GB2239406A - Catalytic gas conversion method - Google Patents

Abstract

A method for the catalytic partial oxidation of methane under relatively mild conditions of temperature, 650 - 900 DEG C, and pressure, 10 - 600 kPa, without the use of steam, gives a product comprising hydrogen and carbon monoxide with little or no steam and carbon dioxide. The catalyst includes a d-block transition metal preferably on a refractory support. Catalyst precursors include mixed metal oxides of formula substantially M2M'2O7 having the pyrochlore structure, where M may be selected from Mg, 8, Al, Ga, Si, Ti, Zr of Hf or a lanthanide and M' is a 'd' block metal.

Description

CATALYTIC GAS CONVERSION METHOD The diminishing reserves of petroleum oil have focused attention on the need to find alternative sources of carbonaceous materials and stimulated considerable interest in the possibility of making more effective use of the world's vast reserves of natural gas. At the present time, only a minor fraction of the available methane is being utilised. In the U.K., for example, it is used both as a fuel and as a feedstock, via steam reforming to synthesis gas (carbon monoxide and hydrogen), for methanol and ammonia synthesis, but in many parts of the world the collection and distribution of methane are uneconomical and it is burnt in situ to form carbon dioxide and water.

There are several known reactions for the oxygenation of methane.

Different catalysts promote these reactions to different extents, but selectivity is normally poor.

This patent application results from our discovery of a class of catalysts that is capable of selectively oxygenating methane to carbon monoxide and hydrogen.

Recently, attempts to convert methane directly into more valuable chemicals have focused on oxidative coupling reactions to yield ethylene and ethane, 1,2,3 and direct oxygenation to methanol and formaldehyde.4,5 Unfortunately, under conditions where the reactions of methane are fast enough to be of interest (typically > 700 C), the formation of CO2 is so favourable (t G < -800 kJ/mol) that partial oxidation to more useful products is difficult to achieve on an economical scale. The non-catalytic, gas-phase partial oxidation of methane to synthesis gas is an established industrial process (e.g. Shell, Texaco6), but operates at very high temperatures ( > 12000C).Synthesis gas mixtures are also formed in two step catalysed reactions using mixtures of methane, water and oxygen which operate at elevated pressures and temperatures in excess of 1000 C.6 This patent application results from our discovery of catalysts that are capable of selectively oxygenating methane to carbon monoxide and hydrogen so that the reaction can be carried out catalytically and at a substantially lower temperature (N7750C). The significance of this result lies in the fact that synthesis gas is a well established feedstock for the synthesis of higher hydrocarbons, alcohols and aldehydes, for example in 7 Fischer-Tropsch catalysis, thus facilitating efficient two-step processes for the conversion of methane to such materials.Equally, one possible application for synthesis gas produced at low pressures, is for use in fuel cell technology.

The overall reaction which is catalysed is:

and this reaction is often described as the partial oxidation of methane.

As noted above, synthesis gas can be made by a number of methods, most of which involve the steam reforming of hydrocarbons or coal.8 Methane, for example, can be converted over a nickel/alumina catalyst9 at 700-8000C, according to:

This reaction is an important source of carbon monoxide and hydrogen, but it is highly endothermic, and leads in addition to the formation of carbon dioxide via the water-gas shift equilibrium: CO + H20 = CO2 + H2. The partial oxidation reaction, by contrast, is mildly exothermic, more selective, and yields an H2/CO ratio that is lower than that obtained by steam reforming. This lower ratio may be highly desirable for certain applications of synthesis gas. Indeed, secondary reformers using CO2 or 0, 2 oxidants are frequently required to reduce the hydrogen content of synthesis gas made by steam reforming.

In FR 1595993, Chimigaz, there is described a method for the catalytic partial oxidation of methane to carbon monoxide plus hydrogen. But the temperatures of 1000 - 12000C were so high as to be uneconomic.

In EPA 303 438, Davy McKee Corporation, there is described a catalytic partial oxidation process for converting a hydrocarbon feedstock to synthesis gas.

The process described uses steam in addition to oxygen and runs at temperatures of 8700C to 10400C and a pressure of about 2760 kPa. Even under optimum conditions, conversion of methane to a product consisting essentially of hydrogen plus carbon monoxide in the substantial absence of steam and carbon dioxide is not achieved, i.e. the (H2 + CO) selectivity of the system is not very good. At lower temperatures and pressures, particularly when using low concentrations of steam, methane conversion and (H2 + CO) selectivity fall off and the catalyst becomes poisoned by carbon deposition.

We have discovered that it is possible to effect the catalytic partial oxidation of methane under relatively mild conditions of temperature (650 - 9000C) and pressure (preferably 10 - 600 kPa i.e. 0.1 - 6 Atmospheres) to give a product consisting essentially of H2 + CO (plus N2 if air is used as the oxidant). The molar H2:H2O and CO:CO2 ratios in the product gas are both at least 8:1. This is achieved without the need to use steam; steam inevitably generates CO2 and thus reduces the (H2 + CO) selectivity of the system.

