenium, Niobium Oxide doped with Molybdenum or Titanium, or Tungsten, Osmium Oxide doped with Rhenium or Tungsten or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium or Niobium, Titanium Oxide doped with Rhenium or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt or Molybdenum or Niobium or Tantalum. In another embodiment, the at least one transition metal oxide doped with at least one type of transition metal is selected from a group consisting of: Molybdenum Oxide doped with Tungsten or Osmium or Titanium or Vanadium, Osmium Oxide doped with Molybdenum and Rhenium Oxide doped with Molybdenum. In some embodiments, the at least one transition metal oxide doped with at least one type of transition metal may be a dioxide, i.e., XO2+Y, wherein X is a transition metal in accordance with the present invention and Y is a dopant transition metal in accordance with the present invention. However, the person of ordinary skill in the art is well aware that the stoichiometry of 1:2 (transition metal atoms to oxygen atoms) is only approximate and in practice the stoichiometry may vary substantially from 2 such as, but not limited to, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.2, 2.3, 2.4 and 2.5, or any range therebetween. In some embodiments, the stoichiometry may vary e.g., because of formation of defects. The surface coverage of the transition metal dopant may vary anywhere between a small fraction in the range of 0.1% dopant atoms / transition metal oxide atoms to a full coverage (i.e., 100% dopant atoms / transition metal oxide atoms) of the dopant atom in the surface layer, effectively creating a full dopant surface layer. In one embodiment, the surface coverage of P15340PC00 the transition metal dopant on the metal oxide surface is selected to be in the range from 5- 15%. In another embodiment, the surface coverage of the transition metal dopant on the metal oxide surface is selected to range from 20-30%. In another embodiment, the surface coverage of the transition metal dopant on the metal oxide surface is selected to range from 45% to 55%. In yet another embodiment, the surface coverage of the transition metal dopant on the metal oxide surface is selected to range from 70% to 80%. In one embodiment, the transition metal dopant covers 98% or more of the metal oxide surface. For different oxides and dopants, the amount of surface coverage may depend on the type of transition metal oxide and the type of dopant, as well as the synthesis procedure. Additionally, the bulk of the doped transition metal oxide may provide a reservoir of dopant atoms. In some embodiments, the reservoir of dopant atoms may comprise one or more types of dopant atoms. In one embodiment, the catalyst surface may comprise a mixture of two or more transition metal oxides doped with two or more different transition metals. In one embodiment, the catalyst surface may comprise at least one surface having a rutile structure, i.e., the at least one catalyst used may comprise a rutile crystal structure. In one embodiment, the catalyst surface may comprise O-terminated bridging sites or OH- terminated bridging sites or H-terminated bridging sites or a combination thereof. One region of the surface may be comprised of one termination and another region of the surface may comprise another. In one embodiment, the catalyst surface may comprise at least one surface having a (110) facet. In one embodiment, ammonia is formed in the electrolytic cell at an electrode potential less than -1.0 V, relative to the reversible hydrogen electrode (RHE). In another embodiment, ammonia is formed in the electrolytic cell at an electrode potential at about -0.5 V, relative to the reversible hydrogen electrode (RHE). In one embodiment, a cyclic varied potential may be used, such that the potential fluctuates between an active potential and a resting potential to generate a cyclic varied potential through the electrolytic cell. In one embodiment, the electrolytic cell may comprise one or more aqueous electrolytic solution. The aqueous electrolytic solution may be acidic, neutral or alkaline. However, in the preferred embodiment, the aqueous solution is neutral. P15340PC00 In one embodiment, the electrolytic solution may comprise an organic protic or aprotic solvent or a miscible mixture thereof, or preferably a water-miscible organic solvent. In one embodiment, nitrogen may be fed to the electrolytic cell by bubbling nitrogen gas to the electrolytic solution in contact with said cathode surface. In another embodiment, feeding N2to the electrolytic cell comprises feeding gaseous nitrogen or air or liquid with dissolved nitrogen to the electrolytic cell. In one embodiment, the source of protons in the formation of ammonia may be water splitting at the anode or H2oxidation reaction at the anode. In one embodiment, the temperature is selected to be in the range from -10° C to 80° C, preferably in the range 20° C to 50°C, more preferably in the range 25°C to 30°C. In the preferred embodiment, ambient temperature may be used. In one embodiment, the pressure is selected to be in the range of 1 to 30 atmospheres, preferably in the range of 1-20 atmospheres, preferably in the range of 1-10 atmospheres, more preferably in the range of 1-5 atmospheres. In the preferred embodiment, the process is operated at atmospheric pressure. In the second aspect of the present invention, a system for generating ammonia is provided, wherein the system comprising at least one electrolytic cell, which comprises at least one cathode having at least one catalytic surface, wherein the at least one catalytic surface comprising at least one transition metal oxide doped with at least one type of transition metal dopant. Therefore, in accordance with the present invention, a system for producing ammonia from nitrogen is provided, wherein the system comprises at least one electrolytic cell, which comprises at least one cathode having at least one catalytic surface, wherein the at least one catalytic surface comprises at least one transition metal oxide doped with at least one type of transition metal dopant selected from the group consisting of: Hafnium Oxide doped with Molybdenum and / or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum and / or Tungsten and / or Hafnium and / or Niobium and / or Nickel and / or Osmium and / or Ruthenium and / or Titanium and / or Vanadium and / or Zirconium, Iridium and / or Rhenium, Niobium Oxide doped with Molybdenum and / or Titanium, and / or Tungsten, Osmium Oxide doped with Rhenium and / or Tungsten and / or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt and / or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium and / or Niobium, Titanium Oxide doped with Rhenium and / or Tungsten, Vanadium Oxide P15340PC00 doped with Rhenium and Tungsten Oxide doped with Cobalt and / or Molybdenum and / or Niobium and / or Tantalum. In one embodiment, the at least one transition metal oxide doped with at least one type of transition metal dopant may be selected from the group consisting of: Hafnium Oxide doped with Molybdenum or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum or Tungsten or Hafnium or Niobium or Nickel or Osmium or Ruthenium or Titanium or Vanadium or Zirconium, Iridium or Rhenium, Niobium Oxide doped with Molybdenum or Titanium, or Tungsten, Osmium Oxide doped with Rhenium or Tungsten or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium or Niobium, Titanium Oxide doped with Rhenium or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt or Molybdenum or Niobium or Tantalum or preferably the at least one transition metal Oxide doped with at least one type of transition metal dopant may be selected from a group consisting of: Molybdenum Oxide doped with Tungsten or Osmium or Titanium or Vanadium, Osmium Oxide doped with Molybdenum, Rhenium Oxide doped with Molybdenum or even more preferably the at least one transition metal oxide doped with at least one transition metal dopant may be selected from a group consisting of: Molybdenum Oxide doped with Tungsten, Rhenium Oxide doped with Molybdenum. In one embodiment, the catalyst surface may comprise a mixture of two or more transition metal Oxides doped with two or more different transition metal dopants. In one embodiment, the catalyst surface in the system comprises at least one surface having a rutile structure. In one embodiment, the catalyst surface may comprise bridging sites with O-termination or bridging sites with OH-terminations or bridging sites with H-terminations or a combination thereof. The catalyst surface may preferably comprise at least one surface having a (110) facet. The electrolytic cell of the system may comprise one or more electrolytic solution. In one embodiment, the electrolytic cell may comprise an acidic, neutral or alkaline aqueous solution. In the preferred embodiment, the electrolytic cell comprises a neutral aqueous solution. In another embodiment, the electrolytic cell comprises an electrolytic solution comprising an organic protic or aprotic solvent, or a miscible mixture thereof, preferably a water-miscible organic solvent. P15340PC00 In one embodiment, the system is configured to produce ammonia in the electrolytic cell at an electrode potential of less than -1.1 V relative to the RHE, more preferably at an electrode potential of less than -0.5 V relative to the RHE. At high electrode potentials, e.g., -1.2 V, or even -1.0 V, a material may be prone to favour the hydrogen evolution reaction over the nitrogen reduction reaction. Accordingly, it is an advantage of the present invention that the materials may be operatable to form ammonia at low-to-moderate electrode potentials such as lower than -1.0 V, or lower than -0.8 V, or lower than -0.6 V, e.g. -0.5 V or lower. In one embodiment, a cyclic varied potential is used such that the potential fluctuates between an active potential and a resting potential to generate a cyclic varied potential through the electrolytic cell. In one embodiment, the pressure during the process selected to be in the range of 1 to 30 atmospheres, preferably in the range of 1-20 atmospheres, preferably in the range of 1-10 atmospheres, more preferably in the range of 1-5 atmospheres. In another embodiment, the pressure is selected to be atmospheric pressure during the process. BRIEF DESCRIPTION OF THE FIGURES The skilled person will understand that the figures, described below, are for illustration purposes only and provide support to the example presented below. The figures are not intended to limit the scope of the present teachings in any way. Figure 1. shows a top view of the doping patterns used, where white unmarked atoms are host metal atoms, black atoms are dopant metal atoms (marked with 'D'), and small white atoms are oxygen atoms (marked with 'O'). The four in-plane oxygen atoms have been removed for clarity. The four possible doping patterns are labelled by (a) where the two host termination atoms are substituted and referred to as Term 2 / T2 in the text, labelled by (b) wherein a single active site atom is replaced by a dopant atom and referred to as Active 1 / A1 in the text, labelled by (c) wherein both active site in the unit cell atoms are replaced by dopant atoms and referred to as Active 2 / A2 in the text, labelled by (d) wherein all four metal atoms in the surface layer are substituted and referred to as All 4 / A4 in text. Figure 2. shows a tilted side view of the different adsorbates studied in the example, with only the top layer of the catalyst shown for clarity. The large and small white circles denote metallic and oxygen atoms (marked with 'M' and 'O'), respectively, nitrogen atoms are shown as black, and hydrogen atoms are denoted by dotted black circles. P15340PC00 Figure 3. shows a top view of the three different surface terminations that are studied for each host-dopant combination in the example. The large / small white circles denote metal / oxygen atoms, while hydrogen atoms are denoted by black dotted circles. The four in plane oxygen atoms from the top layer of the catalyst have been removed for clarity. Figure 4. shows the relative stability of the O (dashed line), OH (dotted line) and H (dashed-dotted line) terminations of the surface bridge sites on MoO2referenced to a surface where the bridge sites are vacant (solid black line). The boxes at the top of the figure signify