Background of the Invention: A variety of methods have been developed for recycling rare and/or valuable reagents and catalysts.

In a conventional methodology, the reagent or catalyst is coupled to an insoluble support prior to its use in a reaction. After the reaction is complete, the reagent or catalyst is isolated and then recycled by isolating the insoluble support. The recycled reagent or catalyst attached to its insoluble support may then be re- employed in further reactions. Examples of this methodology have been developed for a number of reagents and catalysts. (Pittman et al. Comprehensive Organometallic Chemistry; Wilkinson, G., Ed.; Pergamon Press: Oxford, 1982; Mathur et al. Polymers as Aids in Organic Chemistry. Academic Press: New York, 1980; Frechet et al. Tetrahedron 1981, 37, 663; Sherrington et al. Syntheses and separations Using Functionalized Polymers; Wiley: New York, 1988; Bergbreiter et al. J.

Polym. Sci., Polym. Chem. Ed. 1989, 27, 4205).

Although such insoluble reagents and catalysts are

successfully employed in many applications, there are recognized limitations associated with their use.

(Barany et al. In The Peptides, Vol. 2; Gross, E.; Meienhofer, J., Eds.; Academic Press: New York, 1979, p.

1). The insolubility of the reagents and catalysts can be disadvantageous before, during, and after the reaction. Particular disadvantages include the following: (1) Formation of the reagent or catalyst on the support can be more tedious.

(2) Reagent or catalyst characterization is less than routine.

(3) Insolubility can limit the range of substrates and/or the polymer-bound reagent/catalysts utility.

(4) Insolubility can cause the reagent/catalyst reactivity to behave somewhat differently than its soluble (solution) counterpart.

(5) Non-linear kinetics have frequently been observed with insoluble supports.

Because of the above disadvantages, alternative strategies have been developed for recycling valuable reagents and catalysts. In one successful strategy, the reagents or catalysts are supported by soluble homopolymers. (Bayer et al. Angew.Chem. Int. Ed. 1975, 14, 493; Bayer et al. CHEMTECH 1976, 6, 212; Bergbreiter et al. ACS Symposium Series 1986, 308, 17; Bergbreiter et al. J. Am. Chem. Soc. 1987, 109, 174; Bergbreiter et al. CHEMTECH 1987, 17, 686; Phelps et al. Tetrahedron Lett. 1989, 30, 3915; Bergbreiter et al.

J. Org. Chem. 1989, 54, 2726; Bergbreiter et al.

Tetrahedron Lett. 1991, 32, 2731; Doyle et al. J. Org.

Chem. 1992, 57, 6103; Bergbreiter et al. J. Am. Chem.

Soc. 1993, 115, 9295). In this strategy, the reactions are carried out homogeneously, i.e., without an insoluble phase, and the separation of the homopolymer from reaction products is achieved by taking advantage of the properties of the polymer chain. In the arena of combinatorial synthesis, this strategy is termed "Liquid Phase Combinatorial Synthesis" or LPCS. (Han et al.

Proc. Natl. Acad. Sci. USA 1995, 92, 6419.) The key element of LPCS is a linear homopolymer, e.g., polyethylene glycol monomethyl ether (MeO-PEG), which serves a dual role as both a terminal protecting group and a solubilizing agent for any compound(s) synthesized on the support. Combinatorial peptide, small molecule, and peptidomimetic libraries can be synthesized using this approach.(Han et al. J. Am. Chem. Soc. 1996, 118, 2539.) The ligand-accelerated catalytic (LAC) asymmetric dihydroxylation (AD) of olefins based on cinchona alkaloid ligands was described by Sharpless in 1988.

(Jacobsen et al. J. Am. Chem. Soc. 1988, 110, 1968.) Since this seminal report, the AD reaction has been further developed to include application to a wider range of olefins, improved enantiomeric efficiency, and overall simplicity of operation (Kolb et al. Chem. Rev.

1994, 116, 2483). From the standpoint of cost, ligand and/or metal recovery and recycling are of prime interest because the cinchona alkaloid ligand and osmium tetroxide are the most expensive components of the procedure. Olefines have also been asymmetrically dihydoxylated using insoluble polymer bound cinchona alkaloid-ligands. (Kim et al. Tetrahedron Lett.. 1990, 31, 3003; Pini et al. Tetrahedron Lett. 1991, 32, 5175; Lohray et al. Tetrahedron Lett. 1992, 33, 5453; Pini et al. Tetrahedron: Asymmetry 1993, 4, 2351; Lohray et al.

Tetrahedron Lett. 1994, 35, 6559; Pini et al.

Tetrahedron 1994, 50, 11321; Pini et al. Tetrahedron Lett. 1995, 36, 1549; Sung et al. Tetrahedron: Asymmetry 1995, 6, 2687). While this methodology represents a major improvement with respect to convenience and economics, it is deemed less than satisfactory because of increased reaction times, highly variable yields and lower enantioselectivity than had previously been obtained with its solution phase counterpart. A report by Salvadori and co-workers describes an insoluble support that provides improved enantioselectivity.

However, their system still required long reaction times (24 hours) and excess of polymeric ligand. Furthermore, reaction yields were lower than observed under homogeneous reaction conditions. (Petri et al., Chirality 1995, 7, 580.) The problems associated with LAC in which the ligand is localized by attachment to an insoluble polymer can be understood by considering the basic tenet upon which this concept is based (Berrisford et al.

Angew. Chem. Int. Ed. Engl. 1995, 34, 1059). During LAC, the addition of a ligand increases the reaction rate of an already existing catalytic transformation.

Both the ligand-accelerated and the nonaccelerated reactions operate in solution simultaneously and in competition with each other. If the ligand does not have equivalent access to all the reaction compartments where the substrate, metal oxidant and olefin reside, the most fundamental requirement for a successful ligand accelerated catalysis scenario is not met. For the present case, this means that the chiral ligand resides only in the insoluble phase while the OsO4 and olefin are in solution and free to react anywhere. In this situation the optimal LAC conditions can probably never be achieved even when using a large excess of the insoluble polymer-bound ligand. (Petri et al. Chirality 1995, 7, 580.)

Since its discovery in 1988, Sharpless' catalytic asymmetric dihydroxylation (AD) of olefins has been continuously refined by both a development of better ligands and improvements in the secondary oxidant/solvent system. (For review, see: Kolb et al.

Chem. Rev. 1994, 116, 2483; Becker et al. Angew. Chem.

Int. Ed. Engl. 1996, 35, 448). In parallel, Sharpless (Kim et al. Tetrahedron Lett. 1990, 31, 3003) and other groups (Pini et al. Tetrahedron Lett. 1991, 32, 5175; Lohray et al. Tetrahedron Lett. 1992, 33, 5453.; Pini et al. Tetrahedron: Asymmetry 1993, 4, 2351; Lohray et al.

Tetrahedron Lett. 1994, 35, 6559; Pini et al.

Tetrahedron 1994, 50, 11321; Pini et al. Tetrahedron Lett. 1995, 36, 1549; Sung et al. Tetrahedron: Asymmetry 1995, 6, 2687; Petri et al. Chirality 1995, 7, 580; Song et al. Tetrahedron: Asymmetry 1996, 7, 645) have investigated immobilization of these expensive ligands onto insoluble polymer supports so as to aid in their recovery. While this strategy provides a simple but elegant way for automating the AD reaction, it has a number of limitations including prolonged reaction times and, more importantly, a reduction in enantioselectivity What is needed is a soluble polymer-bound ligand which provides all the advantages that an insoluble support can offer including ease of product separation and polymer-bound ligand recover/ reusability while also being as effective as a free ligand both in reactivity and selectivity. Furthermore, the new soluble polymer bound-ligand system should be applicable to other classes of AD ligands for improved enantioselectivity as well as other enantioselective catalytic processes.

Brief Summary of the Invention: The invention relates to polyethylene glycol monomethyl ether-bound alkaloid ligands (termed liquid phase ligands) which are constructed and employed in the

ligand-accelerated catalytic (LAC) Sharp less asymmetric dihydroxylation reaction (AD) with a range of olefins.

These soluble polymer-bound ligands are much more efficient than the corresponding insoluble polymer-bound ligands both in reactivity and selectivity in AD reactions. In addition, these chiral homopolymers provide the same enantioselectivity and reactivity as free ligands in solution. Furthermore, the polymer serves as a scaffold allowing simple product separation, polymer-bound ligand recovery, and recycling of the chiral liquid phase support without a loss of catalytic activity.

