BACKGROUND OF THE INVENTION History has shown the importance of chirality in medicinal applications and consequently the need for catalytic asymmetric induction of pharmaceuticals is increasing. Optically active ß-amino esters and optically active ß-amino acids are key components in the synthesis of many important pharmaceuticals, and numerous techniques have been investigated for their production, see Enantioselective Synthesis of ßAmino Acids, Juaristi, E., Ed., Wiley-VCH: New York, 1997.

One technique, catalytic asymmetric induction of p-amino esters, has faced two difficult barriers to success. First, many Lewis acid catalysts are deactivated or decomposed by the nitrogen atom of the reactants or products, and second, aldimine-chiral Lewis acid complexes are flexible and often have several stable conformers which may result in both enantiomers being produced. Applicants have discovered a process for the catalytic production of optically active p-amino esters and optically active ß-amino acids using vaulted biaryls as chiral ligands of an organometallic catalytic compound. Others have produced optically active p-amino esters by catalytic enantioselective Mannich-type reactions using a BINOL-Zirconium catalyst, see Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc., 1997,119, 7153-7154 and Kobayashi, S.; Ishitani, H.; Ueno, M. J. Am. Chem. Soc., 1998,120, 431-432. However, applicants have discovered a novel catalytic compound having surprisingly increased induction and reduced dependence on temperature as compared to efforts of others.

SUMMARY OF THE INVENTION The purpose of the invention is to provide a catalytic compound comprising 1: 2 metal : VAPOL ligand where the metal is selected from the group consisting of zirconium, titanium and hafnium and both VAPOL ligands are the same enantiomer.

A specific embodiment of the invention is one where the catalytic compound further contains an activator such as N-methylimidazole, imidazole, N-benzoimidazole, pyridine, and dimethylaminopyridine in the molar ratio 1: 2: 1 mental: VAPOL ligand: activator.

Another purpose of the invention is to provide a method of preparing a catalytic compound by forming a reaction mixture containing VAPOL ligand of at least 95% ee, and preferably at least 99% ee, and a metal alkoxide where the metal is zirconium, titanium or hafnium to form a catalytic compound having a 1: 2 metal: VAPOL ligand where both VAPOL ligands are the same enantiomer. In a more specific embodiment, the reaction mixture may also contain an activator such as N-methylimidazole, imidazole, N-benzoimidazole, pyridine, or dimethylaminopyridine to form a catalytic compound having 1: 2: 1 metal: VAPOL ligand: activator and both VAPOL ligands are the same enantiomer.

Yet another purpose of the invention is to provide a process for producing an optically active ß-amino ester by reacting a heteroketene acetal and an imine, at aldol reaction conditions, in the presence of a catalytic compound comprising from 1: 1 to about 1: 2 metal: VAPOL ligand where the metal is zirconium, titanium or hafnium and both VAPOL ligands are the same enantiomer, to form at least one optically active ß-amino ester. A specific embodiment of the invention is one where the reaction is conducted in the presence of an activator such as N-methylimidazole, imidazole, N-benzoimidazole, pyridine, or dimethylaminopyridine. Another specific embodiment is one where the catalytic compound further contains the activator in the molar ratio of at least 1: V: 1 metal: VAPOL ligand: activator where V is 1 or 2.

Still another purpose of the invention is to provide a process for producing an optically active p-amino acid comprising reacting a heteroketene acetal and an imine at aldol reaction conditions in the presence of a catalytic compound comprising from 1: 1 to about 1: 2 metal: VAPOL ligand where the metal is selected from the group consisting of zirconium, titanium and hafnium and both VAPOL ligands are the same enantiomer to form at least one optically active ß-amino ester; and converting the optically active p-amino ester, at saponification conditions, in the presence of a saponification catalyst, to form at least one optically active p-amino acid.

BRIEF DESCRIPTION OF THE DRAWING The figure is a suggested mechanism for the catalytic cycle of the present invention. For ease of understanding, the mechanism is directed to the reaction: The figure demonstrates merely the suggested mechanism of the metal-VAPOL ligand catalytic compound, other mechanisms may be possible.

DETAILED DESCRIPTION OF THE INVENTION In general terms, the invention is a liquid phase process for producing optically active p-amino esters from the asymmetric aldol addition reaction of a heteroketene acetal and an imine using a novel organometallic complex catalytic compound having a metal complexe with one enantiomer of 2,2'-diphenyl- [3, 3'-biphenanthrene]-4, 4'-diol (VAPOL). The specific preferred optically active ligand, represented by S-VAPOL R-VAPOL is fully described in Bao, J.; Wulff, W. D. J. Am. Chem. Soc. 1993,115,3814-3915, and Bao, J.; Wulff, W. D. ; Dominy, J. B.; Fumo, M. J.; Grant, E. B.; Rob, A. C.; Whitcomb, M. C.; Yeung, S.; Ostrander, R. L.; Rheingold, A. L. J. Am. Chem. Soc.

