FIELD OF THE INVENTION The invention relates to the preparation of aldols and other beta-hydroxy carbonyl compounds. More particularly, the invention relates to aldol condensation reactions catalyzed by a base-modified clay.

BACKGROUND OF THE INVENTION Aldol condensation reactions are commercially important for preparing intermediates of compounds used to make lubricants, adhesives, coatings, plastics, fibers, solvents, plasticizers and other additives. An aldol condensation, shown in scheme (1), is the reaction of an aldehyde or ketone having an alpha-hydrogen atom with another aldehyde or ketone to form a beta-hydroxy aldehyde or beta-hydroxy ketone. where R, R2, R3, R4 and RS are each a hydrogen atom or an organic group.

In one application, the carbonyl group of the aldol is hydrogenated to form 1,3-diols, such as neopentyl glycol (NPG), which are in high demand for producing polyesters and polyester resins. The hydrogenation reaction is shown in equation (2):

The aldol may also be is dehydrated, forming a carbon-carbon double bond, which itself may also be subsequently hydrogenated to yield an aldehyde or ketone. These reactions are useful in the production of commercially important products, such as crotonaldehyde, methyl amyl ketone (MAK), methyl isobutyl ketone (MIBK), and 2-ethyl-1-hexanol. The dehydration/hydrogenation reactions are shown in equation (3): Catalysts previously used in aldol condensation reactions are typically moderately to strongly basic. These reactions are typically carried out industrially with homogeneous liquid catalyst solutions such as aqueous sodium hydroxide, aqueous sodium carbonate, aqueous calcium hydroxide, or trialkylamine. For example, a review of the synthesis of NPG by Cornils and Feichtinger in Chem.

Zeitung 100,504 (1976) states that alkali hydroxides, alkali carbonates, and tertiary amines are the most common catalysts for this reaction.

These prior homogeneous liquid catalyst solutions, however, have numerous drawbacks. First, these catalysts fail to adequately minimize side reactions.

Competing side reactions reduce the yield and purity of the desired product. One

common side reaction is a Cannizzaro reaction, where two moles of aldehyde react to produce a mole of alcohol and a mole of carboxylic acid or carboxylic acid salt.

These acids and acid salts must be removed prior to any hydrogenation step or poisoning of the catalyst may occur. Another side reaction is the Tischenko reaction, where two moles of aldehyde react to form the corresponding ester.

A second drawback to the use of homogeneous liquid catalyst solutions is the large amount of salt waste that must be separated, treated, and disposed. Regulatory and economic conditions demand the reduction or elimination of total dissolved solids (TDS) salt waste that is released into the environment.

Third, with liquid catalysts or other catalyst solutions, the step of separating the catalyst from the product is time-consuming and expensive. Use of a catalyst that is easy to separate, therefore, would be highly advantageous.

Several attempts have been made to overcome some or all of the disadvantages of liquid catalysts by using insoluble, heterogeneous solid base catalysts instead.

Dartt et al., Catalysis Today 19,151, (1994), for example, reviews various attempts to modify zeolites for base catalysis. Dartt et al. describe, for instance, how weak basic character can be imparted by ion exchanging zeolites with large alkali metal cations such as cesium or rubidium. Alternatively, strong basic catalysts have been made by depositing alkali metals on supports such as sodium on alumina, silica, or zeolites, and potassium on graphite; however, such catalysts are impractical due to high air-and water-sensitivity.

Reichle, U. S. Pat. No. 4,476,324, describes the use of synthetic hydrotalcite type mineral catalysts for the vapor phase production of isophorone and mesityl oxide from acetone. However, in many commercial processes that involve aldol reactions, it is undesirable to make higher molecular weight products beyond the first condensation. Vaporization also has the drawback of requiring either low reactant concentrations, which results in lower overall production, or high temperatures, which can result in increased byproduct formation, decomposition of the desired product, or coking of the catalyst, depending on the exact reaction conditions.

Holmgren et al., U. S. Patent Nos. 5,144,089 and 5,254,743, describe the use of hydrotalcite type materials in the self-condensation of n-butyraldehyde to 2-ethyl-2-hexenal in the liquid phase. A major drawback of this process is the high temperatures (approximately 200 °C) required to obtain reasonable activity. High temperatures such as these often increase the production of unwanted by-products, decrease selectivity, and degrade the organic reactants and/or products.

Thus, a need exists for a catalyst that can maximize the yield of an aldol condensation reaction while minimizing the formation of unwanted side products and impurities. It is also desirable to find a catalytic process for aldol condensation that would reduce or eliminate the amount of TDS salts waste, can be performed at moderate temperatures and pressures, and would be economically feasible on a commercial scale.

SUMMARY OF THE INVENTION In view of the industry's need for aldols and other beta-hydroxy carbonyl compounds, the invention offers an improved catalyst and an improved process for preparing beta-hydroxy carbonyl compounds via an aldol condensation reaction.

