WO2007093825A1

WO2007093825A1 - Fischer-tropsch catalyst preparation

Info

Publication number: WO2007093825A1
Application number: PCT/GB2007/050058
Authority: WO
Inventors: Clive Derek Lee-Tuffnell
Original assignee: Compactgtl Plc
Priority date: 2006-02-14
Filing date: 2007-02-13
Publication date: 2007-08-23
Also published as: GB0602922D0

Abstract

A catalyst for Fischer-Tropsch synthesis comprises cobalt catalyst metal on a ceramic support. It can be made by (a) contacting a ceramic support with a solution of a cobalt salt; (b) drying and then calcining, to form cobalt oxide; (c) in the presence of a reducing atmosphere comprising hydrogen, raising the temperature gradually to about 375°C, so the cobalt oxide undergoes reduction; (d) gradually oxidising cobalt back to cobalt oxide at an elevated temperature; and (e) again reducing the cobalt oxide to cobalt using hydrogen and the same reducing temperatures as previously. During the first reducing step the temperature is held at a value above the start of the second cobalt oxide reduction peak on a TGA trace. Ruthenium as a promoter can be deposited along with the cobalt, by combining a ruthenium salt with the solution of the cobalt salt.

Description

FISCHER-THOPSCH CATALy-JT PREPARATION

This invention relates to a method of preparing a catalyst suitable for use in a Fischer-Tropsch reactor, and to a catalyst made by this method.

The use of a metal such as iron, cobalt or nickel as a catalyst for a Fischer-Tropsch synthesis is well known. For this synthesis reaction the catalyst would typically be augmented by a promoter such as ruthenium, rhenium or platinum. These metals can be deposited onto a ceramic support by contacting the ceramic with a solution of a metal salt, calcining it to form an oxide, and then reducing it to metal . However the activity of the catalyst has also been found to depend on how it is made, because this can affect the size and distribution of the metal crystallites.

According to the present invention there is provided a method of preparing a Fischer-Tropsch catalyst comprising the steps of: a) contacting a ceramic support with a solution of a salt of catalytic metal, drying and then calcining, to form the metal oxide, wherein the metal of the resulting metal oxide consists primarily of cobalt; b) reducing the metal oxide to metal by heating in the presence of a reducing atmosphere including hydrogen, the heating process being such as to ramp up the temperature to a maximum temperature below 42O⁰C and above a TGA peak by no more than 30 K over a period of several hours, this temperature ramp being no more than l°C/min while reduction is occurring and including holding the temperature for at least an hour at a temperature above the start of the said TGA peak; c) subjecting the resulting metal to an oxidation process at an elevated temperature, by replacing the reducing atmosphere by an atmosphere containing an oxidising agent, and increasing the proportion of the oxidising agent gradually over a period of hours, to ensure formation of metal oxide,- d) and then subjecting the resulting metal oxide to a reduction process at temperatures less than or up to substantially the said maximum temperature so the metal oxide is again reduced to metal .

More preferably, the method comprises the steps of: a) contacting a ceramic support with a solution of a salt of catalytic metal, drying and then calcining, to form the metal oxide, wherein the metal of the resulting metal oxide consists of cobalt and less than 2% by weight of ruthenium; b) in the presence of a reducing atmosphere comprising hydrogen, raising the temperature gradually over a period of at least 2 hours to a temperature no more than 125⁰C to ensure the metal oxide is thoroughly dry, and then gradually ramping the temperature to a maximum temperature less than 42O⁰C which is between 20 and 30 K above a second cobalt oxide reduction TGA peak over a period of at least about 20 hours, this temperature ramp being no more than l°C/min while reduction is occurring and including holding the temperature for at least 3 hours at a holding temperature above the start of the second cobalt oxide reduction TGA peak, and then holding the temperature at the maximum temperature for at least 3 hours so the metal oxide undergoes full reduction,- c) while remaining at the maximum temperature, flushing out the reducing atmosphere and introducing a gas containing oxygen, increasing the proportion of oxygen gradually over a period of at least 3 hours, and then holding the proportion of oxygen for a period of at least 6 hours, to ensure formation of the metal oxide,- d) then flushing out the oxygen-containing gas, lowering the temperature to the holding temperature, introducing a reducing atmosphere, and gradually ramping the temperature up to the maximum temperature, and then holding the temperature at that value for at least 3 hours so the metal oxide again undergoes reduction. The appropriate temperatures at which the temperature should be held during the reduction and oxidation steps are preferably determined by monitoring changes of weight during a thermogravimetric analysis (or thermal gravimetric analysis) (TGA) on a specimen of the ceramic support containing the same metal oxide, in a reducing atmosphere.

