PROCESS FOR PRODUCING GEOMETRIC SHAPED CATALYST BODIES

Abstract

A process for producing geometric shaped catalyst bodies K whose active material is a multielement oxide of stoichiometry [Bi1WbOx]a[Mo12Z1cZ2dFeeZ3fZ4gZ5hOy]1, in which a finely divided oxide Bi1WbOx with the particle size d50A1 and, formed from element sources, a finely divided intimate mixture of stoichiometry Mo12Z1cZ2dFeeZ3fZ4gZ5h with the particle size d90A2 are mixed in a ratio of a:1, this mixture is used to form shaped bodies and these are treated thermally, where (d50A1)0.7·(d90A2)1.5·(a)−1≧820. A shaped catalyst body obtained by the process. A catalyst obtained by grinding the shaped catalyst body. A process for heterogeneously catalyzing the partial gas phase oxidation of an alkane, alkanol, alkanal and/or an alkenal of 3 to 6 carbon atoms using the catalyst.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is a Divisional of U.S. patent application Ser. No. 12/546,883 filed Aug. 25, 2009, and claims priority to U.S. Provisional Application No. 61/096,537 filed Sep. 12, 2008, and German Application No. 10 2008 042 064.6 filed Sep. 12, 2008, the entire contents of which are hereby incorporated by reference.

DESCRIPTION

The present invention relates to a process for producing geometric shaped catalyst bodies K* which comprise, as an active material, a multielement oxide I of the general stoichiometry I

[Bi₁W_bO_x]_a[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_hO_y]₁  (I)

where

• Z¹=one element or more than one element from the group consisting of nickel and cobalt,
• Z²=one element or more than one element from the group consisting of the alkali metals, the alkaline earth metals and thallium,
• Z³=one element or more than one element from the group consisting of zinc, phosphorus, arsenic, boron, antimony, tin, cerium, vanadium, chromium and bismuth,
• Z⁴=one element or more than one element from the group consisting of silicon, aluminum, titanium, tungsten and zirconium,
• Z⁵=one element or more than one element from the group consisting of copper, silver, gold, yttrium, lanthanum and the lanthanides,
• a=0.1 to 3,
• b=0.1 to 10,
• c=1 to 10,
• d=0.01 to 2,
• e=0.01 to 5,
• f=0 to 5,
• g=0 to 10,
• h=0 to 1, and
• x, y=numbers determined by the valency and frequency of the elements in I other than oxygen,
  in which
  • a finely divided mixed oxide Bi₁W_bO_x with a particle diameter d₅₀^A1 reported in the length unit μm, as starting material A1, is preformed with the proviso that 1 μm≦d₅₀^A1≦10 μm;
  • sources of the elements other than oxygen in the component T=[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_hO_y]₁ of the multielement oxide I are used in an aqueous medium to obtain an intimate aqueous mixture M, with the proviso that
    • each of the sources used, in the course of preparation of the aqueous mixture M, passes through a degree of division Q for which its diameter d₉₀^Q is ≦5 μm,
      • and
    • the aqueous mixture M comprises the elements Mo, Z¹, Z², Fe, Z³, Z⁴ and Z⁵ in the stoichiometry I*

Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_h  (I*);

• • the aqueous mixture M, by means of drying and adjusting the degree of division d₉₀^A2, is used to obtain a finely divided starting material A2 with a particle diameter d₉₀^A2 reported in the length unit μm, with the proviso that 200 μm≧d₉₀^A2≧20 μm;
  • starting material A1 and starting material A2, or starting material A1, starting material A2 and finely divided shaping assistant, are mixed with one another to form a finely divided starting material A3, with the proviso that the starting material A3 comprises the elements other than oxygen introduced into the starting material A3 via starting materials A1 and A2 in the multielement oxide I in the stoichiometry I**

[Bi₁W_b]_a[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_h]₁  (I**),

• • finely divided starting material A3 is used to form geometric shaped bodies V, and
  • the shaped bodies V are treated thermally at elevated temperature to obtain the geometric shaped catalyst bodies K*.

The present invention further relates to the use of shaped catalyst bodies K*.

Suitable shaped catalyst bodies K* which comprise a multielement oxide I as an active material and processes for producing such shaped catalyst bodies are known (cf., for example, German application 102007003778.5, EP-A 575 897, WO 2007/017431, WO 02/24620, WO 2005/42459, WO 2005/47224, WO 2005/49200, WO 2005/113127, German application 102008040093.9, German application 102008040094.7 and DE-A 102007005606).

It is additionally known that catalysts K* (geometrically shaped catalyst bodies K*) are suitable for performing heterogeneously catalyzed partial oxidations of alkanes, alkanols, alkenes and/or alkenals having from 3 to 6 carbon atoms in the gas phase.

A full oxidation of an organic compound with molecular oxygen is understood in this document to mean that the organic compound is converted under the reactive action of molecular oxygen such that all carbon present in the organic compound is converted to oxides of carbon and all hydrogen present in the organic compound is converted to oxides of hydrogen. All different conversions of an organic compound under the reactive action of molecular oxygen are summarized in this document as partial oxidations of an organic compound.

In particular, partial oxidations shall be understood in this document to mean those conversions of organic compounds under the reactive action of molecular oxygen in which the organic compound to be oxidized partially, after the reaction has ended, comprises at least one oxygen atom more in chemically bound form than before performance of the partial oxidation.

In this document, the term “partial oxidation”, however, shall also comprise oxidative dehydrogenation and partial ammoxidation, i.e. partial oxidation in the presence of ammonia.

Catalysts K* (geometric shaped catalyst bodies K*) are particularly suitable for performing the heterogeneously catalyzed partial gas phase oxidation of propene to acrolein and of isobutene to methacrolein, and for performing the heterogeneously catalyzed partial gas phase ammoxidation of propene to acrylonitrile and of isobutene to methacrylonitrile.

In general, the heterogeneously catalyzed partial gas phase oxidation of propene (isobutene) to acrolein (methacrolein) forms the first stage of a two-stage heterogeneously catalyzed partial gas phase oxidation of propene (isobutene) to acrylic acid (methacrylic acid), as described by way of example in WO 2005/42459.

By-product formation of acrylic acid (methacrylic acid) which accompanies a heterogeneously catalyzed partial gas phase oxidation of propene (isobutene) to acrolein (methacrolein) is therefore generally not undesired and is normally included with the desired product of value formation.

It is additionally known that the performance of geometric shaped catalyst bodies K*, in the course of continuous operation of a heterogeneously catalyzed partial gas phase oxidation of alkanes, alkanols, alkenes and/or alkenals having from 3 to 6 carbon atoms (for example to the corresponding olefinically unsaturated aldehydes and/or carboxylic acids), decreases with increasing operating time (this applies in particular to the case of a partial gas phase oxidation of propene to acrolein and/or acrylic acid and of isobutene to methacrolein and/or methacrylic acid heterogeneously catalyzed by geometric shaped catalyst bodies K*; however, it is also true for the case of a heterogeneously catalyzed partial gas phase ammoxidation of propene to acrylonitrile and of isobutene to methacrylonitrile). This statement is also valid when the geometric shaped catalyst bodies K* are subjected in the course of continuous operation of the heterogeneously catalyzed partial gas phase oxidation to a periodically repeating regeneration process, as recommended, for example, by specifications WO 2005/42459 and WO 2005/49200.

Primarily, it is the activity of the geometric shaped catalyst bodies K* which is reduced in the course of operation, which usually lasts for several years, of a heterogeneously catalyzed partial gas phase oxidation of an organic compound.

One measure of the activity of the geometric shaped catalyst bodies K* or of a catalyst bed comprising them is that temperature which is required to achieve a particular conversion of the organic compound (for example of the propene or of the isobutene) as the reaction gas mixture comprising the organic compound to be oxidized partially passes through the catalyst bed.

When the activity of the geometric shaped catalyst bodies K* of a catalyst bed comprising them declines to an increasing degree with increasing operating time of the partial oxidation, with otherwise unchanged reaction conditions, an increasingly elevated temperature is required to achieve the same conversion of the organic compound in the course of single pass of the reaction gas mixture through the catalyst bed (when the catalyst bed is present, for example, in the tubes of a tube bundle reactor around which a salt bath flows, the inlet temperature of the salt bath into the tube bundle reactor will normally be increased gradually, under otherwise unchanged operating conditions, with increasing deactivation of the catalyst bed, in order to maintain a partial oxidation conversion based on single pass of the reaction gas mixture through the catalyst bed (cf., for example, EP-A 1 734 030, WO 2007/82827, WO 2005/47224 and WO 2005/42459)).

However, a disadvantage of the above-described procedure is that the catalyst deactivation proceeds increasingly rapidly with increasing reaction temperature until the spent catalyst bed has to be replaced at least partly or fully by a catalyst bed comprising fresh geometric shaped catalyst bodies K* (cf., for example, WO 2004/9525, DE-A 10 2006 00 0996 and WO 2007/77145).

An at least partial or a complete catalyst bed exchange is, however, disadvantageous in that it is necessarily accompanied by an interruption to the preparation of target product.

Furthermore, the preparation of fresh catalyst for an industrial scale preparation of target product is associated with a considerable investment, since both the expenditure for the raw material required for this purpose and the manufacturing complexity are not inconsiderable.

Overall, there is thus a general interest in geometric shaped catalyst bodies K* which, in the course of continuous operation of the heterogeneously catalyzed partial gas phase oxidation, have a minimum deactivation rate.

The causal connections which have to be observed to cause this maximum extension in long-term stability in a preparation of geometric shaped catalyst bodies K* are essentially unknown.

Research projects dedicated to the problem are enormously time-consuming since they address a subject which, at first glance, is displayed only over very long observation periods. Furthermore, errors in the operation of the heterogeneously catalyzed partial gas phase oxidation using one and the same catalyst bed can also cause an increased deactivation rate.

The fundamental basis of the present invention is formed by the observation that geometric shaped catalyst bodies K*, even in the case of identical chemical composition of their active material, can have different deactivation behavior.

A further basis of the present invention is formed by the observation that the deactivation of a fixed catalyst bed proceeds more rapidly under otherwise predefined conditions when the loading of the fixed catalyst bed with reaction gas mixture is increased.

The loading of a fixed catalyst bed which catalyzes a reaction step with reaction gas mixture is understood in this document to mean the amount of reaction gas mixture in standard liters (=l (STP); the volume in liters that the corresponding amount of reaction gas mixture would occupy under standard conditions, i.e. at 0° C. and 1 atm) which is supplied to the fixed catalyst bed, based on the volume of its bed, i.e. on its bed volume (not including pure inert material sections), per hour (→unit=l (STP)/l·h).

The loading may also be based only on one constituent of the reaction gas mixture (for example only on the organic starting compound to be oxidized partially).

It is then the volume of this constituent (for example of the organic starting compound of the partial oxidation) that is fed to the fixed catalyst bed, based on the volume of its bed, per hour.

An additional fundamental basis of the present invention is formed by the experimental finding, made on corresponding reference samples of geometric shaped catalyst bodies K* used for industrial scale production, that it was possible to reflect, in a corresponding rank order, the relative order observed with regard to the deactivation rates within the different catalyst charges of the individual reactors under comparable operating conditions in the course of industrial scale long-term operation when the reference samples were subjected to a stress test requiring a comparatively significantly lower time commitment, wherein, primarily, the same heterogeneously catalyzed gas phase partial oxidation is performed both at a higher temperature and at a higher loading of the fixed catalyst bed with the same reaction gas mixture. The difference between the temperature required to establish the desired partial oxidation conversion under the actually contemplated operating conditions before the performance of the stress test and the same operating conditions, apart from the temperature, to establish the same partial oxidation conversion after the performance of the stress test has been found to be a reliable parameter of rank order with regard to the long-term stability of the catalyst bed.

EP-A 575897 discloses that, in a preparation of geometric shaped catalyst bodies K*, the particle size of the preformed finely divided mixed oxides Bi₁W_bO_x influences the initial activity of such shaped catalyst bodies when they are used for the heterogeneously catalyzed partial gas phase oxidation of propylene to acrolein.

In view of this prior art, it was an object of the present invention to provide geometric shaped bodies K* and a process for preparing them, which, especially in the context of use for a heterogeneously catalyzed partial gas phase oxidation of propene to acrolein, with essentially equal starting activity (after forming has ended), have a reduced deactivation rate in the course of continuous operation of partial oxidation.

This object is achieved by virtue of provision of a process for producing geometric shaped catalyst bodies K which comprise, as an active material, a multielement oxide I of the general stoichiometry I

[Bi₁W_bO_x]_a[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_hO_y]₁  (I)

where Z¹=one element or more than one element from the group consisting of nickel and cobalt,

• Z²=one element or more than one element from the group consisting of the alkali metals, the alkaline earth metals and thallium,
• Z³=one element or more than one element from the group consisting of zinc, phosphorus, arsenic, boron, antimony, tin, cerium, vanadium, chromium and bismuth,
• Z⁴=one element or more than one element from the group consisting of silicon, aluminum, titanium, tungsten and zirconium,
• Z⁵=one element or more than one element from the group consisting of copper, silver, gold, yttrium, lanthanum and the lanthanides (the rare earth metals),
• a=0.1 to 3,
• b=0.1 to 10,
• c=1 to 10,
• d=0.01 to 2,
• e=0.01 to 5,
• f=0 to 5,
• g=0 to 10,
• h=0 to 1, and
• x, y=numbers determined by the valency and frequency of the elements in I other than oxygen,
  in which
  • a finely divided mixed oxide Bi₁W_bO_x with a particle diameter d₅₀^A1 reported in the length unit μm, as starting material A1, is preformed with the proviso that 1 μm≦d₅₀^A1≦10 μm;
  • sources of the elements other than oxygen in the component T=[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_hO_y]₁ of the multielement oxide I are used in an aqueous medium to obtain an intimate aqueous mixture M, with the proviso that
    • each of the sources used, in the course of preparation of the aqueous mixture M, passes through a degree of division Q for which its diameter d₉₀^Q is ≦5 μm,
      • and
    • the aqueous mixture M comprises the elements Mo, Z¹, Z², Fe, Z³, Z⁴ and Z⁵ in the stoichiometry I*

Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_h  (I*);

• • the aqueous mixture M, by means of drying and adjusting the degree of division d₉₀^A2, is used to obtain a finely divided starting material A2 with a particle diameter d₉₀^A2 reported in the length unit μm, with the proviso that 200 μm≧d₉₀^A2≧20 μm;
  • starting material A1 and starting material A2, or starting material A1, starting material A2 and finely divided shaping assistant, are mixed with one another to form a finely divided starting material A3, with the proviso that the starting material A3 comprises the elements other than oxygen introduced into the starting material A3 via starting materials A1 and A2 in the multielement oxide I in the stoichiometry I**

[Bi₁W_b]_a[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_h]₁  (I**),

• • finely divided starting material A3 is used to form geometric shaped bodies V, and
  • the shaped bodies V are treated thermally at elevated temperature to obtain the geometric shaped catalyst bodies K,
    wherein the value F (the stability value F of the geometric shaped catalyst body K) of the product

(d₅₀^A1)^0.7·(d₉₀^A2)^1.5·(a⁻¹)

is ≧820.

Preferably in accordance with the invention, F≧830, preferably ≧840 and even better ≧850.

Particularly advantageously, F≧870 or ≧900, and, particularly advantageously, F≧950 or ≧1000.

Very particularly advantageously, F≧1050, or ≧1100 or ≧1150.

Taking account of the aspect of a satisfactory starting selectivity of target product formation as early as on startup of the catalyst bed, F is preferably ≦2500, frequently ≦2400 or ≦2200.

Favorable values for F are also those which are ≦2000, or ≦1800, or ≦1600 or ≦1500.

In other words, values of F advantageous in accordance with the invention are 2500≧F≧850, or 2450≧F≧900, or 2400≧F≧950.

Values for F which are particularly advantageous in accordance with the invention are 1900≧F≧1000, or 1800≧F≧1050.

Values of F which are very particularly advantageous in accordance with the invention are 1700≧F≧1100, or 1500≧F≧1150.

Advantageously in accordance with the invention, the stoichiometric coefficient a is from 0.2 to 2, particularly advantageously from 0.4 to 1.5 and very particularly advantageously from 0.5 to 1.

Advantageously in accordance with the invention, the particle diameter d₅₀^A1 is 1.2 μm≦d₅₀^A1≦8 μm, particularly advantageously 1.5 μm≦d₅₀^A1≦6 μm, and very particularly advantageously 1.5 μm≦d₅₀^A1≦4 μm, or 2 μm≦d₅₀^A1≦3 μm.

Advantageously in accordance with the invention, the particle diameter d₉₀^A2 is 170 μm≧d₉₀^A1≧30 μm, particularly advantageously 150 μm≧d₉₀^A2≧40 μm, and very particularly advantageously 130 μm≧d₉₀^A2≧50 μm.

To determine particle diameter distributions in dry powders and the particle diameters which can be obtained therefrom, for example d₁₀, d₅₀ and d₉₀, the particular finely divided powder (unless explicitly stated otherwise) was conducted through a dispersing channel into the Sympatec RODOS dry disperser (Sympatec GmbH, System-Partikel-Technik, Am Pulverhaus 1, D-38678 Clausthal-Zellerfeld), dry-dispersed there with compressed air and blown into the test cell in a free jet. In this test cell, the volume-based particle diameter distribution is then determined to ISO 13320 with the Malvern Mastersizer S laser diffraction spectrometer (Malvern Instruments, Worcestershire WR14 1AT, United Kingdom). The particle diameters d_x reported as the measurement results are defined such that X % of the total particle volume consists of particles having this or a smaller diameter. This means that (100−X) % of the total particle volume consists of particles with a diameter >d_x. Unless explicitly stated otherwise in this document, particle diameter determinations and d_x values obtained therefrom, for example d₉₀^Q, d₅₀^A1 and d₉₀^A2, are based on a dispersion pressure employed in the determination (which determines the degree of dispersion of the dry powder during the measurement) of 2 bar absolute.

