IT8349069A1 - CARBONYLATION PROCESS. - Google Patents

Description

Carbonylation process.

SUMMARY

A carboxylic acid, a carboxylic acid anhydride, or a carboxylic acid ester, such as propionic acid, propionic anhydride, or methyl propionate, is prepared by carbonylation of an olefin, such as ethylene, in the presence of water, a carboxylic acid, and/or an alcohol, and in the presence of a halide, wherein a catalyst comprising a component is used.

molybdenum-nickel-alkali metal, tungsten-nickel-alkali metal, or chromium-nickel-alkali metal co-catalyst.

The present invention relates to the carbonylation of olefins to produce carboxylic acids, more particularly monocarboxylic acids, and especially lower alkanoic acids, such as propionic acid, the anhydrides of these acids, and the esters of these acids, especially lower alkanoic esters.

Carboxylic acids have been known as industrial chemicals for many years, and large quantities are used in the manufacture of various products. The production of carboxylic acids by the action of carbon monoxide on olefins (carbonylation) has been described, for example, in Reppe et al., U.S. Patent No. 2,768,968. However, these earlier proposals imply that olefin carbonylation reactions have required the use of very high pressures. Olefin carbonylation processes effective at lower pressures have also been proposed. Craddock et al., U.S. Patent Nos. 3,579,551; 3,579,522; and 3,816,488, for example, describe the carbonylation of olefins in the presence of Group VIII noble metal compounds and complexes, such as iridium and rhodium, in the presence of iodide, under more moderate pressures than those envisaged by Reppe et al. These lower-pressure carbonylation explanations, however, require the use of expensive noble metals. More recently, Belgian patent No. 860,557 proposed the preparation of carboxylic acids by carbonylation of alcohols in the presence of a nickel catalyst promoted by a trivalent phosphorus compound and in the presence of an iodide.

The production of anhydrides by the action of carbon monoxide on olefins has also been described, for example, in the aforementioned Reppe et al. U.S. Patent No. 2,768,968, using very high pressures. Foster et al., U.S. Patent No. 3,852,346, describes the carbonylation of olefins in the presence of Group VIII noble metal compounds such as iridium and rhodium and in the presence of an iodide under more moderate pressures than those provided by Reppe et al. However, this process, like the process in the Craddock et al. patents, requires the use of relatively expensive rare earth metals.

The production of carboxylic acid esters by the action of carbon monoxide on olefins has been described in several patents using processes involving various types of catalysts. For example, Slaugh U.S. Patent No. 3,168,553 shows the reaction of carbon monoxide with an olefinic hydrocarbon in the presence of alcohols using a Group VIIIb transition metal carbonyl catalyst containing cobalt, ruthenium, rhodium, or iridium in complex combination with carbon monoxide and a trialkyl phosphorus. Anderson et al., U.S. Patent No. 3,040,090, produces carbon monoxide, an ethylenically unsaturated compound, and an alcohol in the presence of a Group VIII noble metal chelate. Morris et al., U.S. Patent No. 3,917,677, also discloses a process involving a reaction between carbon monoxide, ethylenically unsaturated compounds, and alcohols that is characterized by the use of a catalyst containing a rhodium component and a tertiary organophosphorus component. This patent contains a discussion of the prior art and the limitations of prior art processes, particularly the low yields achievable therewith. Furthermore, the prior art process generally requires relatively high pressures. Although improved yields are apparently achieved by the process of U.S. Patent No. 3,917,677, this process requires the use of a rhodium catalyst.

U.S. Patent Nos. 4,335,058, 4,354,036, and 4,354,036,

Nos. 4,372,889 illustrate related processes for the carbonylation of olefins and employ a molybdenum-nickel or tungsten-nickel cocatalyst in the presence of a promoter comprising an organophosphorus compound or an organonitrogen compound, such as a tertiary amine phosphine, to produce carboxylic acid and carboxylic acid anhydrides, carboxylic acid esters, and carboxylic acid anhydrides, respectively. While this process involves nickel catalysts that enable the carbonylation of olefins at modest pressures without requiring the use of a noble metal catalyst, and while these processes are highly effective for their intended purpose, there is room for improvement in terms of reaction rate and productivity without requiring the use of organic promoters.

It is accordingly an object of the present invention to provide an improved process for the manufacture of carboxylic acids, especially lower alkanoic acids, such as propionic acid, the anhydrides of these acids, and their esters, which do not require high pressures or Group VIII noble metals and make possible the production of carboxylic acids, anhydrides, and esters, in high yields, in short reaction times, without the need for organic promoters.

