CA1047526A - Process for making 4-acetoxybutyraldehyde and 4-acetoxybutanol - Google Patents

Abstract

Abstract of the Disclosure Hydroformylation of 1-propenyl acetate under hydro-formylating conditions with a cobalt hydroformylation catalyst produces 4-acetoxybutyraldehyde as the predominant component of a mixture with its two isomers, 2-acetoxybutyr-aldehyde and 3-acetoxy-2-methylproplonaldehyde. Some reduction of these aldehydes to their corresponding alcohols also occurs. The yield of these alcohols can be increased to the point where they are the predominant products, either by increasing the reaction temperature, at least for the terminal portion of the reaction period, or by use of phosphine -modified cobalt hydroformylation catalyst.

Description

~o ~5Z6 RD-6469 _ AND 4-ACETOXYBUTANOL

BACKGROUND OF THE INVENTION
Field of the Invention This invention relates to a process for making 4-acetoxybutyraldehyde and 4-acetoxybutanol. More particular-ly, this invention relates to a process for making a mixtureof products which can range in composition from 4-acetoxy-butyraldehyde as the predominant product in a mixture also containing its two isomers, 2-acetoxybutyraldhyde and 3-acetoxy-2-methylprop~nylaldehyde, up to 4-acetoxybutanol as the predor,linant product in a mixture also containing 2-acetoxybutanol and 3-acetoxy-2-methylpropanol which comprises hydroformylating l-propenyl acetate under hydroformylating conditions in the presence of a cobalt hydroformylating catalyst. The choice of hydroformylation conditions and catalyst governs the composition of the products.
Description of the Prior Art ~ eaction of compounds containing aliphatic unsatu-ration with carbon monoxide and hydrogen to form aldehydes which can be hydrogenated, if desired, in either the same or a separate reaction to alcohols is known as the oxo process.
Some redu~tion of the aldehydes to their corresponding alcohols does occur under the mildest reaction conditions.

-1- ~
~P

~Q~7526 RD-6469 Where desired, the yields of the alcohol products can be in-creased to the point where they are the predominant product by increasing the reaction temperature at least for the terminal portion of the reaction period or by using a phos-phine modified cobalt hydroformylation catalyst A more specific adaptation of the oxo process is known as hydroformylation~ Strictly speaking, this latter term defines the addition of a hydrogen atom and a formyl group to the molecule of a compound containing a double bond by reaction with hydrogen and carbon monoxide, the chief product being one or more aldehydes. However, because both gases are present in excess, some hydrogenation of the formyl group to the methy~l group can occur. This latter reaction could more properly be referred to as hydromethyblation. How-ever, hydroformylation is used in the art and in this appli-cation to describe both reactions since it is the choice of conditions and/or catalyst that determines whether aldehydes, alcohols, or mixtures thereof will be the predominant product.
In some cases, the hydroformylation reaction has been extended to acetylenic compounds. The oxo process received considerable attention by the Germans during World War II.
Since that time, it has become industrially important in the United States for the making of aldehydes and alcohols.
~ypical of the U.S. patents in this area are 2,327,006-Roelen; 2,437,600-Gresham et al.; ~,497,303-Gresham et al.;

~ 0~75Z6 RD-6469

2,500,913-Schexnailder, Jr.; 2,564,104-Gresham et al.;
2,670,385-Rosenthal et al.; 3,022,340-Bloch; 3,~39,569-Slaugh et al.; 3,420,898-Van Winkle et al.; 3,496,203-Morris et al.;

