CA2081942C - Production of alkenyl alkanoate catalysts - Google Patents

Abstract

This invention provides a process for producing improved catalysts for the production of alkenyl alkanoates by the reaction of an alkene, an alkanoic acid and an oxygen-containing gas. The catalysts contain palladium, gold and a potassium promoter and are characterized by a reduced sodium content which results in increased catalyst activity. The reduced sodium content is obtained by using essentially sodium-free starting materials in the process for producing the catalysts.

Description

ao81942 PRODUCTION OF ALKENYL ALKANOATE CATALYSTS

FIELD OF THE lNv~NlION
This invention relates to a process for producing catalysts (hereinafter referred to as ~alkenyl alkanoate catalysts") for the production of alkenyl alkanoates from alkenes, alkanoic acids and an oxygen-containing gas.

DESCRIPTION OF THE RELATED ART
Processes for producing alkenyl alkanoate catalysts are known. By way of illustration, British Patent 1,215,210 (National Distillers) discloses a process for the production of olefinically unsaturated carboxylic esters (e.g., vinyl acetate) comprising reacting an olefinically unsaturated compound, a carboxylic acid and oxygen in the presence of a catalyst containing palladium metal and platinum metal and activated with at least one alkali metal or alkaline earth metal hydroxide or organic acid salt or inorganic acid salt.
The alkenyl alkanoate catalysts of National Distillers are produced by: (1) dissolving salts of the metals in conventional solvents, (2) spraying the solutions on an inert carrier or soaking the inert carrier in the solutions, (3) removing the solvent, (4) converting the salts so deposited on the carrier to the free metals: (a) by thermal decomposition, ~.

(b) by reduction with hydrogen or (c) by reduction in suspension using aqueous alkaline formaldehyde, aqueous hydrazine or aqueous or alcoholic sodium borohydride, (S) washing the catalyst with water to remove chlorides (see the National Distillers Examples) and (6) activating the catalyst. In the National Distillers catalyst preparation procedure, there is no precipitation of the metal salts on the carrier prior to their conversion to the free metals.
The National Distillers' alkenyl alkanoate catalysts are activated with a minor amount of at least one alkali metal or alkaline earth metal hydroxide or organic acid salt or inorganic acid salt. The alkali metal or alkaline earth metal salts of weak acids, both organic and inorganic acids are stated to be especially useful as activators.
Sodium, lithium, potassium, rubidium and cesium salts and mi~tures thereof are stated to be most effective as activator and the use of sodium and potassium salts, e.g., sodium and potassium acetates is especially preferred. The salts may have such anions as citrate, acetate, borate, phosphate, tartrate, benzoate or aluminate. National Distillers discloses that alkali metal and alkaline earth metal hydroxides are also effective activators and that the use of halide anions should preferably be avoided, since the presence of halides is stated to deleteriously affect the synthesis reaction.
In both of the National Distillers' Examples the salts were reduced with hydrogen and washed with water to remove chlorides. Then the-catalysts were treated with aqueous solutions containing both sodium acetate and potassium acetate and dried in a rotary 2o~l9~2 evaporator and then under vacuum. Based on the amount of sodium acetate used, the resulting catalysts contained about 0.23 weight percent sodium. The most active catalyst of the National Distillers' E~amples (i.e., the catalyst of E~ample 2H) is disclosed as having an activity of 8.3 grams of vinyl acetate per gram of palladium per hour (equivalent to about 165 grams of vinyl acetate per liter of catalyst per hour, assuming a catalyst density of one gram per milliliter).
As another illustration, Journal of Catalysis, volume 17, pages 366 to 374, 1970 (Nakamura et al.) discloses vinyl acetate catalysts produced by impregnating a carrier (calcined alumina) with an aqueous solution of palladium chloride, evaporating to dryness, reducing with an alkaline hydrazine hydrate solution, washing with distilled water to remove chloride ions, impregnating with a metal salt solution (e.g., a potassium or sodium acetate solution) and drying. Nakamura et al.
reports that impregnating with potassium acetate results in a catalyst activity of 25 grams of vinyl acetate per hour per liter of catalyst while impregnating with sodium acetate results in a catalyst activity of 19 grams of vinyl acetate per hour per liter of catalyst.
As a further illustration, U. S. Patent 3,743,607 (Sennewald et al.) discloses a process for making vinyl acetate from ethylene, acetic acid, and molecular o~ygen or air in the gas phase. A mi~ture of these reactants is passed in contact with a supported catalyst containing metallic palladium, an alkali metal formate or acetate, and metallic gold.

