Method of producing iron

Abstract

The invention provides a method of producing iron, the method comprising: contacting a molten salt electrolyte comprising lithium, carbonate and dissolved iron oxide with a cathode and an anode in an electrolytic cell; and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode, wherein the anode comprises a ferrous metal composition, and wherein iron corrodes from the ferrous metal composition of the anode at a rate less than 20% of the rate of forming solid metallic iron at the cathode.

Description

[0001] Method of producing iron

[0002] [1] The present application claims priority from Australian provisional patent application No. 2024904193 filed on 18 December 2024, the contents of which should be considered to be incorporated into this specification by this reference.

[0003] Technical Field

[0004] [2] The invention relates to a method of producing iron. The method comprises contacting a molten salt electrolyte comprising lithium, carbonate and dissolved iron oxide with a cathode and an anode in an electrolytic cell, and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode. The anode comprises a ferrous metal composition, for example substantially pure iron, and the rate of iron corrosion from the ferrous metal composition is low compared to the rate of forming solid metallic iron at the cathode. The invention further relates to a system for producing iron, a method of conditioning an anode for oxygen evolution, and an anode for oxygen evolution.

[0005] Background of Invention

[0006] [3] Production of iron through electrolysis of iron oxide offers a green alternative for ironmaking which addresses the environmental problems associated with the traditional carbothermal route. Most iron is currently produced in blast furnaces with carbon or other carbon-containing material used as reducing agent for iron oxide, and the resultant CO2 forms a large fraction of total global anthropogenic CO2 emissions. By contrast, the electrolytic approach provides the opportunity to use low carbon-intensity renewable electricity as the energy source for iron oxide reduction. Other advantages which might be achieved in an electrolytic ironmaking process include high energy efficiency, high product purity, utilization of lower grades of iron ore, and decentralized production operated with smaller or modular units.

[0007] [4] Low-temperature (e.g. less than 130°C) iron oxide electrolysis is theoretically more energy efficient than higher temperature processes, but is difficult to implement due to the poor solubility of iron species in the electrolyte. Iron oxide particles can be reduced while suspended in aqueous electrolytes, but this approach faces severe challenges due to competing hydrogen evolution at the cathode, maintaining effective physical contact between the iron oxide and the cathode, high resistivity of the electrolyte, and contamination of the iron product by unreduced iron oxides. High-temperature processes with electrolytes capable of dissolving high concentrations of iron oxide are thus generally preferred. [5] Molten oxide electrolysis (MOE), with similarities to the Hall-Heroult process for aluminium smelting, seeks to produce liquid iron and dioxygen (O2) by utilising a mixture of oxides as an effective solvent for iron oxide. The production of iron above its melting point (1538°C) advantageously allows continuous removal of the liquid iron product from the cell and immediate further metallurgical processing. However, MOE faces significant practical and economic challenges, including substantial energy consumption at the electrolysis temperatures above 1538°C, identification of inert anode materials capable of withstanding the extreme temperature and corrosive environment, transfer of impurities from low-grade iron ores feed to molten iron product, and high capital and operational costs associated with advanced refractory materials.

[0008] [6] Molten salt electrolysis (MSE) offers an alternative approach at an intermediate temperature typically in the range of 600-1000°C. At such temperatures, iron oxide can be solubilised and electrolytically reduced at higher rates than low-temperature processes, but the challenges of operating above the iron melting point are at least partially mitigated.

[0009] [7] Various electrolytes for MSE processes have been investigated, containing molten chloride salts, fluoride salts, carbonate salts, and mixtures thereof. Depending on the solubility of iron oxide in the chosen electrolyte, reduction can be carried out in the solid state (FFC-Cambridge process) or in solution (the dissolution-deposition process). For example, iron oxide pellets have been directly reduced to metallic iron in the solid state at 800°C in a molten chloride electrolyte (NaCI and CaCl2) at a potential less than 1.2V, achieving an efficiency of 95 %. However, separating the metallic iron product from unreacted iron oxide in the pellet remains challenging.

[0010] [8] By contrast, the dissolution-deposition method is expected to produce a high purity solid iron product because the iron oxide precursor is dissolved in the electrolyte. However, iron oxide has undesirably low solubility in molten chloride salts (typically less than 1 wt.% at 850°C) and various molten carbonate salts such as sodium and potassium carbonates (Na2COs and K2CO3).

[0011] [9] A significant development towards the realisation of an industrial MSE process was the discovery that iron oxides such as ferric oxide (Fe2Os) have excellent solubility in molten carbonate electrolytes containing a significant fraction of lithium carbonate (U2CO3). U2CO3 melts at 723°C and can dissolve about 18 wt.% Fe2Os in the presence of 3.4 wt% of added lithium oxide (U2O) at 750°C. Eutectic mixtures of U2CO3 with other carbonates melt at significantly lower temperatures (499°C for binary eutectic Lii.oyNao.gsCOs; 393°C for ternary eutectic LiossNao eiKo^COs) and remain capable of dissolving lower but still significant quantities of iron oxide.

[0010] The presence of lithium is essential for iron solubility, and it is understood that this is due to the formation of lithiated iron oxide (LiFeO2) in solution. In the absence of added U2O, LiFeC>2 formation is accompanied by release of CO2 according to equation (1), whereas the presence of added U2O facilitates LiFeC>2 formation without electrolyte decomposition and CO2 release by equation (2).

[0012] U2CO3 + Fe2C>3 — > 2LiFeC>2 + CO2 equation (1)

[0013] U2O + Fe2C>3 — > 2LiFeC>2 equation (2)

[0014]

[0011] During electrolysis of iron oxides dissolved in lithium carbonate-based electrolytes, LiFeC>2 is believed to be the active electrochemical intermediate, with reduction occurring via equation (3), with cathodic and anodic half reactions according to equations (3a) and (3b) respectively.

[0015] 2LiFeC>2 — > 2Fe + U2O + 3 / 202 equation (3)

[0016] 2LiFeC>2 + 6e_— > 2Fe + U2O + 3O2' equation (3a)

[0017] 3O2' — > 3 / 202 + 6e_equation (3b)

[0018]

[0012] U2O is thus regenerated in solution and available for further dissolution of iron oxide according to equation (2). The overall reduction of ferric oxide is thus according to equation (4), in which U2O is neither consumed nor produced.

[0019] Fe2C>3 — > 2Fe + 3 / 202 equation (4)

[0020]

[0013] The anode plays a critical role in high temperature electrolysis processes, since it must facilitate oxygen evolution according to equation (3b) with high faradaic efficiency and rates yet remain stable against chemical, electrochemical and physical degradation mechanisms at high temperature and overpotential, and in the presence of O2. Carbon anodes, as used in the Hall-Heroult process, are technically feasible but incompatible with decarbonisation goals due to high consumption of carbon to form CO2. Noble metals such as platinum or iridium are suitably inert for non-consumable anodes but are economically unfeasible. Most industrially focused work on MSE has therefore sought to use nickel-based anodes such as nickel metal, nickel alloys or nickel oxide.

[0021]

[0014] Although nickel-based anodes can provide excellent short-term electrochemical performance, they are not satisfactorily stable at higher temperatures and for longer reaction times. Corrosion of nickel or other metals from the anode into the electrolyte not only risks destabilising the anode but may also contaminate the electrodeposited iron product. These issues are a particular concern when operating with a IJ2CO3 single carbonate electrolyte to maximise iron solubility, since electrolysis temperatures of well above 723°C are typically required to maintain the electrolyte in the molten state.

[0015] There is therefore an ongoing need for methods and systems for producing iron which at least partially address one or more of the above-mentioned short-comings, or provide a useful alternative.

[0022]

[0016] A reference herein to a patent document or other matter which is given as prior art is not to be taken as an admission that the document or matter was known or that the information it contains was part of the common general knowledge as at the priority date of any of the claims.

[0023] Summary of Invention

[0024]

[0017] It has now been discovered that iron oxide may be effectively reduced to a high purity iron product by electrolysis in a molten salt electrolyte containing lithium carbonate when using a moderately consumable ferrous metal anode, preferably a substantially pure iron anode. The rate of iron corrosion from the anode has been found acceptably low, particularly relative to the rate of iron electrodeposition on the cathode, even when operating at high temperatures (> 723°C) and with a lithium carbonate single salt electrolyte. In contrast to other choices of anode material, significant corrosion of the anode material into the electrolyte can be tolerated in this process since it does not lead to contamination of the iron product electrodeposited on the cathode.

[0025]

[0018] The rate of iron corrosion from the anode is particularly suppressed when the anode comprises an aluminium-bearing oxidic layer in the assembly of solid oxidic layers at the interface between the ferrous metal composition and the molten salt electrolyte. A discrete aluminium-bearing oxidic layer may be produced on the anode in the early stages of electrolysis, or in a preliminary conditioning step, by providing a dissolved aluminium species in the molten salt electrolyte, for example by contacting the molten salt electrolyte with alumina (AI2O3). Without wishing to be limited by any theory, the aluminium-bearing oxidic layer, which is believed to comprise lithium aluminate (UAIO2), may protect the bulk ferrous metal composition from rapid oxidation of iron while still providing sufficient conductivity to allow effective oxygen evolution at the anode surface.

[0026]

[0019] In accordance with a first aspect, disclosed herein is a method of producing iron, the method comprising: contacting a molten salt electrolyte comprising lithium carbonate and dissolved iron oxide with a cathode and an anode in an electrolytic cell; and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode, wherein the anode comprises a ferrous metal composition, and wherein iron corrodes from the ferrous metal composition of the anode at a rate less than 20% of the rate of forming solid metallic iron at the cathode.

[0020] In some embodiments, the temperature of the molten salt electrolyte is greater than 650°C, such as greater than 700°C, for example greater than 723°C, when flowing the current through the molten salt electrolyte.

[0027]

[0021] In some embodiments, the ferrous metal composition comprises at least 70 wt.% iron, or at least 80 wt.% iron, such as at least 90 wt.% iron, for example at least 95 wt.% iron.

[0028]

[0022] In some embodiments, the ferrous metal composition does not contain nickel. In some embodiments, the ferrous metal composition does not contain nickel, chromium or tin.

[0029]

[0023] In some embodiments, the ferrous metal composition is substantially pure iron.

[0030]

[0024] In some embodiments, the molten salt electrolyte further comprises a dissolved aluminium species during at least part of the time that current is passed through the molten salt electrolyte. In some embodiments, the dissolved aluminium species is MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the dissolved aluminium species comprises UAIO2. In some embodiments, the dissolved aluminium species is not a silicate species, such as MAISiC where M is Li, Na, K. The dissolved aluminium species may be produced by contacting the electrolyte with a suitable aluminium source such as aluminium oxide (AI2O3) or an aluminate.

[0031]

[0025] In some embodiments, the anode is contacted with a molten salt composition comprising lithium, carbonate and a dissolved aluminium species before contacting the molten salt electrolyte with the anode. In some embodiments, the dissolved aluminium species is MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the dissolved aluminium species comprises UAIO2. In some embodiments, the dissolved aluminium species is not a silicate species, such as MAISiC where M is Li, Na, K. The molten salt composition may be the molten salt electrolyte before addition of iron oxide, or it may be a different composition. In some embodiments, a current is passed through the molten salt composition between a cathode and the anode during the contacting.

[0032]

[0026] The anode may comprise one or more solid oxidic layers at an interface between the ferrous metal composition and the molten salt electrolyte during at least part of the time that current is passed through the molten salt electrolyte. In some embodiments, the one or more solid oxidic layers comprise an aluminium-bearing oxidic layer. In some embodiments, the aluminium-bearing oxidic layer comprises MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the aluminium-bearing oxidic layer comprises UAIO2. In some embodiments, the aluminium-bearing oxidic layer does not comprise an aluminosilicate species, such as MAISiC where M is Li, Na, K.

[0027] In some embodiments, the one or more solid oxidic layers further comprise an iron oxide-rich layer adjacent the ferrous metal composition of the anode. The iron oxide-rich layer may thus be located between the aluminium-bearing oxidic layer and the ferrous metal composition of the anode.

[0033]

[0028] In some embodiments, the one or more solid oxidic layers comprise (i) a first iron oxide-rich layer adjacent the ferrous metal composition of the anode, and (ii) a second iron oxide-rich layer proximate to the molten salt electrolyte, wherein the aluminium-bearing oxidic layer separates the first and second iron oxide-rich layers.

[0034]

[0029] In some embodiments, iron corrodes from the ferrous metal composition of the anode at a rate less than 15%, preferably less than 10%, more preferably less than 5%, yet more preferably less than 3%, of the rate of forming solid metallic iron at the cathode.