1996,118,3392-3405; Heller, P.; Goldberg, D. R.; Wulff, W. D. J. Am. Chem. Soc.

1997,119 (43), 10551-10552. For ease of explanation, the optically active ligand described above will be referred to herein as the"VAPOL"ligand. It is preferred that either two R-VAPOL ligands or two S-VAPOL ligands are complexe with a metal (described below) to form the novel catalytic compound. Which enantiomer is chosen depends on the desired enantiomeric product. For example, in a specific reaction, a catalytic compound having two R-VAPOL ligands may asymmetrically catalyze the formation of an S-p-amino ester. Of course, since the designations"R"and"S"are merely naming conventions, the R-VAPOL ligand enantiomers do not always produce"S"enantiomer products. One skilled in the art would readily be able to determine which VAPOL ligand enantiomer is required for the formation of the desired enantiomer product. It is preferred in the present invention that two identical enantiomers of the VAPOL ligand are complexe with a metal. Although not required, it is further preferred that the metal have a +4 valence. The metal is selected from those capable of catalyzing the aldol reaction described herein.

Suitable metals include those in Group IVB of the periodic table, including zirconium, hafnium and titanium. While a 1: 1 molar ratio of metal: VAPOL ligand is successful as a catalytic compound, the preferred molar ratio of metal to VAPOL ligand is 1: 2, and the preferred metal is zirconium. Therefore, the preferred zirconium-VAPOL ligand complex is represented by: The VAPOL ligand is a highly vaulted compound with the diol functionality in a sterically hindered location. The preferred 1: 2 metal: VAPOL ligand molar ratio is surprising since the complex is formed through the metal bonding with each of two very sterically hindered diol functionalities. Examination of 1: 2 zirconium: VAPOL ligand complexes as space filling models reveals that the four oxygen atoms will be greatly favored to be in the same plane. Such a configuration would leave the two open coordination sites of an octahedral zirconium in a trans relationship. This trans relationship of the open coordination sites would lead to an expectation that the 1: 2 zirconium: VAPOL ligand complexes would not be successful for imine substrates that are bidentate since bidentate imine substrates would require two open cis coordination sites, see the catalytic cycle mechanism shown for a BINOL-Zirconium catalyst in Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc., 1997,119,7153 at 7154. However, the expectation is contrary to the discovery presented herein.

These surprising results are further explained via the Figure showing the suggested mechanism for the catalytic cycle of the present invention, which differs significantly from that proposed for a BlNOL-Zirconium catalyst.

The catalytic compound is produced by forming a reaction mixture containing the VAPOL ligand, a metal alkoxide, and optionally an activator in a solvent. It is preferred that the activator be incorporated into the catalytic compound at the time the compound is formed, but it is contemplated that the activator may be added to the catalytic compound at a later time. It is preferred that the mixture be at room temperature. It is also preferred that the VAPOL ligand be of greater than about 95% enantiomeric excess (ee), and also preferred that the VAPOL ligand be of greater than about 99% ee. Enantiomeric excess (ee) is a term well known to one skilled in the art as the mole fraction denoting the ratio of enantiomers in a mixture: [R]- [S] % ee =-» »-x 100 [R] + [S] where [R] and [S] are the concentrations of the (R)-and (S)-enantiomers. It is most preferred that the VAPOL ligand be chemically pure. The metal of the metal alkoxide is zirconium, hafnium or titanium. The activator is discussed in detail below. The mixture is stirred to form the catalytic composite, preferably at room temperature, see Example 1. Either the 1: 1 metal: VAPOL ligand compound or 1: 2 metal: VAPOL ligand may be successfully formed through controlling the amount of VAPOL ligand and metal alkoxide used.

The catalytic compound is used in an amount effective to catalyze the desired aldol reaction. Examples of suitable catalytic compound concentrations include from about 0.5 mole percent to about 20 mole percent of the limiting reagent, and preferred catalytic compound concentrations include from about 2.5 mole percent to about 25 mole percent.

The reactants in the aldol reaction include at least one heteroketene acetal reactant and at least one imine reactant. A preferred imine reactant may be represented by: where the nitrogen of the carbon-nitrogen double bond of the imine is two spacer carbon atoms from X, where X is OH, NH2 or SH. The groups R1, R2, R5 and R6, bonded to the spacer carbon atoms, and R3 and R4, bonded to the carbon of the imine group, can be any suitable functional group derived from compounds including hydrogen, acyclic compounds such as alkanes including linear alkanes, nonlinear alkanes and highly branched alkanes, alkenes, alkynes, and alkadienes, alicyclic compounds including bridged and spiro alicyclic hydrocarbons, mono and polycyclic aromatic compounds, bridged aromatic compounds, hydrocarbon ring assemblies, organic heterocyclic systems, anhydrides, halides, esters, amides, imides, amines, imines, ammonium and other ium groups, ethers, and carbamates. In very general terms, a preferred imine may be described as having the generic formula R3R4C=NCR1 R5-CR2R6X where X and R1-R6 are as defined above.