All data relating to an X-ray diffractogram in this document is based on an X-ray diffractogram obtained using Cu—Kα radiation as the X-radiation (Theta-Theta Bruker D8 Advance diffractometer, tube voltage, 40 kV, tube current: 40 mA, V20 aperture (variable), V20 collimator (variable), detector aperture (0.1 mm), measurement interval (2Θ=2 theta): 0.02°, measurement time per step: 2.4 s, detector: Si semiconductor detector).

The definition of the intensity of a reflection in the X-ray diffractogram is based in this document on the definition laid down in DE-A 198 35 247, and that laid down in DE-A 100 51 419 and DE-A 100 46 672.

In other words, when A¹ denotes the peak location of a reflection 1 and when B¹ denotes the closest pronounced minimum in the line of the X-ray diffractogram when viewed along the intensity axis at right angles to the 2Θ axis (neglecting minima displaying reflection shoulders) to the left of the peak location A¹, and when B² correspondingly denotes the closest pronounced minimum to the right of the peak location A¹, and C¹ denotes the point at which a straight line drawn from the peak location A¹ at right angles to the 2Θ axis intersects with a straight line connecting points B¹ and B², the intensity of the reflection 1 is the length of the straight line section A¹C¹ which extends from the peak location A¹ to the point C¹. The expression “minimum” means a point at which the gradient of slope of a tangent applied to the curve in a base region of the reflection 1 goes from a negative value to a positive value, or a point at which the gradient of slope tends to zero, employing the coordinates of the 2Θ axis and of the intensity axis to determine the gradient of slope. An illustrative performance of an intensity determination is shown by FIG. 6 in DE-A 100 46 672. Detailed remarks regarding intensity determination of X-ray diffraction reflections can also be found in DE-A 101 22 027. Statements regarding half-height widths of diffraction lines are based in this document, correspondingly, on the length of the straight line section which arises between the two points of intersection H¹ and H² when a parallel to the 2Θ axis is drawn in the middle of the straight line section A¹C¹, where H¹, H² mean the first point of intersection in each case of this parallel line with the line in the X-ray diffractogram as defined above to the left and to the right of A¹. In general, the half-height widths of X-ray diffraction reflections of the multielement oxide I active materials are ≦1°, and usually ≦0.5°.

All figures in this document for specific surface areas of solids are based on determinations to DIN 66131 (determination of the specific surface area of solids by gas adsorption (N₂) according to Brunauer-Emmet-Teller (BET)), unless explicitly stated otherwise.

The inventive requirement that each source of the elements other than oxygen in the components T=[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_hO_y]₁ of the multielement oxide i, in the course of preparation of the aqueous mixture M, must pass through a degree of division Q whose diameter d₉₀^Q is ≦5 μm expresses that it is quite possible to proceed from a coarser-grain source (from a coarser-grain starting material). However, on the route of incorporation of such a source into the aqueous mixture M, this source must at least once meet the requirement d₉₀^Q≦5 μm (of course, d₉₀^Q is always >0 μm).

In principle, the condition d₉₀^Q≦5 μm is met when a source is dissolved in a solvent (for example in an aqueous medium; the term “dissolution” is meant here in the sense of molecular or ionic dissolution), and the resulting solution is used to prepare the aqueous mixture M.

This is caused by the fact that, when a source (starting compound, starting substance) is dissolved in a solvent, the source is distributed molecularly or ionically in the solvent. This means that the largest geometric unit of the dissolved starting substance (source) present in the solution unavoidably has “molecular” dimensions which are thus necessarily significantly smaller than 5 μm. It will be appreciated that it is also possible for more than one source (where one source may also comprise more than one element of the component T and hence simultaneously be the source of more than one element) of an element of the component T to be dissolved in one and the same solution, and for the resulting solution to be used to prepare the aqueous mixture M.

However, the requirement d₉₀^Q≦5 μm is also met when a source of an element of the component T is present in a solvent in a colloidal solution.

Colloidal solutions constitute a connection between true (molecular or ionic) solutions and suspensions. In these colloidally dispersed systems there are relatively small accumulations of molecules or atoms which, however, are detectable neither with the naked eye nor with a microscope.

The colloidal solution appears visually to be entirely clear (though it is often colored), since the particles present therein have only a diameter of from 1 to 250 nm (preferably to 150 nm and more preferably to 100 nm), and the corresponding d₉₀^Q is therefore necessarily ≦5 μm. Owing to the small size, a removal of the colloidally dissolved particles by conventional filtration is not possible. They can, however, be separated from their “solvent” by ultrafiltration with membranes of vegetable, animal or synthetic origin (for example parchment, pig's bladder or cellophane). In contrast to the “visually empty” true (molecular or ionic) solutions, a light beam cannot pass through a colloidal solution without being deflected. The light beam is scattered and deflected by the colloidally dissolved particles. In order to keep colloidal solutions stable and to prevent further particle agglomerations, they frequently comprise added wetting and dispersing assistants and other additives.

For example, the element silicon in the process according to the invention can be introduced in the form of a silica sol to prepare the aqueous mixture M. Silica sols are colloidal solutions of amorphous silicon dioxide in water. They are as mobile as water and do not comprise any sedimentable constituents. Their SiO₂ content may be up to 50% by weight and more with a shelf life often of several years (without sedimentation).

However, the requirement d₉₀^Q≦5 μm is also met when a source is, for example, dry-comminuted (for example by grinding) to this particle size.

In principle, such a powder can be used directly as such to prepare the intimate aqueous mixture M. Of course, it can also be suspended in a liquid medium and then used in the form of this suspension to prepare the aqueous mixture M.

Preferably in accordance with the invention, d₉₀^Q for all sources used to prepare the aqueous mixture M (starting compounds, starting substances) is ≦4 μm or ≦3 μm, more preferably ≦2 μm or ≦1 μm and most preferably ≦0.8 μm or ≦0.5 μm. Even better, d₉₀^Q for all sources used to prepare the aqueous mixture M (starting compounds, starting substances) is ≦0.3 μm or ≦0.2 μm.

Particular preference is given to those processes according to the invention in which, in the course of preparation of the aqueous mixture M, all sources used of the elements of the component T pass through the state of a colloidal solution or of a true (molecular or ionic) solution (the resulting aqueous mixtures M shall be referred to in this document as aqueous mixtures M^L).

Very particular preference is given to those processes according to the invention in which, in the course of preparation of the aqueous mixture M, all sources used of the elements other than silicon in the component T pass through the state of a true (molecular or ionic) solution (the resulting aqueous mixtures M shall be referred to in this document as aqueous mixture M^L*). When the aqueous mixture M further comprises a source of the element silicon, it is advantageously a colloidal solution thereof (more preferably a silica sol). Such aqueous mixtures M shall be referred to in this document as aqueous mixtures M^L**.

An intimate aqueous mixture M shall be understood in this document to mean those mixtures M whose component which escapes in gaseous form on conversion from the aqueous mixture M to the finely divided starting material A2 consists of water vapor to an extent of at least 50% of its weight, advantageously to an extent of at least 60% of its weight, particularly advantageously to an extent of at least 70% of its weight, very particularly advantageously to an extent of 80% of its weight and even better to an extent of 90% of its weight. As well as water, the aforementioned component which escapes in gaseous form may also comprise compounds such as HCl, HNO₃, carbon dioxide, ammonia, alcohols (e.g. methanol, ethanol, glycol and glycerol), ketones, for example acetone or other organic compounds which are soluble in water under standard conditions (1 atm, 25° C.).

Useful sources of the elements of the component T of the desired inventive multielement oxide active material I are in principle those compounds which are already oxides and/or those compounds which are convertible to oxides by heating, at least in the presence of molecular oxygen.

As well as the oxides, useful such starting compounds (sources) may in particular be halides, nitrates, formates, oxalates, citrates, acetates, carbonates, amine complexes, ammonium salts and/or hydroxides (and the hydrates of the aforementioned salts).

A favorable Mo source is ammonium heptamolybdate tetrahydrate. However, it is also possible in principle to use molybdenum trioxide, for example. Z¹ sources favorable in accordance with the invention are the nitrates and nitrate hydrates of the Z¹ elements. Z² sources advantageous in accordance with the invention are the hydroxides and nitrates of the Z² elements and hydrates thereof. For the element iron, an iron nitrate hydrate is advantageously used in the process according to the invention.

Silica sol constitutes the Si source preferred in accordance with the invention. Lanthanides preferred in accordance with the invention are Er, Tb, Ho, Eu, Tm, Nd, Lu, Dy, Gd, Ce and Sm. The sources thereof used are preferably, just as in the case of La and Y, the corresponding nitrate hydrates.

In addition to the relevant sources of the elements of component T of the multielement oxide I, it is also possible for substances which decompose and/or are destroyed (converted chemically) to form compounds which escape in gaseous form at least under the conditions of the thermal treatment of the geometric shaped bodies V to form the geometric shaped catalyst bodies K to be incorporated into the particular aqueous mixture M. Substances of this kind may, for example, function as pore formers and be included for the purpose of adjusting the active internal surface area. Useful such (auxiliary) substances include, for example, NH₄OH, (NH₄)₂CO₃, NH₄HCO₃, NH₄NO₃, urea, NH₄CHO₂, H₂CO₃, HNO₃, H₂SO₄, NH₄CH₃CO₂, NH₄Cl, HCl, NH₄HSO₄, (NH₄)₂SO₄, ammonium oxalate, hydrates of the aforementioned compounds and organic substances, for example stearic acid, malonic acid, ammonium salts of the aforementioned acids, starches (e.g. potato starch and corn starch), cellulose, ground nutshells, finely ground polymer (e.g. polyethylene, polypropylene), etc.

Preferably in accordance with the invention, the finely divided starting material A2 is obtained from the aqueous mixture M (especially in the case of an aqueous mixture ML, or M^L*, or M^L**) by spray-drying thereof. This means that the aqueous mixture M in this case is first divided into fine droplets and these are then dried. Preference is given in accordance with the invention to effecting the drying in a hot air stream. In principle, the aforementioned spray drying can also be accomplished using other hot gases (e.g. nitrogen, or nitrogen-diluted air or else other inert gases).

The spray drying can in principle be effected either in cocurrent or in countercurrent flow of the droplets to the hot gas. It is preferably effected in countercurrent flow of the droplets to the hot gas, more preferably in hot air countercurrent flow. Typical gas inlet temperatures are in the range from 250 to 450° C., preferably from 270 to 370° C. Typical gas outlet temperatures are in the range from 100 to 160° C.

Preferably in accordance with the invention, the spray drying is carried out in such a way as to directly give rise to the particle diameter d₉₀^A2 desired for the finely divided starting material A2 as a result of the spray drying (i.e. is the corresponding d₉₀ of the resulting spray powder), such that the resulting spray powder can directly constitute the starting material A2 for use in accordance with the invention.

When the degree of division of the resulting spray powder is too small compared to the desired d₉₀^A2, it can be coarsened in a controlled manner, for example by subsequent compaction, to the degree of division desired for the starting material A2. Conversely, the fineness of the spray powder resulting directly in the spray drying can also be increased if required by grinding to the degree of division desired for the starting material A2.

It will be appreciated that the intimate aqueous mixture M can also first be dried by conventional evaporation (preferably under reduced pressure; the drying temperature generally should not exceed 150° C.), and the resulting dry material can be adjusted by subsequent comminution to the degree of division d₉₀^A2 required in accordance with the invention. In principle, the aqueous mixture M can also be dried in the process according to the invention by freeze drying.

• Z¹ in the process according to the invention is preferably exclusively Co.
• Z² in the process according to the invention is preferably K, Cs and/or Sr, more preferably K.
• Z⁴ in the process according to the invention is preferably Si.

The stoichiometric coefficient b is advantageously from 0.5 to 4 or to 3, particularly advantageously from 1 to 2.5 and very particularly advantageously from 1.5 to 2.5.

The stoichiometric coefficient c is preferably from 3 to 8, particularly advantageously from 4 to 7 and very particularly advantageously from 5 to 6.

The stoichiometric coefficient d is advantageously from 0.02 to 2 and particularly advantageously from 0.03 to 1 or from 0.05 to 0.5.

The stoichiometric coefficient e is advantageously from 0.1 to 4.5, preferably from 0.5 to 4 and more preferably from 1 to 4 or from 2 to 4.

The stoichiometric coefficient g is preferably from >0 to 10, more preferably from 0.1 to 8 or from 0.2 to 7, even more preferably from 0.3 to 6 or from 0.4 to 5, and most advantageously from 0.5 to 3 or from 1 to 3.

The stoichiometric coefficients h and f may both simultaneously be 0, but may also independently assume values other than 0. Advantageously, the component T does not comprise any Bi.

This means that the working examples B1 to B8 and the comparative examples V1 to V6 (including the subsequent use as catalysts for propene partial oxidation) can also be carried out under otherwise unchanged conditions using finely divided starting materials A2 whose corresponding stoichiometry I* is Mo₁₂Co_6.0Fe_3.0Si_1.6K_0.08, or Mo₁₂Co_6.5Fe_3.0Si_1.6K_0.08, or Mo₁₂Co_7.0Fe_3.0Si_1.6K_0.08, or Mo₁₂Co_5.0Fe_3.0Si_1.6K_0.08, or Mo₁₂Co_4.5Fe_3.0Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_2.5Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.5Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_4.0Si_1.6K_0.08, or Mo₁₂Co_7.0Fe_4.0Si_1.6K_0.08, or Mo₁₂Co_6.0Fe_3.5Si_1.6K_0.08, or Mo₁₂Co_7.0Fe_2.0Si_1.6K_0.08, or Mo₁₂Co_6.0Fe_2.5Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Si_0.5K_0.08, or Mo₁₂Co_5.5Fe_3.0Si₃K_0.08, or Mo₁₂Co_5.5Fe_3.0Si_1.6K_0.04, or Mo₁₂Co_5.5Fe_3.0Si_1.6K_0.2, or Mo₁₂Ni_3.0Co_2.5Fe_3.0Si_1.6K_0.08, or Mo₁₂Ni_3.0Co₄Fe_3.0Si_1.6K_0.08, or Mo₁₂Sb_0.2Co_4.2Fe_1.4Zn_0.2W_0.1K_0.06, or Mo₁₂Sb_0.2Co_4.2Fe_1.4Zn_0.2Bi_0.9W_0.1K_0.06, or Mo₁₂Ni_2.8Co_5.2Fe_1.8K_0.1, or Mo₁₂Ni_2.8Co_5.2Fe_1.81311.7K_0.1, or Mo₁₂Co₅Fe₁Ni₃W_0.5K_0.1, or Mo₁₂Co₅Fe₁Ni₃W_0.5Bi₁K_0.1, or Mo₁₂Co_5.5Fe_3.0Bi_0.02Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Bi_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Bi_0.1Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Bi_0.2Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Bi_0.5Si_1.6K_0.08, or Mo₁₂Co₇Fe_3.0Bi_0.06Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Gd_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Y_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Er_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Er_0.25Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Sm_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Eu_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Dy_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Y_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Tb_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Ho_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0Ce_0.05Si_1.6K_0.08, or Mo₁₂Co_5.5Fe_3.0La_0.05Si_1.6K_0.08. Elements which are not present in the finely divided starting material A2-4 of example B4 but are comprised by the above stoichiometries are dissolved in the solution B using nitrate hydrates thereof as the source. In a departure from this, W is added to the solution A as ammonium paratungstate. The stoichiometric coefficient a may in all cases also be 0.5, or 0.7, or 0.8. At the same time, in all aforementioned configurations, d₅₀^A1 may be 2.4 μm and d₉₀^A2 may be 68 μm.

The finely divided mixed oxide Bi₁W_bO_x can be preformed in a manner known per se (cf., for example, EP-A 575 897, DE-A 3338380, EP-A 835, WO 02/24620, WO 2007/017431, German application 102007003778.5, WO 2005/030393 and German application 102008040093.9).

In general, at least one source of the element Bi and at least one source of the element W (i.e. at least one starting compound comprising the element Bi and at least one starting compound comprising the element W) will be mixed intimately with one another in aqueous medium, the aqueous mixture will be dried and the resulting dry material will be calcined (treated thermally) at temperatures in the range from 400 to 900° C. (preferably from 600 to 900° C. and more preferably from 700 to 900° C.), and the particle diameter ca required in accordance with the invention will be established by comminuting the resulting calcination products to obtain the finely divided starting material A1. Useful sources of the Bi and W are in principle those compounds which are already oxides of these elements, or those compounds which are convertible to oxides by heating, at least in the presence of molecular oxygen.

Preferably, water-soluble salts of bismuth such as nitrates, carbonates, hydroxides and/or acetates, with tungstic acid (the essentially water-insoluble tungstic acid is preferably used in the form of fine powder whose d₉₀ is appropriately in application terms ≦5 μm or ≦2 μm, preferably from 0.1 to 1 μm) and/or ammonium salts thereof will be mixed into water, the aqueous mixture will be dried (preferably spray-dried), and the dried material will subsequently be treated thermally as described.

