Method for producing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid on a fixed catalyst bed of a multi-tube reactor

Abstract

The invention relates to a method for producing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid. According to the method, the corresponding ethylenically unsaturated aldehyde and / or the ethylenically unsaturated carboxylic acid is obtained by heterogeneously catalyzed gas phase partial oxidation of at least one alkene with molecular oxygen at an elevated temperature on a fixed catalyst bed of a multi-tube reactor; at least two catalyst layers, each comprising a catalytically active multielement oxide, are arranged in the axial direction of the tubes of the multi-tube reactor in such a way as to obtain a multi-layer filling and a composition of the multielement oxide in one catalyst layer differs from a composition of the multielement oxide in at least one of the other catalyst layers; a stoichiometric ratio L1 of the elements Fe to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed, is greater than a stoichiometric ratio L2 of the elements Fe to Mo in the catalyst layer which is located furthest on the gas outlet side along the tube axis, and a stoichiometric ratio M1 of the elements Bi to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is less than or equal to a stoichiometric ratio M2 of the elements Bi to Mo in the catalyst layer which is located furthest on the gas outlet side along the tube axis.

Description

Process for the preparation of an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid on a fixed catalyst bed of a tube bundle reactor Description The present invention relates to a process for preparing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid on a fixed catalyst bed of a tube bundle reactor having at least two catalyst layers. Furthermore, the present invention relates to the tube bundle reactor used according to the invention and the catalyst bed of the tube bundle reactor. A catalyst bed containing at least two catalyst layers composed of different catalysts is commonly referred to as a structured catalyst bed. Such structured catalyst beds and their use for producing unsaturated aldehydes and / or unsaturated carboxylic acids are known from EP 3 023 406 A1, EP 3 321 247 A1, and EP 1 074 538 A1. The catalyst layers contain catalytically active multielement oxides of different compositions. Multielement oxides containing Mo, Bi, Fe, and Co are known, for example, from EP 3 740 310 A1 and EP 2 731 715 A1. A disadvantage of the production of ethylenically unsaturated aldehydes and / or ethylenically unsaturated carboxylic acids using the known structured catalyst beds in tube bundle reactors is the low yield of the desired product. The object of the present invention was therefore to provide an improved process for producing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid. In particular, the improved process should exhibit higher yields of the desired product with a lower catalyst mass. Accordingly, a process for the preparation of an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid is provided, in which the corresponding ethylenically unsaturated aldehyde and / or the ethylenically unsaturated carboxylic acid is obtained by heterogeneously catalyzed gas-phase partial oxidation of at least one aldehyde with molecular oxygen at elevated temperature on a fixed catalyst bed of a tube bundle reactor, wherein at least two catalyst layers, each comprising a catalytically active multielement oxide, are arranged in the axial direction of the tubes of the tube bundle reactor in such a way that a multilayer filling is achieved and a composition of the multielement oxide in one catalyst layer differs from a composition of the multielement oxide in at least one of the other catalyst layers, characterized in that a stoichiometric ratio Li of the elements Fe to Mo of the catalyst layer,in which the highest temperature occurs within the fixed catalyst bed is greater than a stoichiometric ratio l_2 of the elements Fe to Mo in the catalyst layer arranged furthest along the tube axis on the gas outlet side, and a stoichiometric ratio Mi of the elements Bi to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is less than or equal to a stoichiometric ratio M2 of the elements Bi to Mo in the catalyst layer arranged furthest along the tube axis on the gas outlet side. Preferably, the composition of the catalytically active multielement oxides corresponds to the general formula (I) Moi2Bi a FebCo c NidX e YfZgOn (I) with X = K, Cs and / or Rb, Y = Ca, Sr, Ba, Li, Na, Cr, W, Mn, Cu, Zn, Ga, P, B, As, Sn, Sb, Te, Nb, Ta, Pb, Ce and / or La, Z = Si, Al, Ti, Zr and / or Mg, where a = 0.2 to 2, b = 1 to 4, c = 3 to 9, d = 0 to 4, c + d = 4 to 9.5, e = 0.01 to 0.5, f = 0 to 10, g = 0 to 10, and n = a number determined by the valence and frequency of the elements other than oxygen in the general formula I. Preferably in the general formula (I) X = K, Z = Si, a = 0.4 to 1.5, preferably 0.5 to 1.0, more preferably 0.6 to 0.8, b = 1.3 to 3.7, preferably 1.5 to 3.4, more preferably 1.7 to 3.1, c = 3.8 to 8.5, preferably 4.3 to 8.0, more preferably 4.8 to 7.8, d = 0 to 3.5, preferably 0 to 3.0, more preferably 0 to 2.7, c + d = 5 to 9.0, preferably 6.0 to 8.7, more preferably 6.9 to 8.5, e = 0.02 to 0.3, preferably 0.025 to 0.2, more preferably 0.03 to 0.15, f = 0, g = 0 to 7, preferably 0 to 4, more preferably 0 to 1.8. Preferably, a stoichiometric ratio Ni of the elements K to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is greater than a stoichiometric ratio N2 of the elements K to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side. According to the invention, each tube of the tube bundle reactor has a gas inlet side at which the reaction gas mixture comprising the alkene and the molecular oxygen is introduced into the reaction tube and a gas outlet side at which the at least partially converted reaction gas mixture is discharged from the reaction tube. Preferably, a stoichiometric ratio of the elements Fe to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is from 0.15 to 0.35, preferably 0.18 to 0.32, more preferably 0.22 to 0.28. Preferably, the composition of the catalytically active multielement oxide of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed corresponds to the general formula (II) Moi2Bi a FebCo c NidXeYfZgO n (II). Preferably, in the general formula (II) X = K, Z = Si, a = 0.3 to 1.5, preferably 0.4 to 1.0, more preferably 0.5 to 0.7, b = 1.8 to 3.7, preferably 2.5 to 3.4, more preferably 2.9 to 3.1, c = 3.7 to 8.5, preferably 4.2 to 8.0, more preferably 4.8 to 7.6, d = 0 to 3.5, preferably 0 to 3.0, more preferably 0 to 2.7, c + d = 5 to 8.8, preferably 6.0 to 8.2, more preferably 6.9 to 7.6, e = 0.03 to 0.4, preferably 0.05 to 0.25, more preferably 0.07 to 0.15, f = 0, g = 0 to 7, preferably 0 to 4, more preferably 0 to 1.6. Also preferably, the composition of the catalytically active multielement oxide of the catalyst layer which is arranged furthest along the tube axis on the gas outlet side corresponds to the general formula (III) Moi2Bi a FebCo c NidXeYfZgO n (III). Preferably, in the general formula (III) X = K, Z = Si, a = 0.3 to 1.8, preferably 0.4 to 1.6, more preferably 0.5 to 1.5, b = 1.3 to 3.7, preferably 1.5 to 3.4, more preferably 1.7 to 3.1, c = 4 to 8.5, preferably 5 to 8.2, more preferably 6.0 to 7.8, d = 0 to 3.5, preferably 0 to 3.0, more preferably 0 to 2.7, c + d = 5.0 to 9.0, preferably 6.0 to 8.7, more preferably 6.9 to 8.5, e = 0.02 to 0.33, preferably 0.025 to 0.2, more preferably 0.03 to 0.09, f = 0, g = 0 to 7, preferably 0 to 4, more preferably 0 to 1.7. Preferably, the ethylenically unsaturated aldehyde is acrolein and the ethylenically unsaturated carboxylic acid is acrylic acid. Preferably, the alkene is propylene. Preferably, the catalyst layers are each formed from a solid catalyst body and / or a shell catalyst. Preferably, the ratio of Li to l_2 is from 1.1 to 1.6, preferably from 1.15 to 1.5, more preferably from 1.2 to 1.4. Preferably, the ratio of Mi to M2 is from 0.5 to 1, preferably 0.6 to 1, and more preferably 0.7 to 1. Preferably, the ratio of Ni to N2 is from 1.5 to 3.3, preferably 1.6 to 3.2, more preferably 1.7 to 3.1. Surprisingly, it was found that the structuring of the catalyst bed according to the invention using the claimed element ratios leads to a particularly high yield of valuable product with a lower amount of catalyst. A further subject of the invention is a tube bundle reactor for producing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid by heterogeneously catalyzed gas phase partial oxidation of at least one aldehyde with molecular oxygen at elevated temperature on a fixed catalyst bed of the tube bundle reactor, wherein in the axial direction of the tubes of the tube bundle reactor at least two catalyst layers, each comprising a catalytically active multielement oxide, are arranged such that a multilayer filling is present and a composition of the multielement oxide in one catalyst layer differs from a composition of the multielement oxide in at least one of the other catalyst layers, characterized in that a stoichiometric ratio Li of the elements Fe to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed,is greater than a stoichiometric ratio i-2 of the elements Fe to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side, and a stoichiometric ratio Mi of the elements Bi to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is less than or equal to a stoichiometric ratio M2 of the elements Bi to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side. Furthermore, the invention relates to a fixed catalyst bed, arranged within at least one tube of a tube bundle reactor, for producing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid by heterogeneously catalyzed gas-phase partial oxidation of at least one aldehyde with molecular oxygen at elevated temperature, wherein in the axial direction of the tubes of the tube bundle reactor at least two catalyst layers, each comprising a catalytically active multi-element oxide, are arranged such that a multi-layer filling is present and a composition of the multi-element oxide in one catalyst layer differs from a composition of the multi-element oxide in at least one of the other catalyst layers, characterized in that a stoichiometric ratio Li of the elements Fe to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed,is greater than a stoichiometric ratio L2 of the elements Fe to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side, and a stoichiometric ratio Mi of the elements Bi to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is less than or equal to a stoichiometric ratio M2 of the elements Bi to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side. The production of the full catalyst bodies is described below: The unsupported catalyst bodies obtainable according to the invention are typically formed into geometric shapes, calcined, and used to catalyze the respective heterogeneously catalyzed gas-phase partial oxidation (in particular that of propene to acrolein). In principle, the desired geometry of the unsupported catalysts is not subject to any restrictions. In principle, shaped catalyst bodies can be produced in a simple manner by mixing suitable sources of their elemental constituents (especially those other than oxygen) into a A preferably finely divided dry mixture is produced, the composition of which corresponds to the respective stoichiometry of the shaped unsupported catalyst bodies to be produced, and this, after prior shaping into precursor bodies, which optionally takes place with the use of shaping aids, is calcined at temperatures of 350 to 650°C. The calcination can take place either under an inert gas or under an oxidative atmosphere such as air (or another mixture of inert gas and molecular oxygen, which may also contain reducing components in comparatively smaller proportions), or under a reducing atmosphere (for example a mixture of inert gas, NH3, CO and / or H2, which may also contain oxidizing components in comparatively smaller proportions) or under vacuum. The calcination time can be from a few minutes to a few days and is usually shorter at elevated calcination temperatures. Suitable sources for the elemental constituents of the shaped unsupported catalyst bodies (i.e., starting compounds containing at least one elemental constituent (at least one element contained in the shaped unsupported catalyst body) in a chemically bonded manner) are compounds which are already oxides (which are generally in the solid state under standard conditions (1 atm-abs, 0°C)) (for example, metal oxides) and / or compounds which can be converted into oxides (which are generally in the solid state under standard conditions) by heating (thermal treatment at elevated temperature), at least in the presence of gaseous oxygen and / or gaseous (for example, molecular) oxygen-releasing components. In principle, the oxygen source can be a component of the mixture to be calcined, for example in the form of a peroxide.In general, a starting compound can be the source of several elemental constituents of the full catalyst bodies. In addition to the oxides, such starting compounds (sources) are mainly halides, nitrates, formates, acetates, oxalates, citrates, carbonates, ammine complexes, ammonium salts and / or hydroxides as well as hydrates of the aforementioned salts. Compounds such as NH4OH, (NH^COs, NH4NO3, NH4CHO2, CH3COOH, NH4CH3CO2 and / or ammonium oxalate, which decompose and / or decompose essentially completely into gaseous compounds (e.g. ammonia, CO2, CO, H2O, nitrogen oxides) at the latest during later calcination, can be additionally incorporated into the intimate dry mixture. Organic materials such as stearic acid, malonic acid, ammonium salts of the aforementioned acids, starches (e.g. potato starch, corn starch), ground nutshells and finely divided plastic flour (e.g. polyethylene, polypropylene, etc.) can also be considered as substances that decompose during calcination. A formation (release) of gaseous compounds during the thermal treatment outlined is normally also present when the element sources used to produce the most intimate and preferably finely divided dry mixture are partly of an organic nature (for example in the case of acetates, formates, oxalates and / or citrates) or contain hydroxide ions, carbonate ions, hydrogen carbonate ions, ammonium ions, halide ions, hydrogen phosphate ions and / or nitrate ions, which normally decompose during calcination. The intimate mixing of the starting compounds (sources) for the production of unsupported catalyst bodies can be carried out in dry or wet form. If the mixing is carried out in dry form, the starting compounds (sources) are preferably used as finely divided powders and, after mixing and compaction into the geometric precursor body, are subjected to calcination. According to the invention, however, the intimate mixing of the (element) sources preferably takes place in wet form. The starting compounds are mixed together in the form of solutions and / or suspensions, and the resulting wet (preferably aqueous) mixture M is then dried to form an intimate dry mixture. Water or an aqueous solution is preferably used as the solvent and / or suspending agent, with the resulting wet mixture M being an aqueous mixture M. Particularly intimate dry mixtures are obtained in the mixing process described above when only dissolved sources and / or colloidally dissolved sources of the elemental constituents are used. As already mentioned, a starting compound can be a source for only one or for several elemental constituents. Accordingly, a solution or colloidal solution listed above can contain only one or several elemental constituents of the relevant unsupported catalyst body to be produced in solution. The preferred solvent is, as already mentioned, water. The resulting aqueous mixtures are preferably dried by spray drying. When this document refers to a solution of a source (starting compound, starting substance) in a solvent (especially water), the term "solution" is meant in the sense of a molecular or ionic solution. This means that the largest geometric unit of the dissolved starting substance (source) in the solution necessarily has "molecular" dimensions, and the solution appears "optically empty." In contrast, colloidal solutions represent a bridge between true (molecular and / or ionic) solutions and suspensions. These colloidally dispersed systems contain smaller clusters of molecules or atoms, which, however, are neither visible to the naked eye nor to a microscope. The colloidal solution appears optically completely clear (although often colored), since the particles it contains have a diameter of only 1 to 250 nm (preferably up to 150 nm and especially preferably up to 100 nm). Due to their small size, separation of the colloidally dissolved particles by conventional filtration is not possible. However, they can be separated from their "solvent" by ultrafiltration using membranes of plant, animal, or artificial origin (e.g., parchment, pig bladder, or cellophane). In contrast to "optically empty" true (molecular and / or ionic) solutions, a light beam cannot pass through a colloidal solution without deflection. The light beam is scattered and deflected by the colloidally dissolved particles. To keep colloidal solutions stable and prevent further particle agglomeration, they often contain wetting and dispersing agents, as well as other additives. While elements other than silicon (elemental constituents) of a shaped unsupported catalyst body are preferably introduced from sources present in the form of a solution (particularly preferably dissolved in an aqueous solution) to prepare the wet (preferably aqueous) mixture M, the element silicon is preferably introduced in the form of a silica sol to prepare the wet (preferably aqueous) mixture M. Silica sols are aqueous colloidal solutions of nearly spherical polysilicic acid particles. The diameter of the particles is in the colloidal range and, depending on the type, ranges from 5 to 75 nm. The particles are pore-free. They have a core of SiO2, which is hydroxylated on its surface. The spherical individual particles are not cross-linked. For reasons of stability, a portion of the hydroxyl groups in silica sols is often neutralized with alkali hydroxide and / or ammonium hydroxide. This means that the counterions are sometimes not protons, but alkali ions (e.g., Na + ) and / or NH cations. The SiO2 content of silica sols suitable for producing a wet (preferably aqueous) mixture M can, for example, be 30% to 60% of the weight of the silica sol. Silica sols are normally water-soluble and contain no sedimentable components. They can often be stored for years without sedimentation. Particularly suitable silicon sources are the LUDOX® silica sols from Grace GmbH & KG, In der Hollerecke 1, D-67545 Worms. Their particles are discrete, uniform spheres of silicon dioxide without an internal surface or detectable crystallinity. Their majority is dispersed in an alkaline medium, which reacts with the hydroxylated surface to produce repulsive negative charges. The particle diameters of suitable silica sols can be very narrow (essentially monodisperse) or very wide (polydisperse). A particularly suitable silica sol for the purposes of the invention (for producing a wet (preferably aqueous) mixture M) is the LUDOX TM50 silica sol from Grace. The LUDOX TM50 silica sol has a largely monodisperse (d = 22 nm) particle diameter distribution. Its pH (1 atm-abs, 25°C) is 9.0. The alkali ion that replaces a portion of the hydroxyl protons is Na +The SiO2 content of LUDOX TM50 is 50% of the weight of this hydrogel. The specific surface area of ​​the colloidally dissolved SiO2 particles in LUDOX TM50 is 140 m 2 / g. The mass density (1 atm-abs, 25°C) of LUDOX TM50 is 1.40 g / cm 3 . The titratable alkali content of LUDOX TM50 (calculated as Na2O) is 0.21 wt% (based on the weight of the silica sol). The dynamic viscosity of LUDOX TM50 is 40 mPas (1 atm-abs, 25°C). The Cl' content (calculated as NaCl) of LUDOX TM50 is 0.03 wt.