3,644,529-Tucci et al.; 3,725,493-Deffner et al. Excellent reviews of the oxo and hydroformylation reactions are found, for example, in the Encyclopedia of Chemical Technology, Second Revised Edition, Vol. 14, 373, Interscience Publishers, New York (1967); Chapter 6 in Transition Metal Intermediates in Organic Chemistry, C.WO Bird, Logos Press, London (1967);
Catalysis Rev. 6 [1] 49, 85 (1972); and Chem. Rev. 62, 263 (1962). These reviews and the patent and journal references cited therein, a recent paper by V. M~cho, Effect of Water on the Oxo Synthesis, Chem. Zoesti 25, 49 (1971),and the above-identified patents give a very complete and detailed discus-sion of the hydroformylation reaction and how it is effected, for example, the various catalysts, effects of temperature and pressure and other variables for carrying out the hydro-formylation reaction, the effect of substituents on the olefinic compounds, the various types of olefinic compounds that can be used, etc., In the hydroformylation of alkenes, a mixture of isomers can form since the formyl group can be added to either of the two carbon atoms joined by the double bond.
Furthermore, some isomerization can occur, but -this is not so ~0475Z6 RD-6469 prevalent with the lower alkenes. When the lower alkenes are straight-chain and the olefinic group is terminal, the preponderant yield of aldehyde is formed by the formyl group adding on to ~he terminal carbon atom so that the product is the n-aldehyde. Despite the fact that hydroformylation of alkenes did produce a mixture of isomers, hydroformylation of esters such as acetates was thought to produce only one product. For example, hydroformylation of allyl acetate is reported to produce only 4-acetoxybutyraldehyde~ See, for example, the two articles by Adkins and Krsek, J. Am. Chem.
Soc. 70, 383 (1948) and 71, 3051 (1949).

SUMM~RY OF THE INVENTION
In studying the above reaction, I confirmed that hydroformylation of allyl acetate under the reported hydro-formylation conditions produced the reported 70-75% yield of

4-acetoxybutyraldehyde. However, I unexpectedly found that the balance of the allyl acetate has also been hydroformyl-ated to two products present in the reaction mixture, specifically, the two isomers of 4-acetoxybutyraldehyde, 2-acetoxybutyraldehyde and 3-acetoxy-2-methylpropionaldehyde.
Apparently, previous workers had failed to detect these two isomers.
~s I disclose and claim in my Canad~an application, Serial No. ~Oq, 3O~ filed S~ ~b~ y and assigned ~.047SZ6 to the same assignee as the present invention, it is possible to separate a significant amount of the 4-acetoxybutyraldehyde from its other two isomers in the mixture and the balance of the mixture can be dehydroformylated to produce a mixture of allyl acetate and l-propenyl acetate. Unexpectedly, I have found that l-propenyl acetate can be hydroformylated under hydroformylating conditions to give essentially the same mixture of hydroformylated products as is obtained from allyl acetate provided a cobalt hydroformylating catalyst is used.
In other words, the predominant products produced by hydro-formylation of l-propenyl acetate in this reaction are the completely unexpected products, 4-acetoxybutyraldehyde and/or 4-acetoxybutanol, as a mixture with their respective two iso-mers, 2-acetoxybutyraldehyde and 3-acetoxy-2-methylpropion-aldehyde and/or 2-acetoxybutanol and 3-acetoxy-2-methyl-propanol.
Based on my discovery, it is obvious that l-propenyl acetate, by itself or asa mixture with allyl acetate, can be used in place of allyl acetate as the feedstock for the hydroformylation reaction to produce 4-acetoxybutyraldehyde or 4-acetoxybutanol with very little effect on the actual product mix obtained. A further significance of this discovery is that coupled with my invention in the Canadian application discussed above, the mixture of allyl acetate and l-propenyl acetate, obtained from the thermolysis of the

-5-iO47SZ6 RD-6469 isomeric mixture from which at least some of the desired 4-acetoxybutyraldehyde has been removed, can be used as feed-stock for the hydroformylation reaction thereby providing a means for recycling the undesired products and increasing the overall yield of the desired 4-acetoxybutyraldehyde. As further disclosed in my Canadian application, the 4-acetoxybutyraldehyde can be hydrogenated to produce 4-acetoxybutanol which is readily de-esterified to produce 1,4-butanedioI useful in making polyester resins with dicarboxylic acids, especially tereph~halic acid.

