Fuel cell catalyst layer

Abstract

A proton exchange membrane fuel cell (PEM-FC) anode catalyst layer comprising a hydrogen oxidation reaction catalyst, an ion-conducting material, and an oxygen evolution reaction (OER) catalyst. The o

Description

Field of the Invention The present invention relates to an anode for a proton exchange membrane fuel cell, which layer contains an oxygen evolution reaction catalyst. The invention also provides a new oxygen evolution reaction catalyst. Background of the Invention A fuel cell is an electrochemical cell comprising two electrodes separated by an electrolyte. A fuel, e.g. hydrogen, an alcohol such as methanol or ethanol, or formic acid, is supplied to the anode and an oxidant, e.g. oxygen or air, is supplied to the cathode. Electrochemical reactions occur at the electrodes, and the chemical energy of the fuel and the oxidant is converted to electrical energy and heat. Electrocatalysts are used to promote the electrochemical oxidation of the fuel at the anode and the electrochemical reduction of oxygen at the cathode. Fuel cells are usually classified according to the nature of the electrolyte employed. Often the electrolyte is a solid polymeric membrane, in which the membrane is electronically insulating but ionically conducting. In the proton exchange membrane fuel cell (PEMFC) the membrane is proton conducting, and protons, produced at the anode, are transported across the membrane to the cathode, where they combine with oxygen to form water. An electrolyser is an electrochemical device for electrolysing water to produce high purity hydrogen and oxygen. Electrolysers can operate in both alkaline and acidic systems. Those electrolysers that employ a solid proton-conducting polymer electrolyte membrane, or proton exchange membrane (PEM), are known as proton exchange membrane water electrolysers (PEMWEs). Those electrolysers that utilise a solid anion-conducting polymer electrolyte membrane, or anion exchange membrane (AEM), are known as anion exchange membrane water electrolysers (AEMWEs). A principal component of the fuel cell or water electrolyser is the membrane electrode assembly (MEA). The MEA is typically composed of five layers. The central layer is the polymer ion-conducting membrane. On either side of the ion-conducting membrane there is an electrocatalyst layer, containing an electrocatalyst designed for the specific electrolytic reaction. Finally, adjacent to each electrocatalyst layer there is a gas diffusion layer, and / or a porous transport layer. The gas diffusion layer (or porous transport layer) allows the reactants to reach the electrocatalyst layer and conduct the electric current that is generated by the electrochemical reactions. The gas diffusion layer (or porous transport layer) is porous and electrically conducting. The electrocatalyst layers generally comprise an electrocatalyst material comprising a metal or metal alloy suitable for an oxidation reaction (e.g. fuel oxidation) or a reduction reaction (e.g. oxygen reduction reaction), depending on whether the layer is to be used at the anode or cathode. The electrocatalyst is typically based on platinum or platinum alloyed with one or more other metals. The platinum or platinum alloy catalyst can be in the form of unsupported nanoparticles (such as metal blacks or other unsupported particulate metal powders) but more conventionally the platinum or platinum alloy is deposited as higher surface area nanoparticles onto a high surface area conductive carbon material, such as a carbon black or heat treated versions thereof. Anode catalysts for PEMWEs typically comprise iridium or iridium oxide (IrOx) materials, or oxides containing both iridium and ruthenium. Typically tens or hundreds of MEAs are required to provide enough power for most applications, so multiple MEAs are assembled to make up a fuel cell stack. Field flow plates are used to separate the MEAs. The plates perform several functions: supplying the reactants to the MEAs; removing products; providing electrical connections; and providing physical support. It is known that there are a number of situations in which incorporating a water electrolysis catalyst into a fuel cell at either the anode or the cathode can prove beneficial. For example, WO01 / 15247 and WO2004 / 010521 describe how incorporating an additional or second catalyst composition at the anode for purposes of electrolysing water can improve tolerance of a fuel cell to cell voltage reversal. Cell voltage reversal can occur if a cell receives an inadequate supply of fuel (for example, as a result of blockage of fuel access to a part of that cell). If this occurs, reactions other than fuel oxidation may take place at the fuel cell anode, because the potential on the anode is raised to high values, including water electrolysis and oxidation of anode structural components. Oxidation of anode structural components is undesirable as this can result in significant degradation of the anode. By incorporating a catalyst composition at the anode which promotes the oxygen evolution reaction, degradation of the anode can be reduced or avoided, by promotion of water electrolysis over anode structural component oxidation. Another example of a situation in which promotion of water electrolysis may be beneficial is for fuel cells where it is not practical or economic to provide purging of hydrogen from the anode gas space with an inert gas such as nitrogen during shut down, or when a cell is restarted after being idle for some time. Both of these situations can result in a mixed composition of hydrogen and air on the anode whilst air is present on the cathode. Under these circumstances an internal cell can exist, as described by Tang et al (Journal of Power Sources 158 (2006) 1306-1312), which leads to high potentials on the cathode. The high potentials can cause carbon to oxidise which is highly damaging to the structure of the catalyst layer where the catalyst layer contains carbon. If the cathode layer is able to support oxygen evolution however, the high potentials can be used to drive water electrolysis rather than carbon corrosion. Also, in regenerative fuel cells, the electrodes are bi-functional and both anode and cathode must support two electrochemical reaction types at different times. When operating as a fuel cell the cathode must reduce oxygen and the anode oxidise hydrogen; when operating as an electrolyser the cathode must evolve hydrogen and the anode evolve oxygen. It may therefore be beneficial to incorporate both a traditional hydrogen oxidation reaction catalyst and an oxygen evolution reaction catalyst in the anode of such a fuel cell, because with such an arrangement, the anode can carry out both the hydrogen oxidation and oxygen evolution reactions effectively. Various electrocatalysts for the oxygen evolution reaction are known in the art. For example, in addition to WO01 / 15247 and WO2004 / 010521, WO11 / 021034 discloses catalyst layers comprising an electrocatalyst and an oxygen evolution reaction catalyst, wherein the oxygen evolution reaction catalyst comprises iridium or iridium oxide and one or more metals M or an oxide thereof, wherein M is selected from the group consisting of transition metals and Sn, with the exception of ruthenium. Summary of the Invention The provision of improved cell reversal tolerant fuel cell catalyst layers is desirable, along with the provision of catalysts for use in such layers. The present inventors have provided such a layer which has surprisingly high durability to cell reversal events, and high performance. These advantages lead to numerous benefits, especially at scale, including reduced cost, less exposure to risks regarding availability of metal, increased recyclability, and less risk of negative effects on the chemistry of a catalyst-coated membrane. Accordingly, the present invention provides a proton exchange membrane fuel cell anode catalyst layer comprising a hydrogen oxidation reaction catalyst, an ion-conducting material, and an oxygen evolution reaction catalyst; wherein the oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium, preferably as defined above; and wherein the loading of iridium is 0.02 mg / cm2 of the geometric area of the anode catalyst layer or less. It is surprising, in particular in view of WO01 / 15247, WO2004 / 010521 and WO11 / 021034 that a catalyst layer with such a low loading of iridium has, as shown in the Examples section, greater durability to cell reversal events and greater performance than conventional catalyst layers which contain an oxygen evolution reaction catalyst to facilitate cell reversal tolerance. It is also a surprising advantage in view of disclosure in WO2011 / 021034 that ruthenium dissolution is not observed during operation to the extent it is detrimental in membraneelectrode assemblies containing the catalyst layer of the invention. The examples herein demonstrate that neither ruthenium nor iridium migrate to a detrimental extent from the anode to the cathode during operation. The present invention also provides a new oxygen evolution reaction catalyst, which may be used in a catalyst layer of the invention. The oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium: wherein the oxygen evolution