Process for production of formaldehyde from dimethyl ether

Abstract

Dimethyl ether is converted to formaldehyde using a supported catalyst comprising molybdenum and/or vanadium oxides. The surface density of the oxide(s) ranges from greater than that for the isolated monomeric oxides upwards, so long as there is a substantial absence of bulk crystalline molybdenum and/or vanadium oxide(s). Conversion and selectivity to formaldehyde are improved as compared to data reported for known catalysts. Also disclosed is a catalyst comprising molybdenum and/or vanadium oxides supported on a layer of stannic oxide that is disposed on the surface of a particulate alumina or zirconia support.

Description

FIELD AND BACKGROUND OF THE INVENTION

[0001] This invention relates to a process for production of formaldehyde, and optionally also methyl formate as a co-product, by oxidation of dimethyl ether (DME).

[0002] Formaldehyde is widely used as an intermediate or basic building block in the commercial synthesis of many chemicals. Because of the existence of large reserves of methane worldwide it has been considered desirable for some time to develop processes to convert methane to more valuable chemicals. One such effort has been in the area of direct conversion of methane to formaldehyde via selective oxidation. However, this has not been particularly successful; so far all such processes have resulted in low yields due to the tendency of the formaldehyde so produced being further oxidized to carbon oxides under the severe reaction conditions required for methane oxidation.

[0003] Instead of producing formaldehyde from direct oxidation of methane, formaldehyde is produced from that substance indirectly by first converting methane to synthesis gas (CO and H2), then reacting that to form methanol, which can then be oxidized to produce formaldehyde.

[0004] The oxidation of methanol to formaldehyde has been extensively studied, and is the dominant process today for formaldehyde synthesis, typically using silver- or iron/molybdenum-based catalysts.

[0005] Another possible route to formaldehyde involves the oxidation of dimethyl ether (CH3OCH3) via cleavage of the C—O—C linkages. This process, however, has not been widely studied.

[0006] Dimethyl ether is a generally environmentally benign molecule. Its physical properties resemble those of LPG (liquefied petroleum gases), and dimethyl ether thus can be transported within existing and developing LPG infrastructures. Like methanol, dimethyl ether can be produced from synthesis gas. These characteristics give it the potential to be a new, clean alternative fuel. This potential is expected to lead to the production of substantially larger quantities of dimethyl ether than in the past, thus making it available for use as an intermediate in production of other chemicals, including formaldehyde.

[0007] Several patents disclose processes for producing formaldehyde from dimethyl ether using various catalysts. U.S. Pat. No. 2,075,100 describes such a process using a number of comparatively mild oxidation catalysts including platinum wire or foil, palladium black, and metals such as gold, silver, and copper. Vanadium pentoxide and iron, chromium and uranium sesqui-oxides are termed “very suitable” . U.S. Pat. No. 3,655,771 describes using catalysts containing tungsten oxide, alone or optionally with no more than 10% of an additive. The additives mentioned include bismuth, selenium, molybdenum, vanadium, phosphorus and boron oxides, as well as phosphoric acid, ammonium phosphate and ammonium chloride.

[0008] More recently, U.S. Pat. No. 4,435,602 describes a process for production of formaldehyde from dimethyl ether using naturally occurring manganese nodules as a catalyst. U.S. Pat. No. 4,439,624 describes such a process using an intimate mixture of bismuth, molybdenum and copper oxides, preferably prepared by coprecipitation. U.S. Pat. No. 4,442,307 describes such a process using an intimate mixture of bismuth, molybdenum and iron oxides, similarly prepared. U.S. Pat. No. 6,256,528 describes oxidation of dimethyl ether with a catalyst containing metallic silver to produce a mixture of products including formaldehyde, light alkanes, carbon oxides and water. Information in these patents indicates that formaldehyde was produced with reasonable yields, but that overoxidation of that product to carbon oxides occurred to an undesirable degree.

[0009] As described above, it would be advantageous to provide a process for production of formaldehyde from dimethyl ether with good conversion and good selectivity to formaldehyde. Preferably such a process could be operated without the occurrence of substantial direct oxidation of dimethyl ether to carbon oxides or further oxidation of product formaldehyde to carbon oxides, thus improving the chemical and energy efficiency of the process.

BRIEF SUMMARY OF THE INVENTION

[0010] In brief, this invention comprises a process for the production of formaldehyde by oxidation of dimethyl ether in the presence of a supported catalyst comprising molybdenum oxide, vanadium oxide or a mixture of molybdenum and vanadium oxides. The support is one that substantially does not react with the molybdenum or vanadium oxide to form unreducible mixed oxide(s). Preferred supports are alumina, zirconia, stannic oxide, titania, silica, and mixtures thereof.

[0011] In one preferred embodiment this invention comprises such a process in which the molybdenum and/or vanadium oxides are dispersed on the surface of the support, the surface density of the oxide or oxides on the support is greater than that for the isolated monomeric oxide or oxides, and in which the catalyst is characterized by a substantial absence of bulk crystalline molybdenum and/or vanadium oxides.

[0012] Most preferably the surface density of the molybdenum and/or vanadium oxide or oxides on the support is approximately that of a monolayer of the oxide or oxides at the surface of the support.

[0013] In another preferred embodiment, the catalyst comprises stannic oxide. More preferably in this embodiment, the catalyst comprises a layer of stannic oxide on a particulate support (preferably alumina and/or zirconia) with the molybdenum and/or vanadium being present as an upper layer on the stannic oxide layer. In this embodiment, preferably the surface density of the molybdenum and/or vanadium oxide or oxides on the support is greater than that for the isolated monomeric oxide or oxides, and the catalyst is characterized by a substantial absence of bulk crystalline molybdenum and/or vanadium oxides. Most preferably the surface density of the molybdenum and/or vanadium oxide or oxides on the support is approximately that of a monolayer of the oxide or oxides at the surface of the support.

[0014] Catalysts of the above type in which the molybdenum and/or vanadium oxide is supported on a layer of stannic oxide, the stannic oxide layer being disposed on a particulate alumina and/or zirconia, are novel and form another feature of this invention.

