NANOPARTICLE AND ITS USE AS A WATER-SPLITTING CATALYST

Field of the invention

[0001 ] The invention relates to a nanoparticle and methods for using the nanoparticle to catalyse the splitting of water into hydrogen and/or oxygen. The invention also provides a composition, a material and an electrode comprising the nanoparticle.

Background

[0002] Water-splitting technology to generate hydrogen (hh) has been considered as one of the most promising approaches to convert renewable energy into fuel. Water-splitting processes are considered a sustainable approach to energy generation and storage as they can exploit the abundance of available water and low carbon intensive energy source, such as a renewable source of energy (e.g. solar energy), as inputs. Water-splitting may be carried out in an electrolyser generating hydrogen at the cathode via the hydrogen evolution reaction (HER). Oxygen may be generated at the anode via the oxygen evolution reaction (OER). The current focus is on preparing catalysts for hydrogen evolution as hydrogen is more readily used as a fuel for a variety of applications, such as fuel cells.

[0003] Pt-based catalysts (e.g. Pt/C) are currently the primary option for cathode materials in water electrolysers due to the extremely low overpotential (h) at the onset of hydrogen evolution and large current densities (y). However, Pt-based HER catalysts are not suitable for large-scale applications as they are costly and their supply is not sustainable.

[0004] Recent efforts have focussed on developing water-splitting catalysts based on Earth-abundant metals. Earth-abundant metals exclude Re, Ru, Os, Rh, Ir, Pd, Pt, Ag and Au. One promising class of Earth-abundant metal-based electrocatalysts are the layered double hydroxides (LDHs). LDHs consist of positively charged layers interspersed with negatively charged anions in the interspacial region(s). One example of an LDH is a Ni-Fe LDH, which comprises Ni ²⁺ and Fe ³⁺ cations interspersed with counter anions. Ni-Fe LDHs have been shown to be efficient OER catalysts under alkaline conditions. However, the complex structure of Ni-Fe LDHs means that the catalytic site and mechanism of action are poorly understood, making their further development as electrocatalysts more difficult.

[0005] Accordingly, there is a continuing need to provide alternative HER catalysts. In particular, there is a need to develop catalysts based on Earth-abundant transition metal(s) (e.g. Ni, Co, Fe). In addition, catalysts capable of mediating both HER and OER are desirable. Summary of the invention

[0006] The inventors have developed a nanoparticle that is able to catalyse HER with low overpotentials (h) and favourable current densities (/). Some embodiments of the nanoparticles also catalyse OER with low overpotentials ( h ) and favourable current densities (y). Surprisingly, the nanoparticles provide comparable catalytic performance for HER and/or OER to existing Nobel-metal (Pt, Pd, etc) catalysts. The nanoparticles of the invention may provide sustainable H2 generation through water-splitting.

[0007] In one aspect, the invention provides a nanoparticle comprising:

• a first region comprising a catalytic material;

• a second region comprising an oxygen scavenger; and

• an interface between the first region and the second region.

[0008] In another aspect, the invention provides a composition comprising a plurality of the nanoparticles of the invention.

[0009] In a further aspect, the invention provides a material comprising a plurality of nanoparticles of the invention embedded in a substrate.

[0010] In still a further aspect, the invention provides an electrode comprising the material which comprises a plurality of nanoparticles of the invention on a conductive substrate.

[001 1 ] In yet a further aspect, the invention provides a process of manufacturing a nanoparticle of the invention, the process comprising:

• forming a micelle comprising the catalytic material and the oxygen scavenger

surrounded by a surfactant; and

• heating the micelle to form the nanoparticle.

[0012] In another aspect, the invention provides a method of evolving hydrogen and/or oxygen from water, the method comprising providing an electrochemical cell comprising an anode, a cathode and an electrolyte solution, contacting water with the anode and the cathode, and applying a voltage across the anode and the cathode, wherein at least one of the anode and the cathode comprises a nanoparticle of the invention.

[0013] In a further aspect, the invention provides an electrolyser comprising an anode and a cathode and a power source, wherein at least one of the anode and the cathode comprises a nanoparticle of the invention. [0014] Before describing the present invention in detail, it is to be understood that this invention is not limited to particularly exemplified embodiments, methods of production or use, which may, of course, vary.

[0015] The inventions described and claimed herein have many attributes and

embodiments including, but not limited to, those set forth or described or referenced in this summary section, which is not intended to be all-inclusive. The inventions described and claimed herein are not limited to or by the features or embodiments identified in this summary section, which is included for purposes of overview illustration only and not limitation.

[0016] All publications, patents and patent applications cited herein, whether supra or infra, are hereby incorporated by reference in their entirety. However, publications mentioned herein are cited for the purpose of describing and disclosing the protocols and reagents which are reported in the publications and which might be used in connection with the invention. Nothing herein is to be construed as an admission that the invention is not entitled to antedate such disclosure by virtue of prior invention.

[0017] In this specification where reference has been made to patent specifications, other external documents, or other sources of information, this is generally for the purpose of providing a context for discussing the features of the invention. Unless specifically stated otherwise, reference to such external documents is not to be construed as an admission that such documents, or such sources of information, in any jurisdiction, are prior art, or form part of the common general knowledge in the art.

Brief description of drawing(s)

[0018] The present application will be further described, by way of example only, with reference to the accompanying drawings, in which:

Figure 1a shows a schematic representation of the formation reaction of a nanoparticle of the invention, a Ni-containing nanoparticle and an Fe-containing nanoparticle, and the illustration on the hydrogen evolution reaction (HER) across the interface of the nanoparticle of the invention in alkaline medium;

Figure 1 b shows a scanning transmission electron microscopy (STEM) - high angle annular dark field (HAADF) image of a single nanoparticle of the invention;

Figure 1 c shows high resolution energy-dispersive X-ray spectroscopy (EDS) mapping on HAADF-STEM images of a nanoparticle of the invention, showing the line-scan spectrum, selected area electron diffraction inset (SAED, scale bar - 2 nnr ¹ ) and Ni and Fe elemental mapping; Figure 2a shows an image of the density functional theory (DFT) calculated optimised structure of the interface of a nanoparticle of the invention;

Figure 2b shows a standard free energy diagram of the HER process on the y-Fe ₂ 0 ₃ (31 1 ) and Ni(1 1 1 ) surfaces and the interface between the y-Fe ₂ 0 ₃ (31 1 ) and Ni(1 1 1 ) in a nanoparticle of the invention;

Figure 2c shows a standard free energy diagram of the OER process on the y-Fe ₂ 0 ₃ (31 1 ) and Ni(1 1 1 ) surfaces and the interface between the y-Fe ₂ 0 ₃ (31 1 ) and Ni(1 1 1 ) in a nanoparticle of the invention, and the insets show the optimized structures and the catalytic sites for OERs; Figure 3a shows X-ray diffraction spectra showing the reference stick patterns of NiO (first panel and dotted lines, ICDD: 04-01 1 -9039), Ni (second panel and dotted lines, ICDD: 04-001 - 3331) and y-Fe ₂ C> ₃ (third panel and dotted lines, ICDD:01 -078-6916) (peaks marked with a black spot correspond to carbon fibre paper (CFP) substrate);

Figures 3b-d show X-ray photoelectron spectroscopy (XPS) of Ni-Fe NP measured for (b) Ni 2p, (c) Fe 2p, and (d) O 1 s;

Figure 4a shows linear sweep voltametry (LSV) curves for the hydrogen evolution reaction (HER) catalysed by a nanoparticle of the invention, a nickel-containing nanoparticle, an iron- containing nanoparticle and 20% Pt/C electrode (no /^-correction);

Figure 4b shows Tafel plots for various nanoparticles, including nanoparticles of the invention and benchmark catalysts;

Figure 4c shows LSV curves showing the presence of metal reduction peaks of an electrode comprising nanoparticles of the invention;

Figure 4d shows oxygen evolution reaction (OER) LSV curves for a nanoparticle of the invention, a nickel-iron layered dihydroxide (Ni-Fe LDH) and 20% Ir/C electrode;

Figure 4e shows LSV curves comparing the water-splitting performance of a cell comprising two electrodes comprising nanoparticles of the invention and an Ir/C-Pt/C cell;

Figure 4f shows a chart of cell voltage against time testing the stability of the cell comprising nanoparticles of the invention running at a current of 10 and 20 mA cnr ² compared to the stability of Ir/C-Pt/C cell;