which termination is expected to be present at a given applied potential. Figure 5. shows the predicted surface termination as a function of applied potential for each host metal oxide. For the potential region marked with 'V' for RhO2 vacant bridge sites are favoured. Figure 6. shows the adsorption free energy of N2 plotted against the free energy of H adsorption at U = 0 V vs. RHE for different surface terminations. In each subfigure the boxed region marked as 'Promising region' includes surfaces where the adsorption of N2 is favoured while H adsorption is endothermic. Materials in the bottom right quadrant that fall outside of the box marked 'Promising region' bind N2 unrealistically strongly compared to H, which is explained by structure reconfigurations during relaxation. The large squares show the pure TMOs while the smaller circles denote doped metal oxides. In Figure 6(a), the results are shown for the O-terminated surfaces. In Figure 6(b), the results are shown for the OH- terminated surfaces. In Figure 6(c), the results are shown for the H-terminated surfaces. Figure 7. shows a volcano-plot of the predicted limiting potential of the NRR for the three different surface terminations of the host-dopant combinations that met the first screening criteria (bind N2 selectively at U = 0 V vs. RHE), plotted against a surface reactivity descriptor. The marker type denotes the host metal oxide for each point (see legend). The pure TMOs are shown as large squares marked with the corresponding element. In each subfigure the inset shows a larger view of the top of the volcano, i.e. the materials with a PDS smaller than 0.85 eV, with the dopant atom marked for each point. In Figure 7(a), the results are shown for O-terminated surfaces. In Figure 7(b), the results are shown for the OH-terminated surfaces. In Figure 7(c), the results are shown for the H-terminated surfaces. Figure 8. shows the number of doping patterns that meet the first two screening criteria for each host-dopant combination for the different surface terminations. In Figure 8(a) the results are shown for O-terminated surfaces. In Figure 8(b) the results are shown for OH- terminations. In Figure 8(c), the results are shown for H-termination. A white box denotes no promising doping pattern, a black box denotes one promising doping pattern, and a box with P15340PC00 dashed black lines denotes three promising doping patterns in Figure 8(a) and two promising doping patterns in Figure 8(b) and Figure 8(c). Figure 9. shows the free energy landscape for the vertical NRR pathway on MoO2, WO2, and doping patterns for the MoO2+W host-dopant combination. The v subscript denotes a vertical configuration of a given adsorbate. The magnitude of the potential determining step, which for these surfaces (except for WO2) is the formation of NNHvfrom N2v, is typeset in bold. For WO2, the release of the second NH3(g) molecule is uphill in free energy, reflecting the fact that WO2 overbinds nitrogen containing species. It should be noted that the MoO2 pathway is the one closest to the top and the WO2pathway is the one closest to the bottom of the figure. Figure 10. shows the free energy landscape for the horizontal NRR pathway on MoO2, WO2, and doping patterns for the MoO2+W host-dopant combination. The horizontal pathway, which is a possible route when two adjacent active sites are vacant, goes through NNHh where the h subscript denotes a horizontal configuration. The NNHh intermediate can be reached via a reconfiguration of either N2v to N2h or NNHv to NNHh. The NNHv intermediate is in all cases higher in free energy than the N2h intermediate. Only a pathway going through NNHv is shown for MoO2 as it does not bind N2 in a horizontal configuration. The remaining steps for the horizontal pathway have already been shown for the vertical mechanism. For visual clarity, the T2 and A1H doping patterns are not included. It should be noted that the MoO2 pathway is the one closest to the top and the WO2 pathway is the one closest to the bottom of the figure. Figure 11 shows a volcano-plot of the predicted limiting potential of the NRR for the horizontal pathway of the O terminated surfaces of the host-dopant combinations that met the first screening criteria (bind N2 selectively at U = 0 V vs RHE), plotted against a surface reactivity descriptor. The marker type denotes the host metal oxide for each point (see legend). The pure TMOs are shown as large squares marked with the corresponding element. In each subfigure the inset shows a larger view of the top of the volcano, in this case the materials with a PDS smaller than 0.50 eV, with the dopant atom marked for each point. Figure 12. shows the adsorption energies of N2, H and NNHv−N2v for MoO2, WO2 and the MoO2+W host-dopant combinations compared with those for MoO2 surfaces doped with subsurface W atoms. W2nd means that a Mo atom in the subsurface layer is replaced with a W atom while W3rd means that a Mo atom in the next lowest (third from top) layer is replaced with a W atom. Figure 13. shows the free energy of adsorption for OH and H2O at U=0 V vs. RHE for MoO2, WO2, and all doping patterns for the MoO2+W host-dopant combination. Water P15340PC00 adsorption is in all cases endothermic, while the adsorption of OH is exothermic in all cases, with the adsorption strength increasing from MoO2to WO2. Figure 14. shows the free energy landscape for the vertical NRR pathway on MoO2, WO2, and doping patterns for the MoO2+W host-dopant combination at an applied potential of -0.5 V vs. RHE. The v subscript denotes a vertical configuration of a given adsorbate. The magnitude of the potential determining step, which for these surfaces (except for WO2) is the formation of NNHv from N2v, is typeset in bold. It should be noted that the MoO2 pathway is the one closest to the top and the WO2pathway is the one closest to the bottom of the figure. Figure 15. shows the adsorption free energy of N2 plotted against the free energy of H adsorption at the predicted NRR limiting potential for the O terminated surfaces. In this case, almost all the simulated surfaces are predicted to bind H favourably. The large squares show the pure TMOs while the smaller circles denote doped metal oxides. Figure 16. summarizes the results of the Example (discussed below) in a tabular format. DETAILED DESCRIPTION In the following, exemplary embodiments of the invention will be described. These embodiments are provided to provide further understanding of the invention, without limiting its scope. In the following description, a series of steps may be described. The skilled person will appreciate that unless required by the context, the order of steps is not critical for the resulting configuration and its effect. Further, it will be apparent to the skilled person that irrespective of the order of steps, the presence or absence of time delay between steps, can be present between some or all of the described steps. As used herein, including in the claims, singular forms of terms are to be construed as also including the plural form and vice versa, unless the context indicates otherwise. Thus, it should be noted that as used herein, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. Throughout the description and claims, the terms “comprise”, “including”, “having”, and “contain” and their variations should be understood as meaning “including but not limited to” and are not intended to exclude other components. The present invention also covers the exact terms, features, values and ranges etc. in case these terms, features, values and ranges etc. are used in conjunction with terms such as P15340PC00 about, around, generally, substantially, essentially, at least etc. (i.e., "about 3" shall also cover exactly 3 or "substantially constant" shall also cover exactly constant). The term “at least one” should be understood as meaning “one or more”, and therefore includes both embodiments that include one or multiple components. Furthermore, dependent claims that refer to independent claims that describe features with “at least one” have the same meaning, both when the feature is referred to as “the” and “the at least one”. It will be appreciated that variations to the foregoing embodiments of the invention can be made while still falling within the scope of the invention. Features disclosed in the specification, unless stated otherwise, can be replaced by alternative features serving the same, equivalent or similar purpose. Thus, unless stated otherwise, each feature disclosed represents one example of a generic series of equivalent or similar features. Use of exemplary language, such as “for instance”, “such as”, “for example” and the like, is merely intended to better illustrate the invention and does not indicate a limitation on the scope of the invention unless so claimed. Any steps described in the specification may be performed in any order or simultaneously unless the context clearly indicates otherwise. All of the features and / or steps disclosed in the specification can be combined in any combination, except for combinations where at least some of the features and / or steps are mutually exclusive. In particular, preferred features of the invention are applicable to all aspects of the invention and may be used in any combination. The present invention is based on a large-scale materials study over a wide range of transition metal oxide compounds / candidates wherein the transition metal oxide surfaces comprise different surface terminations and have been doped with various transition metals (referred to as dopants), in various patterns and with a different degree of doping. The study allowed promising catalytic materials for NRR to be identified and found that it is possible to form ammonia in an aqueous solution, at ambient temperature and pressure, with a low applied potential. The invention provides a process and a system for generating ammonia at ambient temperature and pressure. In the process and system of the present invention, an electrolytic cell is used which can be any of a range of conventional commercially suitable and feasible electrolytic cell designs that can accommodate a special purpose cathode in accordance with the invention. Thus, the cell and system may in some embodiments have one or more cathode cells and one or more anode cells. P15340PC00 An electrolytic cell in the present context is an electrochemical cell that undergoes a redox reaction when electrical energy is applied to the cell. The skilled person will appreciate that chemical compounds as described herein are provided by their chemical formula irrespective of their phase or state. In particular, compounds that are present in their gaseous state when present in a pure and isolated form at room temperature (such as N2, H2and NH3) are herein described by their chemical formula. For example, dinitrogen is herein described as N2, whether present as nitrogen gas, as individual molecules, in clusters, bound to surfaces or present as solutes, and the same applies to other molecular species described herein. The proton donor can be any suitable substance that is capable of donating protons in the electrolytic cell. The proton donor can for example be an acid, such as any suitable organic or inorganic acid. The proton donor can be provided in an acidic, neutral or alkaline aqueous solution. The proton donor can also, or alternatively, be provided by H2 oxidation at the anode, i.e., hydrogen can be considered as a source of protons: H2 ^ 2(H++ e-). The electrolytic cell comprises at least three general parts or components, a cathode, an anode, and an electrolyte. The overall cathode reaction can be presented as N2 + 6(H++ e-) ^ 2NH3 The nitrogen species adsorbed on the catalyst surface can be hydrogenated by adding one hydrogen atom at a time, representing a proton from the solution and an electron from the electrode surface. The reaction mechanism can be shown in the following chemical equations below describing the so-called associative mechanism, where an asterisk denotes a surface site: *NNH2+ 4(H++ e−) ⇔ *N + NH3(g) +3(H++e−) [3Aa] *NHNH + 4(H++ e−) ⇔ *NHNH2+ 3(H++ e−) [3B] *N + 3(H++ e−) ⇔ *NH + 2(H++ e−) [4A] P15340PC00 *NHNH2+ 3(H++ e−) ⇔ *NH + NH3(g)+ 2(H++ e−) [4Ba] For the dissociative mechanism the reaction mechanism is according to the below equations: *NH2 + (H++ e−) ⇔ * + NH3(g)