, The vinyl acetate catalysts of Sennewald et al. are produced by impregnating a catalyst carrier with an aqueous solution of a palladium salt and a gold salt and the resulting mixture is evaporated to dryness. The mass so obtained is introduced into an aqueous solution containing an appropriate reducing agent (e.g., hydrazine) that is capable of reducing both the palladium and gold salts to the metallic state. In the Sennewald et al. catalyst preparation procedure, there is no precipitation of the salts on the carrier prior to the reduction. Once the reduction is complete, the catalyst mass is removed from the liquid by filtration and washed with water.
When the reduction is achieved by means of a reducing agent free of alkali (e.g., hydrazine), the catalyst is conveniently impregnated with an about 10%
solution of sodium acetate. The formates or acetates of lithium or potassium can also be used. The catalyst is then dried and is ready for use.
Sennewald et al. discloses that, in the absence of such treatment, despite the gold it contains, the catalyst is found to have a substantially lower activity (e.g. an activity of only 15 grams vinyl acetate per liter of catalyst per hour) instead of the activity of 50 to 120 grams vinyl acetate yield per liter of catalyst per hour disclosed in Sennewald et al. for the Sennewald et al. catalysts. Catalysts which have been reduced by means of a composition comprising sodium formate and formic acid are disclosed to be active, even if no sodium acetate has been added thereto.
Sennewald et al. states that it has unexpectedly been found that the vinyl acetate _ ~ 5 ~ ~I9 space/time yields and, more particularly, the lifetime of the supported catalyst until regeneration thereof, can be substantially increased when the catalyst is impregnated with a solution prepared from a mi~ture of various acetates of sodium, potassium, rhodium or cesium instead of impregnation with a solution of a single alkali metal acetate.
The highest activity reported in the Sennewald et al. E~amples is in Example 11 where a catalyst was impregnated with sodium and potassium acetate as described in E~ample 9 of Sennewald et al. To effect the impregnation, the catalyst was introduced into a solution of the acetates, the supernatant solution was decanted and the catalyst was dried. The catalyst so obtained in Example 11 of Sennewald et al. is reported to contain about 0.8 %
sodium and to have an activity of 146 grams of vinyl acetate per liter of catalyst per hour. The other Sennewald et al. Examples (including E~amples 4, 5 and 12(e) where the catalysts were apparently substantially free of sodium) reported even lower catalyst activities than E~ample 11. The catalysts of Sennewald et al. Examples 4 and 9 were reported to have the same palladium and gold contents and appro~imately the same activities. The catalyst of Example 4 is reported to contain 2.5% potassium (as potassium acetate) while the catalyst of E~ample 9 is reported to contain about 1.8% sodium (as sodium acetate).
U.S. Patent 3,822,308 (Kronig et al.) discloses that particularly active supported catalysts containing palladium and gold for the production of vinyl esters from ethylene, lower _ - 6 - 20819~2 carbo~ylic acids with 2 to 4 carbon atoms and oxygen in the gas phase at elevated temperature and at normal or elevated pressure are obtained by the following procedure: The catalyst support is treated, simultaneously or successively, with or without intermediate drying, with a solution ("Solution A") containing dissolved salts of palladium and gold and, optionally, salts of other metals, and another solution ("Solution B") containing compounds (hereinafter referred to as precipitating agents") which are able to react on the catalyst support with the noble metal salts of the Solution A to form water-insoluble noble metal compounds which are practically free from halogen, sulphur and nitrogen. Solutions A and B (separately or in combination) are used to impregnate the catalyst support in quantities which correspond to from 10 to 110% of the absorptive capacity of the catalyst support for these solutions. The catalyst support impregnated with Solutions A and B is subjected to a time/temperature treatment such that at least 9S% of the impregnated palladium and at least 95% of the impregnated gold are transformed into water-insoluble noble metal compounds. The water-insoluble noble metal compounds are largely transformed by treatment with reducing agents into the noble metals and the water-soluble compounds which are contained in the catalyst are removed by washing, before or after the reduction.
In a preferred embodiment of the Kronig et al. process, alkali metal carbo~ylates (especially alkali metal acetates) are applied on the catalyst - 7 - 20~1 9~2 -before or after the treatment with reducing agents in such quantities that the catalyst, after being dried, contains rom 1 to 30% by weight of alkali metal carboxylate. Examples of the alkali metal carbo~ylates disclosed in Kronig et al. include sodium formate, potassium acetate, sodium acetate, lithium acetate, potassium propionate and potassium butyrate.
The Kronig et al. Examples employing precipitating agents report catalyst activities markedly higher than the activities reported by National Distillers and Sennewald et al. where no precipitating agents are used. Thus, Kronig et al.
Example 1 employs sodium hydro~ide as a precipitating agent and potassium acetate as a promoter and reports a activity of 952 grams of vinyl acetate per hour per liter of catalyst. Kronig et al. E~ample 3 employs potassium carbonate as a precipitating agent and an "alkali metal acetate" as a promoter and reports that the results obtained with the catalyst were comparable to those of Kronig et al. Example 1.
U. S. Patent 4,048,096 (Bissot) discloses a catalyst having a specific activity of at least about 83 grams of vinyl acetate per gram of precious metal per hour measured at 150C. The ~issot vinyl acetate catalyst consists essentially of: (1) a catalyst support having a particle diameter of from about 3 to about 7 mm and a pore volume of from about 0.2 to about 1.5 ml.~g., a 10% by weight water suspension of the catalyst support having a pH of from about 3.0 to about 9.0; (Z) a palladium-gold alloy distributed in a surface layer of the catalyst support, the surface - 8 - 2 081 9~2 layer e~tending less than about 0.5 mm from the surface of the support, the palladium in the alloy being present in an amount of from about 1.5 to about 5.0 grams per liter of catalyst, and the gold being present in an amount of from about 0.5 to about 2.25 grams per liter of catalyst, and (3) from about 5 to about 60 grams per liter of catalyst of alkali metal acetate. Bissot discloses that the palladium is the active catalyst metal and the gold is a catalyst promoter.
Bissot also discloses a process for preparing the Bissot catalyst. Like ~ronig et al., the Bissot process involves precipitation of the metal salts on the catalyst support. The Bissot process comprises: (1) impregnating the catalyst support with aqueous solution of water-soluble palladium and gold compounds, (2) precipitating water-insoluble palladium and gold compounds on the catalyst support by contacting the impregnated catalyst support with a solution of compounds (preferably sodium metasilicate) capable of reacting with the water-soluble palladium and gold compounds to form water-insoluble palladium and gold compounds, (3) converting the water-insoluble palladium and gold compounds into palladium and gold metal on the support by treatment with a reducing agent, (4) washing the catalyst with water, (5) drying the catalyst (see E~ample 1 of Bissot), (6) impregnating the catalyst with alkali metal acetate promoter (e.g., a potassium promoter), and (7) drying the catalyst.