Figure 5a shows a transmission electron microscopy (TEM) image of nanoparticles of the present invention;

Figure 5b shows a size distribution histogram showing the average size of 100 nanoparticles of the invention;

Figure 6a shows a TEM image of comparative iron-containing nanoparticles;

Figure 6b shows a TEM image of comparative nickel-containing nanoparticles;

Figure 7 shows STEM - EDS images of core-shell Ni/NiO nanoparticles in a composition comprising nanoparticles of the invention detected by high resolution-transmission electron microscopy (HR-TEM) (left) and EDS (right). Figure 8 shows LSV curves for HER of nanoparticles of the invention comprising an interface between a first region comprising nickel and a second region comprising iron (III) oxide prepared with different Ni to Fe molar ratios;

Figure 9 shows LSV curves for HER catalysed by a nanoparticle of the invention (Ni-Fe NP), Ni NP, Fe NP and 20% Pt/C in 1 M KOH supported on CFP with /R-correction;

Figure 10 shows LSV curves with HER response from CFP supported nanoparticle of the invention (Ni-Fe NPs) with different amounts of mass loading (estimated from average molecular mass of Ni and Fe from a micellar solution with concentration of 8.5 mg mL ¹ );

Figure 11 shows LSV curve of a comparative iron (III) oxide nanoparticle prepared by thermal reduction showing the presence of two-step reduction peaks of Fe;

Figures 12a-b shows a potential {E-t) trace at a constant applied current to a nanoparticle of the invention of (a) 100 mA cnr ² and (b) 10 mA cm ² ;

Figure 12c shows a STEM-HAADF image of a nanoparticle of the invention following stablity testing shown in Figures 12a-b;

Figure 12d shows LSV curves obtained prior (solid line) and after (dashed line) galvanostatic experiments shown in Figures 12a-b;

Figure 13 shows LSV curves of the oxygen evolution reaction (OER) in 1 M KOH catalysed by a nanoparticle of the invention supported on a carbon fibre paper (CFP) substrate (Ni-Fe JNP/CFP) and comparative fabricated nickel-iron layered dihydroxide on a CFP substrate (NiFe-LDH/CFP);

Figure 14 shows an OER Tafel plot derived from the LSV curve in Figure 3c.

Figures 15a-b show the potential (E-t) traces for a constant applied current to a nanoparticle of the invention at (a) 100 mA cnr ² and (b) 10 mA cm ² ;

Figure 15c shows LSV curves obtained prior (solid line) and after (dashed line) galvanostatic experiments shown in Figures 15a-b;

Figure 15d shows a STEM-HAADF image of a nanoparticle of the invention following the stablity testing shown in Figures 15a-b;

Figure 16 shows an expanded view of the LSV curves shown in Figure 3c across the potential (V i/s. RHE) range of 1.1 to 1.5;

Figure 17 shows a staircase voltammetry plot obtained for OER catalysed by a nanoparticle of the invention on a nickel-foam (NF) substrate in 1 M KOH.

Figure 18a shows an X-ray diffraction (XRD) spectrum of a nanoparticle of the invention (Ni-Fe NP) following HER long-term stability testing (5 hours) at constant j of 20 mA cnr ² (middle line) and 100 mA cnr ² (top line line);

Figure 18b shows an XRD spectrum of a nanoparticle of the invention (Ni-Fe NP) following OER long term stability testing (5 hours) at constant j of 20 mA cnr ² (middle line) and 100 mA cnr ² (top line); Figure 19 shows LSV curves collected at a scan rate of 5 mV s- carried out in a 2-electrode electrolyser cell comparing overall water splitting performance of electrodes comprising a nanoparticle of the invention (Ni ₅ Fei |Ni ₅ Fei) and electrodes comprising Noble-metal catalysts (Pt/C|lr/C);

Figures 20a-c show cyclic voltammetry plots of (a) 2p states of O atom in the y-Fe ₂ 0 ₃ (31 1 ) surface of a nanoparticle of the invention, (b) 3d states of Ni atom in the Ni(1 1 1 ) surface of a nanoparticle of the invention, (c) 2p states of O atom and 3d states of Ni atom at the Ni-Fe interface of a nanoparticle of the invention; and

Figures 21a-b show standard free energy diagrams of the OER process on the (a) Ni side of a Ni-Fe2C>3 interface of a nanoparticle of the invention, and (b) Fe2C>3 side of the interface.

Definitions

[0019] As used therein the term“nanoparticle” relates to a particulate material having a largest dimension measuring less than 1 micrometer.

[0020] As used herein, the term“water-splitting” relates to any process that generates elemental hydrogen or oxygen from water as the starting material. The water-splitting processes described herein are electrolytic in nature. These electrolytic processes involve the hydrogen evolution reaction (HER) at the cathode. Typically electrolytic water-splitting also involves the oxygen evolution reaction (OER) at the anode.

[0021 ] As used herein, the term“oxygen scavenger” relates to any material capable of preventing oxidative passivation of the catalytic material.

[0022] As used herein and in the appended claims, the singular forms“a,”“an,” and“the” include plural reference(s) unless the context clearly dictates otherwise. Thus, for example, a reference to“a nanoparticle” may include a plurality of nanoparticles, and a reference to“a micelle” may be a reference to one or more micelles, and so forth.

[0023] Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be appreciated that any materials and methods similar or equivalent to those described herein can be used to practice or test the invention; the best-known embodiments of the various materials and methods are described.

[0024] The term“(s)” following a noun contemplates the singular or plural form, or both.

[0025] The term“and/or” can mean“and” or“or”. [0026] Unless the context requires otherwise, all percentages referred to herein are percentages by weight of the composition.

[0027] Various features of the invention are described with reference to a certain value, or range of values. These values are intended to relate to the results of the various appropriate measurement techniques, and therefore should be interpreted as including a margin of error inherent in any particular measurement technique. Some of the values referred to herein are denoted by the term“about” to at least in part account for this variability. The term“about”, when used to describe a value, may mean an amount within ±25%, ±10%, ±5%, ±1 % or ±0.1 % of that value.

[0028] The term“comprising” (or variations such as“comprise” or“comprises”) as used in this specification, except where the context requires otherwise due to express language or necessary implication, is used in an inclusive sense, i.e. to specify the presence of the stated features but not to preclude the presence or addition of further features in various embodiments of the invention.

Description of embodiment(s)

[0029] The invention provides a nanoparticle comprising a first region comprising a catalytic material, a second region comprising an oxygen scavenger and an interface between the first region and the second region.

[0030] The nanoparticles of the invention comprise an interface between the first and second regions. The interface allows the first and second regions to interact with each other, which typically requires that the regions are connected physically, for example, the regions may form an intimate junction. The interaction of the first and second regions may involve a charge transfer between the two regions. For example, in some embodiments, the oxygen scavenger may participate in at least a partial charge transfer to the catalytic material. It is believed that the communication across the interface enables the increased catalytic effects of the nanoparticles. Depending on the catalytic material and oxygen scavenger selected, the interface may define a heterojunction, which is formed when both catalytic material and oxygen scavenger are semi conductor materials possessing dissimilar crystalline structures.

[0031 ] The first region of the nanoparticles comprises a catalytic material. Any species capable of catalysing the hydrogen evolution reaction (HER) may be used.

[0032] Nanoparticles comprising catalytic materials have suffered from passivation. For example, nanoparticles comprising a catalytic material have suffered from passivation due to the formation of an oxidized layer of the catalytic material on the nanoparticle surface. The increased surface area inherent for nanoparticles only serves to exacerbate the passivation of the catalytic material. This passivation makes the inclusion of reactive materials capable of catalysing the hydrogen evolution reaction (HER) into nanoparticle catalysts less attractive. The inventors have shown that by forming a nanoparticle with an oxygen scavenger connected with the catalytic material through the interface, this oxidative passivation process can be substantially prevented.

[0033] In some embodiments, the catalytic material is a catalytic metal. The form of the catalytic metal in the nanoparticles may be any form enabling reaction catalysis. For example, the catalytic metal may adopt a metallic form, an ionic form, or a combination thereof. Typically, the catalytic metal will be in a metallic form in the nanoparticles. However, in some

embodiments, the catalytic metal will be in an ionic form. For example, the catalytic metal may be in the form of an oxide, a phosphide, a chalcogenide or a dichalcogenide.