DETAILED DESCRIPTION OF THE INVENTION

[0015] In brief, this invention comprises a process for the production of formaldehyde by oxidation of dimethyl ether in the presence of a supported catalyst comprising molybdenum oxide, vanadium oxide or a mixture of molybdenum and vanadium oxides. Preferably the oxides are supported on alumina, zirconia, stannic oxide (SnO2), titania (TiO2), silica (SiO2), or a mixture of two or more of these oxides. Preferably, the molybdenum and/or vanadium oxides are dispersed on the surface of the support, the surface density of the oxide or oxides on the support is greater than that for the isolated monomeric oxide or oxides, and the catalyst is characterized by a substantial absence of bulk crystalline molybdenum or vanadium oxides.

[0016] Catalysts of this type that comprise molybdenum or vanadium oxides supported on alumina or zirconia are described in several prior publications, for catalyzing the oxidative dehydrogenation of propane to propene. These include Chen, et al., in “Studies in Surface Science and Catalysis”, Vol. 136, pp. 507-512, J. J. Spivey, E. Iglesia and T. M. Fleisch, Ed. (Elsevier Science, B.V., 2001); Chen et al., J. Catalysis 189, 421 (2000), Khodakov et al., J. Catalysis 177, 343 (1998), Chen et al., J. Catalysis 198, 232 (2001) and Chen et al., J. Phys. Chem. B201, 105 (2001). These publications are hereby incorporated herein by reference. However, these publications do not disclose catalysts containing stannic oxide, titania, silica, or other supports.

[0017] In the catalysts of this invention, the molybdenum and/or vanadium oxide is distributed on the surface of the support in what is known as “small domain” distribution. The surface density of the oxide catalyst on the support (measured in units of Mo or V metal atoms per nm2) is chosen so as to be greater than the surface density of the respective isolated monomeric oxide or oxides, but the catalyst overall is characterized by a substantial absence of bulk crystalline molybdenum and/or vanadium oxides (corresponding to the oxide or oxides used in preparing the catalyst). By “bulk crystalline oxides” is meant oxide(s) having a clear X-ray diffraction pattern. The crystallinity can be determined by X-ray diffraction based on the peak intensity ratio between one of the peaks of the supported metal oxide and one of the peaks of the support employed after calibration with a mixture of known amounts of the metal oxide and the support. By “substantial absence” is meant that the supported catalyst contains less than about 5% of bulk crystalline molybdenum and/or vanadium oxide(s).

[0018] Surface densities of the catalysts in this invention are given in terms of nominal surface density. This value is calculated based on the elemental analysis of the molybenum and/or vanadium oxide and on the surface area of the support, i.e., by dividing the number of metal atoms of the catalytic metal (Mo or V) in a given mass of sample by the surface area of the support (calculated from N2 absorption at its normal boiling point using the Brunauer-Emmett-Teller, or BET, equation). Where the metal oxide does not appreciably interact with the support to form a complex (see below), this calculated surface density fairly closely conforms to the actual surface density of the metal atoms on the surface of the support. However, where an appreciable amount of a complex is formed between the metal oxide and the support, the (nominal) surface density represents what that value would be were a complex not formed.

[0019] The surface density of the catalyst affects the catalyst efficiency. At one extreme, catalysts of this type with relatively isolated oxide species, for example monomolybdate or monovanadate species, have relatively few active sites on the support surface. These catalysts tend to retain their oxygen and thus provide rather low reaction rates for the oxidation of dimethyl ether to formaldehyde. At the other extreme, catalysts having bulk crystals may possibly provide reasonable reaction rates per unit surface area. However, they also lack efficiency in the utilization of the oxide catalyst because a substantial amount of the oxide is located within the crystals and is thus not available for catalyzing the reaction. Bulk MoO3 crystals also tend to be nonselective in their functioning, and can cause overreaction to produce carbon oxides rather than the desired products formaldehyde and methyl formate.

[0020] It has been found that the most preferred catalysts for this reaction tend to have a surface density of approximately a monolayer of catalyst on the support. The monolayer surface density depends primarily on the oxide chosen. For molybdenum oxide, the monolayer surface density is˜5.0 Mo atoms per square nanometer of support (Xie et al, Adv. Catal., 37, 1 (1990)). For vanadium oxide, this value is approximately 7.5 V atoms per square nanometer (Centi, Appl. Catal. A, 147, 267 (1996)). The term “monolayer” as used herein is meant to refer to these approximate surface densities. If the catalyst is uniformly dispersed on the support, satisfactory results are obtainable with a preferred surface density of from approximately 50-300% of the monolayer capacity values for alumina supports, and approximately 50-400% of these values for zirconia supports. Overall, a preferred range of surface densities is from about 50 to about 300% of the monolayer capacity, for both molybdenum and vanadium oxides, for all supports usable in this invention.

[0021] The molybdenum or vanadium oxide may be present as the oxide per se, represented by the general formulas MoOx and VOy, where x and y represent general values for oxygen in such molecules. For MoOx, the oxide generally comprises about three oxygen atoms per molybdenum atom; i.e., the general form of the oxide may be represented as MoO3, or molybdenum trioxide. For VOy, the oxide generally comprises about five oxygen atoms per two vanadium atoms, represented by the general formula V2O5, or vanadium pentoxide. However, in a given case the oxide may have an oxygen-to-metal atomic ratio that is not necessarily exactly 3:1 for molybdenum oxides or 5:2 for vanadium oxides.

[0022] In addition, the molybdenum or vanadium oxides may form one or more complexes or compounds with the support. These complexes usually also will be an oxide such as polymolybdates and/or polyvanadates. Such molybdenum complexes may have general formulas such as ZrMo2O8. Vanadium complexes would generally be represented by the formula M2xV2yO(nx+5y) where M is the cationic ion of the support and n is the oxidation state of M, e.g., ZrV2O7. In any case, such complexes of molybdenum and vanadium oxides with the support are considered to be within the definition of the oxide catalysts to which this invention pertains.

[0023] For instance, with molybdenum oxide supported on zirconia, as seen in the examples below, where the Mo surface density is below 6.4 Mo/nm2 the ZrO2 surface is covered predominantly by two-dimensional polymolybdates (irrespective of the temperature of preparation), and the MoOx domain size increases with increasing the Mo surface density. At Mo surface densities above 6.4 Mo/nm2, increase in the Mo surface density leads to the preferential formation of MoO3 or ZrMo2O8 crystallites on the ZrO2 surface after treatment in air at 723 and 773 K or at 873 K, respectively. This makes a fraction of the Mo active centers inaccessible to dimethyl ether reactions and thus, as described below, leads to a monotonic decrease in the primary dimethyl ether reaction rates with increasing Mo surface density (>6.4 Mo/nm2).