[0035]

[0030] In some embodiments, iron corrodes from the ferrous metal composition of the anode at a rate of greater than 1% of the rate of forming solid metallic iron at the cathode.

[0036]

[0031] In some embodiments, iron corrodes from the ferrous metal composition of the anode at a rate of less than 75 mm / year, preferably less than 50 mm / year, such as less than 30 mm / year.

[0037]

[0032] In some embodiments, lithium is present in the molten salt electrolyte in an amount of at least 80 mol%, such as at least 90%, for example at least 95 mol%, of total alkali and alkali earth metals.

[0038]

[0033] In some embodiments, the method comprises adding an iron oxide-rich composition, such as iron ore or a concentrate thereof, to the molten salt electrolyte, thereby providing the dissolved iron oxide.

[0039]

[0034] In some embodiments, the method comprises adding lithium oxide to the molten salt electrolyte.

[0040]

[0035] In some embodiments, at least a portion of the lithium and the dissolved iron oxide is present in the electrolyte as LiFeC>2.

[0041]

[0036] In some embodiments, current is passed through the molten salt electrolyte at a current density of between 100 mA. cm-2and 1000 mA. cm-2, such as between 300 mA. cm-2and 500 mA. cm-2, normalised to the working surface area of the cathode.

[0042]

[0037] In accordance with a second aspect, disclosed herein is a system for producing iron, the system comprising: an electrochemical cell comprising a molten salt electrolyte comprising lithium, carbonate and dissolved iron oxide in contact with a cathode and an anode; and a power supply to apply a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode, wherein the anode comprises a ferrous metal composition.

[0043]

[0038] In some embodiments, the ferrous metal composition comprises at least 70 wt.% iron, or at least 80 wt.% iron, such as at least 90 wt.% iron, for example at least 95 wt.% iron.

[0044]

[0039] In some embodiments, the ferrous metal composition does not contain nickel. In some embodiments, the ferrous metal composition does not contain nickel, chromium or tin.

[0045]

[0040] In some embodiments, the ferrous metal composition is substantially pure iron.

[0046]

[0041] In some embodiments, the molten salt electrolyte further comprises a dissolved aluminium species. In some embodiments, the dissolved aluminium species is MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the dissolved aluminium species comprises UAIO2. In some embodiments, the dissolved aluminium species is not a silicate species, such as MAISiC where M is Li, Na, K. The dissolved aluminium species may be produced by contacting the electrolyte with a suitable aluminium source such as aluminium oxide or an aluminate.

[0047]

[0042] In some embodiments, the anode comprises one or more solid oxidic layers at an interface between the ferrous metal composition and the molten salt electrolyte. In some embodiments, the one or more solid oxidic layers comprises an aluminium-bearing oxidic layer. In some embodiments, the aluminium-bearing oxidic layer comprises MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the aluminium- bearing oxidic layer comprises UAIO2. In some embodiments, the aluminium-bearing oxidic layer does not comprise an aluminosilicate species, such as MAISiC where M is Li, Na, K.

[0048]

[0043] In some embodiments, the one or more solid oxidic layers further comprise an iron oxide-rich layer adjacent the ferrous metal composition of the anode. The iron oxide-rich layer may thus be located between the aluminium-bearing oxidic layer and the ferrous metal composition of the anode.

[0049]

[0044] In some embodiments, the one or more solid oxidic layers comprise (i) a first iron oxide-rich layer adjacent the ferrous metal composition of the anode, and (ii) a second iron oxide-rich layer proximate to the molten salt electrolyte, wherein the aluminium-bearing oxidic layer separates the first and second iron oxide-rich layers.

[0050]

[0045] In some embodiments, lithium is present in the molten salt electrolyte in an amount of at least 80 mol%, such as at least 90%, for example at least 95 mol%, of total alkali and alkali earth metal.

[0051]

[0046] In accordance with a third aspect, disclosed herein is a method of producing iron, the method comprising: contacting a molten salt electrolyte comprising lithium, carbonate, dissolved iron oxide and a dissolved aluminium species with a cathode and an anode in an electrolytic cell, wherein the anode comprises a ferrous metal composition; and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby: (i) forming one or more solid oxidic layers on the ferrous metal composition at an interface between the anode and the molten salt electrolyte, the one or more solid oxidic layers comprising an aluminium-bearing oxidic layer, and (ii) reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode, wherein iron corrodes from the ferrous metal composition of the anode at a rate less than 10% of the rate of forming solid metallic iron at the cathode after forming the one or more solid oxidic layers.

[0052]

[0047] In some embodiments, iron corrodes from the ferrous metal composition of the anode at a rate less than 5%, preferably less than 3%, of the rate of forming solid metallic iron at the cathode after forming the one or more solid oxidic layers.

[0053]

[0048] Various embodiments of the method according to the third aspect may generally have any features as disclosed herein in the context of the first aspect.

[0054]

[0049] In accordance with a fourth aspect, disclosed herein is a method of conditioning an anode for oxygen evolution, the method comprising: contacting a molten salt electrolyte comprising lithium, carbonate, optionally dissolved iron oxide, and a dissolved aluminium species with a cathode and an anode in an electrolytic cell, wherein the anode comprises a ferrous metal composition; and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby providing one or more solid oxidic layers on the ferrous metal composition at an interface between the anode and the molten salt electrolyte, wherein the one or more solid oxidic layers comprise an aluminium-bearing oxidic layer.

[0055]

[0050] The method may be used to condition the anode for oxygen evolution during electrolytic production of iron, particularly in a molten salt electrolyte.

[0056]

[0051] In some embodiments, the dissolved aluminium species is MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the dissolved aluminium species comprises UAIO2. In some embodiments, the dissolved aluminium species is not a silicate species, such as MAISiC where M is Li, Na, K. The dissolved aluminium species may be produced by contacting the electrolyte with a suitable aluminium source such as aluminium oxide (AI2O3) or an aluminate.

[0057]

[0052] In some embodiments, the aluminium-bearing oxidic layer comprises MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the aluminium- bearing oxidic layer comprises UAIO2. In some embodiments, the aluminium-bearing oxidic layer does not comprise an aluminosilicate species, such as MAISiC where M is Li, Na, K.

[0053] In some embodiments, the one or more solid oxidic layers further comprise an iron oxide-rich layer adjacent the ferrous metal composition of the anode. The iron oxide-rich layer may thus be located between the aluminium-bearing oxidic layer and the ferrous metal composition of the anode.

[0058]

[0054] In some embodiments, the one or more solid oxidic layers comprise (i) a first iron oxide-rich layer adjacent the ferrous metal composition of the anode, and (ii) a second iron oxide-rich layer proximate to the molten salt electrolyte, wherein the aluminium-bearing oxidic layer separates the first and second iron oxide-rich layers.

[0059]

[0055] In some embodiments, the ferrous metal composition comprises at least 70 wt.% iron, or at least 80 wt.% iron, such as at least 90 wt.% iron, for example at least 95 wt.% iron.

[0060]

[0056] In some embodiments, the ferrous metal composition does not contain nickel. In some embodiments, the ferrous metal composition does not contain nickel, chromium or tin.

[0061]

[0057] In some embodiments, the ferrous metal composition is substantially pure iron.

[0062]

[0058] In some embodiments, the temperature of the molten salt electrolyte is greater than 650°C, such as greater than 700°C, for example greater than 723°C, when flowing the current through the molten salt electrolyte.

[0063]

[0059] In some embodiments, lithium is present in the molten salt electrolyte in an amount of at least 80 mol%, such as at least 90%, for example at least 95 mol%, of total alkali and alkali earth metals.

[0064]

[0060] In some embodiments, current is passed through the molten salt electrolyte at a current density of between 100 mA. cm-2and 1000 mA. cm-2, such as between 300 mA. cm-2and 500 mA. cm-2, normalised to the working surface area of the cathode.

[0065]

[0061] In accordance with a fourth aspect, disclosed herein is an anode for oxygen evolution, the anode comprising: a ferrous metal composition; and one or more solid oxidic layers at an electrolyte-contacting surface of the ferrous metal composition, the one or more solid oxidic layers comprising an aluminium-bearing oxidic layer.

[0066]

[0062] In some embodiments, the aluminium-bearing oxidic layer comprises MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the aluminium- bearing oxidic layer comprises UAIO2. In some embodiments, the aluminium-bearing oxidic layer does not comprise an aluminosilicate species, such as MAISiC where M is Li, Na, K.

[0067]

[0063] In some embodiments, the one or more solid oxidic layers further comprise an iron oxide-rich layer adjacent the ferrous metal composition of the anode. The iron oxide-rich layer may thus be located between the aluminium-bearing oxidic layer and the ferrous metal composition of the anode.

[0064] In some embodiments, the one or more solid oxidic layers comprise (i) a first iron oxide-rich layer adjacent the ferrous metal composition of the anode, and (ii) a second iron oxide-rich layer proximate to the molten salt electrolyte, wherein the aluminium-bearing oxidic layer separates the first and second iron oxide-rich layers.

[0068]

[0065] Where the terms “comprise”, “comprises” and “comprising” are used in the specification (including the claims) they are to be interpreted as specifying the stated features, integers, steps or components, but not precluding the presence of one or more other features, integers, steps or components, or group thereof.

[0069]

[0066] Further aspects of the invention appear below in the detailed description of the invention.

[0070] Brief Description of Drawings

[0071]

[0067] Embodiments of the invention will herein be illustrated by way of example only with reference to the accompanying drawings in which:

[0072]

[0068] Figure 1 schematically depicts a three-electrode electrolytic cell for iron electrolysis as used in the Examples.

[0073]

[0069] Figure 2 shows cyclic voltammetry (CV) results at 750°C for different lithium carbonate-based electrolytes, as measured in Example 1 .

[0074]

[0070] Figure 3 shows a portion of the cathodic scans of the Figure 2 CV experiments, showing enhanced onset reduction peaks for the negative scan direction.

[0075]

[0071] Figure 4 shows chronoamperometry (CA) results at 750°C and constant potential of -1.4 V vs Ag / AgCI for different lithium carbonate-based electrolytes, as measured in Example 3.

[0076]

[0072] Figure 5 shows CV results with an electrolyte containing U2CO3 + 7 wt.% U2O + 18 wt.% Fe20s at 750°C for different cathode compositions, as measured in Example 4.

[0077]

[0073] Figure 6 shows CA results with an electrolyte containing U2CO3 + 8 wt.% U2O + 18 wt.% Fe20s at 750°C for different cathode compositions, as measured in Example 4.

[0078]

[0074] Figure 7 shows CV results with an electrolyte containing U2CO3 + 18 wt.% Fe2Os at 750°C for different anode compositions, as measured in Example 5.

[0079]

[0075] Figure 8 shows CA results with an electrolyte containing U2CO3 + 18 wt.% Fe2Os at 750°C for different anode compositions, as measured in Example 5.

[0080]

[0076] Figure 9 shows CV results with an electrolyte containing U2CO3 + 10 wt.% Fe2Os at 750°C under different gas atmospheres, as measured in Example 6.

[0077] Figure 10 shows CV results with an electrolyte containing U2CO3 + 10 wt.% Fe2Os at 750°C, with and without pre-fusing of the electrolyte, as measured in Example 7.

[0081]

[0078] Figure 11 shows an XRD pattern of iron deposits produced on the cathode in the CA experiment of Example 3, as measured in Example 8.

[0082]

[0079] Figure 12 shows linear sweep voltammetry (LSV) results with an electrolyte containing Li2COs + 8% U2O + 18 wt.% Fe20s at 750°C with an iron anode, as measured in Example 11.

[0083]

[0080] Figure 13 shows CA results with an electrolyte containing U2CO3 + 8% U2O + 18 wt.% Fe2C>3 at 750°C with an iron anode, as measured with different applied potentials in Example 11.

[0084]

[0081] Figure 14 shows chronopotentiometry (CP) results with an electrolyte containing U2CO3 + 6.7% U2O + 18 wt.% Fe2C>3 at 750°C with an iron anode, as measured over different electrolysis reaction times in Example 12.

[0085]

[0082] Figure 15 is a SEM image of iron deposits produced on a graphite cathode in CA experiments with an electrolyte containing U2CO3 + 18 wt.% Fe2C>3 at 750°C, when using a nickel anode, as measured in Example 13.

[0086]

[0083] Figure 16 is a SEM image of iron deposits produced on a graphite cathode in CA experiments with an electrolyte containing IJ2CO3 + 18 wt.% Fe2C>3 at 750°C, when using a tin oxide anode, as measured in Example 13.