This invention results for our discovery of such catalysts.

The invention provides a method of converting a reactant gas mixture comprising methane and oxygen in a molar ratio of at least 1.7:1 into a product gas mixture comprising H2, CO and optionally H2O and CO2 in which the H2:H2O ratio is at least 8:1 and the CO:CO2 ratio is at least 8:1, which method comprises bringing the reactant gas mixture at a temperature of 650 9000C into contact with a solid catalyst which is either: a) a d-block transition metal on a refractory support; or b) an oxide of a d-block transition metal; or c) a material of the formula Mum' 0 where: yz M is at least one element selected from Mg, B, Al, Ln, Ga, Si, Ti, Zr and Hf, Ln is at least one member of lanthanum and the lanthanide series of elements, M' is a d-block transition metal, and each of the ratios x/z and y/z and (x + y)/z is independently from 0.1 to 8; or d) a catalyst formed by heating b) or c) under the condition of the reaction or under nonoxidising conditions.

Each of the ratios x/z and y/z and (x + y)/z is independently from 0.1 to 8, preferably from 0.2 to 1.0. This definition covers a) particles of metals such as Ni, Ru and Pd on solid oxide supports such as Awl 203, MgO, SiO2 and Ln203 .

b) simple metal oxides such as NiO and RuO2; c) ternary mixed metal oxides such as Pr2Ru207; These metal oxide systems c) may be catalyst precursors, from which the active catalyst is formed by heating under non-oxidizing conditions. The catalyst precursor may decompose to yield the metal on the oxide support.

All these metal oxide systems may be crystalline, monophasic or polyphasic, they may be amorphous, they may be stoichiometric or nonstoichiometric; they may have defect structures. They may be solid solutions. The values of x, y and z may be integral or non-integral. In the mixed metal oxides, the ratio of x to y is not critical and may for example be from 0.001 to 1000.

Some of these mixed metal oxide catalysts are members of a structural class known as pyrochlores (E.

F. Bertant, F. Forrat and M. C. Montmory, Compt. Rend.

(Paris) 249, 829 (1959)).

d-Block transition metals M' are selected from those having atomic number 21-29, 40-47 and 72-79.

Preferably M' is selected from Fe, Os, Co, Rh, Ir, Pd, Pt and particularly Ni and Ru.

The oxide catalyst precursors may be prepared in a number of ways. Typically for M2M'207, an intimate mixture of two metal oxides in the required proportions is heated to a temperature and for a time sufficient to form a homogeneous phase. The heating temperature varies, depending on the nature of the metal oxides used; it is generally at least 25OOC, and may be as high as 1400 C. It is possible simply to mix two or more preformed metal oxide powders. More sophisticated mixing techniques are well known in the ceramic processing field, and are likely to be effective in reducing the time and/or temperature required to achieve homogeneity.

Alternatively the metal oxides, or precursors thereof such as carbonates or nitrates, or organometallic derivatives or any thermally decomposable salts, can be precipitated onto a refractory solid which may itself be massive or particulate. Or one metal oxide or precursor may be precipitated onto the other. For example, Ru3(CO)12 or Rut-indenyl)2 in toluene may be precipitated onto a powdered metal oxide (e.g. PrO2) or onto a high surface area form of metal oxides such as MgO, Al203, SiO2, ZrO2, TiO2, or HfO2. After drying, the coated material is heated, generally in air or oxygen, typically at temperatures up to 4500C, but possibly lower or higher (as described in experiment 19).

The catalyst precursor may need to be preactivated by being heated under conditions which are non-oxidizing or perhaps even reducing. For this purpose, it may be sufficient to pass a reactant gas mixture comprising methane and oxygen over the catalyst precursor, provided that the oxygen content is not too high. Alternatively, the catalyst precursor may be heated, prior to use, in an inert gas such as nitrogen or helium, perhaps together with methane or oxygen, or in a reducing gas such as hydrogen or methane. Or the catalyst precursor may have been formed by heating the mixed metal oxides, not in air or oxygen, but under non-oxidizing or reducing conditions.

In this specification, the usual terms hydrogen, nitrogen and oxygen are used (rather than dihydrogen, dinitrogen and dioxygen) to refer to the dimer molecules generally encountered.

Into contact with this solid catalyst is brought a reactant gas mixture including methane and oxygen. The catalyst performs well in the absence of water, but the presence of traces or small quantities of water vapour in the reactant gases is not deleterious. Alternatively, water may be added in which case the reactions will include those normally associated with steam reforming. If water is present in the reactant gas mixture, it is preferably in a smaller molar proportion than oxygen. Water is preferably not added to the reactant gas mixture. The reactant gas mixture is preferably substantially free of water.