One embodiment of the invention is a process for preparing a beta-hydroxy carbonyl compound via an aldol condensation reaction. This process includes the reacting of a first carbonyl compound having at least one alpha hydrogen with a second carbonyl compound in the presence of a catalytic amount of a base-modified clay, thereby producing a beta-hydroxy-containing carbonyl compound.

Additional embodiments of the invention are processes for preparing a 1,3-diol, a ketone, or an alcohol. The 1,3-diol is prepared by hydrogenating a beta-hydroxy-containing carbonyl compound. The ketone, such as methyl amyl ketone, is prepared by dehydrating a beta-hydroxy-containing carbonyl compound followed by hydrogenation. The alcohol is prepared by further hydrogenating the ketone.

DETAILED DESCRIPTION OF THE INVENTION According to the invention, a beta-hydroxy carbonyl compound is prepared by a process using a base-modified clay to catalyze the conversion of two carbonyl compounds, at least one having an alpha hydrogen, to a beta-hydroxy carbonyl compound via an aldol condensation. This reaction is shown above in equation (1).

The Base-Modified Clays In one embodiment, the invention relates to a base-modified clay having secluded conjugate base sites and exchangeable interstitial cationic spaces.

Preferably, in the base-modified clay, all interstitial hydroxyl groups have been converted to oxide sites, at least one structural hydroxyl group has been converted to an oxide site, and the modified clay contains sufficient conjugate base cations to balance the charge of the oxide sites. In other words, the base-modified clay contains an actual number of sites per unit cell where an alkali metal or other base cation has been exchanged for the hydroxyl proton that exceeds a calculated number of exchangeable sites per unit cell in the unmodified clay.

Any clay having exchangeable hydroxyl sites may be used in this invention.

The term"clays"encompasses not only clays, but also similar exchangeable minerals. Clays are generally microcrystalline, semi-microcrystalline, or amorphous mineral compounds of hydrated silicas and hydrated alumina, magnesias, iron oxides, and/or minor amounts of various other elements many of which exist as point impurities in the silica, alumina, magnesia, and/or iron oxide phases. Clays that are particularly useful in practicing the invention are the layered and exchangeable smectite clays, the partially exchangeable and hydrated illite clays, the hydrated chlorites, and other minerals with expandable and exchangeable interstitial cation spaces. These clays include, but are not limited to, montmorillonites, beidellites, nontronites, saponites, stevensites, sepiolite-palygorskites and hectorites, so-called degraded illites, hydrated chlorites, micas and other exchangeable and solvatable minerals including brucites, aluminas and silica gels.

Clays are typically acidic in their naturally occurring state. Because this acidity would catalyze undesirable side reactions, the clays of the invention are modified by

treating them with a basic cation-containing solution. The acidic or basic character of a particular clay depends largely on the nature of the water of hydration. The water may exist as structurally bound hyrdoxyl groups (OH groups) as well as loosely bound H2O. While intending not to be bound by any particular theory, it is believed that this base-modification serves two purposes. First, it neutralizes substantially all of the acid sites (OH sites) within the clay and replaces them with their conjugate basic sites (O-). Second, the treatment impregnates or dopes the clay with cations from the base.

The base-modified clays of the invention can generally be characterized by the following properties. The crystalline fraction and the crystallographic structure, as shown by X-ray analysis, remains the same between the unmodified clay and the cation-doped clay. The ratios of the framework atoms, such as Si, Al, Mg, and Fe for the cation-doped clay remains substantially the same as the unmodified clay.

However, the base cation is present in the base-modified clay in a concentration ranging from about 0.005 to about 20 weight percent, preferably from about 0.1 to about 15 weight percent, and even more preferably 0.5 to 10 percent weight. At low base concentrations, the catalyst activity remains low since the number of active sites is correspondingly small. At higher base concentrations, the clay acts more and more like a homogeneous base catalyst, and the catalyst selectivity for only primary aldol/Tischenko products diminishes.

It is believed that the cation present is in the form of chemically bound ions, not merely adsorbed, because they remain in the catalyst after repeated washings with solvent. While not desiring to be bound by any theory or scientific explanation, it is believed that the basified clays display unusual basic properties deriving from the location of the base within the highly inaccessible interior of the clay's layer structure and occurring in addition to typical base properties provided by base sites within the readily accessible interstitial cavity or at the clay's surface. It is believed that under the forcing conditions of the catalyst preparation, the unusual base properties form when the reagent base anions penetrate the interior of the clay's layer structure converting the sequestered acidic structural hydroxyl (OH) groups into their Bronstead conjugate base

form, oxide (O-) groups.

