Title of Invention	"A METHOD OF MANUCATURE OF A COBALT CATALYST"
Abstract	A method of manufacturing a catalyst, comprising the steps of (i) forming an aqueous solution of a cobalt ammine carbonate complex, (ii) oxidising said solution by adding a chemical oxidant to the solution such that the concentration of Co(III) in the oxidised solution is greater than the concentration of Co(III) in the un- oxidised solution, (iii) decomposing the cobalt ammine complex by heating the solution to a temperature between 80 and 110 °C for sufficient time to allow an insoluble cobalt compound to precipitate out of the solution, (iv) filtering the precipitated cobalt species from the solution and (v) drying the precipitated cobalt species wherein prior to decomposing the cobalt ammine complex, the solution of cobalt ammine complex is oxidised by adding a chemical oxidant selected from a solution of hydrogen peroxide or a hypochlorite to the solution sufficient to convert from 50-90% of the cobalt calculated as moles of cobalt from Co(II) to Co(IH) such that in the precipitated cobalt species, after drying at a temperature less than 160°C, the ration of cobalt hydrotalcite phase : cobalt spinel is less than 0.6:1, said cobalt hydrotalcite phase and said cobalt spinel phase being measured by X-ray diffractometry and wherein the cobalt spinel comprises Co3O4.

Title of Invention

"A METHOD OF MANUCATURE OF A COBALT CATALYST"

Abstract

A method of manufacturing a catalyst, comprising the steps of (i) forming an aqueous solution of a cobalt ammine carbonate complex, (ii) oxidising said solution by adding a chemical oxidant to the solution such that the concentration of Co(III) in the oxidised solution is greater than the concentration of Co(III) in the un- oxidised solution, (iii) decomposing the cobalt ammine complex by heating the solution to a temperature between 80 and 110 °C for sufficient time to allow an insoluble cobalt compound to precipitate out of the solution, (iv) filtering the precipitated cobalt species from the solution and (v) drying the precipitated cobalt species wherein prior to decomposing the cobalt ammine complex, the solution of cobalt ammine complex is oxidised by adding a chemical oxidant selected from a solution of hydrogen peroxide or a hypochlorite to the solution sufficient to convert from 50-90% of the cobalt calculated as moles of cobalt from Co(II) to Co(IH) such that in the precipitated cobalt species, after drying at a temperature less than 160°C, the ration of cobalt hydrotalcite phase : cobalt spinel is less than 0.6:1, said cobalt hydrotalcite phase and said cobalt spinel phase being measured by X-ray diffractometry and wherein the cobalt spinel comprises Co3O4.