According to the invention, the carbonylation of an olefin is carried out using a molybdenum-nickel-alkali metal, tungsten-nickel-alkali metal, or chromium-nickel-alkali metal cocatalyst in the presence of a halide, preferably an iodide, a bromide and/or a chloride, especially an iodide, and in the presence of a co-reactant which may be water.

a carboxylic acid, or an alcohol. Thus, when one wishes to produce a carboxylic acid, carbonylation is carried out in the presence of water. When one wishes to produce a carboxylic acid anhydride, carbonylation is carried out under essentially anhydrous conditions in the presence of a carboxylic acid, and, to produce an ester, carbonylation is carried out in the presence of an alcohol. The surprising discovery was that this cocatalyst system in such an environment makes it possible to carbonylate olefins, not only at relatively low pressures, but with rapid production, in high yields, of carboxylic acids, anhydrides, and esters.

The apparent effectiveness of the catalyst system of the process of this invention is particularly striking in view of the experimental data reported in published European patent application No. 0035 458 showing the carbonylation of methanol to produce acetic acid and in which experiments using nickel in combination with molybdenum or tungsten or chromium showed no reaction at all, even after two hours. It has also been observed that, when nickel-based catalysts are normally used in carbonylation reactions, there is a tendency for the nickel components to be volatilized and appear in the reaction vapors. It has been surprisingly found that, with the catalyst system of this invention, the volatility of the nickel is strongly suppressed and a highly stable catalyst combination is obtained, especially in the case of the molybdenum-containing co-catalyst, which is the preferred co-catalyst.

Thus, according to the invention, carbon monoxide is reacted with an olefin, such as a lower alkene, in the presence of a halide, such as a hydrocarbyl halide, especially a lower alkyl halide, such as ethyl iodide, and in the presence of the co-catalyst indicated above, and in the presence of water, a carboxylic acid, under anhydrous conditions, or an alcohol.

Propionic acid, propionic anhydride, or a propionic acid ester, for example, can be effectively prepared in representative cases by subjecting ethylene to carbonylation in the presence of appropriate co-reagents and in the presence of the catalyst system of this invention. Propionic acid is employed as a co-reactant in the case of propionic anhydride.

Similarly, other alkanoic acids, such as butyric acid and valeric acid, their anhydrides and esters can be produced by carbonylation of the corresponding alkene, such as propylene, butene-1, butene-2, hexenes, octenes, and the like. Similarly, other alkanoic acids, such as capric acid, caprylic acid, and lauric acid, and similar higher acids, their anhydrides and esters are produced by carbonylation of the corresponding olefin.

The reactant olefin may be any ethylenically unsaturated hydrocarbon having from 2 to 25 carbon atoms, preferably from 2 to about 15 carbon atoms. The ethylenically unsaturated compound has the following general structure:

W CR R

3 4

wherein R^, R^, R^ and R^ are hydrogen or the same or different alkyl, cycloalkyl, aryl, alkaryl, aralkyl, or wherein one of the said R, and R? and one of the said 1 2

R^ and R^ together form a single alkylene group having from 2 to about 8 carbon atoms. R^, R^ and R^ may be branched and may be substituted with substituents that are inert in the reactions of the invention.

Examples of useful ethylenically unsaturated hydrocarbons are ethylene, propylene, butene-l, butene-2, 2-methylbutene-1, cyclobutene, hexene-1, hexene-2. cyclohexene, 3-ethylhexene-l, isobutylene, octene-1, 2-methylhexene-1, ethylcyclohexene, decene-1, cycloheptene, cyclooctene, cyclononene, 3,3-dimethylnonene-l, dodecene-1, undecene-3, 6-propyldecene-l, tetradecene-2, 3-amyldecene-1 etc., hexadecene-1, 4-ethyltridecene-l, octadecene-1, 5,5-dipropyldodecene-l, vinylcyclohexane, allylcyclohexane, styrene, p-methylstyrene, alpha-methylstyrene, p-vinylcumene, beta-vinylnaphthalene, 1,1-diphenylbutene-1, 3-benzylheptene-l, divinylbenzene, 1-allyl-3-vinylbenzene, etc. Of the above olefins, the alpha hydrocarbon olefins and olefins having from 2 to 10 carbon atoms are preferred, e.g., ethylene, propylene, butene-1, hexene-1, heptene-1, octene-1, and the like, i.e., wherein R₁, R₂, R₁, and R₁ are hydrogen or 1 2 3 4

alkyl groups entirely from 1 to 8 carbon atoms, preferably lower alkenes, i.e. alkenes of ?

2 to 6 carbon atoms, especially ethylene.