One aspect of the invention is directed to a PEG bound alkaloid ligand. This PEG bound alkaloid ligand includes a polyethylene glycol which is soluble in aqueous medium and precipitable in aqueous/ether medium, and an alkaloid ligand coupled to the polyethylene glycol. Preferred alkaloid ligands are selected from the following group, viz.: 1,4-bis-(9'-O-dihydroquinidyl)-phthalazine; 1,4-bis-(9'-O-quinidyl)-phthalazine; 3,6-bis-(9'-O-dihydroquinidyl)-pyridazine; 3,6-bis-(9'-O-quinidyl)-pyridazine; 1,4-bis- (9'-O-dihydroquinyl)-phthalazine; 1,4-bis-(9'-O-quinyl)-phthalazine; 3, 6-bis- (9' -0-dihydroquinyl) -pyridazine; 3,6-bis-(9'-0-quinyl)-pyridazine; dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinidine; 4-methoxybenzoyl dihydroquinidine; 4-chlorobenzoyl dihydroquinidine; 2-chlorobenzoyl dihydroquinidine; 4-nitrobenzoyl dihydroquinidine; 3-chlorobenzoyl dihydroquinidine; 2-methoxybenzoyl dihydroquinidine; 3 -methoxybenzoyl dihydroquinidine;

2-naphthoyl-dihydroquinidine; cyclohexanoyl dihydroquinidine; p-phenylbenzoyl dihydroquinidine; dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinine; 4-methoxybenzoyl dihydroquinine; 4-chlorobenzoyl dihydroquinine; 2-chlorobenzoyl dihydroquinine; 4 -nitrobenzoyl dihydroquinine; 3 -chlorobenzoyl dihydroquinine; 2 -methoxybenzoyl dihydroquinine; 3-methoxybenzoyl dihydroquinine; 2-naphthoyl dihydroquinine; cyclohexanoyl dihydroquinine; p-phenylbenzoyl dihydroiquinone; acrylonitrile co-polymer of 9-(4- chlorobenoyloxy)-quinidine; acrylonitrile co-polymer of 11-((2- acryloyloxy)ethyl-sulfinyl)-9-(4- chlorobenoyloxy) -10, 11-dihydroquinidine; acrylonitrile co-polymer of 11-[2- acryloyloxy)ethylsulfonyl]-9-(N,N- dimethylcarbamoyl) -10, 11-dihydroquinidine; acrylonitrile co-polymer of 9-(l0-undecanoyl)- 10,11-dihydroquinidine; and alkaloid ligands represented by the following structures: In a preferred embodiment, the above alkaloid ligands are coupled to said polyethylene glycol by a linkage selected from the group consisting of ester linkage, amide linkage, thoester linkage, ester linkage, thiether linkage, and sulphone linkage. Preferred examples of the PEG bound alkaloid ligand are represented by the following structures:

Another aspect of the invetnion is directed to an improved process for catalyzing a hydroxylation reaction. The process is of the type which includes a step for admixing, within a reaction medium, an olefin, an oxidizing agent, a catalyst for catalyzing the hydroxylation reaction, and an alkaloid ligand for accelerating the catalysis of the hydroxylation reaction. The process is improved by the use of a PEG bound alkaloid ligand. The PEG bound alkaloid ligand is soluble in the reaction medium and precipitable in a precipitation medium. In a preferred embodiment, the hydroxylation reactions are asymmetric dihydroxylation reaction and the PEG bound alkaloid ligand is chiral. A preferred oxidizing agent is a mixture which includes @@@@(CN). potassium ferricvanide III). CHSON

(methanesulfonamide), OSO4 and K2CO3 (potassium carbonate). Another preferred oxidizing agent is a mixture which includes methylmorpholine N-oxide (NMO), tetraethylammonium acetate tetrahydrate, and Iso4. The preferred catalyst is Iso4. Preferred reaction media are acetone/water or butanol/water. Preferred precipitation media include the addition of diethyl ether or cold ethanol to the reaction media.

Another aspect of the invention is directed to a process for catalyzing an hydroxylation reaction. In the first step of this process, an olefin, an oxidizing agent, a catalyst for catalyzing the hydroxylation reaction, and an alkaloid ligand for accelerating the catalysis of the hydroxylation reaction are admixed in a reaction medium. The admixture occurs in a reaction medium under reaction conditions for producing an hydroxylation product. In the second step, a precipitation medium is admixed with the reaction medium of the first step for precipitating the PEG bound alkaloid ligand. In the third step, the precipitated PEG bound alkaloid ligand of the second step is separated from the hydroxylation product so as to recover the PEG bound alkaloid ligand. In a preferred mode, the hydroxylation reaction in the first step is an asymmetric dihydroxylation reaction and the PEG bound alkaloid ligand is chiral.

Another aspect of the invention is directed to a further process for catalyzing an hydroxylation reaction. In the first step of this process, one or more reactants are admixed with a catalyst for catalyzing the hydroxylation reaction and with a PEG bound alkaloid ligand for accelerating the catalysis of the hydroxylation reaction. The admixture occurs occurs in a reaction medium under reaction conditions for producing an hydroxylation product, In the second step,

a precipitation medium is admixed the reaction medium of of the first step for precipitating the PEG bound alkaloid ligand. In the third step, the hydroxylation product of the second step is separated B from the precipitated PEG bound alkaloid ligand for obtaing purified product. In a preferred mode, the hydroxylation reaction is an asymmetric dihydroxylation reaction and the PEG bound alkaloid ligand is chiral.

Brief Description of Figures: Figure 1 illustrates a prior art alkaloid ligand, i.e. compound 1, a PEG radical, i.e., compound 4, alkaloid ligand coupled to a linkage unit, i.e. compound 3, and a PEG bound alkaloid ligand, i.e., compound 2.

Figure 2 illustrates the synthesis of ligands 2 and 3 with the following conditions: (a) TEA, DMAP, glutaric anhydride (60 %); (b) DCC, DMAP, ROH (95 %).

Figure 3 illustrates the comparison of catalytic asymmetric hydroxylations using compound 1-4. The indicated notations are as follows: (a) see Petri et al.

Chirality 1995, 7, 580 for experimental details; (b) results from Petri et al. Chirality 1995, 7, 580; ligand 2, which was recovered from entry 3, was recycled a total of four times; (d) for the purpose of comparison, the reaction was stopped after 5 hours (e) slow addition time of olefin.

Figure 4 illustrates polymer-bound trans-cinnamate esters 7-11 for the Sharpless AD reaction.

Figure 5 illustrates the reaction in which trans- cinnamic acid was immobilized to the four polymeric supports 7-11 (figure 4) and the reactivity of 7-11 was

demonstrated by using the ligands 13 and 14 for the Sharpless AD reaction.

Figure 6 tabulates various AD reactions conducted upon various polymer bound trans-cinnamate esters using the ligands 13 and 14. The indicated notations are as follows: [a] OSO4 equivalents realtive to olefin.

[ligand]/[OsO4] = 2.5. [b] Method A: K3Fe(CN)6 and t- butanol/water = 1/1 system at room temperature. Method B: N-methylmorpholine-N-oxide and acetone/water = 10/1 system at 40C. [c] Determined from the ratio of diol proton signals to the remaining olefin signals by NMR.

[d] Determined through NMR analysis of the derived bis- Mosher ester of the diol.

Figure 7 illustrates the chemical synthesis of ligands 20 and 14.

Figure 8 tabulates various catalytic Asymmetric Dihydroxylation Reactions using Ligand 14. The indicated notations are as follows: (a) For NMO system, the molar ratio of olefin/OsO4/ligand = 1/0.04/0.1, and for K3Fe(CN)6 system, the molar ratio was 1/0.005/0.1.

(b) Number in parenthesis represents results for a free ligand (DHQD)2PHAL from Sharpless et al. J. Org. Chem.

1992, 57, 2768.

Detailed Description of the Invention The invention is directed to the synthesis and use of polyethylene glycol monomethyl ether-bound cinchona alkaloid ligands in the ligand-accelerated catalytic (LAC) Sharpless asymmetric dihydroxylation reaction (AD) with a range of olefins.

Example 1. Synthesis and Exemplary use of Polyethylene glycol monomethyl ether-bound cinchona alkaloid ligands in the AD reaction The following examples are exemplary conditions which demonstrate the versatility of the methodology and are not meant to be restrictive with the models disclosed. Rather, the methodology can be used with symmetrical, nonsymmetrical, substituted and unsubstituted - primary, secondary or tertiary olefins.