reaction catalyst comprises a tetragonal crystalline oxide phase; wherein the oxygen evolution reaction catalyst has a peak between 515 and 525 cm’1 in the Raman spectra. The new oxygen evolution reaction catalyst is prepared by a new method which suitably imparts the physical properties of the catalyst. Accordingly, the present invention also provides a process of producing an oxygen evolution reaction catalyst, wherein the oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium, wherein the process comprises the steps of: i) providing a solution of metal halide salts of ruthenium and iridium; ii) spray drying the solution from i) and collecting the solid product; iii) calcining the solid product from step ii) at a temperature of less than 600 °C; iv) milling the calcined product of step iii); v) calcining the milled product of step iv) at a temperature of greater than 600°C. The present invention also provides an oxygen evolution reaction catalyst obtainable by this process. The present invention also provides: A proton exchange membrane fuel cell gas diffusion electrode comprising a gas diffusion layer and a catalyst layer according to the invention. A proton exchange membrane fuel cell catalyst-coated membrane comprising an ionconducting membrane and a catalyst layer according to the invention. A proton exchange membrane fuel cell membrane-electrode assembly comprising a catalyst layer according to the invention, a gas diffusion electrode according to the invention, or a catalyst-coated membrane according to the invention. A proton exchange membrane fuel cell comprising a catalyst layer according to the invention, a gas diffusion electrode according to the invention, a catalyst-coated membrane according to the invention, or a membrane-electrode assembly according to the invention. The present invention also provides a catalyst-coated transfer substrate comprising a transfer substrate and a catalyst layer according to the invention. Brief Description of the Figures Figure 1 provides an x-ray diffraction pattern for an oxygen evolution reaction catalyst material according to the invention. Figure 2 provides a Raman spectra for an oxygen evolution reaction catalyst material according to the invention along with iridium oxide and ruthenium oxide references. Figure 3 provides a plot of reversal holds vs cell reversal voltage for membrane-electrode assemblies according to the invention and a membrane-electrode assembly containing a benchmark anode catalyst layer, acquired by a procedure emphasising durability comparisons. Figure 4 provides a plot of reversal holds vs cell reversal voltage for membrane-electrode assemblies according to the invention and a membrane-electrode assembly containing a benchmark anode catalyst layer, acquired by a different procedure from that of Fig. 3 with an emphasis on performance comparisons. Figure 5 provides a plot of voltage vs time held at 200 mA / cm2 for membrane-electrode assemblies according to the invention and a membrane-electrode assembly containing a benchmark anode catalyst layer, showing time to failure. Figure 6 also provides a plot of voltage vs time held at 200 mA / cm2 for membrane-electrode assemblies according to the invention and a membrane-electrode assembly containing a benchmark anode catalyst layer, showing time to failure. Figure 7 represents electron probe microanalysis (EPMA) data at end of life for a membraneelectrode assembly according to the invention, showing lack of iridium and ruthenium migration from the anode. Figure 8 represents EPMA data at end of life for a membrane-electrode assembly according to the invention, showing lack of iridium and ruthenium migration form the anode. Detailed Description of the Invention The anode catalyst layer of the invention has an ultra-low loading of iridium, especially in view of the loadings described in WO01 / 15247, WO2004 / 010521 and WO11 / 021034. Thus, the layer benefits from lower metal use which itself has numerous benefits, especially at scale, including reduced cost, less exposure to risks regarding availability of metal, increased recyclability, and less risk of negative effects on the chemistry of a catalyst-coated membrane. Accordingly, in the catalyst layer the loading of iridium is 0.02 mg / cm2 of the geometric area of the anode catalyst layer or less, preferably 0.018 mg / cm2 or less, or 0.015 mg / cm2 or less. The loading of iridium in the catalyst layer may be 0.010 mg / cm2 of the geometric area of the anode catalyst layer or less, 0.007 mg / cm2 or less, or 0.005 mg / cm2 or less. The loading of iridium may be at least 0.0001 mg / cm2 of the geometric area of the anode catalyst layer, or at least 0.001 mg / cm2. The loading of iridium may be at least 0.003, 0.005, 0.008, or 0.010 mg / cm2 of the geometric area of the anode catalyst layer. The hydrogen oxidation reaction catalyst comprises a platinum group metal. Preferably, the catalyst comprises particles of the platinum group metal or an alloy of the platinum group metal, optionally supported on an electrically conductive support. That is, the catalyst can be unsupported particles of the platinum group metal or the alloy of the platinum group metal (e.g. finely divided unsupported metal powder) or may be a supported catalyst wherein the particles (e.g. nanoparticles) are dispersed on an electrically conductive support, such as an electrically conductive carbon support. In the present case, the platinum group metals suitably include platinum, palladium, ruthenium, rhodium or osmium. Most preferably, the anode catalyst is platinum, which may be alloyed with an alloying metal such as other precious metals (e.g. a different platinum group metal) or base metals. A base metal is tin or a transition metal which is not a noble metal. A noble metal is a platinum group metal, silver or gold. Preferred base metals are copper, cobalt, nickel, zinc, iron, titanium, molybdenum, vanadium, manganese, niobium, tantalum, chromium and tin. The alloying metal can be aluminium, yttrium, scandium, gadolinium, lanthanum, tungsten, zirconium, or hafnium. If the hydrogen oxidation reaction catalyst is supported, the loading of platinum group metal particles on the electrically conductive support material may be at least 10 weight %, at least 30 weight %, preferably greater than 50 weight %, more preferably at least 55 weight %, more preferably at least 60 weight % based on the total weight of the platinum group metal plus support. The loading of platinum group metal particles on the electrically conductive support material may be less than or equal to 90 weight %, or less than or equal to 80 weight % based on the total weight of the platinum group metal plus support. The loading of the platinum group metal particles can be determined using inductively coupled plasma mass spectrometry (ICPMS). The term “supported” will be readily understood by a skilled person. For example, it will be understood that the term “supported” includes the particles of the anode catalyst being dispersed on (and / or in the pores of) the support material and bound or fixed to the support material by physical or chemical bonds. For instance, the anode catalyst may be bound or fixed to the support material by way of ionic or covalent bonds, or non-specific interactions such as van der Waals forces. The electrically conductive support may be an electrically conductive carbon support material. Suitably, the electrically conductive carbon support material is a carbon powder which may be, for example, a carbon black or graphitised carbon black for example a commercially available carbon black (such as from Cabot Corp. (Vulcan® XC72R) or Akzo Nobel (the Ketjen® black series)). Another suitable carbon support material is an acetylene black (e.g. those available from Chevron Phillips (Shawinigan Black®) or Denka). The electrically conductive carbon support can be prepared by the method disclosed in WO2013 / 045894. Alternatively, the electrically conductive support can be a metal oxide or a mixed oxide, in particular a conductive mixed oxide such as niobia-doped titania, phosphorus-doped tin oxide and mixed platinum group metal oxides or mixed metal oxides (as disclosed in WO2012 / 080726), a carbide (e.g. tungsten carbide, molybdenum carbide or titanium carbide, suitably tungsten carbide or titanium carbide), a nitride, in particular a conductive nitride (e.g. titanium nitride or titanium aluminium nitride). The anode catalyst layer may have a thickness of greater than or equal to 1 micron, preferably greater than or equal to 2 microns. The anode catalyst layer may have a thickness of less than or equal 15 microns, preferably less than or equal to 10 microns, preferably less than or equal to 5 microns. The hydrogen oxidation reaction catalyst may be present in the anode catalyst layer in a loading of less than or equal to 0.30 mg / cm2of the geometric area of the anode catalyst layer, less than or equal to 0.20 mg / cm2, less than or equal to 0.15 mg / cm2, or less than or equal to 0.10 mg / cm2. The catalyst may be present in the anode catalyst layer in a loading of greater than or equal to 0.02 mg / cm2 of the geometric area of the anode catalyst layer. The ion-conducting material of the anode catalyst layer is preferably an ion-conducting