[0034] The catalytic metal may be a first row transition metal or a combination thereof. In some embodiments, the catalytic metal is an Earth abundant transition metal or a combination thereof. For example, catalytic metal may be an Earth abundant transition metal selected from nickel, cobalt, copper, vanadium and manganese or a combination thereof.

[0035] Some embodiments comprise nickel (Ni) as the catalytic material. Preferably, the catalytic material comprises nickel as metallic nickel, e.g. Ni(1 1 1 ). Metallic Ni has long been recognized as an active HER catalyst in alkaline media. Density functional theory (DFT) calculations have shown that the Gibbs free energy (AGH) of the adsorption of the intermediate hydrogen atom on Ni(1 1 1 ) is -0.27 eV, which is close to that of platinum (Pt; -0.09 eV).

However, the electrochemical reactivity and stability of metallic Ni for HER are compromised by the formation of oxides due to contact with water or air. The formation of nickel oxides (NiO _x ) or nickel hydroxides (Ni(OH) _x ) on a reactive nickel surface is passivating. This issue is aggravated for nickel-containing nanoparticles due to the increased surface area inherent for the nanoparticles, and the corresponding increased catalytic activity.

[0036] The inventors have shown that for nanoparticles comprising an interface between a region comprising nickel as the catalytic material and an oxygen scavenger, the formation of passivating oxidized layer on the nickel is impeded.

[0037] It is believed that the interface between the catalytic material and the oxygen scavenger not only impedes the formation of a passivating layer on the catalytic material during the reductive formation reaction, but also during use. The retention of a reactive catalytic material region enables HER to be catalysed. [0038] The second region of the nanoparticles comprises an oxygen scavenger. Any material capable of preventing the formation of a passivating layer on the catalytic material may be used as oxygen scavenger. In some embodiments, the oxygen scavenger may comprise an ion of a metal of lower electronegativity than the catalytic material. The lower electronegativity allows the oxygen scavenger to be preferentially oxidized to impede formation of a passivating layer on the catalytic material. The precise selection of oxygen scavenger therefore depends on the catalytic material selected.

[0039] In some embodiments, the catalytic material of the nanoparticles remains free of a passivating oxidized layer for at least 5h under electrolytic conditions, for example, at least 6h, 7h, 8h, 9h, 10h, 12h, 18h, 24h or longer.

[0040] Some suitable pairs of catalytic material and oxygen scavenger metal ion are set out in Table 1 .

Table 1 : Pairs of catalytic material and oxygen scavenger metal ion

[0041 ] As shown in Table 1 , the oxygen scavenger may comprise ions of iron (e.g. Fe ⁺ , Fe ²⁺ and Fe ³⁺ ), manganese (e.g. Mn ⁺ or Mn ²⁺ ) or a combination thereof.

[0042] The metal ion may be present in the nanoparticle in any form that enables it to act as oxygen scavenger to the catalytic material. Typically, the metal ion may be in the form of a hydroxide or oxide.

[0043] In some embodiments, the oxygen scavenger comprises a metal oxide. Some metal oxides are able to adopt different morphologies. Any morphology that enables the metal oxide to act as oxygen scavenger to the catalytic material is contemplated. For example, in embodiments comprising iron (III) oxide as oxygen scavenger, it may be provided in the form of

[0044] The inclusion of the oxygen scavenger interfaced with the catalytic material may also enhance the catalytic activity of the catalytic material. It is believed that the enhanced catalytic activity may be due to at least a partial charge transfer effect from the oxygen scavenger to the catalytic material through the interface. The charge transfer may impede oxidation of the oxygen scavenger while inhibiting passivation of the catalytic material. This enhancement was observed, for example, in the nickel-iron oxide nanoparticles described in Examples 1 -3 (referred to herein as“Ni-Fe NP”), the interface between the nickel and iron regions of the nanoparticle not only prevents Ni oxidative passivation, but also modifies the kinetics of HER making the Ni-Fe NP superior as a catalyst to Ni alone. This modification is best shown by the observed HER properties (Figures 4a-c) and also by the comparison of calculated AG _H values for metallic Ni with that for the Ni in the Ni-Fe NPs. The inventors believe the enhancement in HER kinetics is achieved by a charge transfer process between the nickel and iron oxide regions of the Ni-Fe NPs. The LSV curve of Ni-Fe NP shown in Figures 4d and 16 suggests suppression of the Ni oxidation process that occurs prior to the onset of OER, which is common to Ni-based OER catalysts owing to the oxidation of Ni to Ni ³⁺ or Ni ⁴⁺ . This result is further confirmed by the absence of observed Fe reduction peaks during HER catalysed by the Ni-Fe NP, suggesting the presence of a partial-charge-transfer effect from Fe-rich region to Ni- rich region that prevents reduction of the Fe ₂ 0 ₃ that is interconnected to metallic Ni through the interface. These results indicate that in some embodiments, the catalytic material is more difficult to oxidise due to stabilisation of the catalytic material at lower oxidation states when it is interfaced with the oxygen scavenger.

[0045] In some embodiments, the oxygen scavenger is also a catalyst for OER. By coupling an OER catalyst with an enhanced HER catalyst, the nanoparticles of the invention may be used for full water-splitting. Accordingly, in some embodiments, the second region of the nanoparticles comprise an oxygen scavenging OER catalyst. For example, the oxygen scavenging OER catalyst may be selected from an iron oxide (e.g. Fe ₂ 0 ₃ ) and a manganese oxide (e.g. MnO or Mn ₂ 0) or a combination thereof. In some embodiments, the oxygen scavenging OER catalyst is iron (III) oxide.

[0046] Surprisingly, the catalytic activity of oxygen scavenging OER catalysts when included in the nanoparticles of the present invention is increased. This increase in catalytic activity is thought to be achieved through electronic effects (e.g. electron transfer from the catalytic material to the oxygen scavenger) or through a steric effect where the catalytic material brings O ^* (the intermediate product of the rate-limiting step for OER) into contact with the oxygen scavenging OER catalyst at the interface (see Figure 2c and Example 1 ).

[0047] The shape of the nanoparticles will depend at least in part on the particular materials selected for inclusion. For example, the shape of the nanoparticles may be spherical, acyclic, cuboidal, cylindrical, conical, rod-like, toroidal, or any other known nanoparticle shape for the selected materials. It will be appreciated that surfaces of the first region, the second region and across the interface must be accessible to reactants (e.g. water molecules). Therefore, typically, the first and second regions of the nanoparticles do not adopt a core-shell structure, where one region entirely encompasses the other.

[0048] The relative size of the first and second regions may vary. The size of the first region should be such that the oxygen scavenging activity of the second region prevents formation of a passivating layer substantially across the entire surface of the first region. Therefore, the first region may be smaller than the second region. The relative size of the two regions may be determined, for example, by analyzing TEM images of the nanoparticles. In some embodiments, the first region measures less than 50% of the length of the nanoparticle, for example, from 10% to 50% of the length of the nanoparticle.

Preparative methods

[0049] The nanoparticles of the invention may be prepared by any means known in the art provided the interface between a first region comprising the catalytic material and a second region comprising the oxygen scavenger is produced.

[0050] In some embodiments, the nanoparticles may be prepared by thermal reduction.

This approach is scalable and enables precise control over the composition and size of the nanoparticles.

[0051 ] The invention provides a process of manufacturing a nanoparticle, comprising forming a micelle comprising a catalytic material precursor and an oxygen scavenger precursor surrounded by a surfactant; and heating the micelle to form a nanoparticle comprising a first region comprising a catalytic material, a second region comprising an oxygen scavenger and an interface between the first region and the second region. One embodiment of this process is shown in the schematic of Figure 1 a.

[0052] The catalytic material precursor may be any reagent capable of forming the desired catalytic material under thermal reduction conditions. For example, the catalytic material precursor may be an ionic species that is reduced during the thermal reduction process. In some embodiments, the precursor material is a metal ion corresponding to any of the catalytic metals described above, including an ion of an Earth abundant transition metal selected from nickel ions, cobalt ions, copper ions, vanadium ions and manganese ions or a combination thereof.

[0053] The oxygen scavenger precursor may be any reagent capable of forming the desired oxygen scavenger under thermal reduction conditions. For example, the oxygen scavenger precursor may be a metal ion that is oxidized during the process. In some embodiments, the precursor material is a metal ion corresponding to any of the oxygen scavengers described above with a lower oxidation number. For example, the oxygen scavenger precursor may be selected from iron ions (e.g. Fe ⁺ or Fe ²⁺ ), manganese ions (e.g. Mn ⁺ ) or a combination thereof.