[0013] The different parts or components can be provided in separate containers, or they can be provided in a single container. The electrolyte can be an aqueous solution in which ions are dissolved. The aqueous solution can be a neutral, an alkaline or an acidic solution. In some embodiments, the aqueous solution is a neutral solution. In some embodiments, the electrolyte can also be a molten salt, for example a sodium chloride salt. In general terms, the catalyst on the electrode surface should ideally have the following characteristics: It should (a) be chemically stable, it should (b) not become reduced or otherwise consumed during the electrolytic process, it should (c) facilitate the formation of ammonia, and (d) use of the catalyst should lead to the production of minimal amount of hydrogen gas and (e) it should require a low to moderate overpotential. As will be further P15340PC00 described, the doped transition metal oxide catalysts according to the invention fulfil these characteristics. As further illustrated and discussed herein, the catalyst in the process and system of the invention comprises in some embodiments one or more transition metal oxides selected from the group consisting of Hafnium Oxide, Iridium Oxide, Molybdenum Oxide, Niobium Oxide, Osmium Oxide, Platinum Oxide, Rhenium Oxide, Rhodium Oxide, Ruthenium Oxide, Tantalum Oxide, Titanium Oxide, Vanadium Oxide, Tungsten Oxide. These transition metal oxides may be doped by one or more of the dopants selected from the group consisting of Cobalt, Hafnium, Iridium, Molybdenum, Niobium, Nickel, Osmium, Rhenium, Ruthenium, Tantalum, Titanium, Vanadium, Tungsten and Zirconium. Any mixtures and combinations of two or more of these doped transition metal oxides are also applicable in the invention. Accordingly, a mixture or combination of two or more of these catalyst materials may be used to lower the required electrode potential. An advantage of the present invention is that the process can be suitably operated using aqueous electrolytes, such as preferably aqueous solutions with dissolved electrolytes (salts). Thus, in preferred embodiments of the process and system, the electrolytic cell comprises one or more aqueous electrolytic solutions, in one or more cell compartments. Aqueous electrolyte solutions may comprise any of various typical inorganic or organic salts such as but not limited to soluble salts of chloride, nitrate, chlorate bromide, etc. e.g., sodium chloride, potassium chloride, calcium chloride, ammonium chloride, and other suitable salts. The aqueous electrolyte solutions may also comprise anyone, or a combination of, alkali or alkaline earth metal oxides, such as sodium hydroxide, potassium hydroxide, lithium hydroxide, calcium hydroxide, rubidium hydroxide and caesium hydroxide. The aqueous electrolyte solution can also further, or alternatively, comprise one or more organic or inorganic acids. Inorganic acids can include mineral acids that include but are not limited to, hydrochloric acid, nitric acid, phosphoric acid, sulphuric acid, boric acid, hydrofluoric acid, hydrobromic acid, and perchloric acid. The electrolyte can also comprise an organic solvent, preferably an organic solvent miscible in water that is mixed in an aqueous electrolyte. As appears from herein, the essential feature of the present invention concerns the composition and structure of the cathode. Doped transition metal oxides can have a wide variety of surface structures which affect the surface energy of these compounds and influence their chemical properties. The relative acidity and basicity of the atoms present on the surface of metal oxides are also affected by the coordination of the metal cation and the oxide, which alter the catalytic properties of these compounds. P15340PC00 Gas diffusion electrodes (GDE) may be used to facilitate access of gases, in particular N2, to and from the electrodes of the present invention. In one embodiment, the cathode and anode may be a GDE. In another embodiment, only the cathode is a GDE. The GDE may comprise at least one catalyst layer for facilitating and / or catalysing the electrochemical NRR reaction. The GDE may further comprise a gas diffusion layer to provide a pathway for the gas to the electrode surface and hinder and / or delay flooding of the electrode. The GDE may further comprise a gas diffusion media to enable the transport of gases to and from the gas diffusion layer. The gas diffusion media may be hydrophobic and / or microporous to allow passage of gas while repelling liquid electrolyte. Depending on the substance composition of the catalyst, a suitable surface crystal structure may be preferred. Various crystal structures exist for (doped) transition metal oxides and different structures can be obtained at different growth conditions of the catalyst. In some embodiments the catalyst surface comprises at least one surface having a rutile structure. The at least one surface having O-terminated bridge sites or OH-terminated bridge sites or H-terminated bridge sites or a mixture thereof. The catalyst surface comprises in the preferred embodiments at least one surface having an exposed (110) facet. Other crystal structures and facets are as well encompassed within the scope of the invention (see., e.g., International Tables for Crystallography; http: / / it.iucr.org). The surface coverage of dopant atoms on the transition metal oxide surface can vary anywhere from being only a small percentage (less than 1%) to a full coverage of dopant atoms in the surface layer. The surface coverage of dopant atoms in the surface layer of the transition metal oxide may vary between 0.1% and 25%, or vary between 25% and 50%, or vary between 50% and 75%, or vary between 75% to 99%. Preferably, the surface of the doped transition metal oxide may include finite regions of different doping patterns, e.g., regions wherein one or more termination atoms are replaced with dopant atoms and / or regions wherein one or more active site atoms are replaced with a dopant atom. Different regions of doping patterns may lie adjacent to each other or be separated by regular, non-doped, regions of the transition metal oxide surface. During the synthesis of a catalyst, it is not possible to control exactly which doping patterns are formed, but with the host transition metal and the dopant transition metal preferably having similar atomic properties, they may become at least partially miscible, enabling a mixture of host and dopant transition metals, and the surface to be likely to contain patches resembling, but not limited to, all of the doping patterns discussed in the example below, with the relative abundance of each pattern depending on the atomic ratios used during synthesis. Alternatively, the host transition metal atoms of the surface layer of the transition metal oxide may be entirely replaced by dopant atoms. For example, P15340PC00 Molybdenum Oxide material with a surface layer of tungsten dopant atoms. Moreover, dopant atoms may additionally be arranged in the bulk of the transition metal oxide and / or migrate to the bulk of the transition metal oxide. The exact amount of dopant atoms in the bulk of the transition metal oxide depends on the atomic ratios used during synthesis of the catalyst material and the nature of the host transition metal oxide and the dopant. As described in more detail herein, running a current through the electrolytic cell leads to a chemical reaction in which nitrogen reacts with protons to form ammonia. The running of current is achieved by applying a voltage to the cell. The invention makes possible electrolytic production of ammonia at a low electrode potential, which is beneficial in terms of energy efficiency and minimal equipment demands. In certain useful embodiments of the invention, ammonia can be formed at an electrode potential at less than about -1.2 V vs RHE, less than about -1.0 V vs RHE, less than about - 1.0 V vs RHE, less than about -0.9 V vs RHE, less than about -0.8 V vs RHE, less than about -0.7 V vs RHE, less than about -0.6 V vs RHE, less than about -0.5 V vs RHE or less than about -0.4 V vs RHE, less than about -0.3 V vs RHE. In some embodiments, ammonia can be formed at electrode potential in the range of about -0.2 V to about -1.2 V vs RHE, such as in the range of about -0.3 V to about -0.8 V vs RHE, such as in the range of about -0.4 V to about -1.2 V vs RHE, or in the range of about -0.5 V to about -1.2 V vs RHE. The upper limit of the range can be about -0.6 V, about -0.7 V vs RHE, about -0.8 V vs RHE, about -0.9 V vs RHE, about –1.0 V vs RHE, or about -1.1 V vs RHE, about -1.2 V vs RHE. The lower limit of the range can be about -0.2 V vs RHE, about -0.3 V vs RHE, about -0.4 V vs RHE, about -0.5 V vs RHE or about -0.6 V vs RHE. In other words, the process may be operated to form ammonia using the above-mentioned electrode potentials. The present invention may be used with a pulsing electrode potential by applying a cyclic varied potential that fluctuates between an active potential and a resting potential to said electrolytic cell to generate a cyclic varied potential through the electrochemical cell. In one embodiment, the cyclic varied potential is applied in the form of, but not limited to, sinusoidal varied potential, step potential, and intermittent cyclic varied potential. Alternatively, in some embodiments, a pulsing electrode potential can be used by applying a cyclic varied current. The cyclic varied potential comprises active potential pulses in the range from about 0.001 sec. to about 30 sec. with intermittent system specific resting potential for a period of time in the range from 2 sec. to about 180 sec. The cyclic varied potential or current is configured to minimize or eliminate adsorption on electrodes of intermediate species, and the resting potential and resting potential time-period are selected to release adsorbed intermediate P15340PC00 species adsorbed during the active potential time-period. In particular, the cyclic varied potential or current is configured to minimize or eliminate adsorption on electrodes of intermediate species and the resting potential and resting potential time-period are selected to release adsorbed intermediate species adsorbed during the active potential time-period. For example, at anodic potentials, hydrogen and / or protons adsorbed on the catalyst electrode surfaces can be removed from the surface thereby steering the selectivity of the materials towards the NRR. An advantage of the present invention is the efficiency of NH3 formation over H2 formation, i.e., the competition between NRR and HER, which as discussed above has been a challenge in prior art investigations and trials. In certain embodiments of the invention, less than about 50% moles H2 are formed compared to moles NH3 formed, and preferably less than about 40% moles H2, less than about 30% moles H2, less than about 20% moles H2, less than about 10% moles H2, less than about 5% moles H2, less than about 2% moles H2, or less than about 1% moles H2. The system of the invention is suitably designed to accommodate one or more of the above process features. It is an advantage of the invention that the system can be made small, robust and cheaply, such that fertilizer can be produced close to the intended site of use. Ammonia can be used as such as fertilizer, by injecting into soil as gas, although this requires investment by farmers in pressurized storage tanks and injection machinery. Ammonia can also be used to form urea, typically by reacting it with carbon dioxide. Ammonia can be reacted to form nitric acid, which in turn is readily reacted to form ammonium nitrate. Accordingly, systems and processes of the present invention can be readily combined with present solutions for reacting the produced ammonia to other desired products such as but not limited to the abovementioned. NOx and SOx are generic terms for mono-nitrogen and mono-sulphur oxides, such as NO, NO2, SO, SO2 and SO3. These gases are produced during combustion, especially at high temperatures. In areas of high motor vehicle traffic, the amount of these pollutants can be significant. Accordingly, a useful aspect of the invention relates to a system for removing NOx and / or SOx from a stream of gas, by reacting the stream of gas with ammonia that is generated in situ in the stream, or in a system that can be fluidly connected to the stream of gas. The system can comprise a system for generating ammonia as described herein, in particular a system that comprises an electrolytic cell containing a doped transition metal oxide catalyst as described herein. In this context, in situ should be understood as ammonia generation within the system, P15340PC00 for example within the gas stream, or in a compartment within the system that is fluidly connected to the gas stream. The ammonia thus generated, when in contact with the stream of gas, will react with NOxand / or SOxin the stream of gas to convert these toxic species to other molecular species, such as N2, H2O and (NH4)2SO4. In some embodiments, the system can be for use in an automobile engine exhaust or in other engines, where ammonia can be generated in situ by a process according to the present invention, and then used to reduce SOxand / or NOxexhaust gases from the engine. Such system can suitably use electric current produced by conversion from the car engine. Thus, by using electric current from a car engine, ammonia can be generated in situ, and the generated ammonia can thus be allowed to react with SOx and / or NOx from the gas exhaust of the automobile. The ammonia can be generated in the automobile, and subsequently fed into the car exhaust. The ammonia can also be generated in situ within the automobile exhaust system. Thereby, NOx and / or SOx are removed from the car exhaust, reducing the number of pollutants in the exhaust. SELECTED SPECIFIC EMBODIMENTS OF THE INVENTION In the following preferred embodiments, phrases like “approximately 50% surface coverage” refer to a range of surface coverages, i.e., the fraction of dopant atoms to host transition metal atoms (in percentages) arranged in the surface layer, wherein the range surrounds 50%. Hence, an “approximately 50% surface coverage” may mean in the range of 40% to 60% surface coverage, or in the range 45% to 55% surface coverage, or in the range 48% to 52% surface coverage. An “approximately 100% surface coverage” may mean somewhere in the range of 90% to 100% surface coverage, or 95% to 100% surface, or 98% to 100% surface coverage. Moreover, as the skilled person will recognize, the dopant atoms may comprise one or more types of dopant atoms, wherein most of the dopant atoms arranged in the surface is comprised of the mentioned dopant metal. For example, Molybdenum Oxide doped with Osmium is a Molybdenum Oxide that is primarily doped with Osmium, however, this does not exclude that other dopants may be present in the surface and / or in the bulk of the transition metal oxide in smaller amounts. In one embodiment, other types of dopant atoms may be introduced into the doped transition metal oxide as pollutants during synthesis of the catalyst and / or during the operation of the catalyst. In an embodiment, wherein Hafnium Oxide doped with Molybdenum is selected as at least one doped transition metal oxide catalyst, the surface coverage of Molybdenum may be approximately 100%. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to, termination transition metal atoms (also referred to as fully coordinates atoms or bridging atoms) and transition metal active atoms (also referred P15340PC00 to as coordinatively unsaturated sites) being replaced by the dopant transition metal. In one embodiment, the surface may be at least partially O-terminated. In an embodiment, wherein Hafnium Oxide doped with Osmium is selected as at least one doped transition metal oxide catalyst, the surface coverage of Molybdenum may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated. In an embodiment, wherein Hafnium Oxide doped with Molybdenum is selected as at least one doped transition metal oxide catalyst, the surface coverage of Molybdenum may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H-terminated. In an embodiment, wherein Iridium Oxide doped with Rhenium is selected as at least one doped transition metal oxide catalyst, the surface coverage of Rhenium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O- terminated. In an embodiment, wherein Molybdenum Oxide doped with Tantalum is selected as at least one doped transition metal oxide catalyst, the surface coverage of Tantalum may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated. In an embodiment, wherein Molybdenum Oxide doped with Tungsten is selected as at least one doped transition metal oxide catalyst, the surface coverage of Tungsten may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In another embodiment, the Molybdenum Oxide doped with Tungsten may have approximately 50% surface coverage of Tungsten. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active transition metal atoms being replaced by transition metal P15340PC00 dopant atoms. The Molybdenum Oxide doped with Tungsten may have approximately 100% surface coverage, wherein the Molybdenum Oxide doped with Tungsten may comprise various doping patterns such as, but not limited to, termination transition metal atoms and / or transition metal active atoms being replaced by the dopant transition metal. In these embodiments, the surface may be at least partially O-terminated. In an embodiment, wherein Molybdenum Oxide doped with Hafnium is selected as at least one doped transition metal oxide, the surface coverage of Hafnium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may comprise at least partial OH-termination. In an embodiment, wherein Molybdenum Oxide doped with Niobium is selected as at least one doped transition metal oxide, the surface coverage of Niobium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH- terminated. In an embodiment, wherein Molybdenum Oxide doped with Nickel is selected as at least one doped transition metal oxide, the surface coverage of Nickel may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH-terminated. In an embodiment, wherein Molybdenum Oxide doped with Osmium is selected as at least one doped transition metal oxide, the surface coverage of Osmium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In another embodiment, the Molybdenum Oxide doped with Osmium may have approximately 50% surface coverage. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active termination transition metal atoms being replaced by transition metal dopant atoms. In these embodiments, the surface may be at least partially OH-terminated. In an embodiment, wherein Molybdenum Oxide doped with Ruthenium is selected as at least one doped transition metal oxide, the surface coverage of Ruthenium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such P15340PC00 as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH- terminated. In an embodiment, wherein Molybdenum Oxide doped with Titanium is selected as at least one doped transition metal oxide, the surface coverage of Titanium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In another embodiment, the Molybdenum Oxide doped with Titanium may have approximately 50% surface coverage. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily termination transition metal atoms being replaced by transition metal dopant atoms. In these embodiments, the surface may be at least partially OH-terminated. In an embodiment, wherein Molybdenum Oxide doped with Vanadium is selected as at least one doped transition metal oxide, the surface coverage of Vanadium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In another embodiment, the Molybdenum Oxide doped with Vanadium may have approximately 50% surface coverage. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily termination transition metal atoms being replaced by transition metal dopant atoms. In these embodiments, the surface may be at least partially OH-terminated. In an embodiment, wherein Molybdenum Oxide doped with Zirconium is selected as at least one doped transition metal oxide, the surface coverage of Zirconium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may comprise at least partial OH-termination. In an embodiment, wherein Molybdenum Oxide doped with Iridium is selected as at least one doped transition metal oxide, the surface coverage of Iridium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H-terminated. In an embodiment, wherein Molybdenum Oxide doped with Niobium is selected as at least one doped transition metal oxide, the surface coverage of Niobium may be approximately P15340PC00 100%. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to regions of termination transition metal atoms and active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H-terminated. In an embodiment, wherein Molybdenum Oxide doped with Osmium is selected as at least one doped transition metal oxide, the surface coverage of Osmium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H- terminated. In an embodiment, wherein Molybdenum Oxide doped with Rhenium is selected as at least one doped transition metal oxide, the surface coverage of Rhenium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H- terminated. In an embodiment, wherein Molybdenum Oxide doped with Tungsten is selected as at least one doped transition metal oxide, the surface coverage of Tungsten may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H- terminated. In an embodiment, wherein Niobium Oxide doped with Molybdenum is selected as at least one doped transition metal oxide, the surface coverage of Molybdenum may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O- terminated and / or OH-terminated and / or H-terminated. In an embodiment, wherein Niobium Oxide doped with Titanium is selected as at least one doped transition metal oxide, the surface coverage of Titanium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH- terminated. P15340PC00 In an embodiment, wherein Niobium Oxide doped with Tungsten is selected as at least one doped transition metal oxide, the surface coverage of Tungsten may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H-terminated. In an embodiment, wherein Osmium Oxide doped with Rhenium is selected as at least one doped transition metal oxide, the surface coverage of Rhenium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated. In an embodiment, wherein Osmium Oxide doped with Tungsten is selected as at least one doped transition metal oxide, the surface coverage of Tungsten may be approximately 100%. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to regions of both active transition metal atoms and termination transition metal atoms being replaced by transition metal dopant atoms. In another embodiment, the Osmium Oxide doped with Tungsten may have approximately 25% surface coverage. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated and / or at least partially OH-terminated. In an embodiment, wherein Osmium Oxide doped with Molybdenum is selected as at least one doped transition metal oxide, the surface coverage of Molybdenum may be approximately 25% or 50% surface coverage. The approximately 25% or 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH-terminated. In another embodiment, the Osmium Oxide doped with Molybdenum may have approximately 100% surface coverage of Molybdenum. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to, regions of both active transition metal atoms and termination transition metal atoms being replaced by dopant atoms. In this embodiment, the surface may be at least partially H-terminated. In an embodiment, wherein Platinum Oxide doped with Rhenium is selected as at least one doped transition metal oxide, the surface coverage of Rhenium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but P15340PC00 not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH-terminated. In an embodiment, wherein Rhenium Oxide doped with Cobalt is selected as at least one doped transition metal oxide, the surface coverage of Cobalt may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated. In an embodiment, wherein Rhenium Oxide doped with Molybdenum is selected as at least one doped transition metal oxide, the surface coverage of Molybdenum may be approximately 25% or 50% surface coverage. The approximately 25% or 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated and / or at least partially H-terminated. In another embodiment, the Rhenium Oxide doped with Molybdenum may have approximately 100% surface coverage of Molybdenum. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to, both active transition metal atoms and termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated and / or at least partially H-terminated. In an embodiment, wherein Rhodium Oxide doped with Rhenium is selected as at least one doped transition metal oxide, the surface coverage of Rhenium may be approximately 100%. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to regions of both active transition metal atoms and termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated and / or at least partially OH-terminated. In an embodiment, wherein Ruthenium Oxide doped with Rhenium is selected as at least one doped transition metal oxide, the surface coverage of Rhenium may be approximately 100%. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to regions of active transition metal atoms and termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated and / or at least partially H-terminted. In another embodiment, the Ruthenium Oxide doped with Rhenium may have approximately 50% surface coverage of Rhenium. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to, regions of primarily active transition metal atoms being replaced by P15340PC00 dopant transition metal atoms. In one embodiment, the surface may be at least partially OH- terminated. In an embodiment, wherein Tantalum Oxide doped with Vanadium is selected as at least one doped transition metal oxide, the surface coverage of Vanadium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH-terminated. In an embodiment, wherein Tantalum Oxide doped with Niobium is selected as at least one doped transition metal oxide, the surface coverage of Niobium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H-terminated. In an embodiment, wherein Titanium Oxide doped with Tungsten is selected as at least one doped transition metal oxide, the surface coverage of Tungsten may be approximately 100%. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to regions of both active transition metal atoms and termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated and / or at least partially OH terminated. In an embodiment, wherein Titanium Oxide doped with Rhenium is selected as at least one doped transition metal oxide, the surface coverage of Rhenium may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH-terminated and / or at least partially H-terminated. In an embodiment, wherein Vanadium Oxide doped with Rhenium is selected as at least one doped transition metal oxide, the surface coverage of Rhenium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated. In an embodiment, wherein Tungsten Oxide doped with Cobalt is selected as at least one doped transition metal oxide, the surface coverage of Cobalt may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated. P15340PC00 In an embodiment, wherein Tungsten Oxide doped with Molybdenum is selected as at least one doped transition metal oxide, the surface coverage of Molybdenum may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O- terminated. In an embodiment, wherein Tungsten Oxide doped with Niobium is selected as at least one doped transition metal oxide, the surface coverage of Niobium may be approximately 50%. The approximately 50% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially O-terminated and / or at least partially H-terminated. In another embodiment, the tungsten Oxide doped with Niobium may have approximately 100% surface coverage of Niobium. The approximately 100% surface coverage may comprise various doping patterns such as, but not limited to, regions of both active transition metal atoms and termination transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially OH-terminated. In an embodiment, wherein Tungsten Oxide doped with Tantalum is selected as at least one doped transition metal oxide, the surface coverage of Tantalum may be approximately 25%. The approximately 25% surface coverage may comprise various doping patterns such as, but not limited to regions of primarily active transition metal atoms being replaced by transition metal dopant atoms. In one embodiment, the surface may be at least partially H-terminated. The invention will now be illustrated by the following non-limiting example that further describes particular advantages and embodiments of the present invention. EXAMPLE In this example, DFT calculations are used to investigate the NRR catalytic properties of 13 transition metal oxides (TMOs) in the rutile structure (host metal oxides), doped with 14 different transition metals (dopant metals), for a total of 182 combinations. The transition metal oxides considered in this example are Hafnium Oxide (HfO2), Iridium Oxide (IrO2), Molybdenum Oxide (MoO2), Niobium Oxide (NiO2), Osmium Oxide (OsO2), Platinum Oxide (PtO2), Rhenium Oxide (ReO2), Rhodium Oxide (RhO2), Ruthenium Oxide (RuO2), Tantalum Oxide (TaO2), Titanium Oxide (TiO2), Vanadium Oxide (VO2) and Tungsten Oxide (WO2). The dopant atoms may be selected as Cobalt (Co), Hafnium (Hf), Iridium (Ir), Molybdenum (Mo), P15340PC00 Niobium (Nb), Nickel (Ni), Osmium (Os), Rhenium (Re), Ruthenium (Ru), Tantalum (Ta), Titanium (Ti), Vanadium (V), Tungsten (W) and Zirconium (Zr). There are two types of metallic surface atoms in the modelled unit cell of the transition metal oxides, two fully coordinated that are referred to here as the termination atoms (also bridge atoms in the literature) and two with a dangling bond referred to here as the active site atoms, but commonly known as the coordinatively unsaturated site (CUS). For each combination of a host metal oxide and a dopant metal, four different doping patterns of the metallic surface atoms are explored: (i) Termination-2, also referred to as T2, wherein two termination atoms are replaced with dopant atoms. (ii) Active-1, also referred to as A1, wherein one of the active site atoms is replaced with a dopant atom. (iii) Active-2, also referred to as A2, wherein both active site atoms are replaced with dopant atoms. (iv) All-4, also referred to as A4, wherein all four surface metal atoms are replaced with a dopant atom. The doping patterns are displayed in Figure 1. For each surface, the adsorption free energies of N2, H, NNH and NH2 are calculated and used to estimate the activity and selectivity of each surface with the help of thermodynamics descriptors. The simulated adsorbate configurations are shown in Figure 2. Two distinct active sites exist when the A1 doping pattern is simulated, as the adsorbates can either bind on the host metal or the dopant metal. Both cases are considered and referred to in what follows as Active-1-host (abbreviated as A1H) when the adsorbates bind on the host metal, and Active-1-dopant (abbreviated as A1D) when the adsorbates bind on the dopant metal. For each surface, three different surface terminations (occupancy of the bridge sites) are modelled, the oxygen-, hydroxy-, and hydrogen- terminations, marked in Figure 3 and referred to in what follows as O, OH and H terminated surfaces, respectively. The relative stability of the different terminations varies with the applied potential (U) and can be calculated using a surface with vacant bridge sites as a reference point. To predict the likely surface termination at a given applied potential, i.e. which species of O / OH / H occupies the bridge sites, the procedure is as follows; • Stoichiometrically, the (110) surface of rutile is O terminated, i.e. the surface bridge site are occupied by O atoms that can be reduced to OH via: O + H+(aq)+ e- → OH P15340PC00 As this is an electrochemical reaction step it can be driven by an applied potential, i.e., the application of a more cathodic potential drives the reaction to the right, favouring OH termination over O termination. The OH adsorbate can be further reduced to water, leaving an empty surface site, OH + H+(aq)+ e- → H2O(l)+ * where * denotes a vacant bridge site. Finally, the vacant site can be filled by a proton from the solution, * + H+(aq)+ e- → H leading to an H termination. The relative stability of the different terminations can be compared by using the surface with vacant bridge as a reference. This is done for MoO2 in Figure 4 where the black line denotes the reference surface with both the bridge sites in the unit cell left vacant. The stoichiometric O termination is predicted to be the most stable one for all potentials above ca. -0.4 V vs. RHE, at which point the H termination becomes more favourable. However, the O termination is still very stable compared to the vacant bridge site configuration used as a reference, but in order for H to adsorb on the bridge site the O atoms must first be removed. Around -0.75 V the OH termination becomes more stable than the O termination. As the OH species on the bridge sites can be formed directly from O, without the O atom vacating the bridge site, OH termination is expected from around -0.75 V to -1.25 V. At ca. -1.25 V it becomes favourable to reduce OH from the bridge site, leaving it vacant for H to adsorb on it. However, at such negative potentials the reduction of MoO2 to its metallic form would be expected, rendering the H termination relatively uninteresting for MoO2. It should be emphasized that while the change from an O to an OH termination is reversible, i.e., if the potential is cycled back and forth around the value where their predicted stability is the same (ca. -0.75 V in this case), the surface termination would be expected to continuously change (not accounting for barriers). On the other hand, if the adsorbed OH is reduced away from the bridge sites (ca. -1.25 V) and the surface becomes H terminated, a potential of almost +1 V is required to oxidize the H atoms away from the bridge sites. While this is not important for MoO2, due to the untenably negative potentials required to reduce the adsorbed OH from the surface, this is relevant for the host metal oxides where the OH termination is expected to become unstable at relatively low cathodic potentials. Finally, it should be noted that this analysis is only based on thermodynamics, so reaction barriers are ignored. However, P15340PC00 that should not be a problem as the results are only used as a rough prediction of which surface termination is likely present at a given potential. The predicted termination at any given potential of possible interest is shown in Figure 5. For moderately cathodic potentials the host metal oxides shown below to be most promising for doping (MoO2and ReO2) are predicted to be oxygen terminated. For more cathodic potentials the OH termination becomes favoured, leading in general to slightly stronger binding of adsorbates on the active sites, but not significantly different results. Potentials below -1 V vs the reversible hydrogen electrode (RHE), where ReO2 and MoO2 are expected to become H terminated, are of no practical interest due to the expected HER domination as well as the reduction of the transition metal oxide (TMO) surfaces to their pure metallic state. In this example, all three surface terminations are studied, which results in over 2000 doped transition metal oxides to be studied. The computational procedure used in this example is as follows: • Density functional theory (DFT) calculations were used to study the electronic structure of the TMO surfaces. The DFT calculations used the GGA-level RPBE exchange correlation functional as implemented in the Vienna ab initio simulation package (VASP), version 6.3.2. A plane wave basis set with a 400 eV cutoff was used to represent the wavefunctions of the valence electrons, while the projector augmented- wave method was used to describe the core electrons. A TMO geometry structure was considered fully relaxed when the residual forces on every atom were lower than 0.06 eV / Å. The relatively loose force criteria was used to speed up the calculations, as they generally take longer to converge with doped materials. However, test calculations showed no discernible difference in the results obtained using a stricter criteria. Partial occupancies were set using gaussian smearing, with a smearing width of 0.1 eV. A 4x4x1 Monkhorst-Pack k-point grid was used for the Brillouin zone sampling, and maximum symmetry was used to reduce the number of k-points in the calculations. Calculations involving all but three host metal oxides were spin paired, with spin- polarized calculations performed for MoO2, WO2and VO2. For all three, a ferromagnetic spin ordering was assumed for the initial states, but a ferromagnetic ordering is necessary for MoO2and WO2in order to get realistic lattice parameters. Spin polarization is also included for three dopant atoms, i.e Co, Ni and V, due to the known magnetic properties of these atoms. The convergence criterion for electronic steps was 10-5eV. The modelled surfaces are the (110) facet of the rutile structure, with the unit cell consisting of 48 atoms in four layers, with each layer containing 4 metal atoms and 8 oxygen atoms. The bottom two layers were constrained while the P15340PC00 upper two layers and adsorbates are allowed to relax towards a minimum. Periodic boundary conditions were employed in the x and y directions, and a 12 Å vacuum was included between the simulated slabs in the z direction. Additionally, a dipole correction was included to prevent spurious interaction between the repeated slabs. As mentioned in the main text, the surface layer has two distinct sites, the normally coordinatively saturated (and thus occupied) bridge site, onto which the O / OH / H termination atoms adsorb, and the coordinatively unsaturated ontop sites (CUS), referred to here as the active sites, onto which the adsorbates bind. Some transition metal oxides are known to be highly correlated, with a high degree of electron localization, and thus require more complex density functionals. Among the transition metal dioxides in the rutile structure, CrO2 and MnO2 fall within this category, and were therefore not included in the present study. While more accurate lattice constants and some degree of correlation would be captured if the meta-GGA level functionals such as R2SCAN (or SCAN) were used, those marginal improvements would come at a cost, as R2SCAN severely over-binds a number of molecules, including N2 and OH-. Only two of the simulated host TMOs have a non-zero band gap according to our calculations, TiO2 and HfO2. The size of the band gap for HfO2 makes it practically an insulator, and therefore not an interesting electrocatalyst. It is well known that GGA- level calculations severely underestimate band gaps, so in reality not all of the modelled metal oxides are conducting. The free energies presented in this work were constructed using a simplified electrochemical model based on thermochemistry. It takes the most important aspects of electrochemical processes into account via explicit calculations of the adsorption free energies of adsorbates on the catalyst surface, but reaction kinetics and solvation effects are ignored. The free energy of each elementary reaction step is estimated at T = 298 K according to ΔG = ΔE + ΔEZPE - TΔS where ΔE is the DFT calculated energy and ΔEZPE and ΔS are the differences in zero- point energy and entropy, respectively, between the products and reactants. For adsorbates, the latter two terms are obtained by calculating their vibrational frequencies within a harmonic approximation, whereas the zero-point energy and entropy values for gas phase molecules are taken from literature tables. The effects of an applied potential are included using the computational hydrogen electrode (CHE) by shifting the free energy of each reaction step by -neU, where n is the number of electrons being transferred ( ^^ = 1 for all NRR reaction steps), e is the elementary charge and U is the applied potential vs. RHE. Binding energies of adsorbates made P15340PC00 up of N and H atoms are obtained by referencing to the inbox energies of the N2and H2molecules, respectively. However, the same procedure cannot be used when calculating the binding energy of O containing adsorbates, as using the O2molecule as a reference would lead to seriously flawed results. This is due to the well-known over-binding of the O2molecule by GGA-level functionals, explained by the difficulties of GGA-level functionals to properly capture the energies of double and triple bonds (error cancellation makes the N2value useful). To circumvent the over-binding of O2, reference to inbox H2O is used instead, and use free energy corrections for water vapor for the vapour pressure of water at 300 K (where the Gibbs free energy of liquid water and water vapor is equivalent). In this example, the results of a high-throughput screening study, where the NRR activity and selectivity is explored for over 2000 doped transition metal oxide surfaces in the rutile structure are presented. To achieve this, four strict screening criteria that a promising catalyst material must meet are used, namely, (i) The doped transition metal oxide is found to selectively adsorb N2 over H at the hydrogen evolution reaction (HER) equilibrium potential, i.e., ΔGN2 < 0.1 eV and ΔGH > 0 eV at 0 V compared to the reversible hydrogen electrode. (ii) The limiting potential, i.e., the applied potential required for all electrochemical reaction steps to be downhill in free energy, is less than -0.85 V vs the reversible hydrogen electrode, i.e., ΔGN2→NNH < 0.85 eV and ΔGNH2→NH3 < 0.85 eV. (iii) A host-dopant combination is only considered promising if at least one doping pattern fulfils the first two screening criteria. Preferably multiple doping patterns fulfil the first two screening criteria. (iv) The surface is not prone to be poisoned by OH- or water at the applied potentials required for the nitrogen reduction reaction (NRR) to proceed. Additionally, the surface is stable towards reduction to its metallic form. What follows is a stepwise study of the >2000 catalyst candidates using the aforementioned screening criteria to find optimal catalysts capable of catalysing the electrolytic NRR. First screening step – is the surface selective towards N2 adsorption? The first step of the NRR is the non-electrochemical (no net electron transfer) adsorption