[0087]

[0084] Figure 17 is a SEM image of iron deposits produced on a graphite cathode in CA experiments with an electrolyte containing U2CO3 + 18 wt.% Fe2C>3 at 750°C, when using an iron anode, as measured in Example 13.

[0088]

[0085] Figure 18 shows an XRD pattern of an outer oxidic layer on an iron anode after a CA experiment with an electrolyte containing U2CO3 + 8% U2O + 18 wt.% Fe2C>3at 750°C done in Example 11 , as measured in Example 14.

[0089]

[0086] Figures 19-21 are SEM images of a cross-section through an iron anode after a CA experiment with an electrolyte containing IJ2CO3 + 8% U2O + 18 wt.% Fe2C>3 at 750°C done in Example 11 , as measured in Example 14.

[0090]

[0087] Figure 22 is a graph of the elemental composition of indicated areas in Figures 19- 21 , as determined by EDS.

[0091]

[0088] Figure 23 is an EDS line scan along a cross-section through an iron anode after a 2 hour CP experiment with an electrolyte containing U2CO3 + 6.7% U2O + 18 wt.% Fe2C>3 at 750°C, as produced in Example 15.

[0089] Figure 24 is an EDS line scan along a cross-section through an iron anode after a 48 hour CP experiment with an electrolyte containing U2CO3 + 6.7% U2O + 18 wt.% Fe2Os at 750°C, as produced in Example 15.

[0092]

[0090] Figure 25 is a SEM image of a cross-section through an iron anode after a 48 hour CP experiment with an electrolyte containing U2CO3 + 6.7% U2O + 18 wt.% Fe2C>3 at 750°C, as produced in Example 15.

[0093] Detailed Description

[0094] Method of producing iron

[0095]

[0091] The disclosure relates to a method of producing iron. The method comprises contacting a molten salt electrolyte comprising lithium, carbonate and dissolved iron oxide with a cathode and an anode in an electrolytic cell. A potential difference is applied between the cathode and the anode sufficient to pass a current through the molten salt electrolyte. Dissolved iron is thus reduced at the cathode to form solid metallic iron and dioxygen gas is produced at the anode.

[0096]

[0092] The anode comprises a ferrous metal composition, which thus comprises iron as the predominant metallic element. Iron may be corroded from the ferrous metal composition of the anode into the electrolyte during the electrolysis, but at an acceptably low rate. In particular, iron corrodes from the ferrous metal composition of the anode at a rate substantially less than the rate of forming solid metallic iron at the cathode. Iron may corrode from the ferrous metal composition of the anode at a rate less than 20% of the rate of forming solid metallic iron at the cathode, and in some embodiments at substantially lower rates such as less than 10%, or less than 5%.

[0097]

[0093] The disclosure further relates to a system for producing iron, which is typically suitable for performing the disclosed method. The system comprises an electrochemical cell comprising a molten salt electrolyte comprising lithium, carbonate and dissolved iron oxide in contact with a cathode and an anode, and a power supply to apply a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode. The anode comprises a ferrous metal composition.

[0098] Molten salt electrolyte

[0099]

[0094] The molten salt electrolyte used in the methods disclosed herein contains lithium, carbonate and dissolved iron oxide. The molten salt electrolyte may thus comprise lithium carbonate (U2CO3), although it is not excluded that other lithium salts may be added to a molten carbonate electrolyte to provide lithium therein.

[0095] In some embodiments, the molten salt electrolyte further comprises other alkali metals, in particular sodium and / or potassium. The molten salt electrolyte may thus include one or more alkali metal carbonates, such as sodium carbonate (Na2COs) and / or potassium carbonate (K2CO3), in combination with U2CO3.

[0100]

[0096] Combination of Na2COs and / or K2CO3 with U2CO3 may lower the melting point of the molten salt electrolyte compared to a single-carbonate Li2CO3-based electrolyte, thus in principle allowing molten salt electrolysis to be conducted at lower temperatures. For example, the binary eutectic Lii.oyNao.gsCOs melts at 499°C, while the ternary eutectic Lio.ssNao.eiKo^COs melts at 393°C in comparison with U2CO3 which melts at 723°C.

[0101]

[0097] The molten salt electrolyte generally comprises alkali metal carbonate (i.e. U2CO3 together with any Na2COs and / or K2CO3) as the primary molten salt component. Thus, in some embodiments, the molten salt electrolyte comprises at least 50 wt.% alkali metal carbonate, such as at least 60 wt.% or at least 70 wt.% alkali metal carbonate, for example or at least at least 80 wt.% alkali metal carbonate, all based on the total weight of the molten salt electrolyte. Dissolved iron oxide may form a major portion of the remainder of the electrolyte composition.

[0102]

[0098] In some embodiments, lithium is the predominant alkali or alkali earth metal present in the molten salt electrolyte. Lithium carbonate may be the main or only metal carbonate component intentionally present in the molten salt electrolyte. In practice, however, the introduction of small amounts of other alkali or alkali earth metals may be unavoidable, e.g. via the iron oxide feedstock. Thus, in some embodiments, lithium is present in the molten salt electrolyte in an amount of at least 80 mol%, such as at least 90%, or at least 95 mol%, for example as at least 99 mol%, of the total amount of alkali and alkali earth metal in the molten salt electrolyte.

[0103]

[0099] Higher concentrations of U2CO3 in the molten salt electrolyte may be advantageous due to the increased solubility of iron oxide. Since LiFeC>2 is understood to be both the dissolved form of iron oxide and the active electrochemical species in electrolysis, higher lithium and thus LiFeC>2 concentrations will increase the rate of electrolysis to solid metallic iron. By contrast, iron oxide has a very low solubility in pure molten sodium carbonate or potassium carbonate.

[0104]

[0100] However, molten salt electrolytes containing high concentrations of U2CO3 must be maintained at high temperatures, such as above 723°C, to remain in the liquid state, increasing the challenge of identifying a suitable anode capable of acceptably withstanding the electrolysis conditions.

[0105]

[0101] The electrolyte may contain a high concentration of dissolved iron oxide, such as at or near the saturation concentration, e.g. more than 70%, more than 80%, or more than 90%, such as more than 95%, of the saturation concentration. Higher concentrations are expected to increase the rate of iron production. Without wishing to be limited by any theory, high concentrations of dissolved iron oxide may also beneficially lower the dissolution rate of an iron oxide-rich oxidic layer formed in a protective assembly of solid oxidic layers on the ferrous metal of the anode during electrolysis.

[0106]

[0102] The method may include a step of adding an iron oxide-rich composition to the molten salt electrolyte, thereby providing the dissolved iron oxide in the electrolyte. The iron oxide-rich composition may be added to and dissolved in the electrolyte within the electrochemical cell, thus directly replenishing iron oxide which is consumed in the electrolysis. Alternatively, it may be added to and dissolved into the molten salt electrolyte outside of the cell, for example in a dissolution vessel, with the molten salt electrolyte containing dissolved metal oxide being circulated into the electrolytic cell.

[0107]

[0103] The iron oxide-rich composition may be any suitable iron oxide-bearing feedstock suitable for electrolysis. Typically, the composition will include one or both of the two main mineralogically significant forms of iron oxide, being Fe2Os and FesOln some embodiments, the iron oxide-rich composition is an iron ore or a concentrate thereof. Non-limiting examples of suitable iron ores may be hematite-rich iron ore or a magnetite-rich iron ore.

[0108]

[0104] In some embodiments, the molten salt electrolyte further comprises a dissolved aluminium species. The dissolved aluminium species may be MAIO2, where M is one or more alkali metals. In particular, the dissolved aluminium species may comprise UAIO2. The electrolyte may comprise only a low amount of the dissolved aluminium species, such as less than 5 wt.%, based on equivalent AI2O3 as a percentage of the total electrolyte mass. AI2O3 has only a low solubility (as UAIO2) in lithium carbonate-based molten salt electrolytes.

[0109]

[0105] The presence of such dissolved aluminium species in the molten oxide electrolyte has been found to beneficially protect the ferrous metal anode from high rates of corrosion. Without wishing to be limited by any theory, it is proposed that the dissolved aluminium species forms a discrete, protective aluminium-bearing oxidic layer, believed to contain lithium aluminate (UAIO2), in the assembly of solid oxidic layers on the ferrous metal surface.

[0110]

[0106] The dissolved aluminium species may be produced by contacting the molten salt electrolyte with an aluminium source such as aluminium oxide (AI2O3) or a suitable equivalent, such as an aluminate, e.g. MAIO2, where M is one or more alkali metals, suitably Li.

[0111]

[0107] The molten salt electrolyte may comprise the dissolved aluminium species, such as UAIO2, throughout the electrolytic iron reduction reaction. However, given the proposed mode of operation, the dissolved aluminium species may be present during only part of the time that current is passed through the molten salt electrolyte. Once the protective aluminium-bearing oxidic layer is formed, it is expected to persist on the anode even if the molten salt electrolyte is substantially aluminium-free thereafter.

[0112]

[0108] It has been found that addition of aluminosilicates to the molten salt electrolyte has a much smaller protective effect on the ferrous metal composition of the anode than free alumina. As reported by Moradmand et al in Fuel 291 (2021) 120215, aluminosilicates such as kaolinite dissolve in molten alkali carbonates as silicate species, in particular MAISiC where M is Li, Na, K. Without wishing to be bound by any theory, such species are believed to be less capable of forming a protective aluminium-bearing oxidic layer on the ferrous metal anode than UAIO2 due to lower electronegativity. In some embodiments, therefore the molten salt electrolyte comprises a dissolved aluminium species which is not a silicate species.

[0113]

[0109] The molten salt electrolyte may comprise other dissolved metal species, preferably only in minor amounts. Such metal components may include metals derived from gangue minerals in the iron oxide-rich composition. In practice, iron ores will likely contain some aluminium and silicon present in gangue silicate and aluminosilicate species, and other impurity metals. Such species typically have only low solubility in molten carbonate electrolytes and are not expected to significantly influence the solubility of iron oxide or unacceptably affect the electrolysis of dissolved iron oxide to solid metallic iron. Iron ores generally do not contain free alumina.

[0114]

[0110] The method may include a step of adding lithium oxide (U2O) to the molten salt electrolyte. U2O may facilitate the dissolution of iron oxide, for example via equation (2), and reduce the release of CO2 from the electrolyte via equation (1). However, iron oxide dissolution and electrolysis have been found effective in the absence of extrinsically added U2O, and in some scenarios U2O addition may increase the resistivity of the electrolyte. In some embodiments, therefore, U2O is not added to the molten salt electrolyte, although it will be understood that U2O may be formed in situ as an intermediate in the electrolysis reaction via equation (3).

[0115] Electrolytic cell

[0116]

[0111] The system disclosed herein comprises an electrolytic cell comprising a molten salt electrolyte, and a power supply to apply a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte. The method disclosed herein comprises contacting the molten salt electrolyte with a cathode and an anode in the electrolytic cell, and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte.

[0117]

[0112] It will be appreciated that the electrolytic cell generally comprises a vessel suitable to contain the molten salt electrolyte, with the electrodes arranged within the vessel to contact the contained electrolyte. The vessel walls may thus be made from a refractory material which can withstand the high operating temperatures and corrosive environment of electrolysis, and which does not unacceptably contaminate the molten salt electrolyte with impurities.

[0118]

[0113] The cathode of the electrolytic cell generally comprises at least one solid, electrically conductive body which is exposed to the molten salt electrolyte, for example by immersion therein, and configured to receive solid metallic iron deposited thereon. The cathode may thus comprise any electrically conductive material capable of withstanding the operating temperature and exposure to the molten salt electrolyte under electrolytic conditions. Suitable materials may include carbon-based materials such as graphite and metallic compositions.

[0119]

[0114] Preferably, the cathode comprises a ferrous metal composition such as substantially pure iron. Most or all of the solid metallic iron produced by the electrolytic reduction of dissolved iron is deposited on the cathodic working surface. Therefore, the use of a ferrous metal cathode avoids or limits the contamination of the iron product by the cathode composition.

[0120]

[0115] The cathode provides a cathodic working surface in contact with the molten salt electrolyte. The cathode should have a geometry and working surface area of the cathode sufficient to provide the required current density (in A. cm-2) through the cell, and to receive iron deposits. In some embodiments, the working surface area of the cathode is less than the working surface area of the anode.

[0121]

[0116] Optionally, the electrolytic cell comprises a reference electrode, as a third electrode in a 3-electode cell configuration, to measure and / or control the absolute potential of either the cathode or the anode. Alternatively, the electrolytic cell lacks a reference electrode. Such arrangements are common in industrial electrolysers operated at constant or near-constant current conditions.