It is believed that the cationic counterions remain separate, largely within the interstitial cavity bound by ionic forces in order to balance charge but physically too large to fit into the clay's layer structure unless the outer shell of the clay layer is otherwise disrupted. It is believed that the resulting forced charge separation serves to further increase the basicity and further enhance the unusual basic properties. It is also believed that the number of basic sites, and thus the activity, and is directly proportionate to and therefore indicated by the amount of cation contained within the catalyst.

It is believed that lithium is an anomalous exception to basified clays described above due to its small size and propensity to form less ionic or covalent bonds. It is believed that the lithium cation may penetrate the clay's layer structure along with the base anion converting hydroxyl (OH) groups into tightly bound lithium oxide ion pairs, (OLi) groups, and displacing other ions especially in the octahedral layer sites. It is believed that this pervasive infusion explains why lithium loadings are often 2-4 times larger than other metal cations on a gram atom basis.

The base-modified clay contains sufficient metal cations to balance the charge of all oxide sites within the clay. Suitable metal cations include alkaline metals, such as the group 1 alkali metals, the group 2 alkaline earth metals, and various other metals, such as zinc, cesium, rubidium, and thallium. Preferred counterions are sodium, potassium and lithium. A base-modified clay may have a mixture of metal cations. A base-modified clay of the invention should have greater than or equal to 1.2 metal cations per unit cell. The number of metal cations per unit cell will vary depending on the metal used in the base-modification. The amount of metal cation in the clay can be readily determined by elemental analysis or other techniques known on the art. The amount of base-modification can be readily determined by comparing the elemental analysis of the starting clay and that of the base-modified clay. For example in the case of sodium, the concentration of metal cation in the base-modified clay should be greater than about 0.005 weight percent; more preferably, greater than about 0.1 weight percent; and most preferably, greater than 1 weight percent. Sodium-modified clays,

for example, can have up to three or more sodium cations per unit cell giving a sodium concentration of 8 to 9 weight percent.

Lithium-modified clays have been found to have high concentrations of lithium cations. Lithium-modified clays have up to twelve lithium cations per unit cell. The lithium cation concentration may be as high as 10 weight percent in the base-modified clay. Preferably, the lithium cation concentration ranges from 0.1 to 15, and more preferably from 0.5 to 12.

Preferred base-modified clays of the invention not only have secluded conjugate base sites but have cavities or spaces penetratable by the organic substrate adjacent to the base sites. To achieve these ends, the cavities preferably have a gallery width that is less than about 10.2 A, as determined by the X-ray doo, spacing and the clay layer width. More preferably, the gallery width ranges from about 9 A to about 10.2 A. It is also preferable that the cavities are wettable. Unwettable clays with neutral hydrophobic cavities, such as the talcs, or with highly charged cavities having tightly complexing counterions, such as the micas, are generally unsuitable.

The base-modified clay can be used in various forms to catalyze a reaction. For instance, the base-modified clay can be formed into granules, particles, pellets, spheres, tablets, rings, or extrudates and placed into a fixed bed reactor, as is known in the art.

Preferably, the minimum dimension of the catalyst is greater than about 0.02 inches.

In this variation of the invention, the reactants are passed over the catalyst bed and the crude product is collected at the other end. Alternatively, the catalyst can be stirred with the reactants as a slurry and the crude product separated from the catalyst by filtration, gravity settling, or centrifugation. In either case, the fact that the catalyst is a solid material and substantially insoluble in the reaction medium makes separation of the catalyst straightforward. It also avoids many costly environmental problems that are associated with the treatment and disposal of aqueous waste streams containing excess soluble base catalysts.

Preparation of Base-Modified Clays Another embodiment of the invention is a process for preparing a base-modified clay. To prepare a base-modified clay of the invention, the clay is treated with a basic

solution of a compound having a basic anion or an uncharged species capable of abstracting a proton from a hydroxyl group within the clay. The species which extracts the hydroxyl group may, as explained below, be generated after the clay is treated with the compound. The solution may include either salts of a basic anion or other uncharged species in a solvent or as a neat liquid. As discussed above, the basic cation is preferably an alkaline metal cation from group 1 or 2 of the periodic table or other suitable metal cation. Suitable basic anions include, but are not limited to, hydroxide, amide, carbonate, silicate, metasilicate, phosphate, cyanide, borate, tetraborate, hydride, and aluminate. Suitable uncharged species include, but are not limited to, ammonia, amines, molten alkali metals, and alkali metal amalgams. Additional basic anions useful in the invention include other anions that are not intrinsically basic but that can be made basic with further treatment. Mixtures of basic compounds may be used in the base-modification to give a base-modified clay have a mixture of metal cations.