Full Text	Catajvsts This invention relates to catalysts and ir particular to catalysts containing cobali which: are sj ft able: for use in hydrogenation reactions. Catalysis comprising cobalt on s support such as silica or alumina are known in the art for hydrogenation reactions, e.g. for the hydrogenation of aldehydes and nrtriies and for the preparation of hydrocarbons from synthesis gas via the Fischer-Tropsch reaction, in comparison with other catalytic metals such as copper and nickel used for hydrogenation reactions, cobalt is a relatively expensive and so. to obtain the optimum activity, it is desirable that as much as possible of the cobalt present is in an active form accessible to the reactanis. For hydrogenation reactions, the active form of the oobait is elemental cobalt although in the active cataiyst only some, rather than ail, of the cobalt is normally reduced to the elemental form. Hence s useful measureJs the exposed surface area of elemental cobalt per g of total cobalt present. Except where expressly indicated, as used herein, total cobalt contents are expressed as parts by weight of cobalt (calculated as cobalt metal, whether the cobalt is actually present as the metal or is in a combined form, e.g. as cobalt oxides) per 100 parts by weight of the catalyst or precursor thereto. Cobalt catalysts on different carriers are disclosed in "Stoichiometries of H2 and CO Adsorptions on cobalt", Journal of Catalysis 85, pages 63-77 (1984) at page 67, table 1. From the total maximum H2 uptake, it is possible to calculate the cobalt surface area per gram of catalyst and the cobalt surface area per gram of cobalt. US 5 874 381 describes a cobalt or alumina catalyst which contains between 3 and 40% by weight of cobaft and wnich has a relatively high cobalt surface area of above 30 m2/g of total cobalt. As indicated above, the dispersion of the cobalt on the carrier is important since it is the surface of the cobali of the catalyst which is ca'afyiically active. Therefore it is beneficial to maximise the surface area of the metal which is present so as to produce a catalyst which has a high cobalt surface area per unit mass of lotai cobalt. It may be expected that the dispersion of the cobalt on the catalyst would be maximised at relatively low loadings of cobalt and that, as the amount of cobalt contained in the catalyst is increased, the surface area per gram of cobalt wouid decrease because the cobalt becomes more difficult to disperse on the support. The aforementioned US 5 874 381 suggests and exemplifies the production of the catalysts by impregnation of shaped transition alumina particles, e.g. extrudates, with a solution of cobalt ammine carbonate, followed by removal of the excess solution and heating to decompose the cobalt amrnine carbona-.e. We have found that the preparation of cobalt catalysts by the decomposition of cobalt ammine carbonate may be improved. Accordingly the invention provides a method of manufacturing a catalyst, or precursor thereto, comprising the steps of forming an aqueous solution of a cobalt ammine complex, allowing the solution to oxidise such that the concentration of .Co(HI) in the oxidised solution is greater than the concentration of Coflll) in the un-axidised solution, and then decomposing the cobalt ammine complex by heating the solution to a temperature between 80 and 110 "C for sufficient time to allow an insoluble cobalt compound to precipitate out of the solution. We have found that when the cobalt ammine complex solution is allowed to oxidise so that at least some of the coba!t (I!) in the complex is converted to Co(fll), the composition of the insoluble cobalt compound resulting from the decomposition of the cobait ammine complex is readily reducible to cobait metal of high surface area. Without wishing to be bound by theory, we believe that the cobalt species which is precipitated from the solution of complex, e.g. a cobait ammine carbonate complex, comprising Co(lll) species contains a greater amount of Co3C>4 and less cobalt carbonate, cobalt hydroxycarbonate or ammonia than is deposited from the decomposition of a freshly made solution containing more Co(ll) and less Co(lll). Preferred complexes are cobalt ammine carbonate complexes although other compounds may also be used, e.g. formates. The term "cobalt species" is used broadly to include both elemental cobalt and cobalt in combined form, e.g. as compounds such as cobalt oxides and cobalt hydroxycarbonates. The catalyst in its reduced form is useful for catalysing hydrogenation reactions. The catalyst may, however, be provided as a precursor wherein the cobalt is present as one or more compounds, such as oxides or hydroxy carbonates, reducible to elemental cobalt. In this form, the material may be a catalyst precursor and may be treated to reduce the cobalt compounds to metallic cobalt or the material may Itself be a catalyst and used as supplied, e,g, for oxidation reactions. The cobalt surface area figures used herein apply to the material after reduction, but the invention is not limited to the provision of reduced catalyst. Preferably the cobalt ammine complex solution is heated to decompose the complex in the presence of a catalyst support materlai which is selected from standard known supports such as silica (including both synthetic silica and naturally occurring forms of silica such as kieselguhr), alumina, silica-alumina, liiania, zirconia, carbon, coated silicas or aluminas such as titanJa- or zirconia- coated silicas or aluminas for example. The cata'ysi of the invention is particularly suitable for use in Fisclier-Tropsch (F-T) hydrocarbon synthesis and the supports preferred for cobalt catalysts for use sn known cobalt F-T catalysts may be advantageously usec for the- catalysts of the present invention. Preferably ar: alumina support is oresenl, which is most preferably a transition alumina, e.g a gamma, trieta or delta alumina, so that preferred catalysts according to the invention comprise a cobalt species on e transition alumina-support. The support may be in the form of a powder or of E fabricated unit such as a granule, tablet or extrudate. Fabricated units may be in the form.o* elongated cylinders, spheres, lobed shapes or irreaularlv shaped particles, all of which are known in the art of catalyst manufacture. Alternatively the support may be in the form of a coating upon a structure such as a reactor tube wall, honeycomb support, monolith etc. Support materials may contain promoters or other materials such ss binders and may be treated prior to use in the process of the invention, e.g. by drying or calcining. Suitable transition aiurnina may be of the gamma-alumina group, for example a eta-alumina or chi-alumina. These materials may be formed by calcination of aluminium hydroxides si 400 to 750eC and generally have a BET surface area in (he range 150 to 400 nrr/g. Alternatively, the transition alumina may be of the delta-alumina group which includes the high temperature forms such as delta- and theta- aluminas which may be formed by heating a gamma group afumins to z temperature above about SOO°C. The delta-group aluminas generally have a BET surface area in the range 50 to 150 m2/g. The transition aluminas contain less than 0.5 mole of water per-mole of A!2O3, the actual amount of water depending on the temperature to which they have been heated. Alternatively, we have found that suitable catalyst supports may comprise an alpha-alumina. A suitable powder for the catalyst support generally has a surface-weighted mean diameter D[3,2I in the range 1 to 200 um. In certain applications such as for catalysts intended for use in slurry reactions, it is advantageous to use very fine particles which have a surface-weighted mean diameter D[3,2] on average, in the range from 1 to 20 prn, e.g. 1 to 10 (jm. For other applications e.g. as a catalyst for reactions earned out in a fluidised bed, it may be desirable to use larger particle sizes, preferably in the range 50 to 150 um. The term surface-weighted mean diameter D[3>2], otherwise termed ihe Sauter mean diameter, is defined by M. Alderliesten in (he paper "A Nomenclature for Mean Particle Diameters"; Anal. Proc.. vol 21, May 1984, pages 167-172, and is calculated from the particle size analysis which may conveniently be effected by laser diffraction for example using a Malvern Mastersizer. n is preferred (hat a powder support has a relatively large average pore diameter as the use of such supports appears to give catalysts of particularly good selectivity. Preferred supports have an average pore diameter (APD) of at least 10 nm, particularly in the range 15 to 30 nm. (By the term average pore diameter we mean 4 times the pore volume as measured from the riesorption branch of the nitrogen physisorption isotherm at 0.98 relative pressure divided by the BET surface area]. During the production of the compositions of the invention, cobalt compounds are deposited in the pores of (he support, and so the average pore diameter of the composition will be iess than that of the support employed, and decreases as the proportion of cobalt increases. !? is preferred that the catalysts have an average pore diameter of at least 8 nm, preferably above 10 nm and particufarly in the range 10 to 25 nm. When the support is transition alumina, it has been found that, depending on the conditions used, the bulk of the cobalt is .precioitated as cobalt compounds within the pores of the transition alumina and none or only a small proportion of the cobalt is deposited as a coating round the alumina particles. As a result, irrespective of the cobalt content of the composition, the particle size of the compositions of the invention is essentially the same as the particle size of the support, and so the compositions of the invention generally have & surface-weighted mean diameter D[3,2] in the range 1 to 200 urn, in one embodiment preferably less than 100 jjm and particularly iess than 20 urn, e.g. 10 urn or less, and in a second embodiment preferably in the range 50 to 150 urn. On the other hand, since the cobait compounds are primarily precipitated within the pores of the support, the pore volume of the compositions in accordance with the invention will be less than that of the support employed, and will tend to decrease as the cobalt species loading increases. Compositions having a total cobalt content less than 30% by weight preferably have a pore volume of at least 0.5 mi/g whiie compositions having a total cobalt content above 30% by weight, particularly above 40% by weight, preferably have a pore volume of at least 0.3 ml/g, particularly at least 0.4 rn!/g. The compositions of the invention, when in the reduced state, have a cobait surface area of at least 25 rrr/g of cobalt as measured by the H2 chemisorption technique described herein. Preferably the cobalt surface area is greater than 30, more preferably at least 40, especially at least 60 m'/g. The cobalt surface area tends to decrease as higher loadings of cobai! are used, but we have found that when the composition contains 50 to 60% by weight total cobalt in the reduced state, the cobalt surface area achievable is about 80 rrr/g or more. The cobalt surface area is determined by H; chemisorption. This method is used in the Examples, and when a cobalt surface area measurement is mentioned in this specification for the catalysts of the invention (unless otherwise specified). Approximately 0.2 to 0,5 g of sample material is firstly degassed and dried by heating to 140"C at 10"C/min in flowing helium and maintaining 3t 140"C for 60 minutes. The degassed and dried sample is then reduced by heating it from 140"C to 425°C at a rate of 3°C /min under a 50 mf/rnin .flow of hydrogen and then maintaining the hydrogen flow at 425C for 6 hours. Following this reduction, the sample is heated under vacuum to 450°C at 10°C/min and held under these conditions for 2 hours. The sample is then cooled to 15Q~C and maintained for a further 30 minutes under vacuum. The chsmisorptior analysis is then carried out at 15D=C using purs hydrogen gss. An tiLiiornaiic; analysis program is used to measure z full isotherm overtne range 100 rum Hg up »o 760 irtm Hq pressure of hydrogen. Tsie analysis is carried out twice: the firs' measures the "tcjtal" hydrogen uptake (i.e. includes chemisorbed hydrogen and pnysisorbed Hydrogen; and immediately following the first analysis the sample is put under vacuum ( Cobalt surface areas were calculated in ail cases using the following equation; Co surface area = ( 6.023 x 1023 x V x SF x A } / 22414 where V ~ uptake of H2 in ml/g SF - S.loiehiometry factor (assumed 2 for H; cherr.isorption on Co) A = area occupied by one atom of cobalt (assumed 0.0662 nm::) This equation is described in the Operators Manual for the Micromeretics ASAP 2010 Chemi System V 2.01, Appendix C, Part No. 201-42808-01, October 1996. The cobalt ammine complex Is most preferably a cobalt ammine carbonate complex which is formed in situ in aqueous solution by dissolving basic cobalt carbonate in a solution of ammonium carbonate in aqueous ammonium hydroxide, to give a product of the desired cobali content. Alternatively other cobalt salts may be used, including organic salts such as cobait acetate or cobalt formate. A cobalt arnrnine carbonate complex is the product of dissolving basic cobalt carbonate, preferably of empirical formula Co(OH)z.2x(CO3)x in a solution of ammonium carbonate in aqueous ammonium hydroxide, to give a product of the desired cobalt content. The cobali ammtne carbonate solution may be made by dissolving basic cobalt carbonate in an aqueous sofution of ammonium carbonate or ammonium carbamate containing additional ammonium hydroxide. The relative amounts should be such that the pH of the solution is in the ranqe 75 Io12, preferably 9 to 12. The solution preferably contains 0.1 to 2.5 moles of the cobalt complex per litre. As the concentration of cobalt increases, then generally the proportion of carbonate ions reiative to hydroxide ions in the basic cobalt carbonate feed snouSd be increased. Additional ammonium hydroxide solution may be added in order to provide a slurry of handleable viscosity when support particles are mixed in. As an alternative, the cobait ammine carbonate solution may be made by dissolving metallic cobait, preferably in powdered form, in aqueous ammonia of pH 11 -12. in the presence of oxygen or air, either with addition of ammonium carbonate or with addition of CO? gas. The amount of cobalt in the catalyst may tae varied by varying the relative amount of cobalt and support presem in the reaction mixture and by controlling the concentration of the solution of cobalt compound In accordance with the method of the invention, the complex solution is then allowed to oxidise either by ageing in contact with air or an oxygen-rich gas, or by chemical or electrochemical oxidation, in order that the Co(ll) complex is converted, at least in part to a Co(lll) complex. The ageing may 5s accomplished by allowing the solution to stand in an uncovered container for the required time, preferably with stirring. The ageing by stirring in the presence of oxygen should be continued for at least 3 hours and preferably for at least 16 hours. Alternatively the solution may be oxidised by bubbling an oxygen-containing gas stream, e.g. air or oxygen, through the solution, optionally with stirring and in this case ageing may be sufficient after just one hour. Alternative methods of ageing the complex include adding an oxidising agent such as hydrogen peroxide, hypochlorite or by electrolytic ageing methods. The amount of chemical oxidant added to the solution is preferably sufficient to convert from 40% to 100% of the cobait in the unoxidised solution, more preferably from 50 •- 90 % and especially from 60 - 90% of the cobalt, calculated as moles of cobalt and assuming that oxidation from Co2 to Co4* is stoichiometric. For example, where 0,65 moles of hydrogen peroxide is used to oxidise a solution containing 1.7 moles of Co, the conversion of Co2* to Co3" is 76.5%, assuming that one mole of peroxide oxidises two moies of cobalt and that the solution contains Co(il) initially. The amount of conversion of Co24 to Co3+ is likely to increase as the temperature is raised. Therefore the oxidation process may be accelerated by warming the solution but the ageing is normally done at room temperature or slightly above room temperature, e.g. frorri about 18°C to about 35°C. We have found that the oxidation of the aqueous solution of cobait ammine complex is produces an increase in net absorbance of radiation at Amax of the UV/Visibie spectrum occurring between 450 and SOOnm. A,MX represents the height of the peak occurring between 450 and 600nm and is measured in absorbance units relative to an interpolated baseline. The absorbance in this region increases as the extent of oxidation is increased, up to a maximum absorbance when the solution is fully oxidised. In s preferred method of .the invention, the solution of cobalt ammine complex is oxidised until the absorbance at Ariax of the UV/vtsible spectrum occurring between 450 and GOOnm is greater than 35% of the absorbance at A,TO). of a fully oxidised solution. More preferably the absorbence at Am£1), of the UV/visib!e spectrum occurring between 450 and GOOnm is greater than 60%, most preferably greater than 90%, and especially greater than 95%, of the atasorbance at Amax of a fully oxidised solution. The standard measurement conditions utilise a Xenon light source (single beam), a path length of irnrr .-rx: ssmpie temperature of 20 VC - 25 "C « he sample of cobalt ammine complex solirjon is diluted prior to soecirornerry by adding 1 pan of solution to 4 parts 0" & diluent consisting of 3 parts by volume of 30% aqueous ammonia solution to 7 pcrts demineralisec! water. The dJajeni is used as the blank sample in trie UV/visible spectrometry. As a further indication of the oxidation required, we have found that the redox potential of a solution of un-oxidisad cobalt ammine complex containing about 3% cobalt is approximately -SOOrnV a* ambient temperature. We have found that the oxidation of the complex is sufficient her the method of the invention when the redox potential is between 0 V and -200mV, more preverablv from -50 to -150 mV, and mast preferably about -lOOm'v, e.g. from -90 to -130 mV, Although) the redox potential would be expected to vary with the concentration of cobalt in the soiut'on we have found that for cobalt ammine carbonate solutions containing between about 2% cobalt and about 18% cobalt, the redox potentials of trie fresh and fully oxidised solutions vary by less than 5% with concentration over the range of concentration. As a still further indicator for performing the method of the invention, we have found that a sufficiently oxidised solution produces a pink solution when 0.2-- 0.5 ml (i.e. 6 drops) of solution is introduced into 60ml of deionised water at room temperature. Preferably there is no or little precipitation during this test. Supported cobalt catalysts may be made by impregnating a solid support in the form of a powder or a fabricated unit with a solution of the oxidised cobalt ammine carbonate complex. e.g. by spraying (he support with a measured volume of the solution or by dipping the support into a volume of the solution. The impregnated support is then separated from any supernatant or excess solution and dried at a. temperature in the range 60 to 110°C, so that the cobalt complex decomposes to deposit solid cobalt species upon and in the pores of the support. The impregnation and drying may be repeated several times, e.g. up to about five times, depending on the concentration of the solution and the desired cobalt loading of the support. Supported cobalt catalysts and precursors may also be made by slu/rying the powdered support, e.g. transition alumina powder, with the appropriate amount of the oxidised aqueous solution of a cobalt ammine carbonate complex. The slurry is then heated, e.g. to a temperature in the range 60 to 110°C, to cause the cobalt ammine complex to decompose with the evolution of ammonia and carbon dioxide and to deposit an insoluble cobatt compound on She surface, and in the pores, of the support. The support carrying the deposited cobalt compound is then filtered from the aqueous medium and dried. The procedure may be repeated, i.e. the dried product may be re-slurried in a, solution of the cobalt ammine complex, heated, filtered and dried, if required to increase the cobalt content of the product. Usinq this form of the process of the invention, a catalyst having a high cobalt dispersion and a high coba'f loading, e.g. > 10% cobalt, (more preferably > 15% cobalt, by weight) may be prepared in a single deposition step. The time allowed for the precipitation of the cobalt compound is normally about 30 to 200 minutes; the precipitation is usually complete after about 60 to 80 minutes, but the heating of the slurry may be prolonged to include a further precipitate-ageing step. We have found that when the cobalt content is relatively low, e.g. up to about 40% by weight, it is beneficial to use relatively short process times, e.g. by limiting the total heating time. i.e. for both the precipitation and any precipitate-ageing to 200 minutes or less, preferably 'ess than 150 minutes. As the cobalt content of the catalyst is increased, longer process times may be used, e.g. up to about 350 minutes. When the precipitated cobalt compound contains a mixture of Co2* and Co3* a cobalt hydrotateite species may be farmed. The hydrotaicite species has been identified by X-ray diffractometry (XRD) in which the hydrotaicite phase shows a similar diffraction pattern to known Ki/Co hydrotaicite phases. The catalysts of the invention may be distinguished by the ratio of cobalt hydrotaicite to cobalt spinel found in the precipitated cobalt species after drying at less than 160°C i.e. at a temperature iess than the temperature at which Co^Clt is formed from cobalt nitrate by calcination. The cobalt hydrotaicite may.be represented as [Co2+/Co3+](OH)CO3 in which the ratio of Co2+ : Co3+ is about 3:1. The cobalt spinel has an empirical formula of Co3O4 so contains more Co3" than Co2, The amount of hydrotaicite and cobalt spinel may be determined by XRD. Preferred catalysts of the invention have a ratio of cobalt hydrotaicite ; cobalt spinel of less than 0.6:1, mere preferably iess than 0.5:1, especially less than 0.3:1. The cobalt spinel is calculated from the powder diffraction peaks 00304111 at 19.0"28 (=4.587A) and Co3CU 311 at 36.845°28 (=2.4374,4). The cobalt hydrotaicite is estimated from the diffraction pattern in the patterns published by the International Centre for Diffraction Data JCDD No 00-040-0216 for cobalt nickel carbonate hydroxide hydrate Ni 076 Co 0.25 (CO3}0.-.2o (OH)2 -0.38H2O using the ICDD No 00-040-0216 for cobalt nickel carbonate hydroxide hydrate Ni O.rs Co 0.2s (CO3)ai25 (OHV> 'O.SSHaO using the 7.628A (11.591°29}, 3.84,4(23,143°29). 