polymer, preferably a proton-conducting polymer. Preferred ion-conducting polymers are partially- or fully-fluorinated sulphonic acid polymers e.g. perfluorinated sulphonic acid polymers. For example, the ion-conducting polymer may be based on a perfluorinated sulphonic acid material such as Nation® (Chemours Company), Aquivion® (Syensqo), Flemion® (Asahi Glass Group) and Aciplex® (Asahi Kasei Chemicals Corp.). Alternatively, the ion-conducting materials may be based on a sulphonated hydrocarbon polymer, such as those available from FuMA-Tech GmbH as the fumapem® P, E or K series of products, and others. The ion-conducting polymer in the anode catalyst layer suitably has an equivalent weight (EW) in a range of 600 to 1200, wherein EW is the number of grams of dry polymer per mole of sulfonic acid groups when the material is in the acid form. Particular perfluorosulfonic acid ionconducting polymers include the Nation® range available from Chemours Company, especially Nation® 1100EW and DE2020CS, the Aquivion® range available from Syensqo, especially 830 EW. In the case in which the hydrogen oxidation reaction catalyst is supported on a support material, for example on an electrically conductive carbon support material, the loading of ionconducting material, for example proton-conducting polymer, can be expressed in terms of weight of ion-conducting material with respect to total weight of the support material. The ionconducting material may be present in an amount of at least 50 weight % with respect to the weight of the support material (i.e. the weight ratio of ion-conducting material to support material is at least 50:100), at least 60 weight % with respect to the weight of the support material, or at least 70 weight % with respect to the weight of the support material. The ionconducting material may be present in an amount of at most 120 weight % with respect to the weight of the support material (i.e. the weight ratio of ion-conducting material to support material is at most 120:100), at most 100 weight % with respect to the weight of the support material, preferably less than 90 weight % with respect to the weight of the support material, or at most 85 weight % with respect to the weight of the support material. The catalyst layer may be prepared by a number of methods known to those skilled in the art, for example by preparation of an ink and applying the ink to a membrane, gas diffusion layer or transfer substrate by standard methods such as gravure coating, slot die (slot, extrusion) coating (whereby the coating is squeezed out under pressure via a slot onto the substrate), screen printing, rotary screen printing, inkjet printing, spraying, painting, bar coating, pad coating, gap coating techniques such as knife or doctor blade over roll (whereby the coating is applied to the substrate then passes through a split between the knife and a support roller), metering rod application such as with a Meyer bar, and laser induced forward transfer (LIFT). The invention also provides a catalyst-coated transfer substrate comprising a transfer substrate and a catalyst layer as defined herein. Such a catalyst-coated transfer substrate is suitably for decal transfer of the catalyst layer onto another substrate, such as an ionconducting membrane or a gas diffusion layer using techniques known in the art, for example in processes for preparing the catalyst-coated membrane and gas diffusion electrodes described herein. Here, the catalyst layer is deposited onto the transfer substrate, typically by direct coating of a catalyst layer ink onto the substrate. The transfer substrate may be any suitable transfer substrate known to those skilled in the art but is preferably a polymeric material such as polytetrafluoroethylene (PTFE), polyimide, polyvinylidene difluoride (PVDF), or polypropylene (especially biaxially-oriented polypropylene, BOPP) or a polymer-coated paper such as polyurethane coated paper. The transfer substrate could also be a silicone release paper or a metal foil such as aluminium foil. The invention also provides a proton exchange membrane fuel cell catalyst-coated membrane comprising an ion-conducting membrane and the anode catalyst layer as described herein. Here, the catalyst layer is deposited onto an ion-conducting membrane, either by direct coating of a catalyst layer ink onto the membrane, or indirectly by transfer from a carrier or transfer substrate, to form a catalyst-coated membrane. The ion-conducting membrane may suitably have a thickness of 15 microns or less, for example 12 microns or less or 10 microns or less. The thickness will be determined by the intended use. Thickness may be measured by scanning electron microscopy (SEM). SEM analysis is carried out on cross sections of the ionconducting membrane and / or catalyst-coated ion-conducting membrane measured at multiple (for example 10) points. The thickness values are then determined by calculating the arithmetic mean of the measured values. Alternatively, the thickness of the ion-conducting membrane layer may be measured using a low force high precision gauge instrument (e.g. VL-50B Litematic™ available from Mitutoyo (UK) Ltd.), which may give a direct reading of the membrane thickness. A motorised spindle is used to take measurement readings with a measuring force of 0.01 N. At least three readings are taken from different locations on the ion-conducting membrane layer (prior to adding catalyst layers) at a temperature of 20 °C ± 3 °C, and a relative humidity (RH) of 30-50%. The ion-conducting membrane comprises an ion-conducting polymer. The ion-conducting polymer is preferably a proton-conducting polymer. Preferred ion-conducting polymers are partially- or fully-fluorinated sulphonic acid polymers e.g. perfluorinated sulphonic acid polymers. For example, the ion-conducting polymer may be based on a perfluorinated sulphonic acid material such as Nation® (Chemours Company), Aquivion® (Syensqo), Flemion® (Asahi Glass Group) and Aciplex® (Asahi Kasei Chemicals Corp.). Alternatively, the ion-conducting material may be based on a sulphonated hydrocarbon polymer, knows has a hydrocarbon ionomer, such as those available from FuMA-Tech GmbH as the fumapem® P, E or K series of products, and others. The ion-conducting polymer may have an equivalent weight (EW) in a range of 600 to 1200, wherein EW is the number of grams of dry polymer per mole of sulfonic acid groups when the material is in the acid form. Particular perfluorosulfonic acid ion-conducting polymers include the Nation® range available from Chemours Company, especially Nation® and DE2020CS, the Aquivion® range available from Syensqo. The ion-conducting membrane can comprise at least one reinforcing component, or optionally two reinforcing components. The reinforcing component(s) can confer mechanical strength to the ion-conducting membrane. The reinforcing component is preferably a porous material having pores extending through the thickness of the material in the through-plane direction. Preferably the reinforcing component is a porous polymer material. Preferably, the porous polymer material is a porous fluoropolymer material, such as a microporous web or fibres of a polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), ethylene tetrafluoroethylene (ETFE), perfluoroalkoxy alkane (PFA), or fluorinated ethylene propylene (FEP). For example, the planar reinforcing component may comprise electrospun PVDF or forcespun PVDF. The porous polymer material is typically expanded PTFE (ePTFE), for example, such as the microporous web structures of ePTFE supplied by Donaldson Company, Inc., known as Tetratex®, or supplied by other manufacturers. Alternatively, the reinforcing component can comprise a network of fibres (e.g. nanofibres), such as a network comprising polybenzimidazole (PBI) fibres. The network of fibres can be a non-woven mat of fibres (e.g. nanofibres), such as an electrospun mat of fibres or nanofibres. The ion-conducting membrane may also comprise a reinforcing component in the form of a woven fabric, such as a woven fabric formed from polymer threads, such as ePTFE or PEEK threads. Suitable materials are described in US11742507B2 (AGC, INC). The invention also provides a proton exchange membrane fuel cell gas diffusion electrode comprising a gas diffusion layer and a catalyst layer as described herein. Here, the catalyst layer is deposited onto a gas diffusion layer, typically onto a microporous layer on the surface of the gas diffusion layer, either by direct coating of a catalyst layer ink, or indirectly by transfer from a carrier or transfer substrate, to form a gas diffusion electrode. The gas diffusion layer is preferably based on a conventional gas diffusion substrate. Typical gas diffusion substrates include non-woven papers or webs comprising a network of carbon fibres and a thermoset resin binder (e.g. the TGP-H series of carbon fibre paper available from Toray Industries Inc., Japan or the H2315 series available from Freudenberg FCCT KG, Germany, or the Sigracet® series available from SGL Technologies GmbH, Germany or AvCarb® series from Ballard Power Systems Inc., or woven carbon cloths. The carbon paper, web or cloth may be provided with a further treatment prior to being incorporated into a