Preferably the reducing atmosphere in steps (b) and (d) contains no more than 10% hydrogen during the initial stages of reduction; and the concentration of hydrogen may be increased during the later stages of reduction. The concentration of hydrogen may be above 90% whilst the temperature is held at the maximum temperature. The pressure of the reducing atmosphere may be above atmospheric at least during the later stages of reduction, for example at 3 barg. Alternatively the reducing atmosphere may be at atmospheric pressure and at no more than 10% hydrogen throughout the reduction process .

Preferably the temperature ramp in the first reducing step (b) also includes holding the temperature at a temperature above that of the start of the first cobalt oxide reduction TGA peak.

During the reduction process the temperature is preferably raised at no more than l°C/min, for example 0.5 or 0.2°C/min.

The reducing atmosphere is preferably arranged to flow continuously over the substrate, preferably with a space velocity of at least 6000 /hr, more preferably about 8000 /hr. This has the benefit of preventing the development of hot-spots, and also removing water vapour (formed by the reduction process) , so suppressing the formation of aluminates and oxides and hydrothermal ageing of the support if the ceramic comprises alumina. The space velocity, in this specification, is defined as the volume flow rate of the gases supplied to a chamber containing the ceramic support (measured at STP) , divided by the void volume of the chamber .

Preferably the ceramic is in the form of a coating on a metal substrate. This ensures a well-defined thickness in which the active catalyst metal will be evenly distributed and dispersed. For example it might be a 100 μm thick coating on each surface of a metal foil. The foil provides mechanical support and also is a heat conductor. Preferably the metal substrate is a steel alloy that forms an adherent surface coating of aluminium oxide when heated, for example an aluminium-bearing ferritic steel such as iron with 15% chromium, 4% aluminium, and 0.3% yttrium (eg Fecralloy (TM)) . When this metal is heated in air it forms an adherent oxide coating of alumina, which protects the alloy against further oxidation and against corrosion. The substrate may be a wire mesh or a felt sheet, which may be flat, corrugated or pleated, but the preferred substrate is a thin metal foil for example of thickness less than 100 μm (prior to oxidation) . The preferred ceramic is alumina.

Use of a planar catalyst structure combining the catalyst metal (for example, cobalt) doped into a thin film ceramic support (typically between 20 and 150 μm thick, coated on a thin metal foil, felt, foam or sintered structure typically between 50 and 150μm thick) leads to a number of benefits. During the reduction phases it is critical to minimise the partial pressure of water vapour, to avoid the formation of undesired cobalt compounds that are resistant to reduction to metal; the formation of such compounds is promoted by high water vapour partial pressure and by high local temperatures. The cobalt oxide reduction process is driven by the diffusion of hydrogen gas into the pores of the ceramic, and the counter-diffusion of the water vapour produced by the reduction reaction out of those pores. The use of a thin ceramic layer enhances the counter-diffusion process, resulting in lower local concentrations of water vapour and higher concentrations of hydrogen.

An additional benefit arises from the fact that the metal substrate combined with the thin layer of ceramic support enhances the heat transfer to and from the sites at which reduction or oxidation are occurring, so that substantially isothermal conditions can be maintained within the ceramic support during all the steps of the process. This leads to a more uniform distribution of cobalt crystallites within the catalyst, and it also allows a shorter process time than would be necessary with a more conventional pellet or spherical geometry. Thus this planar geometry provides a particularly effective heat and mass transfer environment for the chemical reactions, so that higher temperature ramp rates can be employed than with a conventional catalyst geometry, and that lower reductant and oxidant gas pressures and concentrations can be employed, because the diffusion path lengths are shorter.