2.565A(34,952°2B}, 2.285A(39.401°2e), 1.936A(46.89"28), 1.734A(52,74-7A28), 1.639A(56.065"26), 1.521 A (60.853°28) signals. The cobalt hydrotaicite is most particularly identified using the 7.628A (11.591°29) and 3.84A(23.143e28). The ratio of cobalt hydrotaicite : cobalt spinel is estimated from the ratio of the peak areas. Note that there was no nickel present In the catalysts, the Co/Ni hydrotaicite diffraction pattern was used only to approximate that for the Co^Co3" hydrotaicite. We have found that when the cobalt species is precipitated in the spinel form compared with the hydrotaicite form, the dispersion of the cobalt, and therefore the cobalt metal surface area is greater whan the catalyst is reduced in hydrogen to convert the cobalt compounds into metallic cobalt. It is therefore preferred to maximise the amount of cobalt compound precipitant; in spine! form. In contrast catalysts prepared by impregnation of cooa't nitrate OHIO a support always deposit cobalt as Co' because of tho acidic nature of the cobalt nitrate solution. Therefore no hycirotaicite or spinel is formed and the dried cobalt nitrate must be calcined to form the oxide before reduction in hydrogen. According to a further aspect of the invention we provide a catalyst intermediate comprising a cobalt compound, said cobal compound comprising a Coil!).'Co(!)I) hydrotalcite phase and & Gc-iO/ cobalt spinei phase, wherein the ratio of cobalt hydrotaicite phase : cobalt spinel phase is less than 0.6:1, the amount of said cobalt hydrotalcite phase and said cobalt spine! phase being measured by X-ray diftractometry. The catalyst intermediate may be used as s catalyst but normally is subjected to a further process such as the reduction of the cobalt species in a hydrogen-containing gas to provide a catalyst comprising metallic cobalt. The catalyst intermediate may be obtained using the method of the invention. The catalyst intermediate preferably comprises a support as described above. For some applications it may be desired to incorporate modifiers, such as other metals or compounds thereof, into the catalyst or precursor. This may be effected by Impregnating the dried product with a solution of a compound of the desired modifier that decomposes to the oxide or elemental form upon heating. Examples of such modifiers include alkali metals, precious metafs, and transition metals. Common promoters used in cobalt catalysts for Fischer Tropsch processes include manganese, platinum, ruthenium'and rhenium. If desired, the product may be calcined in air, e,g. at a temperature in the range 200 to 600°C, more preferably 200 to 450°C, to decompose the deposited cobalt compound to cobalt oxide. However we have found that using the method of the invention a .significant part of the cobalt species formed by the decomposition of the cobalt ammine complex is Co3O. The composition may be used in its oxidic siafe, i.e. without reducing the cobalt oxides to metallic cobaft. It may be used as a catalyst in this state for e.g. oxidation reactions. Alternatively and preferably, the catalyst is reduced to an active catalyst containing cobalt metal by the end-user. The composition may alternatively be supplied as a reduced catalyst which has been passivated, so that the cobalt metal is protected from daactivation during storage and transportation. Thus a precursor comprising the support and the unreduced cobsli compound, possibly dispersed in a carrier, may be charged to s hydrogenaiion reactor optionally with the material to be hydroganated and the mixture heated while hydrogen is sparged through the mixture. The catalysts may be used for hydrogenation reactions such as the hydrogenation of oiefinic compounds, e.g. waxes, nitro or nitriie compounds, e.g. the conversion of nitrobenzene to anitine or the conversion of nitriles to amines. They may also be used for the hydrogenation of paraffin waxes to remove traces of unsaturation therein. They may also be useful in s wide range of other reactions, for example the Fischer-Tropsch process, i.e. where hydrogen and carbon monoxide are reacted in the presence of the catalyst to form higher hydrocarbons. This may be part of an overall process for the conversion of natural gas to petroleum compounds wherein the hydrogen / carbon monoxide gas mixture is a synthesis gas formed by steam reforming natural gas. The invention will be further described in the following experimental examples and the accompanying drawings, which are:- Fig 1 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 1. Fig 2 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 2. Fig 3 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 3. Fie 4 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 4. Fig 5: UV-visibie spectrogra'ph of solutions produced in Example 6. Fig 6 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 7. Fig 7 ; Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 8. Fig 8 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 9 & 10. Genera; method pf catalyst preparation for ..Examjjjes 1 _^ 1j3 The cobalt amrnine carbonate complex solution was made up using 1707 g ammonia solution (SG 0.89, 30% ammonia), 198 g ammonium carbonate, 218 g basic cobalt carbonate (46.5% wt% Co, bulk density 0,640 g/ml) and 1877 g demineralised water. The solution was stirred for one hour in order to dissolve the basic cobalt carbonate, The total volume of solution was 4 Hires. The solution was oxidised as describee in the following examples before using (o make rooai: catalysts as oescnbed below, The alumina employed was Purabx™ HP14/150 available from SasoL which is a transittcn-aiumina of the garnma alumina type having a particle size D.x of 4S - 50 prn. The alumina was used as received. The alumina particles and a measured amount of the cobalt ammine carbonate complex soiution were charged to a stirred vessel equipped with a condenser. The relative amounts of alumina and cobalt arnmine carbonate complex solution were calculated to provide a catalyst containing 40% by weight of cobalt mela! in the oxide catalyst. The pH of the aqueous solution was 11.1 The mixture was heated to boiling while stirring and gentle boiling at about QB'C to about 1QCTC was maintained until a pH of between 7.5 and 7,7 was achieved, during which the solution becomes dear. The solid was then filtered off, briefly washed in water and then dried inar at 105CC overnight. The cobalt surface area of the catalysts was measured by reduction in a flowing hydrogen stream at 4250 followed by H; chemisorption at 15Q"C using the standard method described earlier Temperature programmed reduction (TPR) of the dried materials was carried out in a reduction gas stream comprising 5% hydrogen in nitrogen. The sample (between 0.1 and 0.15 g, accurately weighed) was first healed to 120 °C (at 5'C /mm under a reduction gas flow of 25ml/mirt) to remove moisture and heid at 120 "C for 45 minutes. Thereafter the sample was heated from 120 to 1000"C in the reduction gas stream at .a heating rate of 5°C per minute The change in concentration of hydrogen between the inlet gas and the outlet aas was monitored by a katharometer to show the consumption of hydrogen at each temperature. The TPR instrument used was Quantachrome ChernBei© TPR/TPD analyser. The TPR traces are shown in Figs 1 1o 4 and 6 to 8. The cobalt surface area was measured according to the standard hydrogen chemisorption procedure described above. UV-vis spectra were acquired at ambient temperature using a Varian Gary 50 spectropriotorneier, equipped with a Xenon flash lamp; using-a ceil path isngth of 1mm. The sample of cobalt arnmine solution is diluted prior ID spectrometry by adding 1 pan of soiution to 4 Darts of a diluent consisting of 3 pa-is by volume of 30% aqueous ammonia solution to 7 parts dernineralised water. The diluent is used as the blank sample in the UV/visibie spectroinetry. (comparative) The r.obait ammine carbonate solution was used immediately after dissolution of the basic cobalt carbonate solids was complete. The coiaalt crystallite see in the catalyst was measured using X-ray diffraction during an in-situ reduction procedure, The cobalt metal crystallites formed at reduction temperatures between about 400°C and about 6QO°C had an average crystallite size of 6 - 10 nm, based upon the 200 reflection. Example 2 The cobalt ammine carbonate solution was oxidised with stirring in contact with air for 3 hours prior to preparation of the catalyst. Example 3 The cobalt ammine carbonate solution was oxidised with stirring in contact with air for 16 hours and then, stored without stirring In a container for 48 hours, allowing air to enter the container periodically throughout the storage time. The cobalt metal crystalfites formed at reduction temperatures between about 400 degrees and about 600 degrees had an average crystallite size of 4 - 5 nm, based upon the 200 reflection. Example 4 The cobalt ammine carbonate solution was stirred whilst air was continuously bubbled through the solution for one hour prior to preparation of the catalyst. The TPR traces for Examples 1 -2 (in figs 1 - 2) show a reduction taking place between about 350°C and 400°C. This reduction is absent from the trace measured for Example 3 {fig 3) which has the highest cobalt surface area, in Fig 4 some reduction at this temperature appears to take piace, visible as a shoulder on the 30Q°C reduction peak. Reduction between about 350°C and 40CTC is currently believed to be associated with reduction of a cobalt hydrotafcife phase . Example S The cobalt ammine carbonate solution was stirred for one hour and air was bubbled through the solution for 5 minutes every 15 minutes prior to preparation of the catalyst. __ ._ _____ __ j Example f Ageing of sofution j Cobalt rretaf surface _____ None Stir 18 hrs. store 48 hours ! 57,9 Constant air purge 1 hour jtterit air purge 1_ hour ! _52.' Example 5 A cobalt ammine carbonate solution was made up as describee eariie-. One portion, desionateti 5s, was stirred in air-far 3 hours. A second portion,6b: was stirred for 16 hours. A third portion. 6c, was allowed to stand, with occasional air ingress, for 30 days. The solutions « were analysed using UV-visible spectroscopy. The spectra are shown in Fig 5. It can be seen that the portion of the spectrum between about 450 nm and 600 nrn shows an increase in absorbance as the amount of oxidation is increased. Example 7 To a 4 litre baicn of coaalt ammtne carbonate solution made according to the general process given above was added approximately 50 ~ 60 mi of 30% .hydrogen peroxide dropwise with ' stirring until 6 drops (about 0.5 ml) of the solution mixed in 60ml deminerafised water resulted in a pink solution with only "minimal precipitation, indicating that sufficient oxidation of the solution had occurred to give mainly Co3"". The same test on insufficiently oxidised solution results in a blue colour solution and a precipitate. The solution was filtered and catalysts were prepared according to the method given above and the properties were measured using the same methods. The results are shown in Table 2. The TPR trace for the catalyst is shown in • Fig 6. Example 8 A 4-litre batch of cobalt ammine carbonate solution made according to the general process was left standing in static air for 8 days. The solution was filtered and catalysts were prepared according to the method given above and the properties were measured using the same methods. The TPR trace is shown as Fig 7. The results are shown in Table 2. ExarnD:[e_9 A 32 litre batch of cofaait ammine carbonate solution was made using the genera! method above using ingredients in the same proportions as that given for the general method above, The solution was filtered and then a 30% hydrogen peroxide solution was pumped into the solution until the Redox potential changed from approximately -300 (before HjOj addition) to 100 mV. The Redox potential was measured using a Mettler Toledo pH transmitter 2500. Catalysts were prepared according to the method given above and the properties were measured using methods described above. The TPR trace is shown in Fig 8. The results are shown in Table 2 Example 10 (comparison] A second 32 litre batch of cobalt ammine carbonate solution was made and filtered in the same way as described in Example S, Catalysts were made by the general method ai'ven without waiting for the solution to oxidise or adding hydrogen peroxide or other oxidant. The results are shown in Table 2. The TPR trace is shown in Fig 8. Table 2 j Example i Oxidation, of solution Cobalt metal surface area (rrf /g Co) % w/w Co (oxidic) ! 7 " H2O2 to colour change 54.2 41 .8 8 Static air ag_eing_8 days 61.9 35.9 9 H2O2 to -100mV Redox potential , 59.1 42,8 10 None (comparison) 43.5 36.9 Example 11 (comparative) Preparation of-cobalt catalysts supported on alumina extrudates Preparation of cobalt amrnine carbonate solution;- Ammoriiurn carbonate chip (634g), (30-34 w/w% WH3), was weighed into a 3 litre round bottomed flask and ammonia solution i~30%) (1880ml) BDH 'Analar1 Sp.Gr. 0.89 added, The mixture was continually stirred overnight to ensure that the ammonium carbonate chip had dissolved. Cobalt basic carbonate (528g), {45-47 w/w% Co), was added in aliquots over 60 minutes. The final solution was stirred continually for approximately 30 minutes before filtering to remove any traces of insoluble particulate matter. Solution pH was10.95 and Co content was 14.3%. The solution was used immediately after filtering to carry out three impregnations onto an alumina extrudate support. Alumina extrudates of 12mm diameter were calcined at 1050 C for 2 hours. 200g alumina extrudates were placed in a 2 litre round bottom flask. An amount of the prepared cobalt ammine carbonate solution was added to the extrudates such that the extrudates were completely covered. The mixture was swirled occasionally for 5 minutes before-decanting off excess solution. The impregnated extrudates were filtered to remove any remaining solution then dried at'110°C for 2 -4 hours. After drying the impregnation procedure was repeated twice more and the thrice-impregnated extrudates were dried for 16 hours at 110 °C. Exa.rnpjeJ2 A portion of (he solution made in Example 11 was oxidised by allowing it to stand with access to air for 48 hours. 200g of alumina extrudates were impregnated using the method of Example 11 for one impregnation only, After filtration, the impregnated extrudates were dried for 16 hours at 110 °C. Example 13 A portion of the dried extrudates made in Example 12 were impregnated in a further portion of the solution of Example 11 which had been allowed to stand in air for a total period of 21 days. The resulting twice-impregnated extrudates were dried for 16 hours at 110 "C. Eixaniple 14 A portion of the dried extruc'ates maae in Example 13 were impregnated in s further portion of the solution of Example 11 which had been allowed to slant: in air for a total neriocl of 24 days Tne resulting thnce-impregnated extrudafes were dried for 16 hours a: 110 °C. Analysis of the prepared catalysts was carried out as described in Exampie 1 and the results are shown in Table 3. Table 3 \| Exampie ' description ! Cobalt metal surface 1 % w/w Co (oxidic) I area (rn'Vg Co) i I 11 No aqeing 3 dips (comparison) 83.0 10.6 i I 12 2 days ageing, 1 dtp 62.2 6.58 13 21 davs apeiric), 2 dirjs 79,2 10.3 ; ; 14 24 days ageing. 3 dips 88.4 13.6 ! Example 15 Preparation of cobalt catalysts on aJumina extrudates 634g ammonium carbonate chip, ex Brotherton Speciality Products Limited (30-34 X% NH3} and 1880m! ammonia solution (-30%) BDH 'Analar Sp.Gr. 0.8S were placed in the flask. The mixture was continually stirred overnight to ensure that the ammonium carbonate chip had dissolved. 1056g cobalt basic carbonate, ex Shepherd Widnes Ltd. (45-47 /w% Co), was added in aliquots over 60 minutes, whilst continually stirring the solution, and altowed to dissolve. Slow addition was used to prevent any heat buiid up during dissolution of the cobalt powder. The final solution was stirred continually for 48 hours with air access before filtering to remove any undissolved cobalt carbonate. Filtration took 48 hours as the solution was extremely viscous. The solution pH was 10.3 and the Co content was measured as 20.5%. 200g AUOa extrudates (as used in Examples 11 - 14) were placed in a 2 litre round bottomed flask and the impregnation procedure carried out as described in Example 11, The three impregnations were carried out on successive days and so the ageing of the solution for each impregnation was 4, 5 and 6 days. The results are shown in Tabie 4. Example 16 The preparation of Exampie 15 was repeated but the solution was oxidised for s different time. By the time of the third impregnation; it was noted that the solution had become more viscous, possibly due to ioss of ammonia. The results are shown in Table 4. It is notable that it appears possible to prepare a catalyst containing more than 20% of cobalt by this method. Table 4 i Example \| I description ' Cobalt metai surface % w/'w Co (oxidic) I areajrrr/a Co) j ! 15A i 4 days aoeincj 1 dip 81.5 T 9.84 __J ! 15B i 15C f 5 days ageing, 2 dip 87.0 i 13.9 J i 6 days ageing, 3 dips 92.9 1 16.2 ! i 16A 1 5 days ageinc! , 1 dip 96.4 j 11,5 j ! 16B 6 days ageing, 2 dip 83.7 ! 15.1 \| I 18C 17 days ageing, 3 dips 728 ! 20,1 i Example 17 Pt-promoted catalyst Example 15C was repeated and.the resulting catalyst was impregnated with a platinum compound. 0.0806g ptatinum(il) 2,4-pentandionate, equivalent to 0.04- g or Q.1wt% Pi, was dissolved in 27ml acetone. The dried catalyst was placed in a flask and the platinum solution was added drop-wise whilst agitating the fiask gently. Sufficient of the dried catalyst 15C was used so tha' the platinum solution was sufficient to fill the pores of the catalyst with a slight excess. After impregnation, the extrudates were allowed to stand in a fume cupboard for 2 hours to evaporate the solvent prior to drying at 105aC for 16 hours. Analysis gave the following results: Cobalt = 17.6% {by tCP-AES) Platinum = 0-10 % (by ICP-AES) Co metal surface area = 94.5m2/g cobait Example 16 Cobalt on silica extrudates 634g ammonium carbonate chip, (3G-34 7W% NH3) and 1880m! ammonia solution (30%) BDH 'Analar1 Sp.Gr. 0,89 were placed in the flask. The mixture was continually stirred overnight to ensure that the ammonium carbonate chip had dissolved. 1056g cobait basic carbonate, (45-47 w/w% Go), was added in atiquots over 10 hours, whilst continually stirring the solution, and allowed to dissolve with air access for 16 hours. Slow addition was used to prevent any heat build up during dissolution of the cobalt powder. The resulting viscous solution was filtered to remove any undissotved cobalt carbonate. Oxidation of the complex solution was achieved by adding 150mt hydrogen peroxide to the solution after filtration. The solution contained 19.7% Co and had a pH of 11.5, Approximately 100g of silica extrudates, (cylindrical, diameter l.5mrn, length 2-10 mm, KL7200 CY silica extrudates from CRi Kata Leuna.), having a pore volume of 1.1 cm3g~! as measured by water uptake, were placed in a 1 litre round bottom flask. Sufficient quantity of the prepared and oxidized cobalt ammine carbonate solution was added to the extrudates to completely immerse them. The mixture was swirled occasionally for 10 minutes at ambient temperature before filtering off excess solution and drying at 105°C for 16 hours. After dryina the impregnation procedure was repeated. Samples were saved after each impregnation. The properties of the resulting catalysts were measured as noted previously and are shown in Table Exarnate 1J9 Cobalt on zirconia coated silica extrudates 30g of a solution of aqueous ammonium zirconium carbonate solution containing Zr equivalent to 20% ofZrOx (supplied by MEL chemicals of Manchester England) was diluted to 99rnl with demineralised water. The quantity and concentration of solution was calculated to provide approximately Qo of ZrO-> per 1QOg of silica extrudates and to fill 90% of the pore volume of the siiir.s extrudes. lOOc SiG:- exiruciates as used in Example 15 was placed hio the vesse; of z Pascal: Lab-Mixer and tumbled at half spaed far 10 minutes whilst adding trie ZrO; soiution drop-wise. When at the Zr£> solution had been added the vesse! was enclosed and tumbled fcr a further 1C minutes at % speed. Finally, the support was dried at 105CC for 16 hours and calcined at 400"C for 4 hours, During the drying and calcination stages ammonia and carbon dioxide are evolved leading to deposition of ZrO2 onto the support. 100g of the resulting zirconia coated silica extrudates were then impregnated with the cobalt ammine carbonate solution as described in Example 1B. The properties of the resulting caiaivsts were measured as noted previously and are shown in Table 5. The results show that the dispersion of cobalt is greater on the zirconia-coated silica than on the untreated silica extrudates, Table 5 j Example { I Description Cobalt metal surface area (rr2/9 Co) % w/vv Co (oxidic) j % w/w Co I (reduced) L 18 Siiica support, 1 impregnation 32.9 13.5 ] 17.3 i 18 Siiica support, 2 jmgrejinations 39.0 18.5 ! 22.8 18 Silica support, 3 impregnations I 34.8 21.2 1. '29.0 1 ~ "18 Silica support, 4 impregnations 36.2 22.9 31.8 19 ZrOrcoated silica, 1 impregnation 28. 9 13.1 15.6 • -~ig - ZrOz-coaled silica, 2. impregnations 46.1 17.1 23.4 19 ZrOrcoated silica, 3 impregnations 45.2 19.7 27.0 19 ZrGs-coated sritca, 4 impregnations 42.7 21.2 29.0 Exampje_20 Cobalt on silica powder 1880ml of demineralised water and 1920ml of ammonsa were measured into a 5-litre round bottomed flask to which 198g of ammonium carbonate chips were added and stirred at 350rpm until the chips dissolved. Once trie chips had dissolved 218g of cobalt carbonate was added and left to stir over night. The solution was then filtered and left in an open bottle for 2 days to age and then stored in a closed bottle. 2 litres of the solution was measured into a 3-litre round bottom flask on a heating mantle and set to stir. 43,5g of amorphous si!ica powder (Ineos) was added and the pH of the soiution measured. The solution was heated to boiling and tlie pH, temperature and colour of the solution was measured every 15 minutes. When the pH of the solution was found to be in the range of 7.5 to 7.7 the deposition of cobalt was stopped. The remaining solution was filtered off and the filter cake was washed with 2 Hires of dernirieraiised water. The resultant powder was then dried in the oven overnight at 110 °C. Once dry, the powder was put through e 1mm sisvo to break down any large mass of cataiyst. The surface area and average pore diameter (APD) of the silica and the resultant catalysts were measured by nitrogen physisorption. Example 21_ Co on Zr-coated siiice powder 100g of amorphous silica powder from INEOS was weighed into a 1L glass beaker. 41.4g of a solution of zirconium nitrate containing 19.5% ZrCs was weighed into a sample bottle along with 268.2q of demineralised water and mixed. The zirconium nitrate solution was then added tc the silica in small aliquots and stirred by hand until all the zirconium nitrate solution was • worked into the silica. This gave a silica powder at incipient wetness. The sample was then spread evenly on a stainless steel tray and covered and placed in a oven where the temperature was ramped at 2°C per minute to 120°C were it was held for 3 hours and then ramped at 2CC per minute to 450BC where it was held for a further 4 hours. Catalysts were made using the method of Example 20 using the ZrO2-coated silica powder as support instead of the un-coated silica. The % Zr of the coated samples was measured by ICP-AES methodology. The cobalt surface areas of the reduced catalysts were measured by H:» chemisorption as described earlier and the results reported on the oxidic catalyst are shown in Tabie 6. Table 6 Support • Catalyst Reduced cataiyst Oxidic catalyst Example BET PCITA Vnl Co ' BET Particle %Zr (wt %) surface area [0.98des] /i~m3n"1\ Surface area surface area Pore Vol {cmV) APD (A) size D[v, 0.5] %Co (wt %) (mV1) ^iil y } (rrrg"1 cat) (nrg"1) (pm) 20 0 280.9 1.57 25.5 34'1.6 0.34 40 48.9 42.2 21 j 5.3 262,3 1.38 22.4 303.6 I 0.39 39 51.4 42.6~~1 Example 22 Determination of hydrotalcite and spiriei phases in deposited cobalt compound Catalysts containing 20% and 40% cobalt were made using the procedure of Example 1. Samples were made using fresh solutions and oxidised solutions (static air ageing for 8 days). The support used was Puralox HP14/150 gamma alumina and the amount of cobalt in the csfaiyst was varied by varying trie amount of support adcted to the cobalt ammine solution. When the pH of the mixture of Co ammine solution and support had been reduced to about 7.5 by boiling, the solid catalyst particles were collected by filtration and a sample analysed by XRD. The sample was mounted in an XRD holder in its wet state and covered with a PE i dome to D'evern oxidation. XRD analysis was performed on a Siemens D5DOO the'a-thets X-ray o'iffraerometer equipped with a Baltic Scientific Instruments Sol-X Energy Dispersive detector. Conner Ka radiation was used. All scans were done at room temperature. The XRD specimens, were prepared under nitrogen in standard bulk front fill holders. The X-ray beam divergence was controlled by a programmable slit to give 12 mm length illumination of the specimen surface. A 0.6 mm receiving slit was used together with primary soller slits. The data was collected for an angular range of 2-75°26. A rapid scan, for initial measurements with and without a PET dome, step size of 0.1 "29 and a count time per step of 1 second and a slower scan with and without dome with a siep size of 0.05°28 and a count time per step of 1.2 seconds. The scans run with the dome were simply used to confirm that evidence of the oxidation of the sample in the instrument was absent. The Powder Diffraction File™ issued by the International Centre for Diffraction Data, was used as a reference for powder diffraction data. The Go3C>4 spinei pattern from ICDD No 00-043-1003 was used for identification and for calculation of concentrations (Co3O4 111 at 19.G°26 (=4.667A) and Co.,04 311 at 36.845°28 f=2.4374A)). There was no database phase for a Co"'+Co3+ hydrotalcite available, but the pattern for the corresponding Mi/Co phase corresponded quite closely. The "hycrotalcite-type" pattern used for identification and concentration calculations was ICDD No 00-040-0216 for cobalt nickel carbonate hydroxide hydrate Ni 0,7S Co 0,2s (CO3)0.i25 (OH}2 -O.SSHjO using the 7.628A (11.591°2e), 3.6^(23.14326), 2.565A(34.952S128)) 2,285A(39.40r29), 1.93oA(46.89°2e}, 1.734A(52.74"A29), 1.639A(56.065B29), 1,521 A (60.85320) signals. Note that there was no nickel present in the catalysts, the Co/Ni hydrofalciie diffraction pattern was used only to approximate that for the Co2*Cos~ hya'rotaicite. The hydrotaicite phase / spinei phase in the catalysts calculated by this procedure is shown in Table 7, The "wet" samples are measured after filtration but without drying, The "dry" samples are measured after drying In ambient air overnight. Table 7 % Co Solution oxidation Hydrotaicite/spinel ' 4°(^et) 1 •• 1 oxidised j 0.2 ! \|JOJdryJ__ oxidised ; 0,2 1 , 20 (wets ' ~"~ ~" " r — 1 oxidised : 0.5 j 20 (dry) oxidised ; 0.25 ! 40 Not oxidised ! 1.1 i WE CLAIM 1. A method of manufacturing a catalyst, comprising the steps of (i) forming an aqueous solution of a cobalt ammine carbonate complex, (ii) oxidising said solution by adding a chemical oxidant to the solution such that the concentration of Co(III) in the oxidised solution is greater than the concentration of Co(III) in the un- oxidised solution, (iii) decomposing the cobalt ammine complex by heating the solution to a temperature between 80 and 110 °C for sufficient time to allow an insoluble cobalt compound to precipitate out of the solution, (iv) filtering the precipitated cobalt species from the solution and (v) drying the precipitated cobalt species wherein prior to decomposing the cobalt ammine complex, the solution of cobalt ammine complex is oxidised by adding a chemical oxidant selected from a solution of hydrogen peroxide or a hypochlorite to the solution sufficient to convert from 50-90% of the cobalt calculated as moles of cobalt from Co(II) to Co(III) such that in the precipitated cobalt species, after drying at a temperature less than 160°C, the ration of cobalt hydrotalcite phase : cobalt spinel is less than 0.6:1, said cobalt hydrotalcite phase and said cobalt spinel phase being measured by X-ray diffractometry and wherein the cobalt spinel comprises C03O4. 2. A method as claimed in claim 1, wherein the aqueous solution of a cobalt ammine complex is oxidised until the absorbance at max of the UV/visible spectrum occurring between 450 and 600nm is greater than 35% of the absorbance at max of a fully oxidised solution. 3. A method as claimed in claim 2, wherein the aqueous solution of a cobalt ammine complex is oxidised until the absorbance at max of the UV/visible spectrum occurring between 450 and 600nm is greater than 60% of the absorbance at max of a fully oxidised solution. 4. A method as claimed in claim 3, wherein the aqueous solution of a cobalt ammine complex is oxidised until the absorbance at max of the UV/visible spectrum occurring between 450 and 600nm is greater than 90% of the absorbance at max of a fully oxidised solution. 5. A method as claimed in any one of the preceding claims, wherein the cobalt ammine complex solution is mixed with a solid catalyst support prior to heating the solution. 6. A method as claimed in any one of the preceding claims, wherein the catalyst contains from 3 to 85% by weight of total cobalt. 7. A method as claimed in either claim 5 or claim 6, wherein the support material comprises alumina, silica, silica-alumina, zirconia, titania, titania-coated silica, titania-coated alumina, zirconia-coated silica or zirconia-coated alumina. 8. A method as claimed in claim 7, wherein the support material is a powder or fabricated unit comprising alumina. 9. A method as claimed in claim 8, wherein the alumina powder has a mean particle size, D50 in the range 1 µm to 200 µm. 10. A method as claimed in claim 8 or claim 9 wherein the catalyst has a total cobalt content above 20% by weight and wherein the support is a gamma, theta or delta alumina. 11. A method as claimed in any one of the preceding claims, wherein the ratio of cobalt hydrotalcite phase : cobalt spinel is less than 0.5:1. 12. A method as claimed in claim 11, wherein the ratio of cobalt hydrotalcite phase : cobalt spinel is less than 0.3:1. 13. A method as claimed in any one of the preceding claims, further comprising the step of reducing the cobalt species to metallic cobalt in a stream of a hydrogen-containing gas at a temperature in the range 250 - 600 °C.

