The invention relates to a method and an apparatus for removing oxygen from a compound, comprising a metal and oxygen, or from a metal or alloy, by an electrolytic process.

The electro-decomposition or electro-reduction of solid compounds is described in, for example, International Patent Application PCT/GB99/01781 and "Extraction of titanium from solid titanium dioxide in molten salts" by D. J. Fray and G. Z. Chen, TMS (The Minerals, Metals and Materials Society) 2004, pp. 9-17, which are incorporated herein by reference in their entirety. In this process, also known as the FFC process, a solid compound between a metal (which term is used in this document to include both metals and semi- metals) and another substance (for example, oxygen) is electrolytically reduced. A cathode comprising or contacting the solid compound is immersed in or contacted with a fused salt and a voltage is applied between the cathode and an anode, which also contacts the salt, such that the substance dissolves in the salt. This process may be used, for example, to extract the metal from the solid metal compound. A cathode comprising or contacting a mixture of solid metal compounds, or of one or more compounds and one or more metals, may also be processed to form an alloy or an intermetallic compound.

This process is termed electro-decomposition in this document but is also known by other terms including electro-deoxidation or electro-reduction.

Calcium chloride has been commonly used as an electrolyte in the electro- decomposition process as calcium is very electro-positive. The fused salt acts as an electrolyte in the electro-decomposition process. The term fused salt may be used interchangeably with other terms commonly used in the art, such as electrolyte, molten salt and melt.

It is possible to reduce most metal oxides by electro-decomposition using a CaCI ₂ based melt. A problem arises, however, in providing an anode material which is sufficiently inert in a CaCI ₂ -based melt.

At present, there is no suitable material that can be used as an inert anode in chloride melts such as calcium chloride and, therefore, a carbon anode is generally employed. The use of a carbon anode in an electro-decomposition process results in the evolution of CO and/or CO ₂ . Furthermore, the carbon anode is consumed and the product of the electro-decomposition may become contaminated with carbon.

Summary of Invention

The invention provides a method and an apparatus for removing oxygen from a compound comprising a metal and oxygen, or from a metal or alloy, and an anode for an electro-decomposition process, as defined in the appended independent claims to which reference should now be made. Preferred or advantageous features of the invention are defined in dependent sub-claims.

In a first aspect the invention may advantageously provide a method of removing oxygen from a compound comprising a metal and oxygen which comprises the steps of arranging a cathode comprising or contacting the compound in contact with a melt comprising a hydroxide of an alkali metal. An anode is also arranged in contact with the melt and a potential is applied between the anode and the cathode sufficient to remove oxygen from the solid compound. The compound may be an intermetallic compound. In a similar method the cathode comprises or contacts a metal or alloy containing dissolved oxygen, and the applied potential is sufficient to remove oxygen from the metal or alloy. The compound, metal or alloy is preferably solid.

Preferably the melt further comprises an oxide of the alkali metal and this oxide is particularly preferably soluble in the hydroxide. A melt comprising an alkali metal oxide, which is dissolved in an alkali metal hydroxide, is a melt system that may advantageously support oxygen ion conductivity. Good oxygen ion conductivity may be advantageous in efficiently transferring oxygen from the solid compound at the cathode through the melt to be discharged at the anode.

Preferably the potential at the cathode during electro-decomposition is lower (in magnitude) than a potential for continuous evolution of hydrogen or continuous deposition of the alkali metal from the melt. Preferably, the

potential (applied voltage) between the cathode and the anode is lower than a potential for continuous decomposition of the melt.

Although the alkali metal hydroxide may be any alkali metal hydroxide, s preferably it is sodium hydroxide and the alkali metal is sodium, or the hydroxide is potassium hydroxide and the alkali metal is potassium. The oxide is then preferably sodium oxide or potassium oxide respectively. Mixtures of hydroxides and/or oxides comprising different cations may be used.

io Although the solid compound may be any solid compound comprising a metal and oxygen that is less stable than an oxide of the alkali metal, the invention may be particularly advantageous when used to reduce oxides of low stability such as an iron oxide or an oxide of cobalt, nickel, copper, zinc or lead, or when used to remove oxygen from metals or alloys comprising such metals.