- 20819~2 The improvement disclosed in Bissot involves distributing the palladium and gold as an alloy in a surface layer of the catalyst support, the surface layer e~tending less than about 0.5 millimeter from the surface of the support. The impregnating step is carried out with an aqueous solution of palladium and gold compounds and the total volume of the solution is from about 95 to about 100% of the absorptive capacity of the catalyst support. The precipitating step in Bissot is carried out by soaking the wet catalyst support with a solution of an alkali metal silicate, the amount of alkali silicate being such that, after the alkali metal silicate solution has been in contact with the catalyst support for about 12 to 24 hours, the pH of said solution is from about 6.5 to about 9.5.
Bissot does not report the sodium content of the catalysts of the Bissot Examples. Bissot Example 1 reports that the catalyst of that Example had an activity of 560 grams of vinyl acetate per liter of catalyst per hour. In Example III below, two catalysts produced following the disclosure of Example 1 of Bissot were found to have sodium contents of 0.32 and 0.38 weight percent and activities of 551 and 535 grams of vinyl acetate per liter of catalyst per hour.
Despite the foregoing prior art processes, it is desirable to further improve the activity of alkenyl alkanoate catalysts.

- 10 - ' 2~8~9~'2 SUMMARY OF THE INVENTION
This invention is based, in part, on the discovery that the activity of alkenyl alkanoate catalysts is increased if their sodium content is decreased by employing essentially sodium free starting materials in producing the catalyst.
More specifically, this invention provides a process for producing a catalyst that is useful in catalyzing the reaction of an alkene, an alkanoic acid and an o~ygen-containing gas to produce an alkenyl alkanoate and that comprises support particles which are capable of e~changing cations and which are impregnated with palladium, gold and potassium acetate:
said process comprising the steps of:
(a) impregnating the support particles with aqueous solutions of water-soluble palladium and gold compounds;
(b) precipitating water-insoluble palladium and gold compounds onto the support particles from such solutions using a precipitating agent;
(c) converting the precipitated water-insoluble palladium and gold compounds to palladium and gold on the support particles using a reducing agent;
(d) washing the support particles with water;
(e) drying the support particles;
(f) further impregnating the support particles with a potassium promoter; and (g) drying the impregnated particles to produce the catalyst;

20819~2 and said process being conducted by employing essentially sodium-free starting materials in steps (b)and (c) so as to reduce the amount of sodium in the catalyst and thereby to increase the activity of the catalyst.

BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 shows the predicted effect of sodium on the performance of vinyl acetate catalysts produced in accordance with this invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS
In the practice of the process of the present invention, sodium-containing, water-soluble palladium and/or gold compounds can usually be employed since they are usually not used in amounts that result in catalysts having substantial amounts of sodium. The principal sources of sodium in alkenyl alkanoate catalysts are sodium-containing precipitating agents (e.g., sodium metasilicate) and/or sodium-containing promoters or activators (e.g., sodium acetate). To a lesser e~tent, some supports and some reducing agents (e.g., sodium borohydride) can introduce substantial amounts of sodium into the catalyst. Accordingly, in the practice of this invention, essentially sodium-free precipitating agents (e.g., potassium hydro~ide), promoters (e.g., potassium acetate), reducing agents (e.g., hydrazine) and carriers are employed. When using potassium hydro~ide as the precipitating agent, a suitable potassium salt (e.g., potassium acetate) can also be used in the precipitating step to aid in displacement by potassium of any sodium bound on the - 12 - 20319~2 carrier. Preferably the potassium hydroxide and the potassium salt are employed in an aqueous solution.
The salt is used in an amount that provides from 1 to 10 weight percent potassium based on the total weight of the solution. Care should be e~ercised to ensure that the resulting catalyst does not contain so much potassium that catalyst activity is less than desired.
The support particles used in the process of this invention are solid particulate materials that are capable of e~changing cations (e.g., due to the presence of SiOH or AlOH groups), that are capable of being impregnated with palladium, gold and a potassium promoter and that are inert under the conditions used to produce alkenyl alkanoates.
Illustrative of such support particles are particulate silica, alumina, and silica-aluminas.
Silica is the preferred support. The support preferably has a surface area from 100 to 800 square meters per gram.
The aqueous solutions of water-soluble palladium and gold compounds used in the process of this invention include aqueous solutions of any suitable palladium or gold compound such as palladium (II) chloride, sodium tetrachloropalladium (II) (Na2PdC14), palladium (II) nitrate, palladium (II) sulfate, gold (III) chloride or auric (III) acid (HAuC14). The volume of the solution preferably corresponds to from 95 to 100% (more preferably from 98 to 99%) of the pore volume of the support.
The precipitating agents used in the process of the present invention include lithium and - 13 _ i 2 0~1 9q 2 potassium silicates and hydroxides. The precipitating agents are preferably employed in the form of aqueous solutions containing a 1.6 to 1.8 molar excess of the precipitating agents. The volume of such solution used is preferably just sufficient to cover the support particles. To avoid possible degradation of the support, the weight ratio of the precipitating agent to the support should not be too high. By way of illustration, a weight ratio of potassium hydro~ide to support of about 0.08:1 resulted in no noticeable degradation of the support.
The reducing agents used in the process of this invention include ethylene, hydrazine, formaldehyde and hydrogen. The reducing agents are preferably employed in the form of aqueous solutions containing a 50:1 (or more preferably a 10:1) molar e~cess of the reducing agents. If hydrogen is used, it is usually necessary to heat the catalyst to 100 to 300C to complete the reduction.
The potassium promoters used in the process of this invention include potassium alkanoates and any potassium compound that is converted to a potassium alkanoate during the alkenyl alkanoate-forming reaction (i.e., the reaction of ethylene, an alkanoic acid and an oxygen-containing gas in the presence of the catalyst to produce an alkenyl alkanoate). Suitable potassium compounds include potassium acetate, bicarbonate, nitrate and (when a stable support is used) hydroxide. The promoters are preferably applied in the form of aqueous solutions.
The washing step in the process of this invention can be conducted batchwise or ~-16818-2 ` - 14 - 20~19~2 continuously. Continuous washing is more efficient but may not be most suitable for large scale (e.g., plant scale) catalyst production. In continuous washing, the wash water is slowly and continuously passed through the catalyst over a period of time (e.g., from 8 to 24 hours). In batch washing, the catalyst is contacted with the wash water, the mi~ture is allowed to stand (e g., for from 0.5 to 2.0 hours) and the water and catalyst are separated.
In batch washing, several such washes (e.g., from 2 to 10, or preferably from 4 to 6 washes) are often required to reduce the impurity (e.g., halide) content of the catalyst to the desired level.
Temperatures from 20C to 80 and volume ratios of wash water to catalyst of from 2:1 to 100:1 can be used in either batch or continuous washing. The washing step removes certain impurities, particularly chlorides, from the catalyst.
The drying of the catalyst in accordance with step (e) and step (g) of the process of this invention for producing alkenyl alkanoate catalysts can be conducted in any convenient manner. By way of illustration, drying can be conducted at 40C to 120C in a forced air oven for 15 to 30 hours.
The catalyst produced by the process of this invention are useful in catalyzing the reaction of an alkene, an alkanoic acid and an oxygen-containing gas to produce an alkenyl alkanoate and comprise support particles which are capable of exchanging cations and which are impregnat~ed with precipitated and reduced palladium and gold and a potassium promoter, any sodium in the catalyst being preferably present in an _ - 15 _ 20819~2 amount no more than 0.3 weight percent based on the total weight of the catalyst.
The catalysts produced by the process of this invention preferably have a palladium content of greater than 0.25 weight percent based on the total weight of the catalyst, more preferably greater than 0.5 weight percent based on the total weight of the catalyst and most preferably from 0.5 to 1.7 weight percent based on the total weight of the catalyst.
It is preferred that the gold to palladium weight ratio of the catalyst is from 0.2 to 1.5 and, most preferably, from 0.4 to 1.2 The catalysts produced by the process of this invention desirably contain no more than 0.3 weight percent sodium based on the weight of the catalyst. Preferably the catalysts contain no more than 0.2 weight percent sodium and, most preferably, the catalysts contain no more than about 0.1 weight percent sodium based on the weight of the catalyst.
The amount of sodium in the catalysts of this invention will depend primarily on the particular starting materials used and, to a lesser extent, on the number of washes, the volume of wash water and the total washing time.
It is preferred that the catalysts produced by the process of this invention are shell impregnated catalysts wherein a catalyst support has a particle diameter from about 3 to about 7 millimeters and a pore volume of 0.2 to 1.5 milliliters per gram. The palladium and gold are preferably distributed in the outermost 1.0 millimeter thick layer of the catalyst support particles. The catalysts preferably contain from - 16 - 2~ 21 9~ 2 about 1.4 to about 3.8 weight percent (and more preferably from 2 to 3.6 weight percent) of the potassium derived from the potassium promoter.
The catalysts produced by the process of this invention are characterized by their increased catalytic activity. Typically the activity of the catalysts is 5% to 25% greater (in terms of quantity of alkenyl alkanoate produced per unit of catalyst per unit time) than otherwise identical catalysts containing from over 0.3 to about 1.0 weight percent sodium. Although catalyst selectivity (i.e., the tendency to produce alkenyl alkanoates rather than by-products such as carbon dioxide) declines somewhat with decreasing sodium content, that disadvantage is more than offset by increased catalyst activity, particularly in the range of sodium contents found in commercial alkenyl alkanoate catalysts (e.g., up to about 1.0 weight percent sodium).
The process for producing alkenyl alkanoates using the above-described catalysts ("alkenyl alkanoate process") comprises reacting an alkene, an alkanoic acid, and an oxygen-containing gas in the presence of a catalytic amount of catalyst of this invention as described above. The alkenyl alkanoate process is preferably conducted at a temperature from 100C to 250C (and most preferably at a temperature from 140C to 200C) and at a pressure from 15 psi to 300 psi (most preferably at a pressure from 90 pounds per square inch to 150 pounds per square inch.) The alkenyl alkanoate plrocess is preferably conducted continuously in the vapor phase.
Preferred alkanoic acid starting materials used in the alkenyl alkanoate process contain from - 17 - a O ~ 2 two to four carbon atom~ (e.g. ethylene, propylene and n-butene). Preferred products of the alkenyl alkanoate process are vinyl acetate, vinyl propionate, vinyl butyrate, and allyl acetate.

