WO1990008589A1

WO1990008589A1 - Perovskite-type catalyst and processes for using it

Info

Publication number: WO1990008589A1
Application number: PCT/GB1990/000060
Authority: WO
Inventors: Brian Ellis; Alan Ivor Foster
Original assignee: The British Petroleum Company Plc
Priority date: 1989-01-24
Filing date: 1990-01-16
Publication date: 1990-08-09
Also published as: GB8901514D0; AU4845790A

Abstract

A mixed metal oxide having a perovskite-type structure ABO¿3-y¿, where A corresponds to a relatively large cation and B corresponds to a relatively small cation has an oxygen deficiency y of at least 0.1 and has at least two transition metals in the B position, one of which is Cu. The oxide is useful in the conversion of NO to N¿2? in mixtures containing CO and NO.

Description

PEROVSKITE-TYPE CATALYST AND PROCESSES FOR USING IT

The present invention relates to novel mixed oxides with a perovskite-type structure.

Perovskite is a naturally occurring mineral of formula CaTiθ3. However many mixed oxides eg. BaTi03 are known which have the same structure. Perovskite-type mixed oxides correspond to a general formula ABO3. A normally corresponds to a relatively large cation eg. greater than 0.1 nm. B normally corresponds to a smaller cation e.g. smaller than 0.1 nm. Thus the A type cation corresponds to Ba in barium titanate and the B-type ion corresponds to Ti in barium titanate. (see J.B. Goodenough and J.M. Londo in "Landolt-Bδrnstein" Vol 4. (1970)).

In this specification a "perovskite-type" material is one which is isostructural with perovskite, i.e. in which the intensities of the X-ray diffraction lines and their positions substantially correspond to those of perovskite.

Materials having perovskite crystal structures have been investigated as possible catalysts for use in the control of automotive exhaust emissions. The oxides which have been investigated include those in which the cations of 3 or more metals are present in place of the 2 cations present in perovskite itself. These mixed oxides may have normal oxygen stoichiometry. • They may however be made in oxygen-deficient form i.e. where, in the formula ABO_n, n is less than 3.

Gallagher et al, J. American Ceramic Soc, 60(1-2) 28-31 (1977) discloses the preparation of mixed oxides with the perovskite-type structure of formula aMn^__xCu_xθ3_y, and their use in the oxidation of CO as a potential catalyst for the control for automotive emissions. The value of y is not given in all cases but one or two nominal compositions are given in which 3-y is 2.96 or 2.93. The oxygen deficiencies are thus only 0.04 to 0.07, and the compounds are only marginally oxygen deficient.

Aral et al, Applied Catalysis, 2 (1986), 265-276 discloses that methane may be oxidised over various perovskite-type oxide catalysts. Mixed oxide catalysts of the type a^__xA_xB03 are disclosed in which A is for example Sr and B is Co, Mn, and Fe. Perovskite-type materials in which the A position is occupied by La and Sr and the B position is occupied by Co, Mn, or Fe are not usually oxygen deficient except for high Sr concentrations. Even then the oxygen deficiency (corresponding to y in the formula ABθ3__y) is less than 0.1. Mizuno et al, J. Chem. Soc. , Chem. Comm. 1(7) (1988), 887-888 discloses the production of mixed oxides of formula B 2Cu3θ_y. The formula for each unit block is Y o-33 -•^a 0*66 ^^u^ 2-33-Th s material contains only a single transition metal in the B position. The material is used as a catalyst for the reaction of NO and CO to give N2 and CO2.

There is a need for mixed oxide catalysts of perovskite-type structure with improved activity, particularly for certain reactions which are important in the treatment of automotive exhaust emissions. According to the present invention a novel mixed metal oxide has a perovskite-type structure characterised in that it has a general formula ABθ3_y in which A corresponds to a relatively large cation and B corresponds to a relatively smaller cation and y represents a calculated oxygen deficiency (y) of greater than 0.1, there are at least two transition metals in the B position of the perovskite-type structure ABθ3_y, one of which is Cu.