%, and the SO^ content of LUDOX TM50 (calculated as Na2SO4) is 0.08 wt.% (each based on the weight of LUDOX TM50). Of course, in a solution to be used to produce a wet (in particular aqueous) mixture M, at least one element source can be molecularly and / or ionically dissolved and one or more than one other element source can be colloidally dissolved side by side. A favorable source of Mo is ammonium heptamolybdate tetrahydrate. This is primarily due to its excellent solubility in water. According to Ullmann's Encyclopedia of Industrial Chemistry, Volume 22, 2003, WILEY-VCH, pages 320 / 321, a solution of ammonium molybdate tetrahydrate in water at 25°C and 1 atm-abs has a saturation solubility of 30 wt% (calculated as an anhydrous salt). Due to manufacturing processes, ammonium molybdate tetrahydrate may be contaminated in small amounts (usually ppm) with water-insoluble isopolymolybdate (due to process parameters not being strictly adhered to during its production). Dissolving contaminated ammonium heptamolybdate tetrahydrate in water produces an aqueous solution that exhibits a certain turbidity due to the small amounts of finely divided isopolymolybdate present in undissolved form.The applicant's own investigations have shown that even ammonium molybdate tetrahydrate contaminated with isopolymolybdate, the solution of which in water (determined as described in WO 2016 / 147324) has a turbidity of 20 NTU, or of 50 NTU, or of 70 NTU, or of 100 NTU, or of 150 NTU, or of 200 NTU, or of 250 NTU, or of 300 NTU, is suitable for producing unsupported catalyst bodies according to the invention without noticeably impairing their performance when used as catalysts for the heterogeneously catalyzed partial gas phase oxidation of propene to acrolein as the main product and acrolein as the by-product. Other suitable Mo sources are, for example, ammonium orthomolybdate ((NH^MoO^), ammonium dimolybdate ((NH4)2MO2O2), ammonium tetramolybdate dihydrate ((NH4)2Mo4Oi3 x 2 H2O) and ammonium decamolybdate dihydrate ((NH4)4MOIQO32 x 2 H2O). In principle, molybdenum trioxide can also be used. The preferred source of alkali metals in the production of shaped unsupported catalyst bodies is their hydroxides. However, the nitrates of these elements and the hydrates of these nitrates can also be considered as such sources. This means that the preferred potassium source is KOH, but KNO3 or its hydrate can also be used as a potassium source. Bismuth salts are preferably used as a source of Bi, which contains Bi as Bi 3+ Examples of such salts include bismuth(III) oxide, bismuth(III) oxide nitrate (bismuth subnitrate), bismuth(III) halide (e.g. fluoride, chloride, bromide, iodide) and especially bismuth(III) nitrate pentahydrate. Of course, a solution of elemental Bi in aqueous nitric acid can also be used as a Bi source, in which the Bi is present as Bi 3+ In the case of the (inventively preferred) use of an aqueous solution of Bi 3+- Nitrate or its hydrate as a source (such a solution can also be produced by dissolving elemental Bi in aqueous nitric acid), it is advantageous according to the invention if its pH value (1 atm-abs, 25°C) is high, since this prevents undesirable formation of Bi 3+ containing precipitates in the aqueous solution. This pH is preferably < 1, particularly preferably < 0.5. However, this pH is generally > -2, mostly > 0. Such an aqueous solution is expediently nitric acid. This means that its low pH is caused by excess nitric acid (in this case, the molar ratio (HNOS-) / (riBi3+) depends on the molar amount of NOs' (n N o3-) to the molar amount of bis-2-(2-methyl-2-methyl-4-(2-methyl-4 ... + (n B i3+) > 3). Preferred Fe sources are salts of Fe 3+, among which the various iron(II) nitrate hydrates are particularly preferred (see, for example, DE 10 2007 003076 A1). According to the invention, iron(II) nitrate nonahydrate is particularly preferably used as the Fe source for the aforementioned purpose. Of course, salts of Fe 2+ be used as a Fe source. Advantageously, for the production of the unsupported catalyst bodies, based on the total molar amount of Fe contained therein, at least 50 mol%, better at least 75 mol% and preferably at least 95 mol% or 100 mol% is introduced in the form of an Fe source which contains the Fe as Fe 3+ Fe sources can also be used for this purpose, which contain both Fe 3+ as well as Fe 3+ have. Particularly suitable Co sources are its salts, which contain the Co as Co 2+ and / or Co 3+Examples of such sources include cobalt(II) nitrate hexahydrate, CO3O4, CoO, cobalt(II) formate, and cobalt(II) nitrate. The former of these sources is particularly preferred for the aforementioned purpose. Of course, a solution of elemental Co in aqueous nitric acid can also be used as a Co source, in which the Co is present as Co 2+ is present. In the case of the elemental constituent Ni, Ni 2+Salts are used. Particularly suitable are nickel (II) carbonate, nickel (II) sulfate, nickel (II) oxide, nickel (II) acetate, nickel (II) formate, nickel (II) hydroxide, nickel (II) oxalate and nickel (II) nitrate and the respective hydrates of these salts. Hydrates of nickel (II) nitrate (for example its hexahydrate) are particularly preferred as Ni sources. Of course, a solution of elemental Ni in aqueous nitric acid can also be used as Ni source, in which the Ni is present as Ni 2+ Such solutions regularly also contain a certain amount of ammonium nitrate. To improve the solubility of, for example, salts of Fe, Co and / or Ni in an aqueous medium, ammonia (also as its aqueous solution) and / or nitric acid (in particular as its aqueous solution) can be added to the respective solution as required. In principle, the preparation of a wet (e.g. aqueous) mixture M can be carried out in a wide variety of gas atmospheres (e.g. air, argon, nitrogen, water vapor and / or carbon dioxide). According to the invention, the preparation of a wet (e.g. aqueous) mixture M is preferably carried out in air (advantageously, the aqueous mixture M is saturated with air). This applies in particular when salts of Co 2+ and salts of Fe 3+ used. Especially when these salts are nitrates and / or their hydrates. As already mentioned, the wet mixture M according to the invention is preferably an aqueous mixture M, which is particularly advantageously prepared in the following manner. An aqueous solution A is prepared from at least one source of the element Fe, from at least one source of the element Bi, from at least one source of the element Co and optionally at least one source of the element Ni, the pH of which is < 3, preferably < 2, particularly preferably < 1 and very particularly preferably < 0 (pH values ​​of aqueous solutions in this document generally refer (unless explicitly stated otherwise) to a measurement with a glass electrode designed as a combination measuring chain at 1 atm-abs and at the temperature at which the respective aqueous solution is prepared; the calibration of the combination measuring chain required in this regard is carried out under the same conditions using aqueous buffer solutions whose pH is known under these conditions and is close to the desired measured value);The Mettler Toledo pH electrode Inpro 4260 / 425 / Pt 100, which is a combination electrode with an integrated Pt 100 temperature sensor for automatic temperature compensation, is particularly suitable for determining such pH values. As a rule, the pH of the aqueous solution A is not less than -2 and is particularly advantageously in the range -1 to 0. Preferably, the aqueous solution A is an aqueous solution of the nitrates or nitrate hydrates of the aforementioned elements. Particularly preferably, the aqueous solution A is an aqueous solution of these nitrates or nitrate hydrates in aqueous nitric acid. Solutions of the relevant elements in aqueous nitric acid are also suitable as element sources, particularly for preparing such a solution. An aqueous solution B is prepared from at least one source of the element Mo and optionally at least one of the sources of an alkali metal. The pH of the aqueous solution B is advantageously (at 1 atm-abs and the temperature at which the solution B is prepared) < 7. The pH of the aqueous solution B is particularly preferably < 6.5 and very particularly advantageously < 6. As a rule, the pH of the aqueous solution B will be > 3. Favorable solutions B to be used according to the invention have a pH of 4 to 6. According to the invention, the aqueous solution B is preferably used as a source of an alkali metal for preparing the aqueous solution B. Element hydroxide is used (e.g., KOH). The preferred Mo source for preparing an aqueous solution B is ammonium heptamolybdate tetrahydrate ((NH4)eMozO24 x 4 H2O), which is completely soluble in water at 25°C (1 atm-abs) up to saturation solubility (30 wt.%, calculated anhydrous). According to the invention, the total content of the metal constituents Bi, Fe, Co, Ni, etc. in the aqueous solution A is expediently 5 to 20 wt.%, advantageously 10 to 15 wt.%, based on the total amount of the aqueous solution A. According to the invention, the total Mo content of the aqueous solution B is expediently, based on the total amount of the aqueous solution B, 2 to 25 wt.%, advantageously 3 to 20 wt.% and particularly advantageously 5 to 15 wt.%. Subsequently, the aqueous solution A and the aqueous solution B are suitably mixed together. The procedure is advantageously such that the aqueous solution A is continuously stirred into the aqueous solution B. The aqueous solution B is advantageously stirred intensively. According to the invention, the total content of the metal constituents Bi, Fe, Co, Ni, Mo, etc. in the resulting aqueous mixture of aqueous solution A and aqueous solution B is expediently 3 to 20 wt.%, advantageously 5 to 15 wt.%, based on the total amount of the aqueous mixture. The temperature of the initially introduced aqueous solution B and the temperature during stirring in of the aqueous solution A, as well as the temperature of the aqueous solution A itself, is advantageously (preferably throughout the entire mixing process) < 80°C and > 0°C. Preferably, the aforementioned temperatures are < 75°C and > 30°C, and particularly preferably they are < 70°C and > 50°C, or < 65°C and > 55°C. Advantageously, aqueous solutions A and B, as well as the aqueous mixture resulting from stirring aqueous solution A into aqueous solution B, have the same temperature. This is ideally 60°C. The temperatures of aqueous solution A, aqueous solution B, and the resulting aqueous mixtures are preferably constant throughout the described stirring process. For this purpose, thermostatting can be achieved, for example, using a water bath. The working pressure when stirring aqueous solution A into aqueous solution B is advantageously 1 atm-abs (1.01 bar-abs). Preferably, the aqueous solution A is stirred into the initially introduced aqueous solution B within a period of 5 to 60 minutes, particularly preferably within a period of 10 to 30 minutes, and very particularly preferably within a period of 15 to 25 minutes. The resulting aqueous mixture is subsequently stirred, preferably while maintaining the stirring temperature, for 5 to 60 minutes, preferably 10 to 30 minutes, and particularly advantageously 15 to 25 minutes. The pH value of the aqueous mixture of aqueous solution A and aqueous solution B is preferably < 3, better < 2. As a rule, it is > 0. If the unsupported catalyst body contains the elemental constituent Si, then according to the invention, aqueous silica sol is preferably stirred into the aqueous mixture of aqueous solution A and aqueous solution B as a source thereof, it being advantageous to add water to this aqueous mixture prior to this stirring in. Advantageously, both the aqueous silica sol and the water can be added all at once. Both the temperature of the water and the temperature of the aqueous silica sol advantageously correspond to the temperature of the aqueous mixture of aqueous solution A and aqueous solution B. Finally, stirring is advantageously continued for up to 30 minutes. During the stirring, the aforementioned temperature is advantageously maintained. The SiO2 content of the added aqueous silica sol can be 15 to 60% by weight, or 20 to 60% by weight, or 30 to 60% by weight, preferably 40 to 60% by weight.-% and particularly preferably 45 to 55 wt.% (in each case based on its total weight). Instead of placing the aqueous solution B in a thermostatically controlled stirred vessel and then adding the aqueous solution A while stirring, both the aqueous solution B and the aqueous solution A can be continuously fed into the stirred vessel (for example, by means of a “3-way T- In principle, the aqueous solution B can also be continuously stirred into an aqueous solution A. However, this procedure is less preferred. As a rule, the aqueous mixture M obtainable as described is an aqueous suspension (preferably the ratios V described as advantageous are also present in the aqueous mixture M (total molar amount of NH3 + NH contained to molar amount of Mo contained); moreover, the pH of the aqueous mixture M obtainable as described is also advantageously < 3, generally 0 to 2). Aqueous mixtures M obtainable as described advantageously contain no more or less than 60 mol% of the total molar amount of Co and / or Ni contained therein in dissolved form in the aqueous medium (at the temperature and the working pressure at which the aqueous mixture M was produced).Preferably, the above-mentioned proportion AT of the total molar amount of Co and / or Ni present in the aqueous mixture M, dissolved in the aqueous medium of the aqueous mixture M, is <90 mol% and particularly preferably <80 mol%, or <70 mol% or <60 mol%. According to the invention, the total content of Bi, Fe, Mo, etc. in the aqueous mixture M to be dried (preferably spray-dried) is expediently 3 to 20 wt.%, advantageously 5 to 15 wt.%, based on the amount of the aqueous mixture M. As a rule, AT is >3 mol% or <50 mol%. According to the invention, the aqueous mixture M is preferably converted into a finely divided, intimate dry mixture by spray-drying the aqueous mixture M (the drying of the aqueous mixture M preferably takes place as close as possible to its preparation). This means that the aqueous mixture M is first divided into finely divided droplets (sprayed) in a spray dryer and then dried in the spray dryer. Spray-drying preferably takes place in a hot air stream. In principle, however, other hot gases can also be used for the aforementioned spray-drying (e.g., nitrogen or air diluted with nitrogen, as well as other inert gases). Spray drying can occur either in cocurrent or countercurrent flow of the droplets to the hot gas. Typical gas inlet temperatures range from 250 to 450°C, preferably 270 to 370°C. Typical gas outlet temperatures range from 90 to 160°C. Spray drying preferably occurs in cocurrent flow of the droplets to the hot gas. The mean particle diameter of the resulting spray powder is typically 10 to 100 pim, preferably 15 to 60 pim and particularly preferably 25 to 50 pim (the diameter is determined according to ISO 13320-1 by means of light scattering on spray powder dispersed in air (dispersion air pressure 2.0 bar-abs). In general, references to a standard in this document refer to the edition of the standard that was in force on the priority date of this patent application and whose publication date (issue date) has the smallest time difference to the priority date of this patent application. The vibratory mass density (25°C, 1 atm-abs) of the spray powder is typically 500 to 1300 g / l and preferably 700 to 1100 g / l. The loss on ignition of the spray powder (3 h at 600°C (powder temperature) under standing, excess air) is typically 20 to 40 wt.%, preferably 25 to 35 wt.% of its initial weight. Until further processing, the spray powder can be stored temporarily, preferably in hermetically sealed containers (e.g., plastic drums). The storage temperature should not exceed 70°C and is preferably < 50°C. As a general rule, the storage temperature should not fall below -10°C. Since the spray powder is generally hygroscopic, extended contact with humid air should be avoided. Contact with humid air can impair the handling properties of the spray powder (e.g., its flowability) and ultimately reduce the catalytic activity of the solid catalyst bodies produced with it. Of course, the aqueous mixture M can also be dried by other drying methods, such as conventional evaporation (preferably under reduced pressure; the drying temperature will generally not exceed 150 °C). In principle, the drying of an aqueous mixture M can also be achieved by freeze-drying or Spin-Flash® drying. The spray powder can be coarsened, for example, by subsequent precompacting. "Precompacting" means a preliminary compaction of the powder prior to subsequent compaction, which produces the precursor molded body. If the precompacting is carried out dry, finely divided graphite, the dry inorganic fibers, and / or other shaping aids mentioned in this document (e.g., lubricants, reinforcing agents, and / or pore formers) can be mixed into the spray powder prior to precompacting. Mixing is carried out using a mixer, for example, a drum mixer, tumbler mixer, ribbon mixer, blade mixer, or intensive mixer, preferably with a mixer with a shearing action, for example, with a chopper. The inorganic fibers are preferably glass fibers, aluminum oxide fibers, silica fibers, carbon fibers, asbestos, silicon carbide, potassium titanate, or a mixture of the aforementioned fibers. The inorganic fibers are particularly preferably glass fibers. The inorganic fibers preferably have an average fiber length of 25 pim to 750 pim, preferably of 50 pim to 500 pim, more preferably of 75 pim to 400 pim, particularly preferably of 100 pim to 300 pim. The inorganic fibers preferably have an average fiber diameter of 3 pim to 18 pim, preferably of 5 pim to 15 pim, more preferably of 8 pim to 12 pim. The unsupported catalyst shaped body preferably contains from 0.1 to 15% by weight, preferably from 0.5 to 13% by weight, more preferably from 1.0 to 11% by weight, particularly preferably from 2 to 10% by weight, especially preferably from 2.5 to 9% by weight, particularly preferably from 3 to 8% by weight.-%, inorganic fibers, based on the weight of the unsupported catalyst molding. For example, precompaction can be carried out using a calender with two counter-rotating steel rollers. The precompacted material can then be reduced to the appropriate particle size for its intended use. This can be achieved very simply, for example, by forcing the precompacted material through a sieve with a defined mesh size. In principle, precompacting can also be carried out in a wet state. For example, the spray powder can be kneaded with the addition of water. Following kneading, the kneaded mass can be recombined to the desired fineness (see, for example, DE 100 49 873 A1) and dried, depending on the subsequent use. From the finely divided precursor mass (from the finely divided intimate dry mixture of the sources of the elemental constituents) precursor shaped bodies of regular or irregular geometry are formed by compression (compression or compaction) and then the full catalyst shaped bodies are produced by thermal treatment. As further fine-particle shaping aids, lubricants such as graphite, carbon black, polyethylene glycol, polyacrylic acid, stearic acid, starch, mineral oil, vegetable oil, water, boron trifluoride and / or boron nitride can be added to the fine-particle precursor mass before and / or during shaping. Reinforcing agents such as microfibers made of glass, asbestos, silicon carbide or potassium titanate can also be considered as shaping aids, which, after the shaping process, have a beneficial effect on the cohesion of the resulting compressed material (the resulting shaped body) through compaction. Pore-forming agents such as ammonium nitrate, ammonium carbonate, water and / or malonic acid can also be considered as shaping aids. The pore-forming agents decompose or evaporate to form pores during the thermal treatment. The use of lubricants in the context of such shaping can be found, for example, inin the documents DE 10 2007 004961 A1 , WO 2008 / 087116, WO 2005 / 030393, US 2005 / 0131253, WO 2007 / 017431 , DE 10 2007 005606 A1 and in DE 10 2008 040093 A1. Preferably, only fine-particle graphite is used as a lubricant. In particular, the fine-particle graphites recommended in WO 2005 / 030393, US 2005 / 0131253, WO 2008 / 087116, and DE 10 2007 005606 A1 are suitable. This applies in particular to the graphites used in the examples and comparative examples in these documents. Particularly preferred graphites are Asbury 3160 and Asbury 4012 from Asbury Graphite Mills, Inc., New Jersey 08802, USA, and Timrex®T44 from Timcal Ltd., 6743 Bodio, Switzerland. Based on the weight of the finely divided precursor mass to be formed, this can contain, for example, up to 15 wt.