OVERDUE FINES ARE 25¢ PER DAY _ PER ITEM Return to book drop to remove this checkout from your record. THE SYNTHESIS AND FORMATION OF ENOLATE ANIONS OF B-KETOSILANES By Chompoonute Kamthong A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1979 ABSTRACT THE SYNTHESIS AND FORMATION OF ENOLATE ANIONS OF B-KETOSILANES By Chompoonute Kamthong Trimethylsilylmethyllithium, which can be prepared from the reaction of tertrbutyllithium, tetramethylsilane and N,N,N',N'-tetramethyethylenediamine, reacts with a variety of aldehydes to give B-hydroxysilanes in fair to good yields. Oxidation of these compounds to their correspond- ing B-ketosilanes was accomplished by using chromium tri- oxide and pyridine in methylene chloride. Treatment of the B-ketosilanes with such bases as lithium diisopropylamide, potassium hydride and potassium bis- (trimethylsilyl)amide, under kinetically and equilibrium controlled conditions, followed by methylation, gave a mixture of products. After removal of the silyl grouping, the relative quantities of the methylated ketones in the mixture were determined by gas chromatographic analysis. In the case of l-trimethylsilyl-2-butanone, the silicon group was found to enhance the acidity of the adjacent hydrogens presumably by p-d delocalization of the negative Chompoonute Kamthong charge. This d—orbital stabilization effect of the silicon atom, however, proved to be less effective than conjuga— tive delocalization of the anion into the n system of a phenyl group, as noted for l-trimethylsilyl-3-phenyl—2- propanone. To My parents and brother, for so much love and support through so many years, without them this work would not have been possible. 11 ACKNOWLEDGMENTS The author wishes to thank her research preceptor, Dr. Michael w. Rathke for his guidance, patience, under- standing, and support throughout the course of this in- vestigation. The author wishes to thank Dr. William H. Reusch for serving as her first year advisor and her second reader, Dr. Carl H. Brubaker, Jr., for proofreading this Disserta- tion, and Dr. Jack B. Kinsinger for his friendship and interest in her well-being. Appreciation is also extended to Dr. Frederick H. Horne for his understanding and en- couragement. I wish to express my deepest thanks to my parents for their love, dedication, inspiration, and continuing support. To my brother, Pawares, I thank him for his love and en- couragement. I am indebted to many people for making much of this work possible. To Too, Aoy and Pom, thank you for every- thing you have done for me, and your love and confidence in me. To Mike, thank you for caring and being my friend. Thanks, also, go to all the other people at MSU that I have come to know. Your friendship has made my life away from home a pleasant one. I also wish to thank John Emswiler and Dang for their time and assistance in preparing this Dissertation. iii Appreciation is also extended to Tew, Barbara Duhl- Emswiler, Mrs. Deborah Wheaton and Mrs. Bernice Wallace for their kindness and assistance in times of need. Thanks also go to Peri—Anne Warstler for typing this manuscript. Finally, the financial support of Michigan State University and the National Science Foundation is grate- fully acknowledged. iv Chapter LIST OF TABLES. LIST OF FIGURES INTRODUCTION. CHAPTER I - THE SYNTHESIS OF B-KETOSILANES. . . . TABLE OF CONTENTS INTRODUCTION. . . . . . . . . . . . . . . . . REVIEW OF LITERATURE. . . . . . . . . . . . . RESULTS DISCUSSION. EXPERIMENTAL. . . . . . . . . . . . . . . . I. II. III. Materials . . . . . . . . . . . . . . Attempted Syntheses of B- Ketosilanes . . . . . . . . . . . . A. B. Modified Hauser's and Hance's MethOd O O O O O 0 O C O O O O O 0 Reactions of Bis(Trimethylsilyl)— Ketene C O O O O O O O C I O O C 0 Reaction of Ethyl Diethoxy- carbonium Fluoroborate with n-BUtyllithium. o o o o o o o o o Silylation of Cyclohexanone tert-BUDYliminC o o o o o o o o o Silylation and Hydrolysis of Ketone N,N—Dimethylhydrazones . . Reaction of Potassium tert- Butoxide with Tetramethylsilane . Reactions of Lithium Dialkyl- amides with Tetramethylsilane . . . Preparation of B-Hydroxysilanes . . . Page ix k-R’UUH ll 23 26 26 29 29 31 32 32 33 3h 3h 35 Chapter IV. v. Oxidation of B-Hydroxysilanes to B-KGDOSilaneS o o o I o o o o o o A. Moffatt Oxidation . . . . . . B. Modified Moffatt Oxidation. C. Oxidation of Alcohols to Ketones with Polymer Supported Reagent . . . . . . . . . . . D. Preparation of B-Ketosilanes Using Chromium Trioxide-Pyridine Complex in Methylene Chloride Solution. . . . . . . . . . . Product Analysis. . . . . . . . . CHAPTER II - FORMATION OF ENOLATE ANIONS OF B-KETOSILANES o o o o o o o o 0 INTRODUCTION. . . . . . . . . . . . . . REVIEW OF LITERATURE. . . . . . . . . . RESULTS DISCUSSION. . . . . . . . . . . . . . . EXPERIMENTAL. . . . . . . . . . . . . . . I. II. III. IV. Materials . . . . . . . . . . . . Generation of Enolates of the B-KGtOSilanes o o o o o o o o o o A. Preparation of Lithium Diiso- propylamide (LDA) . . . . . . B. Preparation of Lithio Tri- methylsilyl Ketone Enolates . Silylation of the B-Ketosilanes Alkylation of the B-Ketosilanes A. Methylation of B-Ketosilanes and NMR Analysis. . . . . . . vi Page 37 38 38 39 no an 1‘5 145 A7 59 67 67 69 69 69 70 70 7O Chapter VI. VII. VIII. Page B. Benzylation of B-Ketosilanes and NMR Analysis. . . . . . . . . . 71 Methylation of l-Trimethylsilyl- 3-phenyl-2-propanone: Kinetic vs. Equilibrium Conditions. . . . . . . . . 72 A. 10% Excess LDA and l-Trimethyl- silyl-3-phenyl-2-propanone Reactions . . . . . . . . . , , . . 72 B. LDA and 10% Excess l-Trimethyl- silyl-3-phenyl-2-propanone Reactions 0 o o o o a o o o o o o o 72 C. 10% Excess KH and l-Trimethyl- silyl-3-phenyl-2-propanone Reactions . . . . . . . . . . . . . 73 D. KH and 10% Excess l—Trimethyl- silyl-3-phenyl-2-propanone Reaction . . . . . . . . . . . . . 7” E. 10% Excess (Me381)2NK and 1- Trimethylsilyl—3-phenyl-2- propanone Reactions . . . . . . . . 7“ F. (Me3)2NK and 10% Excess l-Tri- methylsilyl-3-phenyl-2-propanone Reactions . . . . . . . . . . . . . 75 Methylation of Phenylacetone. . . . . . 75 A. 10% Excess LDA and Phenyl- acetone . . . . . . . . . . . . . . 75 B. LDA and 10% Excess Phenyl- acetone . . . . . . . . . . . . . . 76 Preparation of Authentic Sample of 2-Phenyl-3-pentanone. . . . . . . . . . 76 Methylation of l-Trimethylsilyl- silyl-2-butanone: Kinetic vs. Equilibrium Conditions . . . . . . . . 77 A. 10% Excess LDA and l-Tri- methylsilyl-2-butanone Reactions . . . . . . . . . . . . . 77 vii Chapter IX. LDA and 10% Excess l-Tri- methylsilyl-2—butanone Reactions 10% Excess KH and l-Tri- methylsilyl-Z-butanone Reactions KH and 10% Excess l-Tri- methylsilyl-2-butanone Reactions Ketone. . A. B. X. Preparation of Authentic Samples of Alkylated Methyl Ethyl Ketone. A. B. XI. BIBLIOGRAPHY. 10% Excess KH and Methyl Ethyl Ketone. KH and 10% Excess Methyl Ethyl Ketone. Methyl Isopropyl Ketone Ethyl Isopropyl Ketone Methylation of Methyl Ethyl and Diisopropyl Ketone. Product Analysis. viii Page 78 79 79 80 80 81 81 81 81 82 8A Table II III IV VI LIST OF TABLES Hydrolysis of Silyl Ketone N,N— Dimethylhydrazones. . . . . Percent Yields of B-Hydroxysilanes. Reactions of l-Trimethylsilyl-B— phenyl-2-propanone with Various Bases . . . . . . Reactions of Phenylacetone with Lithium Diisopropylamide. Reactions of l-Trimethylsilyl-2- butanone with Selected Bases. Reactions of 2-Butanone with HR ix Page 15 20 53 55 56 58 LIST OF FIGURES Figure Page 1 Reaction Apparatus. . . . . . . . . . . . . 30 2 Reaction Apparatus. . . . . . . . . . . . . 36 INTRODUCTION In recent years, use of silicon derivatives, par- ticularly those derived from carbonyl compounds, to effect various synthetic transformations has increased enormously. A number of organosilicon compounds containing a carbonyl functional grouping have been synthesized and their chemi— cal and spectral properties have been investigated. The ability of silyl groups to stabilize electron pairs on 1 on silicon, neighboring atoms is affected by substituents and decreases with increasing distance between the sub- stituent and the silicon atom.2 As a second-row element, silicon has empty d-orbitals in the valence shell which may overlap with p-orbitals on any attached atom or group resulting in a (p+d) n-bond.3 Delocalization of the p— electrons on neighboring groups into the silicon d-orbitals results in stabilization of the delocalized electrons. This interaction is believed to enhance the acidity of silanes through stabilization of carbanions alpha to silicon.