When the drying has been effected by spray drying, the resulting spray powder will advantageously be coarsened before the calcination (for example, advantageously in application terms, will be pasted with addition of up to 20% by weight of water and, for example, extruded by means of an extruder to give extrudates which are easy to handle for calcination purposes; these are subsequently dried and then calcined).

Typically, the thermal treatment is effected in an airstream (for example, in the case of the aforementioned extrudates, in a rotary tube oven as described in DE-A 103 25 487). The division of the resulting calcined mixed oxide to the particle diameter d₅₀^A1 which is essential in accordance with the invention will normally be brought about by grinding in mills. If required, the millbase is subsequently classified to the desired degree of comminution.

Preferred mixed oxides Bi₁W_bO_x formed beforehand in the process according to the invention are the mixed oxides Bi₁W_2.5O₉ (½Bi₂W₂O₉.1.5 WO₃), Bi₁W₃O_10.5 (½Bi₂W₂O₉.2WO₃), Bi₁W₄O_13.5 (½Bi₂W₂O₉.3WO₃), Bi₁W_0.5O₃, Bi₁W₁O_4.5 (½Bi₂W₂O₉), Bi₁W₂O_7.5 (½Bi₂W₂O₉.1WO₃) and Bi₁W_1.5O₆ (½Bi₂W₂O₉.½WO₃), among which Bi₁W₂O_7.5 is very particularly preferred in accordance with the invention (working examples B1 to B8 and comparative examples V1 to V6 (including the use thereof for propene partial oxidation) can therefore also be implemented using Bi₁W_1.5O₆, or Bi₁W_2.5O₉, or Bi₁W₃O_10.5, or Bi₁W₄O_13.5, or Bi₁W_0.5O₃, or Bi₁W₁O_4.5 as the finely divided starting material A1 of particle size corresponding to the finely divided Bi₁W₂O_7.5 used).

In the same way as they are incorporated into the aqueous mixture M, it is also possible to additionally incorporate, into the aqueous mixture of the at least one Bi source and of the at least one W source, in the course of preparation of the mixed oxide Bi₁W_bO_x, substances which, under the conditions of the thermal treatment employed to form the mixed oxide Bi₁W_bO_x, decompose and/or are destroyed (converted chemically) to give compounds which escape in gaseous form. Such substances may, for example, function as pore formers and be included for the purpose of influencing the active internal surface area of the mixed oxide Bi₁W_bO_x. Useful (auxiliary) substances of this kind include, for example, NH₄OH, (NH₄)₂CO₃, NH₄HCO₃, NH₄NO₃, urea, NH₄CHO₂, H₂CO₃, HNO₃, H₂SO₄, NH₄CH₃CO₂, NH₄HSO₄, NH₄Cl, HCl, (NH₄)₂SO₄, ammonium oxalate, hydrates of the aforementioned compounds, and organic substances, for example stearic acid, malonic acid, ammonium salts of the aforementioned acids, starches (e.g. potato starch and corn starch), cellulose, ground nutshells, finely ground polymer (e.g. polyethylene, polypropylene), etc.

In the preparation of the finely divided starting material A3 from the finely divided starting materials A1 and A2, appropriately in application terms, but not necessarily, finely divided shaping assistants are used.

These may already have been mixed into the two finely divided starting materials A1 and A2 or into only one of the two finely divided starting materials A1, A2 before these finely divided starting materials are mixed.

It will be appreciated that the finely divided shaping assistants may also or only (not until this point) be mixed into the finely divided mixture of finely divided starting material A1 and finely divided starting material A2.

The group of the finely divided shaping assistants (especially when they comprise elements Z⁴, the shaping assistants are used with a d₅₀ of >5 μm; they typically do not comprise any elements of the multielement oxide I other than oxygen) includes first the so-called anticaking agents.

These are finely divided materials which are used advantageously in application terms in order to very substantially suppress, in the course of mixing, for example, reagglomeration (“caking together”) of particles within the starting material A1 and/or within the starting material A2, since such a reagglomeration might influence the effective particle diameter. A group of finely divided anticaking agents preferred in accordance with the invention is that of finely divided hydrophobized silicas, especially finely divided hydrophobized synthetic silicas (silicon dioxides). Synthetic silicas can be obtained firstly directly by pyrogenic means from sand and secondly by precipitation reactions from waterglass. Synthetic silicas in particular, owing to their surface OH groups, are hydrophilic, i.e. they are wetted by water. For example, reaction of these surface OH groups with chlorosilanes makes it possible to prepare hydrophobized products both from the fumed silicas and from the precipitated silicas. For example, the hydrophobization can be accomplished by reaction with dimethyldichlorosilane in the presence of water vapor at approx. 400° C. in a fluidized bed reactor (is preferably employed in the case of fumed silicas).

In the case of precipitated silicas in particular, the chlorosilane is added to the precipitation suspension at a temperature of from 50 to 90° C. while stirring vigorously. This is followed by filtration, washing to neutrality with water, drying of the filtercake and heat treatment at from 300 to 400° C. H. Brunner, D. Schutte, Chem. Ing. Techn. 89, 437 (1965) and DT 2435860 and DT 1117245 describe the preparation of hydrophobized fine silicas in detail. Commercial products of hydrophobized precipitated silicas are, for example, the SIPERNAT® brands.

Preferably in accordance with the invention, the finely divided anticaking agent Sipernat® D17 from Degussa or from EVONIK Industries is used. Based on its weight, Sipernat® D17 comprises about 2% by weight of chemically bound carbon and is not wetted by water. Its tapped density (to ISO 787-11) is 150 g/l. Its d₅₀ value is 10 μm (laser diffraction to ISO 13320-1) and the specific surface area (nitrogen adsorption to ISO 5794-1, Annex D) is 100 m²/g.

Advantageously, finely divided anticaking agent, for example Sipernat® D17, is mixed into the finely divided starting material A1 before it is mixed with the finely divided starting material A2 to give the finely divided starting material A3. In general, the amount of finely divided anticaking agent added is from 0.1 to 3% by weight, based on the weight of the finely divided starting material A1.

The addition of anticaking agent also reduces the energy input required for homogeneous mixing of the two starting materials A1 and A2, which has an advantageous effect especially on obtaining the particle size of the finely divided starting material A2 in the course of mixing.

When the inventive shaping of the finely divided starting material A3 to give the geometric shaped bodies V is effected advantageously in accordance with the invention by compaction (or compression), it is appropriate in application terms to add lubricants as further finely divided shaping assistants to the finely divided starting material A3, for example graphite, carbon black, polyethylene glycol, polyacrylic acid, stearic acid, starch, mineral oil, vegetable oil, water, boron trifluoride and/or boron trinitride. Descriptions of additional use of lubricants in the context of a corresponding shaping can be found, for example, in documents DE-A 102007004961, WO 2005/030393, US-A 2005/0131253, WO 2007/017431, DE-A 102007005606 and in German application 102008040093.9. Preference is given in accordance with the invention to using exclusively finely divided graphite as a lubricant. Graphites added with preference are Asbury 3160 and Asbury 4012 from Asbury Graphite Mills, Inc. New Jersey 08802, USA, and Timrex® T44 from Timcal Ltd., 6743 Bodio, Switzerland.

Advantageously, the finely divided graphite (typical d₉₀ values of graphites suitable in accordance with the invention are from 30 to 300 μm) is added first to the mixture of finely divided starting material A1 and finely divided starting material A2. However, it can also be mixed into each of the two finely divided starting materials A1, A2 (or only into one of the two) before the mixing thereof. Based on the weight of the finely divided starting material A3, it may comprise, for example, up to 15% by weight of finely divided lubricant. Usually, the lubricant content in the finely divided starting material A3 is, however, ≦9% by weight, in many cases ≦5% by weight, often ≦4% by weight; this is especially true when the finely divided lubricant is graphite. In general, the aforementioned added amount is ≦0.5% by weight, usually ≦2.5% by weight.

If required, further shaping assistants which may also be added to the finely divided starting material A3 are also finely divided reinforcing agents such as microfibers of glass, asbestos, silicon carbide or potassium titanate, which, after the shaping by compaction has ended, have a beneficial effect on the integrity of the resulting compact (of the resulting shaped body V).

In the course of the inventive thermal treatment of the shaped bodies V, from which the shaped catalyst bodies K accrue, additionally used shaping assistants may either remain in the resulting shaped catalyst body K or escape therefrom at least partly in gaseous form as a result of thermal and/or chemical decomposition to gaseous compounds (e.g. CO, CO₂). Shaping assistants which remain in the shaped catalyst body K, in the course of catalytic use thereof, have an essentially exclusively diluting action on the multielement oxide I active material.

In general, the finely divided starting material A3 is compacted to the desired geometry of the shaped body V (of the geometric shaped catalyst precursor body) through the action of external forces (pressure) on the finely divided precursor mixture. The shaping apparatus to be employed and the shaping method to be employed are not subject to any restriction.

For example, the compactive shaping can be effected by extrusion or tableting. In this case, the finely divided starting material A3 is preferably used dry to the touch. However, it may comprise, for example, up to 10% by weight of its total weight of added substances which are liquid under standard conditions (25° C., 1 atm). It is also possible for the finely divided starting material A3 to comprise solid solvates (e.g. hydrates) which contain such liquid substances in chemically and/or physically bound form. It will be appreciated, however, that the finely divided starting material A3 may also be entirely free of such substances.

A shaping process preferred in accordance with the invention for compacting the finely divided starting material A3 is tableting. The basics of tableting are described, for example, in “Die Tablette”, Handbuch der Entwicklung, Herstellung and Qualitätssicherung [“The Tablet”, Handbook of Development, Production and Quality Assurance], W. A. Ritschel and A. Bauer-Brandl, 2nd edition, Edition Verlag Aulendorf, 2002, and can be applied in an entirely corresponding manner to a tableting process according to the invention.

Advantageously, an inventive tableting is performed as described in the documents WO 2005/030393, German application 102008040093.9, German application 102008040094.7 and WO 2007/017431.

Instead of compacting the finely divided starting material A3 as such directly to the desired geometry of the shaped body V (in a single compaction step), it is frequently appropriate in accordance with the invention to perform, as the first shaping step, first an intermediate compaction in order to coarsen the finely divided starting material A3 (generally to particle diameters of from 100 to 2000 μm, preferably from 150 to 1500 μm, more preferably from 400 to 1250 μm, or from 400 to 1000 μm, or from 400 to 800 μm).

Even before the intermediate compaction, it is possible to add, for example, finely divided lubricant (e.g. graphite). Subsequently, the final shaping is effected on the basis of the coarsened powder, if required with addition beforehand once again, for example, of finely divided lubricant (e.g. graphite) and if appropriate further shaping and/or reinforcing assistants.

Just like the shaping apparatus for use to compact the finely divided starting material A3 and the shaping method to be employed, the desired geometry of the resulting shaped bodies V in the process according to the invention is not subject to any restriction either. In other words, the shaped catalyst precursor bodies (the shaped bodies V) may either have a regular or irregular shape, preference being given in accordance with the invention generally to regularly shaped bodies V.

For example, the shaped body V in the process according to the invention may have spherical geometry. In this case, the sphere diameter may, for example, be from 2 to 10 mm, or from 4 to 8 mm. The geometry of the shaped catalyst precursor body (of the shaped body V) may, however, also be solid cylindrical or hollow cylindrical (annular). In both cases, external diameter (E) and height (H) may, for example, be from 2 to 10 mm, or from 2 or 3 to 8 mm. In the case of solid cylinders, the external diameter may also be from 1 to 10 mm. In the case of hollow cylinders (rings), a wall thickness of from 1 to 3 mm is generally appropriate. It will be appreciated that useful catalyst precursor geometries also include those disclosed and recommended in WO 02/062737.

The shaping pressures employed in the course of compaction of finely divided starting material A3 will, in the process according to the invention, generally be from 50 kg/cm² to 5000 kg/cm². The shaping pressures are preferably from 200 to 3500 kg/cm², more preferably from 600 to 25 000 kg/cm².

Especially in the case of annular shaped bodies V, the shaping compaction in the process according to the invention, following the teaching of the documents German application 102008040093.9, German application 102008040094.7 and WO 2005/030393, shall be carried out in such a way that the side crushing strength SCS of the resulting annular shaped body V is 12 N≦SCS≦25 N. SD is preferably ≧13 N and ≦24 N, or ≧14 N and ≦22 N, and most preferably ≧15 N and ≦20 N.

The experimental determination of the side crushing strength is carried out as described in the documents WO 2005/030393 and WO 2007/017431. It will be appreciated that ringlike shaped bodies V, as recommended by German application 102008040093.9, are very particularly preferred in accordance with the invention. The end face of annular or ringlike shaped bodies V may either be curved or uncurved in the process according to the invention (cf. especially DE-A 102007004961, EP-A 184 790 and German application 102008040093.9). In determining the height of such geometric shaped bodies V, such curvature is not taken into account.

Shaped catalyst bodies K which are obtainable in accordance with the invention and have been produced by thermal treatment of shaped bodies V which have been obtained by compaction of finely divided starting material A3 are referred to as unsupported catalysts (shaped unsupported catalyst bodies K).

Ring geometries of shaped bodies V obtainable by compacting finely divided starting material A3 which are particularly advantageous in accordance with the invention satisfy the condition H/E=0.3 to 0.7. More preferably, H/E=0.4 to 0.6.

It is additionally favorable in accordance with the invention for the aforementioned annular shaped bodies V when the I/E ratio (where I is the internal diameter of the ring geometry) is from 0.3 to 0.7, preferably from 0.4 to 0.7.

Aforementioned ring geometries are particularly advantageous when they simultaneously have one of the advantageous H/E ratios and one of the advantageous I/E ratios. Such possible combinations are, for example, H/E=0.3 to 0.7 and I/E=0.3 to 0.8 or 0.4 to 0.7. Alternatively, H/E may be from 0.4 to 0.6, and I/E may simultaneously be from 0.3 to 0.8 or from 0.4 to 0.7. It is also favorable for the relevant ring geometries when H is from 2 to 6 mm and preferably from 2 to 4 mm. It is additionally advantageous when E for the rings is from 4 to 8 mm, preferably from 4 to 6 mm. The wall thickness of ring geometries preferred in accordance with the invention is from 1 to 1.5 mm.

Possible ring geometries for aforementioned annular shaped bodies V are thus (A×H×I) 5 mm×2 mm×2 mm, or 5 mm×3 mm×2 mm, or 5 mm×3 mm×3 mm, or 5.5 mm×3 mm×3.5 mm, or 6 mm×3 mm×4 mm, or 6.5 mm×3 mm×4.5 mm, or 7 mm×3 mm×5 mm, or 7 mm×7 mm×3 mm, or 7 mm×7 mm×4 mm.

The thermal treatment of inventive shaped bodies V (especially annular shaped bodies V; everything which follows applies especially to their thermal treatment) to obtain the geometric shaped catalyst bodies K is effected, in the process according to the invention, generally at temperatures (in this document, this generally means the temperature within the calcination material) which exceed 350° C. Normally, in the course of the thermal treatment, the temperature of 650° C. is, however, not exceeded. Advantageously in accordance with the invention, in the course of thermal treatment, the temperature of 600° C., preferably the temperature of 550° C. and more preferably the temperature of 500° C. will not be exceeded.

In addition, in the course of the thermal treatment of the shaped bodies V, preferably the temperature of 380° C., advantageously the temperature of 400° C., particularly advantageously the temperature of 420° C. and most preferably the temperature of 440° C. will not be exceeded. At the same time, the thermal treatment can also be divided into several sections over the course of its duration. For example, a thermal treatment can first be carried out at a temperature (phase 1) of from 150 to 350° C., preferably from 220 to 290° C., and then a thermal treatment at a temperature (phase 2) of from 400 to 600° C., preferably from 430 to 550° C.

Normally, the thermal treatment of the shaped bodies V takes several hours (frequently more than 5 h). In many cases, the total duration of the thermal treatment extends to more than 10 h. Usually, in the course of the thermal treatment of the shaped bodies V, treatment times of 45 h or 25 h are not exceeded. The total treatment time is often below 20 h. In principle, the thermal treatment can be carried out at higher temperatures over a shorter treatment time or at not excessively high temperatures over a longer treatment time. In an embodiment of the thermal treatment of the shaped bodies V which is advantageous in accordance with the invention, 465° C. is not exceeded and the treatment time in the 440° C. temperature window extends to from >10 to 20 h. In another embodiment which is advantageous in accordance with the invention (and is preferred for the inventive purposes) of the thermal treatment of the shaped bodies V, 465° C. (but not 500° C.) is exceeded and the treatment time in the temperature window of 465° C. extends to from 2 to 10 h.

This means that the end calcination in all working examples B1 to B8 and in all comparative examples V1 to V6 can also be carried out under otherwise unchanged conditions (including the subsequent use as catalysts for propene partial oxidation) at an end temperature of 450, or 452, or 454, or 456, or 458, or 460, or 462, or 464, or 466, or 468, or 470, or 472, or 474, or 476, or 478, or 480, or 485, or 490, or 495, or 500, or 505 or 510° C.

The end calcination in all working examples B1 to B8 and in all comparative examples V1 to V6 (including the subsequent use as catalysts for the propene partial oxidation) can, however, also be carried out under otherwise unchanged conditions over an end calcination time shortened to 9, or 8, or 7, or 6, or 5, or 4, or 3, or 2, or 1 h, at an end calcination temperature increased in each case by 2, or 4, or 6, or 8, or 10, or 12, or 14, or 16, or 20° C.