It is desired that the end product be optically active, and therefore must contain at least one chiral center. Since R3 and R4 are found in the end product, it is preferred that R3 and R4 not be identical in order to provide a chiral center and thereby maintain the optical activity of the end product. However, it is possible for the end product to contain a chiral center other than one involving R3 and R4 and in that case, R3 and R4 may be identical.

Additionally, R1 and R2, along with the spacer carbon atoms, may be imbedded in a ring, such as found in imines derived from 2-amino phenols or 2-aminocyclohexanol. The ring containing R1 and R2 may be aliphatic or aromatic and may be part of a multi-ring assembly or part of a polycyclic group. The ring may be substituted or non-substituted. In some applications, it may be preferred that the substituting groups be alkyl groups such as a methyl group. It is preferred that R1 and R2, along with the two spacer carbon atoms, be imbedded in a ring. Similarly, it is contemplated that R3 and R4, along with the carbon atom of the imine group, may be imbedded in a ring that may be aliphatic or aromatic and may be part of a multi-ring assembly or part of a polycyclic group. Again, the ring may be substituted or non-substituted.

Although the functional groups R3 and R4 may influence the aldol reaction, their identities are selected based upon their presence in the product p-amino ester.

In other words, the functional groups R3 and R4 are preferably selected according to which specific optically active p-amino ester is desired. The groups R1, R2, R5 and R6 may be selected so that the nitrogen carbon bond may be readily cleaved.

Alternatively, the imine reactant may have nitrogen of the carbon-nitrogen double bond of the imine being three spacer carbons atoms from X as shown by: The functional group X is as described above. The functional groups designated as R may be any group as described for R1 through R6 above, with the exception that the groups are not imbedded in a ring. R3 and R4 are as described above. The groups bonded to the spacer carbon atoms are all designated with R simply for ease of understanding and not to imply that they must all be identical; each may be a different functional group. Again, in very general terms, the imine of this embodiment may be described as having the generic formula R2C=N-CR2-CR2-CR2X where X and R are as described above. Suitable examples of imine reactants include those derived from aldehydes and ortho-aminophenol, see the examples, with preferred imine reactants being those derived from an aldehyde. Procedures to form the imine reactants described above are known to one skilled in the art and are not described in detail here.

The second reactant, the heteroketene acetal reactant, is represented by the preferred structure: where R7 and R8 can be any suitable functional group including those derived from hydrogen, acyclic hydrocarbons such as alkanes, alkenes, alkynes, and alkadienes, alicyclic hydrocarbons including bridged and spiro alicyclic hydrocarbons, mono-and polycyclic aromatic hydrocarbons, bridged aromatic hydrocarbons, hydrocarbon ring assembles, organic heterocyclic systems, acids, anhydrides, heteroatoms such as halides, ethers, thioethers, salts, ethers, imides, amines, imines, azido, other nitrogen containing groups and sulfur containing groups. In the case where R7 and R8 include an aromatic ring, the aromatic ring may be substituted or unsubstituted.

Note that although R7 and R8 may have influence on how the aldol reaction occurs, their identities are selected based upon their presence in the optically active p-amino ester product. In other words, the functional groups R7 and R8 are preferably selected according to which specific optically active p-amino ester is desired.

As discussed above, it is desired that the end product be optically active, and therefore must contain at least one chiral center. R3 and R4, discussed above, if not identical, may comprise part of one chiral center, however, if R3 and R4 are identical another chiral center is needed in the end product. The groups R7 and R8 are also found in the end product and may comprise a chiral center. Therefore, in the case where R3 and R4 are identical it is then preferred that R7 and R8 not be identical in order to provide a chiral center and thereby maintain the optical activity of the end product. In other words, it is preferred that at least one of the pairs of R3-R4 or R7-R8 contain groups that are not identical. It is possible that the end product may contain more than one chiral center, and in that case R3, R4, R7 and R8 may all be different functional groups.