The reaction conditions, in particular temperature, are sufficient to effect conversion into a product gas mixture comprising mainly hydrogen and carbon monoxide. The temperature may be from 650 to 9000C, particularly 700 to 8000C. As shown in experiments 2, 6, and 7, the selectivity to products CO and H2 is very dependent on temperature. Furthermore, the selectivity increases as the temperature increases.

So, though in all cases, all ( < 99.5%) of the oxygen is consumed, the CO:C02 ratio will increase from ca. 1:20 at 3770C to ca. 10:1 at 7770C. Similarly the H2 :H20 ratio will typically increase from ca. 1:4 to ca. 20:1 as the temperature is increased from 377 to 677 C.

Reaction pressure is preferably up to 5.0 MPa (50 Atm), particularly in the range 10 - 600 kPa (0.1 6 Atm). The selectivity of the reaction decreases with increasing pressure. Flow rates are reported below in the examples in terms of gas hourly space velocity (GHSV).

The condition for the gas mixture at any point in the system must obviously be chosen to be non-explosive at the temperature and pressure conditions desired. As is well known in the field, this may be achieved by introducing oxygen into methane (optionally with an inert carrier such as nitrogen) at a rate comparable to its rate of reaction with the methane so that the proportion of unreacted oxygen never rises to a dangerous value. Alternatively and less preferably, methane may be carefully introduced into oxygen or an oxygen-rich carrier gas, at a rate to prevent a dangerous build-up of methane concentration With these provisos, the overall molar ratio of methane to oxygen is preferably at least 1.7, e.g. in the range from 1.7 to 2.3, depending on the pressure.A slight stoichiometric excess of methane ( > 2:1 CH4:O2) may improve selectively by depressing formation of carbon dioxide and water; a slight stoichiometric excess of oxygen may improve the overall percentage conversion of methane. the optimum ratio may be a balance between these conflicting requirements.

It may be convenient to use oxygen in the form of air. In that case, it appears that the nitrogen simply acts as a diluent.

It may be noted that ruthenium and nickel catalysts on refractory supports are used commercially for the steam reforming of methane. We presently believe that steam may be formed as a transient intermediate in the partial oxidation of methane according to our method. On this basis, we expect that supported metal catalysts which are effective for steam reforming methane will also be effective for the method of this invention.

In our hands, the catalysts appear to retain their activity for long periods of time. No diminution in catalyst activity has been observed in reactions lasting for thirty hours. However, as indicated by its X-ray diffraction pattern, the material is not the same after use as it was before use. The nature of the change is not precisely known, but is believed to take place by a pre-activation step, which typically occurs during the first thirty minutes of use at temperatures ca. 1050K with an appropriate reactant gas mix. For example, XRD, AEM, XPS and HRTEM spectroscopic studies were made on samples of the pyrochlore Pr2 Ru 207 both before and after use in the catalytic conversion.

These data showed that after the catalytic reaction there were small particle of ruthenium metal on the surface of the mixed oxide support. We believe that these metal particles contain on their surface the active catalytic sites. The term catalyst is used herein to describe both the initial material placed in the reactor at the outset, and also the material(s) to which the initial material has been transformed as a consequence of pre-activation or of interaction with the reactant gas mixture.

Using the catalysts herein described, we have been able to achieve at least 90% and up to 98% conversion of methane to a product gas mixture in which the ratio H2:H2O was typically greater than 8:1 and often 20:1 or greater; and in which the ratio CO:CO2 was typically at least 8:1.

It is probable that the metal particles contain the active catalytic sites in many of the other catalysts.

Reference is directed to the accompanying drawings, in which each of Figures 1 to 3 is an X-ray powder diffraction pattern of a catalyst of formula substantially Pr2Ru207: Figure 1 shows the catalyst of experiment 2 before use, which contains a small stoichiometric excess of RuO2.

Figure 2 shows the catalyst of experiment 3 before use, which contains a small excess of PrO2.

Although the X-ray powder diffraction pattern of the "Pr2Ru207" catalysts of experiments 2, 3 and 4 show the presence of only traces of phases other than the cubic phase attributed to the pure "pyrochlore" structure (pure Pr2Ru2O7), quantitative measurements by analytical electron microscopy (AEM) show that a range of different oxide phases exist within the furnace product.

Figure 3 shows the catalyst of experiment 3 after use for several hours and in particular, the formation of ruthenium metal.

We have carried out an XPS study of the catalyst system "Pr2Ru207". The samples investigated were: sample A, "Pr2Ru207" formed by heating PrOx and Ru02 at 9000C for iOOh in air. Sample B, the material formed by heating the sample from A at 1050K in a flow of 02. N2 and CH4 under the conditions of experiment2.Sample C, the material formed by heating sample B in pure 2 (lOOmbar) at 900K for 10 minutes. Sample D, the material formed by heating sample C in methane (lOOmbar) at 900K for 10 minutes. Sample E, the material formed by heating sample D in methane (lOombar) at 1100K for 10 minutes. Sample F, pure Ru02 (Aldrich Chemical Co.) as used in the catalyst synthesis.