The thermal treatment (including phase 1 (also referred to as decomposition phase)) of the shaped bodies V can be effected either under inert gas or under an oxidative atmosphere, for example air (or another mixture of inert gas and oxygen), or else under reducing atmosphere (for example a mixture of inert gas, NH₃, CO and/or H₂, or under methane, acrolein, methacrolein). Of course, the thermal treatment can also be performed under reduced pressure. It is also possible to vary the calcination atmosphere over the calcination time. Preferably in accordance with the invention, the thermal treatment of the shaped bodies V is effected in an oxidizing atmosphere. Appropriately in application terms, the atmosphere consists predominantly of stationary or moving air.

In principle, the thermal treatment of the shaped bodies V can be effected in a wide variety of different oven types, for example heatable forced-air chambers (forced-air ovens), tray ovens, rotary tube ovens, belt calciners or shaft ovens. Advantageously in accordance with the invention, the thermal treatment of the shaped bodies V is effected in a belt calcining apparatus, as recommended by DE-A 10046957 and WO 02/24620. This very substantially prevents hotspot formation within the calcination material, by virtue of increased volume flows of calcination atmosphere being conveyed through the calcination material through a gas-permeable conveyor belt which bears the calcination material with the aid of ventilators.

The thermal treatment of the shaped bodies V below 350° C. generally pursues the aim of thermal decomposition of the sources of the elements (of the elemental constituents), present in the shaped bodies V, of the desired multielement oxide I active material of the shaped catalyst bodies K and of any additionally used shaping assistants. This decomposition phase is frequently effected in the course of heating the calcination material to temperatures of 350° C.

In principle, the thermal treatment can be effected as described in US 2005/0131253.

Typically, the side crushing strengths of annular shaped unsupported catalyst bodies K obtainable in accordance with the invention as described are from 5 to 13 N, frequently from 8 to 11 N.

Shaped unsupported catalyst bodies K produced in accordance with the invention need not necessarily be used as such as catalysts for heterogeneously catalyzed partial gas phase oxidations of alkanes, alkanols, alkenes and/or alkenals having from 3 to 6 carbon atoms. Instead, they can also be subjected to grinding and the resulting finely divided material (if appropriate after classification of the resulting finely divided material) can be applied with the aid of a suitable liquid binder (e.g. water) to the surface of a suitable, for example spherical or annular, support body (geometric shaped support body) (for example employing the process principle disclosed in DE-A 2909671 and DE-A 100 51 419). After drying or immediately after application of the active material coating to the support body, the resulting coated catalyst can be used as a catalyst for aforementioned heterogeneously catalyzed gas phase partial oxidations, as described, for example, by WO 02/49757 and DE-A 10122027 for similar active materials.

The support materials used in the above procedure may be customary porous or nonporous aluminum oxides, silicon dioxide, zirconium dioxide, silicon carbide, or silicates such as magnesium or aluminum silicate. The support bodies may have a regular or irregular shape, preference being given to regularly shaped support bodies with distinct surface roughness (for example the spheres or rings already mentioned). It is particularly advantageous to use essentially nonporous, rough-surface rings made of steatite, whose longest dimension (longest direct straight line connecting two points on the surface of the shaped support body) is typically from 2 to 12 mm, frequently from 4 to 10 mm (cf. also DE-A 4442346). The aforementioned longest dimensions are also useful for other shaped support bodies, for example spheres, solid cylinders and other rings.

The layer thickness of the active material coating (powder material) applied to the shaped support body will appropriately be selected within the range from 10 to 1000 μm, preferably in the range from 100 to 700 μm and more preferably in the range from 300 to 500 μm. Possible coating thicknesses are also from 10 to 500 μm or from 200 to 300 μm. The surface roughness R, of the shaped support body is frequently in the range from 40 to 200 μm, in many cases in the range from 40 to 100 μm (determined to DIN 4768 sheet 1 with a “Hommel Tester for DIN-ISO surface parameters” from Hommelwerke, Germany). Appropriately, the support material is nonporous (total volume of the pores based on the volume of the support body 1% by volume).

In principle, the shaping (compaction) of the finely divided starting material A3 to a shaped body V can also be effected by applying the finely divided starting material A3 with the aid of a suitable liquid binder to the surface of a geometric shaped support body as described above. After drying, the resulting shaped precursor bodies V can be treated thermally in the inventive manner to obtain inventive shaped coated catalyst bodies K.

It is also possible to use active material powder obtained by grinding shaped unsupported catalyst bodies K produced in accordance with the invention as such in a fluidized or moving bed for the heterogeneously catalyzed partial gas phase oxidations addressed in these documents.

It is found to be favorable for the production process according to the invention when, for the finely divided starting material A2, also, 10 μm≦d₅₀^A2≦50 μm, preferably 20 μm≦d₅₀^A2≦40 μm.

It is also advantageous for the production process according to the invention when, in addition to the inventive condition for the value F, the condition is satisfied for the value F* of the product (the particle diameters d₅₀^A1 (of the finely divided starting material A1), d₅₀^A2 (of the finely divided starting material A2) again being reported in the length unit μm)

(d₅₀^A1)^0.7·(d₅₀^A2)^0.7·(a⁻¹)

that F*≧15 (preferably ≧18), more preferably 25≧F*≧18.

Additionally found to be favorable in accordance with the invention is a ratio of the particle diameter d₉₀^A2 of the finely divided starting material A2 to the particle diameter d₁₀^A2 of the finely divided starting material A2, i.e. d₉₀^A2:d₁₀^A2 (reported in the same length unit) in the range from 5 to 20, preferably in the range from 10 to 15.

The shaped catalyst bodies K obtainable in accordance with the invention are suitable as catalysts for all heterogeneously catalyzed gas phase oxidations for which the geometric shaped catalyst bodies K* have already been mentioned as suitable within this document. Geometric shaped catalyst bodies K obtainable in accordance with the invention are, however, particularly suitable as catalysts for the partial oxidations of propene to acrolein and of isobutene and/or tert-butanol to methacrolein. This is especially true of inventive annular shaped unsupported catalyst bodies K. The partial oxidation can be carried out, for example, as described in documents DE-A 102007004961, WO 02/49757, WO 02/24620, German application 102008040093.9, WO 2005/030393, EP-A 575 897, WO 2007/082827, WO 2005/113127, WO 2005/047224, WO 2005/042459 and WO 2007/017431.

The ring geometries emphasized individually in this document of the annular unsupported catalysts obtainable as described are found to be advantageous especially when the loading of the catalyst charge with propene, isobutene and/or tert-butanol (or methyl ether thereof) present in the starting reaction gas mixture is ≧130 l (STP)/l of catalyst charge·h (upstream and/or downstream beds of inert material are not considered as belonging to the catalyst charge in this document in considerations of loading).

However, annular shaped unsupported catalyst bodies K obtainable as described still have this advantage (“low deactivation rate”) when the aforementioned loading of the catalyst charge is ≧140 l (STP)/l·h, or ≧150 l (STP)/l·h, or ≧160 l (STP)/l·h. Normally, the aforementioned loading of the catalyst charge will be ≦600 l (STP)/l·h, frequently ≦500 l (STP)/l·h, in many cases ≦400 l (STP)/l·h or ≦350 l (STP)/l·h. Loadings in the range from 160 l (STP)/l·h to 300 or 250 or 200 l (STP)/l·h are particularly appropriate.

It will be appreciated that the annular shaped unsupported catalyst bodies K obtainable as described can be operated in a manner advantageous in accordance with the invention as catalysts for the partial oxidation of propene to acrolein or of isobutene and/or tert-butanol (or methyl ether thereof) to methacrolein also at loadings of the catalyst charge with the starting compound to be oxidized partially of ≦130 l (STP)/l·h, or ≦120 l (STP)/l·h, or ≦110 l (STP)/l·h. In general, this loading will, however, be at values of ≧60 l (STP)/l·h, or ≧70 l (STP)/l·h, or ≧80 l (STP)/l·h.

In principle, the loading of the catalyst charge with the starting compound to be oxidized partially (propene, isobutene and/or tert-butanol (or methyl ether thereof)) can be adjusted by means of two adjusting screws:

a) the loading of the catalyst charge with starting reaction gas mixture (the reaction gas mixture which is supplied to the fixed catalyst bed), and/or

b) the content in the starting reaction gas mixture of the starting compound to be oxidized partially.

The shaped unsupported catalyst bodies K, for example annular shaped unsupported catalyst bodies K, obtainable in accordance with the invention are suitable especially when the loading is adjusted in particular by means of the aforementioned adjusting screw a) at loadings of the catalyst charge with the organic compound to be oxidized partially of above 130 l (STP)/l·h.

The propene content (isobutene content or tert-butanol content (or methyl ether content)) in the starting reaction gas mixture will generally (i.e. essentially irrespective of the loading) be from 4 to 20% by volume, frequently from 5 to 15% by volume, or from 5 to 12% by volume, or from 5 to 8% by volume (based in each case on the total volume of the starting reaction gas mixture).

Frequently, the gas phase partial oxidation process of the partial oxidation catalyzed by the shaped unsupported catalyst bodies K, for example annular shaped unsupported catalyst bodies K, obtainable as described (or other geometric shaped catalyst bodies K) will be performed (essentially irrespective of the loading) at a volume ratio of (organic) compound to be oxidized partially (e.g. propene):oxygen:inert gases (including water vapor) in the starting reaction gas mixture of 1:(1.0 to 3.0):(5 to 25), preferably 1:(1.5 to 2.3):(10 to 20).

Inert gases refer to those gases which remain chemically unchanged in the course of the partial oxidation to an extent of at least 95 mol %, preferably to an extent of at least 98 mol %.

In the above-described starting reaction gas mixtures, the inert gas may consist of molecular nitrogen to an extent of ≧20% by volume, or to an extent of ≧30% by volume, or to an extent of ≧40% by volume, or to an extent of ≧50% by volume, or to an extent of ≧60% by volume, or to an extent of ≧70% by volume or to an extent of ≧80% by volume, or to an extent of ≧90% by volume or to an extent of ≧95% by volume.

However, when the loadings of the catalyst charge with the organic compound to be partially oxidized are ≧150 l (STP)/l·h, it is advisable to use inert diluent gases such as propane, ethane, methane, pentane, butane, CO₂, CO, water vapor and/or noble gases for the starting reaction gas mixture. Generally, these inert gases and their mixtures may also be used even at lower loadings of the catalyst charge with the organic compound to be partially oxidized. Cycle gas may also be used as a diluent gas. Cycle gas refers to the residual gas which remains when the target compound is substantially selectively removed from the product gas mixture of the partial oxidation. It has to be taken into account that the partial oxidations to acrolein or methacrolein using the shaped catalyst bodies K, for example annular shaped catalyst bodies K, obtainable in accordance with the invention may only be the first stage of a two-stage partial oxidation to acrylic acid or methacrylic acid as the actual target compounds, in which case the cycle gas is not usually formed until after the second stage. In such a two-stage partial oxidation, the product gas mixture of the first stage is generally fed as such, optionally after cooling and/or secondary oxygen addition, to the second partial oxidation stage.

In the partial oxidation of propene to acrolein using the shaped catalyst bodies K, for example annular shaped catalyst bodies K, obtainable as described, a typical composition of the starting reaction gas mixture measured at the reactor inlet (irrespective of the loading selected) may comprise, for example, the following components:

• • from 6 to 6.5% by volume of propene,
  • from 1 to 3.5% by volume of H₂O,
  • from 0.2 to 0.5% by volume of CO,
  • from 0.6 to 1.2% by volume of 002,
  • from 0.015 to 0.04% by volume of acrolein,
  • from 10.4 to 11.3% by volume of O₂ and,
  • as the remainder ad 100% by volume, molecular nitrogen,
    or:
  • 5.6% by volume of propene,
  • 10.2% by volume of oxygen,
  • 1.2% by volume of CO_x,
  • 81.3% by volume of N₂ and
  • 1.4% by volume of H₂O.

The former compositions are suitable especially at propene loadings of ≧130 l (STP)/l·h and the latter composition especially at propene loadings of <130 l (STP)/l·h, especially 100 l (STP)/l·h, of the fixed catalyst bed.

As alternative compositions of the starting reaction gas mixture, useful compositions (irrespective of the loading selected) are those which have the following components:

• • from 4 to 25% by volume of propene,
  • from 6 to 70% by volume of propane,
  • from 5 to 60% by volume of H₂O,
  • from 8 to 65% by volume of O₂, and
  • from 0.3 to 20% by volume of H₂;
    or
  • from 4 to 25% by volume of propene,
  • from 6 to 70% by volume of propane,
  • from 0 to 60% by volume of H₂O,
  • from 8 to 16% by volume of O₂,
  • from 0 to 20% by volume of H₂
  • from 0 to 0.5% by volume of CO;
  • from 0 to 1.2% by volume of CO₂;
  • from 0 to 0.04% by volume of acrolein;
  • as the remainder ad 100% by volume, essentially N₁;
    or
  • from 50 to 80% by volume of propane,
  • from 0.1 to 20% by volume of propene,
  • from 0 to 10% by volume of H₂,
  • from 0 to 20% by volume of N₂,
  • from 5 to 15% by volume of H₂O, and
  • sufficient molecular oxygen that the molar ratio of oxygen content to propene content is from 1.5 to 2.5;
    or
  • from 6 to 9% by volume of propene,
  • from 8 to 18% by volume of molecular oxygen,
  • from 6 to 30% by volume of propane, and
  • from 32 to 72% by volume of molecular nitrogen.

However, the starting reaction gas mixture may also have the following composition:

• • from 4 to 15% by volume of propene,
  • from 1.5 to 30% by volume (frequently from 6 to 15% by volume) of water,
  • from ≧0 to 10% by volume (preferably from ≧0 to 5% by volume) of constituents other than propene, water, oxygen and nitrogen, and sufficient molecular oxygen that the molar ratio of molecular oxygen present to molecular propene present is from 1.5 to 2.5, and, as the remainder up to 100% by volume of the total amount, molecular nitrogen.

Another possible starting reaction gas mixture composition may comprise:

• • 6.0% by volume of propene,
  • 60% by volume of air and
  • 34% by volume of H₂O.

Alternatively, starting reaction gas mixtures of the composition according to Example 1 of EP-A 990 636, or according to Example 2 of EP-A 990 636, or according to Example 3 of EP-A 1 106 598, or according to Example 26 of EP-A 1 106 598, or according to Example 53 of EP-A 1 106 598, may also be used.

The shaped catalyst bodies K, for example annular shaped catalyst bodies K, obtainable as described are also suitable for the processes of DE-A 10246119 and DE-A 10245585.

Further starting reaction gas mixtures which are suitable in accordance with the invention may lie within the following composition framework:

• • from 7 to 11% by volume of propene,
  • from 6 to 12% by volume of water,
  • from ≧0 to 5% by volume of constituents other than propene, water, oxygen and nitrogen,
  • sufficient molecular oxygen that the molar ratio of molecular oxygen present to propene present is from 1.6 to 2.2, and as the remainder up to 100% by volume of the total amount of molecular nitrogen.

In the case of methacrolein as the target compound, the starting reaction gas mixture may in particular have the composition described in DE-A 44 07 020.

The reaction temperature for the propene partial oxidation when the shaped catalyst bodies K, for example annular shaped catalyst bodies K, obtainable as described are used is frequently from 300 to 380° C. The same also applies in the case of methacrolein as the target compound.

The reaction pressure for the aforementioned partial oxidations is generally from 0.5 or 1.5 to 3 or 4 bar (what are always meant in this document, unless explicitly stated otherwise, are absolute pressures).

The total loading of the catalyst charge with starting reaction gas mixture in the aforementioned partial oxidations typically amounts to from 1000 to 10 000 l (STP)/l·h, usually to from 1500 to 5000 l (STP)/l·h and often to from 2000 to 4000 l (STP)/l·h.

The propene to be used in the starting reaction gas mixture is in particular polymer-grade propene and chemical-grade propene, as described, for example, in DE-A 102 32 748.

The oxygen source used is normally air.

In the simplest case, the partial oxidation employing the shaped catalyst bodies K, for example annular shaped catalyst bodies K, obtainable as described may be carried out, for example, in a one-zone multiple catalyst tube fixed bed reactor, as described by DE-A 44 31 957, EP-A 700 714 and EP-A 700 893.

Typically, the catalyst tubes in the aforementioned tube bundle reactors are manufactured from ferritic steel and typically have a wall thickness of from 1 to 3 mm. Their internal diameter is generally from 20 to 30 mm, frequently from 21 to 26 mm. A typical catalyst tube length is, for example, 3.20 m. It is appropriate from an application point of view for the number of catalyst tubes accommodated in the tube bundle vessel to be at least 1000, preferably at least 5000. Frequently, the number of catalyst tubes accommodated in the reaction vessel is from 15 000 to 35 000. Tube bundle reactors having a number of catalyst tubes above 40 000 are usually exceptional. Within the vessel, the catalyst tubes are normally arranged in homogeneous distribution, and the distribution is appropriately selected in such a way that the separation of the central internal axes from immediately adjacent catalyst tubes (known as the catalyst tube pitch) is from 35 to 45 mm (cf. EP-B 468 290).