By "calculated oxygen deficiency" we mean the oxygen deficiency y in the formula ABθ3_ calculated on the basis of charge compensation, on the basis of the oxidation state for metals in the A site being maximum value eg 3+ for La, 2+ for Sr, and the oxidation states for the metals in the B position being Cu - 2+, Co ^•= 3.5+, Fe = 3.5+, Ru = 5+, Group VA metals (eg. V) -> 5+, and Group VIIA metals (eg. Mn) ■ 4+. The value of 3.5+ for Co and Fe is used on the basis of the existence of perovskite-type oxides in which this is observed i.e. SrFeθ2.75 and SrCoθ2-75. The catalyst of the present invention is defined on the basis of calculated oxygen deficiency. The calculated oxygen deficiency will generally correspond comparatively closely to the measured oxygen deficiency (where this can be determined) for catalysts subjected to heating during or after preparation to temperatures of not more than 900°C.

Preferred catalysts according to the invention are thus those in which temperatures of 900^βC have not been exceeded during the catalyst preparation or subsequently.

References to the Periodic Table in this specification are references to the Periodic Table published in the Classification Key of the British Patent Office.

The predicted oxygen deficiency is preferably from 0.1 to 0.5, more preferably from 0.2 to 0.5.

The A positions are preferably occupied by La cations together with cations of metals of Groups IA or IIA of the Periodic Table.

An example of a suitable IIA metal is strontium.

The requirement for oxygen-deficiency will generally mean that there will be A site cations from at least two different metals e.g. La and Sr as indicated above. The preferred molar ratio where there are two A site metals is in the range 0.1 to 0.9.

The molar ratio of Cu to the other transition metal is preferably in the range 0.1 to 0.9. The ratio of the A site to the B site cations must be consistent with the requirements for calculated oxygen-deficiency requirements set out above. Examples of the transition metals which may be present in the B position in addition to Cu are Fe, Co, Mn, V and Ru, Group VA and VIIA.

The mixed oxides may be prepared by controlled precipitation by addition of metal salt aqueous solutions (e.g. nitrates) and base aqueous solutions (e.g. sodium or potassium carborates). The precipitates after filtration and washing may be calcined e.g. at temperatures in the range 650-750^βC. The calcination step is preferably carried out in air.

According to the present invention the process for the conversion of NO to N2 by passing a gas containing CO and NO over catalysts at elevated temperatures and in the substantial absence of molecular oxygen is characterised in that the catalyst is a mixed oxide catalyst which has a perovskite-type structure with a calculated oxygen-deficiency y of greater than 0.1, and has at least two transition metals in the B position of the perovskite-type related structure ABθ3_y, one of which is Cu.

Examples of elevated temperatures which may be used are those in the range 100 to 1000"C.

According to another aspect of the present invention a process for the oxidation of oxidisable vapour comprises passing the vapour in an oxygen-containing atmosphere over a catalyst which is a mixed metal oxide which has a perovskite-type structure with a calculated oxygen-deficiency y of greater than 0.1, and has at least two transition metals in the B position of the perovskite-type related structure ABQ3__y, one of which is Cu.

The above described processes are particularly suitable for application to the exhaust gases produced by internal combustion engines, in particular to the exhaust gases produced by automobile engines. They are particularly suitable for application to exhaust gases from gasoline engines.

The catalytic material preferably comprises predominantly perovskite-type material e.g. greater than 90% wt/wt. Preferably it comprises at least 90% wt/wt of perovskite-type material. The catalyst material may, of course, be supported on inactive carriers or supports as is well-known in the art.

The invention will now be described with reference to the following experiments, in which examples of the invention are identified by numbers, and comparative tests, not according to the invention, are identified by letters .

Examples 1-3

Three perovskite-type structures according to the invention and having the composition given in Table 1 were prepared as set out below.

Preparation of perovskite oxides

The aim of the experiment was to co-precipitate the component metals as carbonates/oxides using the method of constant-pH precipitation. In this, a fixed pH is chosen which correlates with the pH required for precipitation of each metal component, information derived from literature sources.