X1 and X2 are heteroatoms such as oxygen, nitrogen or sulfur. The functional groups R9 and R10 can be an alkyl or aryl group, or a silicon-containing constituent, with the silicon containing constituent being preferred. At least one of R9 or R10 must be capable of transferring to the nitrogen atom of the imine reactant, see Examples 4,6, and 7. Which group is transferred to the nitrogen atom of the imine reactant can be predicted by one of ordinary skill in the art. It is preferred that at least one of R9 or R10 be a silicon containing group, with the expectation that the silicon containing group will be the one to migrate. Example 4 discloses data resulting from experimentation using different heteroketene acetal, or more specifically, silyl ketene acetal reactants. The data suggests that steric bulky acetals are preferred, as they result in higher asymmetric induction. The term asymmetric induction is well know to one skilled in the art and is used to describe preferential formation of one enantiomer from a prochiral substrate inducted by, for example, a chiral catalyst and preferential formation of one diastereomer by the creation of a new stereogenic center in a chiral molecule. In very general terms, a preferred heteroketene acetal reactant may be described as having the formula R7R8C=C (X1R9) (X2R10) where X1, X2, and R7-R10 are as defined above.

An activator may be used in the present process to aid in catalysis. The activator is a monodentate ligand capable of binding to the apical site of the catalytic compound to maintain an octahedral geometry. Both the size of the activator and the bacisity of the activator influence the enantioselectivity of the catalytic compound, see Example 5. Too large of an activator seems to inhibit the induction. Suitable activators include N-methylimidazole, imidazole, N-benzoimidazole, pyridine, dimethylaminopyridine, with preferred activators being N-methylimidazole and imidazole. The activator may be incorporated into the catalytic compound at the time the compound is formed, or it may be added later. For instance, the activator may be part of the reaction mixture where the catalytic compound is used. Even if the activator is incorporated at the time the catalytic compound is formed, it may also be added to the reaction mixture in an excess to facilitate a high degree of enantiomeric excess in the product. Regardless of when the activator is incorporated, it is preferred to have the amount of activator in the reaction mixture exceed that of the catalytic compound in order to maintain the preferred molar ratio of 1: 2: 1 metal: VAPOL ligand: activator. A preferred amount of activator to add to a reaction mixture is amounts ranging from about 2 mole percent to about 24 mole percent of the limiting reagent. Note, however, that the activator is not limited to the preferred molar ratio of 1: 2: 1: metal: VAPOL ligand: activator, and instead, may be present in an amount greater than the preferred molar ratio of 1: 2: 1 metal: VAPOL ligand: activator. It is contemplated that the activator may range from about 1: V: 1 metal: VAPOL ligand: activator to about 1: V: 2 metal: VAPOL ligand: activator where V may be 1 or 2.

The reaction of the present invention is performed in the liquid phase, preferably using an organic solvent. The solvent must be capable of solubilizing all components of the reaction including the catalytic compound, the reactants, the activators, and the products. Suitable solvents include toluene and dichloromethane.

The results in Example 2 indicate that toluene is the preferred solvent.

The aldol reaction is conducted in a batch mode or a continuous stirred tank reactor mode at typical aldol reaction conditions including atmospheric pressure and a temperature from as fow as about-50°C to about as high as the temperature at which the liquid reaction mixture begins to vaporize. Most likely the upper temperature limit will be determined by the boiling point of the solvent used and examples of upper temperature limits include from about 25°C to about 60°C.

Surprisingly, Example 2 shows that the enantioselectivity of the present invention remains substantially constant even as the operating temperature is increased.

Experimentation shows that even at room temperature 94% ee was still observed, see Table 2 in Example 2. This lack of dependence on temperature is surprising since other catalytic compounds such as BINOL are quite temperature dependent.

As shown in Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc., 1997,119, 7153 at 7154, Kobayashi, S.; Ishitani, H.; Ueno, M. J. Am. Chem. Soc., 1998,120, 431-432, and Examples 2 and 7 the BINOL and 6,6'-dibromoBINOL catalytic compounds perform best at low temperatures.

The desired product of the above reaction is one enantiomer of a p-amino ester where the amino group is in the position beta to the carbonyl group. A preferred product is represented by: where R1-R8 are as defined above. Q represents either X1-R9 or X2-R10 as defined above. Note that if the imine reactant contained three spacer carbons as opposed to two spacer carbons (as discussed more fully above), the structure of the product ester would be different than shown here. Specifically, one skilled in the art would readily understand that under those circumstances the two-carbon chain with R1, R2, R5 and R6 shown above would be replace by the three-carbon chain with the "R"groups of that chain.

In some applications, a nitrogen substituent will be need to be oxidatively removed to arrive at the p-amino ester, see for example, Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc., The desired product may be separated from the reaction mixture by a variety of separation techniques with the preferred technique for small scale production being chromatographic separation.

Also within the scope of the present invention is the subsequent saponification reaction of converting the product ß-amino ester to a p-amino acid.

Saponification reaction processes and conditions are well known in the art and typically involve a caustic catalytic compound. Typical saponification conditions include temperatures ranging from about 25°C to about 100°C at atmospheric pressure.