The co-reacting carboxylic acid, in the case of anhydride production, can be, in general, any carboxylic acid having from 1 to about 25 carbons and having the formula:

RCOOH

wherein R is hydrogen, alkyl, cycloalkyl, or aryl. Preferably, R has 1 to about 18 carbon atoms, and more preferably R is an alkyl having 1 to about 12 carbon atoms, especially 1 to 6 carbon atoms, such as methyl, ethyl, propyl, isobutyl, hexyl, nonyl, and the like, or is an aryl having 6 to about 9 carbon atoms, such as phenyl, tolyl, and the like.

Examples of useful acids are acetic, propionic, n-butyric, isobutyric, pivalic, n-valeric, n-caproic, stearic, benzoic, phthalic, terephthalic, toluic, 3-phenylhexanoic acid, 2-xylyl palmitic acid, and 4-phenyl-5-isobutyl stearic acid. Preferred acids are fatty acids or alkanoic acids having 2 to 12 carbon atoms, such as acetic, propionic, n-butyric, isobutyric, pivalic, caproic, undecyl acid, and the like. Especially preferred are lower alkanoic acids in which R is an alkyl group of 1 to 6 carbon atoms, especially propionic acid. R may be branched and may be substituted with substituents that are inert in the reactions of the invention.

It is preferred that the reagents are chosen so that the resulting anhydride is a symmetric anhydride, i.e. having two identical acyl groups.

The co-reacting alcohol can be, in general, any alcohol having the formula ROH, where R is alkyl, cycloalkyl, aryl, alkali, or aralkyl, or mixtures thereof.

Preferably, R has 1 to about 18 carbons, and more preferably R is an alkyl having 1 to about 12 carbons, e.g., methyl, ethyl, propyl, butyl, isobutyl, pentyl, hexyl, nonyl, and the like, or is an aralkyl having 7 to about 14 carbons, e.g., benzyl, phenethyl, and the like.

Examples of suitable alcohols include methanol, ethanol, propanol, isopropanol, butanol, tertiary butanol, pentanol, hexanol, 2-ethylhexanol, octanol, decanol, 6-pentadecanol, cyclopentanol, methylcyclopentanol, cyclohexanol, benzyl alcohol, alpha-alpha-dimethyl benzyl alcohol, alpha-alpha-dimethyl benzyl alcohol, alpha-ethylphenethyl alcohol, naphthylcarbinol, xylyl carbinol, tolyl carbinol, and the like.

In more preferred embodiments of the invention, carbon monoxide is reacted with ethylene and water, propionic acid, or methanol in the presence of the halide cocatalyst system of the type described above to produce propionic acid, propionic anhydride, or methyl propionate in reactions which may be expressed as follows:

C H G?+H^<3 > C HgCOOH

2 4 2

C H > C H5COOCOC^K5

2 4 2 5

C_H .+CO+CH-OH > C H COOCH

2 4 3 2 5 3

Carbon monoxide is removed in the vapor phase along with the unreacted olefin when the olefin is normally gaseous, e.g., ethylene, and, if desired, recycled. Normally liquid and relatively volatile components, such as alkyl halide, the normally liquid unreacted olefin and the co-reactant (water, carboxylic acid, or alcohol), and any by-products present in the final product mixture can be readily removed and separated from each other, e.g., by distillation, for recycling, and the net product yield is substantially exclusively the desired carboxylic acid, anhydride, or ester. In the case of liquid-phase reaction, which is preferred, the organic compounds are readily separated from the metal-containing components, e.g., by distillation. The reaction is conveniently carried out in a reaction zone to which the carbon monoxide, olefin, co-reactant, halide, and co-catalyst are fed. When an anhydride is present, the organic compounds are readily separated from the metal-containing components. The desired product is not produced, and as mentioned, the reaction is carried out under essentially anhydrous conditions. In the case of ester production, the liquid-phase reaction is preferably carried out under boiling conditions, so that all volatile components are removed in the vapor phase, leaving the catalyst in the reactor.

As will be evident from the preceding equations, a carbonylation reaction of the type described, selective with respect to the carboxylic acid, carboxylic acid anhydride, or carboxylic acid ester, requires at least one mole of carbon monoxide and one mole of the co-reactant (water, carboxylic acid, or alcohol) per mole (equivalent) of an ethylenically unsaturated bond reacted. Thus, the starting material, the olefin, is normally charged with equimolar amounts of the co-reactant, although more co-reactant may be used.