However, the partial oxidation may also be carried out in a multizone (for example two-zone) multiple catalyst tube fixed bed reactor, as recommended by DE-A 199 10 506, DE-A 103 13 213, DE-A 103 13 208 and EP-A 1 106 598, especially at elevated loadings of the catalyst charge with the organic compound to be partially oxidized. A typical catalyst tube length in the case of a two-zone multiple catalyst tube fixed bed reactor is 3.50 m. Everything else is substantially as described for the one-zone multiple catalyst tube fixed bed reactor. Around the catalyst tubes, within which the catalyst charge is disposed, a heat exchange medium is conducted in each heating zone. Useful such media are, for example, melts of salts such as potassium nitrate, potassium nitrite, sodium nitrite and/or sodium nitrate, or of low-melting metals such as sodium, mercury and also alloys of different metals. The flow rate of the heat exchange medium within the particular heating zone is generally selected in such a way that the temperature of the heat exchange medium rises from the entry point into the temperature zone to the exit point from the temperature zone by from 0 to 15° C., frequently from 1 to 10° C., or from 2 to 8° C., or from 3 to 6° C.

The inlet temperature of the heat exchange medium which, viewed over the particular heating zone, may be conducted in cocurrent or in countercurrent to the reaction gas mixture is preferably selected as recommended in the documents EP-A 1 106 598, DE-A 199 48 523, DE-A 199 48 248, DE-A 103 13 209, EP-A 700 714, DE-A 103 13 208, DE-A 103 13 213, WO 00/53557, WO 00/53558, WO 01/36364, WO 00/53557 and also the other documents cited as prior art in this document. Within the heating zone, the heat exchange medium is preferably conducted in a meandering manner. In general, the multiple catalyst tube fixed bed reactor additionally has thermal tubes for determining the gas temperature in the catalyst bed. Appropriately, the internal diameter of the thermal tubes and the diameter of the internal accommodating sleeve for the thermal element are selected in such a way that the ratio of volume developing heat of reaction to surface area removing heat for the thermal tubes and working tubes is the same or only slightly different.

The pressure drop in the case of working tubes and thermal tubes, based on the same GHSV, should be the same. A pressure drop may be equalized in the case of the thermal tube by adding spalled catalyst to the shaped catalyst bodies. This equalization is appropriately effected homogeneously over the entire thermal tube length.

To prepare the catalyst charge in the catalyst tubes, as already mentioned, it is possible only to use shaped catalyst bodies K, for example annular shaped catalyst bodies K, obtainable as described or, for example also substantially homogeneous mixtures of shaped catalyst bodies K, for example annular shaped catalyst bodies K, obtainable as described and shaped bodies which have no active composition and behave substantially inertly with respect to the heterogeneously catalyzed partial gas phase oxidation. Useful materials for such inert shaped bodies include, for example, porous or nonporous aluminum oxides, silicon dioxide, zirconium dioxide, silicon carbide, silicates such as magnesium or aluminum silicate and/or steatite (for example of the C220 type from CeramTec, Germany).

The geometry of such inert shaped diluent bodies may in principle be as desired. In other words, they may be, for example, spheres, polygons, solid cylinders or else, as, for example, in the case of annular shaped catalyst bodies K, rings. Frequently, the inert shaped diluent bodies selected will be those whose geometry corresponds to that of the shaped catalyst bodies K to be diluted with them. However, along the catalyst charge, the geometry of the shaped catalyst body K may also be changed or shaped catalyst bodies K of different geometry may be used in a substantially homogeneous mixture. In a less preferred procedure, the active composition of the shaped catalyst body K may also be changed along the catalyst charge.

Quite generally, as already mentioned, the catalyst charge is advantageously configured in such a way that the volume-specific (i.e. normalized to the unit of the volume) activity either remains constant or increases (continuously, sharply or stepwise) in the flow direction of the reaction gas mixture.

A reduction in the volume-specific activity may be achieved in a simple manner, for example, by homogeneously diluting a basic amount of shaped catalyst bodies K, for example annular shaped catalyst bodies K, prepared uniformly in accordance with the invention with inert shaped diluent bodies. The higher the proportion of the shaped diluent bodies selected, the lower the active composition or catalyst activity present in a certain volume of the charge. However, a reduction can also be achieved by changing the geometry of the shaped catalyst bodies K obtainable in accordance with the invention in such a way that the amount of active composition present in the unit of the reaction tube internal volume becomes smaller.

For the heterogeneously catalyzed gas phase partial oxidations using annular shaped unsupported catalyst bodies K obtainable as described, the catalyst charge is preferably either configured uniformly with only one type of shaped unsupported catalyst body K over the entire length or structured as follows. At the reactor inlet is positioned, to a length of from 10 to 60%, preferably from 10 to 50%, more preferably from 20 to 40% and most preferably from 25 to 35% (i.e., for example, to a length of from 0.70 to 1.50 m, preferably from 0.90 to 1.20 m), in each case of the total length of the catalyst charge, a substantially homogeneous mixture of annular shaped unsupported catalyst body K obtainable in accordance with the invention and inert shaped diluent bodies (both preferably having substantially the same geometry), the proportion by weight of the shaped diluent bodies (the mass densities of shaped catalyst bodies K and of shaped diluent bodies generally differ only slightly) being normally from 5 to 40% by weight, or from 10 to 40% by weight, or from 20 to 40% by weight, or from 25 to 35% by weight. Downstream of this first charge section, there is then advantageously, up to the end of the length of the catalyst charge (i.e., for example, to a length of from 1.00 to 3.00 m or from 1.00 to 2.70 m, preferably from 1.40 to 3.00 m, or from 2.00 to 3.00 m), either a bed of the annular shaped unsupported catalyst body K obtainable as described which is diluted only to a lesser extent (than in the first section), or, most preferably, an unaccompanied (undiluted) bed of the same annular shaped unsupported catalyst body K which has also been used in the first section. Of course, a constant dilution may also be selected over the entire charge. Charging may also be effected in the first section using only an annular shaped unsupported catalyst body K obtainable in accordance with the invention and having lower active composition density based on its space demand, and, in the second section, using an annular shaped unsupported catalyst body K obtainable in accordance with the invention having higher active composition density based on its space demand (for example 6.5 mm×3 mm×4.5 mm [E×H×I] in the first section, and 5×2×2 mm in the second section).

Overall, in a partial oxidation for preparing acrolein or methacrolein carried out using the shaped catalyst bodies K, for example annular shaped catalyst bodies K, obtainable as described as catalysts, the catalyst charge, the starting reaction gas mixture, the loading and the reaction temperature are generally selected in such a way that, on single pass of the reaction gas mixture through the catalyst charge, a conversion of the organic compound to be partially oxidized (propene, isobutene, tert-butanol or its methyl ether) of at least 90 mol %, or at least 92 mol %, preferably of at least 94 mol %, results. The selectivity of acrolein or methacrolein formation will regularly be ≧80 mol %, or ≧85 mol %. Of course, very low hotspot temperatures are desired.

Finally, it is emphasized that annular shaped unsupported catalyst bodies K obtainable as described also have advantageous fracture behavior in the course of reactor charging.

A fresh catalyst charge (fixed catalyst bed) comprising geometric shaped catalyst bodies K obtainable in accordance with the invention can be started up, for example, as described in DE-A 103 37 788.

In general, activity and selectivity of target product formation initially increase with the operating time of the catalyst charge, before the aging-related reduction thereof sets in.

This forming can also be accelerated by performing it at essentially uniform conversion under increased loading of the catalyst charge with starting reaction gas mixture and, after substantial completing of forming, reducing the loading to its target value.

Otherwise, geometric shaped catalyst bodies K obtainable in accordance with the invention are quite generally suitable as catalysts with a low deactivation rate for gas phase catalytic partial oxidations of alkanes (especially propane), alkanols, alkanals, alkenes and alkenals comprising from 3 to 6 (i.e. 3, 4, 5 or 6) carbon atoms to, for example, olefinically unsaturated aldehydes and/or carboxylic acids, and the corresponding nitriles (ammoxidation, in particular of propene to acrylonitrile and of 2-methylpropene or tert-butanol (or the methyl ether thereof) to methacrylonitrile), and for gas phase catalytic oxidative dehydrogenations of the aforementioned organic compounds comprising 3, 4, 5 or 6 carbon atoms.

When a markedly exothermic heterogeneously catalyzed partial gas phase oxidation over a fixed catalyst bed present in a reaction tube is carried out, which is cooled externally with the aid of a fluid heat carrier (for example a salt melt), the temperature of the reaction gas mixture (and hence also of the fixed catalyst bed), in flow direction of the reaction gas, essentially irrespective of the specific flow regime of the heat carrier relative to the flow direction of the reaction gas mixture, frequently passes through a temperature maximum, which is referred to as the so-called hotspot temperature (cf., for example, Ullmann's Encyclopedia of Industrial Chemistry, VCH, 5. Ed., Volume B4, 1992, page 221).

In order to minimize the magnitude of the hotspot temperature, the activity structuring of the fixed catalyst bed present in the reaction tube is in many cases configured such that, in flow direction of the reaction gas mixture and based on the total length L of the fixed catalyst bed (pure inert material sections are not considered as belonging to the fixed catalyst bed in this consideration), the hotspot range is within about the first quarter of the total length L, and this length region (length section) of the fixed catalyst bed shall therefore be referred to in this document as “hotspot zone”, while all of the rest of the length region of the fixed catalyst bed in flow direction of the reaction gas mixture (¾·L) will be referred to as “main zone”. In general, the deactivation of geometric shaped catalyst bodies K present in the hotspot zone advances more rapidly than that of geometric shaped catalyst bodies K of the same kind which are present in the main zone. Instead of a tube bundle reactor, it is also possible to accommodate the fixed catalyst bed using a thermoplate reactor (cf., for example, DE-A 103 61 456).

In addition, the combination of already published research results (for example in the prior art already cited and in the following scientific publications: O. V. Udalova, D. P. Shashkin, M. D. Shibanova, O. V. Krylov, Kinetics and Catalysis 46 (2005), p. 535-544; D. P. Shashkin, O. V. Udalova, M. D. Shibanova, O. V. Krylov, Kinetics and Catalysis 46 (2005), p. 545-549; Y. Moro-Oka, W. Ueda, Adv. Catal. 40 (1994), p. 233-273; M. W. J. Wolfs, P. A. Batist, J. Catal. 32 (1974) p. 25-36; D.-H. He, W. Ueda, Y. Moro-Oka, Catal. Lett. 12 (1992), p. 35-44; Y. Haykawa, T. Tsunoda, H. Orita, T. Kameyama, H. Takahashi, K. Fukuda, K. Takehira, J. Chem. Soc. Chem. Commun (1987), p. 780-782; M. T. Le, W. J. M. v. Well, P. Stoltze, I. v. Driessche, S. Haste, Appl. Catal. A. Gen. 282 (2005), p. 189-194; W. J. M. v. Well, M. T. Le, N. C. Schiedt, S. Hoste, P. Stolzte, J. Mol. Catal. A. Chemical 256 (2006), p. 1-8 and J. M. M. Millet, G. Coudurier, J. M. Herrmann, J. C. Védrine, J. Catal. 142 (1993), p. 381-391) with the results of the examples and comparative examples performed in this document, and results of in-house experimental studies over and above the examples and comparative examples performed in this application, permits the following statements:

• 1. Studies of geometric shaped catalyst bodies K* by means of scanning electron microscopy (SEM) in combination with energy-dispersive X-ray analysis (EDX), after the thermal treatment of shaped bodies V which is employed to obtain them has ended, indicate that the particles of the finely divided starting material A1 preformed in each case apparently remain unchanged in the resulting shaped catalyst bodies K*.
• 2. Studies of the X-ray diffractograms of geometric shaped catalyst bodies K* and of comparative shaped bodies K*^V produced in an identical manner except without using finely divided starting material A1 indicate that both the component T in the active materials of the shaped catalyst bodies K* and the multielement oxide materials formed in the comparative shaped bodies K*^V consist essentially of the same crystalline phases, there being barely any difference either with regard to the type of these crystalline phases or with regard to their quantitative distribution (% by weight, based on the particular total weight). The particle size of the starting materials A1, A2 influences the above crystal phase formation to an insignificant degree at most.
• 3. In the case of use of geometric shaped catalyst bodies K produced in accordance with the invention and of geometric shaped catalyst bodies K* not produced in accordance with the invention, the preparation of which differed in particular by different values of the particle diameters 4, and dg with essentially identical stoichiometry of the particular component T and essentially identical stoichiometry of the particular starting material A1 and only slightly different stoichiometric coefficients a, as catalysts for a heterogeneously catalyzed partial gas phase oxidation of propene to acrolein, both exhibit, in long-term operation, apart from a barely perceptible loss of Mo, an essentially unchanged empirical stoichiometry of the particular multielement oxide active material I, but significantly different deactivation rates with comparable starting activity. More detailed studies indicate that the particular Mo losses suffered up to the same degree of deactivation differ significantly from one another, but correlate substantially with the operating time of the partial oxidation required in each case up to that point. The bleeding of Mo which has also been reported many times in the technical literature therefore cannot constitute the dominant mechanism of deactivation.
• 4. In the case of isolated production of the crystalline phases identified in the multielement oxide I active materials of unused geometric shaped catalyst bodies K and K* and isolated use of each as the active material for the heterogeneously catalyzed partial gas phase oxidation of propene to acrolein, the resulting activities and selectivities of acrolein formation are well below those of each of the geometric shaped catalyst bodies K and K*. The qualitatively identical result is established in the case of use of comparative shaped bodies K*^V (cf. “2”) as catalysts for the heterogeneously catalyzed partial gas phase oxidation of propene to acrolein.
• 5. In the long-term operation of a partial gas phase oxidation of propene to acrolein catalyzed by geometric shaped catalyst bodies K* and K, a new crystalline phase is formed over time in the multielement oxide I active materials of the shaped catalyst bodies, which is undetectable in the freshly produced, unused shaped catalyst bodies K* and K and which has the stoichiometry Bi₂Mo₃O₁₂.
• 6. The deactivation rate of a fixed catalyst bed present in long-term operation according to “5” correlates with the formation rate of crystalline Bi₂Mo₃O₁₂.
• 7. In the case of isolated preparation of crystalline Bi₂Mo₃O₁₂ and the use thereof as an active material for the heterogeneously catalyzed partial gas phase oxidation of propene to acrolein, the resulting activities are well below those of the geometric shaped catalyst bodies K and K*. The resulting selectivities of acrolein formation are, however, similar to those established in the case of use of geometric shaped catalyst bodies K and K*.

In summary, the above statements permit the following interpretation. The actual catalytically active material which activates the organic substance to be oxidized partially (e.g. the propene) forms a thin layer of an amorphous bismuth molybdate on the inner surface of the component T by virtue of migration processes in the course of production of geometric shaped catalyst bodies K and K*. The oxygen is supplied to this layer from the component T, which is capable of activating the molecular oxygen present in the gase phase. The Bi₁W_bO_x component functions as a Bi depot, and the T component as an Mo depot. On the basis of the migration processes which continue in the course of the partial oxidation, catalytically active amorphous bismuth molybdate is constantly reformed, but is itself simultaneously converted continuously to essentially inactive crystalline Bi₂Mo₃O₁₂. With an increasing total amount of crystalline Bi₂Mo₃O₁₂ already formed, the reformation of the catalytically active amorphous bismuth molybdate and/or the supply thereof with active oxygen is exhausted appreciably. This connection dominates the deactivation of geometric shaped catalyst bodies K* and K. The migration and phase conversion processes which underly the appearance of deactivation are directed both by the stoichiometric coefficient a (by the molar Bi:Mo ratio) and by the particle diameters d₅₀^A1 and d₉₀^A2. Relatively coarse particles in the starting material A2 attenuate the exhaustion process. Small particles in the starting material A2 promote the initial selectivity of target product formation. High values of a promote the formation of crystalline bismuth molybdate.

Appropriately in application terms, inventive shaped catalyst bodies K (and all working examples B1 to B8 and comparative examples V1 to V6) are produced on the industrial scale as described in German applications 102008040093.9 and 102008040094.7 (particularly advantageously according to example I.3. of application 102008040094.7).