of gaseous nitrogen, N2(g) → N2 P15340PC00 where instead of discerning adsorbed species with a “*“, all species are assumed to be adsorbed unless specifically labelled with (g) or (aq). The first step of the competing hydrogen evolution reaction (HER) is the electrochemical adsorption of a proton on the surface, H+(aq)+ e- → H Importantly, the adsorption strength of H can be influenced by applying an external potential, whereas the corresponding changes in the adsorption strength of N2are minimal. A potential NRR catalyst must not bind H too strongly compared to N2. Thus, the first screening criteria imposed is that at the HER equilibrium potential, U = 0 V vs. RHE, the free energy of N2 adsorption is negative while that of H is positive, as is shown in Figure 6(a) for the O- terminated surfaces. For the remainder of the example, all potentials are referred to the RHE, with the RHE notation omitted. In practice, all materials falling in the lower right quadrant of Figures 6(a-c) are deemed as promising. However, results indicating a surface that binds N2 strongly while binding H extremely weakly are not realistic, as both species bind on the same active site. Such artefacts can show up due to surface reconstruction during relaxation of unstable host-dopant combinations. Thus, the promising region is limited to materials for which the difference between the H and N2 adsorption free energy is less than 1 eV. The imposed free energy cut-off is; -1 eV < ΔGN2 < 0.1 eV, with the upper limit set to 0.1 eV instead of 0 eV in order to not eliminate promising candidates during the first screening step due to the inherent uncertainty in DFT and the approximate nature of the model system, while the lower limit is set to screen out the aforementioned unphysical results. For example, out of the 820 different materials for the O-terminated surface, 90 fall in the acceptable region and thus meet the first screening criteria of -1 eV < ΔGN2 < 0.1 eV and 0 eV < ΔGH < 1 eV. Similar, results are obtained for the OH and H terminated surfaces, as is shown in Figures 6(b-c). Importantly, none of the pristine TMOs fulfil the criteria, with N2 binding slightly too weakly on MoO2 and RuO2, much too weakly on NbO2, RhO2 and TaO2, and H binding too strongly on IrO2, OsO2, ReO2 and WO2. The nitrogen molecule desorbs from the HfO2, PtO2, TiO2 and VO2 surfaces. Thus, MoO2 and RuO2, which should not be covered by H at low overpotentials, look most ripe for doping. However, it should be noted that on RuO2, the OH and H terminations are predicted to be more stable than the O terminated one. For the OH termination, 94 out of 820 combinations fall inside the promising region, as is shown in Figure 6(b). For the H termination, as shown in Figure 6(c), 105 of the 817 combinations are deemed promising. For both the OH and H terminations, undoped RuO2 fulfils the first screening criteria, making it potentially an interesting NRR candidate. The OH terminated surfaces generally bind both N2and H stronger than the other O and H terminated ones, likely due to increased hydrogen bonding in the presence of the hydroxyl adsorbate. P15340PC00 Second screening step – Is the candidate an active NRR catalyst? Despite being an important property of a potential NRR catalyst, selectively binding N2at U=0 V does not automatically result in a promising catalyst, the material must also catalyse the reaction itself. When assessing the catalytic activity of a surface towards a given electrochemical reaction, a useful measure is the proton coupled electron transfer step with the highest uphill change in free energy, i.e., the most endothermic step, which is referred to as the potential determining step (abbreviated as PDS). The limiting potential is defined as the applied potential at which all reaction steps become thermoneutral or downhill in free energy (i.e., making the process spontaneous) and is given by the negative of the PDS. In most cases, the PDS for the NRR is either the formation of NNHvfrom N2v, where the subscript v denotes a vertical adsorption configuration, or the formation of NH3(g)from NH2[ 1,2,3], i.e., N2v + H+(aq) + e- → NNHv or NH2 + H+(aq) + e- → NH3(g) Therefore, the PDS of the remaining candidates is taken as the more endothermic one of these two reactions steps. Figures 7(a-c) show the calculated limiting potential for all the materials that fulfil the first screening criteria, with the binding energy of NNHv used as a descriptor for surface reactivity, i.e., a strong binding of NNHv hints at a highly reactive surface and vice versa. Figure 7(a) shows the results for the calculations on the O-terminated surfaces. Figure 7(b) shows the results of the calculations on the OH-terminated surfaces. Figure 7(c) shows the results of the calculations on the H-terminated surfaces. An optimal catalyst must be reactive enough so that the adsorption of N2(g) is favourable, and the formation of NNHv is not too endergonic, but not so reactive that the desorption of NH3(g) becomes endergonic. This is evident in the volcano-like shape of Figure 7(a), where the predicted NRR activity increases (the limiting potential becomes less negative) as ΔENNHV decreases (the reactivity of the surface increases), before peaking at values where ΔENNHV is approximately - 0.5 eV, after which it starts decreasing again, due to the surface becoming too reactive, i.e., it becomes hard to get rid of the NH3(g) product. The inset in Figure 7(a) shows the 21 candidates that bind N2 selectively over H at U = 0 V and have a PDS of less than 0.85 eV, i.e., they fulfil the criteria for NRR activity. Of the pure host metal oxides (shown by square markers), only four pure metal oxides have a predicted NRR limiting potential of less than -1 V, namely, WO2, ReO2, TaO2 and MoO2. As PtO2 does not bind N2 its seemingly promising limiting potential is not of interest. For the OH- and H-terminations, shown in Figure 7(b) and 7(c), respectively, 23 and 16 candidates meet the criteria. The predicted limiting potential on RuO2is around - P15340PC00 1.5 V for all three terminations, rendering what seemed to be a potentially promising catalyst (from criteria 1) impractical. The -1.5 V limiting potential is very similar to what has previously been found both computationally and experimentally (though under the limit of detection) [4] on Ru(0001) [1,5]. For the OH-terminated MoO2and NbO2, the predicted limiting potentials are around -0.5 V (as shown in Figure 7(b)). While this is not relevant for NbO2, due to weak N2 binding, this could be of interest for MoO2, which is predicted to become OH-terminated at around -0.75 V. However, as previously noted, this is only an approximate value, with the OH termination possibly being stabilized at less cathodic potentials. Third screening step – Are there multiple promising doping patterns? For practical realization of NRR, the most interesting combinations are those where more than a single doping pattern is predicted to be promising. In Figures 8(a-c), the number of promising doping patterns for each host-dopant combination is visualized for all three surface terminations. For host-dopant combinations given by dashed / dotted tiles, more than one doping pattern is found to be promising, while for those given in black only a single doping pattern is promising. White tiles show host-dopant combinations with no promising patterns. For the O-terminated surfaces, a few of host-dopant combinations have at least one promising doping pattern as is shown in Figure 8(a), namely, • Hafnium Oxide (HfO2) with Molybdenum (Mo) or Osmium (Os). • Iridium Oxide (IrO2) with Rhenium (Re). • Molybdenum Oxide (MoO2) with Tantalum (Ta) or Tungsten (W). • Niobium Oxide (NbO2) with Molybdenum (Mo). • Osmium Oxide (OsO2) with Rhenium (Re) or Tungsten (W) • Platinum Oxide (PtO2) shows no promising candidate for the O-terminated surfaces. • Rhenium Oxide (ReO2) with Cobalt (Co) or Molybdenum (Mo). • Rhodium Oxide (RhO2) with Rhenium (Re). • Ruthenium Oxide (RuO2) with Rhenium (Re). • Tantalum Oxide (TaO2) shows no promising candidate for the O-terminated surfaces. • Titanium Oxide (TiO2) with Tungsten (W). • Vanadium Oxide (VO2) with Rhenium (Re). P15340PC00 • Tungsten Oxide (WO2) with Cobalt (Co) or Molybdenum (Mo) or Niobium (Nb). For the O-terminated surfaces, two host-dopant combinations have more than one promising doping pattern as is shown in Figure 8(a), namely, • Molybdenum Oxide (MoO2) with tungsten (W) with three combinations • Rhenium Oxide (ReO2) with Molybdenum (Mo) with three combinations. For the OH-terminated surfaces, a few host-dopant combinations have at least one promising doping pattern as is shown in Figure 8(b), namely, • Hafnium Oxide (HfO2) with no promising candidate for the OH-terminated surfaces. • Iridium Oxide (IrO2) with no promising candidates for the OH-terminated surfaces. • Molybdenum Oxide (MoO2) with Hafnium (Hf) or Niobium (Nb) or Nickel (Ni) or Osmium (Os) or Ruthenium (Ru) or Titanium (Ti) or Vanadium (V) or Zr (Zirconium). • Niobium Oxide (NbO2) with Molybdenum (Mo) or Titanium (Ti). • Osmium Oxide (OsO2) with Molybdenum (Mo) or Tungsten (W). • Platinum Oxide (PtO2) with Rhenium (Re). • Rhenium Oxide (ReO2) with no promising candidate for the OH-terminated surfaces. • Rhodium Oxide (RhO2) with Rhenium (Re). • Ruthenium Oxide (RuO2) with Rhenium (Re). • Tantalum Oxide (TaO2) with Vanadium (V). • Titanium Oxide (TiO2) with Rhenium (Re) or Tungsten (W). • Vanadium Oxide (VO2) with no promising candidate for the OH-terminated surfaces. • Tungsten Oxide (WO2) with Niobium (Nb). For the OH-terminated surfaces, four host-dopant combinations have more than one promising doping pattern as is shown in Figure 8(b), namely, • Molybdenum Oxide (MoO2) with Osmium (Os) (two combinations) or Titanium (Ti) (two combinations) or Vanadium (V) (two combinations) • Osmium Oxide (OsO2) with Molybdenum (Mo) (two combinations) P15340PC00 For the H-terminated surfaces, a few of host-dopant combinations have at least one promising doping pattern as is shown in Figure 8(c), namely, • Hafnium Oxide (HfO2) with Molybdenum • Iridium Oxide (IrO2) with no promising candidate for the H-terminated surfaces. • Molybdenum Oxide (MoO2) with Iridium (Ir) or Niobium (Nb) or Osmium (Os) or Rhenium (Re) or Tungsten (W). • Niobium Oxide (NbO2) with Molybdenum (Mo) or Tungsten (W). • Osmium Oxide (OsO2) with Molybdenum (Mo). • Platinum Oxide (PtO2) with no promising candidate for the H-terminated surfaces. • Rhenium Oxide (ReO2) with Molybdenum (Mo). • Rhodium Oxide (RhO2) with no promising candidate for the H-terminated surfaces. • Ruthenium Oxide (RuO2) with Rhenium (Re). • Tantalum Oxide (TaO2) with Niobium (Nb) • Titanium Oxide (TiO2) with Rhenium (Re). • Vanadium Oxide (VO2) with no promising candidate for the H-terminated surfaces. • Tungsten Oxide (WO2) with Molybdenum (Mo) or Tantalum (Ta). For the H-terminated surfaces, one host-dopant combination has more than one promising doping pattern as is shown in Figure 8(c), namely, • Rhenium Oxide (ReO2) with Molybdenum (Mo) with two combinations. As is seen by the above results, for the O-terminated surfaces, shown in Figure 8(a), 21 promising host-dopant pattern combinations are found. Moreover, two host-dopant combinations have more than one promising pattern, namely MoO2 and W with three (four if WO2 + Mo is included) and ReO2 and Mo also with three, but those will be further discussed below. It is noteworthy that the same elements are found to be promising as hosts and as dopants, with Mo being the dopant in six instances and W / Re in five instances each, while the only hosts for which more than two combinations are promising are MoO2, ReO2 and WO2. For the OH-termination, shown in Figure 8(b), 23 promising host-dopant pattern combinations are found. Moreover, nearly half the candidates (11 / 23) involve doping of MoO2, with the promising doping patterns in all cases being T2 (wherein the two termination atoms are P15340PC00 replaced with dopant atoms, i.e.50% surface coverage of dopant atoms) and / or A1H (wherein a single active site atom is replaced with a dopant atom, i.e.25% surface coverage of dopant atoms and wherein the adsorbate binds to the remaining host atom) meaning that the active sites where the reaction takes place are still occupied by Mo atoms. Both doping patterns (T2 and A1H) are promising when MoO2is doped with Ti, V and Os. This is explained by the fact that while binding H too strongly (-0.11 eV) to meet the first screening criteria, pure OH terminated MoO2has a very high predicted NRR activity, or a PDS of 0.5 eV. Thus, doping that slightly weakens the H binding strength results in candidates fulfilling the first two screening criteria. The only other host-dopant combination found to be promising for more than one doping pattern is that of OsO2 with Mo, either A1D (wherein one of the active site atoms is replaced with a dopant atom or a 25% surface coverage of dopant atoms and the adsorbate binds to the dopant) or A2 (wherein both of the active site atoms are replaced with a dopant atom or a 50% surface coverage of dopant atoms). While the OH termination is predicted to be stable on OsO2 at moderately cathodic potential, its strong H binding, and the scarcity of Os prevent OsO2 doped with Mo to be considered as a viable NRR catalyst. For the H- terminated surfaces, as shown in Figure 8(c), 16 promising host-dopant-pattern combinations are found. Moreover, the only host-dopant combination where more than one doping pattern is promising is that of ReO2 doped with Mo. However, the H terminated is not expected on ReO2 for potentials exceeding -1 V, where the reduction of ReO2 to Re would be highly likely. Of the pure host metal oxides, only RhO2, PtO2 and RuO2 are predicted to be H terminated at low cathodic potentials (up to around -0.5 V) with IrO2 and OsO2 also predicted to become H terminated before -1 V. None of the promising patterns for the H-terminated surfaces include the three host metal oxides most likely to be H-terminated, i.e., RhO2, PtO2 and RuO2, but OsO2 doped with Mo is the only combination where H termination is somewhat likely. However, as mentioned earlier, the price of osmium and the binding strength of H on OsO2 partially prevent it from being considered an economical candidate. The