In addition, other commercially available alkaloid ligands can be used in lieu of the disclosed dihydroquinidine hydrochloride 5 (Figure 2). Preferred alkaloid ligands include the following: 1. from U.S. Patent No. 5,260,461, incorporated herein by reference: 1,4-bis-(9'-O-dihydroquinidyl)-phthalazine; 1,4-bis-(9'-O-quinidyl)-phthalazine; 3, 6-bis- (9' -0-dihydroquinidyl) -pyridazine; 3,6-bis-(9'-O-quinidyl)-pyridazine; 1,4-bis- (9'-O-dihydroquinyl)-phthalazine; <BR> <BR> 1, 4-bis- (9' -O-quinyl) -phthalazine; <BR> <BR> <BR> 3, 6-bis- (9 '-0-dihydroquinyl) -pyridazine; <BR> <BR> <BR> 3, 6-bis- (9 '-0-quinyl) -pyridazine; 2. from U.S. Patent No. 4,871,855, incorporated herein by reference: dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinidine; 4-methoxybenzoyl dihydroquinidine;

4-chlorobenzoyl dihydroquinidine; 2-chlorobenzoyl dihydroquinidine; 4-nitrobenzoyl dihydroquinidine; 3 -chlorobenzoyl dihydroquinidine; 2-methoxybenzoyl dihydroquinidine; 3-methoxybenzoyl dihydroquinidine; 2-naphthoyl-dihydroquinidine; cyclohexanoyl dihydroquinidine; p-phenylbenzoyl dihydroquinidine; dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinine; 4-methoxybenzoyl dihydroquinine; 4-chlorobenzoyl dihydroquinine; 2-chlorobenzoyl dihydroquinine; 4-nitrobenzoyl dihydroquinine; 3 -chlorobenzoyl dihydroquinine; 2-methoxybenzoyl dihydroquinine; 3 -methoxybenzoyl dihydroquinine; 2-naphthoyl dihydroquinine; cyclohexanoyl dihydroquinine; p-phenylbenzoyl dihydroiquinone; 3. From U.S. Patent No. 5,126,494, incorporated herein by reference: acrylonitrile co-polymer of 9-(4- chlorobenoyloxy)-quinidine; acrylonitrile co-polymer of 11-((2- acryloyloxy)ethyl-sulfinyl)-9-(4- chlorobenoyloxy) -10, 11-dihydroquinidine; acrylonitrile co-polymer of 11-[2- acryloyloxy)ethylsulfonyl]-9-(N,N- dimethylcarbamoyl)-lo,11-dihydroquinidine; acrylonitrile co-polymer of 9-(10-undecanoyl)- 10, 11-dihydroquinidine; and alkaloid ligands represented by the following structures: Reaction conditions which cover temperature, substrate equivalents, reaction time(s), work-up(s) and buffer solutions (pH levels) may vary, depending on the substrate used, however all proportions are Annroximatels the same. A representative Procedure is

provided in the synthetic protocals.

In an effort to circumvent the problems observed with insoluble supports and LAC, yet, provide the economical and physical advantages (product isolation and reagent recovery) that a polymeric support can offer, we have developed the soluble homopolymer MeO-PEG as a suitable scaffold for the AD reaction. The synthesis of polyethylene glycol monomethyl ether bound cinchona alkaloid ligands (Figure 1) and their successful use in the LAC asymmetric dihydoxylation reaction of various olefins is described as follows.

The synthesis of the MeO-PEG-bound dihydroquinidine ligands is depicted in Figure 2. The commercially available hydroquinidine 5 (Aldrich) was acylated using glutaric anhydride and 4-N,N'-dimethylaminopyridine (DMAP) to provide carboxylic acid 6. This reaction, though simple, provides the linking unit necessary for attachment to the homopolymer MeO-PEG or any other amino or alcohol group. The coupling of acid 6 to polyethylene glycol monomethyl ether and ethyl alcohol in the presence of dicyclohexylcarbodiimide and DMAP produced the homopolymer 2 and its simple diester homologue 3 respectively.

The chiral homopolymer 2 was the archetype used to examine and compare all of the AD reactions investigated. The structural similaritiy of 2 to the insoluble acrylonitrile ligand 1, allowed for a direct comparison between soluble and insoluble supports to be made (Kim et al. Tetrahedron Lett. 1990, 31, 3003) while contrasting the reactivity of ligands 2, 3, and 4 in the AD reaction would delineate any effect that the polyethylene glycol backbone may have on asymmetric induction. In addition, to standardize the comparisons between our soluble ligand support and the insoluble

ligand support, we used the same conditions as reported for the ligand 1- AD catalytic reaction (Kim et. al.

Tetrahedron Lett. 1990, 31, 3003).

The MeO-PEG-supported catalyst 2 is completely soluble in an acetone-water mixture (v/v = 9/1); thus the catalytic reaction is completely homogeneous. Of greater note is that the reaction is complete within the same time frame as that of its solution counterpart with no decrease of yields or enantioselectivity as tablutated in Figure 3. For this methodology to be useful, product isolation, separation, and recovery of the polymer bound ligand must be straightforward and reliable. Upon completion of the AD reaction, the entire mixture was diluted with methylene chloride, dried (anhydrous sodium sulfate) and filtered. Diethyl ether was added to the resulting mixture in order to precipitate MeO-PEG-bound ligand (typically, the MeO- PEG-bound ligand was recovered in >98 % yield). The filtrate contained the dihydroxylated product.

The asymmetric dihydroxylation of a variety olefins by ligands 1-4 is shown in Figure 3. Immediately evident is the fact that MeO-PEG-bound ligand 2 is more efficient than the insoluble polymer bound ligand 1 (Kim et. al. Tetrahedron Lett. 1990, 31, 3003), both in its enantioselectivity and reactivity (entries 1 and 2 Figure 3). What is more, the polymer-bound ligand 2 is easily recovered in near quantitative fashion and recycled several times with no loss of reaction yield or enantioselectivity (entry 3, Figure 3 ). With all four olefins tested, MeO-PEG-bound ligand 2 was as effective as free ligand 3 (compare entries 2 and 4, entries 6 and 7, entries 8 and 9, and entries 10 and 11). These findings strongly suggest that the MeO-PEG back-bone does not influence or affect the observed asymmetric induction (entry 5). Furthermore, these findings

provide direct support for our notion that for successful polymer-bound LAC all components involved in the reaction must be able to interact freely with each other in solution.

In summary, a process has been invented which demonstrates how a chiral ligand can be integrated into a soluble polymeric species so that LAC can operate in an unhindered manner on a polymer support. The soluble polymer-bound ligand provides all the advantages that an insoluble support can offer, while also being as effective as a free ligand both in reactivity and selectivity. This new soluble polymer bound-ligand system should be applicable to other classes of AD ligands for improved enantioselectivity (Kolb et al.

Chem. Rev. 1994, 116, 2483) as well as other enantioselective catalytic processes (Takaya et al. J.

Am. Chem. Soc. 1987, 109, 1596; Zhang et al. J. Am.

Chem. Soc. 1990, 112, 2801; Lowenthal et al. Tetrahedron Lett. 1991, 32, 7373; Li et al. J. Am. Chem. Soc. 1993, 115, 5326; Evans et al. J. Am. Chem. Soc. 1993, 115, 5328; Johnson et al. in Catalytic Asymmetric Synthesis Ed.; VCH: Weinheim, 1993; pp. 103-158; Jacobsen et al.

in Catalytic Asymmetric Synthesis; Ojima, I. Ed.; VCH: Weinheim, 1993; pp. 159-202; Li et al. Angew. Chem., Int. Ed. Engl. 1996, 35, 451).

The MeO-PEG polymer will not only be useful to the research chemist but also for effecting the separation of catalyst from product in homogeneous industrial applications (Steckhan et al. Angew. Chem. Int. Ed.

Engl. 1990, 29, 388; Herrmann et al. Angew. Chem. Int.

Ed. Engl. 1993, 32, 1524; Bergbreiter et al. J. Am.

Chem. Soc. 1993, 115, 9295; Horvath et al. Science 1994, 266, 72). Finally because of its desirable physical properties, this and other ligand accelerated catalysts that are incorporated within liquid phase supports may

find use in automated high through-put synthetic efforts (Frisbee et al. J. Am. Chem. Soc. 1984, 106, 7143; Sabadosa et al. Lab. Rob. Automation 1989, 1, 265; Dewitt et al. Current Opin. Biotech. 1995, 6, 640).

Example 2. Synthesis and Exemplary use of attachment of small organic olefinic moiety to various polymeric matrixs with the Sharpless AD reaction Since the initial discovery of Merrifield (Merrifield et al. J. Am. Chem. Soc. 1963. 85. 2149- 2154) concerning the synthesis of oligopeptides the use of polymers as supports, reagents or even catalysts, for various syntheses mostly related to fine chemistry has increased (Crowley et al. Acc. Chem. Res. 1976, 9, 135- 144; Leznoff et al. Acc. Chem. Res. 1978, 11, 327-333; Manecke et al. Angew. Chem. Int. Ed. Engl. 1978, 17, 657-670; Frechet Tetrahedron 1981, 37, 663-688).

Recently, a high level of activity has been devoted to the field due to the application of combinatorics to the drug discovery arena (Molecular Diversity and Combinatorial Chemistry (eds.: I.M. Chaiken and K.D.

Janda) American Chemical Society, Washington, D.C., 1996; Fruchtel et al. Angew. Chem. Int. ed. Engl. 1996, 35, 17-42; Balkenhohl et al. Angew. Chem. Int. ed.

Engl. 1996, 35, 2288-2337). Although a resurgence of interest in polymers for organic synthesis has been seen, little effort has been paid to the structural parameters of the supports. This being most evident in the realm of automation of organic synthesis.