[0024] For such samples where the surface density was greater than 6.4 Mo/nm2, the areal dimethyl ether reaction rates (per surface area) and primary selectivities approached constant values as the Mo surface density increased. This indicates that the MoO3 or ZrMo2O8 domains at the ZrO2 surface do not change in their local structure or surface properties, while their domain size grows with increasing the Mo surface density. The surface density of 6.4 Mo/nm2 exceeds the theoretical polymolybdate monolayer, which is about 5.0 Mo/nm2 Nevertheless, the catalyst sample with a surface density of 6.4 Mo/nm2 exhibited the highest dimethyl ether reaction rates among the zirconia-supported molybdenum catalyst samples. This appears to be a compromise between reactivity and accessibility of the MoOx sites. The samples having a ZrMo2O8 structure possess a higher reactivity compared to the samples having polymolybdates and MoO3 crystallines at a given Mo surface density, which is believed to be a result of the higher reducibility of the ZrMo2O8 species. The reducibility of the MoOx domains (characterized by a H2 temperature-programmed reduction method) was found also to be dependent on the domain size and structures of the MoOx species. The larger MoOx domains undergo faster reduction compared to the smaller ones, and ZrMo2O8 domains are more reducible than two-dimensional polymolybdate and MoO3 domains at a given Mo surface density, reflecting the difference in the ability of these species to delocalize charge.

[0025] The support may be selected among commonly used supports for such oxide catalysts, including mixtures of such supports, provided it allows or favors the formation of a monolayer of molybdenum and/or vanadium oxide on the surface of the support and otherwise is suitable for use in the production of formaldehyde from dimethyl ether. Some properties may make certain supports unsuitable for use in the process of this invention. For instance, supports that will react with the molybdenum and/or vanadium oxide to form any significant amounts of unreducible mixed bulk oxides, i.e. oxides that would undergo substantial formation of oxygen vacancies at temperatures below about 300-400° C., would in general not be suitable for use in this process. One commonly used catalyst support, magnesium oxide, for example, was tested for suitability in this process and was found unsuitable. Supports that could cause undesired combustion of products to form carbon oxides under the operating conditions of this process, or that contain acid sites that could cause formation of excessive amounts of methanol under the conditions of this process also would not be suitable for use in this invention.

[0026] Preferred supports include alumina, zirconia, stannic oxide, titania, silica, and mixtures of two or more of these oxides. Most preferred are alumina, zirconia and stannic oxide, and mixtures of two or all three of them. The supports that are suitable for use in this process may be used in any of their available forms, including forms that as of the present time might not yet have been developed, or may have been developed but have not yet been commercialized. Both high and low surface area supports may be used, including materials known by the acronym MCM (standing for Mobil Compositions of Matter), e.g., MCM-41. These are recently developed mesoporous materials (often comprising silica) and are described in Kresge, et al., (Nature, 359, 710 (1992)) and by Corma (Chem. Rev., 97, 2373 (1997)). High surface area supports of various physical types are preferred for use in the invention from the point of view of efficiency in that they may produce greater amounts of product per unit mass of overall catalyst.

[0027] Novel catalysts of this invention include those in which the support comprises a layer of stannic oxide disposed on a particulate alumina and/or zirconia, or a layer of zirconia disposed on a particulate alumina, particularly those in which the layer of stannic oxide or zirconia has a surface density close to that of a monolayer of that substance. So, for example, exemplary catalysts may comprise molybdenum and/or vanadium oxide on a near-monolayer of stannic oxide disposed on a particulate (preferably high surface area) alumina.

[0028] The catalysts of the invention are prepared by typical means, for instance by impregnation, particularly incipient wetness impregnation, of the support with an aqueous solution containing molybdenum and/or vanadium, e.g. using a salt such as an ammonium molybdate or vanadate, for instance, ammonium di- or heptamolybdate or ammonium metavanadate. The preparation is carried out so as to disperse the molybdenum and/or vanadium oxide over the surface of the support and the amounts are chosen so as to achieve a desired surface density. Preparation of such catalysts by incipient wetness impregnation is described in the Chen et al. and Khodakov et al. publications mentioned above.

[0029] The primary products of the reaction are formaldehyde and methyl formate. Production of methyl formate can be increased if desired, by decreasing the surface density of metal oxide or choosing a specific support such as stannic oxide and/or zirconia, or it may be decreased (which is generally preferred since formaldehyde is typically the preferred product) by providing a catalyst having a surface density close to the value for a monolayer of catalyst, which, as will be shown below, generally has the highest selectivity to formaldehyde of the catalysts of the invention. However, production of methyl formate is to be expected in such a process, and is not especially detrimental as methyl formate has uses of its own as a chemical intermediate and can readily be separated from the reaction products and forwarded to other process units for such uses.

[0030] Methanol is also produced in processes of this type, but it dehydrates relatively readily to re-form dimethyl ether; the methanol produced can be recovered and recycled. Alternatively, methanol produced by this reaction may be forwarded to another unit, either for production of further formaldehyde using a typical catalyst for that process, or for other uses as a chemical intermediate. Methanol formation therefore can be essentially disregarded in calculating selectivity of the dimethyl ether to formaldehyde.

[0031] As compared with data in patents mentioned above, the process of this invention exhibits both improved conversions of dimethyl ether as well as improved selectivity to formaldehyde, and can achieve these results at lower temperatures. The process of this invention may be operated in general at temperatures of from about 150 to about 400° C., preferably from about 180 to about 350° C., most preferably from about 150 to about 320° C. Operating pressures are about 0.1-100 atm, preferably about 1-20 atm. Residence time generally ranges from about 1 to about 60 seconds.

EXAMPLES

[0032] The following are representative examples of this invention. These examples are provided by way of illustration only and not by way of limitation. Those of skill will readily recognize a variety of noncritical parameters that could be changed or modified to yield essentially similar results.