I ₅

If a precursor material at the cathode comprises a mixture of metal compounds, or a mixture of one or more metal compounds and one or more metals, then an alloy or an intermetallic compound comprising metal species in the precursor material may be produced.

20

The melt may comprise a mixture of hydroxides or oxides of more than one alkali metal, but may additionally contain other anion or cation species. Such species should preferably not, however, be such as to cause corrosion of an inert anode. ₅

The melting point of sodium hydroxide is about 32O ⁰ C. This may advantageously allow the electro-decomposition process to be carried out at low temperatures compared to electro-decomposition in a calcium chloride melt. For example the reaction may proceed at temperatures below 65O ⁰ C or 0 below 500 ⁰ C. The electro-decomposition may particularly advantageously proceed at any temperature at which the electrolyte is molten.

Advantageously, the melting point of sodium hydroxide may be decreased with the addition of a small amount of sodium iodide or sodium bromide. This may s allow electro-decomposition to proceed at even lower temperatures and so

may reduce the energy needed to maintain the melt at an operating temperature and may reduce problems of corrosion in the electro- decomposition apparatus.

The removal of oxygen from the compound(s), metal(s) or alloy(s) at the cathode may involve diffusion. This process is accelerated by increasing temperature and so the rate of reaction may disadvantageously be reduced by operating at low temperature, depending on the materials involved and the geometry of the materials used (such as the particle size of the material at the cathode). Thus, a temperature greater than 500 ⁰ C OR 550°c may advantageously be used. The boiling point of NaOH, for example, is 1390 ⁰ C and so higher temperatures than 500 ⁰ C may in principle be used as long as any resulting reaction rate increase is appropriately balanced against any increase in corrosion of the apparatus at higher temperatures.

Preferably the anode is substantially inert with respect to the melt under operating conditions. For example, if the electro-decomposition is carried out at 65O ⁰ C using a sodium hydroxide melt in which a small proportion of sodium oxide is dissolved, the anode should be substantially inert to this melt at this temperature. Anodes containing nickel or nickel oxide are believed to be substantially inert in caustic melts such as molten sodium hydroxide. Preferably the anode is made from a material that comprises nickel, for example nickel oxide or a nickel alloy or a nickel-rich alloy. An Inconel™ may be suitable. In general, a metal which forms an inert oxide layer may be suitable as an inert anode in a hydroxide melt.

References in this document to inert anodes should be construed, as the skilled person would do, to encompass substantially inert anodes. Thus, an inert anode should be sufficiently inert in practice to be usable for a suitably extended length of time.

Thus, a further aspect of the invention may advantageously provide a method of removing oxygen from a solid compound comprising a metal and oxygen which comprises the steps of arranging a cathode comprising or contacting the solid compound in contact with a melt comprising a hydroxide of an alkali

metal. An anode is also arranged in contact with the melt and a potential is applied between the anode and the cathode sufficient to remove oxygen from the solid compound, the anode comprising a metal or alloy which forms an inert oxide layer, or comprising nickel, nickel oxide or a nickel alloy.

An inert or substantially inert anode may provide a number of advantageous features when compared with carbon anodes as conventionally used in electro- decomposition processes. Gas evolved at the inert anode during electro- decomposition may be substantially pure oxygen. Carbon anodes generally evolve carbon dioxide or carbon monoxide, and these gases may have a deleterious effect on the environment if vented to the atmosphere in the volumes likely to be produced by a commercial industrial plant. Any oxygen produced may be vented to the atmosphere or may be collected as a product of the electro-decomposition process.

As inert anodes do not react during electro-decomposition, they are not consumed by the process or are consumed at an advantageously slow rate. This may allow for longer running times for a cell implementing the process, simpler cell design and lower overall anode costs. In addition, the melt and the product metal may not be contaminated by material from the anode, which may increase the working life of the melt and may reduce the number of postprocessing steps required for the product.