“ Many such carbanions,5 (R381)nH3_nC- have been prepared by several synthetic methods, such as re- action of a-halosilanes with magnesium6’7 or lithium, halogen-metal exchange reactions between a—halosilanes 8 and alkylithiums, addition of alkyllithiums to vinyl- silanes,9 and metalation of silanes with n-butyllithium- tetramethylethylenediamine complex.10 In order to determine the stabilizing effects of the silicon on an adjacent carbanion, enolates of silylated ketones were examined, and the acidity of the alpha hydrogens in such compounds, and in analogous normal ketones were compared. Thus, B-ketosilanes of the type: H O H Iéug3 ,31-'-c_ —R; R = alkyl or aryl I I were synthesized and experiments were designed to de— termine the relative acidity of alpha hydrogens on the two sides of the carbonyl group. CHAPTER I THE SYNTHESIS OF B-KETOSILANES INTRODUCTION The synthesis of B-ketosilanes has been accomplished by several methods. Most of these methods require con- siderable care in product isolation due to the instability of these compounds. B-Ketosilanes are found to be very sensitive to hydrolytic agents,11a readily undergo cleavage of the 81-0 bond by electrophilic and nucleo- philic reagents12 including water and alcohols at room temperature, and decompose during distillation at atmos- pheric pressure.l3 Their facile decomposition was also noted in several infrared and ultraviolet studies. Re- arrangement of a-silylated ketones to isomeric silyl enol ethers occurs during heating or treatment with HgIe, HgBr2, ZnCl2 in ether, R3SiI, or R3SiBr.1u Previous efforts to prepare these compounds are re- viewed in this chapter together with our attempts to apply such methods to the synthesis of B-ketosilanes. Ulti- mately, we found the most convenient route to the B-keto- silanes was oxidation of B-hydroxysilanes. REVIEW OF LITERATURE Most carbonyl compounds with reactive hydrogen in the alpha position will react with strong bases to form resonance stabilized enolate anions. Such enolates react with silicon halides to give either carbon or oxygen silylated products. Hauser and Hance15 reported only C- silyl esters from the reaction of sodio ethyl acetate or sodio ethyl isobutyrate with trimethylchlorosilane. Later Rathke and Sullivan16 found that C-silylation of ester enolates occurs only with acetate esters. Other esters gave predominantly O-silylated products. Carbon silylated products have never been found with ketone enolates. Gilman and Clark17 were the first to report the reaction of triethylchlorosilane with sodio acetoacetic ester enolates. The chemical properties of the product suggested the O-silylated enol structure, I 18 Infrared analysis and this view was confirmed by West. showed that the structure of the trimethylsilyl deriva- tives of ethyl acetoacetate and of acetylacetone corresponds to I rather than g. 0 CH3 0 Si(CH3)3 Z-g-CH=C-O-Si(CH ) Z-g-CH C'O' Z - OH OCH CH 33 '3" "3" 23 CH3 l a Later, Kruger and Rochowl9 used GLC, IR and NMR to show that, in all cases studied, halosilanes react with enolate anions of ketones to produce the corresponding O-silylated enols. That substitution at the oxygen atom takes precedence over substitution at the carbon atom may be due to the relatively high bond energy of the 81-0 bond (108 kcal/mole). -In the case of ester enolates, the alkoxy oxygen competes in the delocalization of electrons as shown below: (O :6:- . u u a u + -C-C-O-R « -C-C=O—R Silylation at carbon is therefore more favorable than in ketone enolates. Since a simple silylation of ketone enolates results in the oxygen-silylated rather than carbon—silylated product, other methods of synthesis of B-ketosilanes were lla attempted required. In l9h7, Whitmore and co-workers synthesis of trimethylsilylacetone from acetyl chloride and the Grignard reagent (CH3)3SiCH2MgCl,11b without success. The resulting yellow solid, on decomposition with water, formed a variety of products including ace— tone. In 1952 Hauser and Hance7 prepared trimethylsilyl- acetone, in 59% yield, from the reaction of 3 and acetic anhydride at -70°C, followed by careful hydrolysis with aqueous ammonium chloride at -lO°C (Eq. 1). o " l)(CH2C)20, -70° 9 (CH3)3SiCH2MgCl e: (CH3)381CH2CCH3 (1) 2)NHuCl, H2O, -1o° ,3 This procedure was used subsequently by several others for the synthesis of a-silylated ketones. However, Krfiger and Rochow reported19 failure to achieve this synthesis. During a study of epoxyethylsilanes Eisch and Trainor2O obtained triphenylsilyacetaldehyde from epoxyethyltriphenyl- silane and magnesium bromide (Eq. 2). ether -—-> ¢3SiCH2CHO (2) -c + ¢BSiCH‘/H2 MgBr2 O No attempts have been made to synthesize B-ketosilanes in a similar manner. Triphenylsilylacetophenone was prepared by Brook and Pierce21 for chemical and spectral studies. A silyl alcohol22 from the reaction of styrene oxide and triphenyl- silylithium was subjected to oxidation under rather severe conditions, according to Eq. 3. ,}K ?H 96% H280“, CrO3/H20 ¢BSiLi+¢-CH-CH2+¢381CH2CH¢ i; r.t. lh (or NBS/col; with pyridine or CaCO3)1b o 3 e381CH2 ¢ (3) 23 When acylsilanes were treated with diazomethane, the major, if not exclusive, products were B-ketosilaneszu (Scheme I). 92' 0 ' 0 [R381- V? -'R J-9R3SiCH2CR (I) 0° t / Wm B R SiCR'+CH Ng—>R3 S-i C- R' 3 2 ether CH UN;\\\\‘ 2 ,OSiR3 ' ”331/17- R ]->CH2 C\R' CH 2 .5. Scheme I The mild conditions of this reaction did not induce iso- merization of the ketones to the enols and the stereo- chemistry of silicon is retained during the formation of both £_and 2.1Uc In 1970 it was reported9 that the reaction of tri- methylsilylmethyl magnesium chloride, 3, with acid chloride affords, after hydrolysis with dilute hydrochloric acid, methyl ketones in reasonable yields (50-70%) (Eq. H). 0 HCl 0 II .. g ll RCCl + R SiCH ——~R -C—SiR -———>RCCH (A) 3 2 . 3 3 Furthermore, a B-silyl carbinol which is a possible pre- cursor of B—ketosilane could be obtained in good yield if». by refluxing ;_with a ketone, followed by work up (Eq. 5). R 1) E:0=o R2 R I (CH3)3SiCH2MgCl -——2———~ Rl-C-CH2Si(CH3)3 (5) 2) H20 OH Hudrlik and Peterson25 devised two interesting routes to S-trimethylsilyl-U-octanone,8_(Scheme II). 1) (0001)2 l) EtLi M8331 CO2H 2) PrZCuLi 2) CO2 03H7 .5. Me Si Me Si 3 “§\ 3 C3“? o\o 3H7 [CD 1) EtLi Me3Si OH ~~———————--.~» >—< m3 CHO C3H7 C3H7 pyridine [N Scheme II The sequence proceeding via carboxylic acid 6 gives a 60% yield of 8 whereas the other gives 8 in 68% yield. A number of other oxidation methods were applied to the 10 alcohol 1, but were unsuccessful. The oxidation procedure shown in Scheme II was used successfully in the prepara- tion of trimethylsilylmethyl cyclohexyl ketone 2 from the corresponding silyl alcohol.26 . Si(CH3)3 An entirely different approach was developed by Corey and co-workers.27 Their approach, as shown in Scheme III, recognizes that N,N-dimethylhydrazones (DMH's), as well as imines, are protected forms of aldehydes and ketones, from which the free carbonyl compounds can be regenerated. A ketone N,N-dimethylhydrazone or imine was prepared by standard techniques.28’ul’u2 Metalation was accomplished by using lithium diisopropylamide (LDA) in tetrahydro— furan at 0°C, followed by reaction with trimethylchloro- silane.27b The corresponding a-silylated ketone was obtained from the N,N—dimethylhydrazone in high yield by oxidative hydrolysis by aqueous sodium periodate at pH 7 and 20-25°C,27a or by a cupric ion—catalyzed hydrolysis.27c We attempted to prepare B-ketosilanes by several meth— ods, including modifications of the procedures described in this chapter. 11 I §NM92 1) LDA, 0°, THF . ENEez c- - -C-C-CH- . .:r- -SiMe I l 2) Me3SiCl, 3.5h ' 3) H20 3 pH N7, NaIOu ++ or pH m8, Cu I ‘3 II -C-C-C-SiMe I I 3 Scheme III RESULTS By using the procedure of Hauser and Hance,7 With slight modifications, tetramethylsilane (TMS) was treated with tert-butyllithium in the presence of N,N,N',N'-tetra— methylethylenediamine (TMEDA) (Eq. 6). hexane (CH3)uSi + tert—BuLi + TMEDA —t:(CH3)3SiCH2Li 25°, 15min. (6) The resulting trimethylsilylmethyllithium was allowed to react further with N-acetylimidazole (Eq. 7) and acetic anhydride (Eq. 8), but the desired product, trimethyl- silylacetone, could not be detected even after a careful N Oflom mpcmw< ofipmaomcmm .mosowmsvzgamguoefiolz.z mCOpmx Hmafim mo mammflopomm .H mflpme 16 In connection with our previous attempts with tetra- methylsilane, it seemed that the highest probability of success would be first to synthesize B-hydroxysilanes by the reaction of 11 with appropriate aldehydes (Eq. 13). . 11* 9H (CH3)BSiCH2M + RCHO-—+~» (CH3)381CHZCHR (13) 11 A survey was taken to find bases and conditions which would efficiently generate 11_from tetramethylsilane. It has been reported10b that tetramethylsilane undergoes metala— tion by an equivalent amount of tegtebutyllithium at 25°C in the presence of 0.25 equivalent of N,N,N',N'-tetra- methylethylenediamine to give “0% trimethylsilylmethyl- lithium after A days. These investigators did not note the effect of using only unactivated Eggtrbutyllithium, n-butyllithium or alkylpotassium. We attempted synthesis of trimethylsily1methy1potassium by a metal exchange re- action of n-butylithium and potassium tertfbutoxide (Eqs. 1“ and 15). n-BuLi + KO+ ++OL1 + n-BuK (11)) .+ - + n-BuK + (CH3)uSi (CH3)3SiCH2K (15) 17 We found that adding propionaldehyde to a solution of n- butyllithium and tetramethylsilane in the presence of potassium tert-butoxide gave no silyl alcohol. The only product shown by GLC was 3—heptanol (Eq. 16). l)5 equiv. n-BuLi, 1 equiv. KO-+3 OH ( ) hexane or THF, -78° ' 5 CH Si 3 C H CHCH CH 3 u 2) CH3CH2CHO, -78° u 9 2 3 3) H30+ (33’501) (l6) Varying the temperature from —78° to 0°C or varying re- action times did not lead to significant changes. When an equivalent amount of potassium Eggtybutoxide was em- ployed, there appeared to be some traces of 1-trimethy1- silyl-2-butanol present. Likewise, only 3-heptanol was obtained from reaction such as Eq. 17. I (CH3)uSi hexane’ r't" 1 h; Cu39CHCH2CH (17) 2)CH3CH2CHO 3 3)H30+ R - Et, i-Pr A successful synthesis of 11 was achieved by using N,N,N',N'-tetramethylethylenediamine (TMEDA) as an ac- dIOa tivating ligand. It has been reporte that protons 18 alpha to silicon undergo metalation in 3 days with equi— valent quantities of n—butyllithium and TMEDA. Trimethyl- silylmethyllithium was prepared from the reaction of £222? butyllithium, tetramethylsilane and TMEDA at room tempera- ture. Treatment of the resulting mixture with aldehydes, followed by hydrolysis gave B-hydroxysilanes in reasonable yields (Eq. 18). hexane tert—BuLi + (CH3)uSi + TMEDA : Me3SiCH‘2’Li+ r.t., 12h 1)RCHO, -78° 9H :Me3SiCH2CHR (18) 2)H3o+, 0° A number of experiments were performed to obtain optimal yields of the silyl alcohols. The reaction requires a 100% excess of tetramethylsilane and the presence of less than a full equivalent of TMEDA (Eq. 19). hexane tert-BuLi + (CH3)uSi + TMEDA ve*1_ r.t., 12h 1 mmol 2 mmol 3/5 mmol 1)RCHO, -780 1 mmol PH *—> Me3SiCH2CHR (19) 2)NHuC1, H2O, 0° 19 The reaction of aldehydes with trimethylsilylmethyllithium at -78°C is very fast. We found that for good