The solvent used in the preparation of the base-modified clays is not critical, provided its dielectric constant or complexing ability is high enough to serve as a solvent for both the entering basic moiety and the leaving cation and provided that it thoroughly wets the clay. Wetting is important for good base exchange. Suitable solvents include, but are not limited to, water, anhydrous ammonia, aqueous ammonia, and polar organic solvents such as ethers, alcohols, amines, ketones, dimethyl sulfoxide, and dimethyl formamide. Preferred solvents include liquid ammonia, dimethyl formamide and, more preferably, water.

One method of preparing a base-modified clay heats the clay in a basic solution for a sufficient amount of time to convert acidic hydroxyl (OH) sites, including interstitial and at least one structural hydroxyl, to oxide sites (O-). The heating is preferabaly with stirring, at reflux, or under pressure. As discussed, the oxide sites are the conjugate bases of the acidic hydroxyl sites. Preferably, the process includes heating a slurry of a clay in a basic solution for an amount of time sufficient to effect an exchange of a base for substantially all acidic sites on the surfaces of the clay. More preferably, the heating should continue until all interstitial hydroxyl groups of the clay have been converted to oxide sites and at least one structural hydroxyl group has been

converted to an oxide site.

The reaction conditions for preparation of the base-modified clay are not critical.

Generally, however, the temperature should be within the range of about 0 °C and about 500 °C, preferably about 40 °C and about 120 °C. Likewise, the reaction time is not critical, but generally ranges from about 0.5 h to about 100 h, preferably at least 6 h. If the heating time is too short, an insufficient of base-modification will take place.

If it is too long, the clay may begin to decay or even decompose. The ratio of basic solution to clay should be greater than about 1: 10 by weight. Preferably the ratio ranges from about 2: 1 to about 100: 1.

Typically, the conversion acidic hydroxyl (OH) sites, including interstitial and at least one structural hydroxyls, to oxide sites (0') can be accomplished by heating for about 4 to 24 hours in a 5-50 percent preferably 10-30 percent aqueous solution of a Brönsted base. As long as the base is present in large excess, the actual concentration is not critical although lower concentration may require longer heating times. Heating times will vary depending on the particular clay and the number of hydroxyl groups to be converted. Preferably the base-modified clay may be prepared by heating (preferably with stirring, at reflux, or under pressure) the clay for 6 to 12 hours in a 5-50 percent aqueous solution of an alkaline hydroxide, such as sodium hydroxide, lithium hydroxide, potassium hydroxide, magnesium hydroxide, calcium hydroxide and the like.

A base-modified clay of the invention may also be prepared by mixing the clay with a solution of a salt or other compound having an alkaline metal in combination with a species that undergoes combustion or pyrolysis to yield a basic species capable of extracting a proton from a hydroxyl group within the clay. A stoichiometric excess of the salt or other compound is first dissolved in an appropriate solvent and then mixed with the clay to treat the clay introducing the salt or other compound into the clay structure. The solvent is then removed or the clay isolated (e. g. by filtration) and the treated clay subjected to pyrolysis and/or combustion. Suitable water-soluble anions for this pyrolysis or combustion technique include, but are not limited to, carbonate, bicarbonate, nitrate, nitrite, azide, aliphatic and aromatic carboxylates, phenoxides,

unsubstituted and substituted cyclopentadienes, and 1,3-dicarbonyl derivatives.

To illustrate, consider alkaline earth hydroxides, which have poor solubility in water. A soluble form of the cation can be dissolved in a solvent, intimately mixed with the clay, then the treated clay is subjected, for example, to pyrolysis or a combustion /pyrolysis process at 150 °C to 1250 ec, thereby releasing small, neutral molecules which are often anhydrides of inorganic acids. The acid anhydrides then react with the hydroxyl sites in the clay affecting the desired base-modification. For example, a metal carbonate would decompose emitting carbon dioxide which, under oxidizing conditions, yields an oxide to extract the hydroxyl proton and, thereby, affecting the basic modification of the clay. Nitrates similarly decompose to an oxide under oxidative conditions. To affect this pyrolysis or combustion, the clay should be heated to a temperature less than its fusing temperature, avoiding a change in molecular structure of the clay. Preferred temperatures ranges from about 300 °C to 500 °C, particularly for clays having a fusing temperature at or just above 600 °C.

Preparation of a Beta-Hvdroxv Carbonyl Compounds One embodiment of the invention provides a process for preparing a beta-hydroxy carbonyl compound, such as an aldol. This compound is prepared by reacting in a reaction vessel two carbonyl compounds, at least one of which has an alpha hydrogen, in the presence of a catalytic amount of a base-modified clay. Scheme (1), repeated below, depicts this reaction.

Suitable first carbonyl compounds include, but are not limited to, those having the formula RICOR2, where R, and R2 are selected from the group consisting of hydrogen; Cl 20 alkyl, C2 20 alkenyl, C2 20 alkynyl; C3 20 cycloalkyl or cycloalkenyl; aryl moieties such as phenyl, substituted phenyl, napthyl, phenanthryl, or biphenylyl; or arylalkyl moieties where the aryl group is phenyl, substituted phenyl, napthyl,

phenanthryl, or biphenylyl and the alkyl group contains from 1 to 20 carbon atoms.