Example 1:

[0033] Production of Formaldehyde from Dimethyl Ether Using Molybdenum Catalysts Supported on Alumina, Zirconia and Stannic Oxide

[0034] Supported MoOx catalysts were prepared by incipient wetness impregnation of ZrO(OH)2, SnO2, or Al2O3, respectively, with aqueous (NH4)2Mo2O8 (99%, Aldrich) solutions (13-15). ZrO(OH)2 was prepared by hydrolysis of aqueous zirconyl chloride solutions (>98%, Aldrich) using NH4OH (14.8 N), followed by drying in ambient air at 393 K overnight. SnO2 was prepared by hydrolysis of tin (IV) chloride pentahydrate (98%, Alfa Aesar) with NH4OH (14.8 N), followed by treatment in flowing dry air at 773 K for 3 h. A commercial source of &ggr;-Al2O3 (Degussa AG) was used without further treatment. All samples were dried after impregnation at 373 K in ambient air and then treated in flowing dry air at 773 K for 3 h. Bulk MoO3 powders were prepared by decomposition of (NH4)2Mo2O8 (99%, Aldrich) in flowing dry air at 773 K for 3 h.

[0035] Dimethyl ether reactions were carried out in a fixed-bed quartz microreactor using catalysts (0.3 g) diluted with quartz powder (1 g) in order to prevent local high temperatures caused by the exothermic nature of the reaction. The reactant mixture consisted of 80 kPa DME and 18 kPa O2, and 2 kPa N2 was used as an internal standard. Reactants and products were analyzed by on-line gas chromatography (Hewlett-Packard 6890 GC) using flame ionization and thermal conductivity detectors and methyl-silicone capillary and Porapak® Q packed columns.

[0036] Table 1 shows catalytic results obtained at 513 K on MoOx domains supported on Al2O3, ZrO2 and SnO2 with similar Mo surface densities (6.4-7.0 Mo-atoms/nm2), alone and compared with results reported in previous patents. Rates and selectivities (in all tables) were measured as a function of DME conversion, which was changed by varying the reactant residence time. DME conversion rates and formaldehyde selectivities were extrapolated to zero reactant residence time in order to obtain primary DME conversion rates and selectivities. DME conversion rates and selectivities are reported in two forms in the results shown in Table 1. One approach considers CH3OH as a DME conversion product; the other approach reports rates and selectivities on a CH3OH-free basis, which seems appropriate in view of the reversible nature of DME conversion to methanol and the pathways available for the ultimate conversion of both CH3OH and DME to HCHO.

[0037] Primary reaction rates (normalized by catalyst mass) were much higher on the three catalysts of the invention than on catalysts previously reported in the patents, even at the lower temperatures used here. Rates were higher on SnO2 and ZrO2 supports than on Al2O3, but the primary formaldehyde selectivity (CH3OH-free) was almost 100% on MoOx/Al2O3. Pure supports showed very low DME conversion rates. A MoO3 sample with relatively low surface area gave a low DME conversion rate (per gram), but its areal rate resembled those on MoOx/Al2O3 and was 2-6 times lower than on MoOx supported on ZrO2 and SnO2. Thus, DME conversion appears to require small MoOx domains with much greater accessibility to reactants than those available in bulk MoO3 samples. Formaldehyde was not detected on MoO3 at 513 K, because of the low DME conversions attained. At higher temperatures (593 K), the primary HCHO selectivity was 52.9% (on a CH3OH-free basis) on bulk MoO3. MoOx/Al2O3 was the most selective catalyst for DME conversion to HCHO. Its primary HCHO selectivity was 79.1% (98.1%, CH3OH-free basis) and CO and CO2 (COx) were not detected as primary products.

[0038] Included in these tests was a catalyst similarly prepared comprising molybdenum oxide supported on magnesium oxide. MgO was prepared by contacting MgO (>98%, Aldrich) with deionized water at 355-365 K for 4 h, followed by treatment in flowing dry air at 773 K for 8 h. As seen from Table 1, however, MoOx domains supported on MgO did not give detectable DME conversion rates, apparently as a result of the formation of mixed metal oxides, which are unable to undergo reduction-oxidation cycles required for catalytic DME conversion turnovers at these temperatures. This support thus appears unsuitable for use with the catalysts in this process. 1 TABLE 1 DME oxidation rates and selectivities on supported MoOx catalysts at 513 K (80.0 kPa, 18 kPa O2 and 2 kPa N2), on bulk MoO3, on pure supports, and on previously reported catalysts. DIMETHYL DME DME Mo Tem- ETHER Conversion Conversion Surface durface pera- Conversion Rated Rated Catalyst area density ture Rated (mol/g- (10−5 Selectivity (%)d Ref- (MoO3 wt %) (m2/g) (Mo/nm2) (K) (mmol/g−cat-h) atom Mo-h) mol/m2-h) CH3OH HCHO MFa DMMb COxc erence MoO3/ZrO2 136.3 6.4 513 17.6 12.2 12.9 22.6 53.4 16.8 0.3 6.9 This (20.7%) (13.6) (9.4) (69.0) (21.7) (0.4) (8.9) study MoO3/SnO2 46.5 6.5 513 18.4 36.5 39.6 13.1 67.3 9.6 0 9.8 This (7.2%) (16.0) (31.7) (77.6) (11.1) (0) (11.3) study MoO3/Al2O3 90.0 7.0 513 5.8 5.7 6.6 19.7 79.9 1.6 0 0 This (15.0%) (4.7) (4.6) (98.1) (1.9) (0) (0) study MoO3/MgO 171.2 5.8 513 not detected not detected — — — — — — This (24.0%) study ZrO2 106.4 — 513 no detected — — — — — — — This study SnO2 48.8 — 513 0.6 — 1.2 29.7 0 0 0 70.3 This study Al2O3 110 — 513 0.2 — 0.2 82.9 0 0 0 17.1 This study MoO3 3.3 — 513 0.2 — 6.1 6.0 0 0 0 94.0 This study eAg — — 887 10.5 — — — 45.8 0 0 30.7 10 fBi—Mo—FeOx — — 773 3.1 — — 0 46.0 0 0 54.0  7 gBi—No—CuOx — — 773 2.9 — — 0 43.0 0 0 55.0  8 hMn nodules ˜230 — 623 1.7 — 0.7 — 49.0 — — —  9 aMF: Methyl formamate; bDMM: dimethoxymethane; cCOx: CO + CO2; dThe data in parentheses are reported on a CH3OH-free basis by excluding CH3OH as a reaction product in view of its reactivity in formaldehyde synthesis and its tendency to dehydrate to re-form CH3OCH3. efrom U.S. Pat. No. 6,265,528; Reactants: 59.9 kPa DME, 8.5 kPa O2, 31.6 kPa N2. ffrom U.S. Pat6. No. 4,442,307; Reactants: 5.0 kPa DME, 20.0 kPa O2, 75.0 kPa N2. gfrom U.S. Pat. No. 4,439,624; reactants same as (f); hfrom U.S. Pat. No. 4,435,602; reactants same as (f).