Typical XPS data are shown in Figures 4-6fand Tables 17-19.

The experiments showed the following properties of the surfaces of the samples A to E.

(i) Over a sampling depth of about 20A the ratio Pr/Ru is approximately the same (i.e. 1.15) for all samples. Table 15. The anomalies in B and E arise from the extra C is peak, only present in these two, which overlaps, and thus artificially increases, the Ru 3d peak. However, for the samples B, C, and D, which have been reacted, the Pr/Ru ratio over the first 7A depth show enrichment of Ru, Table 16. Therefore enrichment of the surface in ruthenium occurs during the initiation of the catalyst.

(ii) The peaks assigned to the Ru in the surface of the catalyst samples which had been reacted with methane are all shifted to lower binding energies than for the sample A, Table 17. This suggests the ruthenium is in a lower oxidation state in the activated catalyst. The Ru 3d peak occurs as a well-resolved doublet which is characteristic for ruthenium metal, as opposed to RuO2. This suggests that ruthenium segregating to the surface during the reaction with methane has been reduced to the metal. The Ru peak of the sample C reacted with oxygen has the appearance analogous to that of Ru02. This surface layer can be reduced again to ruthenium metal by heating under methane, sample D.

The samples B and E which have been exposed to methane at temperatures > 1000K showed substantial surface coverage of carbon, Figures 5,6. The binding energies for the carbon lie below those for carbon bonded to oxygen and above those normally associated with metal carbide species, but are typical for a thick layer of graphitic carbon.

EXPERIMENTAL All the experiments were carried out using 50mg of solid, powdered catalyst, lightly packed between < 20mg of silica wool (MULTILAB) in a straight silica reaction tube of i.d. ca. 4mm. The reaction tube (300mm) was placed in the vertical tube furnace of a LAB CON microreactor and connected to a supply of the gas reaction mixture. The reactant gases, methane (supplied by Union Carbide, Gas and Equipment Ltd.), dioxygen (supplied by Air Products) and dinitrogen (supplied by Air Products) were dried over molecular sieves and passed over the catalyst at a rate of 20-50 ml/min (GHSV of 4 - 7 x 104 hour1). The temperature of the reaction tube was raised from ambient to the required temperature (typically 1050K, unless otherwise stated) over a period of 2 hours.The reaction products were monitored using an on-line HewIett-Packard 5890A gas chromatography apparatus. Separation of all gases was obtained using Helium carrier gas through Porapak Q and sA molecular sieve packed columns, and were detected using a Thermal Conductivity Detector, calibrated on site. In all cases, O2 conversion was > 99.5%, and C, H, O, N mass balances were better than 96%.

The specific details of some experiments are given below.

Experiment 1 (comparative) Catalyst: Ru02 (pure, Aldrich Chemical Company Ltd.) Initial gas pressures ca.: CH4 217 torr 2 108 torr N2 435 torr Total 760 torr = 760 mm Hg = 1 atm = 0.1 MPa GHSV = 7 x 104 hour'l Table 1 Product Partial Pressures/torr Temp/K CH4 converted CO CO2 H2 H20 1050 60% 88 21 165 52 Duration of experiment 4h.

In this prior art method, 60% conversion of methane was achieved, with ratios of CO:C02 and H2:H20 below 5:1.

Experiment 2 The effect of variations of temperature on the conversion of methane and selectivity towards CO and H2.

Catalyst Pr2Ru207 (prepared from an intimate mixture of Pr6011 (0.290g) and + 6Ru02 (0.223g) in an open crucible for 100 hours at 9000C). The X-ray powder diffraction pattern, Figure 1, shows slight excess of RuO2 in the sample.

Reactant gas partial pressures were maintained at ca.: N2 434mm Hg CH4 216mmHg - 2 108mm Hg GHSV maintained at 7 x 104 houff Experiments in chronological order.

Table 2 Temp Co CH4 Product gas partial pressures K converted mm Hg H2 CO CO2 H20 975 48% 138 39 53 46 1000 55% 166 56 46 38 1020 78% 245 108 22 15 1030 82% 258 115 19 11 1040 84% 263 120 16 10 1050 87% 272 126 14 7 It is apparent from Table 2 that increasing the catalyst temperature increased the conversion of methane and also increased the CO:CO2 and H2:H2O ratios in the product gas.

Experiment 3 Catalyst preparation: 0.112g Ru02 and 0.155g Pr02 were intimately mixed in an agate mortar with pestle, and the resulting mixture transferred to can open porcelain crucible and heated in a Gallenkamp muffle furnace at 9500C for 100 hours.

The black solid thus obtained was then put on an Aluminium plate in an X-ray powder diffractometer and its spectrum recorded, Figure 2.