The weight of the cobalt is preferably between 20% and 45% that of the ceramic in which it is distributed, which in the case of a thin alumina coating on a metal foil would be the weight of the alumina coating. Most preferably it is between 25% and 40%, for example 30%. There should also be a promoter combined with the catalyst metal, and this is required in a much smaller proportion, typically no more than a tenth the amount of the catalyst metal . The preferred promoter is ruthenium, and in this case its weight may be only a hundredth that of the catalyst metal . The promoter is preferably co- deposited with the catalyst metal, by mixing a small proportion of a salt of the promoter with the solution of the catalyst metal salt. Consequently the promoter is subjected to calcination and then reduction along with the catalyst metal . In one option the contacting, drying and calcining (step (a) ) is carried out on ceramic material in the form of a powder; the powder is then milled to a suitable particle size of say 10 or 20 μm (if necessary) , combined with a suitable binder, and coated onto a metal substrate prior to the subsequent reduction, oxidation and reduction steps.

The reduction/oxidation/reduction process of the invention produces a highly active catalyst, for example having an activity up to 50% greater than that obtained with a single reduction step.

The invention will now be further and more particularly described, by way of example only, and with reference to the accompanying drawings, in which:

Figure 1 shows graphically a temperature reduction profile obtained by thermogravimetric analysis, in a flow of a reducing atmosphere; and

Figure 2 shows graphically a temperature reduction profile obtained in the same way as that of figure 1, on a different specimen.

The following example is of a cobalt catalyst with a ruthenium promoter, in a porous gamma-alumina coating on a Fecralloy foil. The foil is corrugated, and is intended for use in a flow channel in a compact catalytic reactor for performing Fischer-Tropsch synthesis. The process is described in relation to one foil, but in practice you would usually make several foils at once.

1. Production of Ceramic Layer Containing Cobalt Oxide The annealed and shaped foil is heated to 95O⁰C in air to ensure it has an oxide surface. It is then cooled back to room temperature, cleaned with water, and then degreased for example with acetone. An alumina suspension is made by blending 25 g dispersible alumina (e.g. Dispal 18N4-80 (TM)) with 700 g water to form a sol in a Silverson mixer with a high shear head; a small amount of 10% aqueous ammonia is added to raise the pH to between 8.5 and 9.0, at which pH it gels, becomes white and forms a smooth viscous paste,- then 270 g of gamma- alumina particles (of average size 30 μm) (e.g. Catal (TM) 601 FC), surface area 140 m²/g, containing about 3% lanthana) are added, mixed and well sheared. The weighed foil is heated to about 55O⁰C, and is repeatedly and lightly sprayed with the well -mixed alumina suspension, to form a coating about 120 μm thick on each side. It is then cooled down to room temperature again.

The alumina is then dehydroxylated (calcined) by putting the coated foil in a furnace through which dry air is passed, and heated at l°C/min up to 100⁰C, held for 4 hours at this temperature, and then heated at the same rate up to 55O⁰C, and held for another 4 hours. It is then cooled back to room temperature. The alumina adheres tenaciously to the foil. At this point the coated foil can be weighed, to determine the mass and final thickness (about 80 μm) of alumina.

A mixed salt solution is made by combining 212.6 g of cobalt nitrate hexahydrate with a 1.7 g of ruthenium

(III) 2, 4-pentanedionate (equivalent to a Co: Ru ratio of 100:1), and making up to 750 ml with acetone. Using a microlitre syringe, 25 microlitres of the solution is added dropwise to each 0.01 g of alumina, ensuring that it spreads evenly. On reduction this will provide about 10% (weight) cobalt. The foil is then dried at 8O⁰C in a forced air oven for up to about 4 hours, until it appears dry on the surface. It is then placed in a tube furnace, through which heated dry air is passed at a space velocity of about 3000 /hr, and dried and calcined. This is achieved by raising the temperature of the air and furnace at l°C/min to 12O⁰C and holding the temperature at that for 16 hours,- and then raising the temperature at the same rate to 25O⁰C, and holding it for 6 hours. This ensures that the salts break down, to give the oxides. It is then allowed to cool. Introduction of only a small proportion of cobalt at any one time, and use of a solvent with a comparatively low surface tension, ensures that cobalt particles with a narrow size distribution are deposited within the pores of the alumina.