DETAILS OF THE INVENTION
I have found that l-propenyl acetate can be hydro-formylated using any of the hydroformylating catalysts and the hydroformylating conditions disclosed in the art for the hydroformylation of olefins to the corresponding saturated aldehydes or alcohols. However, only the cobalt hydro-formylation catalysts produce the desired 4-acetoxybutyr-aldehyde or 4-acetoxybutanol. Even rhodium catalysts, which are the next most widely used and active catalysts used for hydroformylation reactions produces 2-acetoxybutyraldehyde with very small yields of its other two isomers.
It is now generally agreed that the hydroformylation reaction mechanism for converting olefins to aldehydes or alcohols involves participation of the metal catalysts no ~0475Z6 matter whether introduced as metal, sal~ or carbonyl, in the form of their hydrocarbonyl. Many transition metals have been found capable of forming the metal hydrocarbonyl. Nickel and iron have been used for the hydroformylation reaction but have been found to give extremely low yields of hydro-formylated products and are therefore considered much less satisfactory than the other metals such as rhod~um, iridium, ruthenium and cobalt. Of these latter metals, cobalt and ~hodium, as their hydrocarbonyls, have proven to be much more desirable as the hydroformylation catalyst and of these latter two, cobalt is so much less expensive that it is the preferred catalyst and i8 the one which the art has amassed the most data on the effect of the various reaction variables. How-ever, insofar as I can determine, my reaction, using cobalt hydroformylation catalysts, is unique in producing, as the predominant product, one in which the formyl group has not added to an unsaturated carbon atom of the starting olefinic ester.
The active form of the catalyst can be prepared by any of the methods used in the prior art. For example, the various metal salts or metal oxide can be reacted with carbon monoxide and hydrogen or the metal powder with carbon mon-oxide to foEm the carbonyl which can then be converted to the hydrocarbonyl by hydrogenation. Actually, these reactions can be carried out right in the reactor in which the hydro-~047526 formylation reaction is to be carried out since the metal, its salts or its carbonyl, either prior to or upon establishing the hydroformylating conditions to be used, will be converted into the hydrocarbonyl. In order to hasten this reacti~n, it is preferable to use cobalt compounds which are soluble in the reaction mixture, for example the carbonyl or salts of carboxylic acids. Therefore, I prefer in using cobalt as the hydroformylation catalyst to introduce it into the reaction mixture in the form of its carbonyl or a salt soluble in the reaction mixture. No matter what form the cobalt is added to the reaction, its active catalytic form is believed to be cobalt hydrocarbonyl, also called cobalt tetracarbonyl hydride.
The form of cobalt present in the hydroformylation reaction mixture is a function of the operating conditions.
Not only are different forms produced in the carbonyl formation reaction itself, but also equilibria exist between various carbonyls and between the carbonyls and the metal which are governed by reaction variables such as temperature, carbon monoxide pressure, hydrogen pressure, liquid-phase cobalt concentration, presence or absence of organic phosphines, etc. Because of this variable nature of the cobalt catalyst in the reaction mixture the various forms are best described as cobalt hydroformylation catalysts.
Actually7 the metal hydrocarbonyl can be used either in catalytic quantities or in stoichiometric quantities since the metal hydrocarbonyl is capable of hydroformylating l~propenyl acetate in the absence of hydrogen and carbon mon-oxide with the extent of reaction being dependent on the amount of metal hydrocarbonyl used in the same manner as discussed by Karapinka et al. for the hydroformylation of propylene in J. Org. Chem. 26, 4187 (1961). Since the use of stoichiometric amounts of the metal catalyst would be waste-ful and the reaction can be equally well performed using the metal hydrocarbonyl in catalytic quantities in conjunction with a carbon monoxide and hydrogen atmosphere, I prefer to use the latter for hydroformylating l-propenyl acetate.
Like the prior art hydroformylation reactions, I
can carry out my hydroformylation of l-propenyl acetate under any of the wide -variety of hydroformylation reaction condi-tions heretofore found suitable in the prior art. Although a wide variety of temperatures and pressures can be used, I
have found that in general a temperature of at least 110 C.
is required to obtain a reasonably fast reaction rate, and preferably the temperature is at least 120 C. where the reaction becomes exothermic and very rapid. In this tempera-ture region, aldehydes are the predominant product. To minimize thermolysis of the 3-acetoxy-2-methylpropionaldehyde to methacrolein, the maximum temperature should not exceed 175-185 C. In this temperature region~ alcohols are the ~0475Z6 RD-6469 predominant product. Likewise, pressures of at least 500 psi and preferably l,000 psi are used to maintain the catalyst in the form of the hydrocarbonyl. For o~timum results, I have found that the preferred temperature is in the range of 125-150 C. for the production of the acetoxyaldehyde productsand 165-190 C. for the production of their corresponding acetoxybutanol products. The preferred pressure is in the range of l,000-5,000 psi although it can be lowered to 500 psi when phosphine modified catalysts are used.
As discussed in the above-referenced article in the Encyclopedia of Chemical Technology, the temperature and pressures used should take into consideration the carbon mon-oxide partial pressures at equilibrium concentrations of dicobalt octacarbonyl at various temperatures. In general, a temperature and pressure should not be chosen whose inter-cept would fall below the uppermost curve of the graph shown on page 376. A good discussion of the catalyst removal and recovery for reuse is given in the same reference on page 383 et seq. and its references. In addition to using the metal hydrocarbonyl catalyst alone, I can, like the prior art, use such catalyst in conjunction with other additives which serve as reaction modifiers if desired,for example,use of other ligands in addition to carbon monoxide such as the phosphines, for example tributylphosphine, or other metals in conjunction with th~ metal catalyst used. For example, palladium has ~ o475Z6 RD-6469 been found to be a promoter or cocatalyst with cobalt and the use of nickel or organic phosphites with cobalt have been found advantageous when the aldehydic products are to be hydrogenated to the corresponding alcohols in the same reac-tor.
When organic phosphines are used to modify the co-balt hydroformylation catalyst, any of the prior art phosphines can be used. The preferred ones are the tri-al~ylphosphines. They stabilize the cobalt carbonyl catalysts so that lower pressures can be used. However, for my reaction, phosphine modified catalysts appear to be less reactive than the same unmodified ca~alyst. The net effect i9 that the hydroformylation reaction with phosphine modified cobalt catalysts requires higher temperatures to initiate than with the same cobalt catalyst without the phosphine.