[0054] Micelle formation may be carried out in any suitable medium able to partition the catalytic material precursor and oxygen scavenger precursor into micelles. Suitable media is typically a hydrophobic solvent, such as a C5 to C10 hydrocarbon solvent, including hexane, pentane, petroleum ether or combinations thereof, an aromatic solvent, including benzene, toluene or combinations thereof, and combinations of these hydrophobic solvents. The media may also include a minor portion of a hydrophilic solvent which may also be incorporated into the micelles or separate into a second phase. Suitable hydrophilic solvents include water (e.g. deionized water), alcohols (e.g. ethanol, methanol, isopropanol, sec-propanol), esters (e.g. diethyl acetate), ketones (e.g. acetone), amides (e.g. dimethyl formamide) and combinations thereof.

[0055] The surfactant may be cationic, anionic, zwitterionic or non-ionic. In some embodiments, the surfactant is an anionic surfactant. The hydrophilic portion of the surfactant may comprise oxygen, nitrogen or sulphur atoms. The hydrophobic portion of the surfactant may comprise a portion that upon reduction forms a conductive material. Suitable portions for forming the conductive material upon reduction include carbon chains, for example, a carbon chain from 8 to 30 carbon atoms in length. By“carbon chain” it is meant a carbon chain uninterrupted by a heteroatom (for example, -(CFfeV where n in an integer from 8-30). The carbon chain may be saturated or unsaturated, for example, comprising 1 -3 degrees of unsaturation. Any unsaturated olefinic bonds in an unsaturated carbon chain may have cis- or trans-configurations. In some embodiments, the surfactant is a fatty acid or a combination of fatty acids. Suitable surfactants include those comprising the following anions: oleate, linoleate, linoleneate, erucate, palmitate, stearate, arachidate or a combination thereof. In some embodiments, the surfactant comprises oleate ions.

[0056] The surfactant may be introduced by first forming a complex of the catalytic material precursor and/or the oxygen scavenger precursor with the surfactant in a hydrophilic medium. Any of the hydrophilic solvents described above may be used as hydrophilic medium. Addition of the hydrophobic solvent described above, will cause the formation of micelles.

[0057] Nanoparticle formation occurs upon heating the micelles to a temperature sufficient to reduce the oxidation number of the precursor of at least one of the catalytic material and the oxygen scavenger. In some embodiments, the method comprises heating the micelles to a temperature of at least about 50°C, for example, at least about 70°C, about 100°C, about 150°C, about 200°C, about 250°C, about 300°C or about 350°C. The temperature may be maintained for a sufficient period of time to cause nanoparticle formation. The length of time that this temperature is maintained will depend on the scale of the reaction, with smaller scale reactions (e.g. providing less than 1 g nanoparticles) requiring shorter reaction times. In some embodiments, the temperature is maintained for about 30 seconds to about 30 minutes. In larger scale embodiments (e.g. embodiments providing 1 g or more of nanoparticles), the temperature is maintained for at least about 1 hour. The rate of temperature increase may also affect the size distribution of the nanoparticles. Accordingly, the temperature may be increased at a rate such that a monodisperse population of nanoparticles is produced. A suitable temperature ramp rate may be about 10°C/min or slower.

[0058] As the nanoparticles are formed upon heating the micelles, the final constitution of the nanoparticles is controlled by the contents of the micelles. In order to form nanoparticles comprising the interface between the two distinct regions, it is believed that one of the catalytic material or oxygen scavenger in the micelle first forms a seed nanoparticle and then the other component forms the other portion on the seed nanoparticle. In some embodiments, the oxygen scavenger forms the seed nanoparticle.

[0059] The ratio of catalytic material to oxygen scavenger used to prepare the micelles has an influence on the properties of the nanoparticles. The molar ratio of catalytic material to oxygen scavenger may be greater than about 1 , for example, the molar ratio may be from about 1 to about 5, including about 1 , about 1.2, about 2, about 3, about 3.6, about 4 or about 5. A molar ratio of 5 corresponds to a ratio of 5:1 for the catalytic material to oxygen scavenger or 5 times the number of moles of catalytic material to the number of moles of the oxygen scavenger. The optimal ratio of catalytic material to oxygen scavenger will ensure that sufficient amounts of the precursors for both the catalytic material and oxygen scavenger are

incorporated into the micelles. If the molar ratio of catalytic material to oxygen scavenger is too low (e.g. ratios of less than 1 ) an excess of nanoparticles comprising the oxygen scavenger only may be produced. If the molar ratio of catalytic material to oxygen scavenger is too high (e.g. ratios of greater than 5) an excess of nanoparticles comprising the catalytic material only may be produced.

[0060] In some embodiments, the nanoparticles provided by the above method are monodisperse. In some embodiments, at least 60%, for example, at least about 65%, about 70%, about 75% or about 80%, of the nanoparticles present in a composition have a size within about 3nm of each other, for example, from about 5nm to about 8nm. In some embodiments, at least 80% of the nanoparticles are less than about 10nm in size. The size and size distribution of the nanoparticles will be influenced by the preparation conditions, including the surfactant, the ratio of catalytic material to oxygen scavenger, the hydrophilic solvent, the hydrophobic solvent, the temperature of reduction and the rate of temperature increase. In some embodiments, the size of a nanoparticle may be determined by direct detection, for example, by TEM techniques. In some embodiments, the size of a nanoparticle may be determined as an average size across a plurality or population of nanoparticles. It will be appreciated that in a composition comprising a population of nanoparticles with an average size of less than 1 micrometer, there will be a substantial number of individual members of that population with an individual size of less than 1 micrometer.

[0061 ] It will also be appreciated that, due to the micellular structure, following heating the nanoparticles may further comprise a layer comprising the surfactant. This layer assists in suspending the nanoparticles in the medium. Suspension of the nanoparticles may assist in deposition of the nanoparticles onto a substrate.

[0062] Accordingly, the invention also provides a composition comprising a plurality of the nanoparticles of the invention. The composition may further comprise a carrier in which the nanoparticles are suspended. The carrier may comprise any of the hydrophilic and hydrophobic solvents described above. In these compositions, the nanoparticles may be monodisperse. In some embodiments, at least 60%, for example, at least about 65%, about 70%, about 75% or about 80%, of the plurality of nanoparticles have a size within about 3nm of each other, for example, from about 5nm to about 8nm. In some embodiments, at least 80% of the plurality of nanoparticles have a size of less than about 10nm.

[0063] The composition may be used to form a material comprising the nanoparticles of the invention on a substrate. Any suitable substrate may be employed depending on the desired application of the nanoparticles. For example, the substrate may be porous or non-porous. It will be appreciated that non-porous substrates the nanoparticles may be applied as a layer (or film) to the surface of the substrate. For porous substrates, the nanoparticles may be applied as a layer on the surface of the substrate or it may be embedded into the substrate. As the nanoparticles are intended to catalyze electrolytic water-splitting, the substrate may be conductive. Suitable conductive substrates include carbon-fibre paper (CFP), nickel foam (NF), stainless steel mesh, nickel mesh, stainless steel plate, nickel plate, titanium mesh, titanium plate, indium/fluorine tin oxides, glassy carbon, carbon cloth, carbon fibre cloth, carbon felt, graphite felt and graphite foam. When the substrate is conductive, the material can be used directly as an electrode. Accordingly, the invention also provides an electrode comprising the nanoparticles on a conductive substrate.

[0064] In some embodiments, the material may comprise a conductive material surrounding each nanoparticle. The formation of a conductive material in contact with the surface of each nanoparticle is advantageous for use in electrochemical reactions as it provides a region of enhanced conductivity surrounding the nanoparticles. For nanoparticles formed with a surfactant comprising a carbon chain, the conductive material may be provided by heating the layer of surfactant surrounding the nanoparticle, for example, during the heating step of the process of manufacturing the nanoparticles.

[0065] In some embodiments, the material comprises the nanoparticles in a mass loading of at least about 1 mg/cm ² , for example, about 2mg/cm ² , about 3mg/cm ² , about 3.65mg/cm ² , about 4mg/cm ² , about 5mg/cm ² , about 6 mg/cm ² , about 6.5 mg/cm ² or greater. In some embodiments the material comprises the nanoparticles with a mass loading from about 1 mg/cm ² to about 50 mg/cm ² , about 2mg/cm ² to about 20mg/cm ² or about 3 mg/cm ² to about 7mg/cm ² .