In carrying out the process of the invention, a wide range of temperatures, e.g. 25° to 350°C are suitable, but temperatures of 100° to 250°C are preferably employed, and more preferred temperatures generally fall in the range of 125° to 225°C. Lower temperatures than those mentioned may be used, but they tend to give reduced reaction rates, and higher temperatures may also be employed, but there is no particular advantage in their use. The reaction time is also not a process parameter and depends largely on the pressure employed, but typical residence times, for example, will generally fall in the range of 0.1 to 20 hours. The reaction is carried out under superatmospheric pressure, but, as previously stated, it is a feature of the invention that excessively high pressures, requiring special high-pressure equipment, are not necessary. In general, the reaction is The reaction is effectively carried out using a partial pressure of carbon monoxide that is preferably at least 15, but less than 2,000 psi, more preferably 15 to 1,000 psi, and particularly 30 to 200 psi, although partial pressures of CO of 1 to 5,000, or even up to 10,000 psi, may also be employed. By establishing the partial pressure of carbon monoxide at the specified values, adequate quantities of this reagent are always present. The total pressure is, of course, that which will provide the desired partial pressure of carbon monoxide, and preferably is that required to maintain the liquid phase, and in this case, the reaction may advantageously be carried out in an autoclave or similar apparatus. At the end of the desired residence time, the reaction mixture is separated into its various constituents, for example by distillation. Preferably, the reaction product is introduced into a distillation zone that can be A fractional distillation column, or series of columns, may be effective for separating the volatile components from the anhydride or ester product and for separating the product from the less volatile catalyst components of the reaction mixture. The boiling points of the volatile components are sufficiently far apart that their separation by conventional distillation presents no particular problem. Likewise, the higher-boiling organic components can be readily distilled away from the metal catalyst components. The cocatalyst thus recovered, including the halide component, can then be combined with fresh quantities of olefin, carbon monoxide, and coreactant and reacted to produce additional quantities of carboxylic acid, anhydride, or ester. When the reaction is carried out under boiling conditions, the effluent is completely in the vapor phase, and after condensation, the components can be separated from each other, as described above.

Although not required, the process can be carried out in the presence of a solvent or diluent. The presence of a higher-boiling solvent or diluent, preferably the product itself, e.g., propionic acid, propionic anhydride, or a propionic acid ester in the case of ethylene carbonylation, will enable more moderate total pressures to be used. Alternatively, the solvent or diluent may be any organic solvent that is inert in the process environment, such as hydrocarbons, e.g., octane, benzene, toluene, xylene tetraline, or carboxylic acids. A carboxylic acid, if used, will preferably correspond to the carboxylic acid fraction in the product to be produced, as it is preferred that the solvent employed be indigenous to the system, e.g., propionic acid in the case of ethylene carbonylation, although other carboxylic acids, e.g., acetic acid, may also be used. A solvent or thinner,

when not the product itself, it is conveniently chosen to have a boiling point sufficiently different from the desired product in the reaction mixture, so that it can be easily separated, as will be apparent to those skilled in the art. Mixtures may be used.

Carbon monoxide is preferably used in a substantially pure form, as commercially available, but inert diluents, such as carbon dioxide, nitrogen, methane, and noble gases, may be present if desired. The presence of inert diluents does not affect the carbonylation reaction, but their presence makes it necessary to increase the total pressure to maintain the desired CO partial pressure. Hydrogen, which may be present as an impurity, is not objectionable and may also tend to stabilize the catalyst. Indeed, to obtain low CO partial pressures, the fed CO can be diluted with hydrogen or any inert gas, such as those mentioned above. Amounts of diluent gas up to 95% can be used.

The co-catalyst components may be employed in any convenient form, that is, in the zero-valent state, or in any higher-valent form. For example, nickel and molybdenum, tungsten, or chromium may be the metals themselves in finely divided form, or either an organic or inorganic compound, which is effective for introducing the co-catalyst components into the reaction system. Thus, typical compounds include the carbonate, oxide, hydroxide, bromide, iodide, chloride, oxyhalide, hydride, lower alkoxide (methoxide), phenoxide, or carboxylates of Mo, W, Cr, or Ni in which the carboxylate ion is derived from an alkanoic acid of 1 to 20 carbon atoms, such as acetates, butyrates, de^ pr;X( laurates, benzoates, and the like. Similarly, complexes of any of the co-catalyst components may be employed, e.g., metal carbonyls and alkyls, as well as chelates, association compounds, and anolic salts. Examples of other complexes include bis-(triphenylphosphine)nickel dicarbonyl, tricyclopentadienyl trinickel dicarbonyl, tetrakis(triphenylphosphite)nickel, and the corresponding complexes of other components, such as molybdenum hexacarbonyl and tungsten hexacarbonyl.