MEA either to make it more wettable (hydrophilic) or more wet-proofed (hydrophobic). The nature of any treatments will depend on the type of fuel cell and the operating conditions that will be used. The substrate can be made more wettable by incorporation of materials such as amorphous carbon blacks via impregnation from liquid suspensions, or can be made more hydrophobic by impregnating the pore structure of the substrate with a colloidal suspension of a polymer such as PTFE or polyfluoroethylenepropylene (FEP), followed by drying and heating above the melting point of the polymer. A microporous layer is typically applied to the gas diffusion substrate on the face that will contact the catalyst layer. The microporous layer typically comprises a mixture of a carbon black and a polymer such as polytetrafluoroethylene (PTFE). The invention also provides a proton exchange membrane fuel cell membrane-electrode assembly comprising a catalyst layer, a gas diffusion electrode or a catalyst-coated membrane as defined herein. The membrane-electrode assembly may be constructed in a number of ways including, but not limited to: (i) An ion-conducting membrane may be sandwiched between two electrodes (one anode and one cathode), the anode gas diffusion electrode as defined herein: (ii) A catalyst-coated membrane as defined herein may be sandwiched between a gas diffusion layer and a cathode gas diffusion electrode, the gas diffusion layer contacting the side of the membrane coated with the anode catalyst layer defined herein: (iii) A catalyst-coated membrane as defined herein which also comprises a proton exchange membrane fuel cell cathode catalyst layer may be sandwiched between two gas diffusion layers. The membrane-electrode assembly may further comprise components that seal and / or reinforce the edge regions of the membrane-electrode assembly which are known to those skilled in the art. The membrane-electrode assembly is assembled by conventional methods known to those skilled in the art. The invention also provides a proton exchange membrane fuel cell comprising a catalyst layer as defined herein, a gas diffusion electrode as defined herein, a catalyst-coated membrane as defined herein, or a membrane-electrode assembly as defined herein. In the anode catalyst layer of the invention, the oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium, for example an iridium and ruthenium mixed metal oxide. Put another way, the oxygen evolution reaction catalyst may suitably comprise a single oxide structure containing both iridium and ruthenium, for example a mixed crystal oxide containing both iridium and ruthenium. The ratio of iridium to ruthenium may typically be in the range of and including 1:9.9 to 4:5, typically 1:9.9 to 2:8, more typically typically 1:9.5 to 1.5:8.5, for example 1:9, as determined, for example, by ICPMS. The atomic ratio of total iridium and ruthenium (Ir + Ru) to oxygen is typically 1:2, as determined, for example, by ICPMS. Preferably, the oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium: wherein the oxygen evolution reaction catalyst comprises a tetragonal crystalline oxide phase; wherein the oxygen evolution reaction catalyst has a peak between 515 and 525 cm-1 in the Raman spectra. The invention also provides this oxygen evolution reaction catalyst perse, which is new in the art. Iridium may suitably be present in the oxygen evolution reaction catalyst in an amount of at least 1 atomic %, preferably at least 5 atomic %, more preferably at least 10 atomic % based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst. Iridium may suitably be present in an amount of at most 20 atomic %, preferably at most 15 atomic % based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst. In a preferred composition, iridium is present in an amount in the range of and including 1 to 20 atomic % based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst. In a more preferred composition, iridium is present in an amount in the range of and including 5 to 15 atomic % based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst. Iridium may be present in an amount of 10 atomic % based on the total atomic composition of iridium and ruthenium in the oxygen evolution reaction catalyst. Ruthenium may suitably be present in the oxygen evolution reaction catalyst of the invention in an amount of 99 atomic % or less, preferably 95 atomic % or less, more preferably 90 atomic % or less based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst. Ruthenium may suitably be present in an amount of 80 atomic % or more, preferably 85 atomic % or more based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst. In a preferred composition, ruthenium is present in an amount in the range of and including 80 to 99 atomic % based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst. In a more preferred composition, ruthenium is present in an amount in the range of and including 85 to 95 atomic % based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst. Ruthenium may be present in an amount of 90 atomic % based on the total atomic composition of iridium and ruthenium in the oxygen evolution reaction catalyst. The amount of iridium and ruthenium may be determined by the molar amount of iridium and ruthenium included in preparation of the material and can be confirmed using inductively coupled plasma mass spectrometry (ICPMS). The oxygen evolution reaction catalyst of the invention may optionally comprise metal species other than iridium and ruthenium, for example in an amount up to and including 5 atomic % based on the total atomic composition of metal species in the oxygen evolution reaction catalyst, preferably less than 1 atomic % based on the total atomic composition of metal species in the oxygen evolution reaction catalyst. Typically, iridium and ruthenium constitute substantially all of the metal species present in the oxygen evolution reaction catalyst. Put another way, the metal species in the oxygen evolution reaction catalyst consist essentially of, preferably consist of, iridium and ruthenium. The atomic ratio of total iridium and ruthenium (Ir + Ru) to oxygen is typically 1:2. The oxygen evolution reaction catalyst may optionally comprise metal species in the metallic form, for example in an amount less than 5 atomic % based on the total atomic composition of metal species in the oxygen evolution reaction catalyst, preferably less than 1 atomic % based on the total atomic composition of metal species in the oxygen evolution reaction catalyst. Such metal species in the metallic form may be iridium. The oxygen evolution reaction catalyst may advantageously contain low levels of halide, preferably chloride. For example, the oxygen evolution reaction catalyst may contain less than 3 weight % halide, preferably chloride, typically less than 2 weight % halide, preferably chloride, by total weight of the oxygen evolution reaction catalyst, which may be measured by conventional halide, preferably chloride, assays. The oxygen evolution reaction catalyst of the invention comprises a tetragonal crystalline oxide phase. This single tetragonal crystalline oxide phase contains iridium and ruthenium and as such may be referred to as a solid solution of iridium and ruthenium. The oxygen evolution reaction catalyst may also comprise other oxide phases, for example amorphous oxide phases. The other oxide phases may comprise an oxide of iridium or an oxide of ruthenium or both, typically IrO2 or RuO2. Suitably, any other oxide phase(s) is / are present in an amount of less than 5 weight %, preferably less than 1 weight % by total weight of the oxygen evolution reaction catalyst. Put another way, the tetragonal crystalline oxide phase is preferably present in an amount of at least 99 weight % by total weight of the oxygen evolution reaction catalyst. Preferably, the oxygen evolution reaction catalyst does not comprise any other crystalline oxide phases, more preferably does not comprise any other oxide phases (amorphous or crystalline). Accordingly, distinct lrO2 and / or RuO2may be present in an amount of less than 5 weight %, suitably less than 1 weight % by total weight of the oxygen evolution reaction catalyst. Preferably, the oxygen evolution reaction catalyst does not contain distinct lrO2 and / or RuO2. As shown in Fig. 1, the x-ray diffraction pattern for the oxygen evolution reaction catalyst of the invention shows that the sample is composed of a single phase, which is an undocumented tetragonal oxide phase which contains ruthenium and iridium, i.e. it is an oxide solid solution containing iridium and ruthenium. The diffraction pattern indicates that no RuO2 or lrO2 is present. The crystalline oxide phase may have a crystallite size of less than or equal to 12.5 nm. The crystalline oxide phase may have a crystallite size of at least 11.5 nm, preferably at least 11.9 nm. Crystallite size may be calculated by the method set out in the Examples section. As shown in Fig. 2, the Raman spectra for the oxygen evolution reaction catalyst of the invention shows a dominant peak which is between the dominant peaks in lr[IV] and Ru[IV] reference materials, which is attributed to the crystalline oxide phase containing both iridium and ruthenium. Accordingly, the