Example 2:

[0039] Additional Tests with Molybdenum Oxide Catalysts Supported on Alumina

[0040] A parallel study showed that the catalytic properties of these MoOx domains depend sensitively on their size and local structure, which were varied by changing the Mo surface density on A12O3 (1.6-11.3 Mo/mn2) and ZrO2 (2.2-30.6 Mo/nm2). Primary DME reaction rates increased from 2.3 to 5.7 mol/g-atom Mo-h as the Mo surface density increased from 1.6 to 7.0 Mo/Mn2 on Al2O3 (Table 2). These rates increased from 0.6 to 12.2 mol/g-atom Mo-h as the Mo surface density increased from 2.2 to 6.4 Mo/nm2 on ZrO2. On both ZrO2 and Al2O3, DME conversion rates per Mo reached maximum values at surface densities of 6-7 Mo/nm2. Results are reported in Table 2.

[0041] X-Ray diffraction and Raman, UV-visible, and X-ray absorption spectroscopies did not detect MoO3 crystallites in samples with surface densities below 7 Mo/nm2. In this Mo surface density range, most, if not all, MoOx species are accessible at surfaces and the DME conversion rates per Mo atom are equivalent to rates per exposed MoOx moiety (i.e. turnover rates). Therefore, the higher reaction rates attained with increasing surface density reflect a higher reactivity of exposed MoOx as the size and dimensionality of MoOx domains increases with increasing Mo surface density. The larger domains formed at higher MoOx surface densities (detected by measurements of their absorption edge energy in the UV-visible spectra) appear to undergo the redox cycles required for DME conversion to HCHO with greater ease than isolated monomolybdate species or smaller two-dimensional polymolybdate domains.

[0042] This interpretation is consistent with the observed decrease in the temperature required for H2 reduction of Mo6+ to Mo4+ in these samples. Ultimately, DME conversion rates (per Mo) decreased at even higher Mo surface densities (>10 Mo/nm2), because the incipient formation of three-dimensional MoO3 clusters leads to increasingly inaccessible MoOx species.

[0043] Primary formaldehyde selectivities also increased monotonically with increasing Mo surface density; they reached 79.1% (98.1% on a CH3OH-free basis) on MoOx/Al2O3 at 11.3 Mo/nm2. Methanol selectivities decreased as the Al2O3 support was covered with MoOx species, and the primary formaldehyde selectivity concurrently increased with increasing Mo surface density. Methyl formate and COx primary selectivities were very low on all Al2O3-supported MoOx samples. On Al2O3-supported samples with surface densities of 1.6 to 11.3 Mo/nm2, the CH3OH-free primary HCHO selectivity was 95.2-98.1% (Table 2). 2 TABLE 2 Effects of surface density of MoOx/Al2O3 catalysts on primary DME reaction rates and primary products at 513 K (80.0 kPa, 18 kPa O2 and 2 kPa N2). MoO3 loading Mo surface density Rated Rated Selecitivyt (%)d (MoO3 wt %) (Mo/nm2) (mol/g-atom MO-h) (10−5 mol/m2-h) CH3OH HCHO MFa DMMb COxc 4.0% 1.6 2.3 0.6 34.1 62.8 2.5 0 0.8 (1.5)d (0.4)d (95.2) (3.7) (0) (1.2) 8.0% 3.4 3.90 2.2 27.0 70.5 2.4 0 0.2 (2.8) (1.6) (96.5) (3.2) (0) (0.3) 10.0% 4.5 5.2 4.0 22.1 76.3 1.5 0 0.1 (4.1) (3.1) (97.9) (1.9) (0) (0.1) 15.0% 7.0 5.7 6.6 19.5 79.1 1.6 0 0 (4.6) (5.3) (98.1) (1.9) (0) (0) 20.0% 11.3 3.8 7.1 19.4 79.0 1.6 0.1 0 (3.1) (5.7) (98.0) (1.9) (0.1) (0) aMF: Methyl formate; bDMM: dimethoxymethane; cCOx: CO + CO2; dThe data in parenteses are reported on a CH3OH-free basis by excluding CH3OH as a reaction product in view of its reactivity in formaldehyde synthesis and its tendency to dehydrate to re-form CH3OCH3.

Example 3:

[0044] Effect of Temperature on Reaction Rates and Primary Products

[0045] Table 3 shows results of a study comparing reaction rates and selectivity at temperatures of 473-533 K (200-250° C.). The catalyst used contained 15 wt. % MoO3 on alumina with a surface density of 7.0 Mo/nm2. As the temperature was increased from 473 to 533 K, the reaction rate increased significantly, and selectivity to formaldehyde (as opposed to methyl formate) also increased significantly. 3 TABLE 3 Effects of temperature on primary DME reaction rates and primary products on MoOx/Al2O3 catalyst (15 wt. % MoO3; 7.0 Mo/nm2) (80.0 kPa, 18 kPa O2 and 2 kPa N2). Tem- pera- Rated Selectivity (%)d ture (mol/g-atom Mo-h) CH3OH HCHO MFa DMMb COxc 473 1.3 24.2 69.2 4.2 0.2 2.1 (1.0) (91.2) (5.5) (0.3) (2.8) 493 2.9 23.8 73.1 2.0 0 1.2 (2.2) (95.9) (2.6) (0) (1.6) 513 5.7 19.5 79.1 1.6 0 0 (4.6) (98.1) (1.9) (0) (0) 533 12.9 17.3 81.0 1.8 0 0 (10.7) (97.9) (2.1) (0) (0) aMF: Methyl formate; bDMM: dimethoxymethane; cCOx: CO + CO2; dThe data in parentheses are reported on a CH3OH-free basis by excluding CH3OH as a reaction product in view of its reactivity in formaldehyde synthesis and its tendency to dehydrate to re-form CH3OCH3.

