The invention relates to a method of forming a powder of a composite material, to a precursor powder suitable for reduction to a powder of a composite material, and to a powder of a composite material. The invention may be particularly advantageous for the production of powders of composite materials such as dispersion hardened alloys, or metal matrix composites (MMC), or metals reinforced with grain-boundary pinning reinforcements. Background

Composites are materials that have at least two separate constituent parts, the properties of the material being influenced by each of its separate constituent parts. Metal matrix composites (MMCs) are materials having a metallic matrix phase and a reinforcement phase, for example carbon fibre reinforced aluminium. The properties of an MMC are influenced greatly by the properties of the matrix metal and the size, shape, and distribution of the reinforcement phase, as well and the nature of the interface between the matrix phase and the reinforcement phase. MMCs are of increasing economic importance due to their superior properties compared with monolithic materials, but are difficult to manufacture with desired properties on a reliable basis.

There are a number of existing methods for the production of MMCs, for example casting, melt infiltration, and powder processing. Each process has advantages and drawbacks. For example, when a MMC is prepared by a casting method, it is difficult to control the distribution and orientation of the reinforcement phase due to melt dynamics and segregation. It is also difficult to wet the surface of very fine reinforcements, which limits the properties that can be achieved. Powder metallurgy methods for manufacture of MMCs involve mixing metal powder particles with reinforcement particles and then

consolidating the mixture. It is difficult to prepare MMCs with a uniform distribution of fine reinforcements by these methods. In recent years, there has been great interest in the direct production of metal by direct reduction of a solid feedstock, for example, a metal-oxide feedstock. One such direct reduction process is the Cambridge FFC ^® electro- decomposition process (as described in WO 99/64638). In the FFC process, a solid compound, for example a metal oxide, is arranged in contact with a cathode in an electrolysis cell comprising a fused salt. A potential is applied between the cathode and the anode of the cell such that the compound is reduced. In the FFC process, the potential that produces the solid compound is lower than a deposition potential for a cation from the fused salt.

Other reduction processes for reducing feedstock in the form of a cathodically connected solid non-metal compound have been proposed, such as the polar™ process described in WO 03/076690 and the process described in WO 03/048399. Non-metal compounds such as metal oxides may also be directly reduced by metallothermic processes, for example the molten salt calciothermic process described in EP1445350.

Conventional implementations of the FFC process, and other solid-state electrolytic reduction processes, typically involve the production of a feedstock in the form of a porous preform or precursor, the porous preform being fabricated from a sintered powder of the solid non-metallic compound to be reduced. This porous preform is then painstakingly coupled to a cathode to enable the reduction to take place. Once a number of preforms have been coupled to the cathode, the cathode can be lowered into the molten salt and the preforms can be reduced.

WO01/62996 proposes the production of MMC components by using an electrolytic reduction. Particles of ceramics such as silicon carbide or titanium diboride are mixed with titanium dioxide. The mixture is then consolidated into a shape and reduced by an FFC process to provide a MMC with a titanium matrix and particles of ceramic reinforcement. Summary of Invention

The invention provides a method of forming a powder of a composite material, a powder of a composite material, and a precursor powder, as defined in the appended independent claims to which reference should now be made.

Preferred or advantageous features of the invention are set out in dependent subclaims.

A method of forming a powder of a composite material, the composite material comprising a plurality of composite powder particles, each composite powder particle comprising both a matrix phase and a reinforcement phase distributed within the matrix phase, comprises a first step of forming a precursor powder comprising a plurality of precursor powder particles with an average particle size of less than 150 micrometres. Each precursor particle comprises a matrix precursor material, comprising a reducible compound comprising a first metal and oxygen, and a reinforcement precursor material. The method comprises a second step of reducing the precursor powder under conditions in which oxygen is removed from the matrix precursor material but in which the reinforcement precursor material does not react. Following reduction, the matrix phase of the composite material comprises the first metal, and the reinforcement phase of the composite consists of the reinforcement precursor material.

The method of the present invention may advantageously produce a powder of a composite material in which the reinforcement phase is distributed uniformly, or homogeneously, throughout the matrix phase. Preferably the matrix phase is a metal matrix, comprising the first metal. The reinforcement phase may comprise regions of reinforcement dispersed within the composite material as discrete regions, grains, or particles, distributed between grains of the matrix phase, or as discrete regions, grains or particles dispersed within individual grains of material forming the matrix phase. The presence of the

reinforcement phase within the matrix phase may advantageously alter one or more physical, electrical and/or thermal characteristics of the matrix phase, such that the properties of the composite material are different from those of the matrix phase or the reinforcement phase when considered alone.

A single particle of composite material may advantageously comprise a plurality of discrete regions, grains, or particles of the reinforcement phase distributed within one or more grains of the matrix phase.

The reinforcement precursor material may be termed a non-reacting reinforcement precursor material, such that the reinforcement precursor material does not react during or after reduction of the matrix precursor material, and the reinforcement phase in the composite material consists of the reinforcement precursor material. That is, the non-reacting reinforcement precursor material is a material capable of not reacting before, during, or after reduction of the matrix precursor material. A non-reacting reinforcement precursor material forms the reinforcement phase without undergoing a chemical reaction during processing, such that the reinforcement precursor material has the same composition as the reinforcement phase in the composite material. The non-reacting reinforcement precursor material thus remains unreacted, or chemically unchanged, during the reduction of the matrix precursor material, and immediately afterwards when the matrix precursor phase has been reduced.

Preferably the matrix precursor material may comprise an oxide of the first metal, or an oxide of an alloy of the first metal and one or more further metals.

The method of forming a powder of a composite material may therefore comprise: a method of producing a non-metallic precursor powder, or feedstock powder, suitable for reduction to form a composite material; and a method of reducing the precursor material to form a composite material. These steps are described separately in more detail, below.

As is known from the prior art, electrolytic processes may be used, for example, to reduce metal compounds or semi-metal compounds to metals, semi-metals, or partially-reduced compounds, or to reduce mixtures of metal compounds to form alloys. In order to avoid repetition, the term metal will be used in this document to encompass all such products, such as metals, semi- metals, alloys, intermetallics, and partially reduced products. The term non- metallic will be used to describe products that lack metallic attributes, including metal compounds, for example metal oxides, that may be reducible to form a metal. The terms feedstock powder and precursor powder will be used interchangeably.

Methods of forming a Precursor Powder

A precursor powder suitable for reduction to form a powder of a composite material may be formed by a number of methods.

The precursor powder may be formed by providing a powder of the matrix precursor material, and a powder of reinforcement precursor material, and mechanically mixing the two powders together. Preferably the powder of the matrix precursor material is combined with the powder of reinforcement precursor material, and the mixture is mechanically deformed in a high-energy environment such as a ball mill. The mechanical deformation causes the reinforcement precursor material to be incorporated into the matrix precursor material by repeated fracturing and cold welding of the powder particles, such that the reinforcement precursor material is contained at the welded interfaces. After the mechanical deformation, the mixture may be consolidated, for example by heat treatment as described below.

The precursor powder may advantageously be formed using a spray forming process. For example, a freeze drying or lyophilisation process may involve formation of a slurry by mixing matrix precursor particles, reinforcement precursor particles and water. The slurry may then be sprayed into a freezer, or other cold environment such as into a tank of liquid nitrogen, where it freezes into particles that contain matrix precursor, reinforcement precursor and ice. The ice may then be removed by sublimation at low pressure leaving the precursor powder particles. Powder particle size may be controlled by altering parameters such as spray velocity. Freeze drying is a commonly used process for forming pharmaceutical powders.

A similar process that may be used to form the precursor powder is spray drying. In this process, a slurry may be formed by mixing matrix precursor particles, reinforcement precursor particles and a volatile liquid such as ethanol. The slurry is then sprayed into a heating chamber where the volatile liquid is boiled off to leave the precursor powder particles. A particularly advantageous process for forming a precursor powder is fluidised-bed spray granulation. This process is described in more detail below.