In an effort to delineate the shortcomings of insoluble polymer bound ligand-accelerated catalysis (LAC), we show how a chiral ligand can be integrated onto a soluble polymeric species so that LAC can operate in an unhindered manner on a polymer support (Han et al.

J. Am. Chem. Soc. 1996, 118, 7632-7633). However, the

converse wherein a small organic moiety is attached to a polymeric matrix and the polymer's influence on the LAC reaction has yet to be investigated. Implications from the successful combination of these two scenario's being that a multipolymer LAC reaction maybe a feasiable target. In this context, we have collected data using both solid and liquid phase polymers to evaluate the influence of the support structure on the Sharpless asymmetric dihydroxylation (AD) reaction. The outgrowth of these findings have allowed us to successfully run a multipolymer Sharpless AD reaction.

A variety of supports have been recently engaged in organic synthesis (Terrett et al. Tetrahedron 1995, 51, 8135-8173; Thompson et al. Chem. Rev. 1996, 555-600; Hermkens et al. Tetrahedron 1996, 52, 4527-4554).

Typically, they may be grouped into three categories.

(1) Minimal cross-linked supports that form well solvated gels (R.B. Merrifield, Angew. Chem. Int. Ed.

Engl. 1985, 24, 799-810); (2) Porous but rigid supports with a high degree of cross linking (Bartholin et al.

Prog. Polym. Sci. 1982, 8, 277-332); (3) Linear soluble polymers also known as liquid phase supports (Geckeler et al. in Advances in Polymer Science, Vol. 121 (Eds.

Spring-Verlag, Berlin, 1995, p. 31; Bayer et al.

Angew. Chem. Int. Ed. Engl. 1991, 30, 113-129; Bergbreiter in Proceedings of the Sixth Annual IUCCP Symposium on Functional Polymers. (Eds. D.E. Bergbreiter and C.R. Martin) Plenum Press, New York, 1989, p. 143.

Two low divinyl benzene-cross-linked polystyrene beads (Merrifield and Wang resin) and the graft copolymer polystyrene-polyethylene glycol (Tentagel) were utilized as category-l supports. The choice of these resins was based on their fundamental differences. Merrifield and Wang resins are relatively hydrophobic matrices while Tentagel is considered to be a hydrophilic tentacle polymer with more "solution-like" characteristics. The

category-3 support used was polyethylene glycol (PEG).

Besides it fulfilling the criteria of a group-3 polymer, this homopolymer was enlisted as it contrasted the chemical reactivity differences between PEG and Tentagel.

Trans-cinnamic acid was immobilized to the four polymeric supports (Figure 4). Using this format insured that all supports would have an olefin tethered as an ester with the composition of the polymer determining the overall reactivity. Furthermore, because the reactivity of ethyl trans-cinnamate (11) is a documented substrate for the AD reaction using the ligand 1,4-phthalazinediyldiether hydroquinidine [(DHQD)2PHAL, 14, Figure 5] Sharpless et al. Org. Chem.

1992, 57, 2768-2771, a direct correlation could be made between solution and polymeric reactions.

Figure 6 shows that the resins structural make-up is the dominant factor influencing the AD reaction.

Overall the amount of metal/ligand used, reaction time required for product conversion, and enantiomeric excess observed varies greatly depending on the support.

Contrasting all four polymeric supports in terms of the AD reaction clearly shows that the non-cross-linked liquid phase resin 10 (entries 6 and 10; Figure 6) provides the best overall results when considering reaction time, conversion percent, enantiomeric excess and metal/ligand ratio. Furthermore, this soluble polymer substrate compares favorably with entry 11 which details our findings for the matrix removed olefin counterpart 11.

Specifically looking at the three insoluble supports Tentagel-9 out performed the Wang-8 and Merrifield-7 resins in the optimum enantioselective oxident, potassium ferricyanide (entries 1-5). However,

the cross-linked polystyrene suppports-8/9, can provide a suitable environment for LAC when NMO is applied as the oxident and acetone/water is the solvent (entries 7 and 8). It may have been anticipated that supports 7 and 8 would show an overall inability to support the AD reaction in the potassium ferricyanide oxident system as these polymers have rather modest swelling properties in the solvent system employed. The consequences of poor swelling would include limited chain mobility which could compromise the location and distribution of catalyst to olefin within these matrices.

Somewhat unexpected was the large amount of metal and ligand that was required for the Tentagel resin-9 to provide a respectable reaction time, conversion and enantioselec-tivity (compare entries 3,4, and 5).

Interestingly there appears to be a threshold of catalyst that is required with Tentagel for any conversion to diol be seen (compare entries 3 and 4).

We investigated this aspect further through the following two experiments. In one reaction conditions as detailed in entry 4 were utilized. After 24h fresh unbound olefin 11 (0.125 mmol.) was added to the reaction mixture as was OSO4 (0.02 eq.relative to the olefin), K2CO3 (3.0 eq relative to the olefin), and K3Fe(CN)6 (3.0 eq relative to the olefin). The reaction mixture was analyzed after 10h and was found to have only obtained 10% conversion of 11 to 12b. In the second reaction unbound olefin 11 was used throughout the same sequence of events, however, in this scenario recharging the reaction mixture with 11 provided complete conversion to 12b in 2.5h. Such results suggest that free ligand becomes entangled in the insoluble support limiting its availability. Thus, while resin-9 is a more hydrophilic copolymer than resins-7 and 8 it still must be considered a "quasi-homgeneous" matrix which is subject to some of the same limitations found

within insoluble resins.

In our defining of the influence of the structure and support texture on the kinetics and enantioselectivty of the AD reaction (vide supra) we identified that a correct combination of soluble and insoluble polymers (i.e. a multipolymeric reaction) could in principle make a tenable AD process. The use of more than a single polymeric reagent/catalyst in reactions either simultaneously or consecutively has been accomplished, but its exploitation has been limited to date (Examples in the area of peptide syntheses include: a) G. Heusel, G. Bovermann, W. Goring, G.

Jung, Angew. Chem. Int. Ed. Engl. 1977, 16, 642-643. b) H. Frank, H. Hagenmaier, Experentia 1975, 31, 131-133.

For examples other than peptides see: a) B.J. Cohen, M.A. Kraus, A. Patchornik, J. Am. Chem. Soc. 1977, 99, 4165-4167. b) C.V. Pittman, L.R. Smith, J. Am. Chem.

Soc. 1975, 97, 1749-1754 c) J.P. Collman, K.M. Kosydar, M. Bressan, W. Lamanna, T. Garrett, J. Am. Chem. Soc.

1984, 106, 2569-2579. d) D.E. Bergbreiter, R. chandran, J. Am. Chem. Soc. 1987, 109, 174-179. e) D. E.

Bergbreiter, R. Chandran, J. Am. Chem. Soc 1985, 107, 4792-4793. f) J. J. Parlow, Tet. Lett. 1995, 36, 1395- 1396. g) F. Svec, J. M. J. Frechet, Science 1996, 273, 205-211. For the detection of highly reactive intermediates using two polymeric reagents simultaneous "Three phase text", see J. Rebek Tetrahedron, 1979, 35, 723-731 and references cited therein). The combination of Tentagel resin-9 and MeO-PEG-[(DHQD)2PHAL] 14, (Figure 5) was investigated as the multipolymer components.

Entry 12 shows that this combination of polymers can successfully support LAC. These two polymers are physically quite different, yet, the outgrowth of this methodology translates into facile product separation and ligand re-isolation. The significance of such technology should have direct application to the

automation of synthetic processes (Frisbee et al. J. Am.

Chem. Soc. 1984, 106, 7143-7145. b) Dewitt et al. Curr.

Biol., 1995, 6, 640-645. c) Kramer et al. Chemtech 1989, 19, 682-688. d) Hobbs Dewitt et al. Acc. Chem.

Res. 1996, 29, 114-122).

From the literature (Terrett et al. Tetrahedron 1995, 51, 8135-8173; Thompson et al. Chem. Rev. 1996, 555-600; Hermkens et al. Tetrahedron 1996, 52, 4527- 4554) as well as from the results reported here a few guidelines seem to be emerging concerning the influence of the structure and support texture on the kinetics and selectivity of polymer supported organic reactions.

These lines appear valid for both supported catalysts as well as for organic moieties attached to these matrices.

The key in the context of both systems can be assimilated back to properties inherent to all polymers: Accessibility or lack of, and microenvironment, the latter often being a prerequisite for the first.

In the field of small molecule combinatorics example 2 demonstrates that multipolymeric reactions should be applicable in the automation of organic synthesis.

Example 3: PEG Approach to the Sharpless Catalytic Asymmetric Dihydroxylation (AD) using a [(DHOD)2PHAL-PEG-OMe] Ligand.