Advantageously the method, when operated for a sufficient time, has as its end product the metal; for example if Fe ₂ O ₃ is the solid compound the product would then be Fe. The electro-decomposition of the solid compound from oxide to pure metal may take place via a number of intermediate compounds. Any of these intermediate compounds may be removed as the product of the process if the process is not run for a sufficient time to allow for complete reduction to the metal.

If the alkali metal were to be produced in metallic form at the cathode and/or to dissolve in the melt, electronic conductivity of the melt may increase. This may disadvantageously reduce the electrical efficiency of the process. The process is advantageously run under conditions such that alkali metal from the

hydroxide does not continuously deposit as a metal at the cathode. The melt may contain more than one species of alkali metal in which case the potential applied between the anode and the cathode is preferably not sufficient for any alkali metal that is present in the melt to deposit continuously as a metal at the cathode. An absence of dissolved alkali metal in the melt is likely to advantageously reduce or substantially eliminate electronic conductivity of the melt.

During electro-decomposition in a hydroxide melt, it is thought that hydrogen may potentially evolve on the cathode at a lower cathode potential than may be required for the reduction of alkali metal ions present in the melt to alkali metal. Any such alkali metal produced at the cathode could potentially dissolve in the melt and may disadvantageously change the properties of the melt, for example by increasing its electronic conductivity. The preferential evolution of hydrogen may advantageously prevent alkali metal ion reduction.

The reaction at the cathode may be monitored by measuring the cathode potential against a reference electrode. Examples of this include a true reference electrode such as a Ag/AgCI electrode, or a psuedo-reference electrode calibrated by a method such as cyclic voltammetry, or a dynamic reference electrode such as an electrode comprising a metal that is the same as an alkali metal species in the melt.

A cell for operating a process according to a particularly-preferred embodiment of the invention would have a melt of sodium hydroxide, containing some sodium oxide, and an anode comprising nickel.

In one embodiment, iron oxide may be electro-decomposed in a melt comprising sodium oxide dissolved in sodium hydroxide. With reference to this embodiment, the following reaction equations and standard electrochemical potentials demonstrate the suitability of the approach disclosed.

The theoretical standard electrochemical potentials for the decomposition reactions of various iron oxides are compiled below. The numbers given below are calculated from tabulated thermodynamic data and are for the temperature

of 600 ⁰ C and unit activities; negative potentials correspond to positive free energies and indicate that energy input is required to enable the reactions to proceed from the left-hand side to the right-hand side.

s 1. Fe ₂ O ₃ = 2 Fe + 1.5 O ₂ E ⁰ = -1.022 V

2. Fe ₃ O ₄ = 3 Fe + 2 O ₂ E ⁰ = -1.075 V

3. FeO = Fe + 0.5 O ₂ E ⁰ = -1.074 V

4. Fe ₂ O ₃ = 2 FeO + 0.5 O ₂ E ⁰ = -0.919 V

5. 3 Fe ₂ O ₃ = 2 Fe ₃ O ₄ + 0.5 O ₂ E° = -0.601 V o 6. Fe ₃ O ₄ = 3 FeO + 0.5 O ₂ E ⁰ = -1.077 V

If during electro-decomposition a mixed oxide containing sodium, iron and oxygen is formed, either chemically or electrochemically, and either before or in the course of electro-decomposition, the following reaction equation and s standard electrochemical potential need to be considered.

7. Na ₂ Fe ₂ O ₄ = 2 Fe + 1.5 O ₂ + Na ₂ O E° = -1.351 V

The stability of the electrolyte is believed to be determined by the following 0 reaction equations and standard electrochemical potentials. The numbers refer to 600 ⁰ C and unit activities.