Specific examples include formaldehyde, acetone, acetaldehyde, propanal, butanal, pentanal, hexanal, cyclic ketone, such cyclobutanone, cyclopentanone, cyclohexanone, cycloheptanone, or cyclooctanone and substituted versions of these cyclic ketones.

Where R, or R2 is an alkyl, alkenyl, or alkynyl, those groups may be straight or branched, unsubstituted or substituted.

Suitable second carbonyl compounds include, but are not limited to, those having the formula R3CHR4COR5, where R3 and R4 are selected from the group consisting of hydrogen, Cl 20 alkyl, C2 20 alkenyl, C2 20 alkynyl, C3 20 cycloalkyl or C2 20 alkenyl; alkoxy moieties having from 1 to 20 carbon atoms; aryl moieties such as phenyl, substituted phenyl, napthyl, phenanthryl, or biphenylyl moiety; or arylalkyl moiety where the aryl group is a phenyl, substituted phenyl, napthyl, phenanthryl, or biphenylyl moiety and the alkyl group contains from 1 to 20 carbon atoms; and where RI is selected from the group consisting of hydrogen, Cl 20 alkyl or C2 20 alkenyl, C2 20 alkynyl, C3 20 cycloalkyl, substituted C3 20 cycloalkyl or C2 20 alkenyl, aryl moieties where aryl is a phenyl, substituted phenyl, napthyl, phenanthryl, or biphenylyl moiety, or an arylalkyl moiety where the aryl group is a phenyl, substituted phenyl, napthyl, phenanthryl, or biphenylyl moiety and the alkyl group contains from 1 to 20 carbon atoms. Where R3, R4, or R5 are alkyl, alkenyl, alkynyl, or cycloalkyl those groups may be straight or branched, unsubstituted or substituted. Specific examples include 2-alkylpropanals, 2-alkylbutanals, 2-alkylpentanals, 2-alkylhexanals, acetaldehyde, propanal, butanal, pentanal, hexanal, formylcyclohexane, formylcyclopentane, formylcylcobutane, formylcyclopropane, 2-formyltetrahydrofuran, 3-formyltetrahydrofuran, 2-formyltetrahydropyran, 3-formyltetrahydropyran, and 4-formyltetrahydropyran.

The reaction conditions for the condensation reaction are generally the same as those employed for base catalyzed aldol-type reactions known in the art. Preferably, both the first and second carbonyl compounds are either in a liquid phase, acting as the reaction solvent or dissolved in a solvent. The reaction temperature can range from about-20 °C to about 450 °C, preferably from about 10 °C to about 150 °C. The

reaction time can be for as long or as short a period as desired, but typically ranges from about 1 minutes to about 50 hours and preferably is about 20 minutes to 6 hours. The pressure should generally be greater than about 7 KPa, preferably greater than about 100 KPa. The upper limit of pressure is that required to keep all reactants and solvents in the liquid phase at the reaction temperature.

If the reactants are in poorly miscible or immiscible liquid phases (e. g., an aqueous phase and an organic phase), then an organic solvent may be used to make the phases of the reaction mixture miscible. Examples of such solvents include, but are not limited to, ethers, alcohols, amines, ketones, dimethyl sulfoxide, dimethyl formamide, or mixtures thereof. Preferably, the organic solvent is an alcohol having the formula ROH, where R is selected from the group consisting of Cl 22 alkyl or C2 l2 alkenyl or C3 12 cycloalkyl; an ether of formula R6OR7, where R6 and R7 are selected from the group consisting of C,, 2 alkyl, C2, 2 alkenyl, C3, 2 cycloalkyl. Preferably the organic solvent is tetrahydrofuran, tetrahydropyran, methanol, ethanol, propanal, isopropanal or diethyl ether. Most preferably the organic solvent is methanol.

The solvent should be added in a large enough amount that the reaction mixture remains homogeneous throughout the course of the reaction, yet small enough that the reactants are not diluted to the point of slowing down the reaction. Typically, the amount of homogenizing solvent ranges from 0 to about 75 % by weight. The preferred range for typical carbonyl-containing reactants is 5 to 50 % by weight with 10 to 40 % by weight being especially preferred.

Preparation of 13-diol Another embodiment of the invention is a process for preparing a 1,3-diol. A beta-hydroxy-containing carbonyl compound is prepared as described above. It is then hydrogenated using techniques known in the art to produce a 1,3-diol having a formula R'C (OH) R2CR3R4CH (OH) R5. A particularly desirable 1,3-diol is neopentyl glycol, the hydrogenated product of isobuytraldehyde and hydroxypivaldehyde.