Example 4:

[0046] Further Experiments with Molybdenum Oxide Catalysts Supported on Zirconia

[0047] A series of molybdenum oxide catalysts supported on zirconia, having a range of surface densities and prepared by calcining at two different temperatures was used to catalyze the production of formaldehyde from dimethyl ether.

[0048] The catalysts were prepared by incipient wetness impregnation of precipitated zirconium oxyhydroxide (ZrO(OH)2) with an aqueous solution of ammonium dimolybdate [(NH4)2Mo2O8] (99%, Aldrich) . Zirconium oxyhydroxide (ZrO(OH)2) was prepared by precipitation of a zirconyl chloride solution (98%, Aldrich) at a constant pH of 10 by controlling the rate of addition of ammonium hydroxide solution (29.8%, Fisher Sci.). After precipitation, the solids were washed with mildly basic ammonium hydroxide solution (pH˜6) until the effluent showed no chloride ions tested by a silver nitrate solution. The resulting solids were dried in air overnight at 393 K (120° C.). The dried solids were impregnated with an aqueous solution of ammonium dimolybdate at room temperature. The Mo6+ concentrations in the solution were varied in order to get desired Mo content in the final catalysts. After impregnation, samples were dried in air at 393 K and treated in dry air at 723, 773 or 873 K (450, 500 and 600° C.) for 3 h.

[0049] The catalysts were systematically characterized by means of powder X-ray diffraction (XRD), diffuse reflectance UV-visible, Raman, and X-ray absorption (XANES/XAFS) spectroscopies. Surface areas were measured by N2 physisorption using standard multipoint BET method. Mo surface density is expressed as the number of Mo atoms per square nanometer BET surface area (Mo/nm2). The catalysts so prepared are listed in Table 4. 4 TABLE 4 Surface areas and Mo surface density for MoOx/ZrO2 catalyst treated at 723, 773 and 873 K. 723 K 773 K 873 K MoO3 Loading Surface area Mo surface density Surface area Mo surface density Surface area Mo surface density Sample (wt. %) (m2/g) (Mo/nm2) (m2/g) (Mo/nm2) (m2/g) (Mo/nm2) 1MoOx/ZrO2  1.0% 118.8 0.3 105.6 0.4 85.6 0.5 6MoOx/ZrO2  5.7% 130.0 1.8 110.3 2.2 97.1 2.5 11MoOx/ZrO2 11.0% 145.9 3.2 132.6 3.5 103.4 4.5 21MoOx/ZrO2 20.7% 153.7 5.6 136.3 6.4 102.7 8.4 29MoOx/ZrO2 29.3% 114.0 10.7 96.9 12.6 64.6 20.0 37MoOx/ZrO2 37.0% 99.6 15.5 73.9 20.9 49.3 31.4 44MoOx/ZrO2   44% 83.5 22.0 60.2 30.6 36.7 50.1

[0050] Tables 5 and 6 show the results using catalysts so prepared and calcined at 773 K and 873 K, respectively. Both catalysts demonstrated very good dimethyl ether conversion rates and selectivity, with better performance being exhibited in general by the catalyst that had been calcined at 873 K. 5 TABLE 5 Effects of surface density of MoOx/ZrO2 catalysts (calcined at 773 K) on primary DME reaction rates and primary products at 513 K (80.0 kPa, 18 kPa O2 and 2 kPa N2). MoO3 loading Mo surface density Rated Rated Selectivity (%)d (MoO3 wt. %) (Mo/nm2) (mol/g-atom No-h) (105 mol/m2-h) CH3OH HCHo MFa DMMb COxc 11.0% 3.5 8.5 4.9 24.0 32.3 30.0 0 13.7 (6.5) (3.7) (42.5) (39.5) (0) (18.1) 20.7% 6.4 12.2 12.9 22.6 53.4 16.8 0.3 6.9 (9.4) (10.0) (68.9) (21.7) (0.4) (8.9) 29.3% 12.6 4.1 8.6 22.3 62.9 8.4 0.2 6.4 (3.2) (6.7) (80.8) (10.7) (0.2) (8.2) 37.0% 20.9 3.2 11.1 23.0 66.9 7.6 0.1 2.4 (2.5) (8.6) (86.9) (9.8) (0.1) (3.1) 44.0% 30.6 2.0 10.4 22.4 68.6 5.0 0 4.0 (1.6) (8.1) (88.4) (6.5) (0) (5.1) aMF: Methyl formate; bDMM: dimethoxymethane; cCOx: CO + CO2; dThe data in parentheses are reported on a CH3OH-free basis by excluding CH3OH as a reaction product in view of its reactivity in formaldehyyde synthesis and its tendency to dehydrate to re-form CH3OCH3.

[0051] 6 TABLE 6 Effects of surface density of MoOx/ZrO2 catalysts (calcined at 873 K) on primary DME reaction rates and primary products at 513 K (80.0 kPa, 18 kPa O2 and 2 kPa N2). MoO3 loading Mo surface density Rated Rated Selectivity (%)d (MoO3 wt. %) (Mo/nm2) (mol/g-atom Mo-h) (10−5 mol/m2-h) CH3OH HCHO MFa DMMb COxc 11.0% 4.5 23.0 17.0 24.9 46.0 23.5 0.4 5.2 (17.3) (13.0) (61.2) (31.3) (0.5) (7.0) 20.7% 8.4 46.5 51.7 23.7 60.1 11.5 0.8 4.0 (35.5) (39.5) (78.6) (15.1) (1.1) (5.2) 29.3% 20.0 19.0 60.0 24.1 59.6 12.2 0.9 3.2 (14.4) (45.5) (78.5) (16.1) (1.2) (4.2) 37.0% 31.4 11.7 61.1 24.9 58.3 12.6 1.1 3.0 (8.8) (45.9) (77.7) (16.8) (1.5) (4.0) 44.0% 50.1 7.4 61.3 27.8 60.1 10.4 0.9 1.3 (5.3) (44.3) (82.7) (14.4) (1.2) (3.8) aMF: Methyl formnate; bDMM: dimethoxymethane; cCOx: CO + CO2; dThe data in parentheses are reported on a CH3OH-free basis by excluding CH3OH as a reaction product in view of its reactivity in formaldehyde synthesis and its tendecy to dehydrate to re-form CH3OCH3.

