The Synthesis of Garner S Aldehyde

1

Truitt

Summer Research 2007:

The Synthesis of Garner’s Aldehyde

By: Kristina J. Truitt

Advisor: Dr. Geisler

Department of Chemistry

ABSTRACT

The overall objective of this project was to form analogs of sphingosine. In making these analogs of sphingosine, the ultimate goal was to hopefully use the analogs as inhibitors of excessive cell growth, such as tumor cells, by inhibiting the enzyme of Protein Kinase C (PKC). Unfortunately, this was not a one step reaction or process. In order to create the analogs, the synthesis of Garner’s aldehyde must be done. During this process of synthesizing it was extremely important that the end product was in its most pure state.

INTRODUCTION

In the year 1984, a man by the name of Dr. Philip Paul Garner reported the first synthesis of 1,1-dimethylethyl 4-formyl-2, 2-dimethyl-oxazolidine-3-carboxylate, which is today known as Garner’s Aldehyde. Dr. Garner received his Ph.D. degree from the University of Pittsburgh in Organic chemistry and later established a research group that focused on the synthesis of natural products and developing new methods for organic synthesis. As a result of the development of his widely used derivative, Garner’s Aldehyde has been used in over 200 reported studies. It has been found that various methods can be used to form Garner’s aldehyde. The first synthesis of 6, by Garner, began with the idea of forming derivates with protecting groups. According to Journal of Chemistry article,

Dr. Garner noted that certain acetamidosugars could be protected as 2, 2-dimethyloxazolidine derivatives, and conceived the idea that a similar protection of the

hydroxy group and the Boc-protected amino-group of serinewould be desirable. His synthesis started with Boc protection of L-serine (2) using di-tert-butyl dicarbonate [(Boc)2O] at pH = 10form N-Boc-serine 3, which was converted to the methyl

ester 4 either by diazomethane 3 or, more conveniently, withMeI and K2CO3 (Scheme 1).4 Compound 4 was then treatedwith Me2C(OMe)2 and TsOH to give the oxazolidine ester5 in 70–89% yield. Direct reduction of ester 5 with DIBAL in toluene then afforded the title aldehyde 1 in 76% yield (Scheme 1).[1]

This scheme differs from the synthesis used within this project because Garner’s synthesis step in making the oxazolidine structure (5, in scheme 1) for the protection of the acidic hydrogens involves a different reagent. Within the procedure followed for the project, the oxazolidine structure (5, in scheme 1) was formed by using dimethoxypropane (50 mL, 400 mmol) and the catalyst borontrifluoride etherate (0.35 mL, 2.8 mmol). The differences in the reagents used in the forming of 5 are due to researchers probably finding modifications of the synthesis to produce, perhaps, a better yield.

For example, Garner’s original synthesis was modified in the first two steps of the scheme by McKillop et al.[2]

It was found that the reversing of the first two steps (Scheme 2), Boc protection and esterification, gave a substantially increased yield of 94% compared to Garner’s yield of 76%. In treating the starting material with HCL, it can be assumed that its addition contributed to the increased yield of methyl serinate9.

Another modification of Garner’s original synthesis focused on the reduction step of compound 5. Instead of using DIBAL for reduction, lithium aluminum hydride was used as a replacement, which was the exact reducing agent used in this project. Within the Journal of Organic Chemistry, Marshall et al proved that with the use of lithium aluminum hydride they yielded 90%. [3] This method of using this particular reducing agent is known as the Swern Oxidation. The Swern Oxidation method is used on alcohols to prevent the use of toxic metals and in using this method, aldehydes and ketones can be prepared from primary and secondary alcohols, such as the compound formed within this project.

With various modifications of Garner’s aldehyde, it can undergo reactions with Grignard reagents, organometallic reagents, reagents for Witting reactions. These are commonly known organic reactions that depend upon the stability of the compound it is being added to. It is said that Garner’s aldehyde is a useful, non-racemic building block of many synthetic reactions of natural compounds. Through the project, Garner’s aldehyde will serve as the starting material of forming sphingosine analogs.