[0122]

[0117] The electrolytic cell is coupled to a power supply suitable to apply the required potential for electrolysis. The power supply is a direct current power supply which is preferably powered by renewable electricity, thus decarbonising the production of iron.

[0123] Anode

[0124]

[0118] The electrolytic cell comprises an anode comprising a ferrous metal composition. As used herein, a ferrous metal composition refers to iron metal or an iron alloy in which iron is the main metal element, typically more than half of the composition by weight.

[0125]

[0119] The anode of the electrolytic cell generally comprises at least one solid, electrically conductive body composed of the ferrous metal composition. The ferrous metal composition is thus exposed to the molten salt electrolyte, for example by immersion therein. The anode should have a geometry and working surface area sufficient to provide the required current density (in A. cm-2) through the cell, even if the anode is partially consumed during the course of the ongoing reaction (at a consumption rate consistent with a moderately consumable anode). In some embodiments, the working surface area of the anode is greater than the working surface area of the cathode, for example at a ratio of at least 1.1 : 1 , or at least 1.2:1 , or at least 1.5:1 , or at least 1 .75: 1 , for example about 2: 1 .

[0126]

[0120] To avoid contamination of the solid metallic iron product derived from the anode, the ferrous metal composition of the anode preferably contains a high fraction of iron. In some embodiments, the ferrous metal composition comprises at least 70 wt.% iron, or at least 80 wt.% iron, or at least 90 wt.% iron, such as at least 95 wt.% iron, for example at least 97 wt.% iron, or at least 98 wt.% iron. In some embodiments, the ferrous metal composition is substantially pure iron. As used herein, substantially pure iron refers to iron metal with at least 99 wt.% iron content, preferably at least 99.5% iron content, and which thus contains only trace impurity elements remaining after refining the iron used to fabricate the anode.

[0127]

[0121] In some embodiments, the ferrous metal composition contains less than 5 wt.% nickel, and preferably less than 1 wt.% nickel, such as less than 0.5 wt.% nickel. In some embodiments, the ferrous metal composition does not contain nickel. Nickel metal anodes have been found to corrode into molten salt electrolytes, such as Li2COs at 750°C, and thus to contaminate the solid metallic iron produced at the cathode with electrodeposited nickel.

[0128]

[0122] In some embodiments, the ferrous metal composition does not contain tin. In some embodiments, the ferrous metal composition does not contain chromium. In some embodiments, the ferrous metal composition does not contain nickel, chromium or tin.

[0129]

[0123] As used herein, a reference to a ferrous metal composition which does not contain an identified metal element means that the ferrous metal composition is functionally free of that element, although trace impurities of the element may be present, i.e. less than 0.5 wt.%, preferably less than 0.2 wt.%.

[0130]

[0124] Surprisingly, it has been found that high iron-content anodes, even substantially pure iron anodes, corrode only slowly when used for oxygen evolution under conditions of iron electrolysis, including at the high temperatures (e.g. 750°C) required when a lithium carbonate- based electrolyte is used. Higher anode corrosion rates were initially expected due to the high susceptibility of iron metal to oxidation under oxygen evolution conditions, forming iron oxide, and the expected transformation of that iron oxide to soluble LiFeC>2 species in the lithium carbonate-based electrolyte.

[0131]

[0125] In particular, the rate of iron corrosion from the anode is sufficiently low relative to the rate of iron electrodeposition on the cathode that anode consumption represents an acceptable and preferably only minor efficiency loss in the electrolysis process. The ferrous metal anode may thus be considered as a moderately consumable anode, in comparison to non-consumable anodes intended to be substantially inert under electrolysis conditions (such as platinum or iridium) or highly consumable anodes (such as carbon). In contrast to other anode materials which might corrode at rates consistent with a moderately consumable anode, corrosion of iron into the electrolyte can be tolerated since it cannot lead to contamination of the iron product electrodeposited on the cathode.

[0132]

[0126] Without wishing to be bound by any theory, it is proposed that the surprisingly low rate of iron corrosion from the ferrous metal is due to the formation of a stable assembly of oxides in one or more solid oxidic layers at the interface between the ferrous metal composition and the molten salt electrolyte.

[0133]

[0127] In some embodiments, the anode comprises a first iron oxide-rich layer adjacent the ferrous metal composition, i.e. as the oxidic layer directly in contact with the unoxidized bulk ferrous metal composition. It is understood based on elemental analysis results that under at least some conditions this iron oxide-rich layer may comprise FesC .

[0134]

[0128] In some embodiments, the anode comprises an aluminium-bearing oxidic layer among the solid oxidic layer(s) at the interface between the ferrous metal composition and the molten salt electrolyte. The aluminium-bearing oxidic layer may comprise MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the aluminium- bearing oxidic layer comprises lithium aluminate (UAIO2).

[0135]

[0129] By careful analysis of the oxidic layers on the anode surface after electrolysis, the inventors observed the presence of a solid oxidic layer containing aluminium, with the aluminium believed to be predominantly in the form of UAIO2 on the basis of XRD analysis. This layer was apparently formed initially as the outer solid oxidic layer on the anode. However, after extended periods of electrolysis, the aluminium-bearing oxidic layer was present as a discrete layer in the assembly of solid oxidic layers on the anode surface, separating a first iron oxide-rich layer adjacent to the ferrous metal composition of the anode and a second, outer iron oxide-rich layer proximate to the molten salt electrolyte. It is believed that the outer iron oxide-rich layer was formed by diffusion of iron species, produced by oxidation of the ferrous metal composition, through the porous aluminium-bearing oxidic layer. Without wishing to be bound by any theory, the inventors propose that the aluminium-bearing oxidic layer protects the underlying iron from excessive rates of corrosion, but is sufficiently porous and conductive (unlike alumina) to permit electrolysis at good rates.

[0130] The presence of aluminium on the anode was unexpected since no aluminium- bearing materials were intentionally present in the experiment, but the aluminium source was ultimately identified as the crucible in which the electrolyte was fused.

[0136]

[0131] The aluminium-bearing oxidic layer in the assembly of oxidic layer(s) at the interface between the ferrous metal composition and the molten salt electrolyte can be produced in any suitable manner. In some embodiments, the aluminium-bearing oxidic layer is produced during electrolytic reduction of iron oxide present in a molten salt electrolyte comprising lithium, carbonate, dissolved iron oxide and a dissolved aluminium species as disclosed herein. For example, the aluminium-bearing oxidic layer may be produced in situ on the ferrous metal composition of the anode during the early stages of electrolysis.

[0137]

[0132] Alternatively, the aluminium-bearing oxidic layer may be produced on the anode during a preliminary electrochemical electrode conditioning step, in the presence or absence of dissolved iron oxide in a molten salt (conditioning) electrolyte. This conditioning method may comprise contacting a molten salt electrolyte comprising lithium, carbonate, optionally dissolved iron oxide, and a dissolved aluminium species (as disclosed herein) with a cathode and the anode comprising a ferrous metal composition. A potential difference is applied between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby providing the one or more solid oxidic layers on the ferrous metal composition at the interface between the anode and the molten salt electrolyte. The one or more solid oxidic layers comprises the aluminium-bearing oxidic layer.

[0138]

[0133] This method may be conducted using the same or different electrochemical cell and cathode, and with the same or different molten salt electrolyte, as subsequently used to produce iron. For example, the electrolyte for conditioning may be the same molten salt electrolyte used to electrolytically produce iron but before dissolution of iron oxide in the electrolyte composition. If a different molten salt electrolyte is used for conditioning, it may comprise U2CO3, optionally in combination with Na2COs and / or K2CO3, as disclosed herein in the context of the molten salt electrolyte for producing iron.

[0139]

[0134] It is also envisaged that the aluminium-bearing oxidic layer may be produced under non-electrolytic conditions, for example by pre-contacting the ferrous metal composition of the anode with a molten salt composition comprising lithium, carbonate and a dissolved aluminium species as disclosed here, without passing current through the electrolyte. Alternatively the ferrous metal composition of the anode may be pre-coated with an aluminate layer, e.g. UAIO2, by another coating technique. The anode with protective aluminium-bearing oxidic layer may then be contacted with the molten salt electrolyte for iron electrolysis. Electrolysis conditions

[0140]

[0135] The method comprises applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode.

[0141]

[0136] The potential difference between the cathode and anode is sufficient to drive the electrolysis reaction according to equation (3) at the operating conditions of the electrolytic cell, but without reaching a potential sufficient for significant carbon formation due to electrochemical reduction of carbonate. Thermodynamic calculations show that the minimum theoretical potential difference required for equation (3) is 1.2 V, and in practice greater potential differences than this will be required to overcome ohmic loss, thermal loss, and polarisation loss and achieve a significant flow of current.

[0142]

[0137] The potential difference may be sufficient to achieve a desired current density or maintain the current density within a desired range. Current density is typically a key performance metric in an industrial electrolyser since it correlates with the production rate of the electrolytic product, in this case metallic iron. In some embodiments, the current density is greater than 1 mA. cm-2, such as greater than 10 mA. cm-2, greater than 50 mA. cm-2, or greater than 100 mA. cm-2, for example greater than 300 mA. cm-2, where the current density units are normalised to the electrochemically active working surface area of the cathode. The inventors have demonstrated by experiment that current densities of 100mA. cm-2can be obtained at a stable potential difference when using an iron anode for extended electrolysis times. Higher current densities will generally be achievable with lower temperature and higher concentrations of dissolved iron oxide. However, excessive current densities may lead to high rates of anode consumption and / or contamination of the iron deposits by carbon formed via carbonate reduction. In some embodiments, therefore, the current density is less than 1000 mA. cm-2, such as less than 600 mA. cm-2, or less than 500 mA. cm-2.

[0143]

[0138] The electrolysis temperature, based on the temperature of the molten salt electrolyte in the electrolytic cell when passing current through the molten salt electrolyte, is high enough to maintain the electrolyte substantially in the liquid state. In principle, a temperature of at least about 400°C is required to melt alkali metal carbonate electrolytes (based on the ternary eutectic Lio.ssNao.eiKo^CCh which melts at 393°C) but in practice significantly higher temperature may be required to solubilise iron oxide and reduce iron at appreciable rates. In some embodiments, the temperature of the molten salt electrolyte is greater than 500°C, or greater than 550°C, or greater than 600°C, or greater than 650°C, such as greater than 700°C.

[0139] U2CO3 melts at 723°C, so that electrolyte compositions containing I^CCh as the predominant or only alkali metal carbonate should be maintained at temperatures higher than this during electrolysis. In some embodiments, therefore the temperature of the molten salt electrolyte is greater than 723°C, such as greater than 740°C. Higher temperatures may increase the solubility of iron oxide in the molten salt electrolyte, but will increase energy consumption and the rate of anode corrosion. Accordingly, in some embodiments the temperature of the molten salt electrolyte is less than 1000°C, such as less than 900°C, or less than 800°C. The inventors have demonstrated stable electrolysis at temperatures of 750°C with a Li2CC>3-based electrolyte. The electrolysis reaction according to equation (4) is endothermic, requiring heat input to maintain the electrolytic cell at the required temperature. At scale, it is expected that most or all of the required heat input will be provided by resistive heating in the electrolytic cell, similar to aluminium electrolysis.

[0144]

[0140] The molten salt electrolyte may be maintained under a gas atmosphere in the electrolytic cell. In some embodiments, the gas atmosphere comprises CO2, such as at least 0.1 bar partial pressure. U2CO3 may decompose at high temperatures to form U2O and CO2, and a gas atmosphere comprising CO2 may suppress this decomposition reaction. While a CO2 atmosphere may thus favour electrolyte stability, it may also reduce dissolution of iron oxide in the electrolyte, since formation of LiFeC>2 requires the presence of U2O from the decomposition of Li2CC>3 or from extraneous U2O addition. Moreover, the presence of CO2 in the cell could cause contamination of the iron product with carbon, since the direct reduction of CO2 to C has a lower deposition voltage than iron deposition. In some embodiments therefore, CO2 is not intentionally added to the gas atmosphere. In some embodiments, the CO2 partial pressure in the gas atmosphere contacted with electrolyte is zero or very low, such as less than 0.1 bar, or less than 0.01 bar.

[0145] Iron corrosion from anode

[0146]

[0141] As already discussed herein, the anode comprises a ferrous metal composition which is not necessarily completely inert but which corrodes (or dissolves) at a rate which is low compared to the rate of iron electrodeposition on the cathode.