% of finely divided lubricant (e.g., graphite) based on its total weight. However, the lubricant content in the finely divided precursor mass to be formed (in the finely divided, intimate dry mixture) is usually < 9 wt.%, frequently < 5 wt.%, and frequently < 4 wt.%; this is particularly the case when the finely divided lubricant is graphite. As a rule, the aforementioned added amount is > 0.5 wt.%, usually > 2.5 wt.%. Typically, the compaction of the finely divided precursor mass (the finely divided, intimate dry mixture), optionally containing shaping aids, to the desired geometry of the precursor molded body is achieved by applying external forces (pressure) to the precursor mass. The molding apparatus or method used for this purpose is not subject to any restrictions. For example, compaction can be achieved by tabletting. The finely divided precursor mass (the finely divided intimate dry mixture) is preferably used in a touch-dry state. However, it can, for example, contain up to 10% of its total weight of substances that are liquid under standard conditions (25 °C, 1 atm-abs (1.01 bar-abs)). The finely divided precursor mass (the finely divided intimate dry mixture) can also contain solid solvates (e.g., hydrates) that contain such liquid substances in chemically and / or physically bound form. Of course, the finely divided precursor mass can also be completely free of such substances. The preferred shaping process by compacting the finely divided precursor mass (the finely divided, intimate dry mixture) is tableting. The principles of tableting are described, for example, in "Die Tablette," a handbook of development, production, and quality assurance, by W.A. Tschel and A. Bauer-Brandl, 2nd edition, Edition Verlag Aulendorf, 2002, and can be applied in a completely analogous manner to a tableting process according to the invention. Tableting is a process of compression agglomeration. The free-flowing feed mixture is introduced into a compression tool with a die between two punches and compacted by uniaxial compression, forming a solid compacted body. Tableting can be divided into four stages: metered insertion, compaction (elastic deformation), plastic deformation, and ejection. Tableting is carried out, for example, on rotary presses or eccentric presses. The outer surface of the tableted catalyst support consists of a circumferential surface corresponding to the inner wall of the die cavity, and a first end face and a second side face corresponding to the operative heads of the punches. The tableted catalyst support can be flat or have curved ends, i.e., at least one of the first side faces and the second side face are curved. Curved side faces can be obtained, for example, by using a concave lower and / or upper punch. If desired, the upper punch and / or the lower punch can comprise projecting pins to form internal passages. It is also possible to provide the pressing punches with a plurality of pins, so that one punch can be manufactured, for example, with four pins to produce shaped bodies with four holes (passages). Typical design features of such pressing tools can be found, for example, in US Pat. No. 8,865,614. Press tools typically consist of a die, an upper punch, a lower punch, and pins (if the formed body has through holes). Suitable materials for press tools include tool steels, tungsten carbide (WC)-based cemented carbides, and ceramic materials. Tool materials with a hardness greater than 55 on the Rockwell C scale are preferred. Examples of tool steel materials are DIN tool steels 1.2210, 1.2343, 1.2436, 1.2379, 1.2601, 1.2080, and 1.25550, as well as high-speed steels, Vanadis 4 Extra from Uddeholm D-40549 Düsseldorf, and Vanadis 8 from Uddeholm D-40549 Düsseldorf. Suitable WC-based materials are described in US Pat. No. 8,865,614. Examples of such WC-based materials are G10-NI from Hartmetall® Gesellschaft in D70497 Stuttgart and htc-KR17® from Hightech-Cerarn®. Examples of ceramic materials are yttrium-stabilized zirconium oxide (YSZ). WC-based cemented carbides and ceramic materials are particularly suitable for tool use, in which a lined die made of WC-based cemented carbide or ceramic is inserted into a steel housing made of tool steel, e.g. 1.2379. The pressing tool typically has a surface coating to improve surface hardness, corrosion resistance, wear resistance, friction, and anti-adhesion properties. Examples of surface coating types include diamond-like carbon (DLC), boron nitride, titanium nitride, chromium nitride, plasma chrome plating, and hard chrome plating. The coating thickness is 1 to 10 μm, preferably 1 to 5 μm. The surface of compression tools that come into contact with the feed mixture and the resulting tablet preferably has a low surface roughness. The arithmetic mean roughness value Ra according to DIN 4768 of compression tool surfaces should preferably be 0.01 to 0.5 μm, particularly preferably 0.02 to 0.3 μm, even more preferably 0.02 to 0.2 μm, and most preferably 0.02 to 0.1 μm. The length of the tip of the lower punch is preferably 2 to 7 mm, more preferably 2 to 6 mm, and most preferably 2.5 to 5 mm. Excessively high tip lengths can lead to high friction, especially if sticking of the precursor shaped body occurs. It is preferred that the upper and lower edges of the tip of the lower punch are sharp rather than rounded. The sharp edge mitigates the jamming of the powder in the gaps at the punch-die interface and the pin-hole interface (in tablet shapes with through holes). Jamming of the powder leads to both sticking and powder leakage. The length of the top straight line of the upper punch is preferably greater than 2 mm and typically ranges from 2 to 10 mm. Unlike the lower punch, a long straight line length does not cause friction problems, as the upper punch is only inserted a few millimeters into the die during tabletting cycles. If the tablet shape has through holes, the lower punch and upper punch have holes to accommodate the pins. The upper punches should have at least one vent hole to allow air to escape from the die cavity during compaction through the upper punch holes to the outside of the upper punch holes. Such upper punches with vent hole(s) are specified in US 2010 / 0010238 (see Figures 4a, 4b, 4c, and 4d). The distance between the die bore and the lower punch outer surface is preferably 3 to 50 pm, more preferably 5 to 35 pm, most preferably 6 to 26 pm. Analogously, the distance between the die hole and the upper punch outer surface is preferably 3 to 50 pm, more preferably 5 to 35 pm, most preferably 6 to 26 pm. The distance is ensured by selecting a suitable combination of the dimensional tolerances of the die and the lower punch. The dimensional tolerance is typically represented according to the ISO shaft tolerance defined in ISO 286-2. Examples of the combinations of the dimensional tolerance of the die bore and the punch outer surface represented in ISO tolerance codes are H6 / f7, H6 / g6, H6 / g7, H7 / g6, H7 / f7, F8 / h6, G7 / h6, F7 / h6 (die bore / lower punch outer surface). In cases where the tablet shape has passages, the compression tool contains pins. The distance between the pin bore of the lower punch and the pin is preferably 3 to 50 μm, more preferably 5 to 35 μm, and most preferably 6 to 26 μm. Analogous to the distance between the die bore and the outside of the punch, this distance is ensured by selecting a suitable combination of the dimensional tolerances of the die and the lower punch. The die bore preferably has a slight taper, starting from a defined depth toward the die top. The tapered portion of the die bore features a gradual increase in bore size toward the die top, creating additional clearance between the die bore wall and the sharp-edged outer side surface of the lower punch. The additional space facilitates the venting of air contained in the compound feed during compression in the die, thereby mitigating powder blow-off and the resulting unstable tabletting due to poor air venting. Another advantage of tapered dies is the ease of ejection after compaction.Compaction into a tablet at the point where the die bore tapers results in a tablet with a slightly tapered outer side surface due to the imprinting effect of the conical die bore. During the ejection phase, in which the lower punch pushes the tablet upwards in the die, the tablet is easily ejected from the die wall due to a slight lifting of the tablet due to the tapered structure. forces the tablet to detach from the die wall. If the die bore is untapered, the tablet will not detach from the die wall, so the entire ejection process (i.e., pushing the tablet up from the depth at which compression occurred to the top of the die) suffers from friction between the outer surface of the tablet and the die wall, and between the straight outer surface of the tip and the die wall. This friction leads to an unfavorably high ejection force. The depth (i.e., depth from the die top) of the die taper should be selected so that the formed tablet is predominantly located in the conical zone before ejection. To achieve this, the depth of the die taper can be oriented by summing the in-die tablet height before elastic recovery (i.e., minimum distance between the upper and lower punches) and the insertion depth of the upper punch. For example, a taper depth of 12 to 15 mm can be used for an in-die tablet height of 12 mm with an insertion depth of 2 mm of the upper punch. The die taper angle is typically 0.1 to 0.6°, and the die bore size increase on the upper side is preferably 0.03 to 0.2 mm, more preferably 0.05 to 0.14 mm. The size increase can be mathematically derived from the taper angle and the taper depth. In cases where the tablet shape has passageways, the compression tool includes pins. The pins are attached to the turret so that the pins are located in the die cavity where the tablet is formed, to exit the tablet's passageways. Analogous to the die, the pins do not move vertically during the tabletting cycle, unlike the upper and lower punches. The vertical plane of the upper end of the pins is equal to or slightly below the level of the upper face of the die. Particularly in cases where the tablet has curved sides, where the lower punching surface has a concave surface, the vertical plane of the upper end of the pin should be slightly below the level of the upper face of the dies to prevent the pins from protruding from the lower punching surface. In cases where the tablet mold has through holes, sticking to the surface of the pins often occurs, leading to disadvantages such as high ejection force due to high friction at the pin-tablet interfaces. This problem is particularly pronounced in multi-threaded molds with multiple pins. Typically, the pins exhibit a higher tendency to stick than the die wall and the tip of the lower punch. In cases where the tablet shape has passages, the pins preferably have a slight taper in the upper area for a defined length. The tapered part of the pins shows a gradual reduction in the pin diameter towards the top of the pin. The main advantage of the tapered pins is the ease of ejection after compaction. Compaction into a tablet at the point where the pins have a taper results in tablet passages with a slightly tapered inner side surface as a result of the imprinting by the tapered pins. The diameters of the tablet passages decrease slightly along the axial axis from bottom to top.During the ejection phase, where the lower punch pushes the tablet upwards in the die while the pins and die remain vertically stationary, tablet ejection from the pins is easy because a slight lift of the tablet forces the tablet to detach from the pins due to the tapered structure. If the pins do not have a taper, the tablet detachment from the pins does not occur, so the entire ejection process (i.e., pushing the tablet up from the depth at which compression onto the upper face of the die) suffers from friction at the tablet-pins interfaces, resulting in a high ejection force. The tapered pins are particularly advantageous when the tablet has multiple passes. The length of the pin taper should be selected so that the formed tablet is predominantly located in the conical zone before ejection. To achieve this, the length of the pin taper can be oriented by summing the height of the in-die tablet before elastic recovery (i.e., minimum distance between the upper and lower punches) and the insertion depth of the upper punch. For example, a tapered length of 12 to 15 mm can be used for an in-die tablet height of 12 mm with an insertion depth of 2 mm of the upper punch. The die taper angle is typically 0.1 to 0.6°, and the reduction in the pin diameter on the upper side is preferably 0.05 to 0.3 mm, more preferably 0.1 to 0.2 mm. The reduction can be mathematically derived from the taper angle and the taper length. Industrial mass production of the tablets is preferably carried out on a rotary tablet press. Commercially available rotary presses can be used for this invention. Examples of rotary tablet presses are the Korsch XT-600 HD, Korsch XT-600, Korsch TPR 700, Korsch TRP 1200, Korsch XL 400 MFP, Kilian RX, and Kilian Synthesis. Rotary presses typically have two compaction rollers to perform a two-stage compaction process, comprising pre-compaction and main compaction. The main compaction pressure is in the range of 5 to 500 MPa, preferably 8 to 400 MPa, more preferably 10 to 300 MPa. The pre-compaction pressure is typically in the range of 5 to 50%, preferably 7 to 40%, more preferably 10 to 35% of the applied main compaction pressure. The compression tool is selected according to the desired geometric dimensions of the compacted body. The size and shape of the compacted body, and thus of the catalyst, are chosen to enable suitable packing of the catalyst bodies obtained from compacted bodies in a reactor tube. The catalysts obtained from the compacted bodies suitable for the catalysts according to the invention are preferably used in reactor tubes with a length of 2 to 5 m, preferably 2.5 to 4 m, more preferably 3.4 m, and an internal diameter of 15 mm to 50 mm, preferably 18 to 40 mm, more preferably 20 to 30 mm, particularly preferably 23 to 29 mm, and especially preferably 25 to 27 mm. The shape of the compacted bodies is not particularly limited and can be any technically feasible form, depending, for example, on the forming process. For example, the support can be a solid tablet or a hollow tablet, such as a hollow cylinder. According to a further embodiment, the support can be characterized by a multi-lobe structure. A multi-lobe structure is intended to refer to a cylinder structure that has a plurality of cavities, e.g., grooves or furrows, that run in the cylinder periphery along the cylinder height. In general, the cavities are arranged substantially equidistantly around the circumference of the cylinder. The compression force during tabletting influences the compaction of the free-flowing feed mixture and thus, for example, the density and / or mechanical stability of the compacted body. In practice, it has proven useful to specifically adjust the lateral compressive strength of the tableted catalyst support by selecting the appropriate compression force and to check it by random sampling. For the purposes of the present invention, the lateral compressive strength is the force that breaks the tableted catalyst support located between two flat, parallel plates, with the two flat, parallel end faces of the catalyst support at right angles to the flat, parallel plates. To improve tabletting properties, the free-flowing feed mixture can be subjected to further processing, e.g., sieving, preheating, and / or pregranulation, i.e., precompaction. A roller compactor, such as a Fitzpatrick Chilsonator®, can be used for pregranulation. Further information on tabletting, particularly on pre-granulation, sieving, lubricants, and tools, can be found in WO 2010 / 000720. Further information on tabletting can be found in Handbook of Powder Technology, Chapter 16: Tabletting, K. Pitt and C. Sinka, Vol. 11, 2007, pp. 735 to 778. Advantageously, tabletting is carried out as described in WO 2005 / 030393, DE 10 2008040093 A1, DE 10 2008 040094 A1 and WO 2007 / 017431. The temperature surrounding the tabletting machine is normally 25 °C. The particle diameters of the precursor mass to be compacted (the finely divided, intimate dry mixture), optionally as a result of pre-coarsening by compaction, are expediently in the range 100 to 2000 pm, preferably 150 to 1500 pm, particularly preferably 200 to 1250 pm, or 300 to 1000 pm, or 400 to 800 pm (shaping aid mixed in prior to compaction is not taken into account). Just as the shaping apparatus to be used for compaction or the shaping method to be applied, the desired geometry of the resulting shaped bodies is not subject to any restrictions in the process according to the invention. A preferred ring geometry (the geometry of an uncalcined green compact and the all-solid catalyst body resulting from it through calcination are normally essentially the same) is 5 mm x 5 mm x 2 mm (outer diameter x height (length) x inner diameter). This is because fixed catalyst beds made up of rings of this geometry result in a particularly low pressure drop in the reaction gas mixture flowing through the fixed catalyst bed (particularly in reaction tubes with an inner diameter of 22 mm to 30 mm). A low pressure drop is particularly advantageous when the partial oxidation catalyzed by the fixed catalyst bed is operated at a high loading of the fixed bed with reaction gas mixture (at a high flow rate of the reaction gas mixture flowing through the fixed catalyst bed).Another preferred ring geometry (it exhibits advantageous bulk behavior, particularly in reaction tubes with a smaller inner diameter (e.g., 21 mm)) is the geometry 5 mm x 3 mm x 2 mm (outer diameter x height (length) x inner diameter). Of course, all geometries disclosed and recommended in WO 02 / 062737 and WO 2015 / 067656 are also possible. Particularly in the case of ring-shaped precursor bodies (precursor bodies are also referred to as green compacts in the literature regardless of their shape), the shaping compaction should advantageously be carried out in such a way that the lateral compressive strength SDF V of the resulting molded body (cf. DE 10 2008 040093 A1, DE 10 2008 040094 A1 and WO 2005 / 030393) the relation 5 N < SDF V s 100 N, preferably 8 N < SDF V 80 N, and particularly preferably 12 N < SDF V 50 N is fulfilled. The experimental determination of the lateral compressive strength is carried out as described in the documents WO 2005 / 030393 and WO 2007 / 017431. Of course, ring-like green compacts, as recommended in DE 10 2008 040093 A1, are particularly preferred according to the invention. The end faces of ring-shaped or ring-like molded bodies can be (both or only one of the two end faces) flat or convex (curved) outwards (cf. in particular DE 10 2007 004961 A1, EP 0 184 790 A2, DE 10 2008 040093 A1 (for example, its paragraph