Spray Granulation Method A precursor powder suitable for reduction to a composite material may be produced by a fluidised-bed spray granulation method comprising the steps of combining a liquid with solid particles of matrix precursor material and particles of reinforcement precursor material to form a mixture, subjecting the mixture to mixing to form a liquid suspension of matrix precursor material, reinforcement precursor material and the liquid, and drying the liquid suspension using a fluidised-bed spray-granulation process to grow a plurality of precursor particles to form the precursor powder. The spray-granulation process grows a plurality of precursor particles, layer by layer, to a predetermined mean particle diameter. Advantageously, the particles can be grown in a substantially spherical morphology. Spherical precursor particles may be reduced to spherical composite particles, which may be particularly desirable in many powder metallurgy applications. Furthermore, spray-granulation may allow a precise control of particle size and particle size distribution, which may advantageously reduce further grading steps and may also reduce wastage. Still further, spray-granulation may allow a precise control of the distribution of reinforcement precursor material within the precursor particles.

The mixing of the mixture may cause the matrix precursor material particles and the particles of reinforcement precursor material to be milled to a fine particle size, for example a mean particle diameter of less than 2 micrometres. For example, the mixing may involve "high-shear" mixing in which a rotor and a stator interact to introduce a high degree of shear to the liquid. This high shear may result in breakdown of brittle particles to a smaller particle size. High- shear mixing is a known technology and the term "high-shear mixer" is a term that would be understood by the skilled person. If the particles added to the liquid have a mean particle size smaller than that achievable by high-shear mixing, however, their particle size will not be reduced any further. The liquid suspension formed by the mixing process may comprise particles of matrix precursor material and particles of reinforcement precursor material that have uniform fineness (i.e. the matrix precursor material particles have a similar average particle size and a similar particle size distribution as the reinforcement precursor material particles). Alternatively, the liquid suspension formed by the mixing process may comprise particles of matrix precursor material and particles of reinforcement precursor material that have different degrees of fineness (i.e. the matrix precursor material particles have a different average particle size and a different particle size distribution to the

reinforcement precursor material particles). Furthermore, there is preferably a uniform, homogeneous, distribution of matrix precursor material and reinforcement precursor material in the liquid suspension.

The liquid that is combined with matrix precursor material particles and particles of reinforcement precursor material to form the mixture may advantageously comprise water and an organic binder, for example an aqueous solution of polyvinyl alcohol (PVA). Many other suitable binders are known. For example, a suitable binder may comprise polyvinylpyrrolidone (PVP) or hydroxyyethylcellulose (HEC). Preferably the mixture is mixed under high shear, for example mixing at a mixing speed of greater than 5000 rpm. By subjecting the mixture of liquid and particles to a high-shear mixing process, the matrix precursor material particles and particles of reinforcement precursor material may be milled to a similar degree of fineness irrespective of any large variations in particle size that may exist prior to high-shear mixing. That is, if there is a wide particle size distribution in the matrix precursor material particles and particles of reinforcement precursor material that are combined with a liquid to form the mixture then the process of high-shear mixing may advantageously mill the particles such that the overall mean particle diameter decreases and the overall particle size distribution becomes narrower. By forming a liquid suspension containing a uniform distribution of fine particles of matrix precursor material and particles of reinforcement precursor material it is, advantageously, possible to use spray-granulation techniques to grow particles to form a precursor powder to a predetermined mean particle diameter.

The step of mixing may preferably be performed in a high-shear mixer having a rotor capable of rotating in excess of 5000 rpm, for example in excess of 6000 rpm or about 6500 rpm. The rotor rotates relative to a stator in a tank containing the mixture. That is, the rotor and stator are in a tank containing the mixture of the liquid and the particles. The difference in the velocity of liquid near the tip of the rotor and liquid adjacent the stator causes an extremely high-shear zone in the liquid. This high-shear mills the particles in the liquid and forms a suspension of particles in the liquid.

A wide range of high-shear mixers are readily available on the market. For example, small volumes may be processed by high-shear mixing using an IKA G45 M dispersing mixer. The ratio of liquid to particles, or powder, of matrix precursor material and reinforcement precursor material in the mixture may be varied. It is preferred, however, that the mixture is between 50 weight % and 70 weight % particles, with the remainder being the liquid. At greater than 70 weight % of particle loading it may be difficult to generate high-shear mixing sufficient to form a homogenised liquid suspension. Suspensions that are formed with greater than 70 % loading are difficult to pump to the spray granulation apparatus. At lower than 50 weight % of particle loading the time taken to build up precursor particles may be excessive. At lower particle loadings the precursor particles produced by spray granulation may be more spherical. Thus, it may be advantageous to maintain the particle loading to between 50 weight % and 60 weight % of the mixture, or between 50 weight % and 55 weight % of the mixture in order to grow substantially spherical particles at a commercially viable rate.

The step of drying the liquid suspension may comprise steps of spraying a portion of the liquid suspension into a heated chamber of a fluidised-bed spray- granulation apparatus such that liquid is removed from individual droplets of the suspension to form a plurality of seed particles, maintaining the plurality of seed particles within the heated chamber by means of a fluidising gas stream, and spraying further portions of the liquid suspension into the heated chamber, droplets of the liquid suspension successively adsorbing to and drying on the plurality of seed particles. By this means a plurality of precursor particles may be grown or agglomerated, layer-by-layer, to a predetermined particle size.

In another aspect, a method of producing a precursor powder suitable for reduction to a composite material may comprise the steps of forming a first liquid suspension comprising a liquid and particles of the reinforcement precursor material, and forming a second liquid suspension comprising a liquid and particles of the matrix precursor material. The method may comprise the further steps of spraying a portion of the first liquid suspension into a heated chamber of a fluidised-bed spray-granulation apparatus such that liquid is removed from individual droplets of the suspension to form a plurality of seed particles, maintaining the plurality of seed particles within the heated chamber by means of a fluidising gas stream, and spraying further portions of the second liquid suspension into the heated chamber, droplets of the liquid suspension successively adsorbing to and drying on the plurality of seed particles, thereby growing particles to form the precursor powder. The order of spraying the first and second liquid suspensions may be varied so that, for example, the second liquid suspension is sprayed first, and/or further portions of the first and second liquid suspensions may be sprayed in an alternating fashion, so as to form alternating layers of matrix precursor material and reinforcement precursor material on the seed particle. In one embodiment the seed particles may be formed of reinforcement precursor material and the second (matrix precursor material containing) liquid suspension adsorbs to and dries on the outside of the seed particles, such that the outer surface of each precursor particle is formed of matrix precursor material.

Fluidised-bed, spray-granulation processes are known, and have particular application in the pharmaceutical industry. When a liquid feed is sprayed into a chamber it is dried to form a seed particle or germ particle. This germ particle is maintained within by a fluidising gas flow. As further liquid feed is sprayed into the chamber it builds up layer-upon-layer on the seed particle. The seed particle grows larger, forming an onion-like structure. A substantially spherical particle is formed. The particles grow until they are too large to be maintained within the fluidised gas stream, after which they drop out of the bottom of the heated chamber. The resultant particle is dry, substantially spherical, and non- dusting.

By controlling the ratio of the matrix precursor material particles and the particles of reinforcement precursor material in the liquid suspension it is possible to control the composition of the precursor powder formed using the method. The matrix precursor material particles and particles of reinforcement precursor material may be advantageously evenly distributed within the precursor powder produced. In an alternative aspect it may be possible to control the composition of the precursor powder by controlling the ratio of the first and second liquid suspensions that are sprayed into the heated chamber.

Process parameters that may be controlled include the fluidising airflow rate and air temperature (inlet temperature). For the purposes of this invention it may be preferred that airflow during spray granulations falls between about 100 and 190 m ³ per hour, preferably between 130 and 170 m ³ per hour. For the purposes of this invention it may be preferred that the air temperature is between 120°C and 190°C, preferably between 130°C and 150°C. The control of process parameters allows granules of predetermined mean particle diameter within the range of about 5 micrometres to about 10 millimetres to be produced. Preferably, the precursor powder is formed with an average particle size of between 10 micrometres and 300 micrometres.

Different parameters may need to be selected to form particles of the same particle size using matrix precursor material(s) and reinforcement precursor material(s) that have different densities. The high degree of control of such processes allows final precursor particle size and particle size distribution to be controlled within extremely close tolerances. Furthermore, process yields may be substantially in excess of 90%. For example, yields in excess of 95% or 98% are achievable when the process is continuously operated.