Generally, these temperatures are as high as the temperature which would give alcohol products with the unmodified cobalt catalysts. Because of this and because the phosphine modi-fied catalysts cause more side reactions, especially with the 3-acetoxy-2~methylpropionaldehyde intermediate, I prefer to use unmodified cobalt hydroformylation catalysts even when carrying out the hydroformylation reaction under alcohol form-ing conditions. However, the highest overall yield of alcohol products is obtained by using the hydroformylation reaction to produce the aldehyde products and using a separate ~047S26 RD-6469 hydrogenation reaction in the absence of carbon monoxide and preferably with a standard hydrogenation catalyst, e.g., a nickel,platinum or palladium catalyst. Not only are the yields better but recycle of the undesired aldehyde products can be effected as previously described. For these reasons, I prefer to hydroformylate the l-propenyl acetate to the acetoxybutyraldehyde mixture.
In the hydroformylation of a single olefinic double bond~ the theoretical ratio is CO:H2=1:1. Generally, a ratio is in the range of 1:1 to 1:2 and is the preferred mole ratio to use even though an excess of these reagents are present to produce the desired pressure. A ratio of 1:2 increases the rate of reaction whereas a ratio of 1:1 increases the proportion of 4-acetoxybutyraldehyde in the product mixture.
As is obvious, by first pressurizing the reactor with one gas to 1/2 the desired pressure and then using the other gas to pressurize to the desired pressure will produce an equal molar ratio of the two gases. Other means, of course, could be used, for example,premixing the gas,using flow meters or other measuring devices.
If the mixture of aldehydic products are to be further hydrogenated in the reaction to produce the co~rresponding acetoxybu~anols, an additional mole of hydrogen is required. Where this reduction is to be carried out either simultaneously with or as a consecutive step following the :10475Z6 RD-64~9 hydroformylation reaction, then the preferred ratio of reactants is at least two moles of hydrogen per mole of carbon monoxide. However, due to the excess gas present to provide the desired pressure, hydrogenation can be accomplished with only one mole of hydrogen per mole of carbon monoxide~ The effect of variation of such process variables as the pressure of the gases, temperatures used, ratio of gaseous reactants, ratio of gaseous reactants to olefin, ratio of catalyst to olefin, etc., are well documented in the prior art. Their effect on the hydroformylation of l-propenyl acetate to 4-acetoxybutyraldehyde or 4-acetoxybutanol i~ not ~art of my invention nor critical thereto, except insofar as these variables and their known effects can be used by those skilled in the art in practicing my invention.
Preferably the reaction is carried out in the liquid phase and in the presence of a solvent in which the catalyst and l~propenyl acetate are soluble. The solvents used should be essentially inert to the reactants and stable under the reaction conditions L Liquid hydrocarbons, especially aromatic hydrocarbons, for example, benzene, toluene, xylene, etc~, are good solvents to use, Concentra-tions of the l-propenyl acetate in the solvent are not critical. I have found that concentrations in the range of 5-50% by weight are convenient to use and aid in controlling the exothermic hydroformylation reaction. Although hydro-~0475Z6 RD-6469 formylation can be carried out in the vapor phase with heterogeneous catalysts, the efficiency of the reaction is so far inferior to that with liquid phase with homogeneous catalyst that there is no incentive to use it.
It is-obvious from what has been said above that my process can be best described as a process for making 4-acetoxybutyraldehyde or 4-acetoxybutanol by hydroformylating l-propenyl acetate under hydroformylating conditions using a cobalt hydroformylating catalyst. Wh~le it is true that the 4-acetoxybutyraldehyde and/or 4-acetoxybutanol is obtained as a mixture with their two isomers, 2-acetoxybutyraldehyde and 3-acetoxy-2-methylpropionaldehyde and/or 2-acetoxybutanol and 3-acetoxy-2-methylpropanol, the 4-acetoxybutyraldehyde and/or 4-acetoxybutanol is the predominant product as well as the most desirable product from the viewpoint of making 1,4-butanediol. However, it is obvious that if any of the other isomers were desired, my process could be used and indeed does produce these isomers. However, in some cases, some of these isomers can be produced in better yields by other means. For example, 2-acetoxybutyraldehyde can be obtained in very high yields from l-propenyl acetate by using rhodium hydroformylation catalyst in place of the cobalt hydroformyl-ation catalysts under the same conditions discussed above for the cobalt catalysts.
As has been mentioned previously, the mixtures of ~ o 47 S Z ~ RD-6469 the aldehydes will also contain some of the corresponding alcohols and the mixturesof alcohols will generally contain some aldehydes. Both reactions can be carried out so that the mixture of aldehydes is essentially free of the alcohols S and vice versa, however, the economics of doing so generally are not favorable.
In general, after the reaction vessel has been charged and pressurized with reactants and catalyst, the temperature is raised until the desired rate of reaction is obtained as noted by the time-rate of decrease in pressure.
Additional hydrogen and carbon monoxide can be charged to restore the pressure, The end of the reaction is readily determined when the pressure remains constant at a constant reaction temperature. When the catalyst for this reaction has been chosen to produce aldehydes, e.g. absence of phos-phine modifiers, using the lowest temperature that will cause a decrease in pressure to occur will maximize aldehyde pro-duction and minimize alcohol production as well as undesired by-product formation, Once the drop in pressure has indi-cated the end of the aldehyde forming reaction, the tempera~ture can be increa~ed to a temperature where a rapid pressure drop will again be noted. This i8 the minimum temperature ; required for this particular catalyst and reaction conditions to cause rapid hydrogenation of the formyl groups to methylol groups. Once determined, the reaction can be carried out in one step at this temperature rather than using t,he two temperature method to produce the alcohols as the predominant products. Use of a single temperature which lies between these two temperatures will cause the ratio of alcohol products to aldehyde products to increase as the temperature increases from the lower to the higher temperature.
When the higher temperatures are used to promote the production of alcohol products,a secondary reaction occurs in that some of the acetoxybutanols, which are mono-acetate esters of butanediols,, disproportionate to somedegree whereby some of the product is deacetylated to the corresponding diol and a corresponding amount of product is acetylated to produce the diacetate ester of the diol. This complicates somewhat the separation of the products. However, this resolves itself if hydrolysis of the esters to the diols or esterification to the diesters is carried out prior to separation.
In order that those skilled in the art may better understand my invention, the following examples are given by way of illustration and not by way of limitation. Tempera-tures are given in degrees Centigrade and pressures are gauge pressures.