The proportion of catalytic metal in the ceramic is preferably arranged to be between 28% and 45%. In this example, the addition of the mixed salt solution, drying, and calcining are repeated three times, so the expected amount of cobalt is about 30% of the weight of the alumina. Preferably the foils are installed in flow channels of a Fischer-Tropsch reactor at this stage, the subsequent steps being performed in the reactor, or the foils may be put in a stainless steel tube.

Before reducing the cobalt oxide, a specimen of the alumina is removed at this stage, and is subjected (in the form of a powder) to a thermogravimetric analysis, heating it at for example 0.5°C/min up to 700⁰C in an atmosphere of argon with 5% hydrogen, while monitoring any changes of weight. Referring to the figures, there are shown resulting differential thermogravimetric graphs as observed with two different specimens. The graphs show the percentage change of mass for every degree rise in temperature (i.e. the rate of change of mass, as a percentage of the total mass, plotted against temperature) , and provides detailed information as to the temperatures at which the two major stages in the reduction occur. In each case there is a small peak P, which is understood to correspond to removal of hydroxyl groups and residual nitrate, this being observed at a temperature in the range 160-180⁰C. There are then two successive larger thermal peaks: the first thermal event, the peak marked A, is believed to correspond to the reduction of Co₃O₄ to CoO; this is referred to as the first cobalt oxide reduction peak. In the example of figure 1 the graph shows that this stage starts at about 16O⁰C, and occurs most vigorously at about 182⁰C, whereas in the example of figure 2 this stage starts at about 18O⁰C and occurs most vigorously at about 200⁰C. The second significant thermal event, the peak marked B, corresponds to the reduction of CoO to Co metal; this is referred to as the second cobalt oxide reduction peak. In the example of figure 1 the graph shows that this stage starts at about 21O⁰C and occurs most vigorously at about 35O⁰C, whereas in the example of figure 2 this second stage starts at about 26O⁰C and occurs most vigorously at about 37O⁰C. After completion of these reduction stages, there are no significant further thermal events.

It will be appreciated that this test need only be carried out if the position and extent of the reduction events are not already known. The shape of the graph, that is to say the characteristic positions, magnitude and shapes of the reduction maxima A and B, is different for different proportions of cobalt and ruthenium on the alumina, and may also be affected by the pH of the alumina dispersion that was deposited, and by the method by which cobalt was impregnated into it (such as the solvent and the organic or inorganic cobalt salt used) . For different foils produced from the same batch of alumina dispersion, and with the same proportions of cobalt and ruthenium deposited in the same way, the graphs can be assumed to also be the same. It will be appreciated that the deposition process described above may be varied for example by using a solution of a different cobalt salt in a different solvent.

2. First Reduction Procedure

The foil (or foils) are then reduced, with a supply of dry argon containing 5% hydrogen passing though the reactor channels (or the tube) flowing though at a space velocity of 8000 /hr. Starting at ambient temperature the temperature is ramped at 3°C/min to 8O⁰C, and held for an hour; the temperature is then increased at 3°C/min to 12O⁰C and held for an hour to ensure the foils are thoroughly dried (in case they may have picked up moisture from the atmosphere) .

In the following explanation it will be assumed that the ceramic has the TGA profile shown in figure 1.

The temperature is then raised at 0.5°C/min to about 17O⁰C (this temperature is set at 10 K below the maximum of the TGA peak in this vicinity, which is the first cobalt oxide reduction peak) , the pressure (in this example) is increased to 3 barg, and the temperature is held for two hours. The temperature is then ramped at 0.2°C/min to 25O⁰C (this temperature being set above the start of the second cobalt oxide reduction peak) , and then held for 16 hours to allow slow initiation of the second reduction step. The temperature is then ramped at 0.2°C/min from 25O⁰C up to about 375⁰C (this maximum temperature being set at 25 K above the maximum of the second cobalt oxide reduction peak) , where it is held for 2 hours. The gas mixture is then replaced with pure hydrogen at the same pressure and the same space velocity as before, and the temperature held at this maximum temperature for at least a further 30 hours.

