WO1996036582A1

WO1996036582A1 - Alkoxylation of alcohols

Info

Publication number: WO1996036582A1
Application number: PCT/US1996/007884
Authority: WO
Inventors: Marguerite A. Schrader; Debra A. Timmers; Alan E. Jackson
Original assignee: The Dow Chemical Company
Priority date: 1995-05-19
Filing date: 1996-05-16
Publication date: 1996-11-21

Abstract

An improved method for alkoxylating alcohols is provided. The alcohol is contacted with an alkene oxide in the presence of an aprotic polar solvent and a catalyst under conditions sufficient to form monoalkyl ethers of a glycol or a polyalkylene glycol.

Description

ALKOXYLATION OF ALCOHOLS

The present invention relates to the alkoxylation of alcohols, and more particularly to the use of polar aprotic solvents in the alkoxylation reactions. The production of glycol ethers and polyglycols typically involves the alkoxylation of alcohols in which an alkene oxide is contacted with an alcohol in the presence of a base as a catalyst. Forthe production of glycol ethers, the reaction is usually carried out with an excess of the alcohol in orderto provide solvency for the catalyst. The alkene oxide to alcohol ratio dictates whether glycol ethers or polyglycols are primarily formed. Known practice in this field is to increase the rate of reaction by either increasing the level of catalyst or increasing the temperature at which the reaction is run. Each of these approaches has drawbacks. Increasing the reaction temperature can result in increased levels of impurities caused by side reactions, and can also cause processing problems if reaction materials are not stable at the higher temperatures. Increasing levels of catalyst typically causes processing problems due to the removal and processing of large amounts of used catalyst. Problems such as these make the current approaches for increasing the rate of alkoxylation of alcohols unacceptable for a number of applications.

The present invention provides a new process for alkoxylating alcohols. The process comprises contacting an alcohol with an alkene oxide in the presence of an aprotic polar solvent and a catalyst, preferably a base, under conditions sufficient to form monoalkyl ether(s) of a glycol or a polyalkylene glycol.

One of the advantages of the present invention is to provide a process of increasing the rate of alkoxylation of alcohols which does not necessarily require the addition of more solid catalyst or the increase in temperatures. It has been discovered that the rate of base catalyzed alkoxylation reactions between alcohols and alkene oxides can be increased by adding an aprotic polar solvent. For the purposes of this invention "alkoxylation" reactions include any reaction of an epoxide group with an alkoxide ion to form monoalkyl ethers of glycol or a polyalkylene glycol of the general formula:

wherein X is the alcohol residue (the alcohol less the reactive OH group(s)); R1 , R2, R3, and R4 are independently a hydrogen atom, an alkyl group, an aryl group, an alkenyl group, a cycloalkyl group or an aralkyl group; m is a positive integer equal to the number of reactive OH groups in the alcohol and n is a positive integer. Two R groups may also be joined to form cyclic structures. The increase in reaction rate has been shown to varying degrees for many different reactions involving different alcohols, alkene oxides, catalysts and aprotic polar solvents.

Many different alcohols are suitable for use in the present invention. Forthe purposes of this invention the term "alcohol" includes all non-carbohydrate compounds having

5 at least one hydroxy group which can be reacted with an epoxide group to form monoalkyl ethers of glycol or a polyalkylene glycol. Monohydroxylated alcohols, such as methanol, ethanol, n-propanol, i-propanol, n-butanol, allyl alcohol and phenol have been shown to be particularly well suited for these reactions, and are preferred. N-butanol is the most preferred alcohol for use in the present invention. For purposes of this invention it is also preferred that

10 the alcohol be liquid at the temperature at which the reaction will take place to facilitate the dissolution of the alcohol.

Alkene oxides, also called alkylene oxides or epoxides, have the general formula:

_₀ wherein Ri, R₂, R₃, and R₄ are independently a hydrogen atom, an alkyl group, an aryl group, an alkenyl group or a cycloalkyl group. Two R groups may also be joined to form cyclic structures. Although any alkene oxide can potentially be used with this invention, alkene oxides having from 2-4 carbon atoms such as ethylene oxide, propylene oxide and butyl ene oxide are preferred. The alkene oxide is preferably employed in an amount to form a ratio of

_.. from 0.1 mole (more preferably 0.25 mole) of oxide per mole of alcohol to 40 moles (more preferably 20 moles) of oxide per mole of alcohol.