The composite powder has an average particle size of 150 micrometres or less. Powders of this particle size may be particularly useful for the production of components by powder metallurgy methods. Powders of such average particle size, in which each powder particle has a matrix phase and a uniformly distributed reinforcement phase, may make new or improved powder metallurgy components possible. Such powders may be particularly suitable for additive manufacturing such as selective laser sintering. Depending on intended use, the average particle size may be lower than 100 micrometres, or lower than 50 micrometres, or lower than 30 micrometres.

Control over the mean particle diameter may allow the formation of precursor powders of specific sizes, which may be used to produce powders of composite material having specific powder properties. WO2014/068267 discusses the production of a metallic powder having specific powder properties for use in specific powder metallurgy processes. Where a powder of composite material is produced, preferred composite powder size ranges following reduction of the precursor powder may vary depending on the desired end use of the composite powder. For example, the following provides an indication of the ranges that are typically preferred for different powder metallurgy processes. In each case, the lower value of the range indicated the D10 particle size and the upper value of the range represents the D90 particle size.

Metal injection moulding (MIM) - particle size range between 5 and 30 micrometres.

Gas dynamic cold spray (GDCS) - particle size range between 15 and 45 micrometres.

Selective laser melting (SLM) - particle size range between 20 and 50 micrometres.

Selective laser sintering (SLS) - particle size range between 20 and 50 micrometres.

Electron beam melting (EBM) - particle size range between 50 and 100 micrometres.

Laser metal deposition (LMD) - particle size range between 50 and 125 micrometres.

Cold isostatic pressing (CIP) - particle size range between 45 and 150 micrometres.

Hot isostatic pressing (HIP) - particle size range between 45 and 200 micrometres.

The actual particle size range of a powder for use in any of the identified processes above may vary outside the stated ranges. These figures are provided as a guideline indicating the preferred or ideal particle size ranges for metal powders used in these processes.

An advantage of the spray-granulation process is that it is possible to form substantially spherical precursor particles. A powder formed from a plurality of such particles may be reduced to a metallic powder of substantially spherical metal particles. Substantially spherical particles are rounded rather than angular and have a low aspect ratio between x, y, and z axes. The aspect ratio is approximately 1 : 1 : 1. Spherical metallic powder particles are advantageous in many powder processing technologies. Currently spherical metallic powders are only produced by processes such as atomisation or spheroidisation of metallic particles. The precursor powder disclosed herein may advantageously be directly reduced to produce a composite powder consisting of a plurality of substantially spherical composite powder particles.

The precursor powder may comprise more than one matrix precursor material and/or more than one reinforcement precursor material. Thus, a first set of matrix precursor material particles and a second set of matrix precursor material particles may be combined with particles of the reinforcement precursor material and the liquid to form the mixture. The first set of matrix precursor material particles have a different composition to the second set of matrix precursor material particles. Preferably, the first set of matrix precursor material particles comprises an oxide of a first metal and the second set of matrix precursor material particles comprises an oxide of a second metal, the first metal being a different metal to the second metal. The mixing of the mixture may cause the different sets of matrix precursor material particles and the particles of reinforcement precursor material to be milled to a fine particle size, for example a mean particle diameter of less than 2 micrometres. If the particles added to the liquid already have a mean particle size smaller than that achievable by high-shear mixing, however, their particle size will not be reduced any further. The liquid suspension formed by the mixing process may comprise matrix precursor material particles of uniform fineness, and particles of reinforcement precursor material of the same or greater fineness.

Furthermore, there is preferably a uniform distribution of each different matrix precursor material and reinforcement precursor material. By controlling the ratio of the first set of matrix precursor material particles and the second set of matrix precursor material particles it may be possible to control the

stoichiometry of the precursor powder formed using the method. The different matrix precursor material particles are evenly distributed within the precursor powder produced and it may be possible to achieve short range stoichiometry in the matrix phase produced following reduction of the precursor powder.

Advantageously, the spray granulation process may be applied to particles in which the first set of matrix precursor material particles and the particles of reinforcement precursor material (and optionally the second set of matrix precursor material particles) have substantially different mean particle sizes. The first set of matrix precursor material particles and the particles of reinforcement precursor material may have mean particle sizes that differ by greater than a factor of 2 (In such a case, the first set of matrix precursor material particles might have a mean particle diameter of 10 micrometres and the particles of reinforcement precursor material might have a mean particle diameter of 5 micrometres.). It may be that the first set of matrix precursor material particles and the particles of reinforcement precursor material have mean particle diameters that differ by a greater factor, such as a factor of 10 or more. It may be that the first set of matrix precursor material particles and the particles of reinforcement precursor material have a mean particle size that differ by greater than a factor of 100. Whenever there is a large disparity in particle sizes between two or more sets of mixed particles it is difficult to produce a precursor of substantially uniform particle size without losing a significant amount of material. The same issue arises where a single set of particles, of a matrix precursor material or of a reinforcement precursor material, has a wide particle size distribution. The proposed process addresses these issues and enables the controllable formation of a precursor powder consisting of substantially even-sized powder particles. Precursor Powder

Precursor powders comprising more complicated compositions may be produced by mixing more than two sets of different matrix precursor material particles with particles of reinforcement precursor material and processing them according to any of the methods described above. For example, the method of producing precursor powder may comprise mixing particles of reinforcement precursor material with three or more sets of matrix precursor material particles, each of the three or more sets having a different

composition.

Precursor powders may similarly contain only one reinforcement precursor material, or two or more sets of particles of different reinforcement precursor materials. It may be advantageous that the matrix precursor material particles that are combined with the liquid to form the mixture are synthetic or refined particles, for example synthetic metal oxide particles. Such particles are typically agglomerations of much finer particles. For example, a vanadium oxide powder particle having a diameter of 100 micrometres may be an agglomeration of much finer sub-particles having particle diameters of, for example, less than 1 micrometre. Such particles may be efficiently milled using a high-shear mixing process. The precursor powder is preferably grown or agglomerated to a predetermined mean particle diameter. The predetermined mean particle diameter of the precursor powder may be anywhere within the range of 5 micrometres to 10 millimetres. Preferable ranges may be between 10 micrometres and 5 millimetres, for example between 20 micrometres and 300 micrometres, or between 50 micrometres and 200 micrometres. Particularly preferably, the precursor powder may have an average particle size of greater than 50 micrometres and/or less than 150 micrometres, or less than 100 micrometres. Advantageously, the process for forming the precursor powder may be controlled such that the precursor powder has a narrow particle size distribution. For example, the width of the particle size distribution may be less than 100 micrometres between a D10 diameter and a D90 diameter. For example, the width of the particle size distribution may be 50 micrometres or less. The ratio of particles of reinforcement precursor material to matrix precursor material particles in the mixture, and thus in the precursor powder, may be varied, depending on the desired composition of the composite material.

Preferably, the ratio of particles of reinforcement precursor material to particles of matrix precursor material may be chosen such that, following reduction, the composite material comprises more than 0.5 volume percent, and less than 20 volume percent of the reinforcement phase. That is, the reinforcement phase makes up between 0.5 and 20 volume percent of the composite material.

Preferably, each particle of the composite material comprises more than 0.5 volume percent, or 1 volume percent, or 2 volume percent and/or less than 5 volume percent, or 10 volume percent, or 15 volume percent of the

reinforcement phase.

Preferably, the particles of the reinforcement precursor material comprised in the precursor powder have an average particle size of between 5 nanometres and 30 micrometres. Particularly preferably, the particles of the reinforcement precursor material have an average particle size of greater than 10

nanometres, or 50 nanometres, or 100 nanometres, or 1 micrometre and/or less than 10 micrometres, or 20 micrometres, or 30 micrometres.

Preferably, the particles of the matrix precursor material comprised in the precursor powder may also have an average particle size of between 5 nanometres and 30 micrometres. Particularly preferably, the particles of the matrix precursor material have an average particle size of greater than 10 nanometres, or 50 nanometres, or 100 nanometres, or 1 micrometre and/or less than 10 micrometres, or 20 micrometres, or 30 micrometres.

The particles of the reinforcement precursor material may have an average particle size that can be related to the average particle size of the precursor powder. For example, the reinforcement precursor material may have an average particle size that is between 5% and 15% of the average particle size of the precursor powder, for example about 10% of the average particle size of the precursor powder. As an example, it may be desired to produce a composite powder that has particular utility as a feedstock in an electron beam melting (EBM) additive manufacturing process. To obtain a composite powder with an average particle size that is optimised for EBM it is preferably to produce a precursor powder having an average particle size of about 100 micrometres. In this circumstance, an advantageous reinforcement precursor material may have an average particle size of between 5 micrometres and 15 micrometres, for example about 10 micrometres.