[0066] In some embodiments, the process further comprises removing any excess medium (or solvent) remaining following reduction. This may be achieved by any means known in the art for separating such components, including evaporation under reduced pressure and

freeze-drying.

Methods of use

[0067] The invention provides a method of evolving hydrogen and/or oxygen from water. The method is carried out in an electrochemical cell, which comprises an anode, a cathode and an electrolyte solution. The method comprises contacting water with the anode and the cathode of the electrochemical cell, and applying a voltage across the anode and the cathode. At least one of the anode and the cathode comprises the nanoparticle of the invention, for example, in the form of an electrode. In some embodiments, the electrolyte solution is an aqueous electrolyte solution. The aqueous electrolyte solution may also be the source of the water. Typically, the water will have an alkaline pH, for example, a pH of at least 8, 9 or 10. In some embodiments, the water comprising a strong base, for example, a hydroxide base such as NaOH or KOH.

[0068] Both the anode and the cathode may comprise the nanoparticles the invention. In these embodiments, the oxygen scavenger may be an oxygen scavenging OER catalyst.

[0069] When used in these methods, the nanoparticles of the invention provide similar catalytic activity to the present leading HER and/or OER catalysts. These leading catalysts are Nobel-metal based - platinum for HER and palladium for OER. Surprisingly, the nanoparticles are able to provide similar catalysis while using Earth-abundant metals.

[0070] In some embodiments, the method of evolving hydrogen proceeds with a Tafel slope of up to 80 mV dec ^-1 , for example up to 79 mV dec ¹ , 78 mV dec ^-1 , 77 mV dec ^-1 , 76 mV dec ^-1 , 75 mV dec ^-1 , 70 mV dec ^-1 , 65 mV dec ^-1 or 60 mV dec ^-1 . [0071 ] The voltage applied across the anode and the cathode may be selected to match the overpotential for the nanoparticles selected. Therefore, the voltage applied may be matched to the overpotential for the nanoparticle under both HER (cathode) and OER (anode) conditions.

[0072] The nanoparticles of the invention require low overpotentials at the cathode to drive HER. In some embodiments, the nanoparticles catalyse HER with an overpotential from about -500mV to about -1 mV, for example, from about -400mV to about -10mV, about -300mV to about -30mV or about -200mV to about -30mV. The voltage applied to the cathode to achieve an overpotential to drive HER may be less than, or equal to, 0V, for example, less than, or equal to, about -1 mV, about -200mV, about -250mV, about -300mV, about -400mV, about -500mV, about -1 V or about -10V. In some embodiments, the voltage applied to the cathode may be from about -10V to about -1 mV, about -1 V to about -1 mV, about -200mV to about -1 mV, or about -180mV to about -40mV. In some embodiments, applying a voltage to the cathode from - 10OmV to -1 mV, for example, from -60mV to -20mV or about -46mV may generate a current density of 10mA/cm ² ; applying a voltage from about -150mV to about -30mV, for example from about -100mV to about -50mV or about -73mV may generate a current density of about 20mA/cm ² ; and applying a voltage from about -200mV to about -100mV, for example, from about -180mV to about -120mV or about -147mV may generate a current density of about 100mA/cm ² .

[0073] The nanoparticles of the invention also require relatively low overpotentials at the anode to drive OER. In some embodiments, the nanoparticles catalyse OER with an

overpotential from about 100mV to about 500mV, for example, from about 150mV to about 400mV, about 150mV to about 300mV or about 200mV to about 280mV. The minimum voltage applied to the anode to achieve an overpotential to drive OER may be at least about 150mV, for example, at least about 180mV or about 200m V. The maximum voltage applied to the anode to achieve an overpotential to drive OER may be up to about 10V, for example, up to about 5V, about 2V, about 1 V, about 330mV, about 300mV or about 240mV. The voltage applied to the anode may be from any of these minimum voltages to any of these maximum voltages. For example, the voltage applied to the anode may be from about 150mV to about 10V, about 150mV to about 5V, about 150mV to about 2V, about 150mV to about 330mV, or about 200mV to about 300mV. In some embodiments, applying a voltage to the anode from about 180mV to about 240mV, such as about 210mV, generates a current density of about 10mA/cm ² ; applying a voltage from about 200mV to about 260mV, such as about 230mV, generates a current density of about 20mA/cm ² ; and applying a voltage from about 250mV to about 320mV, such as about 270mV, generates a current density of about 100mA/cm ² . [0074] It will be appreciated that in a two electrode cell a single voltage is applied across the anode and the cathode. The voltage applied across the anode and the cathode may be sufficient to provide the necessary overpotential at either or both of the anode or the cathode. Therefore, the voltage applied across the anode and cathode may be between any of voltages described above for the cathode and any of the voltages described above for the anode. In some embodiments, the voltage applied across the anode and cathode may be from about 1 .23 V to about 10 V, for example, from about 1 .23 V to about 5 V, or about 1 23V to about 2 V.

[0075] Also provided is an electrolyser comprising an anode and a cathode and a power source. At least one of the anode and the cathode comprises the nanoparticles of the invention. In some embodiments, both anode and cathode of the electrolyser comprise the nanoparticles. In some embodiments, the power source is a low carbon intensive power source. The power source may be a renewable power source, for example, one or more solar panels or wind turbines, or a non-renewable power source, for example, a nuclear reactor.

Examples

[0076] The invention will be further described by way of non-limiting example(s). It will be understood to persons skilled in the art of the invention that many modifications may be made without departing from the spirit and scope of the invention.

Example 1 : Nanoparticle synthesis and characterisation

[0077] Ni-Fe NPs were prepared by dissolving 5mmol of Ni(N0 ₃ ) ₂ .6H ₂ 0 (UNIVAR grade, Ajax Finechem, Australia) and 1 mmol Fe(CI) ₂ .4H ₂ 0 (Sigma-Aldrich, USA) in a mixture of 2ml_ de-ionized water and 1 ml_ ethanol. 4.00g of Na(oleate) (TCI, Japan) was dissolved in a separate mixture of 4ml_ de-ionized water and 3ml_ ethanol. The two mixtures were then mixed in a round- bottomed flask to yield thick and waxy green-yellow substances, and into the mixture, 14ml_ of hexane was added. Upon the addition of hexane, coloured metallic elements were immediately transfered into the hexane phase. The mixture was then stirred at 400 r.p.m. for 60min. Two well -separated layers of water (colorless/opaque) and coloured hexane layer (dark- green) were obtained at the end of stirring.

[0078] Other oleate complexes with different Ni and Fe ratio were made by this method by adjusting the mole ratio between Ni and Fe salts precursors.

[0079] Physical characterisation

[0080] X-ray diffraction spectroscopy (XRD). XRD measurements were performed with PANalytical X’Pert Empyrean instrument equipped with standard Cu anode, K-a wavelength =

1 .54 nm. The typical scan range was 10° to 80°, collected with step size of 0.039° s ¹ . [0081 ] X-ray photoelectron spectroscopy (XPS). XPS measurements were performed with Thermo ESCALAB250i X-ray photoelectron spectrometer, to ensure the results

consistency the scan was performed at 4 different spots.

[0082] Raman spectroscopy. Raman spectra analysis, Renishaw inVia Raman microscope equipped with 514 (green) Ar-ion laser with 1800 I mrrr ¹ was used. All the raman samples was catalysts supported on CFP.

[0083] Transmission electron microscopy (TEM). TEM was carried out with Phillips CM 200 microscope. To prepare TEM samples, the CFP supported catalyst was transferred to Cu- grid by physically scratching the electrode using a sharp knife. The resulting powder was dispersed in absolute ethanol by ultrasonication for 15 mins. The resulting mixture was then drop-casted onto Cu-grid and dried in room temperature.

[0084] Scanning transmission electron microscopy (STEM). STEM was carried out with JEOL JEM-ARM200F equipped with aberration corrected 200 kV electron beam source. To prepare STEM samples, the CFP supported Ni-Fe JNPs was transferred to Cu-grid coated with holey carbon by physically scratching the electron using a sharp knife. The resulting powder was dispersed in absolute ethanol by ultrasonication for 15 mins. The resulting mixture was then drop-casted onto Cu-grid and dried in room temperature.