[0032] ) and DE 10 200804009 A1 (for example its paragraph

[0074] ) and the embodiments individually presented in these documents). When determining / specifying the height of such geometric shaped bodies, such a convex curvature is generally not taken into account in the present document. Rings or ring-like shaped bodies with convexly curved (curved) end faces (preferably both end faces have the same curvature) are advantageous in that fixed catalyst beds of ring-shaped or ring-like shaped bodies with a convexly curved (curved) end face (with otherwise identical geometry) result in a lower pressure drop in the reaction gas mixture flowing through the fixed catalyst bed (especially in the case of fixed catalyst beds in reaction tubes) than fixed catalyst beds of ring-shaped or ring-like shaped bodies with a flat end face.This is especially true when the ring geometry is 5 mm x 5 mm x 2 mm (outer diameter x height (length) x inner diameter). The radius of such a convex curvature is generally 0.4 to 5 times (e.g., 0.8 to 4 times, or 1.2 to 3 times, or 1.6 to 2.6 times) the outer diameter of the circular cylinder of the catalyst ring. As already mentioned, a fixed catalyst bed resulting in a lower pressure drop is particularly advantageous when the partial oxidation catalyzed by the fixed catalyst bed is operated at a high loading of the fixed bed with reaction gas mixture (at a high flow rate of the reaction gas mixture flowing through the fixed catalyst bed). Particularly advantageous ring geometries of molded bodies obtainable by compacting finely divided precursor mass (finely divided intimate dry mixture) satisfy the condition height (length) / outer diameter = H / A = 0.3 to 1.5 or up to 1.2. H / A = 0.5 to 1.1 or up to 1.0 is particularly preferred. Furthermore, for ring-shaped or ring-like green compacts according to the invention, it is advantageous if the ratio l / A (where l is the inner diameter of the ring geometry) is 0.3 to 1.5, preferably 0.6 to 1.1. The aforementioned ring geometries are particularly advantageous if they simultaneously exhibit one of the advantageous H / A ratios and one of the advantageous I / A ratios. Such possible combinations are, for example, H / A = 0.3 to 1.5 or up to 1.2 and I / A = 0.3 to 1.5 or 0.6 to 1.1. Alternatively, H / A can be 0.5 to 1.1 or up to 1.0 and l / A can simultaneously be 0.3 to 1.5 or 0.6 to 1.1. Furthermore, for the relevant ring geometries, it is favorable if H is 2 to 7 mm, and preferably 2 to 6 mm or 3 to 6 mm. Furthermore, it is advantageous if A for the rings is 4 to 8 mm, preferably 4 to 6 mm. The wall thickness of preferred ring geometries is 1 to 2 mm or up to 1.5 mm. Possible ring geometries according to the invention are thus (A x H x I) 5 mm x 5 mm x 2 mm, or 5 mm x 2 mm x 2 mm, or 5 mm x 3 mm x 2 mm, 5 mm x 5 mm x 2.5 mm, or 5 mm x 3 mm x 2.5 mm, or 5.5 mm x 5.5 mm x 2.5 mm, or 5.5 mm x 5.5 mm x 3 mm, or 5 mm x 3 mm x 3 mm, or 5.5 mm x 3 mm x 3.5 mm, or 6 mm x 3 mm x 4 mm, or 6 mm x 6 mm x 3 mm, or 6 mm x 6 mm x 3.5 mm, or 6.5 mm x 3 mm x 4.5 mm, or 7 mm x 3 mm x 5 mm, or 7 mm x 7 mm x 3 mm, or 7 mm x 3 mm x 4mm, or 7mm x 7mm x 4mm. All information in this document on specific surface areas of solids refers to determinations according to DIN 66131 (Determination of the specific surface area of ​​solids by gas adsorption (N2) according to Brunauer-Emmert-Teller (BET)), unless expressly stated otherwise. All information in this document regarding total pore volumes and pore diameter distributions based on these total pore volumes is based on determinations using the mercury porosimetry method using the Auto Pore V 9600 instrument (MicroActive Interactive Dara Analysis Software) from Micromeritics GmbH, D-52072 Aachen (range: 0.5 - 60,000) at 23°C (evaluation using the Washburn equation, assuming a mercury contact angle of 140° and a mercury surface tension of 480 mN / m = 480 dyn / cm). The total pore volume here refers to the total pore volume of pores with a diameter of >0.03 to <300 pm. According to the invention, precursor moldings advantageously have the lowest possible residual moisture content. This is particularly true when the intimate mixing of the various sources of elemental constituents other than oxygen was carried out wet (especially when it was carried out to form an aqueous mixture M). Preferably, the residual moisture content of advantageous green compacts is < 10 wt.%, better < 8 wt.%, even better < 6 wt.%, and most preferably < 4 wt.% or < 2 wt.% (the residual moisture determination can be carried out as described in "Die Bibliothek der Technik", Volume 229, "Thermogravimetric Material Moisture Determination", Fundamentals and Practical Applications, Horst Nagel, Verlag moderne Industrie, (e.g. using a Computrac MAX 5000 XL from Arizona Instruments)). If the green compacts are based on an aqueous mixture M (so that their residual moisture content consists of water), the residual moisture determination is carried out with halogen heating (e.g. halogen moisture analyzer HC103, Mettler Toledo). Against this background, spray drying of a wet (e.g. aqueous) mixture M should be carried out in such a way that the resulting spray powder has the lowest possible residual moisture content. Taking into account the aspect just addressed, the green compacts should be stored, if possible, in the absence of ambient air (containing atmospheric humidity) (preferably storage takes place until calcination under anhydrous inert gas or under pre-dried air or in hermetically sealed containers). Advantageously, the forming / shaping and storage of finely divided, intimate dry mixture is carried out under exclusion of ambient air (containing air humidity) (e.g. under an atmosphere). The calcination of the green compacts normally takes place at temperatures (calcination temperatures) that reach or generally exceed at least 350°C. However, the temperature of 650°C is normally not exceeded during calcination (the term calcination temperature in this document refers to the temperature present in the calcination material (the calcination material advantageously has a calcination temperature that is as uniform (homogeneous) as possible; this applies accordingly to the other calcination conditions)). During calcination, the temperature is advantageously not exceeded at 600°C, preferably at 570°C and frequently at 550°C. Furthermore, the temperature of 650°C is preferably not exceeded during the above calcination. The temperature should exceed 380°C, preferably 400°C, particularly advantageously 420°C, and most preferably 440°C. The calcination process can also be divided into several stages. Advantageously, prior to calcination, a thermal pretreatment is carried out at temperatures of > 120°C and < 350°C, preferably > 150°C and < 320°C, particularly preferably > 170°C and < 290°C. Such a thermal pretreatment is advantageously carried out until the constituents contained within the mass to be thermally treated, which decompose into gaseous compounds under the conditions of the thermal treatment, have been largely (preferably completely) decomposed into gaseous compounds (the time required in this regard can be, for example, 3 to 15 hours, frequently 4 to 10 hours or 5 to 8 hours).This is generally the case when, on the one hand, the molar amount of cations other than metal ions contained in the mass to be calcined subsequently, based on the total molar amount of cations contained, is < 20 mol-% (preferably < 10 mol-%) and, on the other hand, the molar amount of O contained in the same mass. 2 ' different anions, based on the total molar amount of anions contained, is also < 20 mol-% (preferably < 10 mol-%). Favourable temperature windows for the final calcination temperature are therefore in the temperature range 400 to 600°C, or preferably in the temperature range 420 to 570°C, or particularly preferably in the temperature range 450 to 550°C. The total calcination duration is generally more than 0.5 hours, and frequently more than 2 hours. Calcination durations typically do not exceed 45 hours or 30 hours, respectively. The total calcination duration is often less than 25 hours. Generally, a shorter calcination duration is generally sufficient at higher calcination temperatures than at lower calcination temperatures. According to an advantageous embodiment of the calcination according to the invention, 550°C is not exceeded, and the calcination duration in the temperature range > 430°C and < 550°C extends from > 4 to < 25 hours. The entire thermal treatment (including a decomposition phase) of a precursor material (e.g., a green compact) can be carried out under an inert gas or an oxidative atmosphere such as air (or another mixture of inert gas and molecular oxygen), as well as under a reducing atmosphere (e.g., a mixture of inert gas, NH3, CO, and / or H2, or methane). Of course, the thermal treatment can also be carried out under vacuum. The atmosphere can also be varied throughout the thermal treatment. The thermal treatment (especially the calcination phase) preferably takes place in an oxidizing atmosphere. For technical application purposes, this consists predominantly of stationary or (preferably) moving air (particularly preferably, the mass to be thermally treated (the calcination material) is flowed through by an air stream). However, the oxidizing atmosphere can also consist of a static or moving mixture of, for example, 25 vol.% N2 and 75 vol.% air, or 50 vol.% N2 and 50 vol.% air, or 75 vol.% N2 and 25 vol.% air (a treatment atmosphere of 100 vol.% N2 is also possible). In principle, the thermal treatment (e.g., calcination) of the precursor mass (e.g., the green compacts) can be carried out in a wide variety of kiln types, such as heatable forced-air chambers (forced-air kilns, e.g., forced-air shaft kilns), tray kilns, rotary kilns, belt calciners, or shaft kilns. According to the invention, the thermal treatment (e.g., calcination) is advantageously carried out in a belt calcining device, as recommended in DE 100 46 957 A1 and WO 02 / 24620. The formation of hot spots within the material to be treated (within the calcination material) is largely avoided by using fans to convey increased volume flows of calcination atmosphere through the calcination material via a gas-permeable conveyor belt carrying the calcination material (this ensures the most uniform calcination temperature possible in the calcination material). Within the scope of the thermal treatment of the precursor masses (for example the green compacts) to be carried out as described, co-used shaping aids can be used both in the resulting catalyst shaped body remain intact, and also escape at least partially in gaseous form from them through thermal and / or chemical decomposition to form gaseous compounds (e.g., CO, CO2). Shaping aids remaining in the shaped catalyst body act, in the context of catalytic use thereof, essentially exclusively as diluting agents for the active material. In principle, the thermal treatment in this regard can be carried out as described in US 2005 / 0131253. Typically, the lateral crushing strengths of annular unsupported catalyst bodies obtainable according to the invention as described are 4 to 30 N, preferably 6 to 25 N, more preferably 8 to 20 N and particularly preferably 9 to 17 N. The specific (BET) surface area of ​​the solid catalyst body is advantageously 2 to 20 m 2 / g or up to 15 m 2 / g, preferably 3 to 10 m 2 / g and particularly preferably 4 to 8 m 2 / G. The corresponding total pore volume (mercury porosimetry) is advantageously in the range of 0.30 to 0.70 cm 3 / g, preferably in the range 0.3 to 0.6, preferably 0.39 to 0.60, more preferably 0.40 to 0.55 cm 3 / g and particularly preferably in the range 0.41 to 0.50 cm 3 / G. If the pore diameter in pim is plotted on the abscissa and the logarithm of the differential contribution in cm on the ordinate 3 / g of the respective pore diameter to the total pore volume in cm 3 / g, particularly favorable unsupported catalyst bodies according to the invention generally exhibit a substantially monomodal distribution (with only a pronounced maximum). If the contribution of pores with a pore radius < 0.1 pim to the total pore volume is < 0.05 cm 3 / g, particularly good overall target product selectivities result (for example, in the case of a heterogeneously catalyzed partial oxidation of propene to acrolein and / or acrylic acid). In the case that the contribution of such comparatively narrow pores to the total pore volume is > 0.05 cm 3 / g, an advantageous reduction of this contribution can be achieved according to the invention by increasing the calcination time and / or the calcination temperature. Furthermore, it proves to be advantageous for an increased overall target product selectivity if the contribution of pores with a pore radius in the range of 0.1 to 1 pim to the total pore volume, based on the total pore volume, is 85 to 99 vol.%, advantageously 87 to 97 vol.%, particularly preferably 89 to 95 vol.%. Of course, the unsupported catalyst bodies can also be used diluted with inert materials to catalyze heterogeneously catalyzed partial gas-phase oxidations. Suitable inert diluent materials include elemental oxides such as aluminum oxide, silicon dioxide, thorium dioxide, and zirconium dioxide, which are fired at high temperatures and are therefore comparatively low in pores. However, finely divided silicon carbide or finely divided silicates such as magnesium and aluminum silicate or steatite can also be used for the aforementioned purpose. From an application point of view, it is advantageous, for example, to grind the calcined active material into a finely divided powder. This is then mixed with finely divided diluent material, and the resulting mixed powder is formed into a geometric shaped body using a shaping process presented in this document (preferably by tabletting).By subsequent calcination, the latter is then transformed into the corresponding unsupported catalyst body. Of course, the finely divided inert diluent material can also be incorporated, for example, into a wet (e.g., aqueous) mixture M prior to its drying. Furthermore, finely divided inert diluent material can be incorporated into the finely divided dry mixture. However, such procedures are less preferred according to the invention. In particular, shaped catalyst bodies produced according to the described advantageous production processes are characterized by the fact that they essentially have no local centers of elemental oxides. Rather, these elements are largely components of complex, mixed oxomolybdates containing Bi, Fe, and Mo. This has proven advantageous with regard to the inventively sought minimization of undesirable complete combustion of organic reaction gas mixture components during the relevant heterogeneously catalyzed partial oxidations. Furthermore, the procedure for producing full catalysts is as described in WO 2010 / 066645, in order to ensure the most efficient use of materials. Shaped unsupported catalyst bodies according to the invention are suitable not only for catalyzing the heterogeneously catalyzed partial oxidation of propene to acrolein, but generally for catalyzing heterogeneously catalyzed partial gas phase oxidations of alkanes, alkanols, alkenes and / or alkenals having 3 to 6 carbon atoms (partial oxidations are to be understood in this document as meaning, in particular, those reactions of organic compounds under the reactive action of molecular oxygen in which the organic compound to be partially oxidized contains at least one more chemically bonded oxygen atom after the reaction has ended than before the partial oxidation). However, the term partial oxidation is also to be understood in this document as encompassing oxidative dehydrogenation and partial ammoxidation, i.e., partial oxidation in the presence of ammonia. Unsupported catalyst bodies according to the invention are particularly suitable for catalyzing the heterogeneously catalyzed partial gas phase oxidation of propene to acrolein, of isobutene to methacrolein and for catalyzing the heterogeneously catalyzed partial gas phase ammoxidation of propene to acrylonitrile and of isobutene to methacrylonitrile. As already mentioned, the heterogeneously catalyzed partial gas phase oxidation of propene (isobutene and / or tert-butanol) to acrolein (methacrolein) forms the first stage of a two-stage heterogeneously catalyzed partial gas phase oxidation of propene (isobutene and / or tert-butanol) to acrylic acid (methacrylic acid), as described by way of example in WO 2006 / 42459. The formation of acrylic acid (methacrylic acid) as a by-product in a heterogeneously catalyzed partial gas phase oxidation of propene (isobutene) to acrolein (methacrolein) is therefore generally not undesirable and is normally subsumed under the desired formation of the value product. The above applies in particular to annular shaped unsupported catalyst bodies according to the invention. The heterogeneously catalyzed partial oxidation (in particular that of propene to acrolein) can be carried out, for example, as described in the documents DE 10 2007 004961 A1, WO 02 / 49757, WO 02 / 24620, DE 10 2008 040093 A1, WO 2005 / 030393, EP 0 575 897 A1, WO 2007 / 082827, WO 2005 / 113127, WO 2005 / 047224, WO 2005 / 042459, WO 2007 / 017431, DE 10 2008 042060 A1, WO 2008 / 087116, DE 10 2010 048405 A1, DE 10 2009 047291 A1, DE 10 2008 042064 A1, DE 10 2008 042061 A1, WO 2015 / 067656 and DE 10 2008 040094 A1 for similar catalysts (in particular, the procedure can be similar to that described in the exemplary embodiments of these documents). The advantage of the (in particular the ring-shaped) unsupported catalyst bodies obtainable as described also exists when the loading of the fixed catalyst bed of a reactor with propene, isobutene and / or tert. butanol (or its methyl ether) contained in the reaction gas input mixture is > 90 Nl / l fixed catalyst bed • h, or > 110 Nl / l ■ h, or > 120 Nl / l h, or > 130 Nl / l ■ h (pre- and / or subsequent fillings of pure inert material are not considered to be part of the fixed catalyst bed when considering loading in this document; the volume of the fixed catalyst bed is, moreover, its bulk volume in the reactor). Normally, the aforementioned loading of the fixed catalyst bed will be < 400 Nl / l-h, frequently < 300 Nl / l-h, frequently < 250 Nl / l-h or < 230 Nl / l-h. Loadings in the range of > 90 Nl / l-h or > 100 Nl / l-h to < 220 or < 200 or < 190 Nl / l-h are particularly suitable. In this document, the loading of a fixed catalyst bed with reaction gas input mixture is understood to mean the amount of reaction gas input mixture in standard liters (= NI; the volume in liters that the corresponding reaction gas input mixture quantity would occupy at the standard conditions of 0 °C and 1 atm-abs (1.01 bar-abs)) that is fed to the fixed catalyst bed per hour, based on the volume of its bed (bed sections made of pure inert material are not included), i.e., its bed volume (-> unit = Nl / I ■ h). The standard volume units such as NI or Nm 3 In this document, always refer (unless otherwise expressly stated) to the standard conditions of 0 °C and 1 atm-abs (1.01 bar-abs). The loading can also be related to only one component of the reaction gas input mixture (e.g., only to the organic starting compound to be partially oxidized). In this case, it is the volume of this component (e.g., the organic starting compound of the partial oxidation) that is fed to the fixed catalyst bed per hour, based on the volume of its bed. Naturally, shaped catalyst bodies (for example ring-shaped) obtainable according to the invention can be operated advantageously according to the invention as catalysts for the partial oxidation of propene to acrolein or of isobutene and / or tert. butanol (or its methyl ether) to methacrolein even at loadings of the fixed catalyst bed with the starting compound to be partially oxidized of < 130 Nl / l h, or < 120 Nl / l h, or < 110 Nl / l h, or < 100 Nl / l h, or < 90 Nl / l h. As a rule, however, this load will be at values ​​> 20 Nl / I'h, or > 30 Nl / I'h, or > 40 Nl / I'h, or > 50 Nl / I'h, or > 60 Nl / I'h, or > 70 Nl / I h, or > 80 Nl / I h. In principle, the loading of the fixed catalyst bed of a reactor with the starting compound to be partially oxidized (propene, isobutene and / or tert. butanol (or its methyl ether)) can be adjusted via two adjusting screws: a) the loading of the fixed catalyst bed with reaction gas input mixture (the reaction gas mixture that is fed to the fixed catalyst bed), and / or b) the content of the starting compound to be partially oxidized in the reaction gas input mixture. The (for example annular) unsupported catalyst bodies obtainable according to the invention are particularly suitable when, at loadings of above 120 Nl / l ■ h on the fixed catalyst bed with the organic compound to be partially oxidized, the loading is adjusted primarily via the aforementioned adjusting screw a). As a rule, for example, the propene content (isobutene content or tert-butanol content (or the methyl ether content)) in the reaction gas input mixture will be 4 to 10 vol.%, frequently 5 to 9 vol.%, or 5.5 to 8.0 vol.%, or 6.0 to 7.5 vol.%, essentially independent of the loading of the fixed catalyst bed (in each case based on the total volume (flow) of the reaction gas input mixture flowing into the fixed catalyst bed). The gas phase partial oxidation process of the partial oxidation catalyzed by the inventive (for example annular) unsupported catalyst bodies obtainable as described (essentially independent of the load) is frequently carried out with an (organic) compound to be partially oxidized (for example propene): oxygen: inert gases (including water vapor) volume ratio in the reaction gas input mixture of 1:(1.0 to 3.0):(5 to 25), preferably 1:(1.5 to 2.3):(10 to 20). Indifferent gases (or inert gases) are understood to be those gases which remain chemically unchanged to at least 95 mol%, preferably to at least 98 mol%, during the partial oxidation when the reaction gas mixture passes through the catalyst bed once. In the reaction gas input mixtures described above, the indifferent gas can be > 20 vol.%, or > 30 vol.%, or > 40 vol.%, or > 50 vol%, or > 60 vol%, or > 70 vol%, or > 80 vol%, or > 90 vol.%, or > 95 vol.% molecular nitrogen. However, at higher loadings of the reactor's fixed catalyst bed with the organic compound to be partially oxidized (e.g., > 150 Nl / l'h), the use of diluent gases with higher molar heat capacities and / or thermal conductivities, such as propane, ethane, methane, pentane, butane, CO2, CO, water vapor, and / or noble gases, for the reaction gas input mixture is recommended (but not mandatory). In general, however, these inert gases and their mixtures can also be used at lower loadings of the fixed catalyst bed with the organic compound to be partially oxidized. Recycle gas can also be used as a diluent. Recycle gas is the residual gas that remains when the target compound is essentially selectively separated from the product gas mixture of the partial oxidation.It should be taken into account that the partial oxidations to acrolein or methacrolein with the, for example, ring-shaped, unsupported catalyst bodies obtainable according to the invention can only be the first stage of a two-stage partial oxidation to acrylic acid or methacrylic acid as the actual target compounds, so that the cycle gas formation then usually only takes place after the second stage (typical cycle gas compositions for the case of a heterogeneously catalyzed partial oxidation of propene to acrolein and / or acrylic acid are shown in DE 102 32 482 A1 in its paragraphs.