Preferably, the particles of the reinforcement precursor material comprise between 0.5 and 20 percent by volume of each precursor powder particle, for example between 1 volume percent and 10 volume percent, preferably between 2 volume percent and 5 volume percent.

Mean particle diameter or mean particle size may be determined by a number of different techniques. For example, mean particle diameter (mean particle size) may be determined by sieving, laser diffraction, dynamic light scattering or image analysis. While the exact value of the mean particle diameter of a powder particle may differ slightly depending on the measurement technique used to determine the mean value, in practice the values will be of the same order providing the particles do not have an excessively high aspect ratio. For example, the skilled person will appreciate that the same powder may be found to have a mean particle diameter of, say, 150 micrometres if analysed by sieving, but 142 micrometres if analysed by a different technique such as laser diffraction. A preferable technique for determining mean particle diameter is laser diffraction. For example, mean particle diameters may be determined using an analyser such as the Malvern Mastersizer Hydro 2000 MU. Such an analyser may also be used to determine particle size range or particle size distribution. One standard way of defining particle size distribution in a powder is to refer to D10 and D90 values. D10 is the particle size value that 10% of the population of particles lies below. D90 is the particle size value that 90% of the population lies below. A precursor powder that has a wide particle size distribution will have a large difference between D10 and D90 values. Likewise, a precursor powder that has a narrow particle size distribution will have a small difference between D10 and D90 values.

Preferably, the precursor particles, for example the dried particles derived from the spray granulator, are heat treated prior to being reduced to form a composite material. Heat treatment by means of a suitable firing process may remove any traces of organic binder that remain on each individual particle. The loss of organic binder may introduce a degree of porosity to the particle, and this porosity may be particularly advantageous in some methods of reducing the particles to metal, for instance electrolytic reduction methods involving a molten salt. For example, the precursor powder may comprise a plurality of particles, the particles having a porosity of between 5% and 50%, for example from 10 to 30%. In other words, individual particles may have porosity of between 5% and 50%. Heat treatment may also consolidate fine precursor powders, for example powders combined by ball-milling, and may impart a degree of mechanical strength to each particle.

It may be advantageous to heat treat, or fire, the particles at a relatively high temperature in order to produce a particle comprising a homogeneous distribution of matrix precursor material and a reinforcement precursor material. Thus, the particles of the precursor powder may each be a metal oxide-plus-reinforcement precursor material particle having both a

predetermined particle size and a predetermined composition of metallic elements. Thus, the precursor powder may be reduced to form a matrix phase which is a metal, or alloy, having a homogeneous distribution of the

reinforcement phase and a predetermined composition. Thus, it may be desired to heat treat the precursor particles at a temperature of greater than 900°C, for example at between about 1000°C to 1400°C. The method described above preferably produces a free-flowing precursor powder comprising a plurality of precursor powder particles, each precursor powder particle comprising one or more matrix precursor material(s), for example one or more metal oxide(s), and one or more reinforcement precursor material(s), the precursor powder being suitable for reduction to reinforced metal material.

Reduction of Precursor

A method of producing a powder of a composite material may comprise the steps of forming a precursor powder using a method as described herein and reducing the precursor powder to form a powder of the composite material. The precursor powder is directly reduced to form a powder of a composite material, the resulting composite material being a powder comprising a plurality of discrete composite particles, each of the discrete composite particles comprising the reinforcement phase distributed within the matrix phase.

The precursor powder may be reduced by any suitable method, for example by metallothermic reduction.

It may be preferable to reduce the precursor powder by means of electrolytic reduction of the precursor powder in contact with the molten salt. For example, a volume of the precursor powder may be arranged within an electrolysis cell comprising molten salt and a potential may be applied between an anode and a cathode in order to reduce the precursor by removing oxygen from the matrix precursor material. It may be particularly preferred that the reduction of the precursor powder is effected by use of the FFC process wherein the precursor powder is brought into contact with a cathode and a molten salt in an electrolysis cell and a potential is applied between the cathode and the anode such that the precursor is reduced.

A method for producing a composite material may comprise the steps of arranging a cathode and an anode in contact with a molten salt within an electrolysis cell, an upper surface of the cathode supporting a volume of the precursor powder, and a lower surface of the anode being vertically spaced from the precursor powder and the cathode, and applying a potential between the cathode and the anode such that oxygen is removed from the matrix precursor material such that the precursor powder is reduced.

Some reduction processes may only operate when a molten salt or electrolyte used in the process comprises a metallic species (a reactive metal) that forms a more stable oxide than the metallic oxide or compound being reduced. Such information is readily available in the form of thermodynamic data, specifically Gibbs free energy data, and may be conveniently determined from a standard Ellingham diagram or predominance diagram or Gibbs free energy diagram. Thermodynamic data on oxide stability and Ellingham diagrams are available to, and understood by, electrochemists and extractive metallurgists (the skilled person in this case would be well aware of such data and information). Thus, a preferred electrolyte for an electrolytic reduction process may comprise a calcium salt. Calcium forms a more stable oxide than most other metals and may therefore act to facilitate reduction of any metal oxide that is less stable than calcium oxide. In other cases, salts containing other reactive metals may be used. For example, a reduction process according to any aspect of the invention described herein may be performed using a salt comprising lithium, sodium, potassium, rubidium, caesium, magnesium, calcium, strontium, barium, or yttrium. Chlorides or other salts may be used, including mixtures of chlorides or other salts.

By selecting an appropriate electrolyte, almost any metal oxide matrix precursor particles may be capable of reduction using the methods and apparatuses described herein. Naturally occurring minerals containing one or more such oxides may also be reduced. In particular, oxides of beryllium, boron, magnesium, aluminium, silicon, scandium, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, germanium, yttrium, zirconium, niobium, molybdenum, hafnium, tantalum, tungsten, and the lanthanides including lanthanum, cerium, praseodymium, neodymium, samarium, may be reduced, preferably using a molten salt comprising calcium chloride.

The skilled person would be capable of selecting an appropriate electrolyte in which to reduce a particular metal oxide, and in the majority of cases an electrolyte comprising calcium chloride will be suitable.

The step of reducing the precursor powder is carried out under conditions in which oxygen is removed from the matrix precursor material but in which the reinforcement precursor material does not react, such that the reinforcement phase of the resulting composite consists of the reinforcement precursor material. The reaction of the reinforcement precursor material may be avoided, for example, by controlling the voltage across the electrolytic cell, the temperature of the electrolytic cell, and the duration of the reduction process. For example, by lowering the temperature of the salt during reduction, and/or shortening the duration of the reduction, reduction of the reinforcement precursor material may be avoided. FFC type electroreductions may be carried out with a molten salt temperature of between 650°C and 1200°C. For reduction via FFC in a calcium chloride based salt, a typical reduction temperature is 950°C, and a typical reduction duration is 50 hours. It may be advantageous to reduce the precursor powder by FFC at a temperature of 900°C or lower, preferably at a temperature equal to or lower than 850°C, or 800°C, and for a reduction time equal to or less than 50 hours, or 45 hours or 40 hours. Preferably a non-reacting reinforcement precursor material may be selected to form the reinforcement phase without reacting, as described further below.

Following formation of the powder of composite material, the composite material obtained after reduction may be heat treated. For example, the composite material may be spheroidised to produce a powder of composite material comprising a plurality of spheroidised composite particles.

Particularly preferably, the composite material obtained after reduction is consolidated to form a solid shape, for example by pressing and sintering, or additive manufacturing. In a preferred embodiment, a composite material obtained from the method of the present invention may be consolidated to form a solid shape by selective laser sintering, or SLS.

The method of the present invention may be particularly advantageous in the formation of composite powders, which may be termed reinforced metal powders, such as dispersion hardened alloys, metal matrix composites (MMC), or metals reinforced with grain-boundary pinning reinforcements. In a particlarly preferred embodiment, the method of the present invention may be used to form a powder of a composite material, the particles of which comprise a matrix phase comprising an alloy of more than one metal, and a

reinforcement phase. For the purposes of this application intermetallics are considered to be alloys.