Particularly preferred are the elementary forms, compounds which are halides, especially iodides, and organic salts, for example salts of the monocarboxylic acid corresponding to the carboxylic acid fraction in the product to be obtained.

The alkali metal component, for example, a metal from Group IA of the periodic table such as lithium, potassium, sodium, and cesium, is conveniently employed as a compound, especially a salt, and more preferably a halide, such as an iodide. The preferred alkali metal is lithium. The alkali metal component may, however, also be employed as the hydroxide, carboxylate, alkoxide, or in the form of other suitable compounds, as indicated above, in relation to the cocatalyst components, and typical alkali metal components are illustrated by sodium iodide, potassium iodide, cesium iodide, lithium iodide, lithium bromide, lithium chloride, lithium acetate, and lithium hydroxide.

It should be understood that the above-mentioned compounds and complexes are merely illustrative of suitable forms of various co-catalyst components, and are not intended to be limiting.

The specified co-catalyst components used may contain impurities normally associated with commercially available metal or metal compounds and do not need to be further purified.

The amount of each co-catalyst component employed is in no way critical and is not a parameter of the inventive process and can vary over a wide range. As is well known to those skilled in the art, the amount of catalyst used is that which will provide the desired, suitable, and reasonable reaction rate, since the reaction rate is influenced by the amount of catalyst. However, essentially any amount of catalyst will facilitate the basic reaction and can be considered a theoretically effective amount. Typically, however, each catalyst component is employed in an amount of 1 millimole to 1 mole per liter of reaction mixture, preferably 15 millimoles to 500 millimoles per liter, and more preferably 15 millimoles to 150 millimoles per liter.

The ratio of nickel to the molybdenum, tungsten, or chromium co-catalyst component may vary. Typically, it is 1 mole of the nickel component per 0.01 to 100 moles of the second co-catalyst component, i.e., the molybdenum, tungsten, or chromium component. Preferably, the nickel component is used in the amount of 1 mole per 0.1 to 20 moles, more preferably 1 mole per 1 to 10 moles of the second co-catalyst component. Similarly, the ratio of nickel to the alkali metal component may vary, e.g., 1 mole of nickel per 1 to 1000 moles of the alkali metal component, preferably 1 mole per 10 to ?OO moles, and most preferably 1 mole per 20 to 50 moles.

The amount of the halide component can also vary widely, but, in general, will be present in an amount of at least 0.1 mole (expressed as elemental halogen) per mole of nickel. Typically, 1 to 100 moles of halide per mole of nickel are used, preferably 2 to 50 moles per mole. Ordinarily, no more than 200 moles of halide per mole of nickel are used. It should be understood, however, that the halide component need not be added to the system as a hydrocarbyl halide but may be supplied as another organic halide or as a hydrohalide or other inorganic halide, for example, a salt, such as an alkali metal or other metal salt, or even as an elemental halide, such as iodine or bromine.

As previously mentioned, the catalyst system of this invention comprises a halide, especially an iodide, a molybdenum-nickel-alkali metal, tungsten-nickel-alkali metal, or chromium-nickel-alkali metal co-catalyst component. The catalyst system of this invention permits the production of carboxylic acids, carboxylic acid anhydrides, and carboxylic acid esters in high yields in short reaction times, without the use of Group VIII noble metals, and the presence of the alkali metal component along with the molybdenum, tungsten, or chromium component enables good results with relatively small amounts of the co-catalyst components and reduced amounts of nickel compared to prior art processes involving a nickel-containing catalyst.

An embodiment of the catalyst comprising the molybdenum-nickel-alkali metal, tungsten-nickel-alkali metal, or chromium-nickel-alkali metal co-catalyst component and the halide component may be represented by the following formula: X:T:Z:Q, wherein X is molybdenum, tungsten, or chromium, T is nickel, X and T being in a zero-valence form or in the form of a halide, oxide, carboxylate of 1 to 20 carbon atoms, carbonyl, or hydride; Z is a halide source which is hydrogen halide, halogen, alkyl halide, wherein the alkyl group contains 1 to 20 carbon atoms, or alkali metal halide, and Q is the alkali metal component. The preferred alkali metal is lithium, as previously indicated, and is in the form of an iodide, bromide, chloride, or carboxylate as defined for X and T, the molar ratio of X to T being 0.1 - 10:1, the molar ratio of X T to Q being 0.1 - 10:1, and the molar ratio of Z to X T being 0.01 - 0.1:1.