[0147]

[0142] In some embodiments, iron corrodes from the ferrous metal composition of the anode at a rate less of less than 20%, or less than 15%, such as less than 10%, or less than 5%, for example less than 3%, of the rate of forming solid metallic iron at the cathode. Such corrosion rates, particularly in the lower ranges, can be accommodated as a minor efficiency loss in the process, and without risk of contaminating the iron product.

[0148]

[0143] The anodes in the disclosed methods are envisaged as moderately consumable anodes, and may thus have an appreciable rate of corrosion. For example, iron may corrode from the ferrous metal composition of the anode at a rate of greater than 1%, or greater than 2%, of the rate of forming solid metallic iron at the cathode.

[0149]

[0144] It will be appreciated that the rate of iron corrosion may not be consistent at all times during the electrolysis. In particular, the inventors have found by experiment that high rates of corrosion may be experienced in the early stages of electrolysis, for example within an initial conditioning period of e.g. two hours, but that much slower rates of corrosion are obtained thereafter. Without wishing to be limited by any theory, it is proposed that the unconditioned ferrous metal composition of the anode is susceptible to high rates of iron corrosion, but that an assembly of protective oxidic layers form on the metallic anode surface during the conditioning period which subsequently suppresses the rate of iron corrosion. Accordingly, references herein to particular rates of iron corrosion from the ferrous metal composition of the anode, relative to the formation rate of solid metallic iron at the cathode, do not imply that all time periods during the electrolysis exhibit such corrosion rates.

[0150]

[0145] In some embodiments, iron corrodes from the ferrous metal composition of the anode at a rate less than 10%, such as less than 5%, for example less than 3%, of the rate of forming solid metallic iron at the cathode after forming the oxidic layers.

[0151]

[0146] In some embodiments, iron corrodes from the ferrous metal composition of the anode at a rate less of less than 20%, or less than 15%, such as less than 10%, or less than 5%, for example less than 3%, of the rate of forming solid metallic iron at the cathode over a period of at least 24 hours of electrolysis following an initial conditioning period, e.g. of two hours.

[0152]

[0147] In some embodiments, iron corrodes from the ferrous metal composition of the anode at a rate of less than 75 mm / year, such as less than 50 mm / year, for example less than 30 mm / year. Such consumption rates are consistent with a moderately consumable anode which, while not completely inert, can be used for extended electrolysis reactions of acceptable duration.

[0153] Anode for oxygen evolution

[0154]

[0148] In some preferred embodiments, the anode disclosed herein in the context of the method for producing iron comprises a protective aluminium-bearing oxidic layer in the assembly of solid oxidic layers at the surface of the ferrous metal composition. A similar anode structure can be expected to be useful for other electrolytic reactions conducted with a molten salt electrolyte that involve oxygen evolution at the anode.

[0155]

[0149] Therefore, the disclosure also provides an anode for oxygen evolution. The anode comprises a ferrous metal composition and one or more solid oxidic layers at an electrolyte- contacting surface of the ferrous metal composition, the one or more solid oxidic layers comprising an aluminium-bearing oxidic layer.

[0156]

[0150] The aluminium-bearing oxidic layer may comprise MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the aluminium-bearing oxidic layer comprises lithium aluminate (UAIO2). In some embodiments, the aluminium-bearing oxidic layer does not comprise an aluminosilicate species.

[0157]

[0151] The solid oxidic layers at the anode surface may further comprise an iron oxide-rich layer adjacent the ferrous metal composition of the anode. In some embodiments, the solid oxidic layers comprise (i) a first iron oxide-rich layer adjacent the ferrous metal composition of the anode, and (ii) a second iron oxide-rich layer, wherein the aluminium-bearing oxidic layer separates the first and second iron oxide-rich layers.

[0158]

[0152] The ferrous metal composition of the anode may generally be as described in the section on anodes for iron production. In some embodiments, the ferrous metal composition is substantially pure iron.

[0159]

[0153] The anode for oxygen evolution may generally be produced as already disclosed herein for the specific case of iron production. Therefore, also disclosed herein is a method of conditioning an anode for oxygen evolution. The method comprises contacting a molten salt electrolyte comprising lithium, carbonate and a dissolved aluminium species with a cathode and an anode in an electrolytic cell, wherein the anode comprises a ferrous metal composition. A potential difference is applied between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby providing one or more solid oxidic layers on the ferrous metal composition at an interface between the anode and the molten salt electrolyte. The one or more solid oxidic layers comprise an aluminium-bearing oxidic layer.

[0160]

[0154] The molten salt electrolyte may comprise U2CO3, optionally in combination with Na2CC>3 and / or K2CO3, as disclosed herein in the context of the molten salt electrolyte for producing iron. Optionally, the molten salt electrolyte comprises dissolved iron oxide, for example as LiFeO2. The dissolved aluminium species may be MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the dissolved aluminium species comprises UAIO2. In some embodiments, the dissolved aluminium species is not a silicate species, such as MAISiO4 where M is Li, Na, K. The dissolved aluminium species may be produced by contacting the electrolyte with a suitable aluminium source such as aluminium oxide (AI2O3) or an aluminate

[0161]

[0155] The aluminium-bearing oxidic layer may comprise MAIO2, where M is selected from Li, Na, K and combinations thereof. In some embodiments, the aluminium-bearing oxidic layer comprises lithium aluminate (UAIO2). In some embodiments, the aluminium-bearing oxidic layer does not comprise an aluminosilicate species.

[0162]

[0156] The solid oxidic layers may further comprise an iron oxide-rich layer adjacent the ferrous metal composition of the oxygen-evolving anode. In some embodiments the solid oxidic layers may comprise (i) a first iron oxide-rich layer adjacent the ferrous metal composition of the anode, and (ii) a second iron oxide-rich layer proximate to the molten salt electrolyte, wherein the aluminium-bearing oxidic layer separates the first and second iron oxide-rich layers.

[0163]

[0157] The anode for oxygen evolution may be used for any electrolytic reaction conducted with a molten salt electrolyte that involves oxygen evolution at the anode. Non-limiting examples may include electrolytic reductions of reducible species such as (i) metal oxides or other oxidised metallic species (to produce metal or metal alloy), (ii) carbon dioxide and / or or (iii) water.

[0164]

[0158] Therefore also disclosed herein is a method of electrolysis, the method comprising: contacting a molten salt electrolyte comprising lithium carbonate with a cathode and an anode for oxygen evolution, as disclosed herein, in an electrolytic cell; and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing at least one reducible species at the cathode and producing dioxygen gas at the anode.

[0165] EXAMPLES

[0166]

[0159] The present invention is described with reference to the following examples. It is to be understood that the examples are illustrative of and not limiting to the invention described herein.

[0167] Materials and apparatus

[0168] Electrolyte preparation

[0169]

[0160] Electrolytes were prepared by mixing pre-dried components as required for the composition, including Li2COs, Na2COs, K2CO3, Fe2Os, LiFeC>2 and U2O (>99% purity; Sigma Aldrich). To achieve a well-homogenised electrolyte mixture, two different homogenisation process were carried out: physical mixing and thermal fusing. First, 300g of the components were mixed with a mortar and pestle for 10 minutes, then thermally fused in an alumina crucible (99.8% purity AI2O3) in a N2 environment for 20 hours at 800°C.

[0170] Electrolytic cell

[0171]

[0161] A three-electrode cell was used in most electrochemical tests (unless otherwise stated), as schematically depicted in Figure 1. A cylindrical 316 stainless steel reactor vessel (200), heated with heater 201 and containing a ceramic crucible (202) as the electrolyte holder was fitted with the three electrodes passing through the top of the reactor and heating insulation 203, a gas inlet (204) and outlet (206) to control the experiment atmosphere, and a thermocouple to monitor and control internal reactor temperature. Unless otherwise specified, the ceramic crucible was an alumina crucible (99.8% purity AI2O3). The reactor design allowed in-situ insertion and removal of electrodes. The electrodes included working electrode 208, counter electrode 210 and reference electrode 212. The working and counter electrode included ceramic holders 214 to retain various cathode and anode materials for the experiments.

[0172]

[0162] Various cathode materials tested included graphite rod (6 mm diameter, Alfa Aesar, purity > 99%), Ni and Fe foils (1mm thickness, Goodfellow, purity > 99%). Anode materials included SnC>2 rod (12 mm diameter, Goodfellow, purity > 99%), Fe and Ni foils.

[0173]

[0163] An in-house built Ag / AgCI reference electrode (212) was used as the third electrode of the cell, since no commercial reference electrode is available to withstand the corrosivity of the Li2COs electrolyte at high temperature. Alumina containers previously used for reference electrodes did not provide sufficient ionic conductivity to maintain ionic equilibrium between the reference electrode and the electrolyte. Therefore, a 5% yttria-stabilized zirconia (YSZ) tube (216) was selected as the reference container due to its high thermal stability and superior ionic conductivity at temperatures above 700°C. The reference electrode consisted of a silver wire (218) inserted into a eutectic ternary carbonate (32.1 : 33.4 :34.5 wt.% (Li:K:Na)2 CO3) reference electrolyte (220) containing 1 wt.% AgCl. The surface areas of the anode and cathode were recorded for each test, with the counter electrode’s surface area being 2-3 times larger than that of the working electrode to ensure efficient current distribution and avoid mass transfer limitations.

[0174] Iron electrodeposition

[0175]

[0164] Unless otherwise stated, the following conditions were used for Cyclic Voltammetry (CV), Chronoamperometry (CA) and Chronopotentiometry (CP) experiments. The pre-fused electrolyte (prepared as described above) was heated to the required temperature of the electrochemical experiment, and the electrodes were inserted. The open circuit potential (OCP) was measured for three minutes. In the CV experiments, the potential was scanned from the OCP to -1.6 V vs reference electrode at a scan rate of 5 mV / sec. CA was conducted at -1.4V vs reference electrode for the selected time. CP was conducted at a target current density.

[0176]

[0165] After completing the electrochemical tests, all electrodes were removed from the electrolyte and held above the melt under a nitrogen gas atmosphere during cooling. The product deposited on the cathode was washed with Milli-Q water, under stirring with a magnetic stirrer, to remove any residual solidified electrolyte. The metallic products were collected with a magnet and washed with ethanol to remove any water residue on the surface of metallic iron.

[0177] Anode characterisation

[0178]

[0166] After some of the electrolysis experiments, the anode was subjected to analysis. The anode was thus collected after cooling under a nitrogen (N2) atmosphere to avoid oxidation, then gently washed with water for 30 minutes to remove residual solidified electrolyte from the surface. Subsequently, the anode was analyzed using X-ray diffraction (XRD) on a Co radiation source, with scans performed from 10° to 100° in 20. The X'Pert HighScore software was employed to identify the compounds present on the surface.

[0179]

[0167] To precisely investigate the oxide layers, the anode was sectioned longitudinally to expose its cross-section. The cut segment was embedded in resin to stabilize the sample and ensure accessibility to various oxide layers formed during the process. The prepared sample was then examined using scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) to characterize the microstructure and elemental composition of the oxide layers.

[0180] Example 1. Cyclic voltammetry investigations of electrolyte composition

[0181]

[0168] The impact of electrolyte composition on molten salt electrolysis of Fe2Os was investigated by cyclic voltammetry (CV) for Li2CO3-based electrolytes with a graphite cathode (as working electrode), SnC>2 anode (as counter electrode), reference electrode, operating temperature of 750°C and a gas atmosphere of N2. The starting potential was determined by initially measuring open circuit potential (OCP) of the graphite cathode for a period of 3 minutes. The negative vertex potential was held constant for each experiment at -1.6 V vs Ag / AgCI while the positive vertex was controlled to be OCP+0.1 V. The scan rate was 5 mV / s.

[0182]

[0169] The following electrolytes were investigated: (i) U2CO3 + 8 wt.% U2O, (ii) U2CO3 + 18 wt.% Fe2C>3, (iii) U2CO3 + 8 wt.% U2O + 18 wt.% Fe2Os and (iv) U2CO3 + 21 wt.% LiFeC>2, and the results are shown in Figure 2 (full initial scan) and Figure 3 (portion of cathodic scan showing enhanced onset reduction peaks for the negative scan direction).

[0183]

[0170] The iron-free U2CO3 + U2O electrolyte remained stable during the cathodic scan up to -1.5V vs Ag / AgCI, below which electrolyte decomposition occurred as indicated by the rapid increase in current density (peak Ci). On the subsequent anodic scan, the current started to rise at around -1.4V, corresponding to the oxidation of deposited products on the graphite electrode.