In an effort to circumvent the problems associated with heterogeneous reactions, we illustrate in example 1 (vida supra) a homogeneous extention to the AD reaction using a soluble polymer, poly(ethylene glycol) mono- methyl ether (MeO-PEG), bound cinchona alkaloid ligand (Han et al. J. Am. Chem. Soc. 1996, 118, 7633). This liquid-phase methodology (Han et al. J. Am. Chem. Soc.

1996, 118, 2539; Han et al. Methods in Enzymology 1996, 267, 234; Han et al. Proc. Natl. Acad. Sci. USA 1995,

92, 6419) provided all of the advantages that an insoluble polymer has to offer, while also being as effective as the free ligand in terms of both reactivity and enantioselectivity. However, the ligand 5 (Figure 2) utilized was not the most effective in terms of enantioselectivity and reactivity for the solution-phase AD reaction. Therefore in an attempt to improve our liquid-phase approach to the AD reaction we illustrate in this example the synthesis of a (DHQD)2PHAL ligand bound to MeO-PEG-NH2 and its successful use in the AD reaction of various olefins (Figures 7 and 8).

Furthermore, we extend the methodology to include additional ligands which, when bound to the PEG polymer, afford successful use in the AD reaction. The additional ligands include the following alkaloid ligands: 1. from U.S. Patent No. 5,260,461, incorporated herein by reference: <BR> <BR> <BR> 1,4-bis-(9'-O-dihydroquinidyl)-phthalazine; <BR> <BR> <BR> <BR> <BR> <BR> <BR> <BR> 1,4-bis-(9t-O-quinidyl)-phthalazine; <BR> <BR> <BR> <BR> <BR> <BR> <BR> 3, 6-bis- (9 '-0-dihydroquinidyl) -pyridazine; 3,6-bis-(9'-O-quinidyl)-pyridazine; 1,4-bis- (9'-O-dihydroquinyl)-phthalazine; 1,4-bis-(9'-O-quinyl)-phthalazine; <BR> <BR> <BR> <BR> 3, 6-bis- (9 '-0-dihydroquinyl) -pyridazine; <BR> <BR> <BR> <BR> <BR> <BR> <BR> <BR> 3, 6-bis- (9 '-0-quinyl) -pyridazine; 2. from U.S. Patent No. 4,871,855, incorporated herein by reference: dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinidine; 4 -methoxybenzoyl dihydroquinidine; 4-chlorobenzoyl dihydroquinidine; 2-chlorobenzoyl dihydroquinidine; 4-nitrobenzoyl dihydroquinidine; 3-chlorobenzoyl dihydroquinidine; 2-methoxybenzoyl dihydroquinidine; 3 -methoxybenzoyl dihydroquinidine;

2-naphthoyl-dihydroquinidine; cyclohexanoyl dihydroquinidine; p-phenylbenzoyl dihydroquinidine; dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinine; 4-methoxybenzoyl dihydroquinine; 4-chlorobenzoyl dihydroquinine; 2-chlorobenzoyl dihydroquinine; 4-nitrobenzoyl dihydroquinine; 3-chlorobenzoyl dihydroquinine; 2 -methoxybenzoyl dihydroquinine; 3-methoxybenzoyl dihydroquinine; 2-naphthoyl dihydroquinine; cyclohexanoyl dihydroquinine; p-phenylbenzoyl dihydroiquinone; 3. From U.S. Patent No. 5,126,494, incorporated herein by reference: acrylonitrile co-polymer of 9-(4- chlorobenoyloxy)-quinidine; acrylonitrile co-polymer of 11-((2- acryloyloxy)ethyl-sulfinyl)-9-(4- chlorobenoyloxy) -10, 11-dihydroquinidine; acrylonitrile co-polymer of ll-[2- <BR> <BR> <BR> acryloyloxy) ethylsulfonyl] -9- (N,N- <BR> <BR> <BR> <BR> dimethylcarbamoyl) -10, 11-dihydroquinidine; acrylonitrile co-polymer of 9-(10-undecanoyl)- 10, 11-dihydroquinidine; and alkaloid ligands represented by the following structures:

The five step synthetic route to 14 is outlined in Figure 7. In the first step, a mixture of dihydroquinidine, l,4-dichlorophthalazine, KOH, and K2CO3 in dry toluene are refluxes, with a concurrent azeotropic removal of water, to give the mono- substituted chlorophthalazine 17 which upon similar transformation with quinidine provided the di- substituted phthalazine 18 (Amberg et al. J. Org. Chem.

1993, 58, 844; Lohray et al. Tet. Lett. 1994, 35, 6559).

The heating of 18 and 3-mercaptopropionic acid (3-MPA) in the presence of 2,2'-azobisisobutyronitrile (AIBN) in benzene (70 OC) allowed the isolation of 19 as a tan precipitate (Inagaki et al. Bull. Chem. Soc. Jpn. 1987, 60, 4121). The acid 19 was coupled to MeO-PEG-NH2 in

the presence of N,N- dimethylaminopyridine (DMAP) and 1,3-dicyclohexylcarbodiimide (DCC) in methylene chloride (DCM) (Pillai et al. J. Org. Chem. 1980, 45, 5364).

After the reaction was complete, the insoluble urea side-product was removed by filtration and the pegylated ligand 20 was isolated from the reaction mixture by precipitation following a slow addition of diethyl ether. The sulfide 20 was oxidized to the desired sulfone 14 by a mixture of OsO4/N-methylmorpholine-N- oxide (NMO) [in acetone/water (v/v, 2/1)] (Kaldor et al.

Tetrahedron. Lett. 1991, 32, 5043).

Ligand 14 was completely soluble either in t- butanol/water or acetone/water solvent systems allowing the study of homogeneous AD reactions. The AD reaction results of 14 with various olefins are shown in Figure 8 and a number of features are noteworthy. First, it is evident that the t-Butanol/water solvent produces considerably higher ees for all olefins tested consistent with previous reports (Pini et al.

Tetrahedron Lett. 1991, 32, 5175; Lohray et al.

Tetrahedron Lett. 1992, 33, 5453.; Pini et al.

Tetrahedron: Asymmetry 1993, 4, 2351; Lohray et al.

Tetrahedron Lett. 1994, 35, 6559; Pini et al.

Tetrahedron 1994, 50, 11321; Pini et al. Tetrahedron Lett. 1995, 36, 1549; Sung et al. Tetrahedron: Asymmetry 1995, 6, 2687; Petri et al. Chirality 1995, 7, 580; Song et al. Tetrahedron: Asymmetry 1996, 7, 645).

Second, in terms of both reaction time and enantioselectivity, ligand 14 accelerated AD reactions are comparable to its free ligand counterpart, strongly suggesting that the MeO-PEG backbone does not adversely alter either asymmetric induction or the rate of formation of the osmium-ligand-olefin ternary complex.

Finally, the ligand 14 can be isolated in virtually quantitative yield by precipitation using diethyl ether (Han et al. J. Am. Chem. Soc. 1996, 118, 7633).

Combined with our previous example 1 (vida supra: Han et al. J. Am. Chem. Soc. 1996, 118, 7633), the present results demonstrate that MeO-PEG bound ligands behave in a similar fashion to unimmoblized ligands in the AD reaction. What makes this finding even more impressive is that this soluble polymer approach provides the added convenience of ligand recovery and product isolation.

Synthetic Protocals General. Methylene chloride and methanol were dried over CaH2 and powdered magnesium respectively.

Polyethylene glycol monomethyl ether (MeO-PEG, MW. = 5000) was purchased from Aldrich and was dried over P2Os under vacuum before use. All other solvents and chemicals were obtained from commercial sources, and were used without further purification, unless otherwise stated. NMR spectra were obtained on a Bruker AM-300 spectrometer.

Synthesis of Dihydroquinidine Glutarate (Mono-ester) 6 as illustrated in Figure 2. Dihydroquinidine hydrochloride 5 (0.500 g, 1.38 mmol; Aldrich) dissolved in methylene chloride was neutralized by the slow addition of triethylamine (0.140 g, 1.38 mmol) at 4 "C.

This was followed by the addition of 4-(N,N'- dimethyl)aminopyridine (0.170 g, 1.39 mmol) in methylene chloride. To this reaction mixture was slowly added glutaric anhydride (0.315 g, 2.76 mmol). The reaction temperature was raised to room temperature, and the reaction mixture was stirred for an additional 2 h. All volatiles were removed in vacuo, and the resulting residue was purified by column chromatography (silica- gel, methylene chloride: methanol = 9: 1). The dihydroquinidine glutarate (mono-ester) was obtained as a white foam (0.364 g, 60.0 %): 1H-NMR (300 MHz, CDCl3) d 0.95 (t, J = 7 Hz, 3H), 1.35 (m, 1H), 1.5-2.1 (m, 9H), 2.2-2.7 (m, 5H), 3.13 (q, J = 7 Hz, 1H), 3.2-3.6 (m,

3H), 4.11 (s, 3H), 7.28 (d, J = 3 Hz, 1H), 7.40 (dd, J = 9, 3 Hz, 1H), 7.43 (s, 1H), 7.72 (d, J = 4 Hz, 1H), 8.03 (d, J = 9 Hz, 1H), 8.69 (d, J = 4 Hz, lH); 13C-NMR (75 MHz, CDCl3) 6 11.8, 20.1, 21.7, 24.3, 25.5, 26.2, 34.2, 34.4, 35.9, 50.5, 51.2, 57.4, 59.3, 71.5, 102.4, 119.2, 124.2, 127.2, 131.6, 143.0, 144.8, 148.0, 160.4, 172.5, 177.3; HRMS (FAB) calcd for [C25H32N2O5+ H+] 441.2389, found 441.2375.