EXAMPLE 1 To VAPOL ligand (0.11 mmol) and Zr (Oi-C3H7) 4i-C3H70H (0.05 mmol) in toluene (0.9mL) was added N-methylimidazole (0.06 mmol) in toluene (0.1 mL) at room temperature. The mixture was stirred at room temperature for one hour to form the catalytic compound. An imine where R is phenyl (0.25 mmol) as a toluene solution was added to the catalytic compound and the mixture was stirred for an additional 5 minutes. Then the heteroketene acetal reactant (0.3 mmol) (CH3) 2C=C (OSi (CH3) 3) (OCH3) was added as a toluene solution. The resulting mixture was stirred for 15 hours at room temperature. Aqueous NaHC03 was added to quench the reaction. The mixture was extracted twice with 10 mL of dichloromethane. The combined organic layer was concentrated to give the crude product.

The crude product was treated with tetrahydrofuran-1 N hydrochloric acid (10: 1) at 0°C for 30 minutes. Then the reaction was quenched with aqueous NaHCO3 and extracted twice with 10 mL of ethyl acetate. The combined organic layer was washed with brine and concentrated. The desired product was obtained by chromatography using a silica gel column. The optical purity was determined by high pressure liquid chromatography (HPLC) analysis using a chiral column.

EXAMPLE 2 The effectiveness and temperature dependence of the metal-VAPOL ligand catalytic compound of the present invention was investigated and compared to a known asymmetric catalytic compound zirconium-BINOL as described in Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc., and Kobayashi, S.; Ishitani, H.; Ueno, M. J. Am. Chem. Soc., for the aldol reaction: For experiment numbers 1-5, a flame-dried flask was charged with 0.05 mmol Lewis acid and 0.11 mmol ligand in 0.75 mL dichloromethane. For Experiment No. 6, VAPOL ligand and the Lewis acid were stirred in 1 mL of toluene at 25°C for I hour, the solvent was removed in vacuum and the residue warmed at 50°C. 0.05 mm Hg vacuum for 1 hour. N-methylimidazole (0.06 mmol) in 0.25 mL dichloromethane was added at room temperature to all but Experiment 2. The mixture was stirred for 1 hour at room temperature and allowed to remain at room temperature or was cooled to-45°C. Dichloromethane solutions (0.5 mL) of the imine (0.25mmol) and the heteroketene acetal (0.3 mmol) were added and the mixture was continuously stirred. The yield was determined using NMR and percent ee was determined using HPLC with a chiral column. In all experiments, 20 mole percent catalytic compound was used. The results of the experiments are provided in Table 1.

TABLE 1 Experiment No. Ligand Lewis Acid R Temp. Time Yield ee Zr(O(i-C3H7))4phenyl-45°C10h50%80%1R-VAPOL 2 R-BINOL Yb (O (i-C3H7)) 3 phenyl room temp 14 h 60% 0% 3 R-BINOL Yb (O (i-C3H7)) 3 phenyl room temp 14 h 10% 0% Zr(O(i-C3H7))41-naphthyl-45°C15hQuan.80%4S-VAPOL 5 S-VAPOL Zr (O (i-C4H9)) 4 1-naphthyl-45°C 15 h 95% 81% 6 S-VAPOL Zr (O (i-C3H7)) 4 1-naphthyl 45°C 16 h 82% 60% The results demonstrate the effectiveness of the catalytic compound of the present invention at-45°C using dichloromethane as the solvent. Similar experiments were performed to investigate the temperature effect and solvent effect on the zirconium-VAPOL ligand catalytic compound. In Experiment Nos. 1-7 below, a flame dried flask was charged with 0.05 mmol Lewis acid and 0.11 mmol ligand in 0.85 mL toluene. N-methylimidazole (0.06 mmol) in 0.13 mL dichloromethane was added at room temperature. The mixture was stirred for 1 hour at room temperature and allowed to remain at room temperature or was cooled as indicated in Table 2.

Toluene solutions (1 mL) of the imine (0.25 mmol) and the heteroketene acetal (0.3 mmol) were successively added and the mixture was continuously stirred. Solvent was added; for Experiment No. 7 the solvent was 3: 1 toluene: dichloromethane, and for all other experiments the solvent was 15: 1 toluene: dichloromethane. The yield was determined using NMR and percent ee was determined using HPLC with a chiral column. 20 mole percent catalytic compound was used in the experiments.

The results of the experiments are provided in Table 2.