Any of the hydrogenation techniques known in the art may be used in the process of the invention. An aldol can be hydrogenated, for example, by contacting it with a metal catalyst under a hydrogen atmosphere. The base-modified clay catalyst

and an appropriate metal hydrogenation catalyst can be physically mixed together in a fixed bed format or in a stirred tank with the catalysts present as a slurry. In such an embodiment the metal hydrogenation catalyst is impregnated or adsorbed on to the clay catalyst using techniques known in the art. Alternatively, the clay could be used as a support for the metal hydrogenation catalyst. A base-modified clay of the invention may also serve as support for the metal hydrogenation catalyst. These supported clay catalysts may be prepared in the same way as known clay-supported hydrogenation catalysts.

Preferred hydrogenation catalysts for selectively reducing the carbon-carbon double bond include transition metal catalysts having active components of Cr, Co, Ni, and Cr; noble metal catalysts having active components comprised of Ru, Rh, Pd, Re, Ir, and Pt; and various combinations of these metals. These metals can be utilized unsupported, such as Raney nickel, or incorporated on one of many suitable supports including alumina, carbon, silica, zeolites, and clays. Especially preferred catalyst systems are Pd on carbon and Pd on alumina.

Preparation of a Ketone or Alcohol A ketone or an alcohol can also be prepared from a beta-hydroxy-containing carbonyl compound. The compound is first dehydrated using standard techniques known in the art. The resulting a, ß-unsaturated aldehyde or ketone is then hydrogenated as described above, thereby producing a ketone having a formula R'CHWCHR'C (O) R5. An alcohol is prepared by further hydrogenating this ketone as described above, thereby producing an alcohol having a formula R'CHR2CHR3CH (OH) R5.

EXAMPLES The following examples are presented to illustrate, not limit, the disclosed invention. Although the invention is described in its preferred forms and with a certain degree of particularity, it is to be understood that changes can be made without departing from the spirit and scope of the invention. It is also to be understood that these examples do not define the limits under which the base-modified clay catalysts

will perform.

EXAMPLE 1 LiOH-Modified Montmorillonite KSF Clav Preparation A 1-Liter round bottom flask equipped with an overhead mechanical stirrer and a reflux condenser was charged with 201.3 grams of Montmorillonite KSF clay, 600 milliliters deionized water, and 152.7 grams of lithium hydroxide monohydrate (3.640 moles). This mixture was stirred and heated at reflux overnight. The slurry was then allowed to cool to room temperature. The solids were separated from the liquid by centrifugation. The liquid portion was decanted and discarded. The solids were washed by resuspending in 600 milliliters of deionized water and centrifuging.

This sequence was repeated two more times to ensure removal of all soluble species.

The catalyst was dried in an oven at 100 °C with flowing air for 24 hours. The yield of the final material was 214.2 grams. X-ray diffraction analysis of this material exhibited the characteristic pattern for montmorillonite with a doo, basal spacing of 10.2 A, identical to the untreated montmorillonite KSF clay. The lithium content of the material increased from 0.004 weight percent Li to 9.3 % by weight.

EXAMPLE 2 LiOH-Modified Montmorillonite K10 Clay Preparation Example 1 was repeated using 205.5 g of Montmorillonite K10 clay in place of the KSF type clay. All other conditions were essentially the same. The yield of the dried material was 209.7 g. X-ray diffraction analysis of this material exhibited the characteristic pattern for montmorillonite with a dool basal spacing of 10.2 A, identical to the untreated montmorillonite K10 clay. The lithium content of the material increased from 0.002 weight percent Li to 9.0 % by weight. This shows that different types of clay can be modified in the same manner.

EXAMPLE 3 Reaction of Isobutyraldehvde and Formaldehyde with LiOH-Modified KSF Cl A mixture of isobutyraldehyde (78.68 g, 98% pure, 1.069 mol) and formaldehyde (29.98 g, 36.7% aq. solution, 0.366 mol) was placed into a 500 ml

3-neck round bottom flask equipped with overhead stirrer. The LiOH modified KSF type montmorillonite clay from Example 1 (5.09 g) was added and the mixture was allowed to stir at ambient temperature. After 3 hours, the reaction had reached 61% formaldehyde conversion with 95.8% selectivity to hydroxypivaldehyde (HOHPv).

After 24 hours, the conversion was 99% and selectivity was still 95.8%. The observed first order rate constant, kob5, is 0.0228 hr-' (g cat/L)-'where rate=-d CH2O/dt =kobs CH20 catalyst This experiment shows that the LiOH-modified KSF montmorillonite clays are highly active and highly selective catalysts for the desired aldol condensation reaction.