above discussed results are further summarized in Figure 16. Free energy landscape of MoO2 + W – are there other important reaction steps? Of the two most promising O terminated host-dopant combinations, MoO2+W and ReO2+Mo, the former is a more viable potential catalyst, due to the significantly weaker hydrogen binding of pristine MoO2 than ReO2, and the fact that based on Pourbaix diagrams, MoO2 should be more resilient to reduction towards its pure metal state. Also, ReO2 is already predicted to become OH terminated at around -0.5 V, as opposed to -0.75 V for MoO2. To make sure that the chosen activity descriptors are adequate, i.e., that no important reaction steps are being overlooked, the complete free energy landscape of the NRR mechanism on MoO2, WO2and P15340PC00 all doping patterns where MoO2is doped with W is calculated, with the corresponding free energy diagram shown in Figure 9. Importantly, the PDS is in all cases either the formation of NNHvfrom N2vor the formation of NH3(g)from NH2, validating their choice as measures of catalytic activity. For all the surfaces but WO2, the first protonation step is potential determining, but its magnitude varies considerably based on the doping pattern. The trend displayed in Figure 9, where the free energy change for the formation of NNHv from N2v decreases when going from MoO2to WO2, with the final protonation step simultaneously becoming more challenging, nicely demonstrates the challenge of rationally designing an active NRR catalyst. As the nitrogen affinity of the catalyst increases, and thus the binding of N2 becomes stronger, facilitating the first protonation step, the release of ammonia from the surface becomes energetically more challenging. Can the PDS be circumvented via a horizontal mechanism? When two neighbouring active sites are vacant the adsorbed N2 species can reconfigure from a vertical to a horizontal binding position, N2v → N2h In previous scientific work by the inventors [6] it was shown that even for W(110), for which the reaction energy of the reconfiguration step is exothermic by 0.7 eV (i.e., N2h binds 0.7 eV stronger than N2v), the calculated activation energy is 0.34 eV. As the reconfiguration of N2 is not an electrochemical step, the reaction energy cannot be tuned using an applied potential. Thus, it is possible to conclude that the reconfiguration of N2 is highly unlikely to occur on surfaces where the reaction step is endergonic, and even for surfaces where the reaction step is exergonic, there will be a significant accompanying energy barrier. Nevertheless, in theory the possibility of N2 reconfiguration unlocks a different reaction pathway in which the NNHv intermediate is circumvented. Figure 10 shows the corresponding free energy landscape for most of the MoO2+W host-dopant combinations and WO2 (N2h is not stable on MoO2), but those are among the most promising candidates for this pathway as shown in Figure 11. The T2 doping pattern is excluded for clarity, and the difference between A1D and A1H is minimal as both active sites participate in the bonding process. It is evident that for the MoO2+W A1D and A2 doping patterns the reconfiguration step is endothermic, rendering it highly unlikely, whereas for MoO2+W A4 and WO2 it is exothermic so the possibility of a reconfiguration cannot be excluded. However, for the latter two surfaces the formation of NHNH2 becomes potential determining, with an equal (A4) or larger (WO2) PDS than for the vertical pathway. Thus, it is evident that the predicated limiting potentials for the horizontal pathway are at best similar to those obtained for the kinetically much more facile (as discussed below) vertical pathway. P15340PC00 Importantly, the formation of NHNH2cannot be circumvented, as evidenced by the calculated NHNH scission barriers in this example of 0.52 eV and 0.93 eV for WO2and MoO2, respectively. Forth screening step – How stable is W doped MoO2against reconstruction and OH- poisoning? During the synthesis of a catalyst, it is not possible to control exactly which doping patterns are formed but with the atomic properties of Mo and W being quite similar, as the two share a group in the periodic table, they should be highly miscible. Thus, a mixture of MoO2 doped with W is likely to contain patches resembling all the doping patterns, with the relative abundance of each pattern depending on the atomic ratios used during synthesis. With the simulated doping patterns only involving atoms in the surface layer, it is of interest to see how the adsorption properties change when subsurface Mo atoms in MoO2 are replaced by W atoms. Figure 12 shows a comparison of the calculated adsorption free energy of N2v, H and the free energy of the first protonation step, i.e., N2v to NNHv, for the previously calculated MoO2-W host-dopant surfaces, and when subsurface Mo atoms are replaced with W atoms. The effects are small, with the replacement leading to slightly stronger adsorbate binding. To obtain a rough estimate of the likelihood of dopant W atoms migrating between the surface and bulk of MoO2, the total energy of the system with a single W atom at the surface is compared to the total energy when a single W atom is in the second layer. The subsurface W atom is only 0.13 lower in energy, indicating that the MoO2+W host-dopant combination is very stable towards reconstruction via W migration to the bulk. To assess the stability of the MoO2+W surfaces against poisoning from OH- and H2O species from the solution, the adsorption free energy of both species is calculated, with the results shown in Figure 13. While the adsorption of water is endothermic in all cases, the adsorption free energy of OH- ranges from approximately -0.3 eV to -0.8 eV at 0 V, with the adsorption strength increasing with W doping. Thus, at U = 0 V, OH- poisoning could in theory be a problem, while at more cathodic potentials the adsorption of OH- will be endothermic. This underlines the importance of the choice of electrolyte, with neutral electrolytes (i.e., pH of approximately 7) being preferable, where the activity of both H+and OH- is orders of magnitude smaller than the saturation concentration of N2 in an aqueous solution. At moderately negative potentials and in a neutral electrolyte, the poisoning of the surface by OH- could therefore be alleviated. Is NRR likely to prevail over HER on W doped MoO2? P15340PC00 To assess the viability of the NRR on the W doped MoO2surfaces, it must be considered whether additional energy barriers are present, in particular for the most endothermic reaction step, i.e., the PDS. Previous calculations on Ru(0001) [5] and W(110) [6] surfaces showed that the energy barrier obtained for the formation of NNHvfrom N2vis purely thermodynamical, i.e., the NNHvadsorbate is the highest energy state, as well as, the final state. The same behaviour can be expected for the similar reaction step (no relocation or reconfiguration) of going from NNHvto NNH2v. Thus, at applied potentials where the formation of NNH2vfrom NNHvis exothermic, NNHvbecomes the intermediate that is highest in total free energy. To make this even more clear, Figure 14 depicts a free energy diagram for the same host-dopant combinations as in Figure 9, but at an applied potential of U = -0.5 V. In that case, the thermodynamic barriers for NRR on the O-terminated MoO2+W host-dopant combinations are on the order of 0.2-0.4 eV, depending on the doping pattern, the lower half of which is easily surmountable at ambient conditions. For the OH-terminated MoO2+W host-dopant combinations, the NRR activity follows a reverse pattern to that found for the O-termination, with the PDS increasing from around 0.5 eV (for A1D) to 1.1 eV (for A4) with more W doping. This trend is explained by desorption of NH3(g) becoming potential determining when the ratio of surface W atoms in increased. The calculated limiting potentials for the NRR can be compared to those expected for the HER. On pure transition metals, the Tafel mechanism of the HER, which is of Langmuir- Hinshelwood type, is found to be dominant at low overpotentials [7]. There, two adsorbed H atoms combine and desorb as H2(g) in a non-electrochemical reaction step, H + H → H2(g) The Tafel mechanism is highly unlikely to occur on the TMOs due to the comparatively large distance between neighbouring active sites (compared to the pure transition metals). For higher overpotentials, the Heyrovsky mechanism, which is an Eley-Rideal type mechanism, succeeds the Tafel mechanism as the dominant one on pure transition metals (Skúlason, 2010). Contrary to the Tafel mechanism, the desorption of H2(g) is electrochemical in the Heyrovsky mechanism, H + H+(aq) + e- → H2(g) In previous scientific work on transition metal nitrides (TMNs) [8] and on Niobium oxynitride (NbON) thin films [9], the observed onset potentials of hydrogen evolution on relatively bad HER catalysts were at least -0.5 V, but an overpotential of that magnitude is required for the Heyrovsky mechanism to proceed [7]. Based on the fact that the HER on TMOs is expected to take place via the Heyrovsky mechanism, and on the previously reported HER P15340PC00 overpotentials for other transition metal ceramics, there is a chance that certain combinations of MoO2doped with W could be somewhat selective towards the NRR over the HER at moderate overpotentials. The selectivity of the candidates could be tuned towards NRR using potential pulsing, where a switch to more anodic potentials would be utilized to remove adsorbed H. At more cathodic potentials the surface could become OH terminated, leading to somewhat less negative limiting potentials in the case of MoO2. In Figure 15, the adsorption free energy of N2is plotted against the adsorption free energy of H at the predicted NRR limiting potential. As is to be expected, almost all surfaces selectively adsorb H in that case, underlining the importance of some type of potential pulsing. While the HER will likely always be favoured there seems to be a possibility that at certain reaction conditions some current could go towards the reduction of nitrogen (i.e., the NRR), especially if the access to protons is successfully limited by using a neutral electrolyte and a gas diffusion electrode setup. To conclude, in this example, a high-throughput and accurate quantum chemical computational screening was carried out to look for promising nitrogen reduction catalysts among doped TMOs using transition metal dopants. In total, 13 different host TMOs were used and 14 transition metals as dopant atoms, making a total of 182 distinct host-dopant combinations, then including three different surfaces terminated and a total of four different dopant patterns simulated for each host-dopant combination, making over 2000 surfaces / candidates to be screened using a set of four strict screening criteria, listed in the discussion above. Around 100 host-dopant-pattern (out of about 800) combinations fulfil the first screening criteria for each termination. Moreover, only around 20 meet the second screening criteria as well. The most promising combination is MoO2 doped with W, which is expected to be either O or OH terminated, depending on the magnitude of the applied cathodic potential. The predicted limiting potentials vary between -0.5 V and -1.1 V for the different MoO2+W doping patterns and terminations, with the lowest ones obtained for the OH termination. Comparing these potentials to the overpotentials expected for the Heyrovsky mechanism of the HER, it can be concluded that a portion of the applied current is likely to go towards the NRR. Crucially, the applied potential should not significantly exceed -0.5 V, as such potential will always favour the HER. An important step to steer the selectivity of the materials towards the NRR is the application of potential pulsing to remove adsorbed H from the surface. Moreover, it is preferable to use a neutral electrolyte to lower the activity of H+and OH- such that these species must adsorb via water splitting, and a gas diffusion electrode setup to facilitate the access of N2 to the surface of the catalyst. P15340PC00 REFERENCES: [1] Skúlason, E., Bligaard, T., Gudmundsdottir, S., Studt, F., Rossmeisl, J., Abild-Pedersen, F., Vegge, T., Jónsson, H., and Nørskov, J.K. (2012). A theoretical evaluation of possible transition metal electro-catalysts for N2reduction. Phys. Chem. Chem. Phys.14, 1235-1245. [2] Montoya, J.H., Tsai, C., Vojvodic, A., and Nørskov, J.K. (2015). The Challenge of Electrochemical Ammonia Synthesis: A New Perspective on the Role of Nitrogen Scaling Relations. ChemSusChem.8, 2180-2186. [3] Drazevic, E., and Skúlason, E. (2020). Are There Any Overlooked Catalysts for Electrochemical NH3 Synthesis - New Insights from Analysis of Thermochemical Data. iScience.23, 101803. [4] Andersen, S.Z., Colic, V., Yang, S., Schwalbe, J.A., Nielander, A.C., McEnaney, J.M., Enemark-Rasmussen, K., Baker, J.G., Singh, A.R., Rohr, B.A., et al. (2019). A rigorous electrochemical ammonia synthesis protocol with quantitative isotope measurements. Nature. 570, 504-508. [5] Tayyebi, E., Agbhoui, Y., and Skúlason, E. (2019). Elucidating the Mechanism of Electrochemical N2 Reduction at the Ru(0001) electrode. ACS. Catal.9, 11137-11145. [6] Höskuldsson, Á.B., Tayyebi, E., and Skúlason, E. (2021). Computational examination of the kinetics of electrochemical nitrogen reduction and hydrogen evolution on a tungsten electrode. J. Catal.404, 362-370. [7] Skúlason, E., Tripkovic, V., Björketun, M.E., Gudmundsdóttir, S., Karlberg, G., Rossmeisl, J., Bligaard, T., Jónsson, H., and Nørskov, J.K. (2010). J. Phys. Chem. C.114, 18182-18197. [8] Hanifpour, F., Canales, C.P., Fridriksson, E.G., Sveinbjörnsson, A., Tryggvason, T.G., Lewin, E., Magnus, F., Ingason, Á.S., Skúlason, E., and Flosadóttir, H.D. (2022). Investigation into the mechanism of electrochemical nitrogen reduction reaction to ammonia using niobium oxynitride thin film catalyts. Electrochim. Acta.403, 139551. [9] Hanifpour, F., Canales, C.P., Fridriksson, E.G., Sveinbjörnsson, A., Tryggvason, T.G., Yang, J., Arthur, C., Jónsdóttir, S., Garden, A.L., Ólafsson, S., et al. (2022). Operando quantification of ammonia produced from computationally-derived transition metal nitride electro-catalysts. J. Catal.413, 956-967.