8. Na ₂ O = 2 Na+ 0.5 O ₂ E ⁰ = -1.434 V

9. 2 NaOH = Na ₂ O+ 0.5 O ₂ + H ₂ E ⁰ = -1.650 V S 10. 2 NaOH = 2 Na + 0.5 O ₂ + H ₂ O E ⁰ = -2.050 V

11. 2 NaOH = 2 Na + O ₂ + H ₂ E ⁰ = -3.084 V

In the embodiment, the sodium oxide is diluted by the sodium hydroxide and so its actual decomposition potential is lower (more negative) than the one 0 calculated for standard conditions. In quantitative terms, the decomposition potential of sodium oxide becomes more negative by 173 mV at 600°C for each order of magnitude the concentration falls below saturation concentration. In the embodiment it is likely that the concentration will be within one order of magnitude of saturation and so the decomposition potential will be in the order of -1.434 - 0.173 = -1.607 V.

Concerning the sodium hydroxide, the most favourable decomposition reaction leads to the generation of hydrogen gas (reaction 9), while decomposition reactions involving the deposition of sodium metal (reactions 10 and 11 ) 5 require significantly more negative potentials.

As can be seen from the above, electro-decomposition of iron oxides in a sodium hydroxide melt should result in the formation of iron, preferably by applying a cell voltage which is sufficient to cause removal of oxygen from iron o oxides (reactions 1 to 6) but which is not sufficient to cause continuous evolution or deposition of hydrogen or sodium at the cathode (reactions 8 and 9). It should be noted that in order to cause such continuous evolution or deposition of hydrogen or sodium, the cell voltage would have to exceed-the voltages corresponding to reactions 8 or 9 by a sufficient margin in order to s overcome voltage losses in the cell.

The iron product should not react with the sodium hydroxide as the standard free energies for the following reactions are positive.

0 12. Fe + 2 NaOH = FeO + Na ₂ O + H ₂ ΔG° = +111 kJ / moI

13. 3 Fe + 8 NaOH = Fe ₃ O ₄ + 4 Na ₂ O + 4 H ₂ ΔG° = +444 kJ / mol

14. 2 Fe + 6 NaOH = Fe ₂ O ₃ + 3 Na ₂ O + 3 H ₂ ΔG° = +363 kJ / mol

In general, a method embodying the invention may advantageously be used to 5 remove oxygen from a compound, metal or alloy as long as a compound formed between oxygen and a cation in the melt is more stable than the oxygen or oxide in the compound, metal or alloy.

Description of Specific Embodiment o A specific embodiment of the invention will now be described by way of example, with reference to the drawings, in which;

Figure 1 shows a cell for an electro-decomposition process according to an embodiment of the invention. 5

Figure 1 shows a cell 10 for electro-decomposition containing a melt 20 of composition 98% sodium hydroxide and 2% sodium oxide. A cathode 30 in the form of an iron basket 40 containing Fe ₂ O ₃ particles 50 is immersed in the melt. An anode 60 of commercially pure nickel is also immersed in the melt, the anode and the cathode both being connected to a power supply 70.

In operation the melt is heated up to its operating temperature, for example 400 ⁰ C. An operating potential of, for example, 2.5 to 3.0 V is applied between the anode and the cathode. At the operating potential, oxygen in the Fe ₂ O ₃ transfers to the melt and is transported to the anode, where it is evolved as oxygen gas.

In a second embodiment, the melt is heated to an operating temperature of 55O ⁰ C. All the other reaction conditions are described above. This embodiment reduces the Fe ₂ O ₃ to Fe at a higher rate than in the first embodiment. It is believed that this is due to increased diffusion rate in the material at the cathode.

To form the cathode comprising a solid metal compound or solid metal, it may be advantageous to prepare the compound or metal into a porous form, for example by slip-casting (and optionally sintering) the compound or metal in powdered form. In such a structure the material at the cathode should contain interconnected porosity to allow penetration of the melt, and the particle size should be small enough to allow oxygen diffusion. In practice, the material at the cathode may be prepared in any manner and in any geometry, but the reaction rate may be limited by the rate of diffusion of oxygen in the material if disadvantageously thick sections of the material are used.