[0063] and

[0075] ). 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 addition of secondary oxygen (usually as air), to the second partial oxidation stage. In the partial oxidation of propene to acrolein, using the inventive (for example, ring-shaped) unsupported catalyst bodies obtainable as described, a typical composition of the reaction gas input mixture measured at the reactor inlet (regardless of the selected loading) can, for example, contain the following components 5 to 8.0 vol.% propene, 1 to 15 vol% H2O, O to 3 vol.% COx, 0.01 to 0.1 vol.% acrolein, 8 to 15.5 vol.% oxygen and the remainder to 100 vol.% molecular nitrogen; or 6 to 7.5 vol% propene, 1 to 5 vol% H2O, 0.2 to 2 vol% COx, 0.015 to 0.06 vol.% acrolein, 9.4 to 14.5 vol.% oxygen and the remainder to 100 vol.% molecular nitrogen. Alternatively, reaction gas input mixtures of the composition according to Example 1 of EP 0 990 636 A1, or according to Example 2 of EP 0 990 636 A1, or according to Example 3 of EP 1 106 598 A2, or according to Example 26 of EP 1 106 598 A2, or according to Example 53 of EP 1 106 598 A2, or according to the examples of WO 2021 / 013640 can also be used for a propene partial oxidation to acrolein according to the invention. The inventive, for example annular, unsupported catalyst bodies obtainable as described are also suitable for the processes of DE 102 46 119 A1 or DE 102 45 585 A1. The reaction temperature for a heterogeneously catalyzed propene partial oxidation to acrolein according to the invention is frequently from 300 to 450°C, or up to 400°C, or up to 380°C when using the inventive (for example, ring-shaped) unsupported catalyst bodies obtainable as described. A particularly preferred reaction temperature window is from 305 to 365°C. The same applies in the case of methacrolein as the target compound. The reaction pressure for the aforementioned partial oxidations is generally from 0.5 to 5 or up to 4 bar-abs, or preferably from 1.1 or 1.5 to 3.5 or up to 3.3 bar-abs (unless expressly stated otherwise, absolute pressures (-abs) are always meant in this document). The total loading of the fixed catalyst bed with reaction gas input mixture in the aforementioned partial oxidations according to the invention typically amounts to 1000 to 10000 Nl / l-h, mostly to 1500 to 5000 Nl / l-h and often to 2000 to 4000 Nl / l-h. The propene to be used in the reaction gas input mixture is primarily polymer grade propene and chemical grade propene, as described, for example, in WO 2004 / 007405. Air (possibly together with recycle gas) is normally used as the oxygen source. The partial oxidation using the inventive (for example the annular) unsupported catalyst bodies obtainable as described can, in the simplest case, be carried out, for example, in a single-zone tube bundle fixed bed reactor, as described in DE 4431 957 A1, EP 0 700 714 A1 and EP 0 700 893 A1. Typically, the contact tubes in the aforementioned shell-and-tube reactors are made of ferritic steel and typically have a wall thickness of 1 to 3 mm. Their internal diameter is generally 20 to 30 mm, frequently 21 to 26 mm. A typical contact tube length is, for example, 3.50 m, 4.00 m, or 4.50 m. The number of contact tubes accommodated in the shell-and-tube vessel is expediently at least 1000, preferably at least 5000. Frequently, the number of contact tubes accommodated in the reaction vessel is between 15,000 and 40,000. Shell-and-tube reactors with a number of contact tubes above 45,000 are rather the exception. Within the container, the contact tubes are normally arranged in a homogeneous distribution, whereby the distribution is expediently chosen so that the distance between the central inner axes of the contact tubes closest to each other (the so-called contact tube pitch) is 35 to 45 mm (cf. EP 0 468 290 B1).However, the partial oxidation can also be carried out in a multi-zone (e.g., "two-zone") tube bundle fixed-bed reactor, as recommended in DE 199 10 506 A1, DE 103 13213 A1, DE 103 13208 A1, and EP 1 106 598 A2, particularly in the case of increased loading of the catalyst fixed bed of the tube bundle reactor with the organic compound to be partially oxidized. A typical catalyst tube length in the case of a two-zone tube bundle fixed-bed reactor is 3.50 m, 4.00 m, or 4.50 m. Everything else essentially applies as described for the single-zone tube bundle fixed-bed reactor. A heat exchange medium is passed around the contact tubes, within which the catalyst fixed bed is located, in each tempering zone (the single-zone tube bundle fixed bed reactor has only one tempering zone) of the single-zone or multi-zone tube bundle fixed bed reactor.Suitable examples include 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 alloys of various metals. The flow rate of the heat exchange medium within the respective tempering zone is generally selected so that the temperature of the heat exchange medium increases by 0 to 15°C, often 1 to 10°C, or 2 to 8°C, or 3 to 6°C, from the point of entry into the tempering zone to the point of exit from the tempering zone. The inlet temperature of the heat exchange medium, which can be conducted cocurrently or countercurrently to the reaction gas mixture across the respective temperature control zone, is preferably selected as recommended in EP 1 106 598 A2, DE 19948523 A1, DE 19948248 A1, DE 10313209 A1, EP 0 700 714 A1, DE 10313208 A1, DE 10313213 A1, WO 00 / 53557, WO 00 / 53558, WO 01 / 36364, WO 00 / 53557, and the other documents cited in these documents as prior art. Within the temperature control zone, the heat exchange medium is preferably conducted in a meandering pattern. The difference between the highest and lowest temperature of the heat exchange medium located within a tempering zone should, considered over that longitudinal section of the tempering zone in which a catalytically active (not pure inert) section of the fixed catalyst bed is located, advantageously be > 0°C and < 5°C (preferably this difference is small).Typically, the tube-bundle fixed-bed reactor also features thermocouples for determining the temperature of the reaction gas in the catalyst bed (both the thermocouple and the reaction tubes are charged with the same fixed bed). The inner diameter of the thermocouples and the diameter of the thermocouple sleeve (thermocouple sleeve), which is centered inside the tubes and runs parallel to the longitudinal axis of the thermocouple, are selected so that the ratio of the reaction heat-generating volume to the heat-dissipating surface area is the same or only slightly different for the thermocouples and the working tubes. The pressure drop should be the same for working tubes and thermotubes, based on the same GHSV (= the volumetric flow rate of the reaction gas mixture entering the tube divided by the bulk volume of the fixed catalyst bed in the tube). Pressure loss compensation for thermotubes can be achieved, for example, by adding crushed catalyst to the unsupported catalyst bodies. This compensation is preferably carried out homogeneously over the entire length of the thermotube. Furthermore, the filling of thermotubes can be designed as described in EP 0 873783 A1. The process according to the invention for producing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid by heterogeneously catalyzed gas-phase partial oxidation of at least one aldehyde with molecular oxygen in a tube-bundle reactor uses at least two catalyst layers, each comprising a catalytically active multielement oxide. The catalyst layer according to the invention contains at least one catalytically active multielement oxide. The fixed catalyst bed comprises all catalyst layers. According to the invention, at least two catalyst layers are present in the fixed catalyst bed, the number of catalyst layers preferably being 2 to 6, more preferably 2 to 5, particularly preferably 2 to 4, and especially preferably 2 to 3. The number of catalyst layers is preferably determined along the direction of the reaction gas flow from the gas inlet side to the gas outlet side. In principle, the local temperature of the catalyst bed at a given axial position along the reactor tube length / tube axis is the result of simultaneous heat generation and heat removal at that given axial position. Heat generation is due to the exothermic catalytic oxidation reactions taking place at that position. Heat removal (as long as the local temperature of the catalyst bed is higher than the local temperature of the heat exchange medium) is due to heat exchange driven by the temperature difference between the catalyst bed and the heat exchange medium surrounding the reactor tube. The heat generation density, which is mainly influenced by the concentrations of the reactants (e.g., propene and oxygen) and the catalytic activity, varies along the axial position.As a result, the temperature of the catalyst bed (which is essentially identical to the temperature of the reaction gas) exhibits a non-constant profile depending on the position along the reactor tube length / reactor axis, which is referred to as the temperature profile. The local temperature of the catalyst bed at a specific axial position can be determined by measuring the temperature inside the thermowell located in the center of the cross-section of a reactor tube. By measuring the local catalyst bed temperature at different axial positions, a temperature profile is created. The interval of the local bed temperature measurement is preferably <20 cm, more preferably <15 cm, particularly preferably <10 cm, and especially preferably <6 cm. A temperature profile has at least one local maximum, which is called a "local hotspot." The local hotspot with the highest temperature among all local hotspots is called a "hotspot." This "hotspot" is also called a hot spot. In the process according to the invention, the hotspot should not occur in the last catalyst layer (i.e., closest to the reactor gas outlet). If the number of catalyst layers is 2, the hotspot should occur in the first catalyst layer (furthest from the gas inlet side). If the number of catalyst layers is 3, the hotspot should occur in either the first or second catalyst layer relative to the reactor gas inlet. If the number of catalyst layers is greater than 3, the hotspot preferably occurs in either the first or second catalyst layer relative to the reactor gas inlet. According to a preferred embodiment, the hotspot temperature is 360 to 450 °C, preferably 370 to 440 °C, more preferably 380 to 430 °C, particularly preferably 390 to 425 °C, especially preferably 400 to 420 °C. According to a preferred embodiment, the difference between the hotspot temperature and the temperature of the heat exchange medium (at the axial position along the tube axis at which the hotspot occurs) is 40 to 120 °C, preferably 50 to 100 °C, more preferably 60 to 95 °C, particularly preferably 70 to 90 °C. Furthermore, the design is preferably such that this temperature difference increases by > 0°C and < +9°C, preferably < +7°C, more preferably < +5°C, particularly preferably < +3°C when the temperature of the heat exchange medium increases by 1°C at each position of the fixed catalyst bed (see also EP 1 106 598 A1). In order for the hotspot to occur in the desired catalyst layer mentioned above, (1) the ratio of the volume-specific catalytic activities of each catalyst layer and (2) the ratio of the (axial) lengths of the catalyst layers (along the tube axis) must be configured accordingly. The volume-specific catalytic activity is the catalytic activity per reactor tube volume, which includes both the catalyst bodies and the cavity between the bodies. The volume-specific activity can be adjusted by changing the volume-specific activity of the shaped catalyst body itself, by changing the intrinsic activity of the shaped catalyst body by changing the catalyst composition and the processing conditions (e.g. composition, calcination temperature, body density). Alternatively, a reduction in the volume-specific activity can be achieved in a simple manner, for example, by homogeneously diluting a basic amount of uniformly produced (e.g., ring-shaped) unsupported catalyst bodies with inert diluent bodies. The higher the proportion of The more diluted shaped bodies are selected, the lower the active mass or catalyst activity contained in a given volume of the catalyst layer. However, a reduction can also be achieved by changing the geometry of the unsupported catalyst bodies so that the amount of active mass contained per unit of internal reaction tube volume becomes smaller. To prepare the catalyst layer with a reduced volume-specific activity, as already mentioned, only unsupported catalyst bodies available as described (for example, ring-shaped) or, for example, largely homogeneous mixtures of unsupported catalyst bodies available as described (for example, ring-shaped) and shaped bodies containing no active material that are essentially inert with respect to the heterogeneously catalyzed partial gas-phase oxidation can be used. Suitable materials for such inert shaped bodies include, for example, porous or non-porous aluminum oxides, silicon dioxide, zirconium dioxide, silicon carbide, silicates such as magnesium or aluminum silicate, and / or steatite (for example, type C220 from CeramTec, Germany). The BET surface area of ​​the inert shaped bodies is preferably < 0.5 m 2 / g, preferably < 0.3 m 2 / g, more preferably < 0.1 m 2 / g. The lower limit for the BET surface area of ​​the inert molded bodies is preferably 0.001 m 2 / g, more preferably 0.01 m 2 / g. The geometry of such inert diluent bodies is arbitrary. This means that they can be, for example, spheres, polygons, solid cylinders, or even, as in the case of annular catalyst bodies, rings. Inert diluent bodies are often chosen whose geometry corresponds to that of the catalyst bodies they are to be diluted with. The geometry of the catalyst bodies can be changed, or catalyst bodies of different geometries can be used in a largely homogeneous mixture. According to a preferred embodiment, when the number of catalyst layers is 2, the ratio of the volume-specific catalytic activities of the first catalyst layer (located on the gas inlet side) to the second catalyst layer (located on the gas outlet side) is 50 to 95%, preferably 55 to 90%, more preferably 60 to 85%, particularly preferably 65 to 80%. Preferably, the length of the first catalyst layer, based on the gas inlet side, relative to the total length of the fixed catalyst bed (i.e., the sum of the lengths of the first and second catalyst layers) is 15 to 50%, preferably 20 to 45%, more preferably 22 to 43%, particularly preferably 24-40%. According to a preferred embodiment, when the number of catalyst layers is 3 or greater, the fixed catalyst bed is advantageously designed in such a way that the volume-specific activity in the flow direction of the reaction gas mixture either remains constant or increases. When the number of catalyst layers is 3, the preferred ratio of the volume-specific catalytic activities of the second catalyst layer (located midway between the first and third catalyst layers) relative to the third catalyst layer (located on the gas outlet side) is 50 to 95%, preferably 55 to 90%, more preferably 60 to 85%, particularly preferably 65 to 80%. In these cases, the preferred ratio of the volume-specific catalytic activities of the first catalyst layer (on the gas inlet side) relative to the second catalyst layer (midway between the first and third catalyst layers) is 70 to 95%, preferably 75 to 90%, more preferably 78 to 88%. Preferably, the same type of catalytically active multielement oxide is used for the first and second catalyst layers.When the number of catalyst layers is 3, the total length of the first and second catalyst layers relative to the total length of the catalyst bed (ie the sum of the lengths of the first, second and third catalyst layers) is 20 to 60%, preferably 25 to 55%, more preferably 30 to 50%, particularly preferably 35 to 45%. In the flow direction of the reaction gas mixture, upstream of the actual fixed catalyst bed, a bed of inert shaped bodies can be arranged, which, for example, has the purpose of increasing the inlet