l-Propënyl acetate was prepared by the reaction of 10 475~ RD-6469 propionaldehyde and acetic anhydride in the presence of potassium acetate catalyst. The product obtained, bp. 100-L06, was composed of the cis- and trans-isomers in about equal amounts. A mixture of 20.0 grams of this l-propenyl acetate, 1.0 grams of dicobalt octacarbonyl and 60 ml. of benzene was sub~ected to 3,050 psi of 2 1 hydrogen / carbon monoxide in a 300 ml. stirred autoclave. On heating to 12~, the pressure reached 3,600 psi; a rapid reaction began at that point and was complete in about 10 minutes, with a maxi-mum temperature of 155 and a gas uptake of about 700 psi.
The product solution was examined directly by VPCand ~ound to contain 4-acetoxybutyraldehyde, 2-acetoxybutyr-aldehyde and 3-acetoxy-2-methylpropionaldehyde in a 3~1:1~1:1 ratio, in addition to small quantities of the butanediols and their acetates. Also present were very small amounts of methacrolein and acetic acid (the products of dehydroacetoxy-lation of the 3-acetoxy-2-methylpropionaldehyde).
The benzene solution was treated with 5 ml~ of acetic acid and heated at reflux for one hour. The cobaltous acetate that formed was filtered off and the filtrate was distilled. Isolated was 16.4 grams (63% yield) of the mixed aldehydes, bp. 62-89 / 1 mm. The isomer ratio (4-acetoxy-butyraldehyde : 2-acetoxybutyraldehyde : 3-acetoxy-2-methyl-propionaldehyde) at ~his point was 6.3:2.3:1 Methacrolein 2~ and acetic acid were liberated during the distillation thus accounting for the decrease of 3-acetoxy-2-methylpropion-aldehyde.