yields, after warming to 0°C the reaction mixture must be hydrolyzed immediately with 10% ammonium chloride solution. The yields of silyl alcohols were based on Eggtfbutyllithium and aldehydes, and are shown in Table II. A varietycfi'methods for oxidizing secondary alcohols to ketones are available. The reaction of alcohols with dimethyl sulfoxide (DMSO) and dicyclohexylcarbodiimide (DCC) in the presence of certain acids (Moffatt OxidationBO) gives the corresponding aldehydes or ketones, under ex- tremely mild conditions. However, this oxidation method did not seem to be particularly successful in the case of B—hydroxysilanes. An attempt was also made to oxidize the silyl alcohols by applying a modified Moffatt oxida- tion,31 using the water soluble l-cyclohexyl-B-(2-morpho- linoethyl)carbodiimide methoep-toluenesulfonate (CMCM) instead of DCC. Both methods failed to oxidize silyl alcohols to silyl ketones under the reaction conditions reported. Cainelli and co-workers32 reported a polymer supported reagent that is very useful for the oxidation of aochols. The HCrOH form of Amberlyst A-26 was prepared according to 32 The activity of the resin so the published procedure. obtained was determined by oxidation of some sample alcohols in refluxing tetrahydrofuran. It proved to be 20 Table II. Percent Yields of B-Hydroxysilanes. Silyl Alcohols % GLC Yield % Isolated Yield OH Me3SiCH2CHCH3 80 50 OH Me3SiCH2CHCH2CH3 88 58 OH Me3SiCH2CHCH2CH(CH3)2 85 undetermined OH Me3SiCH2CHCH2¢ undetermined HS 21 remarkably effective in oxidizing benzyl alcohol and cyclo- hexanol to benzaldehyde and cyclohexanone in 1 and 3 hours respectively. However, when this procedure was applied to B-hydroxysilanes, only starting materials were detected by GLC. The most satisfactory results were reported by Hudrlik and Peterson.25b The oxidation of silyl alcohols to 8- ketosilanes was achieved by using chromium trioxide and pyridine in methylene chloride solution (Eq. 20). 9H Cr03, pyridine 9 (CH ) SiCH CHCH R ;:(CH ) SiCH CCH R (20) 3 3 2 2 3 3 2 2 0132012 25b.26.33 A few papers reported this oxidation with slightly different quantities, reaction conditions and work up techniques. We found that fairly good yields could be obtained by using less than 2 equivalents of pyridine for each equivalent of chromium trioxide in a large volume of methylene chloride.26 Approximately one sixth equivalent of silyl alcohol diluted in methylene chloride was added to the chromium trioxide-pyridine complex mixture. This general procedure was similar to that described in the 25b The chief drawbacks are the need published report. for a large excess of the oxidizing agents, the handling of large volume of extremely dilute reaction mixture, and 22 the removal of a black tar from the products. Moreover, we discovered another problem during the GLC analysis of products and yields. Although an isolated sample of a silyl ketone showed an NMR spectrum corresponding to pure ketone, when the sample was injected on the gas chroma- tograph a number of peaks were observed. We considered this to be due to decomposition and/or rearrangement of the B-ketosilanes in the gas chromatograph. This, there- fore, created some difficulties in analyzing separations and purifications of our products. Evidently this problem has not been reported in the literature.13 Attempts were made to use different columns, such as carbowax, silicon fluid GE XF-1150 (50% nitrile), diethylene glycol suc- cinate (DEGS), Apiezon and OV-lOl, in order to improve the GLC analysis of the B-ketosilanes, but were unsuccessful. The following B-ketosilanes were obtained by oxidation of the corresponding silyl alcohols using the above pro- cedure: 1. 1-Trimethylsilyl-2-propanone CH3CCH2Si(CH3)3 O 2. 1-Trimethylsily1-2—butanone CH3CH280H281(CH3)3 fl 3. l-Trimethylsilyl—3-pheny1—2-propanone ¢CH2CCHZSi(CH3)3 The yields of isolated B-ketosilanes were approximately 50%. 23 All products exhibited spectral data in accordance with assigned structures. DISCUSSION The fact that certain B-ketosilanes, synthesized in our laboratory, could not be analyzed by GLC caused some difficulties in analysis. During distillation of the final solutions from the chromium trioxide/pyridine oxida- tion of B-hydroxysilanes to B-ketosilanes, samples of the fractions were taken periodically and checked for purity by NMR spectroscopy. However, had we discovered this problem with the GLC analysis of these silyl ketones earlier, we may have found that many attempted syntheses we had considered failures were actually successful. Attempts to prepare B-hydroxysilanes using n-butyl- lithium and potassium tertybutoxide or dialkylamine as bases, Eqs. 16 and 17, failed to metalate tetramethyl- silane. After the failure of the above bases, 33337 butyllithium was tried. It was found, after several trials, that the ratios of tggtybutyllithium, tetramethylsilane and N,N,N',N'-tetramethylethylenediamine, and the reaction conditions were critical to the success of this reaction. Therefore, n—butyllithium and potassium Eggtfbutoxide or dialkylamine might also work under different conditions. Despite our success in the oxidation of the 2U B-hydroxysilanes to B-ketosilanes with Collins reagent, (CSHSN)ZCrO3’ the procedure was considered inconvenient primarily because of the very large excess (customary five- or six-fold) of Collins reagent was used, and the tedious work up and purification of the product. Thus a need still exists for more practical methods. Pyridinium chlorochromate :PCC), CSHSNH+C10rO§, was developed by Corey 3 and co-workers and available commercially from Aldrich. It has become widely used for the oxidation of alcohols to carbonyl compounds. It can be easily and safely prepared by the addition of pyridine to a solution of chromium tri- oxide in 6M hydrochloric acid followed by filtration to obtain the yellow-orange, air stable solid. PCC allows an efficient oxidation in methylene chloride with only 50% excess of oxidant despite the fact that aqueous chloro- chromate is not a very effective oxidizing agent. Yields of aldehydes and ketones obtained with 1.5 molar equiva- lents are reportedly equal to or greater than those obtained with the Collins reagent. However, the mildly acidic character of FCC precludes its use with acid sensitive substrates or products. For cases in which a more neutral reagent is necessary, pyridinium dichromate (PDC), (C5H5NH+)2Cr207- has been shown by Corey and Schmidt35 to be of wide applicability for the oxidation of alcohols to aldehydes, ketones, and, in some cases, carboxylic acids. PDC, a stable solid which can be simply prepared 25 in quantity by dissolving chromium trioxide in a minimum of water, adding pyridine and collecting the precipitated product, is now commercially available. Corey found that the reagent PDC in methylene chloride oxidizes primary alcohols to the corresponding aldehydes and no further, regardless of the nature of the substrate. When DMF was used as solvent, primary alcohols are converted directly to carboxylic acid with no affect on acid- or base-sensi- tive functionality. The oxidation of secondary alcohols such as h-t-butylcyclohexanol can be effected by PDC-DMF (9h% yield) or PDC-CH2C12 (97% yield). Although the literature did not include applications of either PCC or PDC to silyl compounds, on the basis of the results reported thus far, these reagents obviously qualify as potentially better oxidizing reagents for the oxidation of B-hydroxysilanes to the corresponding B-keto- silanes. 26 EXPERIMENTAL 1. Materials Silanes Commercial tetramethylsilane (TMS) was used without purification and stored in a refrigerator. Trimethyl- chlorosilane was freshly distilled and stored under argon. N,N,N',N'-Tetramethylethylenediamine (TMEDA) TMEDA was obtained commercially from Aldrich. Prior to use it was distilled and stored under argon. Organic Bases All of the organolithium bases, amines and potassium tertybutoxide were commercially available. Potassium tertebutoxide was purified by sublimation under vacuum. Methyl magnesium Grignard was prepared from methyl iodide and magnesium metal by the procedure described in the literature.36’37 Methyl Iodide Commercial methyl iodide was distilled and stored over a copper wire in a brown bottle with a septum inlet 27 in a refrigerator. Bis(trimethylsilyl)ketene Bis(trimethylsilyl)ketene was obtained from Rathke's and co—workers' previous work.29 Carbonyl Compounds N-Acetylimidazole was obtained from S.L. Hartzell's 38 Acetic anhydride and all the work in our laboratory. aldehydes and ketones were commercially available and were purified by distillation prior to use. Ethyl Diethoxycarbonium Fluoroborate Ethyl diethoxycarbonium fluoroborate was prepared from triethyl orthopropionate and boron trifluoride etherate according to Meerwein and co-workers.39 Triethyl ortho- propionate was prepared from the commercially available propionitrile, absolute ethanol and dry hydrogen chloride gas (Matheson) as described by S. M. McElvain and J. W. Nelson.“O Imine Cyclohexanene tert-butylimine was prepared from cyclo- hexanone tert-butylamine and titanium tetrachloride by the procedure of H. Weingarten and co-workers."'1 The 28 tert-butylamine and titanium tetrachloride were commer- cially available. N,N-Dimethylhydrazones (DMH) The ketone DMH's were prepared from the corresponding ketones and unsym-dimethylhydrazine in absolute ethanol following the procedure of G. R. Newkome and D. L. Fishel.