[0171] The addition of 18 wt.% Fe2O3 to the U2CO3 + Li2O electrolyte increased the open circuit potential to a more positive value and produced a new reduction peak, C2, ranging from -1.08V to -1.22V. The higher OCP may indicate the formation of a new redox couple and different reaction equilibrium due to dissolution of Fe2O3 in the electrolyte. The new reduction peak indicates that Fe2O3 reduction occurred before decomposition of electrolyte and suggests a one-step reduction process of dissolved Fe2O3 according to Equation (5). By contrast, undissolved or solid Fe2O3 has previously been reported to exhibit two reduction peaks indicating a two-step reduction process according to Equations (6) and (7).

[0184] Fe3+(dissolved) +3e-^ Fe(s) equation (5)

[0185] Fe3+(s) + e_^Fe2+(s) equation (6)

[0186] Fe2+(s) + 2e ^Fe (s) equation (7)

[0187]

[0172] The addition of both Fe2O3 and Li2O into U2CO3 decreased the current density during both negative and positive scans compared to when the electrolyte was charged with only Fe2C>3, indicating an increased resistivity of the electrolyte. A single reduction peak, C3, is still observed during cathodic scanning but at more negative range (decreased from -1.14 V to -1 .2 V). The electrolyte with U2CO3 + 21 wt.% LiFeC>2 provided very similar results to that with □2 03 + 8 wt.% U2O + 18 wt.% Fe2C>3, suggesting that LiFeC>2 forms in situ when dissolving Fe2C>3 in the lithium carbonate-based electrolyte.

[0188]

[0173] Table 1 shows the open circuit potential and reduction onset potential of the electrolytes in the CV experiments. Very similar onset reduction potential (C3), current density, and oxidation peak were observed for all mixture which included Fe2O3. The results are consistent with the hypothesis that LiFeC>2 is the primary electrochemically active species.

[0189] Table 1. Example 2. XRD investigations of electrolyte composition

[0190]

[0174] The in situ formation of LiFeC>2 in the electrolytes formulated with Fe2Os (U2CO3 + 18 wt.% Fe2C>3 and U2CO3 + 8 wt.% U2O + 18 wt.% Fe2Os) was confirmed by X-Ray diffraction results of the electrolytes after different times of fusing (5 hours or 12 hours). LiFeC>2 was identified via the emergence of new peaks at 51 and 75 20 degrees. The presence of U2O did not yield new compounds but the LiFeC>2 peaks in the XRD spectra were enhanced. This suggests that U2O enhances the formation of LiFeC>2 via equations (8) and (9), where equation (5) has a lower theoretical enthalpy value (Fact Sage 8.3).

[0191] U2CO3 +Fe2C>3 — > 2LiFeO2+CC>2 AH—38.19 kJ / mol at 750°C equation (8) l_i2O + Fe2C>3 ^ 2LiFeC>2 AH=-102.18 kJ / mol at 750°C equation (9)

[0192]

[0175] In the absence of U2O, stronger peaks for LiFeC>2 were evident after 12 hours vs 5 hours, suggesting slow reaction kinetics from XRD analysis of fused electrolyte under different heat treatment periods. When U2O was present, there was no significant difference in the 5 hour and 12 hour fused electrolytes, indicating that U2O accelerates the rate of LiFeC>2 formation.

[0193] Example 3. Chronoamperometry investigations of electrolyte composition

[0194]

[0176] The impact of electrolyte composition on molten salt electrolysis of dissolved iron was further investigated by chronoamperometry (CA) with a graphite cathode (as working electrode), SnC>2 anode (as counter electrode), reference electrode, operating temperature of 750°C and a gas atmosphere of N2. A constant potential of -1.4 V vs Ag / AgCI, selected to provide sufficient overpotential for reaction as determined from the CV experiments in Examples 1 , was applied for 4 hours and the current response was measured.

[0195]

[0177] The following electrolytes were investigated: (i) Li2COs + 18 wt.% Fe2Os, (ii) U2CO3 + 8 wt.% U2O + 18 wt.% Fe2C>3 and (ii) U2CO3 + 21 wt.% LiFeC>2, and the results are shown in Figure 4.

[0196]

[0178] All three experiments resulted in a stable plateau current response, indicating that the applied potential and concentration of iron oxide species is sufficient to sustain prolonged electrolysis. Mass transfer did not appear limiting on the reduction process as a stable current was rapidly reached (within 2 minutes) at the beginning of the experiments and stayed substantially rather constant during 4 hours of operation.

[0197]

[0179] The electrolyte with only Fe2Os (U2CO3 + 18 wt.% Fe2Os) provided a higher current density (300mA. cm2) compared to the electrolytes containing U2O (U2CO3 + 8 wt.% U2O + 18 wt.% Fe2C>3 or U2CO3 + 21 wt.% LiFeO2). Added U2O (equivalent to starting with LiFeO2) may increase the resistivity of the system either directly or via increased LiFeC>2 concentration) or modify the melt response of the U2CO3 in a binary component mixture.

[0198] Example 4. Cyclic voltammetry and chronoamperometry investigations of cathode composition

[0199]

[0180] The impact of cathode composition on molten salt electrolysis of Fe2Os was investigated by cyclic voltammetry (CV) at scan rate of 5 mV / s with an electrolyte containing U2CO3 + 7 wt.% U2O + 18 wt.% Fe2C>3, SnC>2 anode (counter electrode), reference electrode, operating temperature of 750°C and a gas atmosphere of N2. The following cathodes (as working electrodes) were compared: metallic iron, metallic nickel and graphite. The results are shown in Figure 5 (portion of cathodic scan showing onset reduction peaks), where the cathodic scan corresponds to the cathodic half-reaction of equation (3a):

[0200] 2LiFeC>2 + 6e_— > 2Fe + U2O + 3O2' equation (3a)

[0201]

[0181] The graphite cathode produced the smallest current density, attributed to its lower electrical conductivity, but advantageously does not show any additional reduction peaks corresponding to surface reactions of the cathode itself.

[0202]

[0182] CV with the nickel cathode showed two cathodic peaks at -1 V (C4) and -1.22 V (C5), likely corresponding to reduction of nickel oxide to nickel and iron oxide to iron respectively. It is expected that the nickel electrode surface was oxidised when placed in contact with the corrosive molten carbonate. The thermodynamic reduction potential difference between nickel oxide and iron oxide suggests about 0.2 V lower reduction potential for nickel compared to iron at standard temperature and pressure (E° (Ni2++2e_^Ni) = -0.22 V, E° (Fe2++2e_^Fe) = -0.44 V), which matches the difference in potential of the observed reduction peaks in the CV with nickel cathode.

[0203]

[0183] The iron cathode provided a higher current density than the graphite electrode from OCP up to emergence of C5 peak, indicating that iron oxides formed after immersion of the cathode in the electrolyte are reduced under polarisation conditions. The surface iron oxide species may be chemically different from iron dissolved in the electrolyte, accounting for the lower reduction potential of these species compared to reduction of the dissolved iron oxide.

[0204]

[0184] There was some variation in the OCP with the three electrodes, with Fe the least negative and graphite the most negative. This variation may be due to formation of different oxide layers on the electrode surface after immersion in the electrolyte prior to CV, with the OCP affected by the equilibrium between cathode species (e.g. Fe / FeO vs C / CO2). However, all three electrodes exhibit a reduction peak at about -1.2 V corresponding to reduction of LiFeO2 species (C5).

[0185] The impact of cathode composition was further investigated by chronoamperometry (CA) at a constant potential of -1.4 V vs Ag / AgCI with an electrolyte containing U2CO3 + 8 wt.% U2O + 18 wt.% Fe2C>3, SnC>2 anode, operating temperature of 750°C and a gas atmosphere of N2, and the current response is seen in Figure 6.

[0205]

[0186] The iron cathode provided the highest current density, suggesting that it may be a preferred cathode material for commercial processes both due to electrochemical performance and compatibility with the electrodeposited iron product. However, fundamental investigations may advantageously be performed with a graphite cathode to better distinguish between species on the cathode formed from the electrolyte vs the cathode itself. The nickel cathode failed in less than two hours due to disconnection from its current collector, and the deposited iron was difficult to separate due to alloying.

[0206] Example 5. Cyclic voltammetry and chronoamperometry investigations of anode composition

[0207]

[0187] The impact of anode composition on molten salt electrolysis of Fe2Os was investigated by cyclic voltammetry (CV) at scan rate of 5 mV / s with an electrolyte containing □2 03 + 18 wt.% Fe2C>3, graphite cathode (as working electrode), operating temperature of 750°C and a gas atmosphere of N2. The effect of the following anodes (as counter electrodes) were initially compared: metallic nickel and tin oxide (SnC>2). The anode reaction is the oxygen evolution reaction according to equation (3b):

[0208] 3O2' — > 3 / 202 + 6e_equation (3b)

[0209]

[0188] The results are shown in Figure 7 (portion of cathodic scan showing onset reduction peaks). The choice of anode material affected the cathodic onset, see peaks Ce and C7 for SnC>2 and Ni respectively (1.09V and 1.18V), suggesting that the graphite cathode is more electrochemically active in presence of SnC>2 anode than Ni anode.

[0210]

[0189] The impact of anode composition was further investigated by chronoamperometry (CA) at a constant potential of -1.4 V vs Ag / AgCI with an electrolyte containing U2CO3 + 18 wt.% Fe2C>3, graphite cathode, operating temperature of 750°C and a gas atmosphere of N2, and the current response is seen in Figure 8. The nickel anode initially provided lower current density, but the current density increased with time and surpassed SnC>2 after 10 minutes. This difference might be attributed to different conductivity of deposited products in presence of different anode materials. Nickel may co-deposited on the cathode together with iron as a result of nickel corrosion from the anode. Example 6. Cyclic voltammetry investigations of gas atmosphere

[0211]

[0190] The impact of gas atmosphere - either N2 or CO2 (at atmospheric pressure) - on molten salt electrolysis of Fe2Os was investigated by cyclic voltammetry (CV) at scan rate of 5 mV / s with an electrolyte containing U2CO3 + 10 wt.% Fe2Os, metallic iron cathode (as working electrode), SnC>2 anode (as counter electrode), reference electrode, and operating temperature of 750°C. The results are shown in Figure 9 (cathodic scan).

[0212]

[0191] It was expected that the CO2 atmosphere may inhibit the decomposition of the U2CO3, which can decompose according to equation 10:

[0213] U2CO3 — > U2O + CO2 equation (10)

[0214]

[0192] However, higher current density was observed with N2 gas atmosphere.

[0215] Example 1. Cyclic voltammetry investigations of electrolyte pre-treatment

[0216]

[0193] The impact of electrolyte heat treatment on molten salt electrolysis of Fe2Os was investigated by cyclic voltammetry (CV) at scan rate of 5 mV / s with an electrolyte containing U2CO3 + 10 wt.% Fe2C>3, metallic iron cathode (working electrode), SnC>2 anode (counter electrode), reference electrode, operating temperature of 750°C and N2 gas atmosphere. In one experiment, the electrolyte components were pre-fused at 800°C for 20 hours, before cooling to room temperature and then re-heating to the required reaction temperature. In another experiment, the electrolyte components were not pre-fused but simply combined and heated to the required reaction temperature and held there for 7 hours before CV (“In situ fused”). The results are shown in Figure 10 (cathodic scan).

[0217]

[0194] Different electrochemical responses were evident in the CV scans. The in-situ fused electrolyte shows only one cathodic peak delayed until -1.6 V, whereas pre-fused electrolyte showed two cathodic peaks at -1.2 V and -1.4V. This different behaviour might be due to incomplete dissolution of iron oxide and formation of lower concentration of LiFeC>2 for in-situ fused electrolyte than pre-fused electrolyte.

[0218] Example 8. XRD characterisation of iron deposits

[0219]

[0195] The electrodeposited product was collected from the graphite cathode after the CA experiment in Example 3 with U2CO3 + 8 wt.% U2O + 18 wt.% Fe2Os electrolyte, washed with water and magnetically collected. The XRD pattern of the deposit is shown in Figure 11 (similar results were obtained with different electrolysis potential, reaction time and electrolyte composition).

[0220]

[0196] The peaks with high intensity at 53, 77, and 99 (20 degree) signify a high crystallinity of metallic iron formed after electrolysis. The peak at 38 (20 degree) corresponds to U2O which may be trapped within iron cavities and thus not washed away. Grinding of the deposit prior to washing may allow more complete U2O removal.