('Atlas of Metal-Ligand Equilibria in Aqueous Solution' by J. Kragten,

Publ. Ellis Horwood Ltd., Press. Wiley 1978)

The range of constant pH levels chosen for these experiments was 9-10. Experimental procedure was as follows:

1. Samples (typically 0.05 mole) of each :component metal nitrate or acetate were dissolved together in 400 ml of water with stirring. The resultant solution was placed in a dosimat (automatic dosing machine) capable of quantifiable delivery (typically up to a maximum volume of 50 ml). Likewise, a solution of potassium or sodium carbonate

(typically 2M in concentration) was placed in a second such dosimat.

Next, distilled water was placed in a large beaker (typically 300 ml) with stirring from a overhead stirrer ("Silverson") plus pH measurement using an immersible pH electrode connected to a pH meter. The starting pH of the water was monitored and was typically

5.5-6.5.

2. The second stage involved the computer-controlled addition of both base and mixed metal component solutions to the water in the beaker so as to bring about precipitation at constant pH. Initially, the pH of the water was increased to the desired level by a controlled addition of base. Next, an aliquot of metal solution (typically 3 ml) was added to the stock solution. This depressed the pH slightly due to the metal nitrate or acetate solution being acidic. The computer then restored the pH to the original level by controlled base addition. The amount of base necessary to achieve this is exactly equal to that needed to precipitate the metal ions present, plus the small extra amount required to raise the pH to the original level. In this way, complete precipitation was ensured without the use of excessive quantities of base which might contaminate the precipitate. The cycle of metal solution/base addition was then repeated by the computer, for the number of cycles required by the operator (typically 100 cycles). During the complete experiment, the pH is never depressed more than 1 pH unit during any one cycle. 3. Filtration of the resultant precipitate/solution gave precipitates ranging in colour from blue-green to olive-green together with a clear filtrate. The washing procedure involved suspension of the wet precipitate in fresh distilled water (typically 600 ml), follows by rapid stirring ("Silverson" stirrer on full-speed) for 10 minutes. Two further filtration/washing cycles were carried out, in order to minimise the concentration of base contaminant in the final precipitate, as shown by XRF analysis. After oven-drying in an air stream at 110^βC for 12 hours, the precipitates were calcined in air for 15 hours at high temperature. The calcination temperature used was typically in the range 650-750"C. Characterisation of oxide catalysts

The products of the above preparation were studied using powder X-ray diffraction and were found to contain predominantly material of perovskite structure (believed to be greater than 90%). The presence of extraneous phases, such as CuO, SrC03, Mn3U was detected in all cases but was sufficiently small to allow subsequent catalyst testing. Analysis by X-ray fluorescence revealed good compositional purity and confirmed that levels of potassium impurity from the preparation process were less than 3 mole %. Catalyst test apparatus and methods The apparatus used for catalytic testing of the samples consisted of three main parts, the inlet gas-handling system, the catalytic reactor with furnace and the gas chromatograph analysis system. The operation of each of these sections can be described as follows: The input gases (10% CO/10% NO in He) were mixed as required using mass flow controllers and could be admitted into the catalytic reactor or analysed separately via a by-pass.

Experiments carried out used a mixture of 2.5% of both CO and NO with a balance of He. The total flow rate of gas was maintained at 140ml/min to give a gas hourly space velocity (GHSV) of 8400. The catalytic reactor consisted of a 3/8" O.D. (9.5 mm outside diameter) 316 stainless steel tube horizontally mounted in a temperature-controlled tubular furnace. Catalyst samples (0.5-1.0 mm particles) occupied a bed volume of 1 ml and were maintained in the correct position in the reactor using loosely-packed silica wool plugs. The furnace temperature was controlled between ambient and a maximum of 700^βC. During heating operations, a temperature ramp not exceeding lO'C/min was maintained.

After passage over the catalyst, part of the exhaust gas was bled off to a Gas Chromatograph (GC), while the rest was vented, in order to restrict the back presure due to the G.C. sampling valves.