It will be apparent that the reaction described above readily tends towards a continuous operation in which the reactants and catalyst are continuously fed to the appropriate reaction zone and the reaction mixture is continuously distilled to separate the organic "-"late constituents and to provide a net product consisting substantially of carboxylic acid, anhydride or ester, with the other organic components being recycled and, in a liquid phase reaction, a fraction containing residual catalyst also being recycled.

It is also apparent that the catalytic reaction involved in the process of the invention can be carried out in the vapor phase, if desired, by appropriate control of the total pressure versus temperature, so that the reactants are in vapor form when in contact with the catalyst. In the case of vapor-phase operation, and in the case of liquid-phase operation, if desired, the catalyst components can be supported, that is, they can be dispersed on a conventional carrier such as alumina, silica, silicon carbide, zirconium oxide, carbon, bauxite, attapulgite clay, and the like. The catalyst components can be applied to the carriers in a conventional manner, such as by impregnating the carrier with a solution of the catalyst component. Carrier concentrations can vary widely, e.g., 0.01 wt.% to 10 wt.% or higher. Typical operating conditions for vapor-phase operation are temperatures of 100°C to 100°C. at 350?C, preferably 150? at 275?C, and more preferably 175? at 255?C, a pressure of 1 to 5000 psia, preferably 59 at 1500 psia, and more preferably 150 at 500 psia, with volumetric velocities of 50 to 10,000 h, preferably 200 to 6000 h, and most preferably 500 to 4000 h (STP). In the case of a supported catalyst, the iodide component is included with the reactants and not on the support.

The following examples will serve to provide a more complete understanding of the invention, but it should be understood that they are given for illustrative purposes only and are not to be construed as limiting the invention. In the examples, all parts are by weight unless otherwise indicated.

EXAMPLE 1

In this example, a magnetically stirred pressure vessel with a glass jacket was employed. The reaction vessel was charged with 12 parts water, 11.9 parts ethyl iodide, 0.76 parts nickel iodide (Nil^), 1.4 parts molybdenum hexacarbonyl, 14.3 parts lithium iodide, and 60 parts ethyl acetate as solvent. The vessel was swept with argon and pressurized to 100 psig with hydrogen and then to 500 psig with carbon monoxide. The vessel was then heated to 180°C with stirring. The pressure was raised to 750 psig with ethylene. The pressure was maintained at 750 psig by charging with a 1:1 mixture of ethylene and carbon monoxide, and the temperature was was held at 180?C. After 1 hour of reaction, G.C. analysis of the reaction mixture showed that propionic acid had formed at the rate of 4.38 moles per liter per hour with all reacted olefins appearing as propionic acid.

EXAMPLE 2

A pressure vessel as described in Example 1 was charged with 12 parts water, 12 parts ethyl iodide, 0.74 parts nickel iodide (Nil), 1.4 parts molybdenum hexacarbonyl, 14.4 parts lithium iodide, and 60 parts tetrahydrofuran solvent. The vessel was swept with argon and pressurized to ?00 psig with hydrogen and then to 500 psig with carbon monoxide. Then, the vessel was heated to 180°C with stirring and the pressure was raised to 800 psig with ethylene. The pressure was maintained at 800 psig by charging with a 1:1 mixture of ethylene and carbon monoxide, and the temperature was maintained at 80°C. After 1 hour of reaction, G.C. analysis showed a 1:1 reaction volume of ethylene and carbon monoxide. of the reaction mixture showed that propionic acid was formed at the rate of 2.97 liters per hour with all of the reactant olefin appearing as propionic acid.

EXAMPLE 3

A pressure vessel as described in Example 1 was charged with 12 parts water, 12 parts iodoethane, 0.72 parts nickel iodide (Nil^), 1.4 parts chromium hexacarbonyl, 14.3 parts lithium iodide, and 60 parts tetrahydrofuran as a solvent. The vessel was swept with argon and pressurized to ?00 psig with hydrogen and then to 500 psig with carbon monoxide. The vessel was heated to 180°C with stirring and the pressure was raised to 900 psig using ethylene. The pressure was maintained at 900 psig by charging a 1:1 mixture of ethylene and carbon monoxide. The temperature was maintained at 180°C. After 1 hour of reaction, G.C. analysis of the reaction mixture showed ? that propionic acid had been formed at the rate of 1.21 moles per liter per hour and that all of the reacted olefin had been converted to propionic acid.