It will be appreciated that the reduction process described above is by way of example only; the reduction process can be modified, for example holding the temperature at 25O⁰C for even longer, for example for 60 hours, and the temperature might be held for longer at the maximum temperature, for example for 46 hours, or each hold may be for a shorter period of only a few hours. And the gas might contain less than 10% hydrogen, and the pressure may be held at or near atmospheric pressure, throughout the reduction process. Indeed the temperature holds at 25O⁰C and 375⁰C can be decreased to 4 hours each, with no change to the activity of the resulting catalyst.

3. Oxidation Procedure

The catalyst-carrying foils are then subjected to a gradual oxidation procedure. While still at the maximum temperature in the reduction procedure (about 375⁰C) , the pressure is reduced to atmospheric and the reducing gases are flushed out with dry helium at a space velocity of 14,000 /hr for 10 minutes. Over a period of six hours, air is gradually introduced into the gas flow, while continuously checking the temperature of the foils,- if the temperature increases by more than 5 K then the proportion of air is reduced to the previous value. The time steps are as shown in the table.

Table

Space velocity Proportion of Dry Air Hold for

14 000 /hr 0.25% 5 min

14 000 /hr 0.60% 10 min

14 000 /hr 0.85% 10 min

14 000 /hr 1.36% 10 min

14 000 /hr 2.10% 20 min

12 850 /hr 2.35% 10 min

12 850 /hr 3.0% 10 min

13 160 /hr 4.6% 20 min

12 050 /hr 8.0% 10 min

13 050 /hr 14.8% 10 min

11 600 /hr 16.7% 10 min 10 150 /hr 19.0% 10 min

8 700 /hr 22.2% 10 min

7 250 /hr 26.7% 20 min

8 200 /hr 35.3% 30 min 9 150 /hr 42.1% 10 min

6 300 /hr 61.5% 10 min

5 100 /hr 76.2% 10 min

3 900 /hr 100% 8 hr

4. Second Reduction Procedure

The foils are then subjected to a second reduction procedure, which is similar to the first reduction procedure. The temperature is lowered to the value of about 25O⁰C mentioned before, and the air is purged out using helium at a space velocity of 8000 /hr. The gas is then changed to dry argon with 5% hydrogen, at the same space velocity, and the temperature is held at 25O⁰C with this reducing atmosphere for 2 hours. The pressure is then raised to 3 barg, and the temperature is then ramped at 0.2°C/min from 25O⁰C up to about 375⁰C (the maximum temperature used previously) , where it is held for 2 hours. The gas mixture is then replaced with pure hydrogen at the same pressure and the same space velocity as before, and the temperature held at this maximum temperature for at least a further 30 hours.

In a modification the gas may be less than 10% hydrogen, and the pressure be at or near atmospheric, throughout the reduction process. The use of this low hydrogen concentration during both the first and second reduction steps has been found to produce as active a catalyst as is produced using higher concentration and pressure. And (as with the first reduction step) the temperature need only be held at 375⁰C for 4 hours. 5. Quality Checks

The resulting catalyst foils, or specimens of the ceramic removed from the surface, are preferably then subjected to measurements of several different parameters, to ensure quality. The ceramic will have mesopores, of characteristic size in the range 2 nm to 20 nm, which provide the majority of sites for the dispersed catalyst metal . Preferably the pores are of size between 8 and 16 nm, more preferably between 12 and 14 nm. For Fischer-Tropsch synthesis it is also necessary for there to be larger mesopores and also macropores, that is to say pores of size at least 50 nm and above. The specific surface area of the catalyst-containing ceramic is preferably about 60-140 m²/g (as measured by the BET gas adsorption technique), for example 90 m²/g. The specific pore volume, as measured by mercury intrusion porosimetry, of the as-supplied particulate alumina is preferably in the range 0.35 to 0.59 cm³/g; while that of the catalyst-containing ceramic is in the range 0.20 to 0.30 cm³/g (as measured by the BET technique), for example 0.24 cm³/g.

The dispersion of cobalt is desirably in the range between 0.06 and 0.15 (that is to say the proportion of cobalt metal atoms which are exposed at surfaces of or within the ceramic) . The extent of reduction of cobalt in the final material is between 90% and 100%. And the hydrogen chemisorption is preferably between 100 and 300 micromoles/g of catalyst-containing ceramic.