TABLE 2 Experiment No. Ligand Lewis Acid R Temp. Time Yield ee 1S-VAPOLZr (0 (t-C, H9))1-naphthyt-45°C10 h50% 96% Zr(O(t-C4H9))41-naphthyl-45°C20h90%97%2S-VAPOL 3 S-VAPOL Zr(O(t-C4H9))4 1-naphthyl -15°C 5. 5 h 90% 97% Zr(O(t-C4H9))41-naphthyl0°C2h90%96%4S-VAPOL Zr(O(t-C4H9))41-naphthylroomtemp.2h90%94%5S-VAPOL 6 S-VAPOL Zr (O (i-C3H7)) 4 phenyl room temp. 15 h 94% 89% Zr(O(t-C4H9))4phenylroomtemp.15h90%15%7R-BINOL From Table 2 it is clear that the zirconium: VAPOL ligand catalytic compound provided at least 89% ee at room temperature which is drastically higher than the Binol-zirconium catalytic compound at the same temperature.

Similar experiments were conducted to evaluate the dichloromethane solvent as compared to a toluene solvent, and to compare the catalytic compound of the present invention to the metal: 6,6'-dibromoBINOL catalytic compound described in Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc., and Kobayashi, S.; Ishitani, H.; Ueno, M. J. Am. Chem. Soc., on the same aldol reaction described above where R is phenyl. The metal: 6,6'-dibromoBINOL catalytic compound was prepared as described in the cited references. The metal : VAPOL ligand catalytic compounds of the present invention were prepared by charging a flame-dried flask with 0.05 mmol Lewis acid and 0.11 mmol VAPOL ligand in 0.85 mL of toluene. N-methylimidazole (0.06 mmol) in 0.13 mL dichloromethane was added at room temperature. The mixture was stirred for 1 hour at room temperature and allowed to react or cooled to a set temperature.

Solutions (1 mL) of aldimine (0.25mmol) and silyl ketene acetal (0.3 mmol) were sequentially added and the mixture was continuously stirred. In all but Experiment No. 5,20 mole % catalytic compound was used and 24 mole % N-methylimidazole was used. In Experiment No. 5,10 mole % catalyst was used and 12 mole percent N-methylimidazole was used. The yield was determined using NMR and percent ee was determined using HPLC with a chiral column. The results of the experiments are provided in Table 3.

TABLE 3 Experiment No. Ligand Lewis Acid Solvent Temp. Time Yield ee R-BINOL Zr (O (t-C4Hg)) 4 CH2C12 25°C 4 h 100% 28% #Zr(O(t-C4H9))4CH2Cl2-45°C19h80%36%2R-BINOL 3 R-6, 6'-DibromoBINOL Zr (O (t-C4Hg)) 4 CH2CI2 25°C 4 h 87% 48% 4 R-6, 6'-DibromoBINOL Zr (O (t-C4H9)) 4 CH2CI2-450C 19 h 87% 86% 5 R-6, 6'-DibromoBINOL Zr (O (t-C4H9)) 4 Toluene 25°C 18 h 95% 62% 6 S-VAPOL Zr (O (I-C3H7)) 4 Toluene 25°C 15 h 94% 89% Zr(O(t-C3H7))4Toluene-45°C20h92%91%7S-VAPOL The results in Table 3 indicate that the catalytic compound of the present invention is not as temperature dependent as the known dibromo-BINOL catalytic compound. For example, Table 3 shows that the induction for the zirconium-VAPOL ligand catalytic compound in toluene falls only from about 91% ee to about 89% ee when the temperature is raised from about-45°C to 25°C whereas the induction from the reaction of the dibromo-BINOL catalytic compound in dichloromethane falls from about 86% ee to about 48% ee over the same temperature change.

EXAMPLE 3 Catalytic compound efficiency was explored through a series of experiments varying the mole percent of the catalytic compound, the mole percent of the activator, the mole percent of the imine reactant, reaction temperature, and reaction time. The aldol reaction conducted was: The solvent was toluene or a volume ratio of 15: 1 toluene: dichloromethane, and the catalytic compounds were generated from VAPOL ligands and Zr (O (i-C3H7)) 4 (i-C3H7OH) or Zr (O (t-C4Hg)) 4 as outlined in Examples 1 and 2. The activator was N-methylimidazole. For Experiments 1-11, the"R"of the imine reactant above is 1-naphthyl, and for Experiments 12-16 the"R"of the imine reactant above is phenyl. The results are listed in Table 4.