“2 The hydrazine was obtained commercially. Carbodiimides The dicyclohexylcarbodiimide (DCC) and the l-cyclo- hexyl—3-(2—morpholinoethyl) carbodiimide metho-p—toluene sulfonate (CMCM) were purchased from Aldrich and used without further purification. Amberlyst A-26 The chloride form of Amberlyst A—26 was obtained from Ventron. Chromium Trioxide The commercial chromium trioxide was placed in a desiccator over phosphorus pentaoxide and evacuated for at least eight hours prior to use. 29 29.132122. Tetrahydrofuran was dried over sodium benzophenone ketyl, distilled and stored under argon over molecular sieves. Anhydrous diethyl ether was dried over lithium aluminum hydride, distilled and stored under argon over molecular sieves. All alkanes, methylene chloride, DMSO and pyridine were dried over molecular sieves prior to 1188. II. Attempted Syntheses of B-Ketosilanes A. Modified Hauser's and Hance's Method7 A dry 50 m1 round-bottomed flask equipped as shown in Figure 1 was flushed with nitrogen and immersed in an ice-water bath. The flask was charged with “.75 ml (10 mmol)cn'2.lM_Eggtrbutyllithium in pentane and 5 m1 of hexane was added. TMS (1.59 ml, 12 mmol) was injected followed by dropwise addition of TMEDA (1.5 ml, 5 mmol) with vigorous stirring. The reaction mixture turned cloudy yel- low and was stirred 10 minutes at 0°C, then allowed to warm up to room temperature and stirred an additional 15 minutes as it became clear. The flask was then cooled to -78°C and 1.1 g (10 mmol) N-acetylimidazole dissolved in 5 m1 THF, was quickly added through a funnel. The dry ice- acetone bath was replaced by ice-water bath and the mix- ture was stirred for 15 minutes. After removing the 30 TO ] -—-- MERCURY BUBBLER GAS ——" INLET VALVE *— RUBBER SEPTUM I MAGNETIC STIRRER I Figure 1. Reaction Apparatus 31 ice-bath, 5 m1 of saturated NHuCl solution and 5 m1 of hexane were added. The hexane layer was separated and washed with water and 10% NaHCO3, then dried over Nazsou. In the experiment with acetic anhydride, the (CH3)3SiCH2Li formed from the reaction of tergrbutyllithium, TMS and TMEDA was transferred by a syringe to the flask containing acetic anhydride in ether cooled to -78°C under nitrogen. The reaction mixture was stirred for 3.5 hours at -78°C. The dry ice-acetone bath was removed and saturated NHuCl solu- tion and hexane were added. The hexane layer was separated and washed as above. B. Reactions of Bis(Trimethylsi1y1)Ketene The following procedure for the reaction of bis(tri- methylsily1)ketene with methyllithium is representative of a number of reactions using bis(trimethylsilyl)ketene. To 1 ml (1 mmol) of methyllithium (0.916M in ether) in a 10 m1 pear-shaped flask with side arm, 0.2 ml (1 mmol) of bis(trimethylsilyl) ketene was added. After stirring for 2 hours at room temperature, the reaction mixture was quenched with 1 ml of acetic acid (1M_in hexane) at 0°C and filtered. GLC analysis (2.5% SE-30) of the neat product mixture established the presence of bis(trimethylsilyl) acetone and l-trimethylsily1-2-trimethylsiloxypropene. 32 C. Reaction of Ethyl Diethoxycarbonium Fluoroborate with n—Butyllithium A dry 50 m1 round-bottomed flask as shown in Figure 1 was flushed with nitrogen and immersed in an ice-water bath. Ethyl diethoxycarbonium fluoroborate (2u.9 mmol) was freshly prepared by Meerwein's method39 in the flask. To this 6 ml pentane and n-butyllithium (12.“ ml of 2.0M solution in hexane, 2U.9 mmol) were added slowly with stirring. After 20 minutes the contents of the flask were quenched with 10 m1 of 1% HBFu and warmed up to room temperature. The organic phase was separated, washed with 10% Na2CO3 and saturated Na2CO3, and dried over K2C03. No traces of 3-heptanone were detected by GLC analysis. D. Silylation of Cyclohexane tert-Butylimine The enolates of cyclohexanone Eggtrbutylimine were prepared by treatment of the imine with the following bases: lithium diisopropylamide, n—butyllithium, n-butyl- lithium/TMEDA, sec-butyllithium/TMEDA and tertrbutyl- lithium/TMEDA. The following procedure is representative. A solution of cyclohexane tertybutylimine (2 ml, 15 mmol) in 25 ml THF was cooled to -78°C and n-butyllithium (9.fl m1 of 1.6M solution in hexane, 15 mmol) and 2.3 ml (15 mmol) TMEDA were added. The flask was warmed up to 25°C then was immersed again in the dry ice-acetone bath 33 with addition of trimethylchlorosilane. The reaction mix- ture was poured into 50 m1 of phosphate buffer solution (pH 7). The organic layer was dried over MgSOu and con- centrated. GLC analysis (5% carbowax) indicated the product was cyclohexane. E. Silylation and Hydrolysis of Ketone N,N-Dimethyl- hydrazones The following procedure for the silylation and hydrolysis of methyl isobutyl ketone DMH is representative. n-Butyl- lithium (3.1 m1 of 1.6M_solution in hexane, 5 mmol) was added to a solution of 0.8 m1 (5 mmol) ketone DMH and 5 ml THF at -78°C. After 20 minutes of stirring and cooling, 0.6 ml of trimethylsilylchloride (5 mmol) was added. The dry ice-acetone bath was removed and reaction flask stirred for 15 minutes. GLC analysis (2.5% SE-30) using tridecane as internal standard showed a 70% yield of silylated ketone DMH. Various hydrolytic agents were used in the attempts to convert the silyl ketone DMH's to silyl ketones. The follow- ing procedure is representative. To a 0.2 ml (1 mmol) sample of the silyl ketone DMH dissolved in 12 m1 of THF, 3 ml of 1§_phosphate buffer solution (pH 7) and a solution of 0.H7 g NaIOu in 5 ml of water was added. The mixture was stirred for 3 hours at room temperature, filtered, extracted with water and methylene chloride and dried over Nazsou. The product that appeared in GLC (2.5% SE—30) 3“ analysis was shown by NMR to be methyl isobutyl ketone DMH. ' F. Reaction of Potassium tert-Butoxide with Tetra- methylsilane Potassium Eggtybutoxide (0.1 g, 1 mmol) was placed in a 25 m1 dry round-bottomed flask equipped as in Figure 1 which then was evacuated for 0.5 hour to insure its dryness. Following the addition of 5 ml of hexane, the flask was cooled in a dry ice-acetone bath and 3.2 ml (5 mmol) of 1.57M of n-butyllithium in hexane was added. Although potassium Egggrbutoxide did not appear to dis- solve, 0.7 ml (5 mmol) of TMS was added and the reaction mixture was allowed to stir at -78° or 0°C with the dura- tion times varied from 15 minutes to 2“ hours. Then o.u m1 (5 mmol) of propionaldehyde was added, followed im- mediately by quenching with water at -78° or 0°C. GLC (2.5% SE-30) analysis of the organic phase showed the presence of 3-heptano1 in 33-50% yield in both cases. G. Reactions of Lithium Dialkylgmides with Tetra— methylsilane Lithium dialkylimide was prepared from n-butyllithium and diisopropylamine or diethylamine. The following pro- cedure is representative. 35 A 25 ml flask equipped with magnetic stirring, septum inlet and mercury bubbler as in Figure l was flame dried under a stream of dry argon. To this flask was added 3.1 ml (5 mmol) of 1.6M_solution of n-butyllithium in hexane and 5 m1 of additional hexane. The flask was cooled in an ice bath and 0.7 ml (5 mmol) of diisopropyl- amine was added dropwise while stirring. The ice bath was removed and after 10 minutes of stirring at room temperature, TMS (0.7 ml, 5.5 mmol) was added. The mixture was stirred for 1 hour at room temperature then cooled at -78°C. After cooling 0.“ ml (5 mmol) of propionaldehyde was added followed immediately by quenching with water. GLC analysis (2.5% SE-30) of the separated and dried (NaZSOu) organic phase established the presence of 3- heptanol. III. Preparation of B-Hydroxysilanes The following procedure for the preparation of 1- trimethylsilyl-2-butanol is representative. A 1000-liter three-neck round-bottomed flask was fitted with a mechanical stirrer, a rubber septum over a glass bushing and a gas inlet tube connected to a mercury bubbler as shown in Figure 2. The system was oven-dried and assembled while warm and flushed with argon. The flask was charged with 2A0 ml of hexane and 215.06 ml (ADO mmol) of 1.86M.solu- tion of tart-butyllithium in pentane and cooled in an 36 .—— TO MECHANICAL STIRRER T0 —-- MERCURY BUBBLER RUBBER —D C -) seen... I Figure 2. Reaction Apparatus 37 ice-water bath. TMS (108.9 m1, 800 mmol) was added followed by slow addition of 36.2 ml (2u0 mmol) of TMEDA while ' stirring vigorously. After 15 minutes, the ice bath was removed and the reaction mixture was stirred over a period _ of 12 hours at room temperature. Then the flask was cooled to -78°C and 28.9 ml (#00 mmol), of propionaldehyde was added slowly. The dry ice-acetone bath was replaced by an ice-water bath and about 100 ml of 10% NHuCl solution was immediately added to the contents of the flask. The hexane extract was washed with water, 10% NaHCO3 solution and dried over Nazsou. The product was concentrated and distilled under reduced pressure to obtain 16.9 g, 58% of colorless liquid, 1-trimethylsily1-2-butanol, bp h5°C (2.5 mm), n25 D 0.935. IV. Oxidation of B-HydroxySilanes to B-Ketosilanes A. Moffatt Oxidation30 To a solution of 0.16 ml (2 mmol) of pyridine, 0.1 ml (1 mmol) of trifluoroacetic acid in 3 ml of benzene, were added 3 ml of DMSO and 0.31 ml (2 mmol) of l-trimethyl- silyl-2-butano1. Dicyclohexylcarbodiimide (1.2 g, 6 mmol), was added quickly through a funnel and the mixture was stirred overnight under nitrogen at room temperature. Ether (50 ml) was then added followed by a solution of oxalic acid (0.5 g, 6 mmol) in methanol (5 ml). After 38 stirring for 30 minutes, water (50 ml) was added and the insoluble DCC was removed by filtration. The organic phase was then extracted twice with 5% NaHCO3 solution and once with water, dried over NaZSOu and concentrated. Only the starting silyl alcohol was recovered as shown by GLC analysis (2.5%. SE-30). B. Modified Moffatt Oxidation31 1-Trimethylsilyl-2-butanol (0.6 g, 3.8 mmol) was dis— solved in 20 ml of benzene (dried over Na wire) and 20 m1 of DMSO, and cooled to u°C. Then pyridine (0.