Synthesis of Ethyl, dihydroquinidine glutarate 3 as illustrated in Figure 2. 1,3-Dicyclohexylcarbodiimide (DCC, 0.153 g, 0.742 mmol) was added to a mixture of absolute ethanol (0.0113 g, 0.245 mmol), DMAP (0.0150 g, 0.123 mmol), and the dihydroquinidine glutarate (mono- ester) 6 (0.324 g, 0.736 mmol) in methylene chloride.

The reaction mixture was stirred until the dihydroquinidine glutarate (mono-ester) 6 disappeared as determined by thin layer chromatography (TLC; the reaction typically took approximately 3 h). The reaction mixture was filtered through celite to allow removal of urea precipitate, and the product was isolated after evaporation of the solvent. The residue was purified by column chromatography (methylenechloride: methanol = 9: 1) to give 3 (0.109 g, 95.1 %): 1H-NMR (300 MHz, CDCl3) d 0.87 (t, J = 7 Hz, 3H), 1.35 (m, 1H), 1.16 (t, J = 7 Hz, 3 H), 1.3-1.6 (m, 7H), 1.75-2.00 (m, 3H), 2.43 (t, J = 7 Hz, 2H), 2.6-3.0 (m, 3H), 3.24 (q, J = 7 Hz, 1H), 3.92 (s, 3H), 4.05 (q, J = 7 Hz, 2H), 6.62 (d, J = 7 Hz, 1H), 7.26 (d, J = 3 Hz, 1H), 7.32 (dd, J = 9, 3 Hz, 1H), 7.42 (d, J = 4 Hz, lH), 7.93 (d, J = 9 Hz, 1H), 8.69 (d, J = 4 Hz, lH); 13C- NMR (75 MHz, CDCl3) d 11.7, 14.1, 19.9, 22.3, 25.2, 25.7, 26.3, 33.0, 33.2, 36.6, 49.6, 50.3, 55.9, 58.7, 60.3, 72.7, 101.1, 118.0, 122.1, 126.6, 131.6, 143.1, 144.5, 147.1, 158.1, 171.4, 172.7; HRMS (FAB) calcd. for [C27H36N2O5 + H+] 469.2702, found 441.2718.

Synthesis of polyethylene glycol, dihydroquinidine glutarate 2 as illustrated in Figure 2. DCC (0.153 g,

0.742 mmol) was added to a mixture of polyethylene glycol (1.23 g, 0.246 mmol), DMAP (0.0150 g, 0.123 mmol), and dihydroquinidine glutarate (mono-ester) 6 (0.324 g, 0.736 mmol) in methylene chloride. The reaction mixture was stirred for 6 h and the urea precipitate was removed by filtration through celite.

Diethyl ether was slowly added to the filtrate with a vigorous stirring, and a precipitate formed from was isolated as a white solid. This material was washed with absolute ethanol and diethyl ether, then dried over P2Os in vacuo (1.29 g, 96.0 %). 1H-NMR (300 MHz, CDCl3) 6 0.98 (t, J = 7 Hz, 3H), 1.40 (m, 1H), 1.6-2.7 (m, 17H), 3.11 (q, J = 7 Hz, 1H), 3.2-3.9 (polyethylene glycol peaks), 4.13 (s, 3H), 4.20 (t, J = 7 Hz, 2H), 7.23 (d, J = 3 Hz, 1H), 7.39 (dd, J = 9, 3 Hz, 1H), 7.59 (s, 1H), 7.80 (s, 1H), 8.05 (d, J = 9 Hz, 1H), 8.65 (d, J = 4 Hz, lH).

Procedure for asymmetric dihydroxylation wherein representative substrates, product yields and ee's are tabulated in Figure 3. A small aliquot (catalytic amount) of OS04 in t-butanol (22 ml of Os04 2.5 wt % solution) was added to a mixture of the polymer catalyst (0.300 g, 54.9 mmol), 4-methylmorpholine N-oxide (0.0386 g, 0.329 mmol), and tetraethylammonium acetate tetrahydrate (0.0574 g, 0.220 mmol) in acetone-water (10/1, v/v, 4 ml) at 4 OC. After stirring for 10 min, the olefin (4 eq. relative to the polymer ligand) was added either in one portion or by a slow addition. The reaction mixture was stirred at 4 "C until the olefin disappeared as judged by TLC. Next, solid sodium metabisulfate (0.500 g) was added, and the mixture was stirred for an additional 5 min, then diluted with methylene chloride (10 ml) and dried over Na2SO4. All solids were removed by filtration and washed three times with methylene chloride (3 x 5 ml). The combined filtrates were evaporated to half volume and diethyl ether was slowly added to the mixture under vigorous

stirring conditions. The precipitate obtained was collected on a glass filter, washed with absolute ethanol/ethyl ether, and dried in vacuo (0.294 g, 98 % recovery of ligand resulted from this procedure).

Reduction in vacuo of the filtrate gave nearly pure dihydroxylated product. The enantiomeric excess of the diol was determined either by HPLC of the bis-acetate of hydrobenzoin on a Pirkle 1A ionic D-phenylglycine column or through NMR analysis of the derived bis-Mosher ester.

General Synthesis of Compounds 7 - 10 (as shown in Figure 4): Trans-cinnamic acid (1.1 equivalents; Aldrich) was coupled to each heterogeneous resin (1.0 equivalents) with the aide of EDC/DMAP coupling agents (1.2 equivalents EDC / 0.10 equivalents DMAP) in 0.10 Molar methylene chloride. For the soluble polymer MeO- PEG, a DCC/DMAP (1.2 equivalents EDC / 0.10 equivalents DMAP) combination was used for the preparation of 10.

Upon reaction completion (approximately 12 hours at 25 "C), the urea formed was removed by filtration. To the filtrate was added diethyl ether and 10 was separated as a white solid from the mixture. The loading of trans- cinnamic acid onto each polymer was determined to be >98 % by measuring the release of cinnamic acid and its methyl ester from the polymers with TFA/CH2Cl (v/v, 95/5) and sodium methoxide in methanol respectively.

Procedure for asymmetric dihydroxylation for olefinic substrates attached to polymeric supports (example using polymer bound olefins) (Figures 5-6): Method A: A small aliquot of OsO4 in t-butanol (0504 2.5 wt % solution, see Figure 6 for the amount used in each reaction) was added to a mixture of the ligand (ligands 13 or 14 of Figure 5 or ligands 2 or 3 of Figure 2), K3Fe(CN)6 (0.250 g, 6 eq. relative to the olefin), K2CO3 (0.052 g, 3 eq.), and methanesulfonamide (0.012 g, 1 eq.) in t-butanol-water (1/1, v/v, 4 ml) at room temperature. Alternatively, the following ligands as described in the following U.S. Patents can be used in lieu of ligands 13, 14, 2, or 3: 1. from U.S. Patent No. 5,260,461, incorporated herein by reference: <BR> <BR> <BR> 1,4-bis-(9'-O-dihydroquinidyl)-phthalazine; <BR> <BR> <BR> <BR> <BR> 1, 4-bis- (9 '-O-quinidyl) -phthalazine; <BR> <BR> <BR> <BR> <BR> 3, 6-bis- (9 '-0-dihydroquinidyl) -pyridazine; 3,6-bis-(9'-O-quinidyl)-pyridazine; 1,4-bis- (9'-O-dihydroquinyl)-phthalazine; 1,4-bis-(9'-O-quinyl)-phthalazine; <BR> <BR> <BR> 3, 6-bis- (9' -0-dihydroquinyl) -pyridazine; <BR> <BR> <BR> <BR> <BR> 3, 6-bis- (9' -0-quinyl) -pyridazine; 2. from U.S. Patent No. 4,871,855, incorporated herein by reference: dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinidine; 4-methoxybenzoyl dihydroquinidine; 4-chlorobenzoyl dihydroquinidine; 2 -chlorobenzoyl dihydroquinidine; 4-nitrobenzoyl dihydroquinidine; 3-chlorobenzoyl dihydroquinidine; 2-methoxybenzoyl dihydroquinidine; 3 -methoxybenzoyl dihydroquinidine; 2-naphthoyl-dihydroquinidine;

cyclohexanoyl dihydroquinidine; p-phenylbenzoyl dihydroquinidine; dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinine; 4-methoxybenzoyl dihydroquinine; 4-chlorobenzoyl dihydroquinine; 2-chlorobenzoyl dihydroquinine; 4-nitrobenzoyl dihydroquinine; 3 -chlorobenzoyl dihydroquinine; 2-methoxybenzoyl dihydroquinine; 3 -methoxybenzoyl dihydroquinine; 2 -naphthoyl dihydroquinine; cyclohexanoyl dihydroquinine; p-phenylbenzoyl dihydroiquinone; 3. From U.S. Patent No. 5,126,494, incorporated herein by reference: acrylonitrile co-polymer of 9-(4- chlorobenoyloxy)-quinidine; acrylonitrile co-polymer of 11-((2- acryloyloxy)ethyl-sulfinyl)-9-(4- chlorobenoyloxy) -10, 11-dihydroquinidine; acrylonitrile co-polymer of 11-[2- <BR> <BR> acryloyloxy) ethylsulfonyl] -9- (N,N- <BR> <BR> <BR> dimethylcarbamoyl) -10, 11-dihydroquinidine; acrylonitrile co-polymer of 9-(l0-undecanoyl)- 10, 11-dihydroquinidine.