The method of the invention enables a highly active Fischer-Tropsch catalyst to be made, the catalyst comprising active catalytic metal in a ceramic support on a metal substrate. The metal substrate suppresses the development of hotspots during use of the catalyst so that the highly active catalyst does not overheat and deteriorate. If the maximum temperature for reduction goes more than about 25 K above the TGA peak, the dispersion becomes less and the activity decreases. If the maximum temperature for reduction is less than about 20 K above the TGA peak, the dispersion will be too high, and the catalyst will be too active and non- selective .

Claims

1. A method of preparing a Fischer-Tropsch catalyst comprising the steps of: a) contacting a ceramic support with a solution of a salt of catalytic metal, drying and then calcining, to form the metal oxide, wherein the metal of the resulting metal oxide consists primarily of cobalt; b) reducing the metal oxide to metal by heating in the presence of a reducing atmosphere including hydrogen, the heating process being such as to ramp up the temperature gradually to a maximum temperature below 42O⁰C and above a TGA peak by no more than 30 K over a period of several hours, this temperature ramp being no more than l°C/min while reduction is a occurring and including holding the temperature for at least an hour at a temperature above the start of the said TGA peak; c) subjecting the resulting metal to an oxidation process at an elevated temperature, by replacing the reducing atmosphere by an atmosphere containing an oxidising agent, and increasing the proportion of the oxidising agent gradually over a period of hours, to ensure formation of metal oxide,- d) and then subjecting the resulting metal oxide to a reduction process at temperatures less than or up to substantially the said maximum temperature so the metal oxide is again reduced to metal .

2. A method of preparing a Fischer-Tropsch catalyst comprising the steps of: a) contacting a ceramic support with a solution of a salt of catalytic metal, drying and then calcining, to form the metal oxide, wherein the metal of the resulting metal oxide consists of cobalt and less than 2% by weight of ruthenium; b) in the presence of a reducing atmosphere comprising hydrogen, raising the temperature gradually over a period of at least 2 hours to a temperature no more than 125⁰C to ensure the metal oxide is thoroughly dry, and then gradually ramping the temperature to a maximum temperature less than 42O⁰C which is between 20 and 30 K above a second cobalt oxide reduction TGA peak over a period of at least about 20 hours, this temperature ramp being no more than l°C/min while reduction is occurring and including holding the temperature for at least 3 hours at a holding temperature above the start of the second cobalt oxide reduction TGA peak, and then holding the temperature at the maximum temperature for at least 3 hours so the metal oxide undergoes full reduction; c) while remaining at the maximum temperature, flushing out the reducing atmosphere and introducing a gas containing oxygen, increasing the proportion of oxygen gradually over a period of at least 3 hours, and then holding the proportion of oxygen for a period of at least 6 hours, to ensure formation of the metal oxide,- d) then flushing out the oxygen-containing gas, lowering the temperature to the holding temperature, introducing a reducing atmosphere, and gradually ramping the temperature up to the maximum temperature, and then holding the temperature at that value for at least 3 hours so the metal oxide again undergoes reduction.

3. A method as claimed in claim 1 or claim 2 wherein throughout steps (b) and (d) the reducing atmosphere contains no more than 10% hydrogen.

4. A method as claimed in claim 1 or claim 2 wherein in steps (b) and (d) the reducing atmosphere contains no more than 10% hydrogen during the initial stages of reduction; and the method involves increasing the concentration of hydrogen during the later stages of reduction.

5. A method as claimed in claim 4 wherein the concentration of hydrogen is above 90% whilst the temperature is at the maximum temperature.

6. A method as claimed in any one of the preceding claims wherein the temperature ramp in the first reducing step (b) also includes holding the temperature at a temperature above that of the start of the first cobalt oxide reduction TGA peak.

7. A method as claimed in any one of the preceding claims wherein in both the reduction steps the reducing atmosphere is arranged to flow continuously over the substrate, with a space velocity of at least 6000 /hr, more preferably about 8000 /hr.

8. A method as claimed in any one of the preceding claims wherein the oxidation step uses air as a source of oxygen .

9. A catalyst made by a method as claimed in any one of the preceding claims.