“? ml), trifluoroacetic acid (6 ml), and 1.5 g of l-cyclohexyl-3- (2-morpholinoethyl) carbodiimide metho-p-toluenesulfonate (CMCM) were added in this order. After stirring 2A hours at A°C, the mixture was poured into ice water and extracted with ether. The ether layer was reshaken with ice water, dried (MgSOu), and concentrated. GLC analysis (2.5%, SE-30) showed only one peak which was the unoxidized alcohol. C. Oxidation of Alcohols to Ketones with Polymer . Supported Reagent The HCrO; form of Amberlyst A-26 was prepared by treat- ment of its chloride form with aqueous solution of chromic acid32 and dried at 50°C under vacuum for 5 hours. One 39 mmole of alcohol was refluxed with 3.5 g of the anion resins in 7.5 ml of THF over periods of l to 9 hours. A sample was taken and analyzed by GLC (2.5% SE-30) at the end of each hour and finally the resin was removed by filtra- tion. When the alcohols were benzyl alcohol and cyclohexanol, the sole products at the end of l and 3 hours were benz— aldehyde and cyclohexanone respectively. However, when the same experiment was applied to l-trimethylsilyl-3- phenyl-3-propanol, GLC analysis shows only the starting silyl alcohol. D. Preparation of B-Ketosilanes Using Chromium Tri- oxide—Pyridine Complex in Methylene Chloride Solu- tion The following procedure for the oxidation of l-trimethyl- silyl-3-pheny1-2-propanol to l—trimethylsilyl-3-phenyl-2- propanone is representative. In a 5 liter three-neck round- bottomed flask fitted with a mechanical stirrer, septum inlet and a mercury bubbler as shown in Figure 2 were placed 166.56 g (1.67 mol) of CrO3. Methylene chloride (3.6 l) was added through a funnel, followed by cooling to 0°C and a dropwise addition of pyridine (252.A8 ml, 3.2 mol). The deep burgundy colored solution was stirred for 1.25 hours at room temperature, then a solution of 53 ml (285.6 mmol) of 1-trimethylsilyl-3-phenyl-2-propanol in about 100 m1 of CH2C12 was slowly added. At this time the mixture must be A0 stirred vigorously as black tar was formed. After 15 minutes, it was filtered through a fritted (medium) funnel, covered with about 0.25 inch of Celite. The filtrate was poured into 10% NHuC1 overlaid with enough ether to result in 2 layers of solution; the ether layer was washed 2 additional times with 10% NHuC1, once or twice with 2M HCl to remove pyridine, and finally, 3 times with saturated NaHCO3. The ether solution was dried over MgSOu, concen- trated and subjected to vacuum distillation yielding about 50% of l-trimethylsily1-3—phenyl—2-propanone, bp 60°-65°C (0.01 mm), h25 D 1.0533. V. Product Analysis All GLC analyses were performed on a Varian 920 Chromatograph. The 2.5% SE-30 on Chromosorb G (Non Acid Wash) was packed into a l/A inch by 6 foot stainless steel column. Also used was 5% carbowax in Chromosorb G (NAW) packed in the same type of column as above. All 1H NMR spectra were taken on a Varian T-60 spectrometer. All compounds except the silyl compounds used TMS as internal standard. The IR Spectra were recorded on a Perkin-Elmer Model 237B Grating Infrared Spectrophotometer. Al Bis(Trimethylsilyl)_Acetone NMR (0010): 50.13 (s,18H), 51.97 (8,3H), 52.08 (s,lH). 1—Trimethylsilyl-2-Trimethylsiloxypropene'(cis and trans) NMR(CClu): 50.10 (s,l8H), 50.27 (s,18H), 51.78 (s,3H), 51.92 (3,3H), 5u.10 (s,2H). l-TrimethylsilyI-Z-butanone N,N-DimethylhydraZOne NMR(CClu) 50.10 (s,9H), 51.01 (t,3H), 52.05 (m,uH). 52.26 (s,6H). l-Trimethylsilyl-u-methy1-2-pentanone N,N-Dimethyl- hydrazone NMR(CC1u): 50.10 (s,9H), 51.0 (s,1H), 50.83 (m,6H), 61.9“ (m,hH), 62.28 (s,6H). 2-Trimethylsilylcyclohexanone N,N-Dimethylhydrazone NMR 13. This indicates that the less hindered alpha protons are removed more rapidly by a strong base. Under equilibrium control, when longer reaction times and excess ketone are employed, the more highly substituted isomer predominates, 142;, 13_> 12. Consequently, a kinetically controlled mixture of enolates is obtained by slowly adding a ketone to an excess of a strong base in an aprotic solvent, whereas the slow addi- tion of a strong base to a ketone or the presence of excess ketone allows the formation of an equilibrium mixture of enolate anions.“5 The observation has been made that equilibration among lithium enolates, in the presence of a proton donor, is much slower than equilibrium among potassium enolates. Equilibration among lithium enolates U7 often requires a reaction period of 30 minutes with 20% or more excess ketone.“3 Among strong bases, the lithium dialkylamides are particularly convenient to prepare and use.u5'u9 It has been reported50 that lithium diisopropylamide reacts with Eggtybutyl trimethylsilylacetate at -78°C producing the lithium ester enolate with the silyl group still intact. Therefore, it appeared that lithium diisopropylamide could be the base of choice to provide us with silyl ketone eno- lates. 51 discovered that potassium Recently, Charles A. Brown hydride shows exceptional reactivity as a strong base toward ketones. In contrast to NaH or LiH, KH in tetrahydrofuran vigorously metalates a wide range of ketones with little or no self-condensation or reduction. Solutions of highly reactive potassium enolates are formed quantitatively in a few minutes at 20°C. Thus potassium hydride may also be an effective base for removing the a-hydrogens in the B-ketosilanes. RESULTS Our first attempts to generate the enolates of B-keto— silanes employed lithium diisopropylamide (LDA) as the strong base. Addition of l-trimethylsilyl-34phenyl-2- propanone, 13, to a pentane solution of LDA at 0°C, “8 followed by removal of solvent and diisopropylamine under reduced pressure, produced a white solid, presumably a mixture of isomeric lithium enolates 15_and 16 and other products (Eq. 23). O I pentane H\ IOLi H\ ,CHZSiMe Me3SiCHchH2¢ + LDA _ 3 v 2 + a O /C C\ C C 25 9 15 min ¢ CH SiMe3 5 ‘0Li 2 15. 1 a 15b + ,0-:0\H + ,0=0: + HN L10 L10 SiMe3 (23) 16a 16b These products were dissolved in deuterated tetrahydro- furan under argon and examined by NMR. A singlet at 61.56 can be assigned to the methylene group adjacent to the sili- con, and a singlet at 6“.5“ corresponds to the olifinic protons in 15g, The presence of unreacted ketone made it difficult to identify and determine the other signals which were thought to be from enolates 15g, 163 and 16p, Trapping the enolates with trimethylchlorosilane gave silylated products 11 and 18. Neither benzene nor toluene could be substituted for deuterated “9 tetrahydrofuran as a solvent, since the silyl derivatives are insoluble in these hydrocarbons. OSiMe3 ' OSiMeB ¢CH I CCHZSiMe3 ¢CHZC 3 CHSiMe3 17a; cis 18a; cis 17b; trans 18b; trans The NMR spectrum of the product mixture proved not to be particularly helpful. Since we were unable to identify the numerous signals, the relative quantities of each product could not be determined. Methylation of 13 by adding methyl iodide to the lithium enolates and then quenching with water gave a mixture of 19 and 10 (Eq. 2“). o THF L10 OLi Me3SiCH2CCH2¢ + LDA : Me3SiCH2C=CH¢ + Me3SiCH-00H25 25°, 15 min 11. 12 15 o 0 CH3I “ . —:~Me3SiCH200H¢ + Me3SiCHCCH2¢ + L11 (2“) 25°, 15 min 19 20 When an excess of 1“ was allowed to react with lithium 50 diisopropylamide overnight or over a period of 36 hours, small changes in the relative quantities of the two methylat- ed products were observed by NMR examination. The same reaction sequence was repeated with benzyl halides as alkylating agents (Eq. 25).. , THF ¢CH2X fl Me3SiCH200H25 + LDA _— 15 + _6_———->Me33i0H200H5 25°, 1h 25°, 1h CH 13 ‘ X=Br, 01 5 I OH 2? + Me3SiCH-CCH2¢ (25) Benzyl halides were used as alkylating agents in the hope that the NMR spectra of the isomers would be easier to identify. This proved not to be the case. Similar experiments were performed with l-trimethyl- silyl-Z-butanone, 21 (Eq. 26). O THF d 0L1 I - 3 I Me3310H200H20H3 + LDA e Me3SiCH20=0H0H3 25°, 15 min 2.1 2.2.. OLi I + Me3SiCH=CCH2CH3 23 (26) 51 Subsequent silylation and alkylation of the enolate anion mixture gg and 13 led to the silylated and methylated products (Eqs. 27 and 28). 22 + _2_3_ ME3SiCl M SiCH 8301:8131 + M SiCH 888M831 : e 8 e I "" 250’ 15 min 3 2 3 3 2 3 (27) CH3I 7 R _2_ + a; ; Me3SiCH2CCHCH3 + Me3SiCHCCH2CH3 25°, 15 min CH3 CH3 2“ 25 (28) Since these experiments were carried out on a small scale; 1131, 1-5 mmoles, no attempts were made to isolate each component in a pure state. Analysis of the products by gas chromatography was misleading because these com- pounds decomposed on the GLC columns. The NMR spectral studies were also not very informative in terms of the relative quantities of the products, and the starting ke- tones. In an effort to obtain a more effective probe for product composition, we removed the silicon in the final step. The effectiveness of this procedure was demonstrated by stirring the silyl ketone 15 in 6H hydrochloric acid for 1 hour and “5 minutes. This cleanly removed the silyl group giving phenylacetone (Eq. 29). _—.. a. 52 9 611 H01, H20 9 Me SiCH con ¢ ton CCH ¢ + Me Si-O-SiMe (29) 3 2 2 1.75 h 3 2 3 3 1“ The products from reactions of lfl_with a variety of bases, followed by methylation can thus be determined by quenching with HCl (Eq. 30). temp. 9 ? CH3I u + base——-> Me SiCH c=CH¢ + Me SiCH=CCH ¢ ———-> -_ reac 3 2 3 2 250 .time 20 min 63 HCl ‘3 R 8 ——»CH CCH¢ + CH CH CCH ¢ + CH CH CCH¢ (30) 3 CH3 26 27 28 With each base, experiments were performed in which the quantities of starting materials, temperature, and reac- tion time were varied. The results of this survey are shown in Table III. We found that the major product in all cases was 3-phenyl-2-butanone g§_which corresponds to the silyl ketone 12. Only small amounts of the dimethylated product gfi appeared as a byproduct. Since we wanted to compare the effects of the silicon to those of the aryl group on the acidity of the adjacent methylene protons, we performed paralleled experiments with non-silylated .mcaoaz Hamum>o co comma ohm mposoopn mo naofiw up .moaoah vac who mvaofim HHo mwmwww no>apmamm nm>apmamm no>fipwamm omwommommo m mmwm we 08 m m m HH as m m m m H as m m m + e mom