oxygen evolution reaction catalyst has a peak between 515 and 525 cm-1, for example a peak at 520 cm-1, in the Raman spectra. Said peak is a dominant peak e.g. it has a relative intensity, normalised to 1, of greater than 0.5, typically greater than 0.8. Raman measurement was carried out on a Renishaw inVia™ confocal Raman microscope, equipped with 532 nm laser for a spectral range of 100 to 3500 cm-1. The sample was heaped into a specialised powder holder plate for the Renishaw instrument. The following settings were used during the Raman spectroscopy measurement; 0.1 % of the total laser power (50mW), 10 seconds exposure, 50 accumulations, *50 objective and analysis depth ~0.7 microns. The Raman data was processed and analysed using Renishaw wire 5.5 software package. In the oxygen evolution reaction catalyst of the invention, the iridium to ruthenium atomic ratio at the surface of the catalyst may be greater than the bulk atomic ratio of iridium to ruthenium. This may be determined by XPS measurement, for example by the method set out in the Examples section. The bulk atomic ratio is the atomic ratio in the entire catalyst and the ratio is in line with the atomic percentages for iridium and ruthenium described above. Accordingly, the bulk atomic ratio of iridium to ruthenium may be in the range of and including 1:9.9 to 2:8, typically 1:9.5 to 1.5:8.5, for example 1:9, as determined, for example, by ICPMS. A greater atomic ratio of iridium to ruthenium at the surface means that XPS measurements reveal a greater ratio of iridium to ruthenium than the ratio known for the bulk from the molar amount of iridium and ruthenium used in preparation of the material and / or ICPMS analysis. XPS measurement may provide an atomic ratio of iridium to ruthenium in the range of and including 1:7 to 1:3, preferably 1:6 to 1:4. Put another way, the average atomic ratio of iridium to ruthenium to a depth of about 7 nm (i.e. the depth to which XPS analyses) in particles of the oxygen evolution reaction catalysts may be in the range of and including 1:7 to 1:3, preferably 1:6 to 1:4. Whilst not wishing to be bound by theory, this may result from non-uniform distribution of iridium and ruthenium sites in the single tetragonal crystalline oxide phase. The majority of the ruthenium at the surface of the catalyst, i.e. greater than 90 atomic %, preferably greater than 95 atomic %, more preferably greater than 99 atomic %, is present in the +4 oxidation state, as determined by XPS. The BET surface area of the oxygen evolution reaction catalyst of the invention may be less than or equal to 8.00 m2 / g. The determination of the specific surface area by the BET method may be carried out by the following process: after degassing to form a clean, solid surface, a nitrogen adsorption isotherm is obtained, whereby the quantity of gas adsorbed is measured as a function of gas pressure, at a constant temperature (usually that of liquid nitrogen at its boiling point at one atmosphere pressure). A plot of l / [Va ((Po / P)-I)] vs P / Po is then constructed for P / P0 values in the range 0.05 to 0.3 (or sometimes as low as 0.2), where Va is the quantity of gas adsorbed at pressure P, and Po is the saturation pressure of the gas. A straight line is fitted to the plot to yield the monolayer volume (Vm ), from the intercept l / Vm C and slope (C l) / Vm C, where C is a constant. The surface area of the sample can be determined from the monolayer volume by correcting for the area occupied by a single adsorbate molecule. More details can be found in 'Analytical Methods in Fine Particle Technology', by Paul A. Webb and Clyde Orr, Micromeritics Instruments Corporation 1997. The present invention also provides a process of producing an oxygen evolution reaction catalyst, suitably the oxygen evolution reaction catalyst of the invention, wherein the oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium, wherein the process comprises the steps of: i) providing a solution of metal halide salts of ruthenium and iridium; ii) spray drying the solution from i) and collecting the solid product; iii) calcining the solid product from step ii) in air at a temperature of less than 600°C; iv) milling the calcined product of step iii); v) calcining the milled product of step iv) in air at a temperature of greater than 600°C. The solution of metal halide salts of ruthenium and iridium is typically formed in an aqueous solvent, typically water, and may be provided by first preparing separate solutions of iridium salt and ruthenium salt and then combining these solutions. A solubilising agent may be used to facilitate dissolution of the salt compound in the solvent. Alternatively or additionally, the mixture, typically in water, may be heated to aid dissolution, typically with stirring, for example to a temperature in the range of 40°C to 80°C. The metal halide salts are preferably metal chloride salts. A suitable iridium chloride salt is IrCh. A suitable ruthenium chloride salt is RuCh. The molar ratio of iridium to ruthenium in this solution provided in step i) will set the atomic ratio of iridium to ruthenium in the product oxygen evolution reaction catalyst. The average solids content of the solution provided in step i) is typically in the range of and including 10 weight % to 30 weight %. The step of spray drying the resultant solution to form a solid product may be performed using the entire solution prepared or, typically, batchwise for example by separating the solution into two or more, typically no more than four, batches for spray drying. Spray drying is typically carried out with an inlet temperature, for example, in the range of and including 150°C to 300°C, typically in the range of and including 200°C to 250°C. Typically, an atomiser pressure of greater than 1 bar and no more than 2 bar is used, for example 1.5 bar. The air flow rate may typically be 5 to 15 kg / hr, for example 7 to 11 kg / hr. After spray drying, the solid product is calcined in step iii). The calcination may be performed on the entire solid product prepared in step ii) or, typically, batchwise for example by separating the solid product into two or more, typically no more than eight, for example no more than six batches. The step iii) calcination is a single step of heating to a temperature of less than 600°C, preferably at most 550°C. The temperature is suitably at least 450°C. Time at the required temperature is typically in the range of and including thirty minutes to two hours. There may also be a step of ramping up the temperature from ambient to the calcination temperature, for example at a rate in the range of and including 5 to 20°C / minute. After the calcination step, the solids are typically left to cool to ambient temperature. After step iii), the calcined solids are milled in step iv). In the case that calcination was carried out in step ii) on batches of solid product, these same batches are typically milled individually. Milling is typically carried out using a ball mill, for example with a sieve having an aperture size in the range of and including 0.05 to 0.15 cm, for example 0.07 to 0.10 cm. This milling step may have the advantage that agglomerates can be broken up before the second calcination step v), allowing for more even calcination. After milling step iv) a second calcination step v) is caried out which is a single step of heating to a temperature of greater than 600°C, preferably at least 650°C. The temperature is suitably at most 850°C, typically at most 750°C. Time at that temperature is typically in the range of and including thirty minutes to two hours. There may also be a step of ramping up the temperature from ambient to the calcination temperature, for example at a rate in the range of and including 5 to 20°C / minute. After the calcination step, the solids are typically left to cool to ambient temperature. In the case that steps iii) and iv) were carried out on batches of solid product, these same batches are typically calcined in step v) individually. The calcined product is allowed to cool to provide the oxygen evolution reaction catalyst. Calcination steps iii) and v) may also be performed with continuous agitation of the powder such as in rotary calcination with or without the inclusion of beads or balls to facilitate the break-up of agglomerates. Calcination steps iii) and v) may be performed in a suitable gaseous atmosphere, for example in air, N2, Ar, He, CO2, CO, O2, H2, and mixtures thereof. Preferably, calcination is performed in an air atmosphere. This process is different from the process disclosed in, for example, WO01 / 15247, WO2004 / 010521 and WO11 / 021034 at least because of the calcination temperatures and the existence of the second calcination step after milling. Without wishing to be bound by theory, it is believed that the different process may impart the new physical and technical properties of the oxygen evolution reaction catalyst of the invention. Examples Synthesis of oxygen evolution reaction catalysts according to the invention To prepare an oxygen evolution catalyst having a 9:1 atomic ratio of Ru to Ir, the following aqueous solutions of Ru and Ir chlorides were prepared, with an overall Ru / lr molar ratio of 9:1: RuCI3 (41.26%) = 1199.4 g (494.9 g, 4.896 