COMPARATIVE EXAMPLE 4 Reaction of Isobu, aldehyde and Formaldehyde with Unmodified K10 Clay The experiment of Example 3 was repeated using 5.02 g of unmodified K10 type montmorillonite clay as the catalyst. After 169 hours very little formaldehyde had reacted, but 26% of the isobutyraldehyde had been consumed. Selectivity to HOHPv was 8.8%. The overall product distribution was 61% triisopropyltrioxane, 29% diisopropyltrioxane, 5% HOHPv, 2% trioxane, and the balance higher boiling compounds. Thus, the untreated montmorillonite clays have low activity and poor selectivity for the desired reaction.

EXAMPLE 5 NaOH-Modification of Montmorillonite K10 Clay The charge to a 1 liter round bottom flask equipped with an overhead stirrer and a reflux condenser was 200.3 grams of Montmorillonite K10 clay, 200 milliliters of 50 percent aqueous sodium hydroxide (d = 1.5329,153 grams NaOH, 3.83 moles), and 400 milliliters of water. Stirring and heating this mixture to reflux for 6 hours completed the base exchange. The workup consisted of a vacuum filtration of the

mixture using a medium porosity glass frit funnel. Washing the filter residue with 3 x 150 milliliter portions of distilled water removed any residual caustic. Finally, air drying and oven drying at 300 °C for 4 hours completed the workup. The yield of the final material was 171.4 grams (85.6 percent). Analysis of this material showed the critical d, oo interstitial spacing increased from 10 A to 12 A as a result of this exchange. The sodium content of the material also increased from 0.4 percent to 1.2 percent.

EXAMPLE 6 KOH-Modification of Montmorillonite K10 Clav Example 5 was repeated using 100 grams of 85 percent potassium hydroxide in place of the caustic solution. All other conditions were the same. Starting with 200 grams of the Montmorillonite K10 catalyst, the yield of the dried, treated product was 174 grams (87 percent).

EXAMPLE 7 NaOH-Modification of Montmorillonite KSF Clav Example 5 was repeated using 200 grams of Montmorillonite KSF in place of Montmorillonite K10. All other conditions were the same. Starting with 200 grams of the Montmorillonite KSF, the yield of the dried, treated product was 171.5 grams (85.8 percent). Elemental analysis showed a sodium content of 11.0 weight percent.

COMPARATIVE EXAMPLE 8 Reaction of Isobut,raldehyde and Formaldehyde with NaOH-Modified KSF Cl The experiment of Example 3 was repeated using as the catalyst 5.32g of the NaOH-modified clay catalyst of Example 7 instead of LiOH. In addition the reaction was carried out at 60 °C to accelerate the slower reaction. After 75 hours, the reaction had reached 50% formaldehyde conversion with 92.6% selectivity to HOHPv. At 168 hours, the conversion was 90% and selectivity was 80.4%. As in

Example 3, the observed rate was first order with kobs equaling 4.03 x 10'"ho' (g cat/L)-', or roughly two orders of magnitude slower than the rate in Example 3.

This shows that while montmorillonite clays modified with other alkali metal hydroxides are selective catalysts for the desired aldol condensation reaction, the activity is significantly lower than those modified with lithium.

EXAMPLE 9 Reaction of Isobutyraldehyde and Formaldehyde with LiOH-Modified KSF Clay and a Homogenizing Solvent A mixture of isobutyraldehyde (375.4 g, 98% pure, 5.10 mol) and formaldehyde (167.6 g, 36.6% aq. solution, 2.04 mol) was placed into a 1 L reaction vessel equipped with a magnetic stir bar and blanketed with nitrogen atmosphere. To this was added 120 ml of methanol, at which point the mixture became homogeneous.

The reaction vessel was placed into a water bath at ambient temperature in order to moderate temperature changes. The LiOH modified KSF type montmorillonite clay from Example 1 (8.01 g) was added and the mixture was allowed to stir at ambient temperature. After 2.5 hours, the reaction had reached 93% formaldehyde conversion with 89.8% selectivity to hydroxypivaldehyde (HOHPv). As in Example 3, the observed rate was first order with kobs equaling 0.106 hr-' (g cat/L)-', or roughly five times faster than the rate in Example 3.

This experiment shows the rate accelerating effect that occurs when the liquid reaction mixture is a single phase as compared to when the liquid reactants are divided between aqueous and organic phases.

EXAMPLES 10-18 Reaction of Formaldehvde and Various Aldehydes with LiOH-Modified KSF Clay In order to demonstrate the versatility of the LiOH-modified montmorillonite clay catalysts, the aldol condensation reaction of formaldehyde with various substrates was performed. For each reaction, the starting aldehyde and formaldehyde in a 2.5: 1 molar ratio were placed in a 250 ml reaction vessel

equipped with a magnetic stir bar and blanketed with nitrogen atmosphere.