[0221] Example 9. SEM-EDS characterisation of deposits from iron-free electrolyte

[0222]

[0197] Chronoamperometry (CA) was conducted with a graphite cathode (working electrode), Ni anode (counter electrode), reference electrode, iron-free electrolyte (U2CO3 + 8 wt.% U2O), operating temperature of 750°C, a gas atmosphere of N2, and constant potential of -1.4 V vs Ag / AgCI for 4 hours. After CA, the deposit was recovered from the cathode and washed with water. Scanning electron microscopy (SEM) imaging showed that the deposit predominantly consists of nano-size spherical carbon particles mixed with graphitic sheet. Energy dispersive X-ray spectroscopy (EDS) confirmed the presence of nickel in the deposit, demonstrating the risk that the anode material can contaminate cathodic deposits when conducting electrolysis in carbonate electrolytes at high temperature.

[0223] Example 10. SEM characterisation of deposits from iron-containing electrolyte

[0224]

[0198] Chronoamperometry (CA) was conducted with a graphite cathode (as working electrode), SnC>2 anode (as counter electrode), iron-containing electrolytes (either U2CO3 + 18 wt.% Fe2C>3 or U2CO3 + 18 wt.% Fe2Os + 8 wt.% U2O), operating temperature of 750°C, a gas atmosphere of N2, and constant potential of -1.4 V vs Ag / AgCI for 4 hours. After CA, the deposit was recovered from the cathode and washed with water.

[0225]

[0199] In the absence of added U2O, SEM imaging showed that the deposit included metallic iron particles with sizes ranging from 300 to 600pm in two different morphologies: big irregular crystal grains and a finer dendritic structure. Some unreduced iron oxide was also apparent, suggesting mass transfer limitations or incomplete dissolution of Fe2Os.

[0226]

[0200] The presence of added U2O significantly changed the morphology and composition of the deposits, with the iron present in longer iron particles with an average size of 600 pm and very smooth surface, and no evidence of residual electrolyte or iron oxide. The deposit was highly crystalline, appeared metallic and was strongly susceptible to magnetic attraction. The addition of U2O appears to enhance the quality of deposited iron on the graphite electrode.

[0227] Example 11. Linear Sweep voltammetry and chronoamperometry investigations of ironbased anode

[0228]

[0201] The role of an iron-based anode (as working electrode in a three-electrode set-up) in molten salt electrolysis of Fe2Os was investigated by linear sweep voltammetry (LSV) with Li2CC>3-based electrolyte (U2CO3 + 8 wt.% U2O + 18 wt.% Fe2Os), graphite cathode (as counter electrode), reference electrode, operating temperature of 750°C and a gas atmosphere of N2. The anode consisted of iron foil cut to a surface area of 1 ,4cm2(total surface are of 2.8cm2over both faces) and the graphite cathode had a surface area of 2.16cm2. The starting potential was the open circuit potential (OCP) and the scan rate was 5 mV / s to a maximum oxidative potential of 1.5 V vs Ag / AgCl. The results are shown in Figure 12.

[0229]

[0202] The first oxidation peak appears at -0.1 V, consistent with the formation of a metal oxide layer. A significant increase in current density, starting from 0.3 V and rising from 0.1 A / cm2to 1.2 A / cm2at 1.5V, highlights the high electrochemical activity of the iron anode under positive overpotential. Based on these observations, two potentials within the oxygen evolution region (0.5 V and 1 V) were selected to determine whether minimal polarization is sufficient for iron deposition or if a higher overpotential is required.

[0230]

[0203] Chronoamperometry (CA) was thus performed at a constant anode potential of either 0.5 V or 1 V vs Ag / AgCl with an electrolyte containing U2CO3 + 8 wt.% U2O + 18 wt.% Fe2C>3, graphite cathode, iron anode, reference electrode, operating temperature of 750°C and a gas atmosphere of N2 and the current density response over the 4 hour reaction is seen in Figure 13.

[0231]

[0204] Both experiments reached a similar value of 0.2 A / cm2after four hours of operation, regardless of the applied potential. However, the higher overpotential resulted in a faster current decline, suggesting that a lower overpotential is preferable for greater stability. Within 100 minutes, the current response dropped to about half its initial value, suggesting that the initial oxide layer may not be stable. However, the current density then stabilised, suggesting the formation of a new and more corrosion-resistant oxide layer as the system approaches equilibrium.

[0232] Example 12. Constant current electrolysis investigations of iron-based anode

[0233]

[0205] Electrolysis of Fe2Os in lithium carbonate-based electrolyte was investigated under constant current conditions with a two-electrode system more representative of industrial scale electrolysers. Chronopotentiometry (CP) was thus performed at a constant current density of 100mA. cm-2with an electrolyte containing U2CO3 + 6.7 wt.% U2O + 18 wt.% Fe2Os, graphite cathode, iron anode, operating temperature of 750°C and a gas atmosphere of N2. The potential difference response over 6 hour and 48 hour reactions is seen in Figure 14.

[0234]

[0206] The results generally showed a stable response during both short- and long-term operation. Some instability was observed during the early part of the 48-hour reaction (between 3 and 18 hours), possibly due to the formation of an inactive, resistive oxide layer on the electrode surface. Despite this, the system eventually returned to a stable operating state which continued for the remainder of the reaction, between 18 and 48 hours. Example 13. Characterisation of cathodic iron deposits produced with Ni, SnO2and Fe anodes

[0235]

[0207] The impact of anode composition on the cathodic deposit morphology and composition was investigated by chronoamperometry (CA) for four hours at a constant potential of -1.4 V vs Ag / AgCI with an electrolyte containing U2CO3 + 18 wt.% Fe2Os, graphite cathode, operating temperature of 750°C and a gas atmosphere of N2. Ni, SnC>2 and Fe anodes (as counter electrodes) were investigated.

[0236]

[0208] After CA, the deposit was recovered from the cathode, washed with water, and imaged by SEM. Selected images are shown in Figure 15 (Ni anode), Figure 16 (SnC>2 anode) and Figure 17 (Fe anode).

[0237]

[0209] Using a Ni counter electrode led to the formation of more dendritic structures composed of thin metallic iron rods and small particles (Figure 15), in contrast to the large, flat metallic iron sheets obtained when using a SnC>2 counter electrode (Figure 16) and the mix of thick iron rods and larger particles obtained when using an Fe counter electrode (Figure 17). Since all experimental conditions are similar except anode (counter electrode) material, these variations may be attributed to the effect of anode corrosion into the electrolyte. Most notably, corrosion of nickel into the electrolyte appears to materially affect the morphology of iron deposits on the cathode, correlating with the instability in current response (c.f. Figure 8).

[0238]

[0210] Energy dispersive X-ray spectroscopy (EDS) mapping was used to investigate the elemental composition of the imaged deposits. Ni and Sn contaminants were detected by EDS in the iron deposits when using the Ni and SnC>2 anodes, respectively. By contrast, pristine iron deposits were obtained when using an Fe anode, due to the absence of any transition metal contaminants introduced by anode oxidation.

[0239] Example 14. Characterisation of oxidic layers on the iron-based anode

[0240]

[0211] Surface morphology and elemental analysis using SEM and EDS on the outer oxide layer present on the iron-based anode surface after CA electrolysis at different potentials (0.5 and 1 V; Example 11) revealed the formation of various iron oxide species. The particle morphology of the outer layer was similar under both high and low overpotentials (0.5V and 1V), but iron-to-oxygen molar ratios (from EDS) varied from point to point in each analysis.

[0241]

[0212] The iron oxide was present in particles which appeared to be encapsulated within a solid matrix primarily composed of oxygen and aluminium. XRD analysis (Figure 23) indicated that the aluminium-bearing compound in the outer oxidic layer is predominantly lithium aluminate (UAIO2). Since no aluminium compounds were intentionally added to the electrolyte, the aluminium in the UAIO2 must be derived from the alumina-based electrolyte crucible. Alumina is slightly soluble in U2CO3 electrolyte at elevated temperatures, forming UAIO2 in solution.

[0242]

[0213] The XRD results (Figure 18) further showed that mixtures of iron-oxygen-lithium compounds were present in the outer oxide layer after electrolysis, including LiFesOs, LiFeC>2, and FesC . By contrast, XRD spectra of the pre-electrolysis iron electrode (held above the electrolyte, without any direct contact) confirmed the initial oxidic layer consists of Fe2Os and FesC , possibly due to low background levels of O2 in the gas atmosphere.

[0243]

[0214] Accessing the inner oxide layers is more challenging, as post-treatments can disrupt the layers and thus confound the analysis. To mitigate this risk, the electrode was crosssectioned and cemented in resin for SEM I EDS analysis of the inner layers.

[0244]

[0215] SEM-EDS mapping was conducted on the cross-section to examine all of the oxidic layers at the electrode-electrolyte interface for the 1 V CA experiment of Example 11. The SEM images in Figures 19, 20 and 21 show the metallic anode composition (100), first layer of oxide on the anode (102) and second layer (104). Figure 22 shows the elemental composition of the investigated areas of Figure 19 (1 - in the first oxide layer 102), Figure 20 (2 - in the first oxide layer 102) and Figure 21 (3 - in the second layer 104), as determined by EDS.

[0245]

[0216] The first oxide layer (102), directly adjacent to the bulk iron metal, is dense and primarily composed of iron and oxygen, with a consistent composition across different positions on the anode surface. The composition is believed to be FesC given the Fe / O mole ratio was measured consistently as 0.76. The next layer (104) is a mixture of electrolyte and iron oxide, providing evidence of FesC dissolution into the electrolyte and suggesting an ongoing interaction between oxide and electrolyte. This is the same layer that was directly analysed on the electrode surface by XRD, as discussed above, and shown to contains LiFesOs, LiFeO2, and FesO4.

[0246] Example 15. Further characterisation of oxide formation and iron corrosion from an ironbased anode

[0247]

[0217] To further investigate the oxide layer formation and iron corrosion from an ironbased anode during the course of extended electrolysis, chronopotentiometry (CP) was performed at a constant current density of 100mA. cm-2with an electrolyte containing Li2COs + 6.7 wt.% U2O + 18 wt.% Fe2C>3, graphite cathode, iron anode, operating temperature of 750°C and a gas atmosphere of N2. CP reactions of 2 hours and 48 hours were conducted. The anode was recovered, cross-sectioned and cemented in resin.

[0218] EDS line scans were conducted to examine the oxide layer composition at varying distances from the anode surface toward the electrolyte, and the results are shown in Figure 23 (2 hours) and Figures 24 and 25 (48 hours).

[0248]

[0219] The thickness of the combined oxide layers increased significantly with longer electrolysis times, from approximately 110 pm after 2 hrs of oxidation to 180 pm after 48 hrs.

[0249]

[0220] After only 2 hrs of oxidation, aluminium and oxygen were observed in the oxide layer closest to the electrolyte. The line scan analysis revealed a first oxide layer close to the electrode with an Fe / O mole ratio ranging from 0.6 to 0.4 (zone 1 , Figure 23). The Fe / O ratio then significantly decreased in the layer closest to the electrolyte, reaching near zero (zone 2, Figure 23), consistent with a UAIO2 layer which limits iron corrosion and dissolution into the electrolyte.

[0250]

[0221] After 48 hrs, a discrete intermediate aluminium / oxygen layer 108 had formed between (i) an inner iron oxide layer 102 (probably FesC ) adjacent to the iron anode body 100 and (ii) an outer iron oxide layer 104 (likely containing LiFesOs, LiFeC>2, and FesC ). It is proposed that this intermediate layer consists of UAIO2, based on the XRD analysis in Example 11 . The formation of an iron oxide-based layer outside the UAIO2 layer (which was initially formed on the anode outer surface, as seen after 2 hours electrolysis) shows that the UAIO2 layer is sufficiently porous to provide controlled ionic conductivity during electrolysis.

[0251]

[0222] The stable potential response observed during electrolysis (c.f. Figure 14) suggests that changes in the oxidic layer structure did not significantly affect the conductivity of the electrode. This is also consistent with the formation of an UAIO2 layer, since UAIO2 is known to be ionically and electrically conductive. In contrast, if AI2O3 formed instead of UAIO2 then conductivity would be disrupted due to its insulating properties.

[0252]

[0223] Without wishing to be bound by any theory, it is proposed that UAIO2 formation on the iron-based anode may act as a protective barrier, protecting the bulk iron from rapid oxidation.

[0253] Example 16. Iron-based anode corrosion rate

[0254]

[0224] The corrosion rate of the iron anode under different electrolysis conditions (current densities, or potentials) in the previous Examples was investigated by measuring the thickness of the iron anode after different reaction times. The measurements were derived from SEM images, and the results are shown in Table 2 below. Table 2

[0255]

[0225] Under a constant current density of 100 mA / cm2, the anode thickness decreased from 996 pm to 853 pm within two hours, and after 48 hours, it further reduced to 560 pm. As seen in Table 2, a high average corrosion rate of about 73 pm / hr was observed in the initial phase of electrolysis, but the consumption rate then moderated over time. Similar corrosion rates were obtained in CA experiments at different applied potentials (for same total electrolysis time), suggesting that the corrosion rate of iron anode is independent of polarization conditions for a given electrolyte composition and temperature.