Full Text

Catajvsts
This invention relates to catalysts and ir particular to catalysts containing cobali which: are sj ft able: for use in hydrogenation reactions.
Catalysis comprising cobalt on s support such as silica or alumina are known in the art for hydrogenation reactions, e.g. for the hydrogenation of aldehydes and nrtriies and for the preparation of hydrocarbons from synthesis gas via the Fischer-Tropsch reaction, in comparison with other catalytic metals such as copper and nickel used for hydrogenation reactions, cobalt is a relatively expensive and so. to obtain the optimum activity, it is desirable that as much as possible of the cobalt present is in an active form accessible to the reactanis.
For hydrogenation reactions, the active form of the oobait is elemental cobalt although in the active cataiyst only some, rather than ail, of the cobalt is normally reduced to the elemental form. Hence s useful measureJs the exposed surface area of elemental cobalt per g of total cobalt present. Except where expressly indicated, as used herein, total cobalt contents are expressed as parts by weight of cobalt (calculated as cobalt metal, whether the cobalt is actually present as the metal or is in a combined form, e.g. as cobalt oxides) per 100 parts by weight of the catalyst or precursor thereto.
Cobalt catalysts on different carriers are disclosed in "Stoichiometries of H2 and CO Adsorptions on cobalt", Journal of Catalysis 85, pages 63-77 (1984) at page 67, table 1. From the total maximum H2 uptake, it is possible to calculate the cobalt surface area per gram of catalyst and the cobalt surface area per gram of cobalt.
US 5 874 381 describes a cobalt or alumina catalyst which contains between 3 and 40% by weight of cobaft and wnich has a relatively high cobalt surface area of above 30 m2/g of total cobalt.
As indicated above, the dispersion of the cobalt on the carrier is important since it is the surface of the cobali of the catalyst which is ca'afyiically active. Therefore it is beneficial to maximise the surface area of the metal which is present so as to produce a catalyst which has a high cobalt surface area per unit mass of lotai cobalt. It may be expected that the dispersion of the cobalt on the catalyst would be maximised at relatively low loadings of cobalt and that, as the amount of cobalt contained in the catalyst is increased, the surface area per gram of cobalt wouid decrease because the cobalt becomes more difficult to disperse on the support.
The aforementioned US 5 874 381 suggests and exemplifies the production of the catalysts by impregnation of shaped transition alumina particles, e.g. extrudates, with a solution of cobalt