50mg of the catalyst was then lightly packed between < 20mg silica wool (Multilab) in a 4mm i.d. vertical silica tube, held in a steel block heated to 1050K. CH4 and 2 (2:1) passed for 3 hours with poor selectivity and conversion.

Nitrogen was passed over the catalyst for 12 hours at this temperature, after which time the gas mixture was changed to a mixture of N2, CH4 and 02. The products were analyzed after allowing ca. 30 minutes initiation time.

A GHSV of 4 x 104 hour'l was used throughout these experiments, and the temperature was maintained at 1050K.

The combined partial pressures of the reaction mixture of gases was one atmosphere throughout this experiment.

These experimental data given in Table 3 were carried out over a period of 30 hours, during which time no change in catalyst activity or selectivity was observable. An X-ray diffraction pattern of the catalyst after use is shown in Figure 3. By comparison with Figure 2, it appears that the catalyst has been significantly modified.

Table 3 Effect of variation of partial pressures of the reactant gases.

Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm Hg 02 converted mmH N2 CH4 2 O2 converted H2 COC2 H20 427 227 110 2.15 90% 297 141 3 2 394 248 118 2.10 91% 311 152 5 1 409 236 115 2.06 93% 302 149 6 2 424 225 111 2.03 94% 300 144 6 2 433 216 110 1.96 95% 288 141 6 5 433 213 114 1.86 96% 289 137 10 7 438 204 118 1.72 98% 306 136 14 14 0 511 248 2.06 88% 474 228 15 9+ 0 536 224 2.39 77% 452 222 8 11 0 587 173 3.39 56% 392 200 2 4 + This experiment ran for 14 hours with no change in activity or selectivity.

The data in Table 3 shows that an increase in the proportion of oxygen relative to methane gives a greater overall conversion of methane, up from 88% to 98% or greater. However this is accompanied by a small reduction in selectivity as indicated by the increasing proportions of C02 and H20 in the products. In this table, the CO:C02 ratios range from approximately 10 to 50; and the H2:H2O ratios range from approximately 20 to 150.

Experiment 4 The same "Pr2Ru207" catalyst as used in experiment 3 was tested at higher pressures.

Experiments at elevated pressures of reactant gases showed conversions and selectivities comparable to those found at 1 atm could be achieved.

Table 4 Total P Initial pressures CH4 CH4 Product partial pressures bar mm g 72 converted mm Hg N2 CH4 2 H2 CO CO2 H20 1.0 0 571 189 3.02 64% 420 211 2 6 1.0 0 587 173 3.39 56% 392 200 2 4 20.0 3480 9650 2060 4.68 28% 3640 1750 457 610 20.0 1080 12110 2000 6.04 23% 3820 1800 434 482 20.0 0 12750 2440 5.22 26% 4160 1940 577 666 Lower CH4 conversion reflects the excess of CH4 over stoichiometry.

Selectivities to CO and H2 were slightly lower at 20 atm. A trace of C2 products was observed.

Experiment S Catalyst Gd2Ru207, prepared by heating Gd203 + 2 Ru02 in a sealed evacuated silica tube at 1 1000C for 100 hours. The X-ray powder diffraction pattern shows that the sample is solely the pure pyrochlore, Gd2Ru207 .

Table 5 Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm Hg G2 converted mm Hg N2 CH4 2 H2 CO CO2 H20 421 229 110 2.07 92% 304 149 4 3 422 226 111 2.03 93% 293 144 6 4 425 222 113 1.97 94% 291 143 6 6 423 222 115 1.93 95% 292 143 8 6 0 512 248 2.07 83% 451 221 18 21 Experiment 6 The same Gd2Ru207 catalyst as used in experiment 5 was tested under conditions of variable temperature. The experiments are in chronological order.

The reactant gas partial pressures were kept constant at: CH4 512 mm Hg 2 248 mm Hg (CH4/02 2.07) Table 6 Temp CH4 Product partial pressures converted mm Hg H2 CO CO2 H20 1050 83% 451 221 18 21 900 57% 324 134 74 71 950 66% 376 169 51 50 1000 73% 412 193 36 36 This experiment shows how selectivity decreases as the temperature is lowered from 1050 K.

Experiment 7 Catalyst Eu2Ru207, prepared by heating Eu203 + 2 RuO2 in a sealed evacuated silica tube at 10000C for 100 hours. The X-ray powder diffraction pattern shows that the sample is mostly the pure pyrochlore, Eu2Ru207, but there are traces of other phases present.

The catalyst was tested under conditions of variable temperature.

The experiments are in chronological order.

The reactant gas partial pressures were kept constant at: CH4 512 mm Hg 02248 mm Hg (CH4/02 2.07) Table 7 Temp CH4 Product partial pressures K converted mm Hg H2 CO CO2 H20 1050 87% 466 229 12 14 1000 80% 434 214 25 26 1020 83% 449 222 20 21 1040 86% 463 227 15 15 1050 87% 467 230 13 13 Experiment 8 A dispersion of Ruthenium on Alumina was prepared by supporting Ru(tl5-CgH7)2 on predried A1203, using an incipient wetness technique with CH2C12 solvent. The uniformly yellow solid was then reduced under a stream of H2 at 2000C for 1 hour and at 8000C for 4 hours.