[0085] Scanning electron microscopy. SEM analysis were carried out with FEI NanoSEM 230 with a 10 kV accelerating voltage.

[0086] Optical microscopy. Microscopy images were taken with Nikon eclipse LV100POL.

[0087] Figures 1 b and 1 c show TEM images of the structure of the Ni-Fe NP. Size distribution analysis in Figure 5b reveals that the Ni-Fe NP size distribution fits to a Gaussian curve, with the majority of particles being about 5nm to about 8nm in diameter. The high resolution image (Figure 1 b), obtained with high angle annular dark-field (HAADF) imaging by scanning tunneling electron microscope (STEM), reveals the interface formed in Ni-Fe NP, dividing the NP into two regions with distinctly different crystal structures and electron densities. The region with higher electron density (appears brighter) reveals a closely packed crystal structure. The darker region exhibits a more sparsely packed, less-dense crystal structure filled with smaller oxygen atoms consistent with the y-Fe2C>3 structure. The selected area electron diffraction (SAED) of Ni-Fe NP in Figure 1c, displays two clear diffraction rings with radius of 2.51 A and 2.10 A, indexed for (311) and (111) in y-Fe203 and Ni crystal lattices, respectively. Energy dispersive X-Ray (EDS) elemental analyses in Figure 1 c (line-scan (left) and map (right)) also shows the interface between the Ni and Fe203 regions.

[0088] As a control experiment, Ni- and Fe-based nanoparticles (denoted as Ni NP and Fe NP, respectively) were also synthesized v/a the above described method modified by omitting one or the other of the metal salt precursors. The TEM image in Figure 6a shows that Fe NPs have evenly distributed particle size of about 6nm to about 7nm, similar to the Ni-Fe NP.

However, Ni NPs have unevenly distributed sizes with the presence of large nanoparticles (about 50nm in diameters, Figure 6b). The nanoparticle size difference suggests that thermal decomposition process of Fe NPs occurs at lower temperatures (about 70°C) and is

instantaneous, while Ni NP occurred at higher temperatures with a more progressive

mechanism. These data suggest that Ni-Fe NP may have been formed by Ni growth on the pre formed Fe NP, rather than simultaneous nucleation.

[0089] X-ray diffraction (XRD) analysis shown in Figure 3a also revealed the unique features of Ni-Fe NP. The XRD pattern of Ni NP indicates the formation of core-shell structure of Ni/NiO as identified by the presence of peaks correspond to both Ni and NiO (ICDD: 04-001 - 3331 ; 04-01 1 -9039), as a result of instantaneous formation of passivating NiO outer shell. The XRD of Fe NP indicates that y-Fe203 (ICDD: 01 -078-6916) was formed. In contrast, in case of Ni-Fe NP, NiO peak(s) was not observed, while all the peaks of Fe ₂ 0 ₃ were still present. The strong peak at 44.5° corresponds to the (1 1 1 ) facet of metallic Ni.

[0090] The oxidation states of Ni and Fe in Ni-Fe NP were further analysed by X-ray photoelectron spectroscopy (XPS). Two main peaks related to Ni _2p exist at about 855.1 eV (Ni _2P 3 _/2 ) and 873.0 eV (Ni _{2pi 2} ) in Ni-Fe NP as shown in Figure 3a. The Ni 2 _{p3 2} peak could be further resolved to three peaks at 852.9eV corresponding to Ni°, and two peaks that arise from the co-formed Ni/NiO core-shelled particle impurity (Figure 7). The EDS analysis (Figure 7) revealed that Fe was not detected in the core-shell Ni/NiO nanoparticles. In contrast, the Ni _{2p3 2} peak at 852.9eV is not present in Ni NP, indicating Ni is preserved in its metallic state in the Ni- Fe NP configuration. As shown in Figure 3c, Fe _2p3/2 exhibits strong peaks at 71 1 4eV and 713.6eV, consistent with Fe ³⁺ oxidation state in Fe ₂ 0 ₃ . The presence of Fe ₂ 0 ₃ is also confirmed by the presence of strong Oi _s at 530.6eV related to metal-bound oxygen species as shown in Figure 3d. The XPS study shows that Ni remains in a metallic state through the integration of Fe in the Ni-Fe NP configuration.

[0091] In Figure 12a, the stability of Ni-Fe NP was also tested under an extreme current condition, the test was configured at JHER 10OmA/cm ² . Constant current analysis at JHER of 10mA/cm ² is shown in Figure 12b. These studies reveal that Ni-Fe NP remains stable, with constant current maintained at relatively low h of 180 mV for an extended period of 10 hours. Additionally, Figure 12d shows that loss of HER activity is minimal validated from the LSV of Ni- Fe NP obtained before and after the stability testing. The stability is confirmed by analysis of a HAADF-STEM image (Figure 12c) which does not indicate any detectable changes in the physical features of Ni-Fe NP. The crystal structure of the Ni-Fe NPs following the HER stability test was also conserved as shown by XRD in Figure 18a.

[0092] Shown in Figures 15a and 15b, are the E-t trace obtained by subjecting the Ni-Fe NPs to the constant currents of 10mA/cm (Fig 15a) and 10OmA/cm ² (Fig 15b) for up to 10 hours. The near constant potential shown in these traces indicates OER stability for the nanoparticles. The electrochemical stability of Ni-Fe NP for OER was further demonstrated by multi-current stepping experiment shown in Figure 17. The multi-current stepping experiment was conducted by setting the initial current density to 50mA/cm ² and increasing the current density by nine increments of 50mA/cm ² every 300s to a final current density of 500mA/cm ² . Results are shown in Figure 17 without iR-correction. The current response was observed to level off quickly on each stepping with high consistency, suggesting excellent mass transport properties as well as physical stability. HAADF-STEM (Figure 15b), LSC (Figure 15c), and XRD (Figure 18b) analyses of Ni-Fe NP following OER stability test show an unchanged structure and OER activity relative to fresh Ni-Fe NP electrode.

[0093] Computational analysis

[0094] All spin-polarized density-functional theory (DFT) computations were carried out using the Vienna ab initio simulation package (VASP) based on the projector augmented wave (PAW) method. Electron-ion interactions were described using standard PAW potentials, with valence configurations of 2s ² 2p ⁶ 3s ² 3cP for Fe (Fe_sv_GW), 2s ² 2 p ⁴ for O (0_Gw_new), and 1 s ¹ for H (H_GW). A plane-wave basis set was employed to expand the smooth part of wave functions with a cut-off kinetic energy of 520 eV. For the electron-electron exchange and correlation interactions, the functional parameterized by Perdew-Burke-Ernzerhhof (PBE), a form of the general gradient approximation (GGA), was used throughout. The GGA+U approach was used with U-J = 3.0 eV for the Fe atoms to take into account the on-site Columbic repulsion between the Fe d electrons in the oxide and model the electronic structure of y-Fe ₂ 0 ₃ .

[0095] To study the mechanistic chemistry of surface reactions, the y-Fe ₂ 0 ₃ (31 1 ) surface was modeled using a slab with eight oxygen layers. And the Ni (1 1 1 ) surface was modelled using a slab with six nickel layers. To build the interface model of the Ni-Fe heterojunction, a y- Fe ₂ C> ₃ (31 1 ) (1 x 1 ) unit cell is combined with a Ni( 11) (4 ^c 2V2) unit cell since their lattices constants along the x direction are close to each other (10.20 A and 9.96 A for the y-Fe ₂ C> ₃ (311) (1 x 1) and Ni(111) (4 ^c 2V2), respectively), which may be one of the driving forces to the formation of the heterojunction due to the small strain at the interface. As such, the lattice constants of the interface model along the x axial is initially set as 10.0 A. And the lattice constants a and b were further optimized to get the stable interface configuration. A sufficiently large vacuum region of 15 A was used for all the systems to ensure the periodic images to be well separated. During the geometry optimizations, all the atoms were allowed to relax.