The analysis utilised a gas chromatrograph and consisted of consecutive sample injections via 6-port air-actuated sampling valves to a thermal conductivity detector. An initial injection onto a lm (1/8" O.D.) (3.2 mm outside diameter) stainless steel column containing 13X molecular sieve allowed separation of N2, O2, CO and NO within 6.5 minutes. At this time, a second injection was made onto a similar column containing "Porapak QS", which was used to separate ^co2» ^N2» ^{and N}2°* Total analysis time was 9 minutes. All chromatographic data was analysed and recorded by a computing integrator.

Catalytic Activity results (CO/NO reaction)

The catalytic activity for the CO/NO reaction was measured as a function of temperature for several novel Cu-M perovskite oxides prepared according to the invention. Table 1 shows the results expressed as the temperature for 50% conversion of CO to CO2 and NO to N2 as a result of CO/NO reaction. All the catalysts are seen to have good activity for promoting this reaction.

The compounds of Examples 1, 2 and 3 had a calculated oxygen-deficiency of respectively of 0.25, 0.33 and 0.38. Comparative Test A

A catalytic activity test was carried out as in Example 1 to 3 but with an empty tube (no catalyst). The results are shown in Table 1.

Table 1 CO/NO activity results for various catalysts prepared according to the invention (expressed as the temperatures for 50% conversion to CO2 and N2).

Conditions

GHSV 8400 h^_1

2.5% CO

2.5% NO

Balance He

Comparative Tests B-E (not according to the Invention)

Various other Perovskites described by other workers (eg. a) L. Wan et al, Catalysis and Automobile Pollution Control, Publ. Elsevier, ed. A. Crucq and A. Frennet (1987). b) P.K. Gallagher et al, Mat. Res. Bull. 1345-52 (1974). c) P.K. Gallagher et al, J. Araer. Ceram. Soc, 60 (1-2) 28-31 (1977) d) Japanese patent unexamined patent specification JP59222230, Matsushita Elect. Co., Filed 27/5/83). as being useful for the removal of carbon monoxide, hydrocarbons and/or nitrogen oxides from exhaust gases were prepared and characterised using the method described in Example 1.

They were tested for catalytic activity in the reaction between CO and NO under the same conditions as were used in Example 1. The results are outlined on Table 2, where it is seen by comparison with Table 1 that these catalysts are considerably less active than those prepared according to the present invention.

A comparison of the results obtained for Examples 1 to 3 with those for Comparative Tests B to E shows that it is not sufficient to have two transition metals, one of which is Cu, in the B position (Comparative Test D), (Comparative Test E). To obtain high activities it is necessary to select a specific combination of catalyst characteristics.

Table 2 CO/NO activity results for various catalysts not corresponding to the invention (expressed as the temperatures for 50% conversion to CO2 and N₂).

Conditions GHSV 8400h^_1 2.5% CO

2.5% NO

Balance He

Examples 4 and 5 Some of the perovskite catalysts prepared according to the invention and described in Example 1 were tested under conditions of increased gas throughout (GHSV) and reduced concentration of the active components (CO and NO). The conditions used (0.5%^v/_v CO, 0.5^V/_V NO,

GHSV 30,000 h^-i) reflected the conditions that might be encountered in an automotive exhaust system. A quartz reactor was used for these experiments, with 3ml of alumina beads as a gas preheater before the 1 ml of catalyst.

The results (presented in Table 3) clearly indicate that the good performance of these catalysts in maintained under these conditions of use.

Comparative Test F

An experiment was carried out as in Examples 4-5 using an empty reactor. (The results are shown in Table 3).

Table 3 CO/NO activity for various catalysts (expressed on the same basis as for Tables 1 and 2).

GHSV 30,000 h^_1 0.5% CO

0.5% NO

Balance Helium

Comparative Example G

The activity under the conditions used for Example 4-5 of a catalyst not according to the invention is shown in Table 4. Its lower activity is apparent; production of nitrous oxide (N2O) was also observed, an undesirable result. The results are shown in Table 4.

Table 4 CO/NO activity results for a Perovskite catalyst not according to the invention, (expressed as the temperature for 50% conversion to CO2 and N₂).