ammine carbonate, followed by removal of the excess solution and heating to decompose the cobalt amrnine carbona-.e. We have found that the preparation of cobalt catalysts by the decomposition of cobalt ammine carbonate may be improved.
Accordingly the invention provides a method of manufacturing a catalyst, or precursor thereto, comprising the steps of forming an aqueous solution of a cobalt ammine complex, allowing the solution to oxidise such that the concentration of .Co(HI) in the oxidised solution is greater than the concentration of Coflll) in the un-axidised solution, and then decomposing the cobalt ammine complex by heating the solution to a temperature between 80 and 110 "C for sufficient time to allow an insoluble cobalt compound to precipitate out of the solution.
We have found that when the cobalt ammine complex solution is allowed to oxidise so that at least some of the coba!t (I!) in the complex is converted to Co(fll), the composition of the insoluble cobalt compound resulting from the decomposition of the cobait ammine complex is readily reducible to cobait metal of high surface area. Without wishing to be bound by theory, we believe that the cobalt species which is precipitated from the solution of complex, e.g. a cobait ammine carbonate complex, comprising Co(lll) species contains a greater amount of Co3C>4 and less cobalt carbonate, cobalt hydroxycarbonate or ammonia than is deposited from the decomposition of a freshly made solution containing more Co(ll) and less Co(lll). Preferred complexes are cobalt ammine carbonate complexes although other compounds may also be used, e.g. formates.
The term "cobalt species" is used broadly to include both elemental cobalt and cobalt in combined form, e.g. as compounds such as cobalt oxides and cobalt hydroxycarbonates. The catalyst in its reduced form is useful for catalysing hydrogenation reactions. The catalyst may, however, be provided as a precursor wherein the cobalt is present as one or more compounds, such as oxides or hydroxy carbonates, reducible to elemental cobalt. In this form, the material may be a catalyst precursor and may be treated to reduce the cobalt compounds to metallic cobalt or the material may Itself be a catalyst and used as supplied, e,g, for oxidation reactions. The cobalt surface area figures used herein apply to the material after reduction, but the invention is not limited to the provision of reduced catalyst.
Preferably the cobalt ammine complex solution is heated to decompose the complex in the presence of a catalyst support materlai which is selected from standard known supports such as silica (including both synthetic silica and naturally occurring forms of silica such as kieselguhr), alumina, silica-alumina, liiania, zirconia, carbon, coated silicas or aluminas such as titanJa- or zirconia- coated silicas or aluminas for example. The cata'ysi of the invention is particularly suitable for use in Fisclier-Tropsch (F-T) hydrocarbon synthesis and the supports preferred for cobalt catalysts for use sn known cobalt F-T catalysts may be advantageously

usec for the- catalysts of the present invention. Preferably ar: alumina support is oresenl, which is most preferably a transition alumina, e.g a gamma, trieta or delta alumina, so that preferred catalysts according to the invention comprise a cobalt species on e transition alumina-support.
The support may be in the form of a powder or of E fabricated unit such as a granule, tablet or extrudate. Fabricated units may be in the form.o* elongated cylinders, spheres, lobed shapes or irreaularlv shaped particles, all of which are known in the art of catalyst manufacture. Alternatively the support may be in the form of a coating upon a structure such as a reactor tube wall, honeycomb support, monolith etc. Support materials may contain promoters or other materials such ss binders and may be treated prior to use in the process of the invention, e.g. by drying or calcining.
Suitable transition aiurnina may be of the gamma-alumina group, for example a eta-alumina or chi-alumina. These materials may be formed by calcination of aluminium hydroxides si 400 to 750eC and generally have a BET surface area in (he range 150 to 400 nrr/g. Alternatively, the transition alumina may be of the delta-alumina group which includes the high temperature forms such as delta- and theta- aluminas which may be formed by heating a gamma group afumins to z temperature above about SOO°C. The delta-group aluminas generally have a BET surface area in the range 50 to 150 m2/g. The transition aluminas contain less than 0.5 mole of water per-mole of A!2O3, the actual amount of water depending on the temperature to which they have been heated. Alternatively, we have found that suitable catalyst supports may comprise an alpha-alumina.
A suitable powder for the catalyst support generally has a surface-weighted mean diameter D[3,2I in the range 1 to 200 um. In certain applications such as for catalysts intended for use in slurry reactions, it is advantageous to use very fine particles which have a surface-weighted mean diameter D[3,2] on average, in the range from 1 to 20 prn, e.g. 1 to 10 (jm. For other applications e.g. as a catalyst for reactions earned out in a fluidised bed, it may be desirable to use larger particle sizes, preferably in the range 50 to 150 um. The term surface-weighted mean diameter D[3>2], otherwise termed ihe Sauter mean diameter, is defined by M. Alderliesten in (he paper "A Nomenclature for Mean Particle Diameters"; Anal. Proc.. vol 21, May 1984, pages 167-172, and is calculated from the particle size analysis which may conveniently be effected by laser diffraction for example using a Malvern Mastersizer.
n is preferred (hat a powder support has a relatively large average pore diameter as the use of such supports appears to give catalysts of particularly good selectivity. Preferred supports have an average pore diameter (APD) of at least 10 nm, particularly in the range 15 to 30 nm. (By the term average pore diameter we mean 4 times the pore volume as measured from the riesorption branch of the nitrogen physisorption isotherm at 0.98 relative pressure divided by

the BET surface area]. During the production of the compositions of the invention, cobalt compounds are deposited in the pores of (he support, and so the average pore diameter of the composition will be iess than that of the support employed, and decreases as the proportion of cobalt increases. !? is preferred that the catalysts have an average pore diameter of at least 8 nm, preferably above 10 nm and particufarly in the range 10 to 25 nm.
When the support is transition alumina, it has been found that, depending on the conditions used, the bulk of the cobalt is .precioitated as cobalt compounds within the pores of the transition alumina and none or only a small proportion of the cobalt is deposited as a coating round the alumina particles. As a result, irrespective of the cobalt content of the composition, the particle size of the compositions of the invention is essentially the same as the particle size of the support, and so the compositions of the invention generally have & surface-weighted mean diameter D[3,2] in the range 1 to 200 urn, in one embodiment preferably less than 100 jjm and particularly iess than 20 urn, e.g. 10 urn or less, and in a second embodiment preferably in the range 50 to 150 urn.
On the other hand, since the cobait compounds are primarily precipitated within the pores of the support, the pore volume of the compositions in accordance with the invention will be less than that of the support employed, and will tend to decrease as the cobalt species loading increases. Compositions having a total cobalt content less than 30% by weight preferably have a pore volume of at least 0.5 mi/g whiie compositions having a total cobalt content above 30% by weight, particularly above 40% by weight, preferably have a pore volume of at least 0.3 ml/g, particularly at least 0.4 rn!/g.
The compositions of the invention, when in the reduced state, have a cobait surface area of at least 25 rrr/g of cobalt as measured by the H2 chemisorption technique described herein. Preferably the cobalt surface area is greater than 30, more preferably at least 40, especially at least 60 m'/g. The cobalt surface area tends to decrease as higher loadings of cobai! are used, but we have found that when the composition contains 50 to 60% by weight total cobalt in the reduced state, the cobalt surface area achievable is about 80 rrr/g or more.
The cobalt surface area is determined by H; chemisorption. This method is used in the Examples, and when a cobalt surface area measurement is mentioned in this specification for the catalysts of the invention (unless otherwise specified). Approximately 0.2 to 0,5 g of sample material is firstly degassed and dried by heating to 140"C at 10"C/min in flowing helium and maintaining 3t 140"C for 60 minutes. The degassed and dried sample is then reduced by heating it from 140"C to 425°C at a rate of 3°C /min under a 50 mf/rnin .flow of hydrogen and then maintaining the hydrogen flow at 425*C for 6 hours. Following this reduction, the sample is heated under vacuum to 450°C at 10°C/min and held under these conditions for 2 hours.

The sample is then cooled to 15Q~C and maintained for a further 30 minutes under vacuum. The chsmisorptior analysis is then carried out at 15D=C using purs hydrogen gss. An tiLiiornaiic; analysis program is used to measure z full isotherm overtne range 100 rum Hg up »o 760 irtm Hq pressure of hydrogen. Tsie analysis is carried out twice: the firs' measures the "tcjtal" hydrogen uptake (i.e. includes chemisorbed hydrogen and pnysisorbed Hydrogen; and immediately following the first analysis the sample is put under vacuum ( Cobalt surface areas were calculated in ail cases using the following equation;
Co surface area = ( 6.023 x 1023 x V x SF x A } / 22414
where V ~ uptake of H2 in ml/g
SF - S.loiehiometry factor (assumed 2 for H; cherr.isorption on Co) A = area occupied by one atom of cobalt (assumed 0.0662 nm::)
This equation is described in the Operators Manual for the Micromeretics ASAP 2010 Chemi
System V 2.01, Appendix C, Part No. 201-42808-01, October 1996.
The cobalt ammine complex Is most preferably a cobalt ammine carbonate complex which is formed in situ in aqueous solution by dissolving basic cobalt carbonate in a solution of ammonium carbonate in aqueous ammonium hydroxide, to give a product of the desired cobali content. Alternatively other cobalt salts may be used, including organic salts such as cobait acetate or cobalt formate.
A cobalt arnrnine carbonate complex is the product of dissolving basic cobalt carbonate, preferably of empirical formula Co(OH)z.2x(CO3)x in a solution of ammonium carbonate in aqueous ammonium hydroxide, to give a product of the desired cobalt content. The cobali ammtne carbonate solution may be made by dissolving basic cobalt carbonate in an aqueous sofution of ammonium carbonate or ammonium carbamate containing additional ammonium hydroxide. The relative amounts should be such that the pH of the solution is in the ranqe 75 Io12, preferably 9 to 12. The solution preferably contains 0.1 to 2.5 moles of the cobalt complex per litre. As the concentration of cobalt increases, then generally the proportion of carbonate ions reiative to hydroxide ions in the basic cobalt carbonate feed snouSd be increased. Additional ammonium hydroxide solution may be added in order to provide a slurry of handleable viscosity when support particles are mixed in.
As an alternative, the cobait ammine carbonate solution may be made by dissolving metallic cobait, preferably in powdered form, in aqueous ammonia of pH 11 -12. in the presence of oxygen or air, either with addition of ammonium carbonate or with addition of CO? gas.