Catalysts used in this invention are preferably bases. For purposes of this invention the term "base" includes any Lewis base. Thus, any substance that is capable of donating a pair of electrons can be used as the catalyst in this reaction. This includes the basic catalysts currently used in the alkoxylation of alcohols, see for example, Jan Chlebicki, "The

30 Synthesis and Properties of Polyoxypropylene Glycol Alkyl Monoethers", Scientific Papers of the Institute of Organic Technology and Plastics, Vol. 21. No. 2, Wroctaw Technical University, 1975. Preferred catalysts include alkali metal hydroxides. Most preferred catalysts include potassium hydroxide, sodium hydroxide, lithium hydroxide and cesium hydroxide, and of all these, potassium hydroxide is most highly preferred. Heterogeneous catalysts may also be successfully

35 used with the present invention. The catalyst is added in an amount of at least 10 ppm based on the weight of alcohol, and preferably no more than 10,000 ppm, more preferably no more than 5,000 ppm. Polar, aprotic solvents are those solvents having a high dielectric constant which do not readily donate protons. For purposes of this invention, high dielectric constant means greater than 15, more preferably greater than 30. The preferred solvents are dimethylsulfoxide (DMSO), n-methylpyrrolidinone (NMP), acetonitrile, dimethylformamide (DMF), dimethylacetamide (DMAC), tetramethylene sulfone, or mixtures thereof. DMSO, and mixtures thereof, are the most preferred polar aprotic solvents. Other possible solvents which may be appropriate under certain reaction conditions include gamma-butyrolactone, dimethylcyanamide, diethylcyanamide, dimethylimidazolidinone (DMI or DMEU), dimethyltetrahydropyrimidinone (DMPU), 2-ethyl-2-oxazoline, hexamethylphosporamide (HMPA), hexamethylphosphorous triamide (HMPT), n-methy-e-caprolactam, 3- methyloxazolidinone, propionitrile, propylene carbonate, tetraethylsulfamide (TES), and tetramethylene sulfoxide. Although the term "solvent" is used herein to describe a class of materials, it should be understood that the use of the term is not intended to indicate that these compounds are necessarily functioning to dissolve any of the reaction materials. While the addition of polar aprotic solvent in a ratio as low as 0.1 : 1 (weight of solvent to weight of alcohol) has been shown to be effective in increasing the rate of reaction, it is preferred that the solvent be added in a ratio of from 0.25: 1 to 1 : 1 , where the weight of alcohol is the total weight of all alcohols present in the feed or initial charge. Thus, the weight of alcohol includes the weight of compounds such as glycol ethers if they are present in the ^ed or initial charge, as these are non-carbohydrate compounds which contain reactive OH groups and so are within the scope of the definition of "alcohol" as used herein.

With some aprotic solvents such as DMSO, it has been found that the addition of small amounts of ketones, aldehydes or other compounds with a carbonyl functionality which are substantially non-reactive, unexpectedly enhance the rate of reaction. Compounds which have been shown to synergistically increase the rate in this manner include acetone, methyl ethyl ketone, methyl isobutyl ketone, 2,6-dimethyl-4-heptanone, cyclohexanone, acetophenone, propionaldehyde and urea. Nitriles, such as acetonitrile, have also shown some synergistic effect when added to DMSO. Early kinetics work with acetone has indicated that increased rates of reaction are observed if some acetone is present (preferably at least 20ppm with respect to the aprotic polar solvent), but that the actual amount of acetone does not seem to matter. Similar rates of reaction were observed with levels ranging from 100 to 10,000 ppm acetone with respect to the weight of the aprotic polar solvent. When the rate enhancing compound is added to the aprotic solvent in large amounts, however, then there is a greater risk of some of the rate-enhancing compound undergoing condensation reactions to form by- products. Accordingly it is preferred that the rate-enhancing compound be added in a range of from 20 ppm to 10,000 ppm with respect to the aprotic polar solvent. It is most preferred that the rate-enhancing compound be added in an amount of about 100 ppm. The reactor used to carry out the reaction of the present invention is not critical. Reactors known in the art such as batch, continuous plug flow or continuous stirred tank type reactors can be used. In a batch reactor the aprotic solvent would be added before the alkene oxide addition. In the continuous reactors the solvent would be added continuously along with the alkene oxide and other reactants. It should be noted that the increased reaction rates provided by this invention make the continuous reactor types such as the continuous stirred tank reactor more attractive and most preferred.