[0052] In other work, the effect of partial pressures on the reaction was investigated on the 44MoOx/ZrO2 catalyst (50.1 Mo/nm2). The reaction rates nearly increased linearly as the DME partial pressure increased from 10 to 40 kPa, and then approached the constant values above 60 kPa. The primary selectivities to methyl formate and dimethoxymethane were almost independent of the DME partial pressure. The primary selectivity to COx decreased from 6.0% to the constant value of 1.5% while the selectivity to HCHO increased from 78.3 to 82.7% with increasing the DME pressure from 10 to 40 kPa.

Example 5:

[0053] Preparation and Use of Supported Vanadium Pentoxide Catalysts

[0054] The catalysts were prepared by incipient wetness impregnation of &ggr;-Al2O3 (Degussa AG) with an aqueous solution of ammonium metavanadate [NH4NO3] (99%, Aldrich) and oxalic acid (Mallinckrodt, analytical grade; NH4/NO3/oxalic acid=0.5 M) (the addition of oxalic acid improves the dissolution of NH4NO3 in water). The V5+ concentrations in the solution were varied in order to get desired V content in the final catalysts. After impregnation, samples were dried in air at 393 K and treated in dry air at 773 K (500° C.) for 3 h.

[0055] On VOx/Al2O3 (8.0 V/nm2) the CH3OH-free primary HCHO selectivity was 99.6% at 513 K and the primary DME reaction rate was 6.8 mol/g-atom V-h. Results are shown in Table 7. 7 TABLE 7 Primary DME reaction rates and primary products on VOx dispersed on different supports at 240° C. (80.0 kPa, 18 kPa O2 and 2 kPa N2). Catalyst V surface density Ratea Ratea Selectivity (%) (V2O5 wt. %) (V/nm2) (mol/g-atom V-h) (10−5 mol/m2-h) CH3OH HCHO MFb DMMc COxd VOx/Al2O3 8.0 8.0 6.7 14.72 84.96 0.26 0 0 (10.0%) (6.8) (5.6) (99.61) (0.3) VOx/ZrO2 6.2 13.1 8.5 5.2 37.9 9.2 0 47.7 (15%) (12.4) (8.1) (40.0) (9.7) (50.3) VOx/MgO 5.5 1.8 1.0 11.0 0 0 0 89.0 (20.0%) aThe data in the parenteses are reported on a CH3OH-free basis by excluding CH3OH as a reaction product in view of its reactivity in formaldehyde synthesis and its tendency to dehydrogenate to re-from CH3OCH3; bMF: methyl formate; cDMM: dimethoxymethane; dCOx: CO + CO2.

Example 6.

[0056] Use of a Mixed Oxide Catalyst.

[0057] Similarly, experiments were conducted using a catalyst containing molybdenum oxide on a support of alumina modified by stannic oxide.

[0058] The supports were prepared by incipient wetness impregnation of alumina (Degussa AG) with an isopropanol solution of tin (IV) isopropoxide (10 w/v in isopropoxide) (98% metal basis, Alfa Aesar). The alumina was treated at 125° C. to remove physisorbed water before the impregnation. The isopropanol solution of tin (IV) isopropoxide was added to the alumina at room temperature under a flow of dry nitrogen. After impregnation at room temperature for 5h, the sample was dried in air at 125° C. overnight, then treated in dry air at 500° C. for 3 h. The amount of tin (IV) isopropoxide added was varied in order to get desired Sn content in the final SnO2/Al2O3 supports. The MoOx/Sn-Al catalysts were prepared by incipient wetness impregnation of the SnO2/Al2O3 support obtained above with an aqueous solution of ammonium heptamolybdate (99%, Aldrich). Impregnated samples were dried in air overnight at 125° C., then treated in dry air at 773 K for 3 h.

[0059] The catalyst retained the good selectivity of the molybdenum oxide/alumina catalysts but had higher activity. Results are reported in Table 8. 8 TABLE 8 Primary DME oxidation rates and selectivities on MoOx supported on SnO2/Al2O3 and on Al2O3 for comarison at 513 K (80.0 kPa DME, 18 kPa O2 and 2 kPa N2). Surface Mo surface DME Conversion Catalyst area density Ratea Selectivity (%)a (MoO3 wt %) (m2/g) (Mo/nm2) (mol/g-atom Mo-h) CH3OH HCHO MFb DMMc COxd MoO3/Al2O3 90.0 7.0 5.7 19.7 79.9 1.6 0 0 (15.0%) (4.6) (98.1) (1.9) (0) (0) MoO3/SnO2 87.9 7.1 15.0 18.6 79.5 2.0 0 0 Al2O3e (12.2) (97.7) (2.4) (0) (0) (15%) aThe data in the parenteses are reported on a CH3OH-free basis by excluding CH3OH as a reaction product in view of its reactivity in formaldehyde synthesis and its tendency to dehydrogenate to re-form CH3OCH3; bMF: methyl formate; cDMM: dimethoxymethane; dCOx: CO + CO2. eSnO2 surface density on Al2O3: 6.1 Sn/nm2.

[0060] All publications and patent applications cited in this specification are herein in corporated by reference as if each individual publication or patent application were specifically and individually indicated to be incorporated by reference.

[0061] Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it will be readily apparent to those of ordinary skill in the art in light of the teachings of this invention that certain changes and modifications may be made thereto without departing from the spirit or scope of the appended claims.

Claims

1. A process for the production of formaldehyde by oxidation of dimethyl ether in the presence of a supported catalyst comprising molybdenum oxide, vanadium oxide or a mixture of molybdenum and vanadium oxides, wherein the support is selected from catalyst supports that allow the formation of monolayers of molybdenum and/or vanadium oxide on the surface of the support but that do not substantially react with the molybdenum and/or vanadium oxide to form unreducible mixed oxides, wherein the molybdenum oxide, vanadium oxide, or mixture of such oxides is dispersed on the surface of the support, the surface density of the oxide or oxides on the support being greater than the surface density of the respective isolated monomeric oxide or oxides, and wherein the catalyst is characterized by a substantial absence of bulk crystalline molybdenum and/or vanadium oxides.