The alkenyl alkanoates produced by the alkenyl alkanoate process are known compounds having known utilities (e.g., vinyl acetate is useful in producing polyvinyl acetate).

, . ~
-L~

2~1942 EXAMPLES
In the following E~ample, the following abbreviations are used:
Abbreviation Meaning Support I Silica beads having an average diameter of 5 to 6 millimeters and containing about 0.1 weight percent sodium. The beads have a surface area from lS0 to 200 square meters per gram and a pore volume from 0.6 to 0.7 milliliters per gram.
Support I contains SiOH
groups that are capable of e~changing cations. Support I is sold by Sud Chemie AG
as "KA-160"
STY* Space Time Yield (a measure of catalyst activity) expressed as grams of vinyl acetate per liter of catalyst per hour.
% Selectivity* Selectivity was calculated as follows: Selectivity =
100 X (moles vinyl acetate)/(moles vinyl acetate + 1/2 X moles CO2) AA Analysis Atomic Adsorption Spectroscopy ICP Inductively Coupled Plasma Optical Emission Spectrometry g VA/l cat/hr Grams of vinyl acetate produced per liter of catalyst per hour *All the values for activities and selectivities reported in the E~amples appearing below are based on the activities and selectivities measured twenty-six hours after full oxygen feed was reached in the Catalyst Test Method described below.

20~1942 .
VA vinyl acetate KOAc potassium acetate EcOAc ethyl acetate NaOAc sodium acetate % percent by weight g grams ml milliliter mm millimeter hrs hours min minute In the following Examples, the following procedures were used:
Catalyst PreParation Procedure Support I (15 g) was added to a solution of Na2PdC14 (35.86% Pd, 0.258 9) and HAuC14 (48.95% Au, 0.094 9) dissolved in 9.0 ml of deionized water. The mixture so formed was gently agitated until all of the moisture was absorbed into the support and then was allowed to stand in a sealed flask for about one hour at room temperature so as to impregnate the support with the palladium and gold salts. The damp catalyst was covered with a solution of potassium hydroxide (0.371 g in 28 ml water) as a precipitating agent. Potassium acetate is used in conjunction with potassium hydroxide in the precipitation step in a preferred embodiment of this invention. After mi~ing for a few seconds, the mi~ture was allowed to stand covered and undisturbed for 23 hours at room 2 0 8 ~ 9 ~ 2 -temperature to deposit water-insoluble palladium and gold compounds on the support. The palladium and gold were then reduced by the addition of 1.0 g of 85% hydrazine hydrate to the above mixture. The mixture was agitated for a few seconds and allowed to stand covered and undisturbed at room temperature for another 23 hours. The supernatant liquid was decanted from the catalyst and the catalyst was rinsed four times with water to remove the small amount of metal sludge present. The catalyst was washed thoroughly by the Column Washing Procedure described below to remove chlorides and residual reagents. The catalyst was dried on a stainless steel screen at 60C in a forced air oven for 20 to 24 hours. The catalyst was analyzed for potassium using AA Analysis. Then the catalyst was impregnated with the desired amount of potassium acetate in water using the impregnation technique described above for the palladium and gold salts. Then the impregnated catalyst was dried at 60C for 20 to 24 hours. The palladium, gold, sodium, and potassium contents in the finished catalysts were determined by ICP
analyses and the sodium and potassium contents were also determined by AA Analyses for greater accuracy.
The catalysts so produced were shell-impregnated (i.e., substantially all of the palladium and gold was present in a shell within 0.5 mm of the surface of the beads of Support I).

Column Washinq Procedure Catalysts were washed or rewashed in a 1.24 inch o.d. g 24 inches glass chromatography ~-16818-2 ~ - 20819~2 column fitted with a Teflon~ stopcock. Typically 15 g of catalyst was added to the column which was then filled with water. The stopcock was adjusted to allow the liquid to flow from the column such that about one liter passed through the catalyst at room temperature over a period of about 24 hours. After this period, the e~cess liquid was drained from the column and the catalyst removed and dried as described above in the Catalyst Preparation Procedure.

Catalyst Test Method The catalyst ~2.5 g samples of 5 to 6 mm catalyst spheres) was diluted with 10.5 ml of 0.5 mm glass beads and the mi~ture was uniformly distributed in both legs of a 316-stainless steel U-tube reactor. The reactor had an outside diameter of 3/8 inch and an overall height of about 6 inches. An ethylene flow of 151 ml/min. was started through the reactor after which the catalyst was heated in an oven maintained at 150C while allowing the system to pressurize to 115 psig. After maintaining at these conditions for 1.5 hours, acetic acid vapor was added to the ethylene and the mi~ture was passed over the catalyst for 45 minutes. Air was added to the feed gas at a slowly increasing rate over a 45-minute period until a total flow of 105 ml./min. was reached. The catalyst was allowed to stabilize for two hours before beginning data collection. The final gas composition was ethylene:acetic acid:o~ygen:nitrogen e 52.9:10.7:7.7:28.7, the total gas hourly space velocity was about 3800 hrl, and the acetic acid liquid hourly space velocity was about 1 20~1942 -hr~l. The product was analyzed by gas chromatography. The run-to-run reproducibility of the microreactors used in these experiments is about ~10 STY units.