temperature of the reaction gas mixture to the temperature of the heat exchange medium. Furthermore, a heterogeneously catalyzed partial oxidation for the production of acrolein (from propene) or methacrolein (from the C4 precursor compounds mentioned in the present document) can advantageously be carried out using annular unsupported catalyst bodies produced according to the invention in complete accordance with the statements in WO 2015 / 067656. Overall, in a partial oxidation for producing acrolein or methacrolein carried out using the (for example ring-shaped) unsupported catalyst bodies obtainable as described as catalysts, the fixed catalyst bed, the reaction gas starting mixture, the space velocity and the reaction temperature are generally selected such that a single pass of the reaction gas mixture through the fixed catalyst bed results in 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 93 mol% or at least 94 mol% or at least 95 mol% or at least 97 mol%, but normally < 99 mol%. The selectivity of the value products (sum of acrolein and acrylic acid or sum of methacrolein and methacrylic acid) will regularly be > 80 mol-%, > 85 mol-% or > 90 mol-%. Finally, it should be noted that shaped unsupported catalyst bodies according to the invention obtainable as described also exhibit advantageous fracture behavior during reactor filling. The commissioning of a fresh fixed catalyst bed according to the invention containing (for example, annular) unsupported catalyst bodies can be carried out, for example, as described in DE 103 37 788 A1 or as described in DE 10 2009 047291 A1. The formation of unsupported catalyst bodies obtainable according to the invention can be accelerated by carrying it out at essentially constant conversion under increased loading of the fixed catalyst bed with reaction gas input mixture. Furthermore, unsupported catalysts obtainable according to the invention are generally suitable for catalyzing the gas phase partial (amm)oxidation of an alkanol, alkanal, aiken, alkane and alkenal containing 3 to 6 (i.e. 3, 4, 5 or 6) C atoms to, for example, olefinically unsaturated aldehydes and / or carboxylic acids and the corresponding nitriles, and for gas phase catalytic oxidative dehydrogenations of the aforementioned organic compounds containing 3, 4, 5 or 6 C atoms. The large-scale production of unsupported catalyst shaped bodies according to the invention is advantageously carried out analogously to that described in DE 10 2008 040093 A1 and DE 10 2008 040094 A1 (particularly advantageously as in the exemplary embodiments of these documents). A coated catalyst preferably consists of a geometric shaped support body and a catalytically active multielement oxide applied to the outer surface of the geometric shaped support body, along with an optional binder. The coated catalyst is prepared by applying the catalytically active multielement oxide and an optional binder to the outer surface of the geometric shaped support body. Examples Production of the solid catalyst bodies K1 to K10: Preparation of the full catalyst precursor body K1 , wherein the active multimetal oxide has the stoichiometry M012 Bio.6 Fea C07 Si 16 Koos O X had. The stirred vessels used were each filled in the presence of ambient air. During stirring / mixing, they were hermetically sealed with a lid that had a pressure relief valve to the atmosphere (1.01 bar-abs). a) Preparation of an aqueous solution B In a temperature-controlled cylindrical stirred vessel (1.75 m 3 430 kg of demineralized water were placed in a stainless steel (EN 1.4541) (internal volume, 1.3 m diameter) and heated to 60°C while stirring (70 rpm). Subsequently, while stirring continuously and maintaining the temperature at 60°C, 0.61 kg of an aqueous potassium hydroxide solution (47.5 wt% KOH) having a temperature of 20°C was added in one minute. While maintaining the temperature at 60°C, 136.2 kg of fine-grained ammonium heptamolybdate tetrahydrate (54.3 wt.% Mo, supplier: NIPPON INORGANIC COLOUR & CHEMICAL 00., LTD., 3-14-1 Funado ltabashi-ku Tokyo 174-0041 Japan) were then stirred in portions with constant stirring at a temperature of 25°C. The resulting aqueous solution (slightly turbid due to minimal insoluble isopolymolybdate impurities) was stirred at 60°C for 60 minutes (70 rpm). b) Preparation of an aqueous solution A In a temperature-controlled cylindrical stirred vessel (1.75 m 3212 kg of an aqueous nitric acid cobalt(II) nitrate solution (12.5 wt% Co, 27 wt% nitrate, prepared by dissolving cobalt metal from MFT Metals % Ferro-Alloys Trading GmbH, D-41474 Viersen, purity > 99.6 wt% Co, < 0.3 wt% Ni, < 100 mg / kg Fe, < 50 mg / kg Cu ​​in aqueous nitric acid) were placed in a stainless steel (EN 1.4541) (internal volume, 1.3 m diameter) and heated to 60°C while stirring (70 rpm). With continued stirring (70 rpm) and continued tempering at 60°C, 78 kg of a 60°C warm iron (l 1 l) nitrate nonahydrate melt (13.8 wt% Fe, < 0.4 wt% alkali metals, < 0.01 wt% chloride, < 0.02 wt% sulfate, from Dr. Paul Lohmann GmbH, D-81857 Emmerthal) were added and stirred for 30 minutes at 60°C (70 rpm). 72.6 kg of a 60°C warm, aqueous, nitric acid bismuth nitrate solution (11.1 wt.% Bi, 13 wt.% nitrate, prepared by dissolving bismuth metal from Sidech SA, BE-1495 Tilly, purity > 99.997 wt.% Bi, < 7 mg / kg Pb, < 5 mg / kg each of Ni, Ag, Fe, < 3 mg / kg each of Cu, Sb, and < 1 mg / kg each of Cd, Zn in aqueous nitric acid) were added to the resulting aqueous solution while maintaining the temperature at 60°C and continuing to stir (70 rpm). c) Mixing of aqueous solution A with aqueous solution B The 60°C warm aqueous solution A was continuously added over 15 minutes to the intensively stirred (70 rpm) 60°C warm aqueous solution B. The resulting aqueous suspension was stirred for a further 15 minutes at 60°C. d) Addition of a silica sol to obtain an aqueous mixture M After stirring was complete, 12.6 kg of silica gel heated to 60°C from Grace GmbH & KG, In der Hollerecke 1, D-67545 Worms, type LUDOX TM50 (47.5 wt% SiO2) was immediately added to the aqueous mixture obtained in c). e) Spray drying of the aqueous mixture M The spray-drying of the aqueous mixture M took place immediately after its preparation. The aqueous mixture (suspension) M, which was continuously stirred at 60°C (also during spray-drying) using a stirrer (40 rpm), was spray-dried in a stainless steel 1.4541 spray tower with a centrifugal atomizer and an atomizer wheel in a hot air cocurrent flow (gas inlet temperature: 350 ± 10°C, gas outlet temperature: 140 ± 5°C, gas flow rate: 2200 ± 100 Nm 3 / h, speed of the atomizer wheel: 20000 rpm). The resulting spray powder was stored in hermetically sealed containers (200 l or 1000 l internal volume, 25°C, atmospheric pressure) until further processing (10 calendar days; shorter or longer intermediate storage of up to 30 calendar days had no influence on the resulting results). The loss on ignition of the resulting spray powder (calcined for 3 h at 600°C (powder temperature) in stagnant, excess air) was 31 ± 2 wt.% of its initial weight. The spray powder exhibited a D50 of 35 ± 10 μm (D50 means that 50% of the particles are smaller than the specified value) at a dispersion pressure of 2.0 bar absolute. f) Production of ring-like precursor bodies 100 kg of spray powder and 1 kg of graphite (grade 3160 from Asbury Graphite Mills Ino., New Jersey 08802) were placed in an inclined mixer (type VIL, filling volume: 200 l, Aachener Misch- und Knetmaschinenfabrik) with mixing and cutting blades (speed of mixing blades: 39 rpm, speed of cutting blades: 3000 rpm) and premixed for 5 minutes. The resulting mixture was then compacted in a compactor type K300 / 200 from Hosokawa Bepex GmbH with concave, grooved smooth rollers (roller diameter 300 mm, roller length 200 mm, gap width: 2.8 mm, roller speed: 3.6 rpm, press force setpoint: 180 kN, length-specific press force setpoint 9 kN / cm). The compacted ribbons were crushed in a shredder (Alexanderwerk type VZ 200, rotation speed 138 rpm) and then granulated in a granulator (Alexanderwerk type RFG 250 DL, equipped with a screen with a mesh size of 800 pim). A vibrating screen (oversize size: 1000 pim, undersize size: 100 pim) with ball sieves (diameter 22 mm) was used to isolate a compacted material with a particle size mostly between 100 pim and 1000 pim. The oversize and undersize particles from the screen were continuously recycled to the compactor. For tabletting, an additional 2.5 wt.% of graphite (grade 3160 from Asbury Graphite Mills Inc., New Jersey 08802) was added to the compact within 2 minutes in a Drais turbulent mixer. The produced granules were then milled using a Korsch XT-600 rotary runner (with 65 EURO B dies) under a dry air atmosphere to form ring-like precursor bodies of geometry A x H xl = 5 mm x 5 mm x 2 mm with a non-curved (i.e. planar) end face of a mass (M v ) of 185 mg. The applied pre-compression force was 0.7 kN, the applied main compression force was 2.7 kN. The rotor speed was 40 rpm. The lateral crushing strength (SDFv) of the resulting ring-like precursor bodies was 22 N. g) Thermal pretreatment and calcination of the ring-like precursor bodies produced in f) The formed rings were placed on the belt of a belt calciner (as described in WO 2002 / 024620) with eight chambers (chamber width 100 cm, chamber length 150 cm). Chambers 1 to 8 each had a fan to generate air circulation and were thermostatted at 150°C, 190°C, 225°C, 380°C, 430°C, 520°C, and 520°C, respectively. Heated air was metered into each chamber. The amount of supply air to chambers 1 to 8 was 130 Nm each. 3 / h, 190 Nm 3 / h, 254 Nm 3 / h, 90 Nm 3 / h, 150 Nm 3 / h, 90 Nm 3 / h, 90 Nm 3 / h and 140 Nm 3 / h. Exhaust air was removed from each chamber using a fan. The amount of exhaust air to chambers 1 to 8 was 136 Nm each. 3 / h, 300 Nm 3 / h, 315 Nm 3 / h, 154 Nm 3 / h, 145 Nm 3 / h, 90 Nm 3 / h, 90 Nm 3 / h and 137 Nm 3 / h. The bed height of chambers 1 to 4 was 40 mm. The bed height of chambers 5 to 8 was 75 mm. Within the chambers, the temporal and spatial temperature deviation from the setpoint was always < 10°C. The belt speeds were such that the residence time in the first four chambers was 105 minutes each and in the fifth to eighth chambers was 270 minutes each. In this way, 2.6 l of full catalyst bodies K1 were produced. Preparation of the full catalyst precursor body K2, wherein the active multimetal oxide had the stoichiometry M012 Bio.6 Fea C075 Ko 14 Ox. The stirred vessels used were filled in the presence of ambient air. During stirring / mixing, they were sealed hermetically with a lid that had a pressure relief valve to the atmosphere (1 atm-abs, 1.01 bar-abs). a) Preparation of an aqueous solution B 3.9 kg of demineralized water were placed in a temperature-controlled cylindrical stirred vessel (10 1 internal volume) made of stainless steel (EN 1.4541) equipped with an anchor stirrer and heated to 60°C while stirring (150 rpm). Subsequently, 27.5 g of a 47 wt% aqueous potassium hydroxide solution (47 wt% KOH) having a temperature of 20°C were added while stirring continuously and maintaining the temperature at 60°C. While maintaining the temperature at 60°C, 1161 g of fine-grained ammonium heptamolybdate tetrahydrate (54.3 wt.% Mo, supplier: NIPPON INORGANIC COLOUR & CHEMICAL CO., LTD., 3-14-1 Funado ltabashi-ku Tokyo 174-0041 (JP); it had a turbidity of 219 NTU as determined according to WO 2016 / 147324) was stirred in portions at a temperature of 25°C, and the resulting aqueous solution (slightly turbid due to minimal insoluble isopolymolybdate impurities) was stirred at 60°C for 20 min (150 rpm). b) Preparation of an aqueous solution A 5763 g of an aqueous nitric acid cobalt(II) nitrate solution (12.6 wt% Co, 27 wt% nitrate (NOy), prepared by dissolving cobalt metal from MFT Metals % Ferro-Alloys Trading GmbH, D-41474 Viersen, purity > 99.6 wt% Co, < 0.3 wt% Ni, < 100 mg / kg Fe, < 50 mg / kg Cu ​​in aqueous nitric acid) were placed in a temperature-controlled cylindrical stirred vessel (5 l internal volume) made of stainless steel (EN 1.4541) equipped with an anchor stirrer and heated to 60°C while stirring (150 rpm). With continued stirring (150 rpm) and continued heating to 60°C, 1994 g of a 60°C warm iron(II) nitrate nonahydrate melt (13.8 wt% Fe, <0.4 wt% alkali metals, <0.01 wt% chloride, <0.02 wt% sulfate, from Dr. Paul Lohmann GmbH, D-81857 Emmerthal) were added and stirred for a further 10 min at 60°C (150 rpm). 618.6 g of a 60°C warm, aqueous, nitric acid bismuth nitrate solution (11.1 wt% Bi, 13 wt%-% Nitrate (NO. 3 ), prepared by dissolving bismuth metal from Sidech SA, BE-1495 Tilly, purity > 99.997 wt.% Bi, < 7 mg / kg Pb, < 5 mg / kg each of Ni, Ag, Fe, < 3 mg / kg each of Cu, Sb and < 1 mg / kg each of Cd, Zn in aqueous nitric acid) and then stirred for a further 10 min at 60°C (150 rpm). c) Mixing of aqueous solution A with aqueous solution B to obtain an aqueous mixture M The 60°C warm aqueous solution A was metered continuously over a period of 15 minutes using a peristaltic pump (type: BVP, company: Ismatec SA, Labortechnik-Analytik, Feldeggstrasse 6, CH-8152 Glattbrugg, setting: 320 scale divisions) into the aqueous solution B, which was now being intensively stirred at 60°C using an Ultra-Turrax (company: Janke & Kunkel GmbH & Co. KG - IKA-Labortechnik, Janke & Kunkel-Str. 10, DE-79219 Staufen, shaft type: 550KR-G45 fine, shaft tube diameter: 25 mm, stator diameter: 45 mm, rotor diameter: 40 mm, setting: level 5). The aqueous solution A was added at the height of the rotor of the Ultra-Turax stirrer, offset by approximately 0.5 to 1 cm from the outer edge of the rotor of the Ultra-Turax stirrer. The resulting aqueous suspension was stirred for a further 15 min at 60 °C. d) Spray drying of the aqueous mixture M The aqueous mixture M was spray-dried immediately after its preparation. The aqueous mixture (suspension) M, which was continuously stirred (also during spray-drying) at 60 °C using an anchor stirrer (150 rpm), was spray-dried in a Mobile Minor™ 2000 (MM-I) spray tower from Niro A / S, Gladsaxevej 305, 2860 Soborg, Denmark, with a centrifugal atomizer type F01 A and an atomizer wheel type SL24-50 in a hot air cocurrent flow (gas inlet temperature: 350 ± 10 °C, gas outlet temperature: 140 ± 5 °C, throughput: 4.7 kg aqueous mixture M / h and a hot air flow with a current of 8.5 Nm 3 / h). The portion of the aqueous mixture M not yet spray-dried was continuously stirred at 60°C. The atomizer wheel speed was set at 25,000 rpm. The resulting spray powder was stored in hermetically sealed containers (200 l internal volume, 25°C, atmospheric pressure) until further processing (10 calendar days; shorter or longer intermediate storage (up to 30 calendar days) had no influence on the resulting results). The loss on ignition of the resulting spray powder (annealed for 3 hours at 600°C (powder temperature) under standing (excess) air) was 31 wt.% of its initial weight. e) Production of ring-like solid catalyst precursor bodies 3 kg spray powder, 3.5 wt.% graphite (grade 3160 from Asbury Graphite Mills Ino., New Jersey 08802), based on the mass of spray powder, and 3 wt.% glass fibers (glass fiber E-Glas type F2F 160 from Profill GmbH Heidelberg, Kichheimerstr. 