Claims

P15340PC00 CLAIMS 1. A process for producing ammonia, said process comprising: a. feeding N2to an electrolytic cell that comprises at least one source of protons, b. allowing the N2to come into contact with a cathode surface in the electrolytic cell, wherein the cathode surface comprises at least one catalyst surface comprising at least one transition metal oxide doped with at least one type of transition metal dopant is selected from the group consisting of: Hafnium Oxide doped with Molybdenum and / or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum and / or Tungsten and / or Hafnium and / or Niobium and / or Nickel and / or Osmium and / or Ruthenium and / or Titanium and / or Vanadium and / or Zirconium, Iridium and / or Rhenium, Niobium Oxide doped with Molybdenum and / or Titanium, and / or Tungsten, Osmium Oxide doped with Rhenium and / or Tungsten and / or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt and / or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium and / or Niobium, Titanium Oxide doped with Rhenium and / or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt and / or Molybdenum and / or Niobium and / or Tantalum, and, c. running a current through said electrolytic cell, whereby nitrogen reacts with protons to form ammonia.

2. The process according to the preceding claim, wherein said at least one transition metal oxide doped with at least one type of transition metal dopant is selected from the group consisting of: Hafnium Oxide doped with Molybdenum or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum or Tungsten or Hafnium or Niobium or Nickel or Osmium or Ruthenium or Titanium or Vanadium or Zirconium, Iridium or Rhenium, Niobium Oxide doped with Molybdenum or Titanium, or Tungsten, Osmium Oxide doped with Rhenium or Tungsten or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium or Niobium, Titanium Oxide doped with Rhenium or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt or Molybdenum or Niobium or Tantalum.P15340PC00 3. The process according to the preceding claim, wherein said at least one transition metal oxide doped with at least one type of transition metal dopant is selected from a group consisting of: Molybdenum Oxide doped with Tungsten or Osmium or Titanium or Vanadium, Osmium Oxide doped with Molybdenum and Rhenium Oxide doped with Molybdenum.

4. The process according to any of the claims 1 to 3, wherein the surface coverage of the transition metal dopant atoms on the metal oxide surface is selected to range from 1- 15% or range from 20-30% or range from 45% to 55% or range from 70% to 80%.

5. The process according to any of the claims 1 to 3, wherein the transition metal dopant covers over 98% of the metal oxide surface, effectively creating a dopant surface layer.

6. The process according to any of the preceding claims, wherein the catalyst surface comprises a mixture of two or more transition metal oxides doped with two or more different transition metals.

7. The process according to any of the preceding claims, wherein the catalyst comprises a rutile crystal structure.

8. The process according to any of the preceding claims, wherein the catalyst surface comprises at least one surface having a (110) facet.

9. The process according to any of the preceding claims, wherein ammonia is formed in the electrolytic cell at an electrode potential less than -1.0 V relative to the reversible hydrogen electrode (RHE) or at an electrode potential less than -0.5 V relative to the reversible hydrogen electrode (RHE).

10. The process according to any of claims 1 to 9, wherein a cyclic varied potential is used that fluctuates between an active potential and a resting potential to generate a cyclic varied potential through the electrolytic cell.

11. The process according to any of the preceding claims, wherein said electrolytic cell comprises one or more aqueous electrolytic solution.P15340PC00 12. The process according to any of claims 1 to 11, wherein said electrolytic cell comprises an electrolytic solution comprising an organic protic or aprotic solvent or a miscible mixture thereof, or preferably a water-miscible organic solvent.

13. The process according to any of the preceding claims, wherein said nitrogen is fed to the electrolytic cell by bubbling nitrogen gas to the electrolytic solution in contact with said cathode surface.

14. The process of any of the preceding claims, wherein the source of protons in the formation of ammonia is water splitting at the anode or H2 oxidation reaction at the anode.

15. The process of any of the preceding claims operated at a temperature in the range from -10° C to 80° C, preferably in the range 20° C to 50°C, more preferably in the range 25°C to 30°C.

16. The process of any of claims 1 to 15, wherein the process is operated at a pressure selected to be in the range of 1 to 30 atmospheres, preferably in the range of 1-20 atmospheres, preferably in the range of 1-10 atmospheres, more preferably in the range of 1-5 atmospheres.

17. A system for generating ammonia, the system comprising at least one electrolytic cell, which comprises at least one cathode having at least one catalytic surface, wherein the at least one catalytic surface comprises at least one transition metal oxide doped with at least one type of transition metal dopant selected from the group consisting of: Hafnium Oxide doped with Molybdenum and / or Osmium, Iridium oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum and / or Tungsten and / or Hafnium and / or Niobium and / or Nickel and / or Osmium and / or Ruthenium and / or Titanium and / or Vanadium and / or Zirconium, Iridium and / or Rhenium, Niobium Oxide doped with Molybdenum and / or Titanium, and / or Tungsten, Osmium Oxide doped with Rhenium and / or Tungsten and / or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt and / or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium and / or Niobium, Titanium Oxide doped with Rhenium and / or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt and / or Molybdenum and / or Niobium and / or Tantalum.P15340PC00 18. The system according to the preceding claim, wherein said at least one transition metal oxide doped with at least one type of transition metal dopant is selected from the group consisting of: Hafnium Oxide doped with Molybdenum or Osmium, Iridium Oxide doped with Rhenium, Molybdenum Oxide doped with Tantalum or Tungsten or Hafnium or Niobium or Nickel or Osmium or Ruthenium or Titanium or Vanadium or Zirconium, Iridium or Rhenium, Niobium Oxide doped with Molybdenum or Titanium, or Tungsten, Osmium Oxide doped with Rhenium or Tungsten or Molybdenum, Platinum Oxide doped with Rhenium, Rhenium Oxide doped with Cobalt or Molybdenum, Rhodium Oxide doped with Rhenium, Ruthenium Oxide doped with Rhenium, Tantalum Oxide doped with Vanadium or Niobium, Titanium Oxide doped with Rhenium or Tungsten, Vanadium Oxide doped with Rhenium and Tungsten Oxide doped with Cobalt or Molybdenum or Niobium or Tantalum.

19. The system according to the preceding claim, wherein said at least one transition metal oxide doped with at least one type of transition metal dopant is selected from a group consisting of: Molybdenum Oxide doped with Tungsten or Osmium or Titanium or Vanadium, Osmium Oxide doped with Molybdenum and Rhenium Oxide doped with Molybdenum.

20. The system according to any of claims 17 to 19, wherein the catalyst surface comprises a mixture of two or more transition metal oxides doped with two or more different transition metals.

21. The system according to any of claims 17 to 20, wherein the catalyst comprises a rutile structure.

22. The system according to any of claims 17 to 21, wherein the catalyst surface comprises at least one surface having a (110) facet.

23. The system according to any of claims 17 to 22, wherein said electrolytic cell comprises one or more electrolytic solutions.

24. The system according to the preceding claim, wherein said electrolytic cell comprises an acidic, neutral or alkaline aqueous solution.P15340PC00 25. The system of claim 23, wherein said electrolytic cell comprises an electrolytic solution comprising an organic protic or aprotic solvent, or a miscible mixture thereof, preferably a water-miscible organic solvent.

26. The system according to any of claims 17 to 25, wherein the system is configured to produce ammonia in the electrolytic cell at an electrode potential of less than -1.0 V relative to the reversible hydrogen electrode (RHE) or preferably at an electrode potential of less than -0.5 V, relative to the reversible hydrogen electrode (RHE).

27. The system according to any of claims 17 to 26, wherein a cyclic varied potential is used that fluctuates between an active potential and a resting potential to generate a cyclic varied potential through the electrolytic cell.

Images