Matrix Precursor Material The matrix precursor material of the precursor powder preferably comprises an oxide comprising one or more first metal selected from the list consisting of beryllium, boron, magnesium, aluminium, silicon, scandium, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, germanium, yttrium, zirconium, niobium, molybdenum, hafnium, tantalum, tungsten, and the lanthanides, including lanthanum, cerium, praseodymium, neodymium, and samarium. For example, the matrix precursor material of the precursor powder may comprise an oxide comprising one or more first metal selected from the list consisting of magnesium, aluminium, silicon, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, zirconium, niobium, molybdenum, hafnium, tantalum, and tungsten.

In some embodiments, the matrix precursor material comprises one or more oxides, at least one of the oxides comprising the first metal and oxygen, such that reducing the precursor powder removes the oxygen from the matrix precursor material to form a composite material comprising an alloy matrix phase comprising the first metal and one or more further metals, and a reinforcement phase.

The precursor powder may comprise at least one additional matrix precursor material, the at least one additional matrix precursor material comprising one or more further metals and oxygen, such that reducing the precursor powder removes the oxygen from the matrix precursor material and the at least one additional matrix precursor material to form a composite material comprising an alloy matrix phase comprising the first metal and one or more further metals, and a reinforcement phase.

Reinforcement Precursor Materials

The reinforcement precursor material is preferably chosen to provide the composite material with properties, for example improved mechanical, thermal or electrical properties, which differ from the properties of the matrix phase in the absence of a reinforcement phase. The composite material may be a reinforced metal material, in which the properties of the composite are a combination of the properties of the reinforcement phase and the matrix phase. Where the reinforcement phase is formed from a material having a hardness greater than that of the metal, for example, the composite material may advantageously have a hardness between that of the reinforcement phase and that of the metal phase.

The reinforcement phase may modify the characteristics of the matrix phase, which is preferably a metal matrix, by, for example, grain boundary pinning or dispersion hardening. The addition of reinforcements of a desired size and a desired quantity may advantageously pin the grain boundaries of the first metal formed during reduction, so as to control the average grain size within the composite material and produce a grain boundary strengthening effect. A higher proportion of the reinforcement phase may produce smaller metal grains within the composite material, for example within a single particle of the composite material. Grain boundary strengthening caused by the

reinforcement phase may advantageously impede dislocation movement and improve the yield strength of the composite material, particularly at high temperatures. The presence of reinforcement phase within the grains of the matrix phase of the composite material may advantageously strengthen the composite by dispersion hardening, as such reinforcements harden, or strengthen, the matrix phase by impeding the movement of dislocations. The reinforcement precursor material may be termed a non-reacting

reinforcement precursor material, such that the reinforcement precursor material does not react during or after reduction of the matrix precursor material, and the reinforcement phase in the composite material consists of the reinforcement precursor material. That is, the non-reacting reinforcement precursor material is a material capable of not reacting before, during or after reduction of the matrix precursor material. The non-reacting reinforcement precursor material forms the reinforcement phase without reacting, such that the reinforcement precursor material is the same material that forms the reinforcement phase in the composite material. The non-reacting reinforcement precursor material thus remains unreacted, or chemically unchanged, during the reduction of the matrix precursor material, and immediately afterwards when the matrix precursor phase has been reduced. Reaction of the non-reacting reinforcement precursor material may be avoided by controlling the conditions under which the precursor powder is reduced, such that oxygen is removed from the matrix precursor material but in which the reinforcement precursor material does not react. The term non-reacting is used because the reinforcement precursor material is not necessarily inert, or unreactive, as the same material may react in different conditions to those experienced during the method of the present invention.

The non-reacting reinforcement precursor materials may comprise a metal compound that is sufficiently inert not to be reduced during reduction of the precursor powder. The non-reacting reinforcement precursor material may be a material that is capable of being reduced, for example and oxide material such as yttria, but that is not reduced during processing because it is electrochemically more stable than the matrix precursor phase and reduction is carried out under conditions that reduce the matrix precursor phase but not the reinforcement precursor phase. For example, reduction may occur in an electrochemical cell at a potential that is sufficient to reduce the matrix phase precursor but is not sufficient to reduce the reinforcement phase precursor. Kinetic considerations may also inhibit the reduction of the non-reacting reinforcement precursor material. Where the precursor material has been formed from first and second liquid suspensions to comprise a core of reinforcement precursor material and an outer portion of matrix precursor material, for example, the matrix precursor material may experience greater exposure to the salt during reduction, while the reinforcement precursor material has little contact with the salt. In this case, the matrix precursor material may be preferentially reduced, while the reinforcement precursor material remains substantially unreacted. Advantageous non-reacting reinforcement precursor materials may include transition-metal carbides or nitrides, or non-reducible intermetallic compounds. Particularly preferably, non-reacting reinforcement precursor materials may be selected from the list consisting of SiC, S13N4, BN, B ₄ C, Y2O3, ScO, titanium aluminide or titanium silicide. Carbon, in particular crystalline carbon such as carbon nanotubes, or graphene, may also act as a non-reacting reinforcement precursor material.

Precursor Powder

A precursor powder, or feedstock powder, suitable for reduction to a powder of a composite material may also be provided. The precursor powder may be formed by a technique such as ball-milling, freeze drying, spray drying, sol-gel processing, or spray granulation, as described above. The precursor powder comprises a plurality of precursor powder particles, each particle comprising a reinforcement precursor material and a matrix precursor material, for example a metal oxide. The precursor particles preferably have a predetermined mean particle diameter. The precursor powder is preferably formed as a free-flowing powder consisting of substantially spherical particles. The precursor powder may have a mean particle diameter of between 50 micrometres and 500 micrometres, for example between 100 micrometres and 250 micrometres. Particularly preferably, the precursor powder has an average particle size of less than 150 micrometres. The precursor powder may comprise an agglomeration, or mixture, of matrix precursor material particles interspersed with particles of reinforcement precursor material. Preferably the reinforcement precursor material is distributed within each precursor powder particle as discrete regions having an average size of between 5 nanometres and 30 micrometres. The matrix precursor material may be distributed within each precursor powder particle as discrete regions having an average size of between 5 nanometres and 30 micrometres. Alternatively, each particle may comprise a core formed from the reinforcement precursor material, which is encapsulated by the matrix precursor material, for example when the precursor is formed by a method using first and second liquid suspensions. Preferably, each precursor powder particle comprises between 0.5 % and 20 % by volume of the reinforcement material, for example between 1 % and 10 % by volume of the reinforcement material, for example between 2% and 5% by volume of the reinforcement material.

Any additional processing features for forming the precursor powder as disclosed above may be combined herein with this aspect of the invention of a precursor powder. Preferably the precursor powder comprises a plurality of precursor powder particles, each particle comprising a reinforcement precursor material and a matrix precursor material, in which the average precursor powder particle size is less than 150 micrometres, and in which the reinforcement precursor material is one or more compound selected from the list consisting of SiC, S13N4, BN, B ₄ C, Y2O3, ScO, titanium aluminide or titanium silicide,

phosphorous, selenium, sulphur tellurium, and carbon, for example carbon in the form of carbon black, carbon nanotubes, or graphene, and the matrix precursor material comprises one or more metal selected from the list consisting of beryllium, boron, magnesium, aluminium, silicon, scandium, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, germanium, yttrium, zirconium, niobium, molybdenum, hafnium, tantalum, tungsten, lanthanum, cerium, praseodymium, neodymium, and samarium.

The precursor powder may comprise the first metal and one or more further metals, such that reducing the precursor powder removes the oxygen from the matrix precursor material to form a composite material comprising an alloy matrix phase comprising the first metal and one or more further metals, and a reinforcement phase. The precursor powder may comprise one or more oxides, at least one of the oxides comprising the first metal and oxygen, such that reducing the precursor powder removes the oxygen from the precursor powder to form a composite material comprising an alloy matrix phase comprising the first metal and one or more further metals, and a reinforcement phase The precursor powder may comprise at least one additional matrix precursor material, the at least one additional matrix precursor material comprising one or more further metals and oxygen, such that reducing the precursor powder removes the oxygen from the matrix precursor material and the at least one additional matrix precursor material to form a composite material comprising an alloy matrix phase comprising the first metal and one or more further metals, and a reinforcement phase. Each precursor particle may comprise a solid solution mixed oxide and particles of reinforcement precursor material. A precursor powder comprising such a mixed oxide may be reduced to form a metallic alloy particle with a homogenous distribution of reinforcements. The precursor powder may be a powder comprising homogeneously dispersed metal oxides and reinforcement precursor material. In a preferred embodiment of the invention, the precursor powder may comprise reinforcement precursor material and titanium, aluminium and vanadium, along with oxygen, and the ratio of the metallic elements in the oxide powder may be between 5.5 wt% and 8 wt% aluminium, between 3.5 wt% and 6 wt% vanadium with the remainder being titanium. Depending on the composition, such a precursor may be suitable to form a reinforced titanium 6-4 (Ti-6AI-4V) alloy.