After stirring for 10 min, the polymer bound olefin (0.125 mmol, 1 eq.) was added in one portion. The reaction mixture was stirred for 24 hrs. To the reaction mixture was added solid sodium metabisulfate (0.400 g), and the mixture was stirred for an additional 5 min. The reaction mixture was filtered, and successively washed with acetone and methylene chloride.

The resin was dried over Na2SO4 under vacuo. i) The resin was suspended in TFA/CH2Cl2 (v/v, 95/5) for 30 min.

After filtering off the resin, the filtrate was

evaporated to dryness. The % conversion of the olefin to the dihydroxylated product was determined by NMR. A comparision of 1H olefinic signals of the unreacted cinnamic acid with aliphatic proton signals of the product diol acid allowed assessment of the percent conversion. ii) The resin was suspended in a methanol solution of sodium methoxide. This mixture was stirred for 2 h at room temperature and the solvent was removed in vacuo. The mixture was separated by silica gel column chromatography (ethyl acetate/hexane, 1/1). The enantiomeric excess of the product diol was determined through NMR analysis of the derived bis-Mosher ester of this methyl ester.

Method B: A similar procedure to Method A was followed except acetone/water solvent and methylmorpholine N-oxide (NMO) were used. The reaction mixture contained a small aliquot of OSO4 in t-butanol (iso, 2.5 wt % solution, 13 pl, 0.01 eq. relative to the olefin), the ligand (0.0025 g, 0.025 eq.), 4-methylmorpholine N-oxide (0.022 g, 1.5 eq.), and tetraethylammonium acetate tetrahydrate (0.033 g, 1.0 eq.) in acetone-water (10/1, v/v, 4 ml) at 4 OC.

Synthesis of compound 18 as illustrated in Figure 7. A mixture of dihydroquinidine 15 (5.00 g, 15.32 mmol, 1 eq; Aldrich), 1,4-dichlorophthalazine 16 (3.66 g, 18.38 mol, 1.2 eq; Aldrich), and K2CO3 (6.35 g, 45.95 mmol, 3 eq) in dry toluene are ref fluxed with a concurrent azeotropic removal of water for 2 hr. Then, 85 % KOH pellets (3.02 g, 45.95 mmol, 3 eq) was added at once and the reaction mixture was ref fluxed until dihydroquinidine disappeared on t.l.c. The light orange solution was cooled to room temeperature, mixed with water, and then extracted with ethyl acetate. The organic layer was dried over sodium sulfate. The removal of solvent and then column chromatogarphy of residue gave the mono-

substituted chlorophthalazine 17 (6.05 g, 81 %), which upon similar transformation with quinidine provided the di-substituted phthalazine 18 (6.76 g, 85 % starting from 5.00 g of 17). 17: 1NMR (300 MHz, CDCl3) 6 0.91 (t, J = 6.7 Hz, 3H), 1.40-1.70 (m, 6H), 1.77 (m, 1H), 2.12 (m, 1H), 2.65-3.00 (m, 4H), 3.43-3.58 (m, 1H), 3.99 (s, 3H), 7.27 (d, J = 7.0 Hz, 1H), 7.34 (dd, J = 9.3 & 2.8 Hz, 1H), 7.44 (d, J = 4.5 Hz, 1H), 7.62 (d, J = 2.7 Hz, 1H), 7.96 (d, J = 9.2 Hz, 1H), 8.00 (m, 2H), 8.13- 8.20 (m, 1H), 8.33-8.42 (m, 1H), 8.64 (d, J =4.5 Hz, lH); l3C NMR (75 MHz, CDCl3) 6 11.9, 23.5, 25.4, 26.1, 27.1, 37.3, 49.9, 50.9, 55.5, 59.9, 101.7, 118.5, 121.4, 121.8, 122.8, 122.9, 125.3, 127.7, 128.2, 131.6, 133.0, 133.3, 144.0, 144.7, 147.0, 147.2, 150.5, 157.7; HRMS (FAB+) calcd for [C2eH29ClN4O2 + W] = 489.2057, found = 489.2051. 18: 1NMR (300 MHz, CDCl3) 6 0.78 (t, J = 6.7 Hz, 3H), 1.20-1.65 (m, 10H), 1.68 (m, 1H), 1.85-2.30 (m, 3H), 2.55-3.00 (m, 8H), 3.39 (m, 2H), 3.88 (s, 6H), 4.97 (m, 2H), 5.90 (m, 1H), 6.95 (d, J = 6.5 Hz, 1H), 7.01 (d, J = 5.9 Hz, 1H), 7.34 (m, 2H), 7.40 (d, J = 4.6 Hz, 1H), 7.42 (d, J = 4.6 Hz, 1H), 7.52 (d, J = 2.7 Hz, 1H), 7.54 (d, J = 2.6 Hz, 1H), 7.91 (m, 2H), 7.96 (d, J = 9.2 Hz, 1H), 7.97 (d, J = 9.2 Hz, 1H), 8.31 (m, 2H), 8.61 (d, J = 4.5 Hz, 1H), 8.62 (d, J = 4.5 Hz, 1H); 13C NMR (75 MHz, CDCl3) 6 11.8, 23.2, 23.4, 25.2, 26.0, 26.2, 26.4, 27.2, 27.7, 37.3, 39.6, 49.4, 49.8, 49.9, 50.8, 55.5, 60.0, 60.2, 76.0, 77.2, 101.9, 102.0, 114.6, 118.2, 118.4, 121.7, 121.8, 122.3, 122.7, 123.0, 127.3, 128.1, 129.0, 131.4, 131.5, 132.0, 132.1, 140.3, 144.6, 144.8, 144.9, 147.3, 156.3, 156.4, 157.5, 157.6; HRMS (FAB+) calcd for [C9sHs2N6o4 + W] = 777.4128, found 777.4104.

Synthesis of Compound 19 as illustrated in Figure 7.

The heating of 18 (6.00 g, 7.73 mmol, 1 eq) and 3- mercaptopropionic acid (3-MPA, 0.821 g, 7.74 mmol, 1 eq) in the presence of 2,2'-azobisisobutyronitrile (AIBN,

0.0635 g, 0.3867 mmol, 0.05 eq) in benzene (70 "C) allowed the isolation of 19 as a tan precipitate (4.43 g, 64 %). This precipitate was already such pure that it was used for the next step without further purification. 19: 1NMR (300 MHz, CDCl3) 6 0.86 (t, J = 6.7 Hz, 3H), 1.25-3.25 (m, 30H), 3.50 (m, 2H), 3.89 (s, 3H), 3.97 (s, 3H), 6.60 (broad s, 1H), 7.30 (m, 3H), 7.44 (t, J = 4.4 Hz, 2H), 7.49 (broad s, 1H), 7.53 (d, J = 2.3, 1H), 7.63 (d, J = 2.4 Hz, 1H), 7.88 (m, 2H), 7.95 (d, J = 9.2 Hz, 1H), 7.96 (d, J = 9.2 Hz, 1H), 8.19 (d, J = 8.1 Hz, 1H), 8.27 (d, J = 7.7 Hz, 1H), 8.59 (d, J = 4.6 Hz, 1H), 8.64 (d, J = 4.5 Hz, 1H); HRMS (FAB+) calcd for [C51H5sN6O6S + W] = 883.4216, found 883.4242.