The present application thus comprises especially the following inventive embodiments:

• 1. A process for producing geometric shaped catalyst bodies K which comprise, as an active material, a multielement oxide I of the general stoichiometry I

[Bi₁W_bO_x]^a[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_hO_y]₁  (I)

where

• • Z¹=one element or more than one element from the group consisting of nickel and cobalt,
  • Z²=one element or more than one element from the group consisting of the alkali metals, the alkaline earth metals and thallium,
  • Z³=one element or more than one element from the group consisting of phosphorus, arsenic, boron, antimony, tin, cerium, vanadium, chromium and bismuth,
  • Z⁴=one element or more than one element from the group consisting of silicon, aluminum, titanium, tungsten and zirconium,
  • Z⁵=one element or more than one element from the group consisting of copper, silver, gold, yttrium, lanthanum and the lanthanides,
  • a=0.1 to 3,
  • b=0.1 to 10,
  • c=1 to 10,
  • d=0.01 to 2,
  • e=0.01 to 5,
  • f=0 to 5,
  • g=0 to 10,
  • h=0 to 1, and
  • x, y=numbers determined by the valency and frequency of the elements in I other than oxygen,
  • in which
    • a finely divided mixed oxide Bi₁W_bO_x with a particle diameter d₅₀^A1 reported in the length unit μm, as starting material A1, is preformed with the proviso that 1 μm≦d₅₀^A1≦10 μm;
    • sources of the elements other than oxygen in the component T=[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_hO_y]₁ of the multielement oxide I are used in an aqueous medium to obtain an intimate aqueous mixture M, with the proviso that
      • each of the sources used, in the course of preparation of the aqueous mixture M, passes through a degree of division Q for which its diameter d₉₀^Q is ≦5 μm,
        • and
      • the aqueous mixture M comprises the elements Mo, Z¹, Z², Fe, Z³, Z⁴ and Z⁵ in the stoichiometry I*

Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_h  (I*);

• • • the aqueous mixture M, by means of drying and adjusting the degree of division d₉₀^A2, is used to obtain a finely divided starting material A2 with a particle diameter d₉₀^A2 reported in the length unit μm, with the proviso that 200 μm≧d₉₀^A2≧20 μm;
    • starting material A1 and starting material A2, or starting material A1, starting material A2 and finely divided shaping assistant, are mixed with one another to form a finely divided starting material A3, with the proviso that the starting material A3 comprises the elements other than oxygen introduced into the starting material A3 via starting materials A1 and A2 in the multielement oxide I in the stoichiometry I**

[Bi₁W_b]_a[Mo₁₂Z¹_cZ²_dFe_eZ³_fZ⁴_gZ⁵_h]₁  (I**)

• • • finely divided starting material A3 is used to form geometric shaped bodies V, and
    • the shaped bodies V are treated thermally at elevated temperature to obtain the geometric shaped catalyst bodies K,
  • wherein the value F of the product

(d₅₀^A1)^0.7·(d₉₀^A2)^1.5·(a⁻¹)

• • is ≧820.
• 2. The process according to embodiment 1, wherein F is ≧830.
• 3. The process according to embodiment 1, wherein F is ≧840.
• 4. The process according to embodiment 1, wherein F is ≧850.
• 5. The process according to embodiment 1, wherein F is ≧870.
• 6. The process according to embodiment 1, wherein F is ≧900.
• 7. The process according to embodiment 1, wherein F is ≧950.
• 8. The process according to embodiment 1, wherein F is ≧1000.
• 9. The process according to embodiment 1, wherein F is ≧1050.
• 10. The process according to embodiment 1, wherein F is ≧1100.
• 11. The process according to embodiment 1, wherein F is ≧1150.
• 12. The process according to any one of embodiments 1 to 11, wherein F is ≧2500.
• 13. The process according to any one of embodiments 1 to 11, wherein F is ≧2400.
• 14. The process according to any one of embodiments 1 to 11, wherein F is ≧2200.
• 15. The process according to any one of embodiments 1 to 11, wherein F is ≧2000.
• 16. The process according to any one of embodiments 1 to 11, wherein F is ≧1900.
• 17. The process according to any one of embodiments 1 to 11, wherein F is ≧1800.
• 18. The process according to any one of embodiments 1 to 11, wherein F is ≧1700.
• 19. The process according to any one of embodiments 1 to 11, wherein F is ≧1500.
• 20. The process according to any one of embodiments 1 to 19, wherein the stoichiometric coefficient a is from 0.2 to 2.
• 21. The process according to any one of embodiments 1 to 19, wherein the stoichiometric coefficient a is from 0.4 to 1.5.
• 22. The process according to any one of embodiments 1 to 19, wherein the stoichiometric coefficient a is from 0.5 to 1.
• 23. The process according to any one of embodiments 1 to 22, wherein 1.2 μm≦d₅₀^A1≦8 μm.
• 24. The process according to any one of embodiments 1 to 22, wherein 1.5 μm≦d₅₀^A1≦6 μm.
• 25. The process according to any one of embodiments 1 to 22, wherein 1.5 μm≦d₅₀^A1≦4 μm.
• 26. The process according to any one of embodiments 1 to 22, wherein 2 μm≦d₅₀^A1≦3 μm.
• 27. The process according to any one of embodiments 1 to 26, wherein 170 μm≧d₉₀^A2≧40 μm.
• 28. The process according to any one of embodiments 1 to 26, wherein 150 μm≧d₉₀^A2≧40 μm.
• 29. The process according to any one of embodiments 1 to 26, wherein 130 μm≧d₉₀^A2≧50 μm.
• 30. The process according to any one of embodiments 1 to 29, wherein d₉₀^Q for each of the sources used is ≦4 μm.
• 31. The process according to any one of embodiments 1 to 29, wherein d₉₀^Q for each of the sources used is ≦3 μm.
• 32. The process according to any one of embodiments 1 to 29, wherein d₉₀^Q for each of the sources used is ≦2 μm.
• 33. The process according to any one of embodiments 1 to 29, wherein d₉₀^Q for each of the sources used is ≦1 μm.
• 34. The process according to any one of embodiments 1 to 29, wherein d₉₀^Q for each of the sources used is ≦0.8 μm.
• 35. The process according to any one of embodiments 1 to 29, wherein d₉₀^Q for each of the sources used is ≦0.5 μm.
• 36. The process according to any one of embodiments 1 to 29, wherein d₉₀^Q for each of the sources used is ≦0.3 μm.
• 37. The process according to any one of embodiments 1 to 29, wherein d₉₀^Q for each of the sources used is ≦0.2 μm.
• 38. The process according to any one of embodiments 1 to 29, wherein each of the sources used, in the course of preparation of the aqueous mixture M, passes through the state of a colloidal solution or of a true solution.
• 39. The process according to any one of embodiments 1 to 29, wherein the component T comprises the element Si and each of the sources of the elements other than silicon used, in the course of preparation of the aqueous mixture M, passes through the state of a true solution and the source of the element Si used is a silica sol.
• 40. The process according to any one of embodiments 1 to 39, wherein the finely divided starting material A2 is obtained by spray-drying the aqueous mixture M.
• 41. The process according to any one of embodiments 1 to 40, wherein the aqueous mixture M comprises at least one auxiliary substance from the group consisting of NH₄OH, (NH₄)₂CO₃, NH₄HCO₃, NH₄NO₃, urea, NH₄CH₀₂, H₂CO₃, HNO₃, H₂SO₄, NH₄CH₃CO₂, NH₄Cl, HCl, NH₄HSO₄, (NH₄)₂SO₄, ammonium oxalate and the hydrates of the aforementioned compounds.
• 42. The process according to any one of embodiments 1 to 41, wherein Z¹=Co.
• 43. The process according to any one of embodiments 1 to 42, wherein Z²=K, Cs and/or Sr.
• 44. The process according to any one of embodiments 1 to 43, wherein Z⁴=Si.
• 45. The process according to any one of embodiments 1 to 44, wherein b is from 0.5 to 3.
• 46. The process according to any one of embodiments 1 to 44, wherein b is from 1 to 2.5.
• 47. The process according to any one of embodiments 1 to 46, wherein c is from 3 to 8.
• 48. The process according to any one of embodiments 1 to 46, wherein c is from 4 to 7.
• 49. The process according to any one of embodiments 1 to 48, wherein d is from 0.02 to 2.
• 50. The process according to any one of embodiments 1 to 48, wherein d is from 0.03 to 1.
• 51. The process according to any one of embodiments 1 to 48, wherein d is from 0.05 to 0.5.
• 52. The process according to any one of embodiments 1 to 51, wherein e is from 0.1 to 4.5.
• 53. The process according to any one of embodiments 1 to 51, wherein e is from 1 to 4.
• 54. The process according to any one of embodiments 1 to 53, wherein g is from 0.1 to 8.
• 55. The process according to any one of embodiments 1 to 53, wherein g is from 0.5 to 3.
• 56. The process according to any one of embodiments 1 to 55, wherein the finely divided mixed oxide Bi₁W_bO_x is the mixed oxide Bi₁W₂O_7.5.
• 57. The process according to any one of embodiments 1 to 56, wherein finely divided starting material A1, finely divided starting material A2 and shaping assistant comprising finely divided hydrophobized silica are mixed with one another to give the finely divided starting material A3.
• 58. The process according to any one of embodiments 1 to 57, wherein finely divided starting material A1, finely divided starting material A2 and shaping assistant comprising finely divided graphite are mixed with one another to give the finely divided starting material A3.
• 59. The process according to any one of embodiments 1 to 58, wherein geometric shaped bodies V are formed with finely divided starting material A3 by compacting the finely divided starting material A3.
• 60. The process according to embodiment 59, wherein the compaction is effected by extrusion or tableting.
• 61. The process according to any one of embodiments 1 to 60, wherein the geometric shaped body V is a ring.
• 62. The process according to embodiment 61, wherein the side crushing strength SCS of the annular shaped body V satisfies the condition 12 N≦SCS≦25N.
• 63. The process according to any one of embodiments 1 to 60, wherein the geometric shaped body V is a sphere.
• 64. The process according to any one of embodiments 1 to 60, wherein the geometric shaped body V is a solid cylinder.
• 65. The process according to embodiment 61 or 62, wherein the external diameter=from 2 to 10 mm, the height=from 2 to 10 mm and the wall thickness of the ring is from 1 to 3 mm.
• 66. The process according to embodiment 63, wherein the sphere diameter is from 2 to 10 mm.
• 67. The process according to embodiment 64, wherein the external diameter=from 1 to 10 mm and the height of the solid cylinder is from 2 to 10 mm.
• 68. The process according to any one of embodiments 1 to 58, wherein geometric shaped bodies V are formed with finely divided starting material A3 by applying the finely divided starting material A3 to the surface of a geometric shaped support body with the aid of a liquid binder.
• 69. The process according to any one of embodiments 1 to 68, wherein the temperature of 350° C. is exceeded and the temperature of 600° C. is not exceeded in the course of thermal treatment of the shaped bodies V.
• 70. The process according to any one of embodiments 1 to 68, wherein the temperature of 420° C. is exceeded and the temperature of 500° C. is not exceeded in the course of thermal treatment of the shaped bodies V.
• 71. The process according to any one of embodiments 1 to 70, wherein the thermal treatment is effected in the presence of air.
• 72. The process according to any one of embodiments 1 to 67 or according to any one of embodiments 69 to 71, wherein the shaped catalyst body K is a shaped unsupported catalyst body K and the process for preparing it is followed by a process for grinding to give finely divided material and the finely divided material is applied to the surface of a geometric shaped support body with the aid of a liquid binder.
• 73. The process according to any one of embodiments 1 to 72, wherein the particle diameter d₅₀^A2 of the finely divided starting material A2 satisfies the condition 10 μm≦d₅₀^A2≦50 μm.
• 74. The process according to any one of embodiments 1 to 72, wherein the particle diameter d₅₀^A2 of the finely divided starting material A2 satisfies the condition 20 μm≦d₅₀^A2≦40 μm.
• 75. The process according to any one of embodiments 1 to 74, wherein the value F* of the product

(d₅₀^A1)^0.7·(d₅₀^A2)^0.7·(a⁻¹)

satisfies the condition F*≧15, where the two particle diameters d₅₀ are reported in the length unit μm.

• 76. The process according to any one of embodiments 1 to 75, wherein the ratio of the particle diameter d₉₀^A2 of the finely divided starting material A2 to the particle diameter d₁₀^A2 of the finely divided starting material A2 is in the range from 5 to 20.
• 77. A shaped catalyst body obtainable by a process according to any one of embodiments 1 to 76.
• 78. A catalyst obtainable by grinding a shaped catalyst body which is a shaped unsupported catalyst body and is obtainable by a process according to any one of embodiments 1 to 67.
• 79. A process for heterogeneously catalyzed partial gas phase oxidation of an alkane, alkanol, alkanal, alkene and/or alkenal which comprises from 3 to 6 carbon atoms over a catalyst bed, wherein said catalyst bed comprises a shaped catalyst body according to embodiment 77 or a catalyst according to embodiment 78.
• 80. The process according to embodiment 79, which is a process for heterogeneously catalyzed partial gas phase oxidation of propene to acrolein.
• 81. The process according to embodiment 79, which is a process for heterogeneously catalyzed partial gas phase oxidation of isobutene to methacrolein.
• 82. The process according to embodiment 79, which is a process for ammoxidation of propene to acrylonitrile or a process for ammoxidation of isobutene to methacrylonitrile.
• 83. The use of a shaped catalyst body according to embodiment 77 or of a catalyst according to embodiment 78 as a catalyst in a process for heterogeneously catalyzed partial gas phase oxidation of an alkane, alkanol, alkanal, alkene and/or alkenal comprising from 3 to 6 carbon atoms.
• 84. The use according to embodiment 83 in a process for heterogeneously catalyzed partial gas phase oxidation of propene to acrolein, of isobutene to methacrolein, or in a process for ammoxidation of propene to acrylonitrile or of isobutene to methacrylonitrile.

EXAMPLES AND COMPARATIVE EXAMPLES

I) Production of inventive annular shaped unsupported catalyst bodies B1 to B8 and of annular comparative shaped unsupported catalyst bodies V1 to V6 with the following stoichiometry of the active material:

[Bi₁W₂O_7.5]_a[Mo₁₂Co_5.5Fe_3.0Si_1.6K_0.08O_x]₁.

• a) Production of finely divided starting materials A1-1 to A1-5 (Bi₁W₂O_7.5=½Bi₂W₂O₉·1WO₃)

In a 1.75 m³ stainless steel jacketed vessel (water for temperaure control flowed through the jacket space) with a cross-beam stirrer, 214.7 kg of tunstic acid at 25° C. (74.1% by weight of W, mean particle size (according to manufacturer determined to ASTM B330) from 0.4 to 0.8 μm, ignition loss (2 h at 750° C. under air) 6-8% by weight, bulk density 5-8 g/inch³, H. C. Starck, D-38615 Goslar) were stirred (70 rpm) in portions into 780 kg of an aqueous bismuth nitrate solution in nitric acid at 25° C. (11.2% by weight of Bi; free nitric acid: 3 to 5% by weight; prepared with nitric acid from bismuth metal from Sidech S.A., 1495 Tilly, Belgium, purity: >99.997% by weight of Bi₁<7 mg/kg of Pb, <5 mg/kg of each of Ni, Ag, Fe, <3 mg/kg of each of Cu, Sb, and <1 mg/kg of each of Cd, Zn) at 25° C. within 20 min. The resulting aqueous mixture was then stirred at 25° C. for another 3 h and then spray-dried. The spray-drying was effected in a Niro FS 15 rotary-disk spray tower in hot air cocurrent at a gas inlet temperature of 300±10° C., a gas outlet temperature of 100±10° C., a disk speed of 18 000 rpm, a throughput of 2001/h and an air rate of 1800 m³ (STP)/h. The resulting spray powder had an ignition loss of 12.8% by weight (calcine under air at 600° C. for 3 h in a porcelain crucible (which had been calcined to constant weight at 900° C.)) and had (at a dispersion pressure of 1.1 bar absolute) a d₅₀ of 28.0 μm (d₁₀=9.1 μm, d₉₀=55.2 μm). Table 1 below gives an overview of the representative d_x values of the spray powder as a function of the absolute dispersing pressure employed:

	TABLE 1
	2 bar	1.5 bar	1.2 bar	1.1 bar
	d10 (μm)	0.91	1.17	3.4	9.1
	d50 (μm)	5.8	8.5	19.7	28.0
	d90 (μm)	27.5	34.3	47.2	55.2

The resulting spray powder was subsequently converted to a paste with 16.7% by weight (based on the weight of the spray powder) of water at 25° C. in a discharging kneader for 30 min, and extruded by means of an extruder to extrudates of diameter 6 mm. These were cut into 6 cm sections, dried under air on a 3-zone belt dryer with a residence time of 120 min per zone at temperatures of 90-95° C. (zone 1), 115° C. (zone 2) and 125° C. (zone 3), and then treated thermally at a temperature in the region of 830° C. (calcined; in a rotary tube oven with air flow (reduced pressure 0.3 mbar, 200 m³ (STP)/h of air, 50 kg/h of extrudate, speed: 1 rpm)). What is important in the exact setting of the calcination temperature is that it has to be oriented to the desired phase composition of the calcination product, but, on the other hand, the calcined material has a BET surface area of 0.2 m²/g. The desired phases are WO₃ (monoclinic) and Bi₂W₂O₉ (orthorhombic); what is undesired here is the presence of γ-Bi₂WO₆ (russellite). Should the content of the γ-Bi₂WO₆ compound after the calcination be more than 5 intensity % (calculated as the ratio (int. r.) of the intensity of the reflection of γ-Bi₂WO₆ in the X-ray powder diffractogram at 2Θ=28.4° (CuKα radiation) to the intensity of the reflection of Bi₂W₂O₉ at 20Θ=30.0°, the preparation should therefore be repeated and the calcination temperature or the residence time at the same calcination temperature should be increased until the value attains or goes below the aforementioned limit. The preformed calcined mixed oxide thus obtained was ground with a 500 BQ biplex crossflow classifying mill from Hosokawa Alpine AG, Augsburg, at 2500 rpm with different throughputs, so as to give rise to the values specified in table 2 for the starting materials A1-1 to A1-5 as parameters of the particle size distribution (measured at a dispersion pressure of 2.0 bar absolute), as the BET surface area (BET) and as the γ-Bi₂WO₆ content (int. r.). FIG. 1 additionally shows, in detail, the particle size distributions of A1-1 to A1-4 measured at a dispersion pressure of 2.0 bar absolute (the abscissa shows the particle diameter d_X^A1 (μm) on a logarithmic scale; the ordinate shows the proportion by volume of the particular finely divided starting material A1 (A1-1 (▴); A1-2 (); A1-3 (♦) and A1-4 (▪)) that has the particular diameter or a smaller diameter (% by volume)). Table 3 shows, by way of example for the starting material A1-2, the influence of the dispersing pressure on the particle size distribution of a starting material A1. In the case of the starting material A1-1, the grinding in the crossflow classifying mill was followed by further grinding in a spiral jet mill; in the case of starting material A1-5, there was exclusively mild milling in an impingement plate mill. Before the further processing described under c), the different finely divided starting materials A1-1 to A1-5 were mixed in portions of 20 kg each in a tilted mixer with mixing and cutting blades (mixing blade speed: 60 rpm, cutting blade speed: 3000 rpm), were mixed homogeneously with 0.5% by weight (based on the particular finely divided starting material A1) of Sipernat® D17 finely divided hydrophobized SiO₂ from Degussa (tapped density: 150 g/l; d₅₀ of the SiO₂ particles (laser diffraction to ISO 13320-1)=10 μm, the specific surface area (nitrogen adsorption to ISO 5794-1, Annex D)=100 m²/g) within 5 min.