EXAMPLE 4

A magnetically stirred pressure vessel with a glass jacket was charged with 12 parts water, 12 parts ethyl iodide, 0.72 parts nickel iodide (Nil), 1.43 parts molybdenum hexacarbonyl, and 14 parts lithium iodide, with 60 parts tetrahydrofuran as the solvent. The vessel was purged with argon and pressurized to 300 psig.

with carbon monoxide. The vessel was then heated to 180°C with stirring and the pressure was raised to 750 psig with ethylene. The pressure was maintained at 750 psig by recharging a 1:1 mixture of ethylene and carbon monoxide, and the temperature

was maintained at 180?C. After 1 hour of reaction, G.C. analysis of the reaction mixture showed that propionic acid was formed at the rate of

4.63 moles per liter per hour and that all 1?olefin

reacted had been converted to propionic acid.

EXAMPLE 5

Example 4 was repeated, except that the molybdenum carbonyl was replaced with an equal amount of tungsten carbonyl. After 1 hour of reaction, G.C. analysis showed that propionic acid

was formed at the rate of 2.6 moles per liter per hour and that all the reacted ethylene had been converted to propionic acid.

EXAMPLE 6

Using a reactor as described in Example 1, the vessel was charged with 11.4 parts water, 11.4 parts ethyl iodide, 0.7 parts nickel iodide (Nil), 1.4 parts molybdenum hexacarbonyl, 18 parts lithium iodide, and 57 parts tetrahydrofuran solvent. The vessel was swept with argon and pressurized with 100 psig ethylene and up to 600 psig carbon monoxide. The vessel was then heated to 180°C with stirring, and the pressure was maintained at 750 psig by charging a 1:1 mixture of ethylene and carbon monoxide. The temperature was maintained at 180°C. After 1 hour of reaction, G.C. analysis of the reaction mixture showed that propionic acid had formed at the rate of 1:1. of 6.3 moles per liter per hour.and that all the reacted ethylene had been converted to propionic acid.

EXAMPLE 7

A reactor as described in Example 1 was charged with 12 parts water, 12 parts ethyl iodide, 0.7 parts nickel iodide, 1.4 parts molybdenum hexacarbonyl, 14 parts cesium iodide, and 60 parts tetrahydrofuran as a solvent. The vessel was purged with argon and pressurized to 100 psig with hydrogen and then to 500 psig with carbon monoxide and heated to 180°C with stirring. The pressure was increased to 900 psig using ethylene, and the pressure was maintained at 900 psig by charging a 1:1 mixture of ethylene and carbon monoxide. The temperature was maintained at 180°C. After 1 hour of reaction, G.C. analysis of the reaction mixture showed that propionic acid had formed at the rate of 1:1. of 1.7 moles per liter per hour and that all the reacted ethylene had been converted to propionic acid.

COMPARATIVE EXAMPLE A

Example 2 was repeated except that no molybdenum was added. After 1 hour of reaction, G.C. analysis showed that the ratio of propionic acid formed was only 0.12 moles per liter per hour.

EXAMPLE 8

A magnetically stirred pressure vessel with a glass jacket was charged with 68 parts propionic acid, 12 parts iodoethane, 0.8 parts nickel iodide, 1.6 parts molybdenum hexacarbonyl, and 16 parts lithium iodide. The vessel was purged with argon and pressurized to 100 psig with hydrogen and then to 500 psig with carbon monoxide. The vessel was heated to 160°C with stirring and the pressure was raised to 800 psig with ethylene. The pressure was maintained at 800 psig by charging a 1:1 mixture of ethylene and carbon monoxide, and the temperature was maintained at 160°C. After 50 minutes of reaction, G.C. analysis of the reaction mixture showed that propionic anhydride had been formed at the rate of 100 psig. of 10 moles per liter per hour and that all the reacted ethylene had been converted to propionic anhydride.

EXAMPLE 9

A magnetically stirred, glass-lined pressure vessel was charged with 68 parts propionic acid, 1.6 parts molybdenum hexacarbonyl, 0.8 parts nickel iodide, 12 parts ethyl iodide, and 16 parts lithium iodide. The vessel was purged with argon and pressurized to 100 psig with hydrogen and then to 500 psig with carbon monoxide. The vessel was heated to 180°C with stirring. The pressure was raised to 750 psig with ethylene, and the pressure was maintained at 750 psig by charging a 1:1 mixture of ethylene and carbon monoxide. The temperature was maintained at 180°C. After 1 hour of reaction, G.C. analysis of the reaction mixture showed that propionic anhydride had formed at the rate of 100 psig. of 7.1 moles per liter per hour and that all the reacted ethylene had been converted to propionic anhydride.