EXAMPLE I
Support I contains about 0.1 weight percent sodium as received from the manufacturer. An additional 0.4 to 0.8 weight percent of sodium is introduced during the precipitation step when sodium hydro~ide or sodium metasilicate is used as the precipitating agent. Three catalysts were prepared using potassium hydroxide as the precipitating agent to reduce the sodium level in the final catalyst.
Varying concentrations of potassium acetate were added to the precipitating solution to further shift the ion-e~change equilibrium toward potassium. The catalysts were prepared from the same master batch having a nominal palladium loading of 0.58 weight percent and an Au/Pd ratio of 0.46. After reduction with hydrazine and washing using the Column Washing Procedure, the catalysts were analyzed for sodium and potassium. Then additional potassium acetate was added as required to give a final potassium acetate content of about 5.3 weight percent. The results given in Table A show that the sodium content of the catalysts was reduced. The sodium content in the catalysts dropped uniformly and catalytic activity improved as the potassium concentration of the precipitating solution was increased.

23 20~19~2 Table A

Low-Sodium Catalyst PreParations % KOAc (a) % Na (b) ~ % Selectivity 0 0.186 576 93.4 2.5 0.121 603 93.2 5.0 0.091 597 93.4 (a) Wt % potassium acetate added to the precipitating solution.
(b) Wt % sodium calculated to be in the finished catalyst. Based on AA Analyses prior to KOAc impregnation and the quantity of KOAc added.

A similar catalyst, prepared using sodium metasilicate as the precipitating agent (in lieu of potassium hydro~ide) had an STY of 544, a selectivity of 93.6 and a sodium content of 0.44 weight percent.
No potassium acetate was added in the precipitation step.

E~ample II
The effect of sodium on the performance of vinyl acetate catalysts was studied using statistically designed e~periments and models were _ - 24 - 20819~2 obtained which are useful in predicting the performance of the vinyl acetate catalysts produced by the process of this invention as well as vinyl acetate catalysts produced by the processes of the two above-mentioned U.S. patent applications filed concurrently herewith. The models predict catalyst activity and selectivity as a function of sodium content, palladium loading, gold to palladium weight ratio, potassium content and catalyst weight. These models and the data from which they were generated are shown in Tables B and C respectively.
Because the degree of conversion has a major effect on both catalyst productivity and selectivity, meaningful comparisons of catalyst variables can only be done at constant conversion. In order to predict the effects of catalyst composition at constant conversion, the Oxygen Conversion Model in Table B
was rearranged to e~press catalyst weight as a function of the palladium content, gold~palladium ratio, potassium content, sodium content and conversion. This catalyst weight term was then used to replace the catalyst weight terms in the STY and Selectivity models. The predicted effects of increasing sodium content on vinyl acetate catalyst activity and selectivity are plotted in Figure 1.

The abbreviations used in Tables B and C
have the following meanings:

Pd Weight percent palladium in the catalyst Au/Pd Weight ratio of gold to palladium in the catalyst Cat.Wt Catalyst weight in grams K Weight percent potassium in the catalyst Na Weight percent sodium in the catalyst STY Space time yield in grams of vinyl acetate per liter of catalyst per hour R2 Correlation coefficients which are indicative of the quality of fit of the data to the models RSD Relative standard derivation EtOAc By- Production of ethyl acetate in moles/
Product Rate kilogram of catalyst/hour ~ Heavies Heavy by-products expressed as a weight By-Products percent of the vinyl acetate produced.
in VA Heavies by-products are defined as all products which elute after acetic acid in the gas chromatographic analytical procedure.
Table D shows the effect of varying sodium on catalyst activity as predicted by the models in Table B.

2~819~2 o o o o o o ll ll ll ll ll ll ~ R ~ Ei ~ ~

.
~ o~ , ,.
O _ ,~, _ ~ -r U~ I ~ ~ 0 l o ` _ o O V N _ ~
_ o O O
e ~ N
O I ^ ~ O ~ -- N
-t y U') -- ' O
N
_ o ~ 0 Z O O ~ O
+ ~ y ~ l cJ~~ `O
~ V~ ~ 0 ~ ~ ^ O 0 J CY J 0 q` 0 0 U7 0 0 U~ O
~: O ~ O
I-J V I I -- I ~ ~ CL N N
-- N -- N ^
Z ~) ~ 0 ~ ~ NN O
~ ~ . o o (~ ~y~ 0 -- O O
O O O I I O ~ -- I +
I C t~ N + N ^ O

Il~ ~O + ~ ~ .O _ O -- . .

o ~~ N N N ~ V
I` U~ ~ N
~ O ^ + O ' 'O O O

O o ~ 0 ~~ ,~ 0 ~ '~
~ _ ~) O O
+-- l ~ +
.. C ~, ~ ' L
¦ ~ ~ ~ r~
t ~ t~ ~ $ r G C~ 10 . 1~ ~

-~ c ~, ~ o ` o o " o "

o ~ ~ *

o U
o O~ L
i ~ ~ . ~

3 _ o ~ _ ,, , o C ~J C

., ,~, U
^ . .
o y ~ C
U~ ~ o _ O
,~
~ , ~ _ o ~ ~ ~ o c I
o ~ ~ . , o Il + ~ , o ~ ~ I ' . O