mol Ru) lrCI3 (51.60%) = 199.77 g (103.08 g, 0.536 mol Ir) lrCI3 (51.20%) =2.97 g (1.52 g, 0.008 mol Ir) The salt solutions were mixed and water (4.0 litre) was added to the mixed salts using a stirring rod to aid dissolution. The solution was warmed to ~60°C and magnetically stirred at this temperature for ~15 min. The heating was switched off but stirring continued over the weekend. The resulting solution was divided between two plastic containers for spray drying. Contents of container 1 (after rinsing) = 2743.5 g Contents of container 2 (after rinsing) = 2816.9 g Both containers were sampled in duplicate to determine the solids content of each. The aliquots were dried overnight in air at 105°C. The average solids content of solution was 22.18%. Spray drying of solution Spray dryer configuration: Co-Current Two Fluid Nozzle Recess - used thick spacer Single point discharge Collecting jars warmed at ~120°C Atomiser air pressure = 1.5 bar Atomiser flow meter air rate = 68% Atomiser air flow rate = 9.4 kg It1 Inlet temperature = 220°C Pump feed setting = 5.0 Collection jars were used to collect the solids. Estimated total solids passed through dryer = 1279 g. Total dried solid collected was 1269.0g (includes any moisture in the product) with an estimated overall yield = 99.2%. Calcination to 500° C The contents of various jars were combined and calcined in five trays in air under the following conditions: 10°C min’1 ramp to 500°C, hold for 1 hour then allow to cool to ambient. The combined batch was placed on a rotating roller bed for 5 hours to ensure complete mixing. The yield of combined material was 588.0 g. Milling The calcined material was milled using a Retsch ball mill with a 0.08 cm sieve. Calcination to 7QCPC The milled material was then calcined under the following conditions: 10°C / min'1 ramp to 700°C, hold for 1 hour, then allow to cool to ambient. The combined batch was placed on a rotating roller bed for 5 hours to ensure complete mixing. The yield of combined material was 588.0 g. BET surface area analysis of samples The BET surface area of the oxygen evolution reaction catalyst produced was determined by the following process: after degassing to form a clean, solid surface, a nitrogen adsorption isotherm is obtained, whereby the quantity of gas adsorbed is measured as a function of gas pressure, at a constant temperature (usually that of liquid nitrogen at its boiling point at one atmosphere pressure). A plot of l / [Va ((Po / P)-I)] vs P / Po is then constructed for P / P0 values in the range 0.05 to 0.3 (or sometimes as low as 0.2), where Va is the quantity of gas adsorbed at pressure P, and Po is the saturation pressure of the gas. A straight line is fitted to the plot to yield the monolayer volume (Vm ), from the intercept l / Vm C and slope (C-I) / Vm C, where C is a constant. The surface area of the sample can be determined from the monolayer volume by correcting for the area occupied by a single adsorbate molecule. More details can be found in 'Analytical Methods in Fine Particle Technology', by Paul A. Webb and Clyde Orr, Micromeritics Instruments Corporation 1997. The BET surface area was 7.60 m2 / g. X-ray diffraction data Fig.1 shows the X-ray diffraction pattern of a material produced according to the invention. There are no documented phases in the ICDD database (Sets 1-50), which exactly characterize the sample. However, the diffraction pattern indicates that no ruthenium oxide (RuO2, PDF No. 40-1290) or Iridium Oxide (IrOz, PDF No. 15-0870) is present. Peak profile analysis shows that the sample is composed of a single phase. The tetragonal phase present is evidently an undocumented Ru / lr / O solid solution. X-ray data collection Powder X-ray diffraction (PXRD) data were collected in reflection geometry using a Bruker AXS D-500 diffractometer using Cu Ka radiation (A = 1.5406 + 1.54439 A) over the 15 <20 <95° range in 0.02° steps. Phase identification was conducted using Bruker AXS Diffrac plus Eva V6 with reference to the PDF-2, Release 2003, database. This showed that the material contains a single crystalline phase with reflection intensities which are consistent with a rutile MO2 phase but reflection positions which do not match lrO2 or RUO2. The reflection broadening shows some variation. Sample fitting Pawley and peak phase refinements were performed using Topas[1] with reflection profiles modelled using a fundamental parameters approach[2] with reference data collected from NIST660 LaB6. The data were separately fitted using a set of peaks with independent sample dependant broadening, for the rutile phase, to obtain crystallite sizes along a selection of crystallographic planes. The crystallite sizes have been calculated using the Scherrer method, with a Scherrer constant of K=0.9.[3] The crystallite size of the sample was 12.2 nanometres. 1. Topas v4.21 v5.0: General Profile and Structure Analysis Software for Powder Diffraction Data, Bruker AXS, Karlsruhe, Germany, (2003-2015). 2. R.W. Cheary and A. Coelho, J. Appl. Cryst. (1992), 25, 109-121 3. P. Debye, P. Scherrer, Physikalische Zeitschrift, (1916) X-Ray photoelectron spectroscopy (XPS) data The atomic ratio of iridium to ruthenium in the catalyst material produced according to the invention having a bulk ratio of 1:9 was 1:5.1 by XPS analysis. This demonstrates that there is a greater amount of ruthenium at the surface of the material, the surface being to a depth of about 7 nm of particles of the material i.e. the depth to which XPS investigates. XPS data were collected using a Thermo Scientific NEXSA Surface Analysis System, comprising an ultra-high vacuum (UHV) chamber operating at a typical base pressure of 2 x 10’09 mbar and a load-lock for sample changes. Thermo Scientific NEXSA is fitted with a microfocused monochromatic Al ka X-ray source (1486.7 eV) with an illuminated spot size of 400 microns * 200 microns on the sample surface. Electronic charge neutralization was achieved using a dual-beam low-energy electron / ion source. XPS data were processed and analysed using Thermo Scientific Avantage software package. The atomic % of the elements present at the surface of the materials was determined from the respective XPS peak areas, using a background subtraction based upon the Shirley method and application of sensitivity factors based on the Scofield method. The surface atomic % ratios were calculated from the atomic % of the elements. Raman spectroscopy data The Raman spectra of a sample according to the invention is shown in Fig. 2. This spectra also provides spectra for lr[l V] and Ru[IV] reference materials. The peak positions of the two reference materials are due to Ru-0 and lr-0 stretching in the structure of Ru [IV] and Ir [IV] respectively. The dominant peaks for the sample of the invention, including at - 520 cm-1, are between the Ir[IV] and Ru[IV] reference materials. The observed peaks in the sample of the invention are contributions from both Ru-0 and lr-0 in the structure. The spectra is indicative of a single mixed oxide structure containing both iridium and ruthenium. Raman measurement was carried out on a Renishaw inVia™ confocal Raman microscope, equipped with 532 nm laser for a spectral range of 100 to 3500 cm-1. The sample was heaped into a specialised powder holder plate for the Renishaw instrument. The following settings were used during the Raman spectroscopy measurement; 0.1 % of the total laser power (50mW), 10 seconds exposure, 50 accumulations, *50 objective and analysis depth ~0.7 microns. The Raman data was processed and analysed using Renishaw wire 5.5 software package. The sample was heaped into the powder holder plate, no sample tape was used. Preparation and testing of membrane electrode assemblies The following membrane electrode assemblies (MEAs) were prepared, having the same membrane and cathode catalyst layers. The anode catalyst layers contained either the oxygen evolution reaction catalyst prepared by the method described above, OER A, or a benchmark iridium tantalum mixed oxide oxygen evolution reaction catalyst OER B prepared using a method as disclosed in WO2011 / 021034. The anode catalyst layers also contained an ionconducting polymer and a Pt / C hydrogen oxidation reaction catalyst. Below, “XX wt% Pt / C, means XX percent platinum by weight of total Pt / C catalyst. Also, “XX wt% ion-conducting polymer” means XX percent ion-conducting polymer by weight of the carbon present in the Pt / C catalyst. For example 90 wt% ion-conducting polymer means a weight ratio of 0.9:1 ionconducting polymer to carbon. MEA1 - OER A, 0.015 mg lr / cm2, 60 wt% Pt / C, 90 wt% ion-conducting polymer MEA2 - OER A, 0.015 mg lr / cm2, 60 wt% Pt / C, 70 wt% ion-conducting polymer MEA3 - OER A, 0.007 mg lr / cm2, 60 wt% Pt / C, 90 wt % ion-conducting polymer MEA4 - OER A, 0.007 mg lr / cm2, 60 wt% Pt / C, 70 wt % ion-conducting polymer MEA5 - OER A, 0.004 mg lr / cm2, 60 wt% Pt / C, 70 wt% ion-conducting polymer MEA6 - OER A, 0.002 mg lr / cm2, 60 wt% Pt / C, 70 wt% ion-conducting polymer MEA7 - OER A, 0.007 mg lr / cm2, 50 wt% Pt / C, 70% wt% ion-conducting polymer MEA8 - OER A, 0.004 mg lr / cm2, 50 wt% Pt / C, 70% wt% ion-conducting polymer MEA9 - OER B, 0.069 mg lr / cm2, 60 wt% Pt / C, 70% wt% ion-conducting polymer Anode