EXAMPLE 10

Example 8 was repeated except that the temperature was held at 140°C and the reaction occurred for 1 hour. GC analysis of the reaction mixture showed that propionic anhydride was formed at a rate of 3.2 moles per liter per hour and that all the reacted ethylene had been converted to propionic anhydride.

EXAMPLE 11

Example 9 was repeated, except that molybdenum hexacarbonyl was replaced with an equal amount of tungsten hexacarbonyl. After 1 hour of reaction, G.C. analysis of the reaction mixture showed that propionic anhydride had formed at the rate of 1.8 moles per liter per hour and that all of the reacted ethylene had been converted to propionic anhydride.

EXAMPLE 12

Example 9 was repeated except that lithium iodide was replaced with an equal amount of potassium iodide. After 2 hours of reaction, G.C. analysis showed that propionic anhydride had been formed at the rate of 0.7 moles per liter per hour and that all of the reacted ethylene had been converted to propionic anhydride.

EXAMPLE 13

Example 9 was repeated, except that the lithium iodide was replaced with an equal amount of cesium iodide. After 1 hour of reaction, G.C. analysis showed that propionic anhydride had been formed at a rate of 1.9 moles per liter per hour and that all the reacted ethylene had been converted to propionic anhydride.

EXAMPLE 14

Example 9 was repeated, except that no hydrogen was charged after 1 hour of reaction. GC analysis of the reaction effluent showed that propionic anhydride had formed at a rate of 2.3 moles per liter per hour and that all the reacted ethylene had been converted to propionic anhydride.

COMPARATIVE EXAMPLE B

Example 9 was repeated except that lithium iodide was not charged. After 1 hour of reaction, G.C. analysis showed that propionic anhydride had formed at a rate of only 0.2 moles per liter per hour.

COMPARATIVE EXAMPLE C

Example 9 was repeated, except that no molybdenum hexacarbonyl was charged. After 1 hour of reaction, G.C. analysis showed that propionic anhydride had formed at the rate of 0.5 moles per liter per hour.

EXAMPLE 15

In this example, a magnetically stirred Parr Hastelloy flask with a glass jacket was used as the reaction vessel. The flask was charged with tetrahydrofuran (57 parts), ethyl iodide (12 parts), nickel iodide (0.68 parts), molybdenum hexacarbonyl (1.4 parts), lithium iodide (13.6 parts), and methanol (16 parts), purged with argon, and pressurized to 400 psig with carbon monoxide. The flask was heated to 180°C with stirring, and then the flask was charged with ethylene to bring the pressure up to 800 psig. The pressure was maintained at 800 psig by recharging carbon monoxide and ethylene in equal amounts as needed, and the temperature was maintained at 180°C. After ? hour of reaction, G.C. analysis of the reaction effluent showed ? that methyl propionate was produced at the rate of 1.75 moles per liter per hour.

EXAMPLE 16

Example 15 was repeated except that the molybdenum hexacarbonyl was replaced with an equal amount of tungsten hexacarbonyl. Methyl propionate was found to have formed at a rate of 0.55 moles per liter per hour.

CLAIMERS

1. A process for the preparation of a carboxylic acid, a carboxylic acid anhydride, or a carboxylic acid ester, comprising reacting an olefin with carbon monoxide in the presence of water, a carboxylic acid, and/or an alcohol, in the presence of a halide, and in the presence of a catalyst, wherein the catalyst comprises a molybdenum-nickel-alkali metal, a tungsten-nickel-alkali metal, or a chromium-nickel-alkali metal co-catalyst component.

2. A process as defined in claim 1, wherein the co-catalyst component comprises molybdenum-nickel-alkali metal.

3. A process as defined in claim 1, wherein the alkali metal is lithium

4. A process as defined in claim 3, wherein the co-catalyst comprises molybdenum nickel lithium.

5. A liquid phase carbonylation catalyst comprising a molybdenum-nickel-alkali metal, tungsten-nickel-alkali metal, or chromium-nickel-alkali metal co-catalyst component and a halide component represented by the following formula: X:T:Z:Q, wherein X is molybdenum, tungsten, or chromium, T is nickel, X and T being in the zero-valence form, or in the form of a halide, an oxide, a carboxylate of 1 to 20 carbon atoms, a carbonyl, or a hydride; Z is a halide source which is a hydrogen halide, halogen, an alkyl halide, where the alkyl group contains 1 to 20 carbon atoms, or an alkali metal halide, and Q is the alkali metal component, and in the form of an iodide, bromide, chloride, or carboxylate as defined for X and T, the molar ratio of X to T being 0.1 -10:1, the molar ratio of X T to Q being 0.1 -10:1, and the molar ratio of Z to X T being 0.01 -0.1:1.