The amount of cobalt in the catalyst may tae varied by varying the relative amount of cobalt and support presem in the reaction mixture and by controlling the concentration of the solution of cobalt compound
In accordance with the method of the invention, the complex solution is then allowed to oxidise either by ageing in contact with air or an oxygen-rich gas, or by chemical or electrochemical oxidation, in order that the Co(ll) complex is converted, at least in part to a Co(lll) complex.
The ageing may 5s accomplished by allowing the solution to stand in an uncovered container for the required time, preferably with stirring. The ageing by stirring in the presence of oxygen should be continued for at least 3 hours and preferably for at least 16 hours. Alternatively the solution may be oxidised by bubbling an oxygen-containing gas stream, e.g. air or oxygen, through the solution, optionally with stirring and in this case ageing may be sufficient after just one hour. Alternative methods of ageing the complex include adding an oxidising agent such as hydrogen peroxide, hypochlorite or by electrolytic ageing methods. The amount of chemical oxidant added to the solution is preferably sufficient to convert from 40% to 100% of the cobait in the unoxidised solution, more preferably from 50 •- 90 % and especially from 60 - 90% of the cobalt, calculated as moles of cobalt and assuming that oxidation from Co2* to Co4* is stoichiometric. For example, where 0,65 moles of hydrogen peroxide is used to oxidise a solution containing 1.7 moles of Co, the conversion of Co2* to Co3" is 76.5%, assuming that one mole of peroxide oxidises two moies of cobalt and that the solution contains Co(il) initially. The amount of conversion of Co24 to Co3+ is likely to increase as the temperature is raised. Therefore the oxidation process may be accelerated by warming the solution but the ageing is normally done at room temperature or slightly above room temperature, e.g. frorri about 18°C to about 35°C.
We have found that the oxidation of the aqueous solution of cobait ammine complex is produces an increase in net absorbance of radiation at Amax of the UV/Visibie spectrum occurring between 450 and SOOnm. A,MX represents the height of the peak occurring between 450 and 600nm and is measured in absorbance units relative to an interpolated baseline. The absorbance in this region increases as the extent of oxidation is increased, up to a maximum absorbance when the solution is fully oxidised. In s preferred method of .the invention, the solution of cobalt ammine complex is oxidised until the absorbance at Ariax of the UV/vtsible spectrum occurring between 450 and GOOnm is greater than 35% of the absorbance at A,TO). of a fully oxidised solution. More preferably the absorbence at Am£1), of the UV/visib!e spectrum occurring between 450 and GOOnm is greater than 60%, most preferably greater than 90%, and especially greater than 95%, of the atasorbance at Amax of a fully oxidised solution. The standard measurement conditions utilise a Xenon light source (single beam), a path length of

irnrr .-rx: ssmpie temperature of 20 VC - 25 "C « he sample of cobalt ammine complex solirjon is diluted prior to soecirornerry by adding 1 pan of solution to 4 parts 0" & diluent consisting of 3 parts by volume of 30% aqueous ammonia solution to 7 pcrts demineralisec! water. The dJajeni is used as the blank sample in trie UV/visible spectrometry.
As a further indication of the oxidation required, we have found that the redox potential of a solution of un-oxidisad cobalt ammine complex containing about 3% cobalt is approximately -SOOrnV a* ambient temperature. We have found that the oxidation of the complex is sufficient her the method of the invention when the redox potential is between 0 V and -200mV, more preverablv from -50 to -150 mV, and mast preferably about -lOOm'v, e.g. from -90 to -130 mV, Although) the redox potential would be expected to vary with the concentration of cobalt in the soiut'on we have found that for cobalt ammine carbonate solutions containing between about 2% cobalt and about 18% cobalt, the redox potentials of trie fresh and fully oxidised solutions vary by less than 5% with concentration over the range of concentration.
As a still further indicator for performing the method of the invention, we have found that a sufficiently oxidised solution produces a pink solution when 0.2-- 0.5 ml (i.e. 6 drops) of solution is introduced into 60ml of deionised water at room temperature. Preferably there is no or little precipitation during this test.
Supported cobalt catalysts may be made by impregnating a solid support in the form of a powder or a fabricated unit with a solution of the oxidised cobalt ammine carbonate complex. e.g. by spraying (he support with a measured volume of the solution or by dipping the support into a volume of the solution. The impregnated support is then separated from any supernatant or excess solution and dried at a. temperature in the range 60 to 110°C, so that the cobalt complex decomposes to deposit solid cobalt species upon and in the pores of the support. The impregnation and drying may be repeated several times, e.g. up to about five times, depending on the concentration of the solution and the desired cobalt loading of the support.
Supported cobalt catalysts and precursors may also be made by slu/rying the powdered support, e.g. transition alumina powder, with the appropriate amount of the oxidised aqueous solution of a cobalt ammine carbonate complex. The slurry is then heated, e.g. to a temperature in the range 60 to 110°C, to cause the cobalt ammine complex to decompose with the evolution of ammonia and carbon dioxide and to deposit an insoluble cobatt compound on She surface, and in the pores, of the support. The support carrying the deposited cobalt compound is then filtered from the aqueous medium and dried. The procedure may be repeated, i.e. the dried product may be re-slurried in a, solution of the cobalt ammine complex, heated, filtered and dried, if required to increase the cobalt content of the product. Usinq this form of the process of the invention, a catalyst having a high cobalt dispersion and a high

coba'f loading, e.g. > 10% cobalt, (more preferably > 15% cobalt, by weight) may be prepared in a single deposition step.
The time allowed for the precipitation of the cobalt compound is normally about 30 to 200 minutes; the precipitation is usually complete after about 60 to 80 minutes, but the heating of the slurry may be prolonged to include a further precipitate-ageing step. We have found that when the cobalt content is relatively low, e.g. up to about 40% by weight, it is beneficial to use relatively short process times, e.g. by limiting the total heating time. i.e. for both the precipitation and any precipitate-ageing to 200 minutes or less, preferably 'ess than 150 minutes. As the cobalt content of the catalyst is increased, longer process times may be used, e.g. up to about 350 minutes.
When the precipitated cobalt compound contains a mixture of Co2* and Co3* a cobalt hydrotateite species may be farmed. The hydrotaicite species has been identified by X-ray diffractometry (XRD) in which the hydrotaicite phase shows a similar diffraction pattern to known Ki/Co hydrotaicite phases. The catalysts of the invention may be distinguished by the ratio of cobalt hydrotaicite to cobalt spinel found in the precipitated cobalt species after drying at less than 160°C i.e. at a temperature iess than the temperature at which Co^Clt is formed from cobalt nitrate by calcination. The cobalt hydrotaicite may.be represented as [Co2+/Co3+](OH)CO3 in which the ratio of Co2+ : Co3+ is about 3:1. The cobalt spinel has an empirical formula of Co3O4 so contains more Co3" than Co2*, The amount of hydrotaicite and cobalt spinel may be determined by XRD. Preferred catalysts of the invention have a ratio of cobalt hydrotaicite ; cobalt spinel of less than 0.6:1, mere preferably iess than 0.5:1, especially less than 0.3:1. The cobalt spinel is calculated from the powder diffraction peaks 00304111 at 19.0"28 (=4.587A) and Co3CU 311 at 36.845°28 (=2.4374,4). The cobalt hydrotaicite is estimated from the diffraction pattern in the patterns published by the International Centre for Diffraction Data JCDD No 00-040-0216 for cobalt nickel carbonate hydroxide hydrate Ni 076 Co 0.25 (CO3}0.-.2o (OH)2 -0.38H2O using the ICDD No 00-040-0216 for cobalt nickel carbonate hydroxide hydrate Ni O.rs Co 0.2s (CO3)ai25 (OHV> 'O.SSHaO using the 7.628A (11.591°29}, 3.84,4(23,143°29). 2.565A(34,952°2B}, 2.285A(39.401°2e), 1.936A(46.89"28), 1.734A(52,74-7A28), 1.639A(56.065"26), 1.521 A (60.853°28) signals. The cobalt hydrotaicite is most particularly identified using the 7.628A (11.591°29) and 3.84A(23.143e28). The ratio of cobalt hydrotaicite : cobalt spinel is estimated from the ratio of the peak areas. Note that there was no nickel present In the catalysts, the Co/Ni hydrotaicite diffraction pattern was used only to approximate that for the Co^Co3*" hydrotaicite.
We have found that when the cobalt species is precipitated in the spinel form compared with the hydrotaicite form, the dispersion of the cobalt, and therefore the cobalt metal surface area is greater whan the catalyst is reduced in hydrogen to convert the cobalt compounds into metallic cobalt. It is therefore preferred to maximise the amount of cobalt compound

precipitant; in spine! form. In contrast catalysts prepared by impregnation of cooa't nitrate OHIO a support always deposit cobalt as Co*' because of tho acidic nature of the cobalt nitrate solution. Therefore no hycirotaicite or spinel is formed and the dried cobalt nitrate must be calcined to form the oxide before reduction in hydrogen.
According to a further aspect of the invention we provide a catalyst intermediate comprising a cobalt compound, said cobal* compound comprising a Coil!).'Co(!)I) hydrotalcite phase and & Gc-iO/ cobalt spinei phase, wherein the ratio of cobalt hydrotaicite phase : cobalt spinel phase is less than 0.6:1, the amount of said cobalt hydrotalcite phase and said cobalt spine! phase being measured by X-ray diftractometry. The catalyst intermediate may be used as s catalyst but normally is subjected to a further process such as the reduction of the cobalt species in a hydrogen-containing gas to provide a catalyst comprising metallic cobalt. The catalyst intermediate may be obtained using the method of the invention. The catalyst intermediate preferably comprises a support as described above.
For some applications it may be desired to incorporate modifiers, such as other metals or compounds thereof, into the catalyst or precursor. This may be effected by Impregnating the dried product with a solution of a compound of the desired modifier that decomposes to the oxide or elemental form upon heating. Examples of such modifiers include alkali metals, precious metafs, and transition metals. Common promoters used in cobalt catalysts for Fischer Tropsch processes include manganese, platinum, ruthenium'and rhenium.
If desired, the product may be calcined in air, e,g. at a temperature in the range 200 to 600°C, more preferably 200 to 450°C, to decompose the deposited cobalt compound to cobalt oxide. However we have found that using the method of the invention a .significant part of the cobalt species formed by the decomposition of the cobalt ammine complex is Co3O. The composition may be used in its oxidic siafe, i.e. without reducing the cobalt oxides to metallic cobaft. It may be used as a catalyst in this state for e.g. oxidation reactions. Alternatively and preferably, the catalyst is reduced to an active catalyst containing cobalt metal by the end-user. The composition may alternatively be supplied as a reduced catalyst which has been passivated, so that the cobalt metal is protected from daactivation during storage and

transportation. Thus a precursor comprising the support and the unreduced cobsli compound, possibly dispersed in a carrier, may be charged to s hydrogenaiion reactor optionally with the material to be hydroganated and the mixture heated while hydrogen is sparged through the mixture.
The catalysts may be used for hydrogenation reactions such as the hydrogenation of oiefinic compounds, e.g. waxes, nitro or nitriie compounds, e.g. the conversion of nitrobenzene to anitine or the conversion of nitriles to amines. They may also be used for the hydrogenation of paraffin waxes to remove traces of unsaturation therein. They may also be useful in s wide range of other reactions, for example the Fischer-Tropsch process, i.e. where hydrogen and carbon monoxide are reacted in the presence of the catalyst to form higher hydrocarbons. This may be part of an overall process for the conversion of natural gas to petroleum compounds wherein the hydrogen / carbon monoxide gas mixture is a synthesis gas formed by steam reforming natural gas.
The invention will be further described in the following experimental examples and the accompanying drawings, which are:-
Fig 1 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 1.
Fig 2 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 2.
Fig 3 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 3.
Fie 4 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 4.
Fig 5: UV-visibie spectrogra'ph of solutions produced in Example 6.
Fig 6 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 7.
Fig 7 ; Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 8.
Fig 8 : Temperature programmed reduction trace of signal vs temperature for catalyst made according to Example 9 & 10.
Genera; method pf catalyst preparation for ..Examjjjes 1 _^ 1j3
The cobalt amrnine carbonate complex solution was made up using 1707 g ammonia solution (SG 0.89, 30% ammonia), 198 g ammonium carbonate, 218 g basic cobalt carbonate (46.5% wt% Co, bulk density 0,640 g/ml) and 1877 g demineralised water. The solution was stirred for one hour in order to dissolve the basic cobalt carbonate, The total volume of solution was 4

Hires. The solution was oxidised as describee in the following examples before using (o make rooai: catalysts as oescnbed below,
The alumina employed was Purabx™ HP14/150 available from SasoL which is a transittcn-aiumina of the garnma alumina type having a particle size D.x of 4S - 50 prn. The alumina was used as received.
The alumina particles and a measured amount of the cobalt ammine carbonate complex soiution were charged to a stirred vessel equipped with a condenser. The relative amounts of alumina and cobalt arnmine carbonate complex solution were calculated to provide a catalyst containing 40% by weight of cobalt mela! in the oxide catalyst. The pH of the aqueous solution was 11.1 The mixture was heated to boiling while stirring and gentle boiling at about QB'C to about 1QCTC was maintained until a pH of between 7.5 and 7,7 was achieved, during which the solution becomes dear. The solid was then filtered off, briefly washed in water and then dried inar at 105CC overnight.
The cobalt surface area of the catalysts was measured by reduction in a flowing hydrogen stream at 425*0 followed by H; chemisorption at 15Q"C using the standard method described
earlier
Temperature programmed reduction (TPR) of the dried materials was carried out in a reduction gas stream comprising 5% hydrogen in nitrogen. The sample (between 0.1 and 0.15 g, accurately weighed) was first healed to 120 °C (at 5'C /mm under a reduction gas flow of 25ml/mirt) to remove moisture and heid at 120 "C for 45 minutes. Thereafter the sample was heated from 120 to 1000"C in the reduction gas stream at .a heating rate of 5°C per minute The change in concentration of hydrogen between the inlet gas and the outlet aas was monitored by a katharometer to show the consumption of hydrogen at each temperature. The TPR instrument used was Quantachrome ChernBei© TPR/TPD analyser. The TPR traces are shown in Figs 1 1o 4 and 6 to 8.
The cobalt surface area was measured according to the standard hydrogen chemisorption procedure described above.
UV-vis spectra were acquired at ambient temperature using a Varian Gary 50 spectropriotorneier, equipped with a Xenon flash lamp; using-a ceil path isngth of 1mm. The sample of cobalt arnmine solution is diluted prior ID spectrometry by adding 1 pan of soiution to 4 Darts of a diluent consisting of 3 pa-is by volume of 30% aqueous ammonia solution to 7 parts dernineralised water. The diluent is used as the blank sample in the UV/visibie spectroinetry.