METHODS

RXN A: N- [(1, 1-dimethylethoxy)carbonyl]-L-serine methyl ester

(1)(2)

The materials used in this experiment included a 500 ml, 3-necked, round-bottomed flask, a magnetic stirring bar, a magnetic stirring apparatus, thermometer, reflux condenser, and a pressure-equalizing dropping funnel. Before starting the experiment, all the required glassware, including the graduated cylinders used for measurements, were washed with acetone and placed in the oven to dry. All the glassware had to be placed in the oven to ensure that all the glassware contained absolutely no water, to prevent the contamination of the product.

To a 500-mL three-necked flask was added a magnetic stirring bar, Methyl serinate hydrochloride (10.0g, 64.3 mmol), and tetrahydrofuran (100 mL). The mixture within the flask was then swirled by hand to initial the mixing of the contents within the flask. The flask was then equipped with a thermometer, a reflux condenser and a dropping funnel. The reflux condenser was protected by a calcium-filled drying tube to prevent moisture from destroying the solution in the flask. The dropping funnel was charged with exactly 97% di-tert-butyl dicarbonate (14.3g, 63.6mmol) and tetrahydrofuran (100 mL). A nitrogen line was also connected to the dropping funnel through a rubber septum that covered the top opening of the dropping funnel (Note 1). The white mixture that began to form within the flask was then placed in an ice bath to cool. During the cooling and stirring period, the solution of 97% di-tert-butyl dicarbonate within the dropping funnel was added drop wise over the hour period. When an hour passed the solution was stirred for an additional ten minutes. The ice water bath was then removed and the mixture was stirred overnight for approximately 17 hours.

After being stirred overnight, the reaction mixture was then warmed at 60˚C for three hours. During the warming process, the temperature was monitored to control the temperature of the solution, while also being stirred on the magnetic stirring apparatus. The temperature was controlled through the use of the heating mantle and was monitored by observing the temperature reading on the thermometer. The solution was then removed from the magnetic stirring apparatus and heating mantle after the nitrogen flow was turned off. The contents within the flask were placed on the rotary evaporator to remove the solvent and then partitioned between diethyl ether (200 mL) and saturated aqueous sodium bicarbonate (250 mL). The solution was transferred to an acetone washed and oven dried separatory funnel in order to separate the bottom aqueous layer from the top organic layer by the method of extraction. The aqueous layer was extracted from the separatory funnel into an Erlenmeyer flask and the organic layer was then extracted from the separatory funnel into a separate Erlenmeyer flask. The funnel was closed and the aqueous layer was placed back into the separatory funnel to be washed with three portions of 75-mL of diethyl ether (Note 2). The organic layers were placed in one round-bottomed flask that had been previously weighed and placed on the rotary evaporator to remove all the solvent from the solution. The solution left within the flask, after the removing of the solvent, was placed under the vacuum overnight and weight of the contents within the flask were calculated. For further analysis of the product, 2, a TLC analysis was taken (Note 3).

RXN B: 3-(1, 1-Dimethylethyl) 4-methyl-(S)-2, 2-dimethyloxazolidine-3, 4-

dicarboxylate

(2)(3)

The product from reaction A, N-Boc-L-serine methyl ester was used to start this reaction. Instead of using the entire product retrieved from reaction A, only N-Boc-L-serine methyl ester (10g, 45.6 mmol) was used. The remaining product was left in a separate flask. Exactly acetone (165 mL) was added to 2, 2-dimethoxypropane (50 mL, 400 mmol) and boron trifluoride (0.35 mL, 2.8 mmol) (Note 4). The reaction mixture resulted in a dark orange-red solution, which indicated that a reaction had begun within the flask. The dark solution was then stirred for 2.5 hours to ensure that the reaction went to completion.

After being stirred for 2.5 hours, the solution was analyzed through TLC analysis in 1:1 ethyl acetate: hexane. After the TLC analysis, the solution was charged with 99% triethylamine (0.9 mL). The solution lost its dark color and became a pale yellow color. The solution was placed on the rotary evaporator to remove the solvent. A tick syrup substance resulted from the rotary evaporator. The substance was then partitioned between diethyl ether (150 mL) and saturated aqueous sodium bicarbonate solution (250 mL). The purpose of the partitioning was to retrieve the organic layer. The aqueous layer was washed with 2- 75 mL portions of Diethyl ether. The organic layer was then dried with 2-5 scoops (depending on how proportionate the scoops are) of anhydrous sodium sulfate. The solution retrieved from the drying step was then placed on the rotary evaporator again to remove the solvent. The flask was then weighed and further calculations were made. The final product, oxazolidine methyl ester, was placed under the vacuum overnight.