[0096] In this work, the Brillouin-zone integrations were conducted using Monkhorst-Pack grids of special points. A gamma-centered (2 c 4 c 1 ), (5 c 5 c 1 ) and (5 c 5 c 1 ) /(-point grid was used for the y-Fe ₂ C> ₃ (31 1 ) (1 c 1 ), Ni (1 1 1 ) (2 c 2) surface cells and the interface model. The H ₂ , 0 ₂ and H ₂ 0 molecules were calculated in a 20 c 20 c 20 A ³ box. The convergence criterion for the electronic self-consistent loop was set to 1 O ^-4 eV. And the atomic structures were optimized until the residual forces were below 0.01 eV A ¹ . Based on DFT results, the magnetic moments for Fe ³⁺ and Ni° metals are around 4.0 and 0.7 m _B , respectively, which indicates that the Fe ³⁺ has a high spin state with five unpaired electrons. The antiferromagnetic (AFM) structures are considered for the y-Fe ₂ 0 ₃ based on the superexchange model.

[0097] To provide insights on the HER and OER performance of the Ni-Fe NP, first- principles DFT calculations were conducted. The interface model between the y-Fe ₂ 0 ₃ (31 1 ) and Ni (1 1 1 ) surfaces (see Figure 2a) was built based on the experimentally observed STEM- HAADF image of a single Ni-Fe NP (see Figure 1 b). These calculations suggest that the Fe and Ni atoms are coupled via the bridge O atoms at the interface. These calculations further suggest that Ni atoms at the interface are oxidized since their partial charge densities are transferred to bridge O atoms, as evidenced by the alteration of Bader charge of the relevant atoms listed in Table 2. The strong interaction between Ni and O at the interface can further be confirmed by their overlapped O 2 p and Ni 3d states around the Fermi energy level (see Figure 20).

Table 2. Bader charge of O and Ni atoms in the y-Fe ₂ C> ₃ (31 1 ), Ni(1 1 1 ) and the interface between the first and second regions of the nanoparticles.

Y-Fe ₂ 0 ₃ (311) Ni(111) Interface

Ni 0.0 0.4

[0098] The AGH is a key descriptor used to evaluate the HER performance of an electrocatalyst. Under the standard condition, the overall HER pathway includes two steps: first, adsorption of hydrogen on the catalytic site ( ^* ) from the initial state (H ⁺ + e- + ^* ), second, release the product hydrogen (½H ). The total energies of H ⁺ + e- and ½H ₂ are equal. Therefore, the Gibbs free energy of the adsorption of the intermediate hydrogen on a catalyst {AGH) is the key descriptor of the HER activity of the catalyst and is described in Equation 1 : DQH = DE _H +DZRE - TAS Equation 1

[0099] In Equation 1 , DE _H represents the binding energy, DZRE represents the zero point energy change and AS represents the entropy change of H adsorption. The calculated value of DZRE - TAS is about 0.24 eV for HER catalysed by the Ni-Fe NPs. DE _H is calculated using Equation 2:

DEH = E _su rf ₊ H— Esurf— ½ EH2 Equation 2

[0100] In Equation 2, E _SUrf+H represents the total energy of the system with one adsorbed H atom in each unit cell, E _surf represents the energy of the bare surface and E _H 2 represents the energy of the H ₂ gas molecule.

[0101 ] The DFT results reveal that the H atom can be either adsorbed on the top site of O atoms in y-Fe2C>3 (31 1 ) or the face-centred cubic (f.c.c.) site of Ni(1 1 1 ) with the AGH of -0.62 and -0.31 eV, respectively. The negative values suggest that the activity of both surfaces are too high for HER. At the interface of the Ni-Fe, both the interfacial O and Ni atoms have the more optimal activity to the HER as demonstrated by the corresponding AGH of -0.27eV and -0.14eV, respectively (Figure 2b). It suggests that both the activities of the interfacial Ni and O atoms are reduced due to the formation of Ni-0 bond and thereafter the charge transfer, which is beneficial to the HERs, as observed in experiments (Figure 4a).

[0102] The OER mechanism includes four fundamental steps in an acidic solution. This mechanism may be described in the following reaction scheme (showing the reaction intermediates in bold):

2H ₂ 0 OH* + H ⁺ + e- + H ₂ 0 « 0* + 2H ⁺ + 2e- + H ₂ 0 OOH* + 3H ⁺ + 3e- H 0 ₂ + 4H ⁺ + 4e ^~

[0103] In an alkaline electrolyte, the anode reaction may be described by the following scheme (showing the reaction intermediates in bold):

40H- OH ^* + 30H- + e < O ^* + 20H + 2e ^~ + H ₂ 0 OOH ^* + OH + 3e H 0 ₂ + 2H ₂ 0 + 4e

[0104] As such, the Gibbs free energy of the adsorption of intermediates including OH ^* , O ^* and OOH ^* can be used to calculate the change of Gibbs free energy of each fundamental step (AGn, n =1 - 4) of OER, which is approximately independent to the pH value of the solution. [0105] A similar method (AG = DE +DZRE - TAS) to calculate the Gibbs free energy of the adsorption of intermediates including OH ^* , O ^* and OOH ^* were also employed. The entropies and zero point energies (ZPEs) used in the construction of the reaction free energies are obtained from the previous results. AG _n is finally calculated using Equations 3 to 6:

AGi= AGOH Equation 3

AG ₂ = AGO* - AGOH Equation 4

AG ₃ = AGOOH* - AGO Equation 5

AG ₄ = - E02 + 2EH2 - 2EH20 - AGOOH Equation 6

[0106] In Equation 6, EH2 represents the energy of the H ₂ gas molecule, E02 represents the energy of the 0 ₂ gas molecule, and EH20 represents the energy of the H ₂ 0 gas molecule.

[0107] To understand the OER activities, the change of the Gibbs free energies, AG _n (n=1 ,4), for the four OER fundamental steps were calculated according to Equations 3 to 6. The magnitude of overpotential (q) has been demonstrated to be the difference of the practical potential (maximum AG _n over the charge e) and the standard Nernstian potential (1 .23/1 .25 V vs. reversible hydrogen electrode (RHE) for the experiments and DFT calculations, respectively). Thus, the AG _n values of intermediates need to reach the minimal difference between each other to reduce the q. The calculated AG _n on the y-Fe ₂ 0 ₃ (31 1 ), Ni(1 1 1 ) and heterojunction are shown in Figures 2b, 2c and 2d. These theoretical results indicate that the y- Fe ₂ 0 ₃ itself is a good OER electrocatalyst with the theoretical q of 0.55 eV. These calculations reveal that the rate-determining step is the formation of the O ^* intermediate, as the binding strength of Fe and O is relatively weak.

[0108] As a comparison, the Ni metal possesses a large calculated q of 2.05 V with the rate-determining step for the formation of OOH intermediate. This is largely due to much stronger calculated binding between the surface Ni atom and the O intermediate (see Table 3; Figures 21 a and 21 b). At the interface of Ni-Fe NPs, the intermediate O atom adsorbs on the surface at the bridge site between Fe and Ni. As such, the binding energy of O is optimized to attain the smallest theoretical q of 0.28 V, which greatly matches the experimental observation (see Figures 4d and 16). Our theoretical studies also demonstrate that only the interface has the superior OER performance (Figures 2b and 2c). Table 3. Binding energy of O (E _b = E _tot - E _{su f} - E ₀ , where the E _tot , E _surf and E ₀ represent the energies of the surface with the adsorbed O atom, bare surface and isolated O atom, respectively. The more negative values means a stronger binding strength) on the g- Fe ₂ C> ₃ (31 1 ), Ni(1 1 1 ) and their heterojunction.

Y-Fe ₂ 0 ₃ (311) Ni(111) Ni-Fe

heterojunction

Example 2: Electrode preparation

[0109] Electrodes were prepared by employing 3-D conductive substrate material, such as carbon fiber paper (CFP, Goodfellow, UK) and nickel foam (NF, thickness: 1.6mm, bulk density: 0.45g/cm ³ , Goodfellow, UK). The substrates were modified with metal oleate complexes dissolved hexane solution via drop casting. For the preparation of Ni/Fe NP onto CFP electrode, Ni(oleate)2 and Fe(oleate)2 containing hexane solutions were mixed and drop-casted onto CFP substrate. The substrate was then dried under bench-top conditions until all the remaining hexane was evaporated and a waxy film was formed at the electrode surface. The modified substrates were subsequently annealed in Ar-protected horizontal tube furnace at 350°C for 2 hours at a ramp rate of 10°C/min with an Ar-flow rate of 5ml_/min. Upon completion, the annealing chamber was allowed to cool down naturally to room temperature. The electrodes were washed ultrasonically with Milli-Q water to remove the loosely bound nanoparticles for 5minutes. Catalyst mass loadings (ml.) of CFP and NF modified with metal nanoparticles were 3.75mg/cm ² and 6.5mg/cm ² , respectively, unless otherwise stated.