mo mo + emoo mmoIlquauHIe mowumofim m: + emou w moflm ms .an9 mmwm + e mow moan m: Hm 0 mo Io o.mommm m30fi9m> sud: ozocwaoHMImI HaconQImI HmafimamnmeHABI H no mcofipomom .HHH magma 5h analog of £3 (Eq. 31). ¢CH2CCH3 + LDA 23¢CH-CCH3 + ¢CH2080H2 reaction time CHBI : _§_ + g_ + L (31) 25°, 20 min Lithium enolates of phenylacetone were prepared by stirring the ketone with lithium diisopropylamide in tetrahydro- furan solution at 25°C over various time periods. Subse- quent methylation by methyl iodide gave quantitative over- all yields of products. Table IV lists the results of these experiments, and as expected, gfi was the major product. We investigated the reactions of 1-trimethylsilyl-2- butanone with such bases as lithium diisopropylamide and potassium hydride, followed by methylation (Eq. 32). The results, after hydrolysis of 32_and 3; with 6§_hydrochloric acid, are tabulated in Table V. It is interesting to note that the reactions of 22 with lithium diisopropyl- amide at room temperature led predominantly to ketone 35, and at reflux temperature gave mostly 3;. In the cases with RE, the only product obtained from reactions at .OOHOHO Hamho>o no woman who mposcopn mo vHoH» up .moaoam vac mum OUHOHO HHo coHpommm oapmm >Hs m u OOHOHH O OOHOHH a OOHOHH m>HpmHmm m>HpmHom o>HpmHmm 5 OmmemOmmO + O mmO m m . mm 2H5 ON .omw m m m OOH» OOHOOOOO m m +OO mO mO + OmOm TO m O mOOI mO + OmOIO mO «mg + O mOO mO O H mO HHO HHO 0mm x .mOHEszaonaomHHQ Edanpaq saws mcopmowazcmcm yo macapowom .>H canoe .OUHOHO HHmnm>o no woman mum mpospopa mo UHOH» Rn .OOHOHO oqu mam OOHOHO HHOO 56 AOOHOOOO AOOzOIOOeO II OOH II OO O OO OOH H\H.H AOOHOOOV AOOOOIOOOO II OO OH OO O H OOH H\H.H Axsammnv O mp HN O» O H OOH H\H.H II OOH II OO O ON OO H\H.H II OOH mmoOOO mm OHE OH OO H\H.H II OOH II OO O ON Ox H.H\H II OOH OOOOOO Om OH2 OH Ox H.H\H II O mm OO O ON OOH H\H.H II O NO O: OH5 OH OOH HxH.H II HH mm m: O ON OOH H.H\H II NH mm mm OH2 OH OOH H.H\H mOONOOwOONAmOOO mOONOOwNOOmOO NAmmOOOmeOO O O.OHOHH OOHO OOOO OOOO .0 OO O o Hampm>o COHpommm \OQOpmx O O OHOH» O O OHOHO O OOHOHO . OHOOO m>prHmm o>HumHmm m>HpmHmm . .>Hsvm mOO O N mOONOOOOONAmOOV + mOONOOONmOmOO + mOOOWOmOO.IIIJHI OHS ON a = w HOO 2O .omN m m m m m OOH» OOHOOOOOO m N m m IINIII. OONOOOIOOHO AmOOO + OOOOIONOOHO A OOOI, Ommm + OONOOO OOHO A OOO H 20 W W 0mm % .mmmmm pmpomamm npfiz OCOCOHSQINIHOHHmaznmeHpBIH no mcofipommm .> magma 57 25°C was 3-pentanone 3;. o 0' H temperature I CCHQCH3 + base ~e~Me3SiCHZC=CHCH3 react. time Me SiCH 3 2 29 30 O + Me3SiCH=CCH20H ~O~Me SiCH 00H0H 3 25°, 20 min 3 2 H 3 3 31 22. 9 6N H01 u + Me331gHCCH20H3-—-;—;—>CH3CCHCH3 + CH3CH2CCH20H3 (32) H 3 CH3 3;. 25 ii For comparison, 2—butanone was treated with various quantities of KH for times of 15 minutes or 2“ hours prior to the methylation (Eq. 33). 9 ether 9K 0H 00H 0H + KH H:0H 0=0HCH + 3 2 3 25°, react. 3 3 time 9K CH3I + CH2-CCH20H3 O: 13 + 3__ (33) 25°, 20 min .OUHOHO HHwHo>o no woman mum mpospona mo OHOHM an .OOHOHO 0A6 mam OOHOHO HHOO 58 OH OO ON O ON H\H.H HN OH OO 2H8 OH H\H.H HO OO OO O ON H.H\H mm 3. m» CHE ma H.H\.O OOONOOwNOOOOO NAOOOOOOwOOO O O.OHOHH OEHO OOHpommm OmmeOOOO O O HHOOOOO OHOOO >HOOO O O.OHOHH O O.OHOHH OOONOOONOOOOO + O O OOH» O . . O O O OH5 ON OON O N N O O OOHOOOOO OON O N O OOOOO OOI, O OO OOOI OO + OOOOIO OOII OO + OO OOO OO O H OO Om OO .33.... m .mx Sufi: mcocmpsmlm mo mCOHpommm .H> mHan 59 The results, which are listed in Table VI, show 3£_as the major product both kinetically and at equilibrium. When lithium diisoprOpylamide was employed, we were not able to find an appropriate solvent other than tetrahydro- furan, which complicated the GLC analysis of the products. DISCUSSION The methods now available for the formation of enol acetates or trimethylsilyl enol ethers (Eq. 3“) sometimes ? 9 excess MeBSiCl 9SiMe3 Cl)SiMe3 RCH=CCH2R' + RCH2C=CHR' (3h) allow a particular structural isomer to be isolated in high yield, and easily converted back to the starting ketone. However, it is often very difficult to obtain a good separa- tion of the two isomeric trimethylsilyl enol ethers from an unsymmetrical ketone without recourse to either an efficient fractional distillation (through a spinning- ban column) or preparative gas chromatography.52 This difficulty, together with the possible decomposition of the silylated ketones, led us to consider alkylation of the enolate mixture as a procedure which would allow us 60 to determine each individual product. The alkylation method is, in practice, complicated by the fact that even under kinetically controlled conditions, enolate equilibrium can occur once some monoalkylated pro- duct (an un-ionized ketone such as 3Q) is produced in the reaction mixture, leading to dialkylated products (Eqs. 35 and 36). If this equilibration rate is comparable to the 2 excess base; DME ’ + 3.6. RX _. .3]. 3.8. O O I R I . . . . 1) Base or florfl 3.2. L ;» dialkylated products 2) RX lg RX 32 O31_.32 (36) enolate equilibration rate of reaction of the initial enolate with the alkylating agent, the amounts of 32_and 59 in the final mixture would not represent the relative quantities of the enolates 31_and 3§_initially formed. However, it was found that lithium eno- lates undergo alkylation with reactive alkylating agents such as methyl iodide faster than enolate equilibration occurs.uua’53a 61 Lithium enolates offer the best compromise of reasonable reaction rates with alkylating agents, and relatively slow equilibration rates among enolates. The fact that less highly substituted alkali metal enolates (such as 31) may sometimes react more slowly with alkyl halides than their analogs having additional a-substituents has been noted in several studies.53 This problem, under condi- tions of kinetic control, was lessened by the use of excess methyl iodide. Methylation at the highly substituted position (such as 3Q) can be accomplished easily since both the position of enolate equilibrium and the rate of reaction with the alkylating agent favor this site.52 It was noted that the yields of methylated products of phenylacetone in Table IV (96-100% of 3-phenyl-2-butanone) were much better than yields of its trimethylsilyl enol ethers (Bu-61%) reported in the literature.“6 We believe .that by using methyl iodide to alkylate the enolate mix- ture of B-ketosilanes in our experiments gave a close ap- proximation of the relative quantities of the isomeric enolates formed in the reactions of the silyl ketones with bases under various conditions. When potassium hydride or potassium bis(trimethylsilyl)- amide was used as a base to drprotonate the ketones, the enolate mixtures obtained were enriched in the more highly substituted component, even under kinetically controlled conditions. This may be attributed to the relatively 62 faster reaction of the highly reactive potassium enolates with the starting ketones, as compared to the slower rate of enolate formation, resulting in equilibrium mixtures of the enolate anions. On the other hand, the lithium enolates equilibrate slowly at room temperature and gave mostly kinetically favored product. However, increasing the polarity of the solvent, accompanied by heating, ap- peared to increase the rate of equilibration, giving 75- 1001 (GLC, relative) yield of the thermodynamically favored product in the case of l-trimethylsilyl-Z-butanone. Comparison of the products obtained from the reactions of l-trimethylsilyl-3-pheny1-2-propanone with various bases (Eq. 37), shown in Table III, with those of phenyl- acetone and lithium diisopropylamide, shown in Table IV, under kinetic and equilibrium conditions shows 3-phenyl-2- butanone was the major product in all cases. There are 3 base ? (CH3 ) 3SiCI-IzCCH2 G ——O ( CH3 ) 3SiCHZC=CH-. l” "2 O- “3 two features here to be considered. The l-trimethylsilyl— 3-phenyl-2-propanone has on one a-carbon, a silicon whose 63 d-orbitals are available for overlapping if the carbanion is formed on that side, and an aromatic ring is attached to the other a-methylene group. In the studies of carbon acids, Bordwell and coworkers?4 found an abundance of kinetic and equilibrium evidence that a-phenyl substitution increases the acidity of a hydrogen atom attached to carbon. The large acidifying effects of a phenyl substituent, ApKH - 7.2 (measured in DMSO), on acetone must be due primarily to its ability to stabilize the anion by delocalization of charge. This view is supported by the relatively small acidifying effects of phenyl that have been observed in molecules where steric crowding prevents effective overlap between p-orbital of the carbanion and the w-system of the phenyl group.5u For example, 9-phenylxanthene was found to be only 2.h pK units more acidic than xanthene in DMSO.5u On the other hand, silicon atom has empty d-orbitals in the valence shell. Two of these are of w-symmetry relative to the tetrahedral a-bonds of the saturated element, and so can combine with the n-orbitals of any attached atom or group; if the latter w-orbitals contain .electron-pairs, (as in the halogen atoms, or the dimethyl- amino group), the energy of the delocalized electrons will be lowered by this interaction, and a (p+d) w-bond will 36 result: 6h ‘9 9 O 6‘ 9 Si Other Atom H-Bond If a proton is abstracted from the methylene group next to silicon in compound $3, the resulting enolate anion will be stabilized by delocalization of the w-electrons into the vacant d-orbitals of silicon. The greater yield of 3-phenyl-2-butanone, gg, com- pared with pheny1-2-butanone, 21, indicated that a greater amount of fig_ was present in the mixture of isomeric eno- lates (Eq. 37) than 3} at equilibrium. In the case of l- trimethylsilyl-3-phenyl-2-propanone, as well as phenyl acetone, the predominant formation of the enolate where the negative charge was on the carbon adjacent to the phenyl- group, was attributed to the (p+p) w-conjugation, resulting from delocalization of the charge with the aromatic «- system. Evidently, when both silicon and phenyl groups are present in a compound such as 13, the acidifying effect exerted by the phenyl is larger than that of the silicon, resulting in a greater yield of g§_than 21 in reaction 30. However, when the phenyl is replaced by a methyl group, 65 as in l-trimethylsilyl-Z-butanone, an equilibrium controlled reaction gave predominantly diethylketone, 33 in Eq.