Alternative precursor powders may be suitable for reduction to a reinforced titanium-aluminide alloy powder or a reinforced titanium-tantalum alloy powder or a reinforced tantalum-tungsten alloy powder or a reinforced tantalum- aluminium alloy powder.

Composite Material

A powder of a composite material, which may be termed a reinforced metal material, may also be provided, in which the powder of composite material comprises a plurality of composite particles, each particle comprising a reinforcement phase and a matrix phase comprising a first metal. The composite material is formed by reducing a precursor powder according to the method and precursor powder as described above. Features of the composite material may be as described in relation to the other aspects of the invention. A single particle of composite material may advantageously comprise a plurality of grains of the matrix phase and a plurality of grains of the

reinforcement phase.

The composite material is in the form of a powder, and preferably comprises discrete regions of reinforcement phase having an average size of between 5 nanometres and 30 micrometres. Preferably, particles or regions of the reinforcement phase may have an average size that can be related to the average particle size of the composite powder. For example, the reinforcement phase may have an average size that is between 5% and 20% of the average particle size of the composite powder, for example between 8% and 15% of the average particle size of the composite powder, for example about 10% of the average particle size of the composite powder.

Preferably each particle of the composite material contains between 0.5 and 20 volume percent of the reinforcement phase, that is the reinforcement phase forms between 0.5 and 20 volume percent of the composite material.

Preferably, each particle of the composite material comprises more than 0.5 volume percent, or 1 volume percent, or 5 volume percent and/or less than 10 volume percent, or 15 volume percent, or 20 volume percent of the

reinforcement phase. Particularly preferably each particle of the composite material contains between 1 and 10 volume percent of the reinforcement phase, for example between 2 and 5 volume percent.

Preferably, each particle of the composite material comprises a substantially uniform distribution of reinforcement phase.

Preferably the reinforcement phase may be one or more compound selected from the list consisting of SiC, S13N4, BN, B ₄ C, Y2O3, ScO, titanium aluminide or titanium silicide, and carbon, for example crystalline carbon in the form of carbon nanotubes, or graphene.

The matrix phase preferably comprises one or more metal selected from the list consisting of beryllium, boron, magnesium, aluminium, silicon, scandium, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, germanium, yttrium, zirconium, niobium, molybdenum, hafnium, tantalum, tungsten, lanthanum, cerium, praseodymium, neodymium, and samarium. Preferably, the composite material comprises a particle reinforced material, a dispersion hardened material, a metal matrix composite (MMC), and/or a metal matrix with grain-boundary pinning reinforcements.

The composite material is preferably a particle- re info reed composite material, in which reinforcement particles reinforce a metal matrix. Such a material may be termed a particle-reinforced metal matrix composite (MMC). Depending on the size of the reinforcement regions, or particles, the particle-reinforced composite material may be either a large-particle composite, or a dispersion- strengthened composite (otherwise known as a dispersion-hardened composite).

The composite material may be in the form of a large particle composite, for example, when the reinforcement phase comprises reinforcement particles having a mean particle size of greater than 0.1 micrometres, or 0.5

micrometres, or 1 micrometre and/or less than 10 micrometres, or 20 micrometres, or 30 micrometres. In a large-particle composite the

reinforcement particles form discrete grains, or regions, of reinforcement which bond to metal grains and reinforce the metal matrix phase. The reinforcement particles of a large-particle composite may advantageously pin the grain boundaries of the metal matrix, impeding the movement of dislocations through the metal matrix, and thus increasing the hardness and yield strength of the composite material. The composite material may form a dispersion-strengthened composite where the reinforcement phase comprises small reinforcement particles, for example particles with a mean particle size of greater than 5 nanometres, or 10 nanometres, or 20 nanometres, and/or less than 80 nanometres, or 90 nanometres, or 100 nanometres. In a dispersion-strengthened composite, the reinforcement particles are dispersed within individual grains of the metal matrix phase, and act to inhibit the movement of dislocations, thus increasing the hardness and yield strength of the composite material. A composite material may be provided, the material comprising a composite material powder consolidated to a solid shape, for example by pressing and sintering, or by additive manufacturing processes such as SLS.

It may be advantageous that a composite comprises a powder of discrete particles in which each particle has a matrix phase that is a metal, or an alloy comprising a metal, selected from the list consisting of titanium, tantalum, aluminium, silicon, vanadium, and titanium aluminide, and a reinforcement phase that is carbon in the form of carbon nanotubes or graphene. Carbon nanotubes with dimensions of between 2.5 and 50 nanometres diameter and between 1 and 30 micrometres in length may preferably be used.

For example, a composite comprising a matrix phase of tantalum or tantalum alloy and a reinforcement phase comprising between 2 volume percent and 5 volume percent of carbon nanotubes may be particularly useful in high heat flux applications.

For example, a composite comprising a matrix phase of titanium or titanium alloy, such as a titanium/aluminium/vanadium alloy, and a reinforcement phase comprising between 2 volume percent and 5 volume percent of carbon nanotubes may be particularly useful in high strength applications.

For example, a composite comprising a matrix phase of titanium aluminide, such as TiAl, and a reinforcement phase comprising between 2 volume percent and 5 volume percent of carbon nanotubes may be particularly useful in high temperature, lightweight applications such as turbochargers.

For example, a composite comprising a matrix phase of silicon, and a reinforcement phase comprising between 2 volume percent and 5 volume percent of carbon nanotubes may be particularly useful as an electrode in battery applications.

It may be advantageous that a composite comprises a powder of discrete particles in which each particle has a matrix phase that is a metal, or an alloy comprising a metal, selected from the list consisting of titanium or aluminium, and a reinforcement phase that is silicon carbide in the form of whiskers or particles. Silicon carbide whiskers preferably have dimensions of between 0.5 and 2 micrometres diameter and between 5 and 30 micrometres in length. For example, a composite comprising a matrix phase of titanium or titanium alloy, such as a titanium/aluminium/vanadium alloy, and a reinforcement phase comprising between 2 volume percent and 5 volume percent of silicon carbide particles may be particularly useful in high strength/ high wear resistance applications such as tool facings.

Specific Embodiments of the Invention

Specific embodiments of the invention will now be described with reference to figures in which:

Figure 1 is a schematic diagram illustrating a high-shear mixing apparatus suitable for use in an embodiment of the invention,

Figure 2 is a schematic diagram illustrating a fluidised-bed spray-granulation apparatus that may be used in embodiments of the invention,

Figure 3 is a schematic diagram illustrating the formation of precursor powder particles embodying the invention using a spray-granulation process, Figure 4 is a schematic diagram illustrating an electrolysis apparatus arranged to reduce a precursor powder according to an embodiment of the invention,

Figure 5 is a schematic cross-sectional view illustrating additional detail of the cathode structure of the electrolysis apparatus of Figure 4,

Figure 6 is a plan view of the cathode illustrated in Figure 5.

The invention relates to a method of forming a composite material. The invention also relates to a precursor powder and a composite material, or reinforced metal material, formed by the method.

The method of forming a powder of a composite material includes the steps of forming a precursor powder, and reducing the precursor powder to form the composite material. As described above, the process of forming the precursor powder may be carried out in a number of ways. The precursor powder may be formed, for example, by mechanically mixing powders of the matrix precursor material and the reinforcement precursor material in a ball mill, or by a sol-gel process, optionally followed by consolidating the precursor powder by heat treatment prior to reduction. As an exemplary embodiment, the process of forming the precursor powder by mixing and fluidised-bed spray-granulation will now be discussed in general terms.