[01 10] As a control experiment, the noble metal modified electrodes were prepared through drop-casting method. An ink of 20% Pt/C was prepared by the addition 10mg of 20% Pt/C or 20% Ir/C (Premetek, USA) into 1 96ml_ of 50% ethanol solution containing 0.04ml_ 5% Nation binder solution (Sigma-Aldrich, USA). The ink was then loaded onto substrate and dried under vacuum at 40°C, resulting in catalyst mass loading of 3.0mg/cm ² .

[01 1 1 ] All electrochemical measurements were carried out with a CHI 760 electrochemical workstation (CH Instruments, USA). Electrochemical measurements were performed with a standard configuration of three-electrode cell configuration composed of working electrode, graphite rod counter electrode and standard calomel electrode (SCE) as the reference electrode, unless otherwise stated. For all of the electrochemical measurement, the recorded SCE potential was converted into RFIE using the following formula, ERFIE = ESCE + 0.241 + 0.059 x pH. OER and HER polarisation curves were recorded at scan rate of 5 mV s-1 , in 1 M KOH at 25°C with 95% iR corrections, unless otherwise stated. [01 12] For HER testing, a three electrode cell was constructed using a working electrode modified with Ni-Fe NPs with a mass loading of 3.65 mg/cm ² on a carbon fiber paper electrode, a graphite rod counter electrode and a standard calomel electrode (SCE) reference electrode with 1 M KOH electrolyte solution.

[01 13] For OER testing, a three electrode cell was constructed using a working electrode modified with Ni-Fe NPs with a mass loading of 6.5 mg/cm ² on a nickel foam electrode, a graphite rod counter electrode and a standard calomel electrode (SCE) reference electrode with 1 M KOH electrolyte solution.

Example 2.1 : Hydrogen evolution reaction (HER) catalysis

[01 14] Figure 4a shows the HER activity of Ni-Fe NP supported on carbon fiber paper (CFP) by a linear sweep voltammetry (LSV), collected at a scan rate (v) = 5 mV S ¹ , in 1 M KOH. The catalyst mass loading (Figure 10) and mole ratio of Ni to Fe (Figure 8) were optimized based on the HER performance, and the ratio of 5:1 was chosen for HER testing. As evidenced by Figure 4a, the Ni-Fe NP shows an HER onset potential of 0.0 V vs. RHE with concomitant gas evolution observed on electrode surface, indicating H2 production. An overpotential ( q ) of only 100 mV is required to achieve a current density (y) of l OmAcnr ² (corresponding to about 10% solar-to-fuel conversion efficiency), without iR corrections (46 mV with iR corrections, Figure 9). This HER performance is comparable to the benchmark HER electrocatalyst (20% Pt/C) supported on CFP at the same catalyst loading, and outperform other known non-precious metal based HER catalysts. In comparison, significantly higher HER overpotentials are required for Ni NP (260 mV) and Fe NP (410 mV) to achieve l OmAcnr ² , suggesting the distinct HER catalytic activity of the Ni-Fe NPs (Table 4). The Ni-Fe NP also exhibit excellent electrochemical stability for prolonged HER of 10 hours, as shown by Figure 12.

Table 4. Summary of HER catalytic activity for Ni-Fe NPs and control Ni NPs and FeNPs.

Tafel

(mV) / mV dec Substrate* Notes

Catalyst

10 mA cm ² 20 mA cm ² 100 mA cm ²

Ni-Fe NPs -46 -73 -147 58 CFP

Ni ₃ Fei -154 -214 -323 124 CFP

N^Fe, -288 -348 -470 251 CFP 1 0 M KOH, m,: 3 75 mg cm ²

Ni NP -190 -248 -402 167 CFP

Fe NP -320 -370 -476 186 CFP

Notes: (A) CFP = carbon fibre paper

[01 15] The Tafel plot in Figure 4b shows that Ni-Fe NP exhibits a slope of 58mVdec \ and an exchange current density of 1.58 c 10 ^-3 A/cm ² . These results show that the nanoparticles of the invention produce similar values compared with current leading Pt/C catalysts in acidic media. For 20% Pt/C-CFP, a Tafel slope of 50 mV/dec is obtained, consistent to the Volmer- Heyrovsky HER mechanism. This similarity in the Tafel slope values suggests that HER at Ni-Fe NPs also proceeds via the Volmer-Heyrovsky mechanism. The slope value of Ni-Fe NPs is also distinctly different from previously reported Ni-Fe/nanocarbon catalyst for HER (typically above 80 mV dec ¹ ), suggesting HER mechanism due to the distinct interface between the nickel-containing portion and the iron oxide-containing portion in the Ni-Fe NPs.

[01 16] The role of the interface between Ni and Fe2C>3 was evaluated by the preparation of a physical mixture of Ni NP and Fe NP (denoted as Ni/Fe NP) at a molar ratio of 5:1 (Ni:Fe).

The LSV of the Ni/Fe NP mixture exhibited a very different characteristic from the Ni-Fe NP despite of the same catalyst loading used. It is shown in Figure 4c that Ni/Fe NP mixture requires 1 12 mV higher i _j than Ni-Fe NP to achieve a HER j of 10mAcrrr ² . Two additional reduction peaks were observed at 0 V and -0.2 V for the Ni/Fe NP mixture, identical to reduction peaks observed for on the reduction process of Fe ³⁺ in Fe NP (Figure 1 1 ).

Example 2.2: Oxygen evolution reaction (OER) catalysis

[01 17] The Ni-Fe NPs also exhibit high catalytic activity towards OER. Ni-Fe NPs were supported on nickel foam (NF; Figure 4d) or carbon fibre paper (CFP; Figure 13), and their OER performance was examined in 1 M KOH. Nickel-iron layered double hydroxide (NiFe-LDH) electrocatalyst was also synthesized according to established methodology to provide a comparison. Figure 4d shows that the OER process catalysed by Ni-Fe NP initiates at about 1 .41 V vs. RHE. A current density of 10 rmAcnr ² , can be achieved at a low A; of 210 mV. Higher j = 20 mA/cm ² and 100 mA/cm ² were achieved at h of 230 mV and 270 mV, respectively (Table 5). The values are comparable to the most efficient OER catalysts in 1 M KOH, including NiFe LDH and 20% Ir/C. Figure 14 shows the Tafel plots derived from the LSV in Figure 4d. These Tafel plots reveal that Ni-Fe NP exhibits the lowest slope, and the linearity of the plot is also maintained at high j, indicating fast electron transfer and mass transport properties of the catalyst. The Ni-Fe NP also exhibits excellent electrochemical stability for prolonged OER (Figure 15).

Table 5. Summary of OER catalytic activity

Example 3: Electrolytic water-splitting

[01 18] Two-electrode water electrolysis cell was constructed from two nickel foam (NF) electrodes modified with Ni-Fe NP with mass loading of 6.5 mg/cm ² in 1 M KOH. Both counter and reference electrode connections were connected to a single Ni-Fe NP cathode, while working electrode was connected to a separate Ni-Fe NP anode (see Figure 4e).

[01 19] The linear sweeping voltammetry (LSV) of the Ni-Fe NP cell is recorded and compared to a similar cell constructed using benchmark noble metal catalysts, 20% Ir/C (OER) and 20% Pt/C (HER). All voltammetry was collected in 1 M KOH, with scan rate of 5 mV s ^_1 . Shown in Figure 4e, the Ni-Fe NP cell displayed a superior performance than the Ir/C | Pt/C cell, the cell potential required for Ni-Fe NP cell to achieve 10 mA cnr ² is only 1 .47 V (1.55 V without /F?-correction, Figure 19; Table 6). The stability of the Ni-Fe NP catalyst was tested by conducting bulk water electrolysis at j = 10 mA cnr ² and 20 mA cnr ² , respectively, over 24 hours (Figure 4f). In comparison, the cell voltage using benchmark noble metal catalysts, 20% Ir/C (OER) and 20% Pt/C (HER) voltage is significantly higher and shows gradual degradation after 2 hours of testing at 10mAcrrr ² . The energy efficiency of the Ni-Fe NP cell is calculated to be 83.7% with iR correction and 73.4% without iR correction.

Table 6. Summary of results from alkaline water electrolyser experiments in 2-electrode configuration.