-32, and indicates that_more of the enolate 33, Me3SiCH-8CH20H3, than 39, Me3SiCH23-CHCH3, had been formed prior to methyla- tion, while the more highly substituted enolate was favored in the case of methyl ethyl ketone, producing more of methyl isopropyl ketone, 33, than diethyl ketone, 33, (Eq. 33). The methylene protons alpha to the tri- methylsilyl group are thus more acidic than those adjacent to a methyl group. In the studies of alkyl effects on a-methylene protons 55 of acetone, Bordwell and co-workers reported an increase in acidity, caused by substitution of a hydrogen with a methyl, by 1.3 pH units (in the gas phase). Although an exact pK value for methyl ethyl ketone in solution is not available, acidifying effects of methyl are reportedly smaller for ketones in DMSO than in the gas phase. The methyl group is assumed to enhance the acidity of the alpha methylene protons by its polarizability and hyperconjuga- tion. If we assume from Bordwell's report that the pH of the methylene protons in phenyl acetone is ~19.3 (DK acetone - 26.5, ApKH - 7.2)5u in DMSO, then together with the findings from our experiments we are led to believe that the acidifying effect of the silicon is between those of the methyl and phenyl groups. Consequently, the alpha methylene protons next to silicon in the B-ketosilanes 66 should have pH value between 19.3 and 25.2, if measured in DMSO. In tetrahydrofuran, in which most of our experi- ments were run, we might expect ApKH of a-phenyl substituted acetone to be smaller than 7.2, since ApKH in diglyme for acetophenone/phenylacetophenone is only 3.2, compared to 7.2 in DMSO.5u Thus the limits for the pH value of the B-ketosilanes, if measured in THF or diglyme, might be expected to be different from those measured in DMSO. 67 EXPERIMENTAL I. Materials Ketosilanes The B-ketosilanes prepared in Chapter I and already purified by vacuum distillation were stored under argon in brown bottles. Trimethylchlorosilane The commercially available trimethylchlorosilane was freshly distilled and stored under argon. Organic Bases Tert-butyllithium was stored under argon in a re- frigerator. Diisopropylamine was distilled from calcium hydride and stored under argon. Potassium Hydride Potassium hydride was commercially available from Ventron Corp. as a 25-30% mineral oil dispersion. The dispersion was standardized by measuring the gas given off when a sample of known volume was treated with water. 68 Potassium Bis(Trimethylsilyl) Amide Potassium bis(trimethylsilyl) amide was prepared by adding distilled bis(trimethylsilyl) amine to a THF solu- tion of equivalent amount of potassium hydride with cooling and vigorous stirring.”b After filtration and drying, white solid amide was stored under argon. Bis(trimethylsilyl) amine or l,l,l,3,3,3-hexamethyl disilazane was obtained in 98% pure liquid from Aldrich and was purified by distillation. Organic Halides Methyl iodide was distilled and stored in a brown bottle over copper wire with a septum inlet in a refrigera- tor. Other organic halides were purified by distillation prior to use. Ketones Commercial phenylacetone and 2-butanone were used without further purification. Organic Solvents Pentane, hexane and methylene chloride were dried over molecular sieves prior to use. Anhydrous diethyl ether was distilled from lithium aluminum hydride and 69 stored under argon over molecular sieves. Tetrahydro- furan was distilled from sodium benzophenone ketyl and stored under argon over molecular sieves. Dimethylsul- foxide was dried over molecular sieves prior to use. II. Generation of Enolates of the B-KetosiAQnes A. Preparation of Lithium Diisopropylamide (LDA) A dry round-bottomed flask with side arm, equipped as shown in Figure l was flushed with argon, then charged with 0.5 m1 of pentane and cooled to 0°C. Then 0.31 ml (0.5 mmol) of 1.6M solution of n-butyllithium in hexane was added, followed by dropwise addition of 0.07 ml (0.5 mmol) of diisoprpylamine. The reaction mixture was warmed to room temperature and stirred for 10 minutes. The solvent was then removed under reduced pressure, leav- ing a white solid. The LDA just prepared must be used immediately. B. Preparation of Lithio Trimethylsilyl Ketone Enolateg The following procedure for the reaction of LDA and 1-trimethylsilyl-3-phenyl-2-propanone is representative. LDA (0.5 mmol) was prepared in a 25 ml flask in the manner described above without removing pentane. The flask was cooled to 0°C and 0.11 ml (0.56 mmol) of l-trimethylsilyl- 3-phenyl-2-propanone was added slowly with stirring. The 7O ice bath was removed and the reaction mixture stirred for 15 minutes. A vacuum was then applied to remove the pen- tane and the amine. The remaining white solid was dis- solved in 0.5 ml of deuterated THF and analyzed under argon by NMR. Then an additional 2011 (0.1 mmol) Of the silyl ketone was added to the NMR sample and another NMR spectrum was taken after about 8 hours. III. Silylation of the g-Ketosilanes The enolates of l-trimethylsilyl-3-phenyl-2-propanone (0.10 ml, 0.5 mmol) were prepared following procedure IIB. Before the THF-d8 solution of enolates was analyzed by NMR, 0.06 ml (0.5 mmol) of trimethylchlorosilane was added and the solution stirred for 15 minutes, followed by removal of LiCl by filtration. No more ketone was added to NMR samples. IV. Alkylation of the_§—Ketosilanes A. Methylation of §:Ketosilanes and NMR Analysis The following procedure for the methylation of l-tri- methylsi1y1-3-phenyl-2-propanone is representative. LDA was prepared from the reaction of n-butyllithium (0.63 ml of 1.6M solution in hexane, 1 mmol) and 0.1" ml (0.1 mmol) 71 of diisopropylamine as described in procedure IIA. THF (1 ml) was added and 0.20 ml (1 mmol) of l-trimethylsilyl- 3-phenyl-2-propanone was added. The reaction mixture was allowed to stir for 15 minutes at room temperature, cooled to 050, then methyl iodide (0.12 ml, 2 mmol) was added and stirred for 15 minutes at room temperature. Then it was extracted with pentane and water, dried over MgSOu and concentrated. NMR analysis showed a mixture of pro- ducts. Two similar experiments were performed with 10% excess ketone (0.22 ml, 1.1 mmol) allowed to react with the solu- tion of LDA overnight and over a period of 36 hours. B. Benzylation of B-Ketosilanes and NMR Analysis Procedure IVA was followed, except that 0.23 ml (2 mmol) of benzyl chloride was substituted for methyl iodide The l-trimethylsilyl-3-phenyl-2—propanone (0.2 ml, 1 mmol) was stirred with LDA (1 mmol), for 1 hour prior to alkyla- tion in one experiment, U8 hours in another, where 10% excess silyl ketone (0.22 ml, 1.1 mmol) was used. NMR analysis of the product mixtures showed complex spectra. 72 V. M n of l-T imeth lsil - - hen 1-2- r anone: Kinetic vs. Equilibrium Conditions A. 10% Excess LDA and l-Trimethylsilyl-3-phenyl-2- propanone Reactions To a 5 ml THF solution of 3.3 mmol of LDA at 0°C, prepared as described in procedure IIA, was added 0.59 ml (3 mmol) of l-trimethylsilyl—3-pheny1-2-propanone, followed by stirring at room temperature. The reaction time for this step was varied for each experiment (5 min, lb and 2hh). Then the flask was immersed in an ice—water bath and 0.36 ml (6 mmol) of methyl iodide was added. The reaction mixture was stirred for 20 minutes'at room temperature then diluted with pentane. To the diluted mixture 3 ml of 63 H01 was added and it was stirred for 2 hours at room temperature. After extraction, the organic layer was washed with water and filtered through a 15 ml fritted funnel covered with 0.25 inch of anhydrous MgSOu and concentrated. GLC analysis (2.5% SE-30) using 1 mmol of tridecane as internal standard of all three experiments established the predominant presence of 3-pheny1-2-butanone over other products. B. LDA and 10% Excess l-Trimethylsily -3-phenyl-2- propanone Reactions Three experiments were performed exactly as described 73 in procedure VA, except the quantities of LDA and the silyl ketone used are as follows: 3 mmol of LDA from 1.88 ml (3 mmol) of 1.6M,n-butyllithium in hexane and 0.fl2 ml (3 mmol) of diisopropylamine in 5 ml pentane, and 0.65 ml (3.3 mmol) of the silyl ketone. The same results as above were obtained by GLC analysis. C. 301 Excess KH and l-Trimethylsilyl-3-phenyl-2- propanone Reactions A dry 50 m1 round-bottomed flask equipped with mag- netic stirring, septum inlet and a gas inlet valve was attached to a mineral oil filled gas buret and a mercury bubbler, was flushed with argon and charged with 0.8 ml (3.3 mmol) of H.11M_KH dispersion in mineral oil. The flask was immersed in a water bath maintained at 25°C. THF (5 ml) was injected followed by dropwise addition of l-trimethylsilyl—3-pheny1-2—propanone (0.59 ml, 3 mmol). A total of U2 ml (1.88 mmol) of hydrogen gas evolved in 5 minutes. Following completion of the reaction (1h), 0.36 ml (6 mmol) of methyl iodide was added at 25°C and after stirring for 20 minutes it was quenched with 3 ml of 6§_HCl. The mixture was stirred for 2 hours, extracted with ether, washed with water, dried over MgSOu and con- centrated. The major peak in GLC traces (2.5% SE-30) was 3-phenyl-2-butanone. 