Table 8 Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm Hg 52 converted mm Hg N2 CH4 2 H2 CO C 2 H20 389 254 117 2.17 89% 311 155 3 2 391 247 122 2.02 93% 312 154 6 4 O 512 248 2.06 87% 467 229 12 16 Experiment 9 Englehard E catalyst (4871), was tested as recieved from Englehard.

The catalyst is 0.5% Ruthenium on Alumina pellets. The pellets are cylindrical, ca. 3.5 mm high and ca. 3.5 mm diameter. 20 such pellets were loaded into a silica tube of ca. 8 mm i.d. (packed at either end with silica wool), which was then heated to ca. 1050K (+ 15K), and connected to a gas stream of the appropriate reactant gases. The GHSV was maintained at ca. 104 hours (+ 20%).

Table 9 Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm lig (92 converted mm Hg N2 CH4 O2 H2 CO CO2 H20 418 240 101 2.37 78% 271 133 4 7 424 234 102 2.29 79% 266 130 6 9 426 226 108 2.09 86% 274 134 9 9 287 321 152 2.11 86% 348 172 10 9 0 634 126 5.02 37% 314 162 3 7 0 508 252 2.01 83% 448 197 23 27 Experiment 10 A 1% by weight loading of Ruthenium on Alumina was prepared: 0.500g F20 Alumina (Phase Sep), 80 - 100 mesh, was predried in a muffle furnace at 2000C for 12 hours and 8000C for 24 hours. 0.0 lOg anhydrous Ruthenium trichloride was dissolved in the minimum amount of methanol, predried over magnesium turnings.Aliquots of the solution were added to the alumina under a dry nitrogen atmosphere, each time until the alumina was just "wet", then the solvent was removed under reduced pressure at room temperature. In this way, a completely uniform dispersion of RuCl3 on A1203 was obtained. The solid was then transferred under dry nitrogen into a silica tube, which was then attached to a hydrogen supply, and the solid treated under a hydrogen stream at 4000C for 14 hours, and at 8000C for 4 hours. The chlorine content of the catalyst was tested by microanalysis, and it was found that the Ruthenium was only partly reduced, since there is some 0.5% residual chlorine.

Table 10 Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm Hg 2 converted mm Hg N2 CH4 2 H2 CO CO2 H20 433 222 104 2.13 91% 286 142 3 2 421 229 110 2.09 92% 295 146 3 3 428 223 109 2.04 94% 294 145 4 3 432 217 110 1.97 95% 290 143 6 4 0 512 248 2.07 88% 473 233 11 12 0 514 250 2.04 89% 474 232 12 t This experiment run for 24 hours, with no noticeable change in methane conversion or product selectivity.

Experiment 11 A 0.1% by weight loading of Ruthenium on Alumina was prepared: l.SOg F20 Alumina (Phase Sep) 80 - 100 mesh was dried in a muffle furnace for 12 hours at 2000C and for 24 hours at 8000C. 0.003g anhydrous RuC13 was added to the alumina under dry nitrogen, dissolved in the minimum volume of dry methanol to just "wet" the alumina completely. The methanol was then removed under reduced pressure at room temperature. The uniform dispersion thus produced was then reduced under a steady hydrogen stream for 14 hours at 4000C and for 10 hours at 8000C. The chlorine content of the catalyst was tested by microanalysis and it was found that almost all of the chlorine still remains in the catalyst.This is reflected in the results of passing the N2/CH4/02 gas mixture over the catalyst, as the results are far from steady statc.

A gas mixture containing 391 torr N2, 251 torr CH4 and 118 torr 2 (CH4/02 = 2.12) was passed over the catalyst at 1050K, and 20 mI/min (GHSV = ca. 2 x 104 hour~ 1) .

Table 11 Time % CH4 Product partial pressures hour converted mm Hg H2 CO CO2 H20 1 34 55 28 52 95 4 42 110 53 43 72 11 52 162 79 34 55 15 59 190 94 29 42 19 64 214 105 24 35 24 67 227 110 22 31 26 69 232 115 19 30 The catalyst improves over the first 26 hours on stream. In general, this is clearly a much less efficient catalyst than any previously tested. This suggests that there may be an insufficient amount of Ruthenium for formation of an active and selective catalyst.

Experiment 12 A 1% by weight loading of Rhodium on Alumina was prepared, as in experiment 10, except with RhC13 instead of RuC13. Microanalysis showed that there was some 0.8% residual chlorine.