The reaction can be run at similar temperatures to current alkoxylation reactions known in the art (see, for example V. I. Emel'yanov etal., "Optimization of the Process of Combined Synthesis of the Butyl Ethers of Mono- and Diethylene Glycols", Khimicheskava Promyshlennost, Vol. 21, No. 12, pp 898, 1989, or U. S. Pat. 3,972,948 to G. J. Laemmle et al.) The reaction temperature should be such that the solvent is stable at the chosen temperature. The preferred temperature range to carry out the reaction is from 80°Cto 200°C, more preferred from 125°Cto 175°C, and most preferred around 150°C. The invention will become more clearly understood by considering the following examples, in which all percentages are by weight unless otherwise noted. As unreacted oxide raises the pressure of the reactor due to its vapor pressure at the reaction temperature, safety factors dictated the feed rates of the alkene oxide in the runs not embodying the invention as each reaction vessel has an inherent pressure limitation. The feed rates established during these comparative examples were then roughly duplicated and levels of unreacted oxide were monitored in order to get an indication of the relative reaction rates. It should be noted that the faster the oxide reacts, the faster the oxide can generally be added, and so the reaction times of the examples embodying the invention have been artificially extended for the purposes of comparison. Accordingly, the reported reductions in reaction rates are approximations based primarily on peak oxide levels. Chemometric modeling was used to correct for the variations in times for oxide addition. It should be noted that the present invention increases the reaction rate by as much as 25 percent or more. Example A (Comparative)

A batch reactor was used to carry out the reactions in this and subsequent Examples. The reactor was a series 45422-liter stainless steel reactor manufartured bythe Parr Instrument Company, equipped with an agitator, a cooling coil utilizing cooling water, a line for nitrogen padding, pressure relief venting, a sample line through which a sample could be cooled to ice temperatures, and an oxide addition line. The reactor was heated using an external electric heating mantle. The oxide addition line included an oxide addition cylinder. This oxide addition cylinder was charged with a desired amount of alkene oxide then was pressurized with nitrogen. The addition rate was then controlled by adjusting a needle valve located between the cylinder and the reactor. The reactor also included a strip chart recorder so that reactor temperature, pressure and the weight of the alkene oxide added could be monitored as a function of time.

Four hundred grams of n-butanol and 1.1 grams of potassium hydroxide were added to the reactor, stirred and heated to approximately 150°C. One hundred and fifty grams of propylene oxide (PO) was then continuously added over a period of approximately 0.63 hour. Samples were taken every 15 to 30 minutes during the reaction. The reaction was deemed to be complete when the propylene oxide content in the sample was less than 0.05 percent as determined by a gas chromatographic (GC) analysis. For this example, the reaction was deemed complete in 2 hours. Peak oxide levels observed during the reaction were 5.1 percent.

Example 1

The same reaction as in Example A was carried out with the exception that 400 grams of DMSO and 1.2 grams of acetone were added to the initial batch reaction charge. Propylene oxide was added such that the pressure in the reactor vessel remained at safe levels. The feed was complete in about 0.75 hour, and the reaction was complete in 1.0 total hours, with a peak oxide level of 0.75 percent, after adjusting forthe presence of the solvent. This represents approximately a 6 fold increase in the reaction rate over the method generally described in Example A. Afterthe reaction was complete the solvent was separated by d i sti 11 ati on and retu rned to the reactor to be recycl ed .

Example 2

Four hundred grams of n-butanol and 1.1 grams of potassium hydroxide were mixed and placed in the reactor along with 200 grams of DMSO and 1.0 gram of acetophenone. This mixture was heated to approximately 150°C. Next, 150 grams of propylene oxide was added over half an hour. The reaction was complete in 0.83 hour. The peak amount of propylene oxide observed during the reaction was 1.57 percent, adjusted forthe presence of the solvent. This represents a 4.8 fold increase in reaction rate over the method generally described in Example A.

Example 3

Four hundred grams of n-butanol and 1.1 grams of potassium hydroxide were mixed and placed in the reactor along with 200 grams of DMSO and 1.0 gram of propionaldehyde. This mixture was heated to approximately 150°C. Next, 150 grams of propylene oxide was added over 0.45 hour. The reaction was complete in 0.83 hour. The peak amount of propylene oxide (PO) observed during the reaction was 2.2 percent, adjusted forthe presence of the solvent. This represents a 3.4 fold increase in reaction rate over the method generally described in Example A. Example 4

A charge of 400 grams of n-butanol, 1.1 grams of potassium hydroxide, and 400 grams of DMAC were added to the reactor described in Example 1, stirred and heated to 150^CC. Then, 150 grams of propylene oxide were added. Addition was complete in 0.5 hour, and the reaction was complete in 1 hour. The peak PO level during the reaction was 2.20 percent after adjustment for solvent level for comparison purposes. This represents approximately a 3.5 fold increase over the method generally described in Example A.