For further purification of the product, column chromatography was used with the eluent 1:4 ethyl acetate: hexane. An extremely small piece of cotton and a thin layer of sand were placed at the bottom of the column to prevent leakage. Approximately silica gel (200 mL) and 1:4 ethyl acetate: hexane mixed was poured into the column and any of the silica gel that hardened on the sides of the inner part of the column was washed with the solvent, 1:4 ethyl acetate: hexane (Note 5). The layer was silica gel within the column was washed with 1:4 ethyl acetate: hexane by pouring an approximate 3-inch layer of the solvent into the column and draining it into a flask. Once the solvent layer was no longer on top of the silica gel, the product was added to the column by the use of a long-stemmed pipette. The end tip of the pipette was placed along the inner sides of the column and the product was added in a circular motion to ensure the even distribution of the product on top of the silica gel layer.

Once the entire product was added to the column, an approximate 5 to 6 inch layer of 1:4 ethyl acetate: hexane was added to the column for the collectionof fractions. In collecting the fractions, TLC analysis was done on each fraction to separate wanted product from unwanted product. The amount of fractions collected depended upon the spotting on the TLC plates. When spots no longer began to appear on the TLC plates, it was assumed that the entire product had been washed from the column into the various flasks used for collection (Note 6).

RXN C:N-[(1, 1-Dimethylethoxy) carbonyl]-N, O- isopropylidene-L- serinol

(3)(4)

Using the product from Reaction B, oxazolidine methyl ester, a mixture was made within a 500 ml, 3-neck, round bottomed flask. The 3-neck flask was filled with the solution from reaction B, 100 ml of tetrahydrofuran, and 2.16g of lithium aluminum hydride. The 3-neck flask was equipped with a thermometer, a reflux condenser and a dropping funnel. The dropping funnel attached to the flask was charged with 50 ml of tetrahydrofuran and 9.90g of oxazolidine ester. The solution in the dropping funnel was added to the flask dropwise for a period of twenty minutes. After the addition of the solution within the dropping funnel, the funnel was washed with 2- 3ml portions of tetrahydrofuran, which was suspended for a further twenty minutes. A TLC analysis was done to show the progression of the reaction.

The reaction was then placed in an ice-water bath while 20 ml of 10% KOH was added dropwise for ten minutes. (NOTE: The reaction was exothermic.) The reaction underwent additional stirring for one hour at room temperature. The resulting white precipitate was removed by filtration through a celite pad, which was rinsed with 3- 30 ml portions of diethyl ether. The organic phase was washed with 100 ml of aqueous phosphate buffer and the aqueous layer was extracted. The aqueous layer was then extracted with 3- 30 ml portions of diethyl ether. The organic phases were combined and were dried with anhydrous magnesium sulfate. Finally, the solution was placed under a rotary evaporator and then under the vacuum. The color of the product was not exactly a pale yellow.