Table 12 Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm Plg 52 converted mm eg N2 CH4 2 H2 CO CO2 H2O 450 210 100 2.09 92% 273 139 4 4 428 223 109 2.04 94% 287 140 5 5 Experiment 13 A 1% by weight loading of Palladium on Alumina was prepared, as in experiment 10, with a few differences. The PdC12 used was obtained from Johnson Matthey Plc, and was not pure, but was supplied with an accurate metal assay of 59.87% Pd. 0.017g of this was dissolved in the minimum volume of concentrated HCl, and added, via an incipient wetness technique, to 1 .000g of predried alumina. This was then treated under flowing hydrogen as before.

Table 13 Reactant gas partial pressures C114 CH4 Product gas partial pressures mm Hg converted mm Hg N2 CH4 2 H2 CO C02 H20 435 221 104 2.13 90% 274 143 4 2 442 213 105 2.03 92% 269 140 6 4 432 219 109 2.01 93% 277 144 6 5 Experiment 14 A commercial Nickel-based steam reforming catalyst (CRG'F', approximate elemental composition Ni6Al2O9) was obtained from the London Research Station, British Gas Plc. The pellets were crushed to a powder in an agate mortar with pestle, in air.

Table 14 Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm s v 2 convertec mm r g N2 CH4 2 H2 CO CO2 H20 420 229 111 2.07 93% 290 149 5 4 437 213 110 1.93 96% 277 143 8 6 0 516 244 2.11 88% 475 234 9 9t t This experiment was run for 16 hours with no detectable loss of activity or selectivity.

Experiment 15 Catalyst La2MgPtO6, prepared by heating La203, MgO and PtO2 in air at 1473K for 200 hours. The catalyst has a perovskite structure as checked by powder X-ray diffraction. The catalyst performance gradually improved up to 20 hours under the flowing reactant gas mixture, when the following data was taken.

Table 15 Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm Hg ' converted mm Hg N2 CH4 2 H2 CO CO2 H20 423 225 112 2.02 80% 241 125 15 18 Experiment 16 Catalyst NiO (BDH Chemicals Ltd.), calcined in air at 1150K for 50 hours.

Table 16 Reactant gas partial pressures CH4 CH4 Product gas partial pressures mm Hg O2 converted mm Hg N2 CH4 2 H2 CO CO2 H2O 442 218 100 2.18 72% 212 110 16 18 Table 17 XPS Pr 4d/Ru 3d peak height ratios, see Figure 4 Sample Pr/Ru A as prepared 1.16 B catalytic 3.83 C O2/900 K 1.14 D CH4/900 K 1.15 E CH411 100 K 1.36 Table 18 XPS Ru M4 5VV/Pr 3d 5/2 peak height ratios, see Figure 5 Sample Height (Ru M4,5VV)/height (Pr 3d 512) A As prepared 0.12 C O2/900K 0.22 D CH4/900 K 0.24 Table 19 XPS Peak energy shifts (eV), see Figures 5,6 Sample Ru 3d Ru 3p Pr4d Ols F Ru02 280.8 462.8 - 529.7 A As prepared +0.9 +0.8 116 -0.5/+1.1 B catalytic -0.9 -1.2 0 +1.7 C O2/900 K -0.5 0 0 -0.1/+1.5 D CH4/900 K -0.9 -1.3 +0.1 -0.6/+2.0 E CH4/1100K -0.9 -1.5 +0.3 +0.1 Experiment 17 Table 20 Showing that a number of different supported metals and mixed metal oxides are efficient catalysts for the partial oxidation reaction.

Gas feed, CH4:O2:N2 = 2:1:4. GHSV = 4x104 hour-1. Temp = 1050 K.

% Methane % CH converted to Catalvst converted CO 1% Ru/Al203 94 97 99 1% Rh/A12O3 94 97 99 1% Pd/Al203 93 96 98 1% Pt/A12O3 95 96 99 Ni/Al2O3 94 97 99 Pr2Ru207 94 97 99 Eu2Ir2O7 94 96 98 La2MgPtO6 80 89 93 CRG 'F' steam reforming catalyst, ex. British Gas Experiment 18 Table 21, showing how the methane conversion and selectivity to partial oxidation products vary with methane:oxygen ratio.

Gas feed contains approx. 57% N2 diluent. GHSV = 4x104 hour-1.

Temp = 1050 K. Catalyst = Pr2Ru2O7.

% Methane % CH4convened to CH,/O, ratio converted CO 1.72 98 91 96 1.86 96 93 98 1.96 95 96 98 2.00 94 97 99 2.06 93 97 99 2.10 91 97 100 2.15 90 98 100 Experiment 19 Table 22, showing how the methane conversion and selectivity to partial oxidation products varv with reaction pressure.

No N2 diluent. GHSV = 4x104hour-1. Temp = 1050 K.

Catalyst = Dy2Ru2O7 Pressure % Methane % CH4 converted to atm. CHAlO ratio convened CO H2 1 3.4 56 99 100 5 3.8 39 91 91 10 4.1 38 90 90 15 4.5 33 86 88 20 4.5 30 85 88 Experiment 20 Table 23,showing how the methane conversion and selectivity to partial oxidation products varv with reaction temperature.