catalyst layers were prepared by forming inks containing a per-fluoro sulphonic acid (PFSA) ionomer dispersed in a water, a 50 wt% or 60 wt% Pt / C electrocatalyst material, and the oxygen evolution reaction catalyst. This mixture was mechanically agitated using an overhead stirrer until all of the catalyst had been wetted and dispersed in the liquid. The ink was then processed through an Eiger ball mill to form a well dispersed ink. Milled ink was back-diluted with water / propan-1-ol mixture for coating. The platinum loading was 0.08 mgPt / cm2, the loading of the oxygen evolution reaction catalyst of the invention or comparative oxygen evolution reaction catalyst was as described above for MEA1 to MEA9. Cathode catalyst layers contained a 50 wt% Pt / C electrocatalyst in which the carbon support is a carbon specifically designed for use in a fuel cell as described in WO2013 / 045894. Cathode catalyst layers were prepared by forming inks containing a PFSA ionomer dispersed in a water / ethanol mixture and the 50 wt% Pt / C electrocatalyst. This mixture was mechanically agitated using an overhead stirrer until all of the catalyst material had been wetted and dispersed in the liquid. The ink was then processed through an Eiger ball mill to form a well dispersed ink. The platinum loading was 0.4 mgPt / cm2. Catalyst coated ion-conducting membranes of 217 cm2 active area were prepared by depositing the anode ink and cathode inks onto PTFE sheets to form catalyst layers and transferring the appropriate layers to either side of PFSA reinforced membranes (15 micron thickness). A gas diffusion layer was applied to each face of each catalyst coated ion-conducting membrane to form the complete membrane electrode assemblies. The gas diffusion layer used was a carbon fibre paper with a hydrophobic microporous layer containing carbon and PTFE applied to the face in contact with the catalyst coated ion-conducting membrane. Membrane Electrode Assembly Reversal Testing For the data in Fig. 3, the MEAs were tested in a single cell format. First the MEAs were conditioned using a cathode starvation protocol where the MEA was held at 80°C, 100 kPa pressure and 100% RH on both anode and cathode. A current of 500 mAcm-2 was drawn from the cell and the cathode stoichiometry was cycled between 2.0 and 0.0, whilst the anode stoichiometry was kept constant at 1.5. Following 17 of these cathode starvation events the MEA was then kept with a constant cathode stoichiometry of 2.0 and a current of 500 mAcm-2 for 2 hours. Following this, a re-conditioning protocol was run where the MEA was held at the experimental conditions of 65°C, Ambient pressure and 50% RH for 1 hour. At the experimental conditions a beginning-of-life (BOL) polarisation curve between open circuit voltage (OCV) and 2000 mAcnr2 was completed to assess performance at the beginning of life. Next, the MEA was subject to cell reversal cycling where the MEA was held at 200 mAcnr2, and the anode gas stream was switched to N2 and held for 5 minutes, followed by switching this gas stream back to H2 and drawing a current of 500 mAcnr2 for 15 minutes (a reversal hold). This cell reversal cycle was repeated six times before repeating the polarisation curve performance assessment. This sequence of six reversal cycles followed by a polarisation curve performance test was repeated until the performance of the MEA during the performance test was less than 0.35V at 1 Acm-2 or the cell voltage during the reversal holds was less than -1.2V. Fig. 3 is a plot of reversal voltage at 200 mA / cm2 against number of reversal holds for membrane electrode assemblies according to the invention MEA1, MEA3 and MEA4 along with MEA9 containing the benchmark iridium tantalum mixed oxide oxygen evolution reaction catalyst. The less negative potentials overtime forMEAl, MEA3 and MEA4 as compared with MEA9 means superior durability than the benchmark even at greater reversal holds. Therefore, the MEAs are less affected by reversal holds. Moreover, this improvement for MEA1, MEA3 and MEA 4 occurs with a far lower loading of iridium in the layer than in MEA9. That is, 0.015 and 0.007 mg / cm2 a compared with 0.069 mg / cm2. For the data in Fig. 4, the MEAs were tested in a single cell format. First the MEAs were conditioned using a cathode starvation protocol where the MEA was held at 80°C, 100 kPa pressure and 100% RH on both anode and cathode. A current of 500 mAcnr2 was drawn from the cell and the cathode stoichiometry was cycled between 2.0 and 0.0, whilst the anode stoichiometry was kept constant at 1.5. Following 17 of these cathode starvation events the MEA was then kept with a constant cathode stoichiometry of 2.0 and a current of 500 mAcnr 2 for 2 hours. Following this, a re-conditioning protocol was run where the MEA was held at the experimental conditions of 65°C, Ambient pressure and 50% RH for 1 hour. At the experimental conditions a BOL polarisation curve between OCV and 2000 mAcnr2 was completed to assess performance at the beginning of life. Next, the MEA was subject to cell reversal cycling where the MEA was held at 200 mAcm'2, and the anode gas stream was switched to N2 and held for 5 minutes, followed by switching this gas stream back to H2 and drawing a current of 500 mAcm-2 for 15 minutes (a reversal hold). This cell reversal cycle was repeated 6 times before repeating the polarisation curve performance assessment. Fig. 4 is a plot of reversal voltage at 200 mA / cm2 against number of reversal holds for a membrane electrode assemblies according to the invention MEA1, MEA3 and MEA4 along with MEA9 containing the benchmark iridium tantalum mixed oxide oxygen evolution reaction catalyst. It can be seen by the shallower gradient for all three of MEA1, MEA3 and MEA4 than MEA9 that MEA1, MEA3 and MEA4 demonstrate better performance. This means that less carbon corrosion is occurring showing that use of the catalyst layer containing the oxygen evolution reaction catalyst of the invention provides an inherently more active membrane electrode assembly. Figs. 5 and 6 show time to failure for various MEAs, failure being when the potential reaches -1.2V and as such carbon corrosion is dominating the cell potential. The tests are sustained reversal hold test. For acquiring the data, the MEAs were tested in a single cell format. First the MEAs were conditioned by running for an hour, with hydrogen (1.5 stoich.) at the anode and oxygen (2.0 stoich.) at the cathode, at 80°C, 100 kPa pressure, 100% RH on both anode and cathode at a current density of 500 mAcm-2. A current density of 500 mAcm-2 to 2000 mAcm-2 back to 500 mAcm-2 was then drawn with three minutes at each current density for a total of one hour. This step was then repeated. A further conditioning step was carried out for an hour with hydrogen (1.5 stoich.) at the anode and oxygen (2.0 stoich.) at the cathode, at 60°C, 150 kPa pressure, 100% RH on both anode and cathode a current density of 500 mAcm-2. Then, the reversal hold was performed by holding under the same conditions but at a current density of 200 mAcm-2 until the potential reaches -1.2V and the test aborts. Time to failure was recorded. Fig. 5 shows that MEA1, 2, 3, 4 and 5 last substantially longer before failure than MEA9 which contains the benchmark OER catalyst and has a much higher loading of iridium in the anode catalyst layer. Even MEA6, which only has a loading of 0.002 mg lr / cm2 has comparable durability to MEA9, with ~35 times less iridium in the anode layer. Fig. 5 also shows that the resistance of the part to reversal events can be optimised by the percentage of ion-conducting polymer present in the anode layer. MEA2 and MEA4, which differ only in having less ionconducting polymer in the anode layer as compared with MEA1 and MEA3 respectively, show greater resistance to cell reversal. Fig. 6 shows that the loading of platinum on carbon has an effect on the resistance of the part to cell reversal events. MEA3 and MEA5 show greater resistance than MEA7 and MEA8 respectively, with the only difference being that the Pt / C hydrogen oxidation reaction catalyst in the anodes of MEA3 and MEA5 have a higher Pt / C loading. Iridium and ruthenium migration analysis Fig. 7 are electron probe microanalysis (EPMA) images shows that no iridium or ruthenium migrates from the anode to the cathode during the period of operation for MEA1 under sustained reversal hold conditions as described above (Fig.7a is a Ru trace at end of life, Fig. 7b is a Ir trace at end of life). Fig. 8 shows that no iridium or ruthenium migrates from the anode to the cathode during the period of operation for MEA3 under a sustained reversal hold conditions as described above (Fig. 8a is a Ru trace at end of life, Fig. 8b is a Ir trace at end of life). Measurements for EPMA were undertaken using a Jeol JXA-iHP200F electron probe microanalyzer with five wavelength dispersive spectrometers and one SDD (silicon drift diode) EDX detector. Quantitative maps were run by first acquiring standards in Probe for EPMA, acquiring mapping data for the actual sample and then converting the intensity maps to quantitative maps using Calcimage. Mapping data was collected at 20keV with a current of 200nA and a 100ms pixel dwell time. The step size was 0.3 microns.