(comparative)
The r.obait ammine carbonate solution was used immediately after dissolution of the basic cobalt carbonate solids was complete. The coiaalt crystallite see in the catalyst was measured using X-ray diffraction during an in-situ reduction procedure, The cobalt metal crystallites formed at reduction temperatures between about 400°C and about 6QO°C had an average crystallite size of 6 - 10 nm, based upon the 200 reflection.
Example 2
The cobalt ammine carbonate solution was oxidised with stirring in contact with air for 3 hours
prior to preparation of the catalyst.
Example 3
The cobalt ammine carbonate solution was oxidised with stirring in contact with air for 16 hours and then, stored without stirring In a container for 48 hours, allowing air to enter the container periodically throughout the storage time. The cobalt metal crystalfites formed at reduction temperatures between about 400 degrees and about 600 degrees had an average crystallite size of 4 - 5 nm, based upon the 200 reflection.
Example 4
The cobalt ammine carbonate solution was stirred whilst air was continuously bubbled through
the solution for one hour prior to preparation of the catalyst.
The TPR traces for Examples 1 -2 (in figs 1 - 2) show a reduction taking place between about 350°C and 400°C. This reduction is absent from the trace measured for Example 3 {fig 3) which has the highest cobalt surface area, in Fig 4 some reduction at this temperature appears to take piace, visible as a shoulder on the 30Q°C reduction peak. Reduction between about 350°C and 40CTC is currently believed to be associated with reduction of a cobalt hydrotafcife phase .
Example S
The cobalt ammine carbonate solution was stirred for one hour and air was bubbled through
the solution for 5 minutes every 15 minutes prior to preparation of the catalyst.
__ ._ _____ __
j Example f Ageing of sofution j Cobalt rretaf surface
_____ None

Stir 18 hrs. store 48 hours ! 57,9
Constant air purge 1 hour jtterit air purge 1_ hour ! _52.'

Example 5
A cobalt ammine carbonate solution was made up as describee eariie-. One portion,
desionateti 5s, was stirred in air-far 3 hours. A second portion,6b: was stirred for 16 hours. A
third portion. 6c, was allowed to stand, with occasional air ingress, for 30 days. The solutions
« were analysed using UV-visible spectroscopy. The spectra are shown in Fig 5. It can be seen
that the portion of the spectrum between about 450 nm and 600 nrn shows an increase in absorbance as the amount of oxidation is increased.
Example 7
To a 4 litre baicn of coaalt ammtne carbonate solution made according to the general process given above was added approximately 50 ~ 60 mi of 30% .hydrogen peroxide dropwise with ' stirring until 6 drops (about 0.5 ml) of the solution mixed in 60ml deminerafised water resulted in a pink solution with only "minimal precipitation, indicating that sufficient oxidation of the solution had occurred to give mainly Co3"". The same test on insufficiently oxidised solution results in a blue colour solution and a precipitate. The solution was filtered and catalysts were prepared according to the method given above and the properties were measured using the same methods. The results are shown in Table 2. The TPR trace for the catalyst is shown in • Fig 6.
Example 8
A 4-litre batch of cobalt ammine carbonate solution made according to the general process was left standing in static air for 8 days. The solution was filtered and catalysts were prepared according to the method given above and the properties were measured using the same methods. The TPR trace is shown as Fig 7. The results are shown in Table 2.
ExarnD:[e_9
A 32 litre batch of cofaait ammine carbonate solution was made using the genera! method above using ingredients in the same proportions as that given for the general method above, The solution was filtered and then a 30% hydrogen peroxide solution was pumped into the solution until the Redox potential changed from approximately -300 (before HjOj addition) to
100 mV. The Redox potential was measured using a Mettler Toledo pH transmitter 2500.
Catalysts were prepared according to the method given above and the properties were measured using methods described above. The TPR trace is shown in Fig 8. The results are shown in Table 2
Example 10 (comparison]
A second 32 litre batch of cobalt ammine carbonate solution was made and filtered in the same
way as described in Example S, Catalysts were made by the general method ai'ven without

waiting for the solution to oxidise or adding hydrogen peroxide or other oxidant. The results are shown in Table 2. The TPR trace is shown in Fig 8.
Table 2

j Example
i Oxidation, of solution Cobalt metal surface area (rrf /g Co) % w/w Co (oxidic)
! 7 " H2O2 to colour change 54.2 41 .8
8 Static air ag_eing_8 days 61.9 35.9
9 H2O2 to -100mV Redox potential , 59.1 42,8
10 None (comparison) 43.5 36.9
Example 11 (comparative) Preparation of-cobalt catalysts supported on alumina extrudates Preparation of cobalt amrnine carbonate solution;-
Ammoriiurn carbonate chip (634g), (30-34 w/w% WH3), was weighed into a 3 litre round bottomed flask and ammonia solution i~30%) (1880ml) BDH 'Analar1 Sp.Gr. 0.89 added, The mixture was continually stirred overnight to ensure that the ammonium carbonate chip had dissolved. Cobalt basic carbonate (528g), {45-47 w/w% Co), was added in aliquots over 60 minutes. The final solution was stirred continually for approximately 30 minutes before filtering to remove any traces of insoluble particulate matter. Solution pH was10.95 and Co content was 14.3%. The solution was used immediately after filtering to carry out three impregnations onto an alumina extrudate support.
Alumina extrudates of 12mm diameter were calcined at 1050 *C for 2 hours. 200g alumina extrudates were placed in a 2 litre round bottom flask. An amount of the prepared cobalt ammine carbonate solution was added to the extrudates such that the extrudates were completely covered. The mixture was swirled occasionally for 5 minutes before-decanting off excess solution. The impregnated extrudates were filtered to remove any remaining solution then dried at'110°C for 2 -4 hours. After drying the impregnation procedure was repeated twice more and the thrice-impregnated extrudates were dried for 16 hours at 110 °C.
Exa.rnpjeJ2
A portion of (he solution made in Example 11 was oxidised by allowing it to stand with access to air for 48 hours. 200g of alumina extrudates were impregnated using the method of Example 11 for one impregnation only, After filtration, the impregnated extrudates were dried for 16 hours at 110 °C.
Example 13
A portion of the dried extrudates made in Example 12 were impregnated in a further portion of the solution of Example 11 which had been allowed to stand in air for a total period of 21 days. The resulting twice-impregnated extrudates were dried for 16 hours at 110 "C.
Eixaniple 14

A portion of the dried extruc'ates maae in Example 13 were impregnated in s further portion of the solution of Example 11 which had been allowed to slant: in air for a total neriocl of 24 days Tne resulting thnce-impregnated extrudafes were dried for 16 hours a: 110 °C. Analysis of the prepared catalysts was carried out as described in Exampie 1 and the results are shown in
Table 3.
Table 3

| Exampie ' description
! Cobalt metal surface 1 % w/w Co (oxidic)

I area (rn'Vg Co) i
I 11 No aqeing 3 dips (comparison) 83.0 10.6 i
I 12 2 days ageing, 1 dtp 62.2 6.58
13 21 davs apeiric), 2 dirjs 79,2 10.3 ;
; 14 24 days ageing. 3 dips 88.4 13.6 !
Example 15 Preparation of cobalt catalysts on aJumina extrudates
634g ammonium carbonate chip, ex Brotherton Speciality Products Limited (30-34 X% NH3} and 1880m! ammonia solution (-30%) BDH 'Analar Sp.Gr. 0.8S were placed in the flask. The mixture was continually stirred overnight to ensure that the ammonium carbonate chip had dissolved. 1056g cobalt basic carbonate, ex Shepherd Widnes Ltd. (45-47 */w% Co), was added in aliquots over 60 minutes, whilst continually stirring the solution, and altowed to dissolve. Slow addition was used to prevent any heat buiid up during dissolution of the cobalt powder. The final solution was stirred continually for 48 hours with air access before filtering to remove any undissolved cobalt carbonate. Filtration took 48 hours as the solution was extremely viscous. The solution pH was 10.3 and the Co content was measured as 20.5%.
200g AUOa extrudates (as used in Examples 11 - 14) were placed in a 2 litre round bottomed flask and the impregnation procedure carried out as described in Example 11, The three impregnations were carried out on successive days and so the ageing of the solution for each impregnation was 4, 5 and 6 days. The results are shown in Tabie 4.
Example 16
The preparation of Exampie 15 was repeated but the solution was oxidised for s different time. By the time of the third impregnation; it was noted that the solution had become more viscous, possibly due to ioss of ammonia. The results are shown in Table 4. It is notable that it appears possible to prepare a catalyst containing more than 20% of cobalt by this method.
Table 4

i Example | I description ' Cobalt metai surface % w/'w Co (oxidic) I areajrrr/a Co) j
! 15A i 4 days aoeincj 1 dip 81.5 T 9.84 __J
! 15B i 15C f 5 days ageing, 2 dip 87.0 i 13.9 J

i 6 days ageing, 3 dips 92.9 1 16.2 !
i 16A 1 5 days ageinc! , 1 dip 96.4 j 11,5 j
! 16B 6 days ageing, 2 dip 83.7 ! 15.1 |
I 18C 17 days ageing, 3 dips 728 ! 20,1 i

Example 17 Pt-promoted catalyst
Example 15C was repeated and.the resulting catalyst was impregnated with a platinum
compound. 0.0806g ptatinum(il) 2,4-pentandionate, equivalent to 0.04- g or Q.1wt% Pi, was
dissolved in 27ml acetone. The dried catalyst was placed in a flask and the platinum solution
was added drop-wise whilst agitating the fiask gently. Sufficient of the dried catalyst 15C was
used so tha' the platinum solution was sufficient to fill the pores of the catalyst with a slight
excess. After impregnation, the extrudates were allowed to stand in a fume cupboard for 2
hours to evaporate the solvent prior to drying at 105aC for 16 hours. Analysis gave the
following results: Cobalt = 17.6% {by tCP-AES)
Platinum = 0-10 % (by ICP-AES)
Co metal surface area = 94.5m2/g cobait
Example 16 Cobalt on silica extrudates
634g ammonium carbonate chip, (3G-34 *7W% NH3) and 1880m! ammonia solution (30%) BDH 'Analar1 Sp.Gr. 0,89 were placed in the flask. The mixture was continually stirred overnight to ensure that the ammonium carbonate chip had dissolved. 1056g cobait basic carbonate, (45-47 w/w% Go), was added in atiquots over 10 hours, whilst continually stirring the solution, and allowed to dissolve with air access for 16 hours. Slow addition was used to prevent any heat build up during dissolution of the cobalt powder. The resulting viscous solution was filtered to remove any undissotved cobalt carbonate. Oxidation of the complex solution was achieved by adding 150mt hydrogen peroxide to the solution after filtration. The solution contained 19.7% Co and had a pH of 11.5,
Approximately 100g of silica extrudates, (cylindrical, diameter l.5mrn, length 2-10 mm, KL7200 CY silica extrudates from CRi Kata Leuna.), having a pore volume of 1.1 cm3g~! as measured by water uptake, were placed in a 1 litre round bottom flask. Sufficient quantity of the prepared and oxidized cobalt ammine carbonate solution was added to the extrudates to completely immerse them. The mixture was swirled occasionally for 10 minutes at ambient temperature before filtering off excess solution and drying at 105°C for 16 hours. After dryina the impregnation procedure was repeated. Samples were saved after each impregnation. The properties of the resulting catalysts were measured as noted previously and are shown in Table
Exarnate 1J9 Cobalt on zirconia coated silica extrudates
30g of a solution of aqueous ammonium zirconium carbonate solution containing Zr equivalent to 20% ofZrOx (supplied by MEL chemicals of Manchester England) was diluted to 99rnl with demineralised water. The quantity and concentration of solution was calculated to provide approximately Qo of ZrO-> per 1QOg of silica extrudates and to fill 90% of the pore volume of the

siiir.s extrudes. lOOc SiG:- exiruciates as used in Example 15 was placed hio the vesse; of z Pascal: Lab-Mixer and tumbled at half spaed far 10 minutes whilst adding trie ZrO; soiution drop-wise. When at the Zr£> solution had been added the vesse! was enclosed and tumbled fcr a further 1C minutes at % speed. Finally, the support was dried at 105CC for 16 hours and calcined at 400"C for 4 hours, During the drying and calcination stages ammonia and carbon dioxide are evolved leading to deposition of ZrO2 onto the support.
100g of the resulting zirconia coated silica extrudates were then impregnated with the cobalt ammine carbonate solution as described in Example 1B. The properties of the resulting caiaivsts were measured as noted previously and are shown in Table 5. The results show that the dispersion of cobalt is greater on the zirconia-coated silica than on the untreated silica extrudates,
Table 5