1 D-69214 Eppelheim, nominal fiber size: 10 μm diameter, 200 μm length, BET surface area 0.4 m 2 / g) were mixed in a mixer equipped with a chopper (type L5 / 10 from Gebrüder Lödige Maschinenbau GmbH, Elsener Straße 7-9 33102 Paderborn, Germany) for 10 min, whereby the chopper blade was used intermittently for a total of 1 min. The powder mixture produced as described was then compacted (tableted) using a Kilian E 150+ rotary press (with 21 EURO D dies) under dry air to form ring-like precursor bodies with the dimensions A x H x I = 5 mm x 5 mm x 2 mm and a non-curved (i.e., planar) end face, a mass (Mv) of 160 mg (body density 1.94 g / ml). The applied pressing force (P v ) was 1.6 kN. The rotor speed was 15 - 20 rpm. The lateral crush strength (SDF V ) of the resulting ring-like multimetal oxide precursor bodies was not measured. f) Thermal pretreatment and calcination of the ring-like unsupported catalyst precursor bodies produced in e) 1000 g of the prepared full catalyst precursor bodies were evenly distributed on four grids arranged side by side, each with a square base area of ​​150 mm x 150 mm (piling height: 15 mm), in a forced-air shaft furnace (Nabertherm GmbH, D-28865 Lilienthal; furnace model S60 / 65A) through which 4500 Nl / h of previously dried air (which had an inlet temperature of 140 °C) flowed (the forced-air furnace was located in an environment with a temperature of 25 °C).Subsequently, while maintaining the air flow (including its inlet temperature), the temperature in the forced-air shaft furnace was varied as follows (the temperature specifications refer to the temperature in the respective applied bulk material; this was determined using four thermocouples, each located at the geometric center of the four grids in the center of the bulk material applied to the respective grid; one of the thermocouples provided the actual value for temperature control of the forced-air shaft furnace; the other thermocouples confirmed that the temperatures were identical within the interval ± 0.1 °C). The temperature increases occurred essentially linearly over time. Within 72 minutes, heating took place from 25 °C to 130 °C. This temperature was maintained for 72 minutes and then increased to 190 °C within 36 minutes. The 190 °C was maintained for 144 minutes before the temperature was increased to 220 °C within 36 minutes.The temperature was maintained at 220 °C for 72 min before being increased to 380 °C over a period of 93 min. The temperature was maintained at 380 °C for 187 min before being increased to 430 °C over a period of 93 min. The temperature was maintained at 430 °C for 187 min before being increased to the final calcination temperature of 480 °C over a period of 93 min. This temperature was maintained for 467 min. The temperature was then cooled to 25 °C over a period of 12 h. For this purpose, both the heating of the forced-air shaft kiln and the air flow preheater were switched off (the air flow of 4500 Nl / h was maintained, however; the air flow inlet temperature was then 25 °C). The obtained ring-like unsupported catalyst bodies K2 were stored in an airtight container at a temperature of 25°C and at atmospheric pressure until their use for catalyzing a heterogeneously catalyzed partial oxidation (for example, from propene to acrolein). Preparation of the full catalyst precursor body K3, wherein the active multimetal oxide had the stoichiometry M012 Bio.6 Fe24 C07.7 Ko os OX. The preparation of the fully formed catalyst body K3 was carried out in the same way as the preparation of the fully formed catalyst body K2. However, the added amounts of the sources of the elemental constituents other than Mo were adjusted to the initial stoichiometry M012 Bio 6 Fe24 Co77 Ko os Ox present in the aqueous mixture M. The amount of glass fiber in step e) was adjusted to 4.0 wt.% based on the amount of spray powder. The tableting conditions in step e) were as follows: M v = 150 mg (body density 1.83 g / ml) Pv = 0.8 kN SDFv = 14 N The final calcination temperature in step f) was adjusted to 470°C. Preparation of the full catalyst precursor body K4, wherein the active multimetal oxide had the stoichiometry M012 Bio.6 Fea C049 Ni26 Ko.12 Ox. The preparation of the fully formed catalyst body K4 was carried out analogously to the preparation of the fully formed catalyst body K2. However, the added amounts of the sources of the elemental constituents other than Mo were adjusted to the initial stoichiometry M012 Bio 6 Fe3 C049 Ni26 Ko 12 Ox present in the aqueous mixture M. The nickel component was introduced into solution A in production step b). A nitric acid solution of nickel(II) nitrate (13.8 wt% Ni, 0.6 wt% free nitric acid, prepared by dissolving nickel metal in aqueous nitric acid) was used as the nickel source. The nickel(II) nitrate solution was added 10 minutes after the addition of the bismuth nitrate solution. Solution A was stirred at 60 °C for 10 minutes after the addition of the nickel(II) nitrate solution. The production step e) for the formation of the complete catalyst precursor body was modified as follows: 3 kg of the resulting spray powder were mixed with 2 wt.% graphite (type 3160 from Asbury Graphite Mills Inc., New Jersey 08802), based on the weight of the spray powder, in a rotating drum mixer (drum diameter 650 mm) at 30 rpm for 30 minutes. The powder mixture produced as described was then compacted (tableted) using a Kilian E 150+ rotary press (with 21 EURO D dies) in a dry air atmosphere to form ring-like precursor bodies with a geometry of A x H x I = 5 mm x 3 mm x 2 mm and a non-curved (i.e., planar) end face with a mass of 102 mg. The applied press force was 2.0 kN. The resulting tablets were crushed on a vibrating screen crusher equipped with a sieve (mesh size 800 μm) and iron balls. The crushed granules were mixed with 1.5 wt.% of the same type 3160 graphite in a rotating drum mixer (30 rpm, 15 min.) to obtain granules. The resulting granules were then compacted (pressed into tablets) using a Kilian E150+ rotary mixer (with 21 EURO D dies) in a dry air atmosphere into ring-shaped precursor bodies with a geometry of A x H x I = 5 mm x 5 mm x 2 mm, a non-curved end face, and a mass (Mv) of 185 mg (body density 2.24 g / ml). The applied compression force was 3.4 kN. The lateral compressive strength (SDFv) of the resulting annular multimetal oxide precursor bodies was 25 N. The precursor bodies were calcined as in production step f) to produce K2, with the final calcination temperature set to 500°C. Preparation of the full catalyst precursor body K5, wherein the active multimetal oxide has the stoichiometry M012 Bio.6 Fe24 Co6 i Nii 6 KQ04 O X had. The preparation of the fully formed catalyst body K5 was carried out analogously to the preparation of the fully formed catalyst body K4. However, the added amounts of the sources of the elemental constituents other than Mo were adjusted to the initial stoichiometry M012 Bio 6 Fe24 C061 Nii.6 K004 Ox present in the aqueous mixture M. The tableting conditions in step e) were as follows: M v = 190 mg (body density 2.30 g / ml) Pv = 3.2 kN SDFv = 25 N The precursor bodies were calcined as in production step f) to produce K4, with the final calcination temperature set to 500°C. Preparation of the full catalyst precursor body K6, wherein the active multimetal oxide has the stoichiometry M012 Bii,4 Fe.8Co6.i Ni2.5 Koos O X had. The preparation of the fully formed catalyst body K6 was carried out analogously to the preparation of the fully formed catalyst body K4. However, the added amounts of the sources of the elemental constituents other than Mo were adjusted to the initial stoichiometry M012 Bii,4 Fe s C06.1 M2.5 Ko os Ox present in the aqueous mixture M. The tableting conditions in step e) were as follows: M v = 196 mg (body density 2.38 g / ml) Pv = 2.9 kN SDF V = 25 N The precursor bodies were calcined as in production step f) to produce K4, with the final calcination temperature set to 500°C. Preparation of the full catalyst precursor body K7, wherein the active multimetal oxide had the stoichiometry M012 Bio.6 Fea C049 Ni26 Ko.12 Ox. The preparation of the fully formed catalyst body K7 was carried out analogously to the preparation of the fully formed catalyst body K2. However, the added amounts of the sources of the elemental constituents other than Mo were adjusted to the initial stoichiometry M012 Bio 6 Fes C049 Ni26 Ko.12 Ox present in the aqueous mixture M. The nickel component was introduced into solution A in production step b). A nitric acid solution of nickel(II) nitrate (13.8 wt% Ni, 0.6 wt% free nitric acid, prepared by dissolving nickel metal in aqueous nitric acid) was used as the nickel source. The nickel(II) nitrate solution was added 10 minutes after the addition of the bismuth nitrate solution. Solution A was stirred at 60 °C for 10 minutes after the addition of the nickel(II) nitrate solution. The amount of glass fiber in step e) was adjusted to 0 wt%, i.e. no glass fibers were added. The tableting conditions in step e) were as follows: Mv = 150 mg (body density 1.82 g / ml) Pv = 1.2 kN The precursor bodies were calcined as in production step f) to produce K2, with the final calcination temperature set to 540°C. Preparation of the full catalyst precursor body K8, wherein the active multimetal oxide had the stoichiometry M012 Bio.6 Fe24 Co6 i Nii 6 Ko04 OX. The preparation of the fully formed catalyst body K8 was carried out analogously to the preparation of the fully formed catalyst body K7. However, the added amounts of the sources of the elemental constituents other than Mo were adjusted to the initial stoichiometry M012 Bio 6 Fe24 C06.1 Nii.6 K004 Ox present in the aqueous mixture M. The tableting conditions in step e) were as follows: Mv = 150 mg (body density 1.82 g / ml) Pv = 1.1 kN The precursor moldings were calcined as in production step f) to produce K2, with the final calcination temperature set to 510°C. Preparation of the full catalyst precursor body K9, wherein the active multimetal oxide had the stoichiometry M012 Bio.6 Fea C049 Ni26 Ko.12 Ox. The preparation of the fully formed catalyst body K9 was carried out analogously to the preparation of the fully formed catalyst body K7. However, the added amounts of the sources of the elemental constituents other than Mo were adjusted to the initial stoichiometry M012 Bio 6 Fe3 C049 Ni26 Ko.12 Ox present in the aqueous mixture M. The amount of glass fiber in step e) was adjusted to 3 wt% based on the amount of spray powder. The tableting conditions in step e) were as follows: Mv = 160 mg (body density 1.94 g / ml) Pv = 1.6 kN The precursor bodies were calcined as in production step f) to produce K2, with the final calcination temperature set to 500°C. Preparation of the full catalyst precursor body K10, wherein the active multimetal oxide had the stoichiometry M012 Bio 6 Fe24 C06.1 Nii 6 Ko 04 OX. The preparation of the fully formed catalyst body K10 was carried out analogously to the preparation of the fully formed catalyst body K9. However, the added amounts of the sources of the elemental constituents other than Mo were adjusted to the initial stoichiometry M012 Bio 6 Fe24 C061 Nii.6 K004 Ox present in the aqueous mixture M. The tableting conditions in step e) were as follows: Mv = 160 mg (body density 1.94 g / ml) Pv = 1.6 kN The precursor bodies were calcined as in production step f) to produce K2, with the final calcination temperature set to 500°C. All calcined catalysts K1 - K10 had a ring geometry A x H x I = 4.9 mm x 4.9 mm x 1.9 mm. The specific surface area (BET) of the calcined catalysts was determined by nitrogen adsorption at 77 K on a Micromeritics ASAP 2420. Prior to measurement, the samples were degassed at 200 °C for 15 hours under vacuum. The specific surface area A s was determined according to the multipoint Brunauer-Emmett-Teller method (BET) in the pressure range P / P0 = 0.05-0.20 assuming an N2 cross-sectional area of ​​16.2 Ä 2 calculated. The pore volume (PV) of the calcined catalysts was determined by mercury porosimetry at room temperature with a pressure range of 0.0034–420 MPa on a Micromeritics AutoPore V9600 (software: MicroActive 1.03.01). The pore size distribution was calculated from the intrusion volume as a function of the intrusion pressure, assuming a contact angle of 140°. 0 and a surface tension of 485 mN / m. Table 1 shows the physical properties of the obtained full catalyst bodies and the stoichiometry. Table 1 : BET = BET surface area PV = total pore volume of pores with a diameter of >0.03 to <300 pm 5 PV (0.01-0.1) = volumetric fraction of pores with a diameter of >0.01 to <0.1 pm in the total pore volume PV (0.1-1) = volumetric fraction of pores with a diameter of >0.1 to <1 pm in the total pore volume PV (1-10) = volumetric fraction of pores with a diameter of >1 to <10 pm in the total pore volume PV (10-300): = volumetric fraction of pores with a diameter of >10 to <300 pm in the total pore volume Catalysis experiments on the selective oxidation of propene to acrolein and acrylic acid A reaction tube (stainless steel type 1.4541 (EU standard number EN 10088-3); 33.7 mm outer diameter; 2 mm wall thickness; 29.7 mm inner diameter; 400 cm length, 4 mm thermal sleeve) was used for the test. The reaction tube was surrounded by two stirred and externally electrically heated salt baths (mixture of 53 wt.% potassium nitrate, 40 wt.% sodium nitrite and 7 wt.% sodium nitrate; 50 kg molten salt) over its length (the flow velocity at the tube was 3 m 3 / h (in the plane perpendicular to the longitudinal axis of the tube). The first salt bath covered a length of 0–191 cm, starting from the lower end of the reaction tube. The second salt bath, on the other hand, covered a length of 191–400 cm, starting from the lower end of the reaction tube. The temperatures of the salt baths were controlled independently, creating two temperature zones (Zone 1: lower part; Zone 2: upper part). The reaction tube was loaded from top to bottom as follows: Section 1 : 50 cm length Pre-filling of steatite rings with a geometry of 7 mm x 3 mm x 4 mm (outer diameter x length x inner diameter; Steatite C220 from CeramTec); Section 2: 300 cm length; Catalyst bed consisting of two layers (see Table 2; layer 1 is closest to the reactor inlet); Section 3: 20 cm length Refill from the same steatite rings as in section 1 ; Section 4: 30 cm length Empty pipe. The structure of the reaction zone is shown in Table 2. To adjust the relative volume-specific catalyst activity, solid catalyst bodies were diluted with inert steatite rings measuring 5 mm x 5 mm x 2 mm (outer diameter x length x inner diameter, C220 steatite from CeramTec). Table 2 shows the composition of the respective catalyst layers and the corresponding ratios Fe / Mo, Bi / Mo and K / Mo. The reactor feed gas consisted of a mixture of propene (technical or polymer grade), air, nitrogen, and water. The gas mixture was preheated to 200 °C. The concentration of propene and oxygen is shown in Table 4. The ratio of water vapor to propene was 0.3. The reactor feed gas was fed from the bottom to the top of the reactor tube (upflow). The temperature in the catalyst bed was continuously measured by a thermocouple placed in a thermowell located inside the reactor tube and moved from the bottom to the top of the reactor bed using a traction machine. The maximum temperature of this measurement corresponded to the hot spot temperature T H . In this document, the propene conversion (X (mol-%)) is understood as: Number of moles of propene converted X = x 100 mol%. Number of moles of propene dosed In this document, the yield of the desired product formation (Y (mol-%)) is defined as: Number of moles of propene converted to acrolein and acrylic acid Y = - x 100 mol-% Number of moles of propene dosed (the conversion figures are based on a single pass of the reaction gas mixture through the fixed catalyst bed). The catalytic performance data are shown in Table 3. Figures 1 to 5 show the temperature profiles of selected examples. Temperature T in the fixed catalyst bed, measured in the thermal sleeve, is plotted against the axial position P of the fixed catalyst bed in the direction from the gas inlet side to the gas outlet side. Figure 1 shows the temperature profile of Example E1. Figure 2 shows the temperature profile of Example E4. Figure 3 shows the temperature profile of Example E5. Figure 4 shows the temperature profile of Example E6. Figure 5 shows the temperature profile of Example E7. A comparison of the examples shows that at the same propene loading (130 or 180 Nl / l h) a significantly higher yield of valuable product (with lower catalyst mass) of the process according to the invention is achieved. Table 2: Filling of the reaction tubes z: Example C: Comparison example Table 3: Composition of the catalyst layers E: Example C: Comparison example Table 4: Reaction conditions / yield E: Example; C: Comparison example Propene load = quotient of propene flow rate (Nl / h) and empty pipe volume of the reaction zone (I) 5 Selectivity = Value product selectivity (acrolein + acrylic acid) Yield = yield of valuable product (acrolein + acrylic acid)