TABLE 2
Start-
ing
material		d10A1/	d50A1/	d90A1/	BET/	Int. r./
1	Milling	μm	μm	μm	m2 · g−1	Int.-%
A1-1	Additional	0.78	1.60	3.3	1.9	4
	spriral jet mill
A1-2	Classifying	0.92	2.10	4.7	1.2	4
	mill, low
	throughput
A1-3	Classifying	1.05	2.45	5.9	0.8	3
	mill, moderate
	throughput
A1-4	Classifying	1.10	3.0	13.9	0.5	4
	mill, high
	throughput
A1-5	Impingement	1.45	5.0	21	0.1	0
	plate mill

TABLE 3
Dispersion pressure/bar absolute	d10A1-2 /μm	d50A1-2/μm	d90A1-2/μm
1.1	0.70	1.25	3.0
2.0	0.92	2.10	4.7
4.5	1.20	4.5	62

• b) Production of finely divided starting materials A2-1 to A2-7 (Mo₁₂Co_5.5Fe_3.0Si_1.6K_0.08)

A solution A was prepared by metering 1.075 kg of an aqueous potassium hydroxide solution (47.5% by weight KOH) at a temperature of 60° C. and subsequently, via a differential metering balance at a metering rate of 600 kg/h, 237.1 kg of ammonium heptamolybdate tetrahydrate at a temperature of 25° C. (white crystals with a particle size d of <1 mm, 81.5% by weight of MoO₃, 7.0-8.5% by weight of NH₃, max. 150 mg/kg of alkali metals, H. C. Starck, D-38642 Goslar) into 6601 of water at a temperature of 60° C. in a water-heated 1.75 m³ stainless steel jacketed vessel with a crossbeam stirrer at 60° C. with stirring (70 rpm) within one minute, and the resulting slightly cloudy solution was stirred at 60° C. for 60 min (70 rpm).

A solution B was prepared by initially charging a water-heated 1.75 m³ stainless steel jacketed vessel with a crossbeam stirrer at 60° C. with 282.0 kg of an aqueous cobalt(II) nitrate solution at a temperature of 60° C. (12.5% by weight of Co, prepared with nitric acid from cobalt metal from MFT Metals & Ferro-Alloys Trading GmbH, D-41747 Viersen, purity >99.6% by weight, <0.3% by weight of Ni, <100 mg/kg of Fe, <50 mg/kg of Cu), and 142.0 kg of an iron(III) nitrate nonahydrate melt at 60° C. (13.8% by weight of Fe, <0.4% by weight of alkali metals, <0.01% by weight of chloride, <0.02% by weight of sulfate, Dr. Paul Lohmann GmbH, D-81857 Emmerthal) were metered into it with stirring (70 rpm). Subsequently, the mixture was stirred for a further 30 minutes while maintaining the 60° C. Then, while maintaining the 60° C., solution B was discharged into the initially charged solution A and stirred at 70 rpm at 60° C. for a further 15 minutes. Subsequently, 19.9 kg of a Ludox TM 50 silica sol from Grace at 25° C. (50.1% by weight of SiO₂, density: 1.29 g/ml, pH 8.5 to 9.5, alkali metal content max. 0.5% by weight) were added to the resulting aqueous mixture which was then stirred at 70 rpm at 60° C. for a further 15 minutes.

This was followed by spray-drying in a Niro FS-15 rotary spray tower in hot air countercurrent (gas inlet temperature: 350±10° C., gas outlet temperature: 140±5° C., throughput: 270 kg/h). By establishing the disk speeds and air rates specified in table 4, the spray powders (the finely divided starting materials) A2-1 to A2-6 were obtained with the particle size distributions (measured at a dispersion pressure of 2.0 bar absolute) and ignition losses IL (ignite at 600° C. under air for 3 h) likewise specified there. FIG. 2 additionally shows, in detail, the particle size distributions of selected starting materials A2 measured at a dispersion pressure of 2.0 bar absolute (the abscissa shows the particle diameter d_X^A2 (μm) on a logarithmic scale; the ordinate shows the proportion by volume of the particular finely divided starting material A2 (A2-1 (▴); A2-2 (♦) and A2-4 (▪)) that has the particular diameter or a smaller diameter (% by volume). For the starting material A2-6, table 5 also shows, by way of example, the influence of the dispersion pressure on the particle size distribution of a starting material A2. The starting material A2-7 was produced by repeated classification of the starting material A2-2 with the aim of obtaining an essentially homogeneous particle size of 30 μm.

TABLE 4
Start-
ing
mate-	Atomizer	Air rate/	IL/%
rial	disk	m3	by	d10A2/	d50A2/	d90A2/
2	speed/rpm	(STP) · h−1	wt.	μm	μm	μm
A2-1	18 000	2100	30.0	3.9	20.3	47.1
A2-2	15 000	2100	30.0	4.5	26.3	59.6
A2-3	14 000	2100	30.0	5.6	32.5	69.1
A2-4	12 500	2100	30.0	7.0	39.4	80.3
A2-5	18 000	2000	31.5	3.2	19.7	51.5
A2-6	18 000	2200	30.5	5.2	23.6	49.5
A2-7	multiple classification of A2-2	30.0	26.1	30.0	35.0

TABLE 5
Dispersion pressure/bar absolute	d10A2-6/μm	d50A2-6/μm	d90 A2-6/μm
1.1	9.9	28.5	56.3
2.0	5.2	23.6	49.5

• c) Production of the annular shaped unsupported catalyst bodies B1 to B8 and V1 to V6

The finely divided starting materials A1-1 to A1-5 comprising added Sipernat® D17 were mixed homogeneously with the finely divided starting materials A2-1 to A2-7 in the combinations specified in table 6 in amounts required for multimetal oxide active materials of the stoichiometries likewise specified there (the particular coefficient a is reported) (total amount: in each case 3 kg) in an Eirich intensive mixer (R02 type, capacity: 3-5 l, power: 1.9 kW, Maschinenfabrik Gustav Eirich GmbH & Co KG, D-74736 Hardheim) with bladed heads rotating counter to the plate (plate speed: 44 rpm, bladed head speed: 2500 rpm) within 5 min. Based on the aforementioned overall material, in each case 1% by weight of TIMREX® T44 graphite from Timcal AG (d₅₀=20.8 μm) was additionally mixed in homogeneously at a speed of approx. 30 rpm in a drum hoop mixer (wheel diameter: 650 mm, drum volume: 5 l) within 30 minutes. The resulting mixture was then compacted in a laboratory calender with two contrarotatory steel rollers at a pressure of 9 bar and forced through a screen with a mesh size of 0.8 mm. The compactate was then mixed in the above drum hoop mixer with, based on its weight, a further 2.5% by weight of the graphite specified at a speed of approx. 30 rpm within 30 minutes, and then, as described in German application 102008040093.9, compacted in a Kilian S100 rotary tableting press (9-die tableting machine) (from Kilian, D-50735 Cologne) under a nitrogen atmosphere to give annular shaped unsupported catalyst precursor bodies (shaped bodies V) of geometry 5×3×2 mm (A (external diameter)×H (height)×I (internal diameter)) with a side crushing strength SCS of 20±1 N and a mass of 125 mg (fill height: 7.5-9 mm, pressing force: 3.0-3.5 kN).

For the final thermal treatment, in each case 1000 g of the annular shaped unsupported catalyst precursor bodies V produced in each case, divided homogeneously between 4 grids arranged alongside one another with a square base area of in each case 150 mm×150 mm (bed height: approx. 15 mm), were heated in a forced-air oven (from Heraeus Instruments GmbH, D-63450 Hanau, type: K 750/2) through which 650 l (STP)/h of air heated to initially 140° C. flowed (instead of the airstream, it is also possible to employ a gas stream composed of 25% by volume of N₂ and 75% by volume of air, or of 50% by volume of N₂ and 50% by volume of air, or of 75% by volume of N₂ and 25% by volume of air) first from 25° C. to 185° C. within 120 min. This temperature was maintained for 1 h and then increased to 225° C. within 50 min. The 225° C. was maintained for 2 h, before the temperature was increased further to 270° C. within 23 min. This temperature was likewise maintained for 1 h, before it was increased with a heating ramp of 1° C./min to the end temperature specified in table 6 (T_{end calcination}). This end temperautre was maintained for 10 hours. This was followed by cooling to 25° C. within approx. 12 h. The end temperature was in all cases selected within the (463±6)° C. range, so as to result in essentially the same starting activity (T^S_210h=328±3° C.) in all cases in the performance of the stress test described under II. Table 6 additionally shows the important other production features of the annular shaped unsupported catalyst bodies B1 to B8 and V1 to V6 produced.

TABLE 6
Production features of the different annular shaped
unsupported catalyst bodies with the stoichiometry
[Bi1W2O7.5]a [Mo12Co5.5Fe3.0Si1.6K0.08Ox]1 of their
multielement oxide active material
	Powder	d50A1/	Powder	d90A2/			Tend calcination/
	A1	μm	A2	μm	a	F	° C.
V1	A1-1	1.60	A2-1	47.1	1.0	449	457
B1	A1-1	1.60	A2-4	80.3	1.0	1000	457
V2	A1-2	2.10	A2-1	47.1	1.0	543	459
B2	A1-2	2.10	A2-4	80.3	1.0	1210	459
V3	A1-3	2.45	A2-2	59.6	1.1	783	462
B3	A1-3	2.45	A2-2	59.6	1.0	862	462
B4	A1-3	2.45	A2-2	59.6	0.7	1231	465
B5	A1-3	2.45	A2-2	59.6	0.5	1723	468
B6	A1-4	3.0	A2-2	59.6	1.0	993	465
B7	A1-4	3.0	A2-3	69.1	1.0	1239	465
B8	A1-4	3.0	A2-4	80.3	1.0	1553	465
V4	A1-3	2.45	A2-5	51.5	0.96	721	464
V5	A1-3	2.45	A2-6	49.5	0.8	815	460
V6	A1-5	5.0	A2-7	35.0	1.0	639	465

II. Determination of the long-term stability of the annular shaped unsupported catalyst bodies B1 to B8 and V1 to V6 in a gas phase partial oxidation of propene to acrolein catalyzed by them

To determine the long-term stability of the annular shaped unsupported catalyst bodies B1 to B8 and V1 to V6 in a gas phase partial oxidation of propene to acrolein catalyzed by them, the annular shaped unsupported catalyst bodies B1 to B8 and V1 to V6 were subjected to the stress test described below.

A reaction tube (V2A steel; external diameter 21 mm, wall thickness 3 mm, internal diameter 15 mm, length 120 cm) was charged in each case as follows from the top downward in the flow direction of the reaction gas mixture:

• Section 1: length approx. 30 cm
  • 40 g of steatite spheres (C220 steatite from CeramTec) with a diameter of from 1.5 to 2.0 mm as a preliminary bed (heating zone).
• Section 2: length approx. 70 cm
  • A homogeneous mixture of 90 g of the particular annular shaped unsupported catalyst body and 10 g of steatite rings (C220 steatite from CeramTec) of the same ring geometry as the particular shaped unsupported catalyst body as a fixed catalyst bed.

The temperature of the reaction tube was in each case controlled by means of a molecular nitrogen-sparged salt bath having the salt bath temperature T^S (° C.) required in each case (53% by weight of potassium nitrate, 40% by weight of sodium nitrite and 7% by weight of sodium nitrate).

During the performance of the stress test, the reaction tube was charged continuously with a starting reaction gas mixture (charge gas mixture of air, polymer-grade propene and molecular nitrogen) of the following composition:

• • 5% by volume of propene (polymer-grade),
  • 9.5% by volume of molecular nitrogen, and
  • 85.5% by volume of N₂.

For the purpose of forming the fixed catalyst bed freshly introduced into the reaction tube, T^S was adjusted over the first 210 h of operating time at a volume flow of the starting reaction gas mixture conducted through the reaction tube (the inlet temperature of which into the reaction tube was approx. 30° C.) of 90 l (STP)/h such that the propene conversion C^P on single pass of the charge gas mixture through the reaction tube was continuously about 95 mol % (the starting selectivities of overall target product formation (acrolein+acrylic acid) which were established after an operating time of 90 h were (in mol % of the molar amount of propene converted in single pass through the reaction tube): 96 (V1); 94.5 (B1); 95.8 (V2); 94.5 (B2); 94.5 (V3); 94.6 (B3); 94.2 (B4); 93.8 (B5); 94.4 (B6); 94.5 (B7); 93.6 (B8); 94.9 (V4); 94.3 (V5) and 94.9 (V6)). The pressure on entry into the reaction tube was 1.2 bar absolute. The salt bath temperature T_210h^S (° C.) required to achieve the target conversion (95 mol %) at the end of the forming phase is a measure of the particular starting activity of the particular fixed catalyst bed. Table 7 indicates that the annular shaped unsupported catalyst bodies B1 to B8 and V1 to V6 have an essentially identical starting activity.

After the forming, the volume flow of starting reaction gas mixture conducted into the reaction tube was increased to 200 l (STP)/h and maintained for a period of 216 h. Within the aforementioned period, the salt bath temperature T^S, irrespective of the propene conversion which is established in this case, was kept at the constant value of 380° C. (the pressure on entry into the reaction tube was 1.6 bar absolute).

After the aforementioned stress phase had ended, the volume flow of starting reaction gas mixture conducted into the reaction tube was reduced again to the 90 l (STP)/h envisaged for actual normal operation of the propene partial oxidation. While maintaining this volume flow over 94 further operating hours, the salt bath temperature T^S was adjusted again such that the envisaged target conversion of propene (C^P) (based on single pass of the starting reaction gas mixture through the reaction tube) of about 95 mol % was established.

The salt bath temperature T_520h^S (° C.) required to attain the target conversion (95 mol %) at the end of the total of 520 operating hours which had then passed reflects the end activity of the particular fixed catalyst bed which is present in each case at the end of the stress test.

The difference T_520h^S (° C.)−T_210h^S (° C.)=ΔT^S determines the relative order within the annular shaped catalyst bodies B1 to B8 and V1 to V6 with regard to their long-term stability in the propene partial oxidation to acrolein (main product) and acrylic acid (by-product) catalyzed by them. The smaller ΔT^S is, the lower is the deactivation rate of the fixed catalyst bed and of the annular shaped unsupported catalyst bodies present therein in the course of operation of the partial oxidation. Table 7 indicates that the greater the inventive stability value F of the particular annular shaped unsupported catalyst body, the lower the deactivation rate of the particular annular shaped unsupported catalyst body with essentially uniform starting activity (T_210h^S (° C.)) and essentially uniform end selectivity S_520h of overall target product formation (acrolein+acrylic acid, in mol % of the molar amount of propene converted in single pass through the reaction tube).

	TABLE 7
	Shaped
	catalyst
	body	F	T210 hS/° C.	ΔTS/° C.	S520 h/mol %
	V1	449	326	14	96.7
	B1	1000	330	10	96.4
	V2	543	326	12	96.7
	B2	1210	328	8	97.0
	V3	783	327	11	96.8
	B3	862	326	9	96.7
	B4	1231	325	5	96.7
	B5	1723	325	2	96.8
	B6	993	328	7	96.9
	B7	1239	329	6	96.9
	B8	1553	331	4	96.8
	V4	721	328	11	96.8
	V5	815	326	11	96.7
	V6	639	325	12	96.6

FIG. 3 (the abscissa shows d₉₀^A2 (in μm) and the ordinate ΔT^S (in ° C.)) indicates, as an evaluation of the annular shaped unsupported catalyst bodies V1, B1, V2, B2, B3, B6, B7, B8 and V6, viewed in each case from left to right, for each use of the same finely divided starting material A1 (A1-1 (▴); A1-2 (◯); A1-3 (♦); A1-4 (▪) and A1-5 (X)) and with a constant coefficient a and the same chemical properties of the finely divided starting material A2, a coarsening of the finely divided starting material A2 (an increasing d₉₀^A2 can significantly reduce the deactivation rate (ΔT^S) with essentially homogeneous starting activity (T_210h^S).

Conversely, consideration of the same FIG. 3 viewed in each case from the top downward indicates that, for each use of the same finely divided starting material A2 and with a constant coefficient a and the same chemical properties of the finely divided starting material A1, a coarsening of the finely divided starting material A1 (an increasing d₅₀^A1) can likewise reduce the deactivation rate (ΔT^S) with essentially uniform starting activity (T_210h^S). In this context, the dependence of the deactivation rate on d₉₀^A2 is especially marked. Remarkably, the annular comparative unsupported catalyst V6 also fits seamlessly into this system, having been produced deliberately using a comparatively monodisperse finely divided starting material A2. The fact that the comparative unsupported catalyst V6 would not fit in a correspondingly seamless manner into a corresponding plot against d₅₀^A2 leads to the conclusion that the presence of a fraction of comparatively large particles in the finely divided starting material A2 is essential for the attenuation of the deactivation rate. In other words, in the case of an essentially homogeneous particle size of the finely divided starting material A2, this particle size should be selected so as to give rise to comparatively coarse particles for the inventive purpose.