j Example
{
I Description Cobalt metal surface area (rr2/9 Co) % w/vv Co (oxidic) j % w/w Co I (reduced)
L 18 Siiica support, 1 impregnation 32.9 13.5 ] 17.3
i 18 Siiica support, 2 jmgrejinations 39.0 18.5 ! 22.8
18 Silica support, 3 impregnations I 34.8 21.2 1. '29.0 1
~ "18 Silica support, 4 impregnations 36.2 22.9 31.8
19 ZrOrcoated silica, 1 impregnation 28. 9 13.1 15.6
• -~ig - ZrOz-coaled silica, 2. impregnations 46.1 17.1 23.4
19 ZrOrcoated silica, 3 impregnations 45.2 19.7 27.0
19 ZrGs-coated sritca, 4 impregnations 42.7 21.2 29.0
Exampje_20 Cobalt on silica powder
1880ml of demineralised water and 1920ml of ammonsa were measured into a 5-litre round bottomed flask to which 198g of ammonium carbonate chips were added and stirred at 350rpm until the chips dissolved. Once trie chips had dissolved 218g of cobalt carbonate was added and left to stir over night. The solution was then filtered and left in an open bottle for 2 days to age and then stored in a closed bottle.
2 litres of the solution was measured into a 3-litre round bottom flask on a heating mantle and set to stir. 43,5g of amorphous si!ica powder (Ineos) was added and the pH of the soiution measured. The solution was heated to boiling and tlie pH, temperature and colour of the solution was measured every 15 minutes. When the pH of the solution was found to be in the range of 7.5 to 7.7 the deposition of cobalt was stopped. The remaining solution was filtered off and the filter cake was washed with 2 Hires of dernirieraiised water. The resultant powder

was then dried in the oven overnight at 110 °C. Once dry, the powder was put through e 1mm sisvo to break down any large mass of cataiyst.
The surface area and average pore diameter (APD) of the silica and the resultant catalysts were measured by nitrogen physisorption.
Example 21_ Co on Zr-coated siiice powder
100g of amorphous silica powder from INEOS was weighed into a 1L glass beaker. 41.4g of a solution of zirconium nitrate containing 19.5% ZrCs was weighed into a sample bottle along with 268.2q of demineralised water and mixed. The zirconium nitrate solution was then added tc the silica in small aliquots and stirred by hand until all the zirconium nitrate solution was • worked into the silica. This gave a silica powder at incipient wetness. The sample was then spread evenly on a stainless steel tray and covered and placed in a oven where the temperature was ramped at 2°C per minute to 120°C were it was held for 3 hours and then ramped at 2CC per minute to 450BC where it was held for a further 4 hours.
Catalysts were made using the method of Example 20 using the ZrO2-coated silica powder as support instead of the un-coated silica.
The % Zr of the coated samples was measured by ICP-AES methodology. The cobalt surface areas of the reduced catalysts were measured by H:» chemisorption as described earlier and the results reported on the oxidic catalyst are shown in Tabie 6.
Table 6

Support • Catalyst
Reduced cataiyst Oxidic catalyst
Example BET PCITA Vnl Co ' BET Particle
%Zr (wt %) surface area [0.98des]
/i~m3n"1\ Surface
area surface area Pore Vol {cmV) APD (A) size D[v, 0.5] %Co (wt %)
(mV1) ^iil y } (rrrg"1 cat) (nrg"1) (pm)
20 0 280.9 1.57 25.5 34'1.6 0.34 40 48.9 42.2
21 j 5.3 262,3 1.38 22.4 303.6 I 0.39 39 51.4 42.6~~1
Example 22 Determination of hydrotalcite and spiriei phases in deposited cobalt compound Catalysts containing 20% and 40% cobalt were made using the procedure of Example 1. Samples were made using fresh solutions and oxidised solutions (static air ageing for 8 days). The support used was Puralox HP14/150 gamma alumina and the amount of cobalt in the csfaiyst was varied by varying trie amount of support adcted to the cobalt ammine solution. When the pH of the mixture of Co ammine solution and support had been reduced to about 7.5 by boiling, the solid catalyst particles were collected by filtration and a sample analysed by XRD.

The sample was mounted in an XRD holder in its wet state and covered with a PE i dome to D'evern oxidation. XRD analysis was performed on a Siemens D5DOO the'a-thets X-ray o'iffraerometer equipped with a Baltic Scientific Instruments Sol-X Energy Dispersive detector. Conner Ka radiation was used. All scans were done at room temperature. The XRD specimens, were prepared under nitrogen in standard bulk front fill holders.
The X-ray beam divergence was controlled by a programmable slit to give 12 mm length illumination of the specimen surface. A 0.6 mm receiving slit was used together with primary soller slits. The data was collected for an angular range of 2-75°26. A rapid scan, for initial measurements with and without a PET dome, step size of 0.1 "29 and a count time per step of 1 second and a slower scan with and without dome with a siep size of 0.05°28 and a count time per step of 1.2 seconds. The scans run with the dome were simply used to confirm that evidence of the oxidation of the sample in the instrument was absent.
The Powder Diffraction File™ issued by the International Centre for Diffraction Data, was used as a reference for powder diffraction data. The Go3C>4 spinei pattern from ICDD No 00-043-1003 was used for identification and for calculation of concentrations (Co3O4 111 at 19.G°26 (=4.667A) and Co.,04 311 at 36.845°28 f=2.4374A)).
There was no database phase for a Co"'+Co3+ hydrotalcite available, but the pattern for the corresponding Mi/Co phase corresponded quite closely. The "hycrotalcite-type" pattern used for identification and concentration calculations was ICDD No 00-040-0216 for cobalt nickel carbonate hydroxide hydrate Ni 0,7S Co 0,2s (CO3)0.i25 (OH}2 -O.SSHjO using the 7.628A (11.591°2e), 3.6^(23.143*26), 2.565A(34.952S128)) 2,285A(39.40r29), 1.93oA(46.89°2e}, 1.734A(52.74"A29), 1.639A(56.065B29), 1,521 A (60.853*20) signals. Note that there was no nickel present in the catalysts, the Co/Ni hydrofalciie diffraction pattern was used only to approximate that for the Co2*Cos~ hya'rotaicite.
The hydrotaicite phase / spinei phase in the catalysts calculated by this procedure is shown in Table 7, The "wet" samples are measured after filtration but without drying, The "dry" samples are measured after drying In ambient air overnight.
Table 7
% Co Solution oxidation Hydrotaicite/spinel
' 4°(^et) 1 •• 1 oxidised j 0.2 !
|JOJdryJ__ oxidised ; 0,2 1
, 20 (wets ' ~"~ ~" " r — 1
oxidised : 0.5 j
20 (dry) oxidised ; 0.25 !
40 Not oxidised ! 1.1 i

WE CLAIM
1. A method of manufacturing a catalyst, comprising the steps of
(i) forming an aqueous solution of a cobalt ammine carbonate complex,
(ii) oxidising said solution by adding a chemical oxidant to the solution such that the concentration of Co(III) in the oxidised solution is greater than the concentration of Co(III) in the un- oxidised solution,
(iii) decomposing the cobalt ammine complex by heating the solution to a temperature between 80 and 110 °C for sufficient time to allow an insoluble cobalt compound to precipitate out of the solution,
(iv) filtering the precipitated cobalt species from the solution and
(v) drying the precipitated cobalt species
wherein prior to decomposing the cobalt ammine complex, the solution of cobalt ammine complex is oxidised by adding a chemical oxidant selected from a solution of hydrogen peroxide or a hypochlorite to the solution sufficient to convert from 50-90% of the cobalt calculated as moles of cobalt from Co(II) to Co(III) such that in the precipitated cobalt species, after drying at a temperature less than 160°C, the ration of cobalt hydrotalcite phase : cobalt spinel is less than 0.6:1, said cobalt hydrotalcite phase and said cobalt spinel phase being measured by X-ray diffractometry and wherein the cobalt spinel comprises C03O4.
2. A method as claimed in claim 1, wherein the aqueous solution of a cobalt ammine complex is oxidised until the absorbance at max of the UV/visible spectrum occurring between 450 and 600nm is greater than 35% of the absorbance at max of a fully oxidised solution.
3. A method as claimed in claim 2, wherein the aqueous solution of a cobalt ammine complex is oxidised until the absorbance at max of the UV/visible spectrum occurring

between 450 and 600nm is greater than 60% of the absorbance at max of a fully oxidised solution.
4. A method as claimed in claim 3, wherein the aqueous solution of a cobalt ammine complex is oxidised until the absorbance at max of the UV/visible spectrum occurring between 450 and 600nm is greater than 90% of the absorbance at max of a fully oxidised solution.
5. A method as claimed in any one of the preceding claims, wherein the cobalt ammine complex solution is mixed with a solid catalyst support prior to heating the solution.
6. A method as claimed in any one of the preceding claims, wherein the catalyst contains from 3 to 85% by weight of total cobalt.
7. A method as claimed in either claim 5 or claim 6, wherein the support material comprises alumina, silica, silica-alumina, zirconia, titania, titania-coated silica, titania-coated alumina, zirconia-coated silica or zirconia-coated alumina.
8. A method as claimed in claim 7, wherein the support material is a powder or fabricated unit comprising alumina.
9. A method as claimed in claim 8, wherein the alumina powder has a mean particle size, D50 in the range 1 µm to 200 µm.
10. A method as claimed in claim 8 or claim 9 wherein the catalyst has a total cobalt content above 20% by weight and wherein the support is a gamma, theta or delta alumina.

11. A method as claimed in any one of the preceding claims, wherein the ratio of cobalt hydrotalcite phase : cobalt spinel is less than 0.5:1.
12. A method as claimed in claim 11, wherein the ratio of cobalt hydrotalcite phase : cobalt spinel is less than 0.3:1.
13. A method as claimed in any one of the preceding claims, further comprising the step of reducing the cobalt species to metallic cobalt in a stream of a hydrogen-containing gas at a temperature in the range 250 - 600 °C.

Documents:

6229-DELNP-2006-Abstract-(27-02-2012).pdf

6229-delnp-2006-abstract.pdf

6229-DELNP-2006-Claims-(23-08-2012).pdf

6229-DELNP-2006-Claims-(27-02-2012).pdf

6229-delnp-2006-claims.pdf

6229-delnp-2006-correspondence others-(09-05-2008).pdf

6229-DELNP-2006-Correspondence Others-(23-08-2012).pdf

6229-DELNP-2006-Correspondence Others-(27-02-2012).pdf

6229-delnp-2006-Correspondence Others-(30-08-2012).pdf

6229-delnp-2006-correspondence-others.pdf

6229-delnp-2006-description (complete).pdf

6229-DELNP-2006-Drawings-(27-02-2012).pdf

6229-delnp-2006-drawings.pdf

6229-DELNP-2006-Form-1-(27-02-2012).pdf

6229-delnp-2006-form-1.pdf

6229-delnp-2006-form-18-(09-05-2008).pdf

6229-DELNP-2006-Form-2-(27-02-2012).pdf

6229-delnp-2006-form-2.pdf

6229-DELNP-2006-Form-3-(23-08-2012).pdf

6229-DELNP-2006-Form-3-(27-02-2012).pdf

6229-delnp-2006-Form-3-(30-08-2012).pdf

6229-delnp-2006-form-3.pdf

6229-delnp-2006-form-5.pdf

6229-DELNP-2006-GPA-(23-08-2012).pdf

6229-DELNP-2006-GPA-(27-02-2012).pdf

6229-delnp-2006-gpa.pdf

6229-delnp-2006-pct-304.pdf

6229-delnp-2006-pct-409.pdf

6229-delnp-2006-pct-request form.pdf

6229-delnp-2006-pct-search report.pdf

6229-DELNP-2006-Petition-137-(27-02-2012).pdf

« Previous Patent

Next Patent »

Patent Number

254373

Indian Patent Application Number

6229/DELNP/2006

PG Journal Number

44/2012

Publication Date

02-Nov-2012

Grant Date

29-Oct-2012

Date of Filing

25-Oct-2006

Name of Patentee

JOHNSON MATTHEY PLC.

Applicant Address

40-42 HATTON GARDEN LONDON EC 1 N 8EE ,UNITED KINGDOM.

Inventors:

#	Inventor's Name	Inventor's Address
1	CORNELIS MARTINUS LOK	29 MONTAGU HARRIER GUISBOROUGH CLEVELAND TS14 8PB UNITED KINGDOM
2	JILL TURNER	128 LOW LANE BROOKFIELD ,MIDDLESBROUGH,CLEVELAND ,TS5 8EE UNITED KINGDOM

PCT International Classification Number

B01J 23/75

PCT International Application Number

PCT/GB05/001780

PCT International Filing date

2005-05-10

PCT Conventions:

#	PCT Application Number	Date of Convention	Priority Country
1	0410408.9	2004-05-11	U.K.