TABLE 4 Catalytic Experiment @mine compound, Activator, Temp. Time No. (M) mol-% mol-% (°C) (hr) Yield, % Ee, % 1 0. 125 20 24 25 2 90 94 2 0. 167 10 12 25 3.5 85 94 56255659230.2 4 0. 125 2 2 25 19 48 77 5 0. 125 0. 5 0. 5 25 26 3 20 0.10.12526--60.125 7 0. 125 2 24 25 19 43 95 8 0. 5 2 24 25 22 64 93 9 1 2 24 43 18 78 >90 10 1 1 24 43 18 70 >90 11 1 0. 5 24 43 18 52 >90 12 0. 125 20 24 25 15 94 89 13 0. 5 2 24 40 6 100 86 14 1 0. 5 24 40 19 60 85 15 1 0. 5 100 40 19 57 83 16 0. 125 20 24-45 20 92 91 The results of Table 4 demonstrate that the zirconium-VAPOL ligand catalytic compound is effective in concentrations as low as 0.5 mole %. The results also demonstrate that the induction does not significantly change when the temperature is raised from about 20°C to about 40°C. At low catalytic compound loading, 2 mole %, the percent ee values are achieved when 24 mole % activator is used.

EXAMPLE 4 Different imine substrates and heteroketene acetals were tested using the catalytic compound and process of the present invention. The imine used is represented by: # where R is as shown in Table 5. The heteroketene actals used are shown in Table 5. A flame dried flask was charged with Zr (O (i-C3H7)) 4 (i-C3H70H) (0.05 mmol) and VAPOL ligand (0.11 mmol) in toluene (1 mL) and N-methylimidazole (0.06 mmol) in dichloromethane (0.13 mL) was added at room temperature and the mixture stirred for one hour. Then imine in toluene (1 mL) and heteroketene acetal (0.3 mmol) were successively added. The catalytic compound was present at 20 mole % of the amount of the limiting substrate, the imine, except for the last two experiments where the catalytic compound was present at 2 mole percent. The N-methylimidazole was present at 24 mole % of the amount of the limiting substrate, the imine. The results of the experiments are presented in Table 5.

TABLE 5 R Heteroketene acetal Mmol Temp. Time Yield ee Imine Phenyl (CH3) 2C=C (OSi (CH3)3)(OCH3) 0.25 room temp. 2 h 93% 89% 1-Naphthyl (CH3) 2C=C (OSi (CH3) 3) (OCH3) 0.25 room temp. 2 h quan. 94% p-CI-Phenyl (CH3) 2C=C (OS@ (CH3) 3) (OCH3) 0.25 room temp. 2 h 95% 79% p-Me-Phenyl (CH3) 2C=C (OSi (CH3) 3) (OCH3) 0. 25 room temp. 2 h 90% 89% 1-Naphthyl CH2=C (OSi (CH3) 3) (SC2H,) 0. 25 room temp. 2 h 85% 29% p-CI-Phenyl CH2=C (OSi (CH3) 3) (SC2H5) 0. 25 room temp. 2 h 93% 40% 1-Naphthyl CH2=C (OSi (CH3) 3) (S (t-C4H9)) 0.25 room temp. 2 h 90% 71% 1-Naphthyl CH2=C (OSi (C4Hg) 3) (OCH3) 0.24 h - - 1-Naphthyl CH2=C (OSi (CH3) 3) (S (t-C4H9)) 025 room temp. 24 h 1-Naphthyl CH2=C(OSi(CH3)3)(S(t-C4H9)) 0.25 room temp. 24 h 1-Naphthyl CH2=C (OSi (CH3) 3) (S (t-C4Hg)) 0.125 room temp. 15 h 90% 71% Phenyl CH2=C (OSi (CH3) 3) (S (t-C4H9)) 0.125 room temp. 15 h 85% 65% 0.125-45°C24hquan.77%1-NaphthylCH2=C(OSi(CH3)3)(S(t-C4H9)) Phenyl CH2=C(OSi(CH3)3)(S(t-C4H9)) 0.125 hquan.71%24 Phenyl CH2=C(OSi(CH3)3)(SC2H5) 0.125 room temp. 14 h 41 % 79% Phenyl CH2=C(OSi(CH3)3)(SC2H5) 0.125 -80°C 16 h 54% 51% 1-Naphthyl (CH3) 2C=C (OSi (CH3) 3) (OCH3) 0. 5 room temp. 22 h 64% 93% Phenyl (CH3) 2C=C (OSi (CH3) 3) (OCH3) 0.5 40°C 6 h quan. 86% From the results in Table 5 the heteroketene acetals without the gem-dimethyl group provided lower percent ee values, and the CH2=C (OSi (CH3) 3) (S (t-C4H9)) gave higher percent ee values than the CH2=C (OSi (CH3)3)(SC2H5), implying that the steric bulky acetal will provide the higher percent ee.

EXAMPLE 5 Different activators were investigated based on the imino aldol reactions of N-benzylidene-2-amino phenol with H2C=C (OSi (CH3) 3) (S (t-C4H9)). The catalytic compound was obtained from VAPOL ligand and Zr(O(i-C3H7))4, and was used in the used at 24 mole %. The solvent was toluene and the reaction temperature was 25°C. The results are presented in Table 6.