RESULTS

N- [(1, 1-dimethylethoxy) carbonyl]-L-serine methyl ester

1. In a three-neck flask bathed in an iced bath, methyl serinate hydrochloride (10g, 64.3 mmol), tetrahydrofuran (100 mL) were stirred with a solution of 97% di-tert-butyl dicarbonate (14,3g, 63.6 mmol) and tetrahydrofuran (100 mL).After stirring for an hour and ten minutes, the ice bath was removed the reaction mixture continued to stir for 17 hours. During the cooling and stirring period, the solution of 97% di-tert-butyl dicarbonate within the dropping funnel was added drop wise over the hour period. After being stirred overnight, the reaction mixture was then warmed at 60˚C for three hours. During the warming process, the temperature on the thermometer rose to 70˚C. The heat was reduced to drop the temperature, but it did not drop to the required temperature of 50˚C. Once the heating process was completed, the contents within the flask were placed on the rotary evaporator at 40˚C to reduce the volume of the solution. Partitioning of the compound 2 was done between diethyl ether (200 mL) and saturated aqueous sodium bicarbonate (250 mL). The aqueous layer was extracted from the separatory funnel into an Erlenmeyer flaskwith three portions of 75-mL of diethyl ether (Note 2) and the organic layer was collected for a further step.The organic layer was then dried with 2-5 scoops (depending on how proportionate the scoops are) of anhydrous sodium sulfate. For further analysis of the product, 2, a TLC analysis was taken (Note 3). The first TLC analysis was done on Compound 2 to determine whether the product contained any contaminants that could possibly destroy the product itself. The plates were done on silica plates that allowed Rƒ values to be calculated to determine the distance the solute traveled with the solvent. The eluent 1:1 ethyl acetate: hexane was used for the TLC plates. Three trials were run to correct errors made throughout the reaction. The yield wanted for compound 2 ranged from 93-97%. The yield of the first trial was not calculated, the yield of the second trial was 133.53%, and the yield of the third trial was 95.31%.

3-(1, 1-Dimethylethyl) 4-methyl-(S)-2, 2-dimethyloxazolidine-3, 4-dicarboxylate

2. To the flask, which contained compound 2 (10g, 45.6 mmol), was added acetone (165 mL),2, 2-dimethoxypropane (50 mL, 400 mmol) and boron trifluoride (0.35 mL, 2.8 mmol). The reaction mixture resulted in a dark orange-red solution, which indicated that a reaction had begun. The dark solution was then stirred for 2.5 hours to ensure that the reaction went to completion. After being stirred for 2.5 hours, the solution was analyzed through TLC analysis in 1:1 ethyl acetate: hexane. After the TLC analysis, the solution was charged with 99% triethylamine (0.9 mL). The solution lost its dark color and became a pale yellow color. The solution was placed on the rotary evaporator to remove the solvent. A thick syrup substance resulted from the rotary evaporator. Partitioning of the compound 3 was done between diethyl ether (150 mL) and saturated aqueous sodium bicarbonate solution (250 mL). The purpose of the partitioning was to retrieve the organic layer. The aqueous layer was washed with 2- 75 mL portions of diethyl ether. The organic layer was then dried with 2-5 scoops (depending on how proportionate the scoops are) of anhydrous sodium sulfate. The solution retrieved from the drying step was then placed on the rotary evaporator again to remove the solvent. The final product, compound 3, was placed under the vacuum overnight. The yield from Trial 1 could not be calculated for compound 3 because of error in the first reaction. The yield of compound 3 from Trial 2 was 115.24%. The yield of compound 3 from Trial 3 was 83.59%. The best yield for this reaction was from Trial 3.

N-[(1,1-Dimethylethoxy) carbonyl]-N,O- isopropylidene-L- serinol

3. To a flask was added tetrahydrofuan (50 mL) and lithium hydride (2.16 g, 57.0 mmol). The dropping funnel that was equipped to the flask was charged with oxazolidine methyl ester (9.90g, 38.2 mmol) and tetrahydrofuran (50 mL) and the contents within the dropping funnel were added dropwise to the solution in the flask over a period of twenty minutes. After twenty minutes the dropping funnel was washed with two 3-mL portions of tetrahydrofuran and the solution within the flask was stirred for an additional twenty minutes. A TLC analysis was done to note the complete formation of thealcohol, but the Rf value was not calculated. Four spots showed on the TLC plate in this analysis, and the alcohol was one of the two spots at the very bottom of the plate. The reaction mixture was placed in an ice-water bath and then additionally stirred for another hour. The product resulted in a white precipitate that was filtered through a celite pad by rinsing it with three- 30- mL portions of diethyl ether. The organic layer retrieved was then washed with 100 mL of aqueous phosphate buffer (pH 7). An extraction was done to remove the aqueous layer by washing the solution with diethyl ether (3*30mL). The compound 4 was then placed on the rotary evaporator to remove the solvent and was later placed under the vacuum. The percent yield of compound 4 was 125. 3% (11.60g), which differed from the expected yield range of 93-96%. The TLC and NMR analysis of compound 4 was too impure to continue to Reaction D. (Note 7)