GHSV ** 30,000 h^_1 0.5% CO 0.5% NO

Balance Helium Examples 6-8

Some of the perovskites prepared according to the invention and described in Examples 1-3 were tested for their activity in the oxidation of carbon monoxide and hydrocarbons that are difficult and easy respectively to oxidise. The conditions used (GHSV ^•*■ 30,000h^~l, 0.5% of CO, 0.1% methane 0.1% propylene and 1.0% by volume oxygen were selected to be representative of the type of exhaust gas with which an oxidation catalyst might be employed.

The results (summarised in Table 5) indicate that all the catalysts tested had good activity for CO oxidation, while in addition those containing Fe and Co as the second "B site" element also had good activity for hydrocarbon oxidation. Comparative Test H

This was a comparative test carried out as in Examples 6 to 8 but using an empty tube containing no catalyst. The results are given in Table 5.

Table 5 Hydrocarbon or CO oxidation activity for Perovskite catalysts prepared according to the Invention.

(expressed as the temperature for 50% conversion of methane, propylene and carbon monoxide).

GHSV = 30,000 h^_1

0.5% CO

0.1% methane

0.1% propylene

1.0% oxygen

Balance helium

Comparative Examples I-J

Perovskite catalysts not according to the invention (see comparative examples B and E) were tested under the same conditions as used in Examples 6-8. The results (Table 6) show that these catalysts are not as active as those prepared according to the invention and containing Fe or Co (Table 5). This is notwithstanding the fact that the perovskite La ι__x Sr_x COJ.Q O3 (x - about 0.3) is claimed to be particularly active for methane oxidation (H. Akai, T. Yamada, K. Eguchi and T. Seijama; Applied Catalysis 2 (1986), 265).

TABLE 6 Hydrocarbon and CO oxidation activity for Perovskite catalysts not according to the invention.

(expressed as the temperature for 50% conversion of methane, propylene and carbon monoxide)

GHSV = 30,000 h^_1 0.5% CO 0.1% methane 0.1% propylene 1.0% oxygen Balance Helium

Claims

UClaims:

1. A mixed metal oxide having a perovskite-type structure characterised in that it has a general formula of the type AB03__y in which A corresponds to a relatively large cation and B corresponds to a relatively smaller cation, and y represents a calculated oxygen deficiency greater than 0.1, and there are at least two transition metals in the B position of the perovskite-type structure one of which is Cu.

2. A mixed oxide according to claim 1 wherein the calculated oxygen deficiency (y) is in the range of 0.1 to 0.5.

3. A mixed oxide according to claim 2 wherein the calculated oxygen deficiency is in the range of 0.2 to 0.5.

4. An oxide according to one of the preceding claims wherein the A positions are occupied by La cations together with cations of metals of groups IA or IIA of the periodic table.

5. An oxide according to claim 4 wherein the A positions are occupied by La cations and Sr cations.

6. An oxide acording to any one of the preceding claims wherein the molar ratio of Cu to the other transition metal in the B position is in the range 0.1 to 0.9.

7. An oxide according to any one of the preceding claims wherein the transition metals present in the B position in addition Cu are one or more Fe, Co, Mn, and Ru, group VA and VIIA.

8. An oxide according to any one of the preceding claims which has not been subjected to a temperature of 900"C during preparation or subsequently.

9. A process for the conversion of NO to N by passing a gas containing CO and NO over catalysts at elevated temperatures and in the substantial absence of molecular oxygen is characterised in that the catalyst is a mixed oxide catalyst which has a perovskite-type structure with a calculated oxygen-deficiency y of greater than 0.1, and has at least two transition metals in the B position of the perovskite-type related structure ABθ3_y, one of which is Cu.

10. A process for the oxidation of oxidizable vapour comprises passing the vapour in an oxygen-containing atmosphere over a catalyst which is a mixed metal oxide which has a perovskite-type structure with a calculated oxygen-deficiency y of greater than 0.1, and has at least two transition metals in the B position of the perovskite-type related structure ABθ3__y, one of which is Cu.

11. A process according to either of Claims 9 or 10 wherein the elevated temperature is a temperature in the range 100^βC to 1000 "C .

12. The process according to Claim 9 wherein the gas containing CO and NO is an exhaust gas from a gasoline engine.