Figure 1 is a schematic illustration of a mixing apparatus 10 that may be suitable for forming the liquid suspension in embodiments of the invention. In the specific embodiment of Figure 1 the mixing apparatus 10 is a high-shear mixing apparatus. A high-shear mixing apparatus may be preferable to a low- shear mixing apparatus when forming some specific precursor powders. The high-shear mixing apparatus 10 comprises a tank 1 1 containing a mixture of liquid and particles 12, the particles consisting of matrix precursor material particles and/or reinforcement precursor material particles. A high-shear mixer 20 is arranged in contact with the liquid mixture 12. The high-shear mixer includes a motor 21 , a connecting shaft 22, a rotor 23 and a stator 24. The rotor 23 is separated from the stator 24 by a narrow gap 25. In use, the motor 21 causes the rotor 23 to rotate at speeds of typically between 5000 rpm and 10000 rpm. As the rotor rotates the stator remains static and the differences in velocity of the liquid in the region of the rotor and the stator result in high-shear within the liquid.

According to an aspect of the present invention, the liquid mixture 12 consists of an aqueous solution of a binder, such as polyvinyl alcohol (PVA), a proportion of particles of matrix precursor particles, and a proportion of particles of reinforcement precursor material. The high-shear forces set up in the liquid mixture 12 cause milling of the matrix precursor particles and reinforcement precursor material particles. Various parameters may be altered to influence the final particle size of the milled particles. For example, parameters such as rotational speed of the rotor 23, distance of the gap 25 between the rotor and the stator, and proportion of particles to liquid in the liquid mixture 12, as well as mixing time, may all be varied to influence the particle size of the matrix precursor material and reinforcement precursor material resulting from the high-shear mixing process. The high-shear mixing forms a suspension of finely milled particles in a liquid, for example, an aqueous solution of PVA.

The relative quantities of matrix precursor material and reinforcement precursor material added to the mixture are chosen so as to determine the composition of the precursor powder formed by spray granulation, and therefore the composition of the composite material obtained following reduction.

A range of suitable high-shear mixers or batch mixers are known and commercially available. For example, IKA ^® manufacture a wide range of batch mixers for forming suspensions of pharmaceutical products. These high-shear mixers may be suitable for use in embodiments of the present invention.

Figure 2 is a schematic illustration of a fluidised-bed spray-granulation apparatus 30. The apparatus includes a heated chamber through which an upwardly directed stream of hot air 32 is passed. A nozzle 33 allows droplets of a liquid suspension 34 to be injected into the heated chamber 31. The droplets of the liquid suspension 34 may be supplied directly from a high-shear mixing apparatus 10 or may be transferred to a separate holding tank prior to injection. Once in the heated chamber 31 the droplets of liquid suspension 34 are dried and form solid particles 35. These solid particles 35 are maintained within the heated chamber 31 by the fluidising action of the heated airstream 32. As further droplets of liquid suspension 34 are injected into the heated chamber 31 they adsorb to existing particles 35 and dry, thereby increasing the diameter of the particles 35. As the particles grow (agglomerate), they eventually reach a mass that is too great for them to remain in a fluidised state within the chamber. Once the particles reach this diameter they drop towards the bottom of the chamber and are collected.

The size, shape, and mass of the collected particles 36 and the size distribution of the collected particles 36 can be influenced by controlling parameters such as particle loading of the liquid suspension, injection pressure and initial droplet size, and flow rate of the fluidising airflow.

The use of spray-granulation technology enables a large degree of flexibility in controlling particle sizes and particle size distributions. Through control and optimisation of the process parameters, particle sizes within a range of 10 micrometres to 10 millimetres may be achieved. The system has the advantage that any undersize particles that have passed through the spray- granulator may be returned to the heated chamber 31 for further growth.

Furthermore, any oversize particles may be returned to the high-shear mixer. Thus, process yields should comfortably exceed 90%, and preferably exceed 95%, or exceed 98%.

Figure 3 is an illustration depicting the growth, or agglomeration, of particles within a fluidised-bed spray-granulation apparatus. Droplets of the liquid suspension 34, once injected into the heated chamber of the spray-granulation apparatus, swiftly dry to form small seed particles 35. As discussed in relation to Figure 2, these seed particles 35 are fluidised by a stream of heated air. Subsequent droplets of the liquid suspension 37 adsorb to the surface of the seed particles 35. These additional liquid droplets swiftly coat the surface of the seed particle and dry to add a layer of thickness to the seed particle. Over a period of time, more and more droplets adsorb to the surface of the fluidised particles and form a layered, onion-like, particle 36. Figure 3 illustrates a cutaway of a fully formed particle 36 showing the layered structure.

Once the particles have reached the predetermined particle size they are collected from the spray-granulation apparatus. It is then preferable that the particles are heat treated in order to drive off any remaining binder from the particles and to provide some mechanical stability. Thus, the collected particles 36 may be subjected to a heat treatment regime. The heat treatment may be a two-step regime comprising, for example, heating to 500°C and holding for a period of time followed by heating to 1000°C and holding for a further period of time.

Precursor powder particles formed according to a spray granulation method of the present invention can be seen by SEM imaging to be substantially the same size and relatively porous. The porosity of the powder may enhance the reduction of the powder using molten salt reduction processes.

After the particles have been collected and, if required, heat treated, the precursor powder may be reduced, so that the oxygen is removed from the matrix precursor material, to form a powder of composite material, containing a matrix phase comprising a first metal, and a reinforcement phase.

Figure 4 illustrates an electrolysis apparatus 110 configured for use in performing a reduction of a precursor powder. The apparatus comprises a stainless steel cathode 120 and a carbon anode 130 situated within a housing 140 of an electrolysis cell. The anode 130 is disposed above, and spatially separated from, the cathode 120. In certain embodiments the housing 140 may contain 500 kg of an electrolyte, for example a calcium chloride based molten salt electrolyte 150, the electrolyte comprising CaC and 0.4 wt % CaO. Both the anode 130 and the cathode 120 are arranged in contact with the molten salt 150. Both the anode 130 and the cathode 120 are coupled to a power supply 160 so that a potential can be applied between the cathode and the anode.

The cathode 120 and the anode 130 are both substantially horizontally oriented, with an upper surface of the cathode 120 facing towards a lower surface of the anode 130.

The cathode 120 incorporates a rim 170 that extends upwards from a perimeter of the cathode and acts as a retaining barrier for a precursor powder 190 supported on an upper surface of the cathode. The rim 170 is integral with, and formed from the same material as, the cathode. In other embodiments, the rim may be formed from a different material to the cathode, for example from an electrically insulating material. The structure of the cathode may be seen in more detail in Figure 5 and Figure 6. The rim 170 is in the form of a hoop having a diameter of 30 cm. A first supporting cross-member 175 extends across a diameter of the rim. The cathode also comprises a mesh-supporting member 171 , which is in the form of a hoop having the same diameter as the rim 170. The mesh-supporting member has a second supporting cross-member 176 of the same dimensions as the supporting cross-member 175 on the rim 170. A mesh 180 is supported by being sandwiched between the rim 170 and the mesh-supporting member 171 (the mesh 180 is shown as the dotted line in Figure 10). The mesh 180 comprises a stainless steel cloth of mesh-size 100 that is held in tension by the rim 170 and the mesh-supporting member. The cross-member 175 is disposed against a lower surface of the mesh 180 and acts to support the mesh. An upper surface of the mesh 180 acts as the upper surface of the cathode.

The stainless steel cloth forming the mesh 180 is fabricated from 30

micrometre thick wires of 304 grade stainless steel that have been woven to form a cloth having square holes with a 150 micrometre aperture. The mesh size may be varied and should, in general, be of smaller diameter than the mean particle diameter of the precursor powder that is being reduced. The mesh 180, cross-member 175 and rim 170 that form the cathode are all electrically conductive. In other embodiments, the mesh may be the only electrically conductive component of the cathode.

In use, the precursor powder may be reduced by applying a potential between the cathode 120 and the anode 130 sufficient to remove oxygen from the matrix precursor material in the precursor powder 190. The composite material remaining after reduction can be removed and washed to separate the composite material from any remaining salt. Example 1

As a first specific example, a silicon-carbide-particle reinforced Ti-AI-V alloy composite was formed according to the present invention. The initial starting oxide powders used to form the precursor powder for reduction were T1O2, AI2O3 and V2O5, and the reinforcement precursor material used was silicon carbide (SiC) particles. The T1O2 oxide powder had a mean particle size of 1.2 micrometres. The AI2O3 oxide powder had a mean particle size of

8.2 micrometres. The V2O5 powder had a mean particle size of

97.7 micrometres. The SiC particles had a mean particle size of 100 nanometres. It would normally be extremely difficult to combine these three oxide powders and the SiC particles in a suitable ratio to form a composite powder with an alloy matrix phase containing low proportions of Al and V, for example a Ti-6AI-4V alloy matrix phase, due to the large disparity between the initial particle sizes of the starting oxides. While large pellets having a desirable ratio of the oxides may be formed, it is clear that the short range composition is unlikely to proximate that required to form a Ti-6AI-4V alloy.