Synthesis of Compound 20 as illustrated in Figure 7. A mixture of acid 19 (1.00 g, 1.13 mmol, 1.5 eq), MeO-PEG- NH2 (3.78 g, 0.756 mmol, 1 eq), N,N- dimethylaminopyridine (DMAP, 0.0231 g, 0.189 mmol, 0.25 eq) and 1,3-dicyclohexylcarbodiimide (DCC, 0.233 g, 1.13 mmol, 1.5 eq) in methylene chloride (DCM) was stirred overnight at room temperature. After the reaction was complete, the insoluble urea side-product was removed by filtration and the pegylated ligand 20 (4.34 g, 98 t) was isolated from the reaction mixture by precipitation following a slow addition of diethyl ether. 20: 1NMR (300 MHz, CDCl3) 6 0.87 (t, J = 6.7 Hz, 3H), 1.25-2.85 (m, 30H), 3.20-3.75 (PEG peaks), 3.84 (t, J = 6.5 Hz, 2H), 3.88 (s, 6H), 6.59 (broad t, 1H), 6.90 (d, J = 6.6 Hz, 1H), 6.97 (d, J = 5.7Hz, 1H), 7.30 (dd, J = 9.2 & 2.4 Hz, 1H), 7.32 (dd, J = 9.0 & 2.4 Hz, 1H), 7.39 (t, J = 4.6 Hz, 2H), 7.47 (d, J = 2.5 Hz, 1H), 7.51 (d, J = 2.5 Hz, 1H), 7.90 (m, 2H), 7.93 (d, J = 9.0 Hz, 1H), 7.94 (d, J = 9.2 Hz, 1H), 8.28 (m, 2H), 8.59 (d, J = 4.5 Hz, 2H).

Synthesis of Compound 14 as illustrated in Figure 7. To the mixture of the sulfide 20 (4.00 g, 0.682 mmol, leq),

and N-methylmorpholine-N-oxide (NMO, 0.24 g, 2.046 mmol, 3 eq) in acetone/water (v/v, 2/1) was added OSO4 (1 mol %, 0.01 eq). The resulting reaction mixture was stirred overnight at room temperature, and then quenched with sodium bisulfide (100 mg). The reaction mixture was dried over sodium sulfate, and solid was filtered off.

Slow addition of ether to the filtrate separated 14 as a white precipitate (3.90 g, 97 %). 14: 1NMR (300 MHz, CDCl3) 0.79 (3H), 1.10-3.00 (30H), 3.10-3.80 (PEG peaks), 3.80-4.00 (8H), 6.70 (1H), 6.95 (1H), 7.05 (1H), 7.20-7.45 (7H), 7.50 (1H), 7.80-8.10 (4H), 8.28 (1H), 8.37 (1H), 8.55 (2H). Peak mutiplicity was not further characterized due to line-broadening.

General procedure for the attachment of PEG soluble support to the ligand: A mixture of ligand acid (1.00 g, 1.13 mmol, 1.5 eq., MeO-PEG-NH2 (3.78 g, 0.756 mmol, 1 eq; Aldrich/ Sigma/ Fluka, etc.), N,N- dimethylaminopyridine (DMAP, 0.0231 g, 0.189 mmol, 0.25 eq) and 1,3- dicyclohexylcarbodiimide (DCC, 0.233 g, 1.13 mmol, 1.5 eq) in methylene chloride (DCM) was stirred overnight at room temperature. After the reaction was complete, the insoluble urea side-product was removed by filtration and the pegylated ligand (4.34 g, 98 %) was isolated from the reaction mixture by precipitation following a slow addition of diethyl ether. Ligand acid can be obtained as described in the following U.S. Patents: 1. from U.S. Patent No. 5,260,461, incorporated herein by reference, ligand acids may be obtained for the following alkaloids: 1,4-bis-(9'-O-dihydroquinidyl)-phthalazine; 1,4-bis-(9'-O-quinidyl)-phthalazine; 3, 6-bis- (9 '-0-dihydroquinidyl) -pyridazine; 3,6-bis-(9'-O-quinidyl)-pyridazine; 1,4-bis- (9'-O-dihydroquinyl)-phthalazine; 1,4-bis-(9'-O-quinyl)-phthalazine;

3, 6-bis- (9' -0-dihydroquinyl) -pyridazine; and 3, 6-bis- (9' -0-quinyl) -pyridazine; 2. from U.S. Patent No. 4,871,855, incorporated herein by reference, ligand acids may be obtained for the following alkaloids: dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinidine; 4-methoxybenzoyl dihydroquinidine; 4-chlorobenzoyl dihydroquinidine; 2-chlorobenzoyl dihydroquinidine; 4-nitrobenzoyl dihydroquinidine; 3-chlorobenzoyl dihydroquinidine; 2-methoxybenzoyl dihydroquinidine; 3-methoxybenzoyl dihydroquinidine; 2-naphthoyl-dihydroquinidine; cyclohexanoyl dihydroquinidine; p-phenylbenzoyl dihydroquinidine; dimethylcarbamoyl dihydroquinidine; benzoyl dihydroquinine; 4 -methoxybenzoyl dihydroquinine; 4-chlorobenzoyl dihydroquinine; 2 -chlorobenzoyl dihydroquinine; 4-nitrobenzoyl dihydroquinine; 3-chlorobenzoyl dihydroquinine; 2-methoxybenzoyl dihydroquinine; 3 -methoxybenzoyl dihydroquinine; 2-naphthoyl dihydroquinine; cyclohexanoyl dihydroquinine; and p-phenylbenzoyl dihydroiquinone; 3. From U.S. Patent No. 5,126,494, incorporated herein by reference, ligand acids may be obtained for the following alkaloids: acrylonitrile co-polymer of 9-(4- chlorobenoyloxy)-quinidine; acrylonitrile co-polymer of 11-((2- acryloyloxy)ethyl-sulfinyl)-9-(4- chlorobenoyloxy) -10, 11-dihydroquinidine; acrylonitrile co-polymer of 11-[2- acryloyloxy)ethylsulfonyl]-9-(N,N- dimethylcarbamoyl)-10,11-dihydroquinidine; and acrylonitrile co-polymer of 9-(10-undecanoyl)- 10, 11-dihydroquinidine.

General procedure for asymmetric dihydroxylation using ligands attached to PEG support: Method A: A small aliquot of OsO4 in t-butanol (iso4 2.5 wt % solution, 0.01-0.1 eq) was added to a mixture of K3Fe(CN)6 (0.250 g, 6 eq. relative to the olefin), K2CO3 (0.052 g, 3 eq.), methanesulfonamide (0.012 g, 1 eq.), and the PEG-bound alkaloid ligand 14 (0.025-0.25 eq.) or any of the PEG-bound alkaloid ligands described above, in t-butanol-water (1/1, v/v, 4 ml) at room temperature.

After stirring for 10 min, the olefin substrate (0.125 mmol, 1 eq.was added in one portion. Permissible substrates include all classes of primary, secondary, tertiary and quaternary substituted olefins described in the prior art for the AD reaction. The reaction mixture was stirred for 24 hours. The reaction mixture was stirred at 4"C or room temperature until the olefin disappeared as judged by TLC. Next, solid sodium metabisulfate (0.500 g) was added, and the mixture was stirred for an additional 5 min, then diluted with methylene chloride (10 ml) and dried over Na2SO. All solids were removed by filtration and washed three times with methylene chloride (3 x 5 ml). The combined filtrates were evaporated to half volume and diethyl ether was slowly added to the mixture under vigorous stirring conditions. The precipitate obtained was collected on a glass filter, washed with absolute ethanol/ethyl ether, and dried in vacuo (0.294 g, 98 % recovery of ligand resulted from this procedure).

Reduction in vacuo of the filtrate gave nearly pure dihydroxylated product. The enantiomeric excess of the diol was determined either by HPLC of the bis-acetate of hydrobenzoin on a Pirkle 1A ionic D-phenylglycine column or through NMR analysis of the derived bis-Mosher ester.

Method B: A small aliquot (catalytic amount) of OSO4 in t-

butanol (0.01-0.1 eq) was added to a mixture of the polymer catalyst (PEG-ligand; see Method A for examples of compounds used) (0.025-0.25 eq), 4-methylmorpholine N-oxide (1.5 eq), and tetraethylammonium acetate tetrahydrate (1 eq) in acetone-water (10/1, v/v, 4 ml) at 4 "C. After stirring for 10 min, the olefin (1 eq) was added either in one portion or by a slow addition.

The reaction mixture was stirred at 4 "C until the olefin disappeared as judged by TLC. Next, solid sodium metabisulfate (0.500 g) was added, and the mixture was stirred for an additional 5 min, then diluted with methylene chloride (10 ml) and dried over Na2SO4. All solids were removed by filtration and washed three times with methylene chloride (3 x 5 ml). The combined filtrates were evaporated to half volume and diethyl ether was slowly added to the mixture under vigorous stirring conditions. The precipitate obtained was collected on a glass filter, washed with absolute ethanol/ethyl ether, and dried in vacuo (0.294 g, 98 % recovery of ligand resulted from this procedure).

Reduction in vacuo of the filtrate gave nearly pure dihydroxylated product. The enantiomeric excess of the diol was determined either by HPLC of the bis-acetate of hydrobenzoin on a Pirkle 1A ionic D-phenylglycine column or through NMR analysis of the derived bis-Mosher ester.