2. A process according to claim 1 in which the surface density of the oxide or oxides on the support is from about 50% of the surface density of a monolayer of the oxide or oxides to about 300% of the surface density of a monolayer of the oxide or oxides.

3. A process according to claim 1 in which the surface density of the oxide or oxides on the support is approximately that of a monolayer of said oxide or oxides.

4. A process according to claim 1 in which the support is selected from alumina, zirconia, stannic oxide, titania, silica, and mixtures thereof.

5. A process according to claim 1 in which the oxide comprises molybdenum oxide.

6. A process according to claim 5 in which the support comprises alumina.

7. A process according to claim 5 in which the support comprises zirconia.

8. A process according to claim 5 in which the support comprises stannic oxide.

9. A process according to claim 8 in which the support comprises a layer of stannic oxide disposed on a particulate alumina or zirconia.

10. A process according to claim 5 in which the surface density of the molybdenum oxide is from about 1.5 to about 20 Mo/nm2.

11. A process according to claim 6 in which the surface density of the molybdenum oxide is from about 50% of the surface density of a molybdenum oxide monolayer on the alumina support to about 300% of the surface density of a molybdenum oxide monolayer on the support.

12. A process according to claim 7 in which the surface density of the molybdenum oxide is from about 1.5 to about 50 Mo/nm2.

13. A process according to claim 7 in which the surface density of the molybdenum oxide is from about 50% of the surface density of a molybdenum oxide monolayer on the zirconia support to about 400% of the surface density of a molybdenum oxide monolayer on the support.

14. A process according to claim 5 in which the surface density of the molybdenum oxide is approximately that of a monolayer of molybdenum oxide on the support.

15. A process according to claim 1 in which the oxide is vanadium oxide.

16. A process according to claim 15 in which the support comprises alumina.

17. A process according to claim 15 in which the support comprises zirconia.

18. A process according to claim 15 in which the support comprises stannic oxide.

19. A process according to claim 18 in which the support comprises a layer of stannic oxide disposed on particulate alumina or zirconia.

20. A process according to claim 15 in which the surface density of the vanadium oxide on the support is from about 50% of the surface density of a monolayer of vanadium oxide to about 300% of the surface density of a monolayer of vanadium oxide.

21. A process according to claim 1 in which methyl formate is a co-product with the formaldehyde.

22. A process according to claim 1 in which selectivity to formaldehyde is 50% or greater.

23. A process for the production of formaldehyde by oxidation of dimethyl ether in the presence of a supported catalyst comprising molybdenum oxide, vanadium oxide or a mixture of molybdenum and vanadium oxides, wherein the support is selected from alumina, zirconia, stannic oxide, titania and silica, and mixtures thereof, wherein the surface density of the oxide or oxides on the support is from about 50% of the surface density of a monolayer of the oxide or oxides to about 300% of the surface density of a monolayer of the oxide or oxides, and wherein the catalyst is characterized by a substantial absence of bulk crystalline molybdenum and/or vanadium oxides.

24. A process according to claim 23 in which the surface density of the molybdenum or vanadium oxide or mixture of said oxides on the support is approximately that of a monolayer of said oxide or oxides on the support.

25. A process according to claim 23 in which the oxide is molybdenum oxide.

26. A process according to claim 25 in which the support comprises alumina.

27. A process according to claim 25 in which the support comprises zirconia.

28. A process according to claim 25 in which the catalyst comprises a layer of zirconia disposed on particulate alumina.

29. A process according to claim 25 in which the support comprises stannic oxide.

30. A process according to claim 29 in which the support comprises a layer of stannic oxide disposed on particulate alumina or zirconia.

31. A process according to claim 26 in which the surface density of the molybdenum oxide is from about 1.5 to about 20 Mo/nm2.

32. A process according to claim 26 in which the surface density of the molybdenum oxide is from about 50% of the surface density of a molybdenum oxide monolayer to about 300% of the surface density of a molybdenum oxide monolayer.

33. A process according to claim 27 in which the surface density of the molybdenum oxide is from about 1.5 to about 50 Mo/nm2.

34. A process according to claim 27 in which the surface density of the molybdenum oxide is from about 50% of the surface density of a molybdenum oxide.

35. A process according to claim 25 in which the surface density of the molybdenum oxide is approximately that of a monolayer of molybdenum oxide.

36. A process according to claim 23 in which the oxide is vanadium oxide.

37. A process according to claim 36 in which the support comprises alumina.

38. A process according to claim 36 in which the support comprises zirconia.

39. A process according to claim 36 in which the support comprises stannic oxide.

40. A process according to claim 39 in which the support comprises a layer of stannic oxide disposed on a particulate alumina or zirconia.

41. A process according to claim 36 in which the surface density of the vanadium oxide on the support is from about 50% of the surface density of a monolayer of vanadium oxide to about 300% of the surface density of a monolayer of vanadium oxide.

42. A process according to claim 23 in which methyl formate is a co-product with the formaldehyde.

43. A process according to claim 22 in which selectivity to formaldehyde is 50% or greater.

44. A catalyst comprising molybdenum oxide, vanadium oxide, or a mixture of molybdenum oxide and vanadium oxide supported on a layer of stannic oxide, the stannic oxide layer being disposed on a particulate alumina or zirconia support, in which the surface density of the molybdenum and/or vanadium oxide or oxides on the support is greater than that for the respective monomeric isolated oxide or oxides, and the catalyst is characterized by a substantial absence of bulk crystalline molybdenum and/or vanadium oxides.

45. A catalyst according to claim 44 in which the surface density of the oxide or oxides on the support is from about 50% of the surface density of a monolayer of the oxide or oxides to about 300% of the surface density of a monolayer of the oxide or oxides.

46. A catalyst according to claim 44 in which the surface density of the molybdenum and/or vanadium oxide or oxides on the support is approximately that of a monolayer of the oxide or oxides at the surface of the support.

47. A catalyst according to claim 44 comprising molybdenum oxide supported on a layer of stannic oxide that is disposed on a particulate alumina support, and in which the surface density of the molybdenum oxide is from about 1.5 to about 20 Mo/nm2.

48. A catalyst according to claim 44 comprising molybdenum oxide supported on a layer of stannic oxide that is disposed on a particulate alumina support, and in which the surface density of the molybdenum oxide is approximately that of a monolayer of the oxide or oxides at the surface of the support.