FIG. 4 indicates, as an evaluation of the annular shaped unsupported catalyst precursor bodies V3, B3, B4 and B5 (the abscissa shows the stoichiometric coefficient a and the ordinate ΔT^S (in ° C.)), that, in the case of use of the same finely divided starting material A1 (A1-3) and the same finely divided starting material A2 (A2-2), the deactivation rate (ΔT^S) decreases significantly with decreasing stoichiometric coefficient a. This means that even the sole measure of reducing the molar n_Bi/n_Mo ratio (n=molar amount of a particular element indicated) in the multielement oxide which forms the active material for the relevant shaped unsupported catalyst body allows ΔT^S to be reduced from 11 to 2 at essentially the same starting activity (T_210h^S=(326±1)° C.). Maintenance of the end calcination temperature employed in the production of V3 in the production of B4 and B5 would further increase the ΔT^S difference.

Table 8 shows the extent within which, for the same chemical properties of the finely divided starting materials A1 and A2 but variable particle size thereof and variable stoichiometric coefficient a and otherwise the same mode of production as for the annular shaped unsupported catalyst bodies B1 to B8 and V1 to V6, inventive stability values F can be established.

TABLE 8
d50A1/μm	d50A2/μm	d90A2/μm	a	F
2	36.3	72.3	1	999
2	38.6	78	1	1119
2.2	36.3	72.3	1	1068
2.2	38.6	78	1	1196
2.2	27.2	59.2	0.9	879
2.2	29.1	60.2	0.9	901
2.2	36.3	72.3	0.9	1186
2.2	38.6	78	0.9	1329
2.2	27.2	59.2	0.8	989
2.2	29.1	60.2	0.8	1014
2.2	36.3	72.3	0.8	1334
2.2	38.6	78	0.8	1495
2.2	27.2	59.2	0.7	1130
2.2	29.1	60.2	0.7	1159
2.2	36.3	72.3	0.7	1525
2.2	38.6	78	0.7	1709
2.2	27.2	59.2	0.6	1318
2.2	29.1	60.2	0.6	1352
2.2	36.3	72.3	0.6	1779
2.2	38.6	78	0.6	1994
2.4	27.2	59.2	1	841
2.4	29.1	60.2	1	862
2.4	36.3	72.3	1	1135
2.4	38.6	78	1	1271
2.4	27.2	59.2	0.9	934
2.4	29.1	60.2	0.9	958
2.4	36.3	72.3	0.9	1261
2.4	38.6	78	0.9	1413
2.4	27.2	59.2	0.8	1051
2.4	29.1	60.2	0.8	1078
2.4	36.3	72.3	0.8	1418
2.4	38.6	78	0.8	1589
2.4	27.2	59.2	0.7	1201
2.4	29.1	60.2	0.7	1232
2.4	36.3	72.3	0.7	1621
2.4	38.6	78	0.7	1816
2.4	27.2	59.2	0.6	1401
2.4	29.1	60.2	0.6	1437
2.4	36.3	72.3	0.6	1891
2.4	38.6	78	0.6	2119
2.6	27.2	59.2	1	889
2.6	29.1	60.2	1	912
2.6	36.3	72.3	1	1200
2.6	38.6	78	1	1345
2.6	27.2	59.2	0.9	988
2.6	29.1	60.2	0.9	1013
2.6	36.3	72.3	0.9	1333
2.6	38.6	78	0.9	1494
2.6	27.2	59.2	0.8	1111
2.6	29.1	60.2	0.8	1140
2.6	36.3	72.3	0.8	1500
2.6	38.6	78	0.8	1681
2.6	27.2	59.2	0.7	1270
2.6	29.1	60.2	0.7	1303
2.6	36.3	72.3	0.7	1714
2.6	38.6	78	0.7	1921
2.6	27.2	59.2	0.6	1482
2.6	29.1	60.2	0.6	1520
2.6	36.3	72.3	0.6	2000
2.6	38.6	78	0.6	2241
2.8	27.2	59.2	1	936
2.8	29.1	60.2	1	960
2.8	36.3	72.3	1	1264
2.8	38.6	78	1	1416
2.8	27.2	59.2	0.9	1041
2.8	29.1	60.2	0.9	1067
2.8	36.3	72.3	0.9	1404
2.8	38.6	78	0.9	1574
2.8	27.2	59.2	0.8	1171
2.8	29.1	60.2	0.8	1200
2.8	36.3	72.3	0.8	1580
2.8	38.6	78	0.8	1770
2.8	27.2	59.2	0.7	1338
2.8	29.1	60.2	0.7	1372
2.8	36.3	72.3	0.7	1806
2.8	38.6	78	0.7	2023
2.8	27.2	59.2	0.6	1561
2.8	29.1	60.2	0.6	1600
2.8	36.3	72.3	0.6	2107
2.8	38.6	78	0.6	2360

FIG. 5 shows, in overview, a plot of the ΔT^S (in ° C. as the ordinate) determined in the described stress test of the annular shaped unsupported catalyst bodies B1 to B8 (♦) and V1 to V6 (⋄) as a function of the corresponding stability values F (as the abscissa).

III. Determination of the crystal phase composition of an annular shaped catalyst body K produced in accordance with the invention (F=1091) and of an annular shaped catalyst body K* not produced in accordance with the invention (F=803) in their deactivated state and in the unused state

The deactivation was effected in both cases in the course of use thereof as catalysts for the heterogeneously catalyzed partial gas phase oxidation of propylene to acrolein under comparable operating conditions in salt bath-cooled tubular reactors, as described in documents WO 2005/42459 and WO 2005/49200. The two annular shaped catalyst bodies were produced as described under I. In the case of the shaped catalyst body K, d₉₀^A1 was 2 μm, d₉₀^A2 was 72 μm and a was 0.91. In the case of the shaped catalyst body K*, d₅₀^A1; was 2 μm, d₉₀^A2 was 60 μm and a was 0.94.

The starting activity of the two shaped catalyst bodies determined as described in II. was within the range of T_210h^S=317±5° C.

FIG. 6 shows, by way of example, the X-ray powder diffractogram of the annular shaped catalyst body K in the unused state and FIG. 7 shows, by way of example, the X-ray powder diffractogram of the annular shaped catalyst body K in a deactivated state, having been deactivated in the course of the heterogeneously catalyzed partial gas phase oxidation of propene to acrolein (deinstalled sample from the hotspot zone). The abscissa shows the diffraction angle in the 2Θ scale (2 theta scale) and the ordinate shows the absolute intensity of the diffracted X-radiation.

Analysis of the X-ray diffractogram (cf., for example, Moderne Röntgenbeugung [Modern X-ray Diffraction], Spieβ-Schwarzer-Behnken-Teichert, Vieweg+Teubner (2005); R. A. Young, The Rietveld Methode, IUCr, Oxford University Press (1995); and H. Krischner, Röntgenstrukturanalyse and Rietveldmethode [X-Ray structural analysis and Rietveld method], Vieweg Lehrbuch, 5th edition, Brunswick (1994)) allows both the qualitative and the quantitative crystal phase composition to be determined. The basis of the determination of the qualitative crystal phase composition is the fact that the X-ray diffractogram of a mixture of crystal phases is a weighted sum of the X-ray diffractograms of the different individual crystal phases present therein. Since the element composition of the multielement oxide I active material is known, it is possible to limit those crystal phases which can form in principle in the presence of the aforementioned elements.

The X-ray powder diffractograms thereof are available to the public in corresponding data collections (e.g. The International Centre of Diffraction Data (ICDD) with its headquarters in 12 Campus Boulevard, Newton Square, Pa., USA, 19073-3273). Simple comparison of the positions of the reflections present in the X-ray diffractogram with those from the data collection thus allows the qualitative crystal phase composition (i.e. the type of the individual phases present) to be determined. The crystal structures (of the particular phases) which are additionally required to quantitatively determine the crystal phase composition can likewise be taken from corresponding databases (for example the Inorganic Crystal Structure Database (ICSD) of the Fachinformationszentrum (FIZ) D-76102 Karlsruhe).

The conventional method of quantitative phase composition was to find, in the X-ray diffractogram, well-separated reflections of the particular individual crystal phases present and to determine the relative proportions by weight of the individual phases from their relative intensities with reference to standards.

In more complicated cases with many phases, however, the separation of individual diffraction lines is often impossible. Here, the overlap of the powder diffractograms of those crystal phases which have been found in the qualitative phase analysis is therefore simulated according to Rietveld. This envelope is matched optimally to the X-ray powder diffractogram of the mixture measured by the “least squares” method, in which the parameters of the structure models underlying the individual powder diagrams are varied (multiphase Rietveld method).

Instead of the intensities of the individual lines of the particular crystal phases, the intensities of all lines of the powder diffractograms of the particular crystal phases are thus taken into account in the form of a phase-specific “scaling factor”.

According to a relationship set up by Hill and Howard (J. Appl. Cryst. 20 (1987) 467-474), the particular scaling factor is linked via the constants which are characteristic of the crystal structure (the construction of its unit cell) of the corresponding phase (these constants being deposited in the ICSD database) to the relative proportion by weight of the particular crystal phase in the phase mixture, which allows the relative proportions by weight to be determined from the scaling factors. The reason for the considerable advantage of the multiphase Rietveld method for analysis of the quantitative phase composition is that it does not require the additional use of material standards in the recording of the X-ray diffractogram, since the “standard” in this method consists in the known crystal structure of the individual phase.

In the present case, the X-ray diffractogram is analyzed with the aid of the DIFFRAC^plus TOPAS software package developed for this purpose by Bruker AXS GmbH, D-76187 Karlsruhe.

For the unused annular shaped catalyst bodies K, K*, the following crystal phase compositions were obtained (in % by weight based on the particular total weight):

	Crystal phase	K	K*
	β-(Co, FeII)MoO4	44	38
	Bi1W1O4.5	13	11
	WO3	8	7
	MoO3	9	7
	FeIII2(MoO4)3	20	29
	α-(Co, FeII)MoO4	7	8
	CoWO4	0	0
	Bi2Mo3O12	0	0

After operation of the particular heterogeneously catalyzed propene partial oxidation over a plurality of years (>2 years in both cases), a corresponding crystal phase analysis at a comparable degree of deactivation of the particular fixed catalyst bed gave the following results (the number “1” indicates that the analysis is based on a deinstalled sample from the hotspot zone of the fixed catalyst bed; the number “2” indicates that the analysis is based on a deinstalled sample from the main zone of the fixed catalyst bed), the figures reported again being % by weight based on the particular total weight:

	Crystal phase	K-1	K*-1	K-2	K*-2
	β-(Co, FeII)MoO4	46	46	47	38
	WO3	2	7	7	8
	α-(Co, FeII)MoO4	8	13	6	11
	MoO3	0	4	7	8
	Bi1W1O4.5	−0.7/J	−2.5/J	−0.6/J	−1.5/J
	FeIII2(MoO4)3	−2/J	−9/J	−0.7/J	−3.5/J
	CoWO4	+2.7/J	+3/J	+0.6/J	±0/J
	Bi2Mo3O12	+1.1/J	+3.5/J	+0.7/J	+2.5/J

In the case of the last 4 crystal phases, the average formation (+) and degradation (−) rates are reported in % by weight/year.

The significantly higher rate of formation of Bi₂Mo₃O₁₂ in the case of the annular shaped catalyst body K* correlates particularly well with its significantly higher rate of deactivation.

With regard to the quantitative Co and Fe^II contents in the particular γ-, β-MoO₄ phases and the variation in these proportions over the particular operating time, no precise statement is possible.

The formation of CoWO₄ and Bi₂Mo₃O₁₂ with degradation of Bi₁W₁O_4.5 indicates that migration processes have taken place between the starting materials of the annular unsupported catalysts. In addition to a variety of different other migration and phase conversion processes, the influence of which on the catalytic activity is comparatively limited, Bi₂Mo₃O₁₂ formation dominates the deactivation. Processes which should be emphasized among these other processes are the loss of MoO₃ via the gas phase, the reduction of the trivalent iron present in the Fe^III₂ (MoO₄)₃, associated with an enrichment of the iron in the β-(Co, Fe^II)MoO₄, and the formation of the Co-rich phases α-(Co, Fe^II)MoO₄ and CoWO₄, and the crystal growth in the β-(Co, Fe_II)MoO₄ phase which can be discerned from the change in the line width (approx. 40 to 60 nm in the case of the unused annular shaped catalyst bodies K, K* and approx. 70 to 110 nm in the case of the deactivated annular shaped catalyst bodies K, K*).

In addition to the above remarks, FIG. 8 shows, for various annular shaped catalyst bodies produced in a manner corresponding to the above annular shaped unsupported catalyst bodies K and K* (their starting activities determined for their unused state as described in II. were within the range of T_210h^S=317±5° C.; their stability values F were in the range of 800≦F≦1500), the proportion by weight of Bi₂Mo₃O₁₂ (plotted on the ordinate in % by weight (based on the total weight of the particular shaped unsupported catalyst body) and determined as described from the corresponding X-ray diffractogram) that they had in various states of deactivation present during a partial gas phase oxidation of propene to acrolein catalyzed by them (the shaped unsupported catalyst bodies studied in each case for this purpose were taken either from the main zone (▪) or from the hotspot zone (▴) and then in each case subjected to a determination of T_210h^S as described in II. for unused shaped unsupported catalyst bodies; the abscissa shows the T_210h^S in ° C. which reproduces the particular state of activity).

FIG. 8 demonstrates that the greater the proportion by weight of Bi₂Mo₃O₁₂ present, the greater the degree of deactivation (the T_210h^S).

U.S. Provisional Patent Application No. 61/096,537, filed Sep. 12, 2008, is incorporated into the present application by reference. With respect to the abovementioned teachings, numerous changes to and deviations from the present invention are possible. It can therefore be assumed that the invention may be implemented differently than specifically described herein within the scope of the appended claims.

Claims

1. (canceled)

2. A shaped catalyst body comprising, as an active material, a multielement oxide having stoichiometry (I): wherein the shaped catalyst body is produced by a process comprising:

[Bi1WbOx]a[Mo12Z1cZ2dFeeZ3fZ4gZ5hOy]1  (I)

wherein:

Z1 is nickel, cobalt, or a mixture thereof;

Z2 is an alkali metal, an alkaline earth metal, thallium, or any mixture thereof;

Z3 is zinc, phosphorus, arsenic, boron, antimony, tin, cerium, vanadium, chromium, bismuth, or any mixture thereof;

Z4 is silicon, aluminum, titanium, tungsten, zirconium, or any mixture thereof;

Z5 is copper, silver, gold, yttrium, lanthanum, a lanthanide, or any mixture thereof;

a is 0.1 to 3;

b is 0.1 to 10;

c is 1 to 10;

d is 0.01 to 2;

e is 0.01 to 5;

f is 0 to 5;

g is 0 to 10;

h is 0 to 1; and

x and y are numbers determined by the valency and frequency of the elements in stoichiometry (I) other than oxygen,

(I) preforming a finely divided mixed oxide Bi1WbOx having a particle diameter d50A1 reported in the length unit μm, as starting material A1, with the proviso that 1 μm≦d50A1≦10 μm;

(II) mixing sources of the elements other than oxygen in the component T=[Mo12Z1cZ2dFeeZ3fZ4gZ5hOy]1 of the multielement oxide in an aqueous medium, to obtain an intimate aqueous mixture M, with the proviso that each of the sources employed, in the course of preparation of the aqueous mixture M, passes through a degree of division Q for which its diameter d90Q is ≦5 μm, and the aqueous mixture M comprises the elements Mo, Z1, Z2, Fe, Z3, Z4, and Z5 in the stoichiometry (I*) Mo12Z1cZ2dFeeZ3fZ4gZ5h  (I*);

(III) drying and adjusting the degree of division d90A2 of the aqueous mixture M, to obtain a finely divided starting material A2 having a particle diameter d90A2 reported in the length unit μm, with the proviso that 200 μm≧d90A2≧20 μm;

(IV) mixing starting material A1 and starting material A2, or starting material A1, starting material A2 and finely divided shaping assistant, to form a finely divided starting material A3, with the proviso that the starting material A3 comprises the elements other than oxygen introduced into the starting material A3 via starting materials A1 and A2 in the multielement oxide I in the stoichiometry (I**): [Bi1Wb]a[Mo12Z1cZ2dFeeZ3fZ4gZ5h]1  (I**);

(V) forming a geometric shaped body V with the finely divided starting material A3; and

(VI) thermally treating the shaped body V at an elevated temperature, to obtain the shaped catalyst body,

wherein the value F of the product (d50A1)0.7·(d90A2)1.5·(a−1)

is ≧820.

3. A catalyst obtained by grinding the shaped catalyst body of claim 2.

4. A process for heterogeneously catalyzed partial gas phase oxidation of an alkane, alkanol, alkanal, alkene, alkenal, or any mixture thereof, which comprises from 3 to 6 carbon atoms over a catalyst bed, wherein the catalyst bed comprises the shaped catalyst body of claim 2.

5. The process of claim 4, which is a process for heterogeneously catalyzed partial gas phase oxidation of propene to acrolein.

6. The process of claim 4, which is a process for heterogeneously catalyzed partial gas phase oxidation of isobutene to methacrolein.

7. The process of claim 4, which is a process for ammoxidation of propene to acrylonitrile or a process for ammoxidation of isobutene to methacrylonitrile.