A total of 4800 grams of mixed oxide powder and 96 grams of SiC particles was used. Of the 4800 grams of mixed oxide powder, 4195 grams was T1O2, 402 grams was AI2O3, and 203 grams V2O5. This corresponds to a proportion of metallic elements of 85.7% Ti, 8.2% Al, and 4.1 % V, and a proportion of SiC reinforcing particles of 2%. A precursor of this composition was expected to produce, upon reduction by the FFC process, a matrix alloy of composition approximating Ti-6AI-4V, comprising an even distribution of SiC particle reinforcements. It is noted that the proportion of aluminium in the feedstock powder was deliberately increased above 6% in order to account for losses of aluminium during the reduction process. The oxide powders and SiC particles were mixed with a liquid to form a slurry having 60 % solid phase. The liquid consisted of an aqueous solution of demineralised water and PVA. The proportion of PVA was dependent on the total solid loading. Thus, the proportion of PVA was 2.5 wt % with respect to the total solid loading. The liquid mixture was then subjected to high-shear mixing for 15 minutes at a rotation speed of 6500 rpm. Shear-mixing was achieved using an IKA dispersion mixer model no. G45M.

After shear-mixing the liquid, SiC, and oxide mixture had formed a suspension of milled oxide particles and SiC particles in the liquid/binder. The particle size of the oxides had been substantially homogenised, with the oxide particles in the suspension having diameters in the region of 2 micrometres or less. As the SiC particles had a particle size lower than 2 micrometres prior to high-shear mixing, the particle size of the SiC was not reduced any further by high-shear mixing. This liquid suspension was then subjected to a fluidised-bed spray- granulation process to produce solid metal-oxide-plus-SiC precursor particles. A Glatt Procell Labsystem spray-granulator was used to form the precursor particles. Process parameters were set such that the fluidising airflow through the apparatus was 150 m ³ per hour, the air temperature was 120°C. The liquid suspension was sprayed into the chamber at a spraying pressure of 3 bar and a spray rate of 57 grams per minute. These parameters were selected to provide a mean precursor particle size within the range of 100 and 150 micrometres.

After spray-granulation the collected precursor particles were heat treated. Heat treatment was carried out in order to remove organic components of the PVA binder and to impart mechanical strength to each individual precursor powder particle. In order to heat treat the precursor particles produced by spray granulation they were heated to a temperature of 550°C at a rate of 3°C per minute in an electrically heated furnace. The precursor particles were maintained at this temperature for a period of 1 hour to remove traces of the organic binder. The precursor particles were then heated to 1000°C at 3°C per minute and held at that temperature for a further 2 hours before cooling to room temperature.

The precursor powder was reduced to a reinforced metallic alloy powder using apparatus of the type discussed above in relation to Figure 4. Approximately 20 grams of the precursor powder was arranged on the upper surface of the cathode 20 and in contact with the molten salt 150. The precursor powder 190 was supported by the mesh 180 of the cathode. The depth of the precursor powder 190 was approximately 1 centimetre. The precursor powder was not consolidated prior to reduction.

The cathode potential was measured with reference to a reference electrode (not shown) and potentiostatically controlled throughout the reduction process.

The temperature of the molten salt 50 (CaC and 0.4 wt% CaO), the cathode potential, and the duration of the reduction time were controlled during the reduction process to preferentially reduce the metal oxide in the precursor, and to avoid reaction of the SiC particles.

Thermal currents and gas lift deflects generated by buoyancy of gases generated at the anode (predominantly CO and CO2) cause the molten salt to circulate within the cell and generate a flow of molten salt through the bed of precursor powder. At the end of the reduction time the cell was cooled and the cathode was removed and washed to free salt from the reduced precursor powder.

The reduced precursor powder formed a composite powder of silicon carbide reinforced Ti-6AI-4V alloy, in which titanium, aluminium and vanadium were homogenously distributed and alloyed throughout the metal matrix phase. Silicon carbide particles were distributed evenly throughout the particles of the composite.

This specific example produces a composite having a Ti-6-4 alloy matrix and a SiC reinforcement phase having reinforcement particles of about 100 nm particle size. Clearly, larger SiC particles may be used to increase the size of the reinforcement phase. If very large particles or whiskers of SiC are required it may be necessary to use a two-stage mixing regime. In a first mixing stage the matrix precursor components could be mixed and milled using high-shear, and in a second mixing stage the reinforcement precursor material may be incorporated with low-shear mixing. This prevents milling of the reinforcement.

Example 2 As a second example, a boron carbide reinforced titanium powder was produced according to a method of the present invention. The precursor powder was formed using T1O2 oxide powder having a mean particle size of 1.2 microns, and boron carbide (B ₄ C) particles having a mean diameter of 10 micrometers.

In order to avoid, or reduce, the breaking up of the boron carbide particles during mixing, the precursor powder was prepared from a first liquid suspension containing the T1O2, and a second liquid suspension containing the boron carbide particles. In order to form the first liquid suspension, 287.5g pigment grade Ti02was added to 102.6g of liquid consisting of an aqueous solution of demineralised water and 5% PVA. The mixture was then subjected to high-shear mixing for 15 minutes at a rotation speed of 6500 rpm. The second liquid suspension was formed by adding 25g of boron carbide particles to 10g of 5% PVA solution, following high-shear mixing for 15 minutes at a rotation speed of 6500 rpm to form a suspension.

These liquid suspension were then subjected to a fluidised-bed spray- granulation process to produce solid precursor particles by spraying a portion of the first liquid suspension into a heated chamber of a fluidised-bed spray- granulation apparatus to form a plurality of seed particles, and spraying further portions of the first and second liquid suspensions into the heated chamber, thereby growing particles to form the precursor powder. The spray granulation parameters were substantially as described in relation to the Example 1 , and were selected to provide a mean precursor particle size within the range of 100 and 150 micrometres.

After spray-granulation the collected oxide particles were heat treated to form a precursor powder as described above in relation to Example 1. The precursor powder was then reduced using the reduction process described in relation to Example 1 , such that the T1O2 was reduced to titanium, while the boron carbide particles did not react. The reduced powder was then collected and washed.

The reduced precursor powder formed a reinforced titanium powder comprising a titanium matrix reinforced by boron carbide particles, in which each particle of the reinforced titanium powder comprises titanium metal and boron carbide reinforcements.

Example 3

As a third example, a carbon nanotube reinforced tantalum powder was produced according to the present invention. The precursor powder was formed using tantalum oxide (Ta20s) powder agglomerate having a mean particle size of 300 micrometres, and carbon nanotubes having a mean diameter of 50 nanometres and a mean length of 10 micrometres.

A total of 1000 grams of tantalum oxide powder was used, together with 20 grams of carbon nanotubes, such that the initial powder comprised

approximately 98 % Ta20s and 2 % carbon nanotubes.

The carbon nanotubes and oxide powders were mixed with an aqueous solution of demineralised water and PVA to form a slurry having 59.5 % solid oxide and 1.2 % carbon. The slurry was then subjected to high-shear mixing and spray granulation as described in relation to Example 1. The spray granulation parameters were selected to provide a particle size distribution within the range of 100 and 150 micrometres. The product of spray granulation was a free-flowing powder of grey granules. After spray-granulation the collected particles were heat treated to form a precursor powder as described above in relation to Example 1. The precursor powder was then reduced using the FFC process using the apparatus described in relation to Example 1. During reduction the molten salt 50 (CaC and 0.4 wt% CaO) was maintained at a temperature of 800°C and a potential was applied between the anode and the cathode.

During reduction of the precursor powder, the cathode potential is controlled such that tantalum oxide is reduced to tantalum metal.

Following reduction, the reduced powder was collected and washed. The powder produced was a reinforced tantalum powder in which each particle contained carbon nanotube reinforcements dispersed within a tantalum matrix.