[0256]

[0226] Based on the corrosion rate between 2 and 48 hours, it can be estimated that the iron corrosion rate from the anode surface was 31.13 mm / year, equating to 9 mg / hr of iron consumption. By contrast, the theoretical consumption of iron to produce iron oxide (Fe2Os) would be 236 mg / h if all of the current during CP caused iron oxidation with 100% efficiency. The faradaic efficiency of the iron anode toward oxygen evolution was thus 96%, with only 4% contribution to iron oxide formation.

[0257]

[0227] Following electrolysis, the iron deposits on the cathode were collected and weighed to determine the ratio of iron produced versus the amount corroded from the anode. The iron production at the cathode corresponded to 175 mg / hr, implying an overall iron production efficiency of about 70%. Only a minor efficiency loss of 4% is attributed to iron consumption at the anode. These results demonstrate that, while the iron anode undergoes some consumption during electrolysis, it primarily serves as an efficient oxygen evolution electrode, providing sufficient electrons to sustain the reduction process at the cathode.

[0258] Example 17. Investigations of the role of aluminium

[0259]

[0228] To further investigate the role of aluminium species on oxide layer formation and iron corrosion from an iron-based anode, chronopotentiometry (CP) reactions of 4 hours duration were performed in a two-electrode configuration at a constant current density of 100mA. cm-2with Li2CO3-based electrolytes containing different aluminium sources, a graphite cathode, iron anode, operating temperature of 750°C and a gas atmosphere of N2. After the reactions, the anode was recovered, cross-sectioned and cemented in resin, and the rates of iron corrosion were determined as described in Example 16. The results are shown in Table 3.

[0260]

[0229] Iron ore used in Experiments B and C was fines from a hematite-goethite ore (Pilbara region, Western Australia), composition: 60.5 wt% Fe, 2.3 wt% Si, 1.3% Al. The aluminium is predominantly present in kaolinite (Al2Si2Os(OH)4) and no free alumina (AI2O3) is present.

[0261] Table 3

[0262]

[0230] In Experiment A, a graphite crucible was used as the electrolyte holder and the electrolyte was U2CO3 containing Fe2Os, thereby eliminating all sources of aluminium. The anode corrosion rate was measured as 75 pm / h, the highest iron consumption rate observed. In Experiment B, a graphite crucible was again used but the source of iron in the electrolyte was iron ore, thereby introducing aluminium in the form of gangue aluminosilicates, primarily kaolinite. The corrosion rate decreased slightly to 62 pm / h under otherwise identical conditions. In Experiment C, an alumina crucible was used, thereby introducing aluminium to the electrolyte both as alumina (from the crucible) and aluminosilicate (from the ore). The corrosion rate was substantially reduced, to only 18 pm / h. As comparison, Table 3 also shows as Experiment D the 36 pm / h corrosion rate (described in Example 16) over 6 hours electrolysis when using an alumina crucible and an electrolyte containing Fe2Os, so that the only source of aluminium was alumina (from the crucible).

[0263]

[0231] SEM / EDS line-scanning characterization was conducted on the cross-sectioned post-reaction anodes, as described in Example 15, to observe structural and compositional changes within the anode surface layer. The results confirmed the presence of aluminium in a discrete, coherent layer (but not in the other oxidic layers) at the surface of the anode from Experiments C and D (layer 108 per Figure 25, after 48 hours), but no such aluminium-bearing layer was evident in the anodes from Experiments A and B. For Experiment B, aluminium was detected only in scattered regions on the anode surface at very low concentration (approximately 0.6 wt%), likely due to residual solidified electrolyte contamination rather than interfacial layer formation.

[0264]

[0232] The results correlate with the observed corrosion rate and demonstrate that the introduction of aluminium as alumina (AI2O3) to the electrolyte provides a very significant protective effect to the iron-based anode, whereas only a much smaller effect is evident when aluminosilicate (gangue species in iron ore) was introduced. This effect corresponds to the formation of a discrete aluminium-bearing layer, believed to contain UAIO2.

[0265]

[0233] Without wishing to be limited by any theory, this behaviour may be attributed to differences in electronegativity and bond character among the electrolyte species. The iron anode, being positively charged during electrolysis, preferentially attracts species with stronger electronegative interactions. Lithium aluminate (UAIO2; containing Li-0 and AI-0 ionic bonds) produced by dissolution of alumina (AI2O3) exhibits more ionic character and therefore stronger electrostatic attraction. In contrast, lithium aluminosilicate (LiAISiC ) produced by dissolution of aluminosilicates contains more covalent bonding (Li— O, Al-O, Si-O). The more electronegative aluminium species in electrolytes containing alumina thus adsorb and form a coherent surface layer, while in aluminosilicate-containing electrolytes this attraction is weaker, preventing formation of a distinct protective layer.

[0266] Example 18. Mixed carbonate electrolytes

[0267]

[0234] Iron oxide reduction was also investigated in a eutectic ternary carbonate mixture comprising 32.0 wt% U2CO3, 33.5 wt% Na2CO3, and 34.5 wt% K2CO3. Chronoamperometry (CA) was conducted with a graphite cathode (working electrode), SnC>2 (counter electrode), reference electrode, electrolyte containing Fe2Os (U2CO3: Na2COs: K2CO3 + 10 wt.% Fe2Os), operating temperature of 750°C, a gas atmosphere of N2, and constant potential of -1.4 V vs Ag / AgCI for 4 hours. For comparison, the same experiment was conducted but with lithium carbonate electrolyte (U2CO3 + 10 wt.% Fe2Os). After CA, the deposits were recovered from the cathodes, washed with water, and analysed by SEM and EDS.

[0268]

[0235] High-purity iron was successfully produced on the cathode in both experiments. The purity of the iron deposits was unaffected but the particle size and morphology changed significantly. The deposits formed in the ternary carbonate were noticeably smaller and exhibited a needle-like shape, consistent with a reduced dissolution capacity of iron oxide species in the ternary carbonate compared with U2CO3. The effect of reduced iron concentration in the ternary electrolyte could be mitigated by adding U2O to the melt or by operating at higher temperatures.

[0269]

[0236] Those skilled in the art will appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. It is understood that the invention includes all such variations and modifications which fall within the spirit and scope of the present invention.

Claims

Claims1. A method of producing iron, the method comprising: contacting a molten salt electrolyte comprising lithium, carbonate and dissolved iron oxide with a cathode and an anode in an electrolytic cell; and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode, wherein the anode comprises a ferrous metal composition, and wherein iron corrodes from the ferrous metal composition of the anode at a rate less than 20% of the rate of forming solid metallic iron at the cathode.

2. The method according to claim 1 , wherein the temperature of the molten salt electrolyte is greater than 700°C when flowing the current through the molten salt electrolyte.

3. The method according to claim 1 or claim 2, wherein the ferrous metal composition comprises at least 70 wt.% iron.

4. The method according to any one of claims 1 to 3, wherein the ferrous metal composition is substantially pure iron.

5. The method according to any one of claims 1 to 4, wherein (i) the molten salt electrolyte further comprises a dissolved aluminium species during at least part of the time that current is passed through the molten salt electrolyte and / or (ii) the anode is contacted with a molten salt composition comprising lithium, carbonate and a dissolved aluminium species before contacting the molten salt electrolyte with the anode.

6. The method according to claim 5, wherein the dissolved aluminium species comprises LiAIO2.

7. The method according to any one of claims 1 to 6, wherein the anode comprises one or more solid oxidic layers at an interface between the ferrous metal composition and the molten salt electrolyte, the one or more solid oxidic layers comprising an aluminium- bearing oxidic layer.

8. The method according to claim 7, wherein the aluminium-bearing oxidic layer comprises LiAIO2.

9. The method according to claim 7 or claim 8, wherein the one or more solid oxidic layers further comprise an iron oxide-rich layer adjacent the ferrous metal composition of the anode.

10. The method according to any one of claims 7 to 9, wherein the one or more solid oxidic layers comprise (i) a first iron oxide-rich layer adjacent the ferrous metal composition of the anode, and (ii) a second iron oxide-rich layer proximate to the molten salt electrolyte, wherein the aluminium-bearing oxidic layer separates the first and second iron oxide-rich layers.

11. The method according to any one of claims 1 to 10, wherein iron corrodes from the ferrous metal composition of the anode at a rate less than 5% of the rate of forming solid metallic iron at the cathode.

12. The method according to any one of claims 1 to 11, wherein lithium is present in the molten salt electrolyte in an amount of at least 90% of total alkali and alkali earth metals.

13. The method according to any one of claims 1 to 12, comprising adding lithium oxide to the molten salt electrolyte.

14. The method according to any one of claims 1 to 13, wherein at least a portion of the lithium and the dissolved iron oxide is present in the electrolyte as LiFeO2.

15. The method according to any one of claims 1 to 14, wherein current is passed through the molten salt electrolyte at a current density of between 100 mA. cm-2and 1000 mA. cm-2normalised to the working surface area of the cathode.

16. A system for producing iron, the system comprising: an electrochemical cell comprising a molten salt electrolyte comprising lithium, carbonate and dissolved iron oxide in contact with a cathode and an anode; and a power supply to apply a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode, wherein the anode comprises a ferrous metal composition.

17. The system according to claim 16, wherein the ferrous metal composition is substantially pure iron.

18. The system according to claim 16 or claim 17, wherein the molten salt electrolyte further comprises a dissolved aluminium species.

19. The system according to claim 18, wherein the dissolved aluminium species is UAIO2.

20. The system according to any one of claims 16 to 19, wherein the anode comprises one or more solid oxidic layers at an interface between the ferrous metal composition and the molten salt electrolyte, the one or more solid oxidic layers comprising an aluminium- bearing oxidic layer.

21. The system according to claim 20, wherein the aluminium-bearing oxidic layer comprises LiAIO2.

22. The system according to claim 20 or claim 21, wherein the one or more solid oxidic layers further comprise an iron oxide-rich layer adjacent the ferrous metal composition of the anode.

23. The system according to any one of claims 20 to 22, wherein the one or more solid oxidic layers comprise (i) a first iron oxide-rich layer adjacent the ferrous metal composition of the anode, and (ii) a second iron oxide-rich layer proximate to the molten salt electrolyte, wherein the aluminium-bearing oxidic layer separates the first and second iron oxide-rich layers.

24. A method of producing iron, the method comprising: contacting a molten salt electrolyte comprising lithium, carbonate, dissolved iron oxide and a dissolved aluminium species with a cathode and an anode in an electrolytic cell, wherein the anode comprises a ferrous metal composition; and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby (i) forming one or more solid oxidic layers on the ferrous metal composition at an interface between the anode and the molten salt electrolyte, the one or more solid oxidic layers comprising an aluminium-bearing oxidic layer, and (ii) reducing iron at the cathode to form solid metallic iron and producing dioxygen gas at the anode,wherein iron corrodes from the ferrous metal composition of the anode at a rate less than 10% of the rate of forming solid metallic iron at the cathode after forming the one or more solid oxidic layers.

25. A method of conditioning an anode for oxygen evolution, the method comprising: contacting a molten salt electrolyte comprising lithium, carbonate, optionally dissolved iron oxide, and a dissolved aluminium species with a cathode and an anode in an electrolytic cell, wherein the anode comprises a ferrous metal composition; and applying a potential difference between the cathode and the anode sufficient to pass a current through the molten salt electrolyte, thereby providing one or more solid oxidic layers on the ferrous metal composition at an interface between the anode and the molten salt electrolyte, wherein the one or more solid oxidic layers comprise an aluminium- bearing oxidic layer.

26. The method according to claim 25, wherein the aluminium-bearing oxidic layer comprises LiAIO2.

27. The method according to claim 25 or claim 26, wherein the one or more solid oxidic layers further comprises an iron oxide-rich layer adjacent the ferrous metal composition of the oxygen-evolving anode.

28. The method according to any one of claims 25 to 27, wherein the ferrous metal composition is substantially pure iron.

29. An anode for oxygen evolution, the anode comprising: a ferrous metal composition; and one or more solid oxidic layers at an electrolyte-contacting surface of the ferrous metal composition, the one or more solid oxidic layers comprising an aluminium-bearing oxidic layer.

30. The anode according to claim 29, wherein the aluminium-bearing oxidic layer comprises LiAIO2.

Images