Claims

1. A proton exchange membrane fuel cell anode catalyst layer comprising a hydrogen oxidation reaction catalyst, an ion-conducting material, and an oxygen evolution reaction catalyst;wherein the oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium; andwherein the loading of iridium is 0.02 mg / cm2 of the geometric area of the anode catalyst layer or less.

2. A proton exchange membrane fuel cell anode catalyst layer according to claim 1, wherein the hydrogen oxidation reaction catalyst comprises a platinum group metal catalyst supported on an electrically conductive carbon support material.

3. A proton exchange membrane fuel cell anode catalyst layer according to claim 2, wherein the ion-conducting material is present in an amount of less than 90 weight % with respect to the weight of the electrically conductive carbon support material.

4. A proton exchange membrane fuel cell anode catalyst layer according to any of claims 2 to 3, wherein the loading of platinum group metal particles on the electrically conductive carbon support material is greater than 50 weight % based on the total weight of the platinum group metal plus support.

5. A proton exchange membrane fuel cell anode catalyst layer according to any of claims 1 to 4, having a thickness of less than or equal to 5 microns.

6. A catalyst-coated transfer substrate comprising a transfer substrate and a catalyst layer according to any of claims 1 to 5.

7. A proton exchange membrane fuel cell gas diffusion electrode comprising a gas diffusion layer and a catalyst layer according to any of claims 1 to 5.

8. A proton exchange membrane fuel cell catalyst-coated membrane comprising an ionconducting membrane and a catalyst layer according to any of claims 1 to 5.

9. A proton exchange membrane fuel cell membrane-electrode assembly comprising a catalyst layer according to any of claims 1 to 5, a gas diffusion electrode according to claim 7, or a catalyst-coated membrane according to claim 8.

10. A proton exchange membrane fuel cell comprising a catalyst layer according to any of claims 1 to 5, a gas diffusion electrode according to claim 7, a catalyst-coated membrane according to claim 8, or a membrane-electrode assembly according to claim 9.

11. An oxygen evolution reaction catalyst, wherein the oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium:wherein the oxygen evolution reaction catalyst comprises a tetragonal crystalline oxide phase;wherein the oxygen evolution reaction catalyst has a peak between 515 and 525 cm’1 in the Raman spectra.

12. The oxygen evolution reaction catalyst according to claim 11, wherein ruthenium is present in 80 atomic % or more based on the total atomic composition of iridium and ruthenium species in the oxygen evolution reaction catalyst.

13. The oxygen evolution reaction catalyst according to claim 11 or claim 12, wherein the tetragonal crystalline oxide phase is a single tetragonal crystalline oxide phase containing iridium and ruthenium.

14. The oxygen evolution reaction catalyst according to any of claims 11 to 13, wherein the iridium to ruthenium atomic ratio at the surface of the catalyst is greater than the bulk atomic ratio of iridium to ruthenium.

15. A process of producing an oxygen evolution reaction catalyst, wherein the oxygen evolution reaction catalyst is an oxide material comprising iridium and ruthenium, wherein the process comprises the steps of:i) providing a solution of metal halide salts of ruthenium and iridium;ii) spray drying the solution from i) and collecting the solid product;iii) calcining the solid product from step ii) at a temperature of less than 600 °C;iv) milling the calcined product of step iii);v) calcining the milled product of step iv) at a temperature of greater than 600 °C.

16. A process according to claim 15, wherein the oxygen evolution reaction catalyst is as defined in any of claims 11 to 14.

17. An oxygen evolution reaction catalyst obtainable by the process according to claim 15.

18. A proton exchange membrane fuel cell anode catalyst layer according to any of claims 1 to 5, wherein the oxygen evolution reaction catalyst is as defined in any of claims 11 to 14 or 17.27

Images