F. ‘3“.‘LS .IA #3\\\” L ‘ «1‘1” \‘ ' K V; " "WI/5”.” bveaoug mes: 25¢ For «a per ite- RETQRNING LIBRARY MATERIALS. Place in bookre turn to move charge from circulation records THE SYNTHESIS AND REACTIONS OF A HINDERED SECONDARY AMINE By Zakaria A. Fataftah A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1980 ABSTRACT THE SYNTHESIS AND REACTIONS OF A HINDERED SECONDARY AMINE By Zakaria A. Fataftah The unsaturated secondary amine, bis-(l-ethynylcyclohexyl) amine, lg, was successfully synthesized by coupling l-chloro-l-ethynyl- cyclohexane with l-ethynylcyclohexylamine in 65% yield. Raney nickel hydrogenation of l§_gave the corresponding saturated amine, bis-(1- ethylcyclohexyl) amine, 11, in 80% yield. Several reactions of £1 with typical electrophiles, including methyl iodide, boron trifluoride etherate, N-chlorosuccinimide, bromine and trimethylchlorosilane were investigated. The lithio-bis-(l-ethylcyclohexyl) amide, 21, was prepared by reaction of lz_with n-butyllithium or sec-butyllithium in hexane at room temperature for several days. Amide gz_was formed at a much faster rate (less than 5 minutes), in the presence of an equivalent amount of N,N,N,N-tetramethylethylenediamine. Several reactions of 21 with very weak acids including methyl iodide, toluene, a-methylstyrene and organoboranes were investigated. Amide gz_reacted with 2-methy1-3—pentanone and 3-methy1cyclo- hexanone to give almost exclusively the less substituted lithium enolate. The stereochemistry of the enolates formed by deprotonation of a series of ketones with a variety of lithium dialkylamide bases was Zakaria A. Fataftah investigated. Under kinetically controlled conditions, deprotonation of 3-pentanone and 2-methyl-3-pentanone gave mainly the E-enolate. The Z-enolate was the major product under equilibration conditions. Conditions required to obtain either thermodynamic or kinetic control of ketone deprotonation reactions were investigated. A mechanism was proposed for enolates equilibration. To my wife, Nahla, and my daughter, Joanne. ii ACKNOWLEDGMENTS I wish to extend my appreciation to a fine chemist and a fine friend, Dr. Midhael W. Rathke, for his guidance, assistance and inspira- tion through this investigation. Thanks are also given to Dr. William H. Reusch for serving as Second Reader. Special thanks are extended to my loving parents for all their encouragement, and I hope that their expectations for me have been adequately fulfilled. I wish to thank my wife, Nahla, for her patience, interest and encouragement. Appreciation is extended to Ihor Kopka for being a very pleasant partner for part of this project. I also wish to thank past and present members of Dr. Rathke's research group who have always provided an enjoyable atmosphere in the lab. Finally, I would like to express my gratitude to Yarmouk Univer- sity for providing me with the full scholarship which made it possible for me to continue my graduate studies and complete this work. iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . viii LIST OF FIGURES . . . . . . . . . . . . . ix LIST OF ABBREVIATIONS . . . . . . . . . . . x CHAPTER I THE SYNTHESIS AND REACTIONS OF STERICALLY HINDERED SECONDARY AMINES INTRODUCTION . . . . . . . . . . . . . . 2 RESULTS . . . . . . . . . . . . . . . 7 Synthesis of bis(1-ethylcyclohexyl) amine, _1_Z . . . . . 7 Reactions of bis(1-Ethylcyclohexyl) amine l1 . . . . . 8 Reactions of secondary amines with methyl iodide . . . . 8 Reaction of Secondary Amines With BF3 ' OEt2 . . . . . 9 Reaction of 11 with Trimethylchlorosilane . . . . . . 9 Reaction of H with N-Chlorosuccinimide (NCS) . . . . . 9 Reaction of Secondary Amines With Bromine . . . . . . 10 Reaction of N-Bromoamines with Sodium . . . . . . 11 Reaction of 11 with n-Butyllithium . . . . . . . 13 Reaction of L2 With Other Organolithium Reagants . . . . 14 Analysis for the Formation of the Li-Amide _2_Z_ . . . . . 15 Reaction of lz_with n-BuLi in the Presence of N,N,N,N-tetra- methylethylenediamine . . . . . . . . . . . 16 Reaction of a With Methyl Iodide . . . . . . . . 17 Reaction of a With Toluene . . . . . . . . . 18 iv TABLE OF CONTENTS--Continued Reaction ofng'With a-Methylstyrene . . . Reaction of 21 With Triethylboron . . . . Reaction of gZ_With 9-Methyl BBN . . . . Reaction of 21 and Other Li-Amides With 2-Methyl-3-pentanone Reaction of gz_and Other Li-Amides with 3-Methylcyclohexanone DISCUSSION . . . . . . . . . . . EXPERIMENTAL . . . . . . . . . . I. Materials . . . . . . . II. Preparation of bis(1-ethylcyclohexyl) amine (11) a. Preparation of l-chloro-ethynylcyclohexane (14) b. Preparation of Bis(l-ethynylcyclohexyl) amine (lg) . c. Hydrogenation of bis(l-ethynylcyclohexylamine) with W2 Raney Nickel Catalyst . . . . . III. Reactions of bis(l-ethylcyclohexyl) amine (11) Reaction of lz_with Methyl Iodide . . . Reaction of 11 with BF ° OEt 3 2 O O 0 Reaction of lz_with trimethylchlorosilane Reaction of_lz with N—Chlorosuccinimide (NCS) Reaction of lz_with Bromine . . . . . Reaction of N-Brom-amines with Sodium . Reaction of lz_with Organolithium Reagents . Reaction of lz_with n—Butyllithium in the Presence of MDA O I O O O O O O I 0 Analysis for Li—Amide Formation . . IV. Reactions of the lithio-bis(l-ethylcyclohexyl) amide (21) o o o o o o o o c 0 Reaction of the Amide gz_with Methyl Iodide . V Page 19 19 20 20 22 24 36 36 37 37 38 38 4O 4O 41 41 42 42 43 44 44 45 45 45 TABLE OF CONTENTS--Continued Page Reaction of ZZ_With Toluene .. . . . . . . . 46 Reaction of gz_with arMethylstyrene . . . . . . 46 Reaction of gz_with triethylboron . . . . . . 47 Preparation of B—methyl-9—borabicyclo[8.3.Ijnonane . . 47 Reaction of gz_with B-methyl-9BBN . . . . . . 48 Reaction of gz_and other Li-amides with Unsymmetrical ketones . . . . . . . . . . . . . 48 CHAPTER II STEREOCHEMISTRY OF ENOLATE FORMATION INTRODUCTION . . . . . . . . . . . . . 50 RESULTS . . . . . . . . . . . . . . 59 A Study of the Concentration Effects on the E and Z Enolate Mtio O O O O O O O O O O O O O O 61 A Study of the Stability of the Enolates, 39 and 31, and the Corresponding Silyl Ethers, g; and 3; . . . . . . 61 A Study of the Temperature Effect on the E and Z Enolate Ratio 0 O O O O O O O O O O O O O 61 A Study of the Solvent Effect on the E and Z Enolate Ratio . 63 A Study of the Effect of Excess Ketone on the E and Z Enolate Mtio O O O O O O O I O I O O O I 63 A Study of