TABLE 6 Experiment No. hoursYield,Time, % Ee, % 1 N-Methylimidazole 6480 2 1, 2-Dimethylimidazole 19 100 10 3 Imidazole 19 90 63 4 N-Benzylimidazole 2 85 36 5 N-Methylbenzoimidazole 2 90 12 6 1-Methylbenzotriazole 2 87 5 Different activators were also investigated based on the imino aldol reactions of N-benzylidene-2-amino phenol with (CH3) 2C=C (OSi (CH3) 3) (OCH3). The catalytic compound was obtained from VAPOL ligand and Zr (O (i-C3H7)) 4, and was used in the reaction at 20 mol %. The imine concentration was 0.125 M. The solvent was toluene and the reaction temperature was 25°C. The results are presented in Table 7.

TABLE 7 Experiment No. Activator Time, hours Yield, % ee, % 1 None 3 91 34 2 Pyridine 16 84 52 3 4- (N, N-Dimethylamino) pyridine 20 79 82 393894N-Methylimidazole From the results above, it appears that the effectiveness of the activator may be influence by both the size of the activator and the bacisity of the activator.

N-Methylimidazole and imidazole are the preferred activators resulting in the highest percent enantiomeric excess.

EXAMPLE 6 Different imine compounds were investigated in the following reactions using the catalytic compound of the present invention. The catalytic compound was obtained from VAPOL ligand and Zr (O (i-C3H7)) 4, and was used in the reaction at 20 mole %. The imine concentration was 0.125 M, and the activator, N-methylimidazote was used at 24 mole %. The solvent was toluene and the reaction temperature was 25°C. The results are provided in Table 8.

TABLE 8 Experiment No. @mineTime, hYield, % ee, % 1 R1=R2=R3=R4=H 15 94 89 2 R1=R2=R3=H R4=CH3 2 80 36 3 R1=R3=R4=H R2=CH3 2 97 91 R3=CH31593954R1=R2=R4=H 5 R1=R3= CH3 R2=R4=H 15 100 >98 6 R2=R3=R4=H R1=CH3 16 89 >99 7 R1=R2=R4=H R3=CI 12 78 60 R3=CH3O16----8R1=R2=R4=H The results in Table 8 demonstrate that with the methyl group in the ortho position to the hydroxyl group of the imine, the percent ee improved dramatically and is therefore the preferred configuration.

EXAMPLE 7 Comparison experiments were performed using the catalytic compound of the present invention and the 6,6'-dibromoBINOL-containing catalytic compound as described in Kobayashi, S.; Ishitani, H.; Ueno, M. J. Am. Chem. Soc., 1998,120, 431-432 and Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc., 1997,119, 7153 and different imines and heteroketene acetals to form the phenylaminoester, Structure I below. The 6,6'-dibromoBINOL-containing catalytic compound was prepared from 2 equivalents of 6,6'-dibromoBINOL and 1 equivalent of Zr (O (t-C4H9)) 4, and is represented as"B"in Table 9 below. The catalytic compound of the present invention was generated from 2 equivalents of VAPOL ligand and 1 equivalent of Zr (O (i-C3H7)) 4, and is represented as"A"in Table 9 below. In all experiments, the activator was N-methylimidazole. All but the first two experiments were conducted in toluene; the first two experiments were conducted in dichloromethane. The results of the experiments are presented in Table 9. The structures of the product and reactants used are as follows: TABLE 9 Heteroketene Catalytic Mole % Yield, Imine acetal Compound activator Temp., C Time, h % ee, % l I A 12-45 19 63 64 A242547128IIVIII A1225189347IIVIII B242515100>98IIVIII ix IX B 10 25 19 11 91 11 IX B 10 25 19 16 85 B202519----IIVIII B2060191890IIVIII V Vlil B 12 25 16 22 92 B122516----IIVII IV Vl l B 24 25 16 27 70 IV IX B 24 25 16 47 80 B2425167264VIVIII EXAMPLE 8 The synthesis of the side chain in the Taxol pharmaceutical drug was successfully performed using the catalytic compound and process of the present invention.

Two different imine reactants were used, Experiment 1 used XI and Experiments 2-4 used 11 shown above. The catalytic compounds were generated from VAPOL ligand and Zr(O(i-C3H7))4#(i-C3H7OH) as outlined in Examples 1 and 2. The solvent was toluene and the activator was N-methylimidazole. The results of the experiments are provided in Table 10.

TABLE 10 Activator,ExperimentCatalytic No. compound, mol-% mol-% Temp Time Yield, % ee, % 1 10 12 25°C 15 hr. 89 71 1225°C16hr.8140210 2425°C14hr.7174320 4 20 24-80°C 16 hr.--