,~ ~ o ~
~o ~ c o ~ U
, ~ ~ U

-~, .............................
a U~
o~o ~o ~n ~ ~ ~ ~ ~ ~ ~ o ~ o o ~r ~ ~ ~ ~ o ~ ~ ~ ~ ~ ~ ~ ~ co ~ u~
..............................

o -~ ~ ~ ~r Ln u~ ~ ~ o u~ ~D O ~ O ~D ~ O ~ ~D ~ ~ ~ ~ ~ r~l ~ ~ ~ o ~ u~ u~ o ~ a v ..............................
1~5 ~ I O O ~ O O N ~ N O O ~ O O ~1 (`I N ~ --1 ~ ~1 ~ ~ ~ ~ O O r~l Z ~>
E~
U~
Z ..............................
o~C OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

~ o o o o o o o o ~ ~ ~ N C~ ~1 O~q tl- O O O ~ ~ t` O O u~
.........................

O O O L~ 1 ~ ~I CO 1~ 0 ~n ~ ~I N CO CO ~` t~ ~` t` ~0 'C 1 ) ~ o u~ ~1 It'1 o o In In O 11') ~ Il~ O ) C~ t` ~` O U') O O Ir) O U~
..............................
OOO_~OO~O~O~_~OO~O~O_IOOOO~O_I~O~O

`-- 208~94~

H
J- H
~q J~ , .,1 ~ 1 U
~r H
O ~ ~ Il~ d ~ ~` ~ r ~ u~ ~ ~1 ul ~ ~ aD ~ ~ ~ ~ ~ u~ ~I ~.
~, .......... ............... J~

o\
~ Oq 1 ,,~ t' ~ IS~ ~D C~ al co o 1~ OD 0 0 ~ ~1 ~ O ao ~1 ~ ~ In o a, 0 0 0 ~ U h ~1 ~ U. ~ "
, r _ ~
O ~ ~0 a V aD ~1 o ~) ~1 ~ ~ It) a~ al ~ o ~ a) a~ N ~) ~ ^
O o ~ ~ ~ ~ ~1~ r ~ ~ _ ~D OD ~ o ~ ~ r ~ ~ r 1 0\o ~ In u~ d ~ ~ ~ ~ In ~ ~ ,~ d d ~ d d d d d d d d ~ d d d ~ ~ ul Id u -- ~ r~ r~ ~ ~ ~D ~ d OD ~ Ln ~ u C) O O ~ ~1 ~ r-1 ~ ~1 0 N ~ ~`1 ~ (~ ~ ~'1 C`~ ~ ~'1 ~`1 C`l ~ ~ ~ ~ ~ - R O ~ I
ai ,, ~
P o ~ c~ o o a~ O ~ ~1, - ~ r-a ~o ~ n ~ ~ d ~ P d ~1 ~ ~ ~ ,1 O~ ~ ~ ~ ~ ~ ~n ~1 0 ,~
E z .......... ~.........................
o\o oooooooooo ,oooooooooo~oo~ ~
~ J ~C
a) a ~ ~
R ~5 O
~D ~` ~ ~ O O O O O O cn ~ aD aD ~ ~ a~ ~ o ~ ~ u~ ~ /V Ul o\l a3~0 r ~ I u.
~ ~ V ~r a rc ~ ~ I
o~ ~,~a~ r .......... ...............
aC oooooooooo ooooooooOOOOOOO,u,~
c U Ul U ~ ! a U~ -I
.~ I
d aD O O ~ ~1 ~ ` r ~ 1~ d ~ D ~D d d 1~ d 1l~ o D t~ ul I O r- a u~ o ~ a:~ ~1 ~ r t~ r r ~ n u~ In ~ ~ n O P~ Ir ~V ~
~ ............. ............... ,Iva o\o ~~ ul~a~ R
a~ ~
~~ ~V ~ Ul F ~
. _ ~; ~ Id JJ
O --I ~ ~ d u~ o ~ ~ ~ ~ u~ ~o r a~ o ~ ~ ~ d u~ ~ G R ~
Z ~ ~ ~ r~ ~ ~ ~ ~ ~ d ~ ~ ~ ~ ~ d d d ~ Ln ~ u~ -~ ~ rJ

2Q8I S~2 TABLE D

Predicted Effect of Sodium on Catalyst* Activity % Na STY % Improvement**

0.1 665 0.0 0.2 649 2.5 0.3 633 5.1 0.4 618 7.6 0.5 603 10.3 0.6 581 14.5 0.7 574 15.9 0.8 560 18.8 0.9 546 21.8 * For a catalyst composition set at: O. 58% Pd, Au/Pd =
0.45 and 2.2% K and an o~ygen conversion set at 35%.

*~ Predicted percent change in STY resulting from decrease in sodium content from amount shown in first column to 0.1%.