Gas feed, CH4:02:N2 2:1:0. GHSV = 4x104 hour-1.

Catalyst = Yb2Ru2O7 Temperature % Methane % CH converted to K converted CO 1050 83 91 95 900 53 59 79 800 39 28 59 700 31 8 33 650 29 5 25 Experiment 21 Table 24, showing how the methane conversion and selectivity to partial oxidation products vary with reactant gas space velocity.

Gas feed, CH4:O2:N2 = 2:1:0. Temp = 1050 K. Catalyst = Pr2Ru2O7 GHSV % Methane % CH4 converted to hour converted CO 6x 103 90 95 98 4x 104 88 93 97 6 x 104 81 89 94 8 x 104 73 82 90 2x 105 68 76 87 References.

1. Keller, G.E. & Bhasin, M. M. J. Catal. 73, 9-19 (1982).

2. Hutchings G. J., Scurell M. S. & Woodhouse, J. R. Chem. Soc.

Rev. 18, 251-283 (1989).

3. Ashcroft, A.T., Cheetham, A.K., Green, M. L. H., Grey, C.P. & Vemon, P. D. F. J. Chem. Soc. Chem. Commun. 21, 1667-1669(1989).

4. Gesser, H. D., Hunter, N. R. & Prakash, C. B. Chem. Rev. 85, 235 244 (1985).

5. Spencer, N.D. & Pereira, C.J. J. Catal. 116, 399-406 (1989).

6. "Encyclopedia of Chemical Technology", Ed. Kirk. R.E. and Othmer, D.F. 3rd Edition, Wiley Interscience, N.Y., 1980, Vol. 12, 952.

7. For example, Henrici-Olivé, G. & Oliv, S. Angew. Chem. Int. Ed.

Eng. 15, 136-141(1976).

8. "Catalysis in C1 Chemistry", Ed. Keim, W., D. Reidel Publ. Co., Dordrecht, 1983.

9. Rostrup-Nielsen. J. R. in "Catalysis, Science & Technology, Vol.5" (ed. Anderson, J.R. & Boudart, M. (Springer, Berlin 1984).

and Topp-Jdrgensen. J. in "Methane Conversion" (ed. Bibby, D. M., Chang, C. D., Howe, R.F. & Yurchak, S.) Elsevier, 1988, p. 293.

Claims (12)

1. A method of converting a reactant gas mixture comprising methane and oxygen in a molar ratio of at least 1.7:1 into a product gas mixture comprising H2, CO and optionally H2O and CO2 in which the H 2:H2O ratio is at least 8:1 and the CO:CO2 ratio is at least 8:1, which method comprises bringing the reactant gas mixture at a temperature of 650 - 9000C into contact with a solid catalyst which is either: a) a d-block transition metal on a refractory support; or b) an oxide of a d-block transition metal; or c) a material of the formula MxM' y0z where: M is at least one element selected from Mg, B, Al, Ln, Ga, Si, Ti, Zr and Hf, Ln is at least one member of lanthanum and the lanthanide series of elements, M' is a d-block transition metal, and each of the ratios x/z and y/z and (x + y)/z is independently from 0.1 to 8; or d) a catalyst formed by heating b) or c) under the condition of the reaction or under nonoxidising conditions.

2. A method as claimed in Claim 1, wherein the reactant gas mixture is at a pressure of 10 - 600 kPa.

3. A method as claimed in Claim 1 or Claim 2, wherein water is not added to the reactant gas mixture.

4. A method as claimed in any one of Claims 1 to 3, wherein the molar ratio of methane to oxygen in the reactant gas mixture is from 1.7:1 to 2.3:1.

5. A method as claimed in any one of Claims 1 to 4, wherein M is Ln.

6. A method as claimed in any one of Claims 1 to 5, wherein M' is selected from Fe, Ru, Os, Co, Rh, Ir, Ni, Pd, and Pt.

7. A method as claimed in any one of Claims 1 to 6, wherein the formula of the material c) is substantially M2M'207, e.g. Sm2Ru2O7 and Pr2Ru2O7.

8. A method as claimed in any one of the Claims 1 to 6, wherein the catalyst is a metal M' supported on a refractory oxide of an element M.

9. A method as claimed in any one of Claims 1 to 8, wherein the catalyst has been activated under the conditions of the catalytic reaction or under nonoxidizing conditions.

10. A method as claimed in any one of Claims 1 to 9, wherein the reactant gas mixture contains nitrogen.

11. A method as claimed in any one of Claims 1 to 10, wherein each of the ratios x/z and y/z and (x + y)/z is independently from 0.2 to 1.00.

12. A method as claimed in any one of Claims 1 to 11, wherein at least 90% of the methane is oxidised.