~ 0~ 75Z6 RD-6469 ~MPLE 2 A mixture of 15.0 grams of l-propenyl acetate, 0.75 grams of dicobalt octacarbonyl and 45 ml. of benzene was sub-~ected to l,OQ0 psi of 1:1 hydrogen / carbon monoxide and heated to 125. At that point an exothermic reaction and rapid gas uptake began. The temperature reached 150 before subsiding; gas was replenished to maintain the pressure at 1,000-1,200 psi. Reaction was complete within 15 minutes.
A direct quantitative VPC analysis o~ the products (diphenylmethane added as internal standard) showed the presence of lL.2 grams of 4-acetoxybutyraldehyde (57% yield), 2.2 grams of 3-acetoxy-2-methylpropionaldehyde (11% yield), and 1.3 grams of 2-acetoxybutyraldehyde (7% yield). A
significant amount of butanediol and butanediol acetates (about 10% total yield, almost exclusively the 1,4-compounds) was also produced. Thus the yield of 1,4-butanediol precursors was about 67%.

A series of hydroformylation reactions were carried out with l-propenyl acetate, allyl acetate, and mixtures of the two (25% in benzene, with 1.0 grams of dicobalt octa-carbonyl per 20 grams of olefin, at about 3,000 psi of 2:1 hydrogen / carbon monoxide and 125-150). The gas uptake in each case was complete within 10 minutes of onset of reaction.

-1~--104';t526 The products were subjected to direct quantitative glplc analysis as in the above case. The results are summarized in Table I. Minor amounts of butanediol and butanediol acetates were also produced.

_ TABLE I
StartlnR Olofin Product Ylelds Allyl l-Propenyl 4-Acetoxy- 3-Ac~toxy-2~methyl- 2-Acotoxy-EXamP10 Ac~ta~e Ac-tato butyraldohyde ProD~onaltehyde butyraldehyde 3 100 X O Z 63 ~ 16 ~ 12 X

The ~ollowing example illustrates that rhodium catalysts are not the equivalent of the cobalt catalysts.
The catalyst used is one of the most active rhodium hydro-formylation catalysts.

A mixture of 20.0 grams of l-propenyl acetate, 0.25 grams of rhodium bis(triphenylphosphine)dicarbonyl chloride [RhCl(C0)2(PPh3)2], 0.035 grams of triethylamine, and 60 ml, of benzene was subjected to 1,000 psi of 1:1 hydrogen / carbon monoxide and heated at 75-100 for one hour. Gas was replenished to maintain the pressure at 1,000-1,200 psi.
Direct quantitative VPC analysis of the products showed the presence of 13.1 grams of 2-acetoxybutyr--19o ~047SZ6 aldehyde (50% of theoretical), 0.2 grams of 3-acetoxy-2-methylpropionaldehyde (1~), 0.2 grams of 4-acetoxybutyr-aldehyde (1%), and 1.4 grams of residual l-propenyl acetate (7% unconverted). Also detected were 1.1 gram of acetic acid (9%) and an approximately corresponding amount of methacrolein.

A mixture of 100.8 grams of 4-acetoxybutyraldehyde, 16.8 grams of 3-acetoxy-2-methylpropionaldehyde, 19.6 grams of 2-acetoxybutyraldehyde, and 2.8 grams of acetic acid was subjected to rapid distillation through a heated 300 mm.
Vigreaux column. About half of the material was taken over at 80-104 / 10 mm. As determined by quantitative VPC
analysis, the distillate was composed of 26.3 grams of 4-acetoxybutyraldehyde, 12.8 grams of 3-acetoxy-2-methyl-propionaldehyde, 19.1 grams of 2-acetoxybutyraldehyde, and 3.4 grams of acetic acid, and the residue contained 70.7 grams of 4-acetoxybutyraldehyde and 1.4 grams of 3-acetoxy-2-methylpropionaldehyde. Some methacrolein was collected in a cold trap.
The residue was 98% pure 4-acetoxybutyraldehyde.
It could be still further purified by redistillation.

A mixture of 20.0 grams of l-propenyl acetate, 1.0 104752~ RD-6469 gram of dicobaltoctacarbonyl and 60 ml. of benzene was sub-jected to 3,000 psi of 2:1 hydrogen / carbon monoxide and heated as in Example 7. External heating was maintained so that the temperature rose to 180 during the course of the reaction. After one hour at 180-185, the mixture was cooled and vented. VPC examination of the products showed the presence of a complex mixture containing 4-acetoxybutanol, 3-acetoxy-2-methylpropanol, 2-acetoxybutanol, and their respective diol and diacetate disproportionation products.
The benzene solution was combined with 40 ml. of acetic anhydride and heated at 80 for one hour Cobaltous acetate precipitated and was filtered off, Quantitative VPC
analysis (diphenylmethane internal standard) of the solution showed the presence of 17.4 grams of 1,4-butanediol diacetate (50% yield), 5.6 grams of 2-methyl-1,3-propanediol diacetate (16%), and 7.7 grams of 1,2-butanediol diacetate (22%).

A mixture of 20.0 grams of l-propenyl acetate, 1.0 grams of dicobalt oc~acarbonyl, 1.30 grams of tri-n-butyl-phosphine and ~0 ml. of benzene was subjected to 1,200 psi of 2:1 hydrogen / carbon monoxide and heated. Upon reaching 180 gas uptake began. The mixture was maintained at 180-190 and 1,000-1,200 psi for 30 minutes, then cooled and vented, -2~L-~ 7 5 ~ 6 RD-6469 The reaction product was a complex mixture which when examined directly by VPC was found to contain as the major constituents 4-acetoxybutanol, 2-acetoxybutanol, and their respective diol and diacetate disproportionation products Also present was about 3 grams of acetic acid.
The benzene solution was combined with 40 ml. of acetic anhydride and heated at 80 for one hour. The acetylated mixture was then sub~ected to quantitative VPC
analysis as in the above example and was found to contain 13.6 grams of 1,4-butanediol diacetate (39% yield), 4.2 grams of 1,2-butanediol diacetate (12%), 0.2 grams of 2-methyl-1,3-propanediol diacetate (0.6%), and 3.9 grams of n-butyl acetate (17%).
The above examples have clearly demonstrated the best mode known to me of carrying my invention into effect.
As will be readily understood by those skilled in the art, variations can be made in practicing my invention as clearly taught in the balance of the specification and by the prior art on hydroformylation without departing from the true intended scope of my invention.
My invention can be used as an independent process for the conversion of l-propenyl acetate into the mixture of acetoxyaldehydes which are useful in and of themselves as chemical compounds or as intermediates in the preparation of the corresponding acetoxybutanols which can be de esterified 10475~6 RD-6469 to the corresponding butanediols. My process also has further utility when combined with the dehydroformylation process disclosed and claimed in my copending application referenced above, wherein I disclose that at least some of the 4-acetoxybutyraldehyde product may be separated from its other two isomers and the latter products dehydroformylated to produce a mixture of l-propenyl acetate and allyl acetate which then, by the process disclosed herein, can be hydro-formylated to again produce the mixture which is predominant-ly 4-acetoxybutyraldehyde. By repetition of this related process on the mixture of aldehyde products, the net effect, except for minor process losses, is the hydroformylation of essentially all of the l-propenyl acetate to 4-acetoxybutyr-aldehyde, This hydroformylation step also is an important step in the process of making l,4-butanediol by oxidatively coupling propylene and acetic acid to produce allyl acetate as an intermediate product which is hydroormylated to produce the above mixture of aldehydic products which can then be treated as described above to produce increased yields of 4-acetoxybutyraldehyde which is hydrogenated and de-esterified to 1,4-butanediol. Thus, this invention provides an improvement when combined with my related inven-tion resulting in a net overall increase in the production of 1,4-butanediol from propylene and acetic acid and oxygen.

1~ ~75Z6 RD-6469 Such an over-all improved process is disclosed and claimed in my Canadian application cross-referenced above.
These and other modifications of this invention and its uses as will be readily discerned by those skilled in the art, based on the teachings of the prior art herein incorporated by reference and the specific teachings of this application can be employed within the scope of the invention.
The invention is intended to include all such modifications and variations as are embraced within the following claims.

Claims (5)

The embodiments of the invention in which an exclu-sive property or privilege is claimed are defined as follows:

1. The process for making 4-acetoxybutyraldehyde or 4-acetoxybutanol which comprises hydroformylating 1-propenyl acetate under hydroformylating conditions in the presence of a cobalt hydroformylation catalyst.

2. The process of claim 1, wherein the 1-propenyl acetate is present in a mixture with allyl acetate.

3. The process of claim 1, wherein the cobalt hydroformylation catalyst is introduced into the reaction mixture as dicobalt octacarbonyl or soluble cobalt salt.

4. The process of claim 2, wherein the cobalt hydroformylation catalyst is introduced into the reaction mixture as dicobalt octacarbonyl or soluble cobalt salt.

5. The process of claim 1, for making 4- acetoxy -butyraldehyde wherein the active catalyst for the hydro -formylation reaction is cobalt hydrocarbonyl.