- 31 - Z 081 9~ 2 EXAMPLE III
The procedure of E~ample 1 of U.S. Patent 4,098,096 (Bissot) was repeated as follows: Two preparations (Runs 1 and 2) were made, each employing 159 of Support I, 0.315 g of Na2PdC14, 0.085 g of HAuC14, 0.585 g of Na2SiO3.9H2O, 0.59 g of 85%
hydrazine hydrate, and 0.823 g of potassium acetate.
Since the e~act washing procedure is not disclosed in E~ample 1 of Bissot the catalyst of Run 1 was washed using the Column Washing Procedure for 16 hours using 23 ml of H2O per gram of catalyst whereas the catalyst of Run 2 was similarly washed but with 31 ml of H2O per gram of catalyst. The catalysts of Runs 1 and 2 were analyzed by ICP for palladium and gold and by AA Analysis for potassium and sodium. The e~perimental error of the sodium determination is estimated to be about + 0.01 relative percent. The catalyst of Run 1 was analyzed in duplicate. The results are shown in Table E.

TABLE E
%Pd %Au %K %Na ~Y Selectivity Run 1 0.544 0.201 2.34 0.32 550 93.9 Run 1 0.556 0.204 2.35 0.32 Run 2 0.552 0.195 2.34 0.38 535 93.7 Bissot 0.578* 0.242* 2.08* ---- 560** 93**

* Calculated based on the data in E~ample 1 of Bissot ** Disclosed in Example 1 of Bissot EXAMPLE IV
The measured activities of three commercial catalysts (Catalysts X, Y and Z) are shown in Table F. In Table F, the activities of Catalysts ~ and Y
are compared to the predicted activities of catalysts of this invention having the same composition (~Model Catalysts"). The predicted activities were determined from the models of Table B and assuming a 0.15% level of sodium. The Model Catalysts had markedly higher predicted activities. A similar comparison could not be made for Catalyst Z because its composition is outside the range of the models of Table B.
Catalyst X was prepared using the Catalyst Preparation Procedure. The preparation of Catalyst X
differed from the preparation of Catalyst Y in that, in the preparation of Catalyst Y: (1) the catalyst was dried before precipitation and (2) the precipitating agent used was sodium hydro~ide rather than sodium metasilicate. The reduction, washing, drying and potassium acetate impregnation steps were the sa~e for both catalyst preparations.
With reference to Catalyst Y, the suffi~es A, B and C in Table F denote different preparations ("lotsn) of nominally the same catalyst and, with respect to Catalysts Y and Z, the suffixes 1 and 2 denote duplicate analyses of different samples from the same lot of the catalysts.
Catalyst Z has a high palladium content and uses a cadmium co-catalyst rather than gold. Cadmium is significantly more toxic than gold. In addition, Catalyst Z is prepared by a process substantially 33 20819~2 different from the process of this invention, that is, Catalyst Z is prepared by impregnating a support with a solution of palladium, cadmium and potassium acetates and drying. There are no precipitation, reduction or washing steps in the process used to produce Catalyst Z.

TABLE F
%Pd %Au ~Ç~ %K %Na STY
Catalyst X~ 0.530.22 0 2.36 0.54 272 Model Catalyst 0.53 0.22 0 2.36 0.15 589 Catalyst YA-l* 0.490.19 0 2.29 0.60 360 Model Catalyst 0.49 0.19 0 2.29 0.15 546 Catalyst YA-2* 0.490.19 0 2.31 0.60 360 Model Catalyst 0.49 0.19 0 2.31 0.15 545 Catalyst YB* 0.630.24 0 2.27 0.70 386 Model Catalyst 0.63 0.24 0 2.27 0.15 669 Catalyst YC-l* 0.610.29 0 2.24 0.69 395 Model Catalyst 0.61 0.24 0 2.24 0.15 653 Catalyst YC-2* 0.610.26 0 2.18 0.70 395 Model Catalyst 0.61 0.26 0 2.18 0.15 658 Catalyst Z-l* 2.16 0 1.88 1.89 0.08 685 Model Catalyst Outside range of models Catalyst Z-2* 2.16 0 1.89 1.92 0.09 685 Model Catalyst Outside range of models *Comparative Catalysts

Claims (6)

1. A process for producing a catalyst that is useful in catalyzing the reaction of an alkene, an alkanoic acid and an oxygen-containing gas to produce an alkenyl alkanoate and that comprises support particles which are capable of exchanging cations and which are impregnated with palladium, gold and potassium acetate, said process comprising the steps of:
(a) impregnating the support particles with aqueous solutions of water-soluble palladium and gold compounds;
(b) precipitating water-insoluble palladium and gold compounds onto the support particles from such solutions using a precipitating agent;
(c) converting the precipitated water-insoluble palladium and gold compounds to palladium and gold on the support particles using a reducing agent;
(d) washing the support particles with water;
(e) drying the support particles;
(f) further impregnating the support particles with a potassium promoter; and (g) drying the impregnated particles to produce the catalyst; and said process being conducted by employing essentially sodium-free starting materials in steps (b) and (c) so as to reduce the amount of sodium in the catalyst and thereby to increase the activity of the catalyst.

2. A process as claimed in claim 1 wherein potassium hydroxide is used as the precipitating agent in step (b).

3. A process as claimed in claim 2 wherein potassium salt is used to aid in the displacement of sodium bound on the carrier by the potassium precipitating agent.

4. A process as claimed in claim 1 wherein the catalyst produced by the process has no more than 0.3 weight percent sodium based on the weight of the catalyst.

5. A process as claimed in claim 1 wherein the catalyst produced by the process has no more than 0.2 weight percent sodium based on the weight of the catalyst.

6. A process as claimed in claim 1 wherein the catalyst produced by the process has no more than 0.1 weight percent sodium based on the weight of the catalyst.