Methanol was slowly added until the reaction mixture became homogeneous. The LiOH modified KSF type montmorillonite clay from Example 1 was added in the average amount of 7.5 g catalyst per mole formaldehyde. The mixture was allowed to stir at ambient temperature with the results shown in Table I, below. The observed first order rate constant, kobs, has been corrected for concentration effects due to varying amounts of methanol needed to make the reaction mixtures homogeneous. As can been seen, the LiOH modified clays show good activity and selectivity to a wide variety of aldehyde substrates.

Table I Observed rate P-hydroxy-Selectivity constant Ex. Starting aldehyde aldehyde product (%) (hr~l g cat/L'') n Urbi 10 89.8 0.106 isobutyraldehyde hydroxypivaldehyd e 0 o ! t OH 0 11 94.9 0.145 2- methylbutyraldehyd e Ou 0 0 OH O 12 93.2 0.000869 2- ethylbutyraldehyde 13 89.1 0.0410 0 t ! OH O 2- methylvaleraldehyde 14 1 86.8 0.0139 0 1 OH O OH 0 2-ethylhexanal 0--, O Q 0 15 OH O 84.4 0.461 3-formyl- tetrahydrofuran 0 0 j ruz 16 82.1 0.00470 OH 0 2-formyl- tetrahydrofuran i ICI 17 1 39. 6 0. 238 OH u 2-phenyl- propionaldehyde 18 O OH O 89.1 0.0405 Oh O formylcyclohexane

EXAMPLE 19 Reaction of Acetone and n-Butyraldehyde with LiOH-Modified KSF Clay under Reducing Conditions The LiOH modified KSF type montmorillonite clay from Example 1 (11.67 g) and a commercial 5% palladium on activated carbon catalyst (1.24 g) were placed into a 300 ml stainless steel autoclave equipped with gas entrainment impeller and cooling coils. This autoclave was also charged with acetone (87.1 g, 1.50 mol), nbutyraldehyde (36.0 g, 0.50 mol), and water (15.0 g, 0.83 mol). The vessel was purged with hydrogen and then placed under a 400 psig (2760 KPa) hydrogen atmosphere. The temperature was held at 100 °C with vigorous stirring for 7 hours.

During the course of the reaction, hydrogen was consumed at a linear rate of 0.046 mol/hr. At the end of this time, conversion of n-butyraldehyde had reached about 70%. The product, analyzed by FID GC, exhibited the following selectivities: MAK, 56.4%; 2-ethylhexanal, 36.0%; MIBK, 0.0%; over-condensation products,

7.6%.

This experiment demonstrates the utility of the LiOH-modified clays for another industrially valuable aldol reaction. It is of interest to note that significant amounts of 2-ethylhexanal are produced using this catalyst. 2-Ethylhexanal is a desirable product since it can easily be converted into the commercially valuable compound, 2-EH. In addition, only trace amounts, if any, of the less desirable MIBK are produced. In a typical commercial process, which generally uses aqueous sodium hydroxide as the aldol catalyst, no 2-ethylhexanal is produced but a large percentage of MIBK is produced. In addition, larger amounts of undesirable over-condensation products are typically generated in the commercial process.

COMPARATIVE EXAMPLE 20 Reaction of Acetone and n-Butyraldehyde with Unmodified KSF Clay under Reducing Conditions The experiment of Example 13 was repeated with 10.50 g of unmodified KSF type montmorillonite as the aldol catalyst. In addition, 0.75 g of the 5% Pd on activated carbon was used. After 5 hours at 100 °C and 400 psig (2760 KPa) H2 atmosphere, n-butyraldehyde conversion was only about 30%. The n-butyraldehyde that reacted was converted into three products: n-butanol (73.6%), di-n-butyl ether (20.1%), and 1,1-dibutoxybutane (6.3%). This shows that the untreated montmorillonite clays do not catalyze the desired aldol reaction.

COMPARATIVE EXAMPLE 21 Reaction of Acetone and n-Butyraldehyde with NaOH-Modified KSF Clay under Reducing Conditions The experiment of Example 13 was repeated with 10.07 g of K10 type montmorillonite that had been treated with NaOH instead of LiOH. In addition, 0.85 g of the 5% Pd on activated carbon was used. After 5 hours at reaction conditions (100 °C, 400 psig (2760 KPa) H2 atmosphere), almost no hydrogen had been consumed (0.045 mol).

Analysis of the reaction mixture at the end of the experiment showed that

practically no reaction had taken place. The only species besides starting materials that could be detected were trace amounts of n-butyl-n-butyrate, 2-ethylhexaldehyde, 2-ethyl-1,3-hexanediol-3-n-butyrate, and 2-ethyl-1,3-hexanediol-1-n-butyrate. This shows that montmorillonite clays modified with other alkali metal hydroxides are not active or selective catalysts for the desired aldol condensation reaction.