7“ Another experiment using exactly the same quantities was performed at ~78°C and maintained at -78°C through quenching. The reaction time for the enolates and methyl iodide at -78°C was 30 minutes. It was shown by GLC that enolates formation did not occur under these conditions. D. 0 Exces l—T imeth i 1- - h - - propanone Reaction l-Trimethylsilyl-3-phenyl-2-propanone (0.65 ml, 3.3 mmol) was treated with 0.73 ml (3 mmol) of h.llM_KH dis- persion in mineral oil at 25°C followed by methylation according to the above procedure, VC. GLC analysis showed 3-phenyl-2-butanone as the only monoalkylated product. E. 101 Excess,(Me331)2NK#§nd l-Trimethylsilyl-3- phenyl-23propanone Reactions Potassium bis(trimethylsilyl) amide (0.62 g, 3.11 mmol) was weighed in a glove bag under argon and trans- ferred into a 25 ml round-bottomed flask. THF (5 ml) was added, followed by 0.55 ml (2.83 mmol) of l-trimethylsilyl- 3-phenyl-2-propanone at 25°C. The reaction mixture was stirred for 1 hour in room temperature water bath. Sub- sequent methylation and work up procedures were done at 25°C in the same manner described previously in procedure VC. GLC analysis showed 3-pheny1-2-butanone as the major 75 product. The exact same experiment was repeated except the temperature was held at -78°C and 30 minutes were allowed for the reaction of enolates and methyl iodide. The 3- phenyl-Z-butanone was the only product shown by GLC. F. (Me3Si)2 NK and 10% Excess l-Trimethylsilyl-B- phenyl-Zepropanone Reactions l-Trimethylsilyl-3-phenyl-2-propanone (0.63 ml, 3.27 mmol) was treated with 0.59 g (2.9 mmol) of (Me331)2 NR and then methylated with methyl iodide following pro- cedure VE at 25° and -78°C. The results obtained from GLC analysis were similar to those of Section VE. VI. Methylation of Phenylacetone A. 10% Excess LDA and Phenylacetone To a solution of 5.5 mmol of LDA in 8 ml of pentane, was added 0.66 ml (5 mmol)of phenylacetone at 0°C. The solution was warmed to room temperature, stirred for 15 minutes, cooled back to 0°C and 0.62 ml (10 mmol) of methyl iodide was added. After 20 minutes of stirring at room temperature, the reaction mixture was quenched with ZflbHCl, extracted with pentane and water, dried over MgSOu and concentrated. 76 B. LDA and 10% Excess Phenylacetone Procedure VIA above was followed for the reaction of phenylacetone (0.73 ml, 5.5 mmol) and 5 mmol of LDA, followed by methylation. Both the 15 minute and 2h hour experiments gave exclusively 3-phenyl-2-butanone as shown by GLC. VII. Preparation of Authentic Sample of 2-Phenyl-3- pentanone To a 5 ml THF solution of 3 mmol of LDA prepared according to procedure IIA, was added 0.59 ml (3 mmol) of l-trimethylsilyl-3-phenyl-2-propanone at 0°C. The reaction mixture was stirred at room temperature for 3 hours, then cooled to 0°C and 0.36 ml (6 mmol) of methyl iodide was added. After stirring for 20 minutes at 25°C, most of the solvent and excess methyl iodide were removed by vacuum and the mixture was filtered. Then the contents of the flask was dissolved in a small volume of THF. LDA (3.3 mmol) was prepared in a separate flask. Then the solu- tion in the first flask was taken up into a syringe and transferred to the LDA flask at 0°C with stirring. The reaction mixture was stirred for 3 hours at 25°C, cooled to 0°C again and 0.36 ml (6 mmol) methyl iodide was added. After 20 minutes of stirring at room temperature, 6§_HCl was added while cooling and the solution was stirred for 77 2 hours at 25°C. Then it was extracted with pentane, washed with water, dried over MgSOu and concentrated. The 2-phenyl-3-pentanone was collected from GLC (2.5% SE-BO). VIII. Methylation of l-Trimethylsilyl-2-butanone: Kinetic vs. Equilibrium Condition; A. 10% Excess LDA and l-Trimethylsilyl-2-butanone Reactions LDA (2.2 mmol) was prepared in a 25 m1 round-bottomed flask following procedure IIA, and was dissolved in 3.5 m1 of THF. The solution was cooled at 0°C, 0.32 ml (2 mmol) of 1-trimethylsily1-2-butanone was added and stirred for 15 minutes at room temperature. Then it was cooled to 0°C again and 0.25 mi (u mmol) of methyl iodide was added. The ice bath was removed and the reaction mixture stirred for 20 minutes. Then a vacuum was applied to remove THF whose peak in the GLC came on top of the products. The residue was dissolved in 2 ml of pentane, quenched with 5 m1 of 63 H01 and stirred for 2 hours at 25°C. Then it was extracted with pentane, washed with water, dried (MgSOu) and the solvent was evaporated partially by rotary evapora- tor and 25°C water bath. Care must be taken to prevent evaporation of relatively low boiling products (BO-125°C). GLC analysis (6 foot column - 10% SE-30 on Chromosorb G NAW) 78 indicated the ratio of methyl isopropyl ketone and diethyl ketone to be 7.3:1, using 1 mmol of octane as internal standard. The same experiment was repeated allowing the ketone and LDA to stir for 2“ hours prior to methylation. Similar results were obtained from GLC analysis. B. LDA and 10% Excess l—Trimetgylsilyl-2-Butanone Reactions A mixture of 0.35 ml (2.2 mmol) of l-trimethylsily1-2- butanone and 3.5 ml THF solution containing 2 mmol of LDA was stirred at 25°C for 15 minutes, followed by methylation, removal of THF and quenching as done in procedure VIIIA. Relative GLC yield of methyl isopropyl ketone was increased to 92%. Another experiment, identical to above, but allowing the silyl ketone to react with LDA over a period of 2A hours, gave almost identical results. Three other experiments were performed with the same quantities of starting materials. The solution of silyl ketone and LDA in one flask was refluxed for 1 hour under argon prior to methylation. In the second and third flasks, 0.1“ m1 (2 mmol) of DMSO was added to the THF solution of silyl ketone and LDA, and the reaction mixtures were also refluxed for 1 hour and “0 hours under argon. After methylation and removal of THF as done previously, 3 ml 79 of pentane were added to the residues, but the residues appeared to be quite insoluble until 63_HC1 was added. After 2 hours of stirring at room temperature, the resulting pentane extracts were washed with water, dried (MgSOu) and concentrated. GLC analysis established the predominant presence of diethyl ketone over other products in all 08.868 I C. 10% Excess KH and 1-Trimethylsilyl-2-butanone Reactions Following the procedure described in procedure VC, substituting 8 ml of anhydrous diethyl ether for THF, one flask of 0.H8 ml (3 mmol) of l-trimethylsily1-2-,butanone and ethereal suspension of 0.81 ml (3.3 mmol) of “.113 RH dispersion in mineral oil was stirred for 15 minutes at room temperature while the other flask with the same quantities of starting materials was stirred for 2h hours at 25°C. Following completion, the reaction mixtures were methylated, quenched and worked up as described previously in procedure VC. Diethyl ketone's presence was exclusive as indicated by GLC analysis of the aliquots of both flasks. D. KH and 10% Excess l-Trimethylsilyl-2-butanone Reactions Two flasks, each with 0.53 ml (3.3 mmol) silyl ketone, 0.73 ml (3 mmol) of H.11M KH suspension in oil and 5 ml of 80 ether, were stirred for 15 minutes and 2” hours at room temperature following the same procedure as above, VIIIC. The resulting solutions after methylation and quenching were analyzed by GLC and found to contain 100% yield of diethyl ketone relative to the overall yield. IX. Methylation of Methyl Ethyl Ketone A. 105 Excess KH and Methyl Ethyl Ketone Two flasks were set up as described in procedure VC. Each was charged with 1.35 ml (5.5 mmol) of h.llM'KH suspension in mineral oil followed by addition of 8 m1 of diethyl ether and a dropwise addition of 0.fl5 m1 (5 mmol) of the ketone with stirring. The water bath temperature was maintained at 25°C throughout the reaction time which was 15 minutes for one flask and 2" hours for another. Then the flasks were cooled to 0°C and to each was added 0.u ml (6.67 mmol) of methyl iodide. The reaction mixtures were stirred for an additional 20 minutes at room tempera- ture, quenched with 23’HC1 and extracted with ether. The organic layer was washed with water, dried over MgSOu and the solvent was partially evaporated. The major peak in GLC traces was methyl isopropyl ketone. 81 B. KH and 10% Excess Methyl Ethyl Ketone Procedure IXA was followed using in each flask, 0.50 ml (5.5 mmol) of the ketone, 1.22 ml (5 mmol) of “.11!_ RH suspension in oil. GLC analysis showed an increase in the relative yield of the methyl isopropyl ketone. X. Preparation of Authentic Samples of Alkylated Methyl Ethyl Ketone A. Methyl Isoprppyl Ketone A 25 ml round-bottomed flask equipped as shown in Figure l was charged with 1.22 ml (5 mmol) of “.11M,KH suspension in mineral oil and 8 m1 of dry anhydrous diethyl ether. Methyl ethyl ketone (0.5 ml, 5.5 mmol) was added and the reaction mixture was stirred for 30 minutes. Then 0.& ml (6.67 mmol) of methyl iodide was added at 0°C. After stirr- ing for 20 minutes at room temperature, it was quenched with 2§,HCl, extracted with ether, washed with water, dried (MgSOu) and concentrated. The methyl isopropyl ketone was collected pure from GLC (2.5% SE-30), n25D 0.9u. B. Ethyl Isopropyl Ketone and Diisopropyl Ketone Procedure XA was followed using 0.53 ml (5 mmol) of diethyl ketone, 1.U6 ml (6 mmol) of h.llM’KH suspension in oil and 8 ml of THF. Ethyl iSOpropyl ketone (n25D 0.93) 82 were collected from GLC. XI. Product Analysis The products were isolated by preparative GLC and identified with authentic samples. The GLC yields of methylated products of l-trimethylsilyl-3-phenyl-2-propanone and phenylacetone were determined by use of tridecane as internal standard and a 0.25 inch by 6 foot stainless steel column packed with 2.5% SE-30 on Chromosorb G (NAW). The same type of column packed with 10% SE-30 on Chromosorb G (NAW) and internal standard octane were used for determin- ing the methylated products of l-trimethylsilyl-2-butanone and methyl ethyl ketone. A11 1H NMR spectra were taken on a Varian T—60 spectrometer, with some using TMS as external standard. 3-Phenyl-2-butanone NMR