Claims

Patent claims 1. A process for the preparation of an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid, in which the corresponding ethylenically unsaturated aldehyde and / or the ethylenically unsaturated carboxylic acid is obtained by heterogeneously catalyzed gas-phase partial oxidation of at least one aldehyde with molecular oxygen at elevated temperature on a fixed catalyst bed of a tube bundle reactor, wherein in the axial direction of the tubes of the tube bundle reactor at least two catalyst layers, each comprising a catalytically active multielement oxide, are arranged such that a multilayer filling is achieved and a composition of the multielement oxide in one catalyst layer differs from a composition of the multielement oxide in at least one of the other catalyst layers, characterized in that a stoichiometric ratio Li of the elements Fe to Mo of the catalyst layer,in which the highest temperature occurs within the fixed catalyst bed is greater than a stoichiometric ratio l_2 of the elements Fe to Mo in the catalyst layer arranged furthest along the tube axis on the gas outlet side, and a stoichiometric ratio Mi of the elements Bi to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is less than or equal to a stoichiometric ratio M2 of the elements Bi to Mo in the catalyst layer arranged furthest along the tube axis on the gas outlet side.

2. Process according to claim 1, characterized in that the composition of the catalytically active multielement oxides corresponds to the general formula (I) Moi2Bi a FebCo c NidX e YfZgOn (I) with X = K, Cs and / or Rb, Y = Ca, Sr, Ba, Li, Na, Cr, W, Mn, Cu, Zn, Ga, P, B, As, Sn, Sb, Te, Nb, Ta, Pb, Ce and / or La, Z = Si, Al, Ti, Zr and / or Mg, where a = 0.2 to 2, b = 1 to 4, c = 3 to 9, d = 0 to 4, c + d = 4 to 9.5, e = 0.01 to 0.5, f = 0 to 10, g = 0 to 10, and n = a number determined by the valence and frequency of the elements other than oxygen in the general formula I.

3. Process according to claim 1 or 2, characterized in that a stoichiometric ratio Ni of the elements K to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is greater than a stoichiometric ratio N2 of the elements K to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side.

4. Process according to one of claims 1 to 3, characterized in that a stoichiometric ratio of the elements Fe to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is from 0.15 to 0.

35.

5. Process according to one of claims 1 to 4, characterized in that the composition of the catalytically active multielement oxide of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed corresponds to the general formula (II) Moi2Bi a Fe b Co c NidXeYfZgOn (II).

6. Process according to one of claims 1 to 4, characterized in that the composition of the catalytically active multielement oxide of the catalyst layer arranged furthest along the tube axis on the gas outlet side corresponds to the general formula (III) Moi2Bi a FebCo c NidXeYfZgO n (III).

7. Process according to one of claims 1 to 6, characterized in that the ethylenically unsaturated aldehyde is acrolein and the ethylenically unsaturated carboxylic acid is acrylic acid.

8. Process according to one of claims 1 to 7, characterized in that the alkene is propylene.

9. Process according to one of claims 1 to 8, characterized in that the catalyst layers are each formed from a solid catalyst and / or a coated catalyst.

10. Process according to one of claims 1 to 9, characterized in that the ratio of Li to l_2 is from 1.1 to 1.

6.

11. Process according to one of claims 1 to 10, characterized in that the ratio of Mi to M2 is from 0.5 to 1.

12. Process according to one of claims 3 to 11, characterized in that the ratio of Ni to N2 is from 1.5 to 3.

3.

13. Process according to one of claims 2 to 12, characterized in that in the general formula (I) X = K, Z = Si, a = 0.4 to 1.5, b = 1.7 to 3.1, c = 4.8 to 7.8, d = 0 to 2.7, c + d = 6.9 to 8.5, e = 0.03 to 0.15, f = 0, g = 0 to 1.

8.

14. Process according to one of claims 5 to 13, characterized in that in the general formula (II) X = K, Z = Si, a = 0.5 to 0.7, b = 2.9 to 3.1, c = 4.8 to 7.6, d = 0 to 2.7, c + d = 6.9 to 7.6, e = 0.07 to 0.15, f = 0, g = 0 to 1.

6.

15. Process according to one of claims 6 to 14, characterized in that in the general formula (III) X = K, Z = Si, a = 0.5 to 1.5, b = 1.7 to 3.1, c = 4.8 to 7.8, d = 0 to 2.7, c + d = 6.9 to 8.5, e = 0.03 to 0.09, f = 0, g = 0 to 1.

8.

16. A tube bundle reactor for producing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid by heterogeneously catalyzed gas-phase partial oxidation of at least one aldehyde with molecular oxygen at elevated temperature on a fixed catalyst bed of the tube bundle reactor, wherein in the axial direction of the tubes of the tube bundle reactor at least two catalyst layers, each comprising a catalytically active multielement oxide, are arranged such that a multilayer filling is present and a composition of the multielement oxide in one catalyst layer differs from a composition of the multielement oxide in at least one of the other catalyst layers, characterized in that a stoichiometric ratio Li of the elements Fe to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed,is greater than a stoichiometric ratio L2 of the elements Fe to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side, and a stoichiometric ratio Mi of the elements Bi to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is less than or equal to a stoichiometric ratio M2, of the elements Bi to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side.

17. A fixed catalyst bed arranged within at least one tube of a tube bundle reactor for producing an ethylenically unsaturated aldehyde and / or an ethylenically unsaturated carboxylic acid by heterogeneously catalyzed gas-phase partial oxidation of at least one aldehyde with molecular oxygen at elevated temperature, wherein in the axial direction of the tubes of the tube bundle reactor at least two catalyst layers, each comprising a catalytically active multielement oxide, are arranged such that a multilayer filling is present and a composition of the multielement oxide in one catalyst layer differs from a composition of the multielement oxide in at least one of the other catalyst layers, characterized in that a stoichiometric ratio Li of the elements Fe to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed,is greater than a stoichiometric ratio l_2 of the elements Fe to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side, and a stoichiometric ratio Mi of the elements Bi to Mo of the catalyst layer in which the highest temperature occurs within the fixed catalyst bed is less than or equal to a stoichiometric ratio M2 of the elements Bi to Mo in the catalyst layer which is arranged furthest along the tube axis on the gas outlet side.

Images