Example 5

The same reaction as in Example A was carried out with the exception that 400 grams of NMP was added to the initial reaction charge. Propylene oxide was added such that the pressure in the reactor vessel remained at safe levels. The feed was complete in approximately 0.75 hour, and the reaction was complete in 1.5 hours. The peak propylene oxide level observed during the reaction was 2.2 percent, adjusted for the presence of the solvent. This represents approximately a 2.5 fold increase in the reaction rate over the method generally described in Example A.

Example 6 The same reaction as described in Example 1 was performed, with the exception that no acetone was added to the reaction charge. Propylene oxide was added over a 0.58 hour period. At 2.5 hours, the adjusted propylene oxide concentration was 0.27 percent. The peak propylene oxide concentration (adjusted for comparison purposes forthe amount of solvent present) during this reaction was 18.7 percent. This indicated a rate that was approximately 0.04 the reaction rate of the method of Example 1.

Example 7

Six hundred grams of n-butanol and 1.65 grams of potassium hydroxide with 296 grams of DMSO and 1.75 grams of acetone were added to a reactor substantially as described in Example 1 and heated to 150°C. Then, 280 grams of butylene oxide were added over a 60 minute period. This reaction was complete in a total of 3.75 hours and peak butylene oxide levels during the reaction reached 2.51 percent. These values indicated approximately 3.8 fold increase over a similar reaction run without added polar aprotic solvent.

Example 8

Sodium hydroxide, in an amount of 0.616 gram, was dissolved in 400 grams of n- butanol, then added along with 200 grams of DMSO and 1.0 gram of acetone to the reactor. The mixture was then heated to 150°C. 150 grams of propylene oxide was then added over 35 minutes. The reaction was complete in 45 minutes. Maximum PO concentration reached during the reaction was 1.9 percent after adjusting forthe presence of the solvent, indicating a three fold increase over a similar reaction run without polar aprotic solvent and acetone.

Example 9

One hundred and fifty grams of propylene oxide were reacted with 400 grams of n-butanol in the presence of 8.30 grams of 10 percent potassium oxide supported on carbon as a catalyst. Also in the mixture was 398.8 grams DMSO and 1.2 grams of acetone. The propylene oxide was added to the reactor over a period of 2 hours. This reaction was estimated to be o about 2.1 times faster than the same reaction run without added polar aprotic solvent and acetone.

Example 10

One gram of potassium hydroxide was dissolved in 200 grams of methanol. This 5 mixture was added along with 100 grams of DMSO and 0.5 gram of acetone to the reactor and heated to 150°C. An amount of 150 grams of propylene oxide was then added over 20 minutes. The reaction was complete in 30 minutes. A maximum oxide concentration of 1.05 percent (after adjusting forthe presence of the solvent) was observed. This indicated approximately a three fold increase in reaction rate over a similar reaction run without DMSO and acetone. 0

Example 11

Two hundred forty grams of n-butanol, 160 grams of propylene glycol n-butyl ether and 0.9 gram of potassium hydroxide were mixed and added to the reactor. Two hundred grams of DMSO and 1.0 gram of acetone were then added to the mixture. The total 5 reaction charge was then heated to approximately 150°C. Three hundred twenty-three grams of propylene oxide were added over a 0.9 hour time period. After 3 hours, the reaction mixture contained 0.35 percent, after adjustment for he solvent level. The peak oxide level observed during the reaction was 12.95 percent. This is a 2.4 fold increase over a similar reaction run without any polar aprotic solvent and acetone. 0

Example 12

An amount of 1.1 grams of KOH was dissolved in 400 grams of n-butanol, to which was then added 200 grams of NMP and 200 grams of acetonitrile. This mixture was heated to 150°C. Then, 150 grams of propylene oxide was added over a 42 minute period. 5 Based on comparison with a model of a neat reaction, this solvated reaction was about 1.9 times faster than a neat reaction. By comparison, reactions with 400 grams acetonitrile as a solvent resulted in a 1.6 times rate increase, while reactions with 400 grams NMP as a solvent resulted in a 2.6 times rate increase. This indicates the performance of a 50/50 mixture of these solvents was part way between the performance of either acetonitrile or NMP by itself.

Example 13 An amount of 2.24 grams CsOH was dissolved in 400 gram n-butanol. Two hundred grams of DMSO and 1.0 grams of acetone were added to the mixture before heating to 150°C. One hundred fifty grams of PO were then added over 32 minutes. The peak oxide level was 2.05 percent, and the reaction was complete in 0.75 hours. This indicated a rate increase of 2.8 times over the reaction run without any polar aprotic solvent and acetone.

Example 14

A reaction was run, in which 0.163 gram of LiOH was dissolved in 400 grams n- butanol. Two hundred grams of DMSO and 1.0 gram of acetone were added to the mixture before heating to 150°C. Then, 150 grams of propylene oxide were added over a 30 minute period. The peak oxide level was 7.56 percent, and the reaction was complete in 1.75 hour. This indicated a rate increase of 2.25 times over the neat reaction.

Example 15

An amount of 0.14 gram of KOH was dissolved in 400 grams n-butanol. To this, 75 grams of DMSO and 0.26 gram of acetone were added. This mixture was heated to 175°C. In one run, 65 grams of ethylene oxide (EO) were added in 0.5 hour. The peak oxide level was 0.6 percent and the reaction was complete in 1.5 hours. In a second reaction, 65 grams of ethylene oxide was fed in 0.5 hour. The peak EO level was 0.44 percent and the reaction was complete in 1.0 hour. By comparison of modeling data, it can be concluded that the solvated runs were 1.8 to 2.5 times faster than comparable reactions run without the addition of polar aprotic solvents.

Example 16

Two grams of potassium hydroxide were dissolved in 200 grams of tripropylene glycol n-butyl ether and loaded in the Parr reactor along with 200 grams of DMSO and 1.0 gram of acetone. After heating to 150°C, propylene oxide addition was started. An amount of 420 grams of PO was added over a 2 hour period. At 2.25 hours, no unreacted PO remained. Analytical results showed a peak oxide level of was 1.5 percent at 1.5 hours. By comparison of peak oxide levels, this indicated an 8.4 times rate increase over the same reaction run without the presence of any polar aprotic solvent and acetone.

It should be realized by one of ordinary skill in the art that the invention is not limited to the exact configuration or methods illustrated above, but that various changes and modifications may be made without departing from the spirit and scope of the invention as described within the following claims.

Claims

I . A process of alkoxylating an alcohol comprising contacting the alcohol with an alkene oxide in the presence of an aprotic polar solvent and a catalyst, under conditions sufficient to form monoalkyl ethers of a glycol or a polyalkylene glycol. 2. The process of Claim 1 wherein the catalyst is selected from the hydroxides of alkali metals.

3. The process of Claim 1 wherein the aprotic polar solvent is selected from the group consisting of dimethylsulfoxide, n-methylpyrrolidinone, acetonitrile, dimethylformamide, dimethylacetamide, and tetramethylene sulfone, or mixtures thereof. 4. The process of Claim 1 wherein the aprotic polar solvent is dimethylsulfoxide or solvent mixtures containing it.

5. The process of Claim 1 or 4 wherein an aldehyde or ketone or nitrile is added to the aprotic polar solvent.

6. The process of Claim 5 wherein the aldehyde or ketone or nitrile is selected from the group consisting of acetone, methyl ethyl ketone, methyl isobutyl ketone, 2,6- dimethyl-4-heptanone, cyclohexanone, acetophenone, propionaldehyde, acetonitrile and urea, or mixtures thereof.

7. The process of Claim 5 wherein the aldehyde or ketone or nitrile is added in an amount of from 20 ppm to 10,000 ppm with respect to the dimethylsulfoxide. 8. The process of Claim 1 wherein the alcohol is monohydroxylated.

9. The process of Claim 8 wherein the alcohol is selected from the group consisting of methanol, ethanol, n-propanol, i-propanol, n-butanol, allyl alcohol, phenol, and mixtures thereof, and the alkene oxide is selected from the group consisting of ethylene oxide, propylene oxide and butylene oxide. 10. The process of Claim 1 wherein the aprotic polar solvent is added in an amount of from 0.1 gram aprotic polar solvent per gram of alcohol to 1 gram of aprotic polar solvent per gram of alcohol.

I I . The process of Claim 5 wherein the alcohol is n-butanol, the alkene oxide is propylene oxide, the reaction is carried out at a temperature off from 80°C to 200°C, the aprotic polar solvent is dimethylsulfoxide (DMSO) and is present in an amount of from 0.1 gram to 1 gram DMSO per gram of n-butanol, and the ketone is acetone and is present in an amount of from 20 ppm to 10,000 ppm with respect to the DMSO.

12. In a process for alkoxylating an alcohol where the alcohol is contacted with an alkene oxide in the presence of a catalyst, the improvement comprising adding an amount of aprotic polar solvent, and the aprotic polar solvent is able to increase the rate of reaction between the alcohol and the alkene oxide by at least 25 percent when compared to the same process without the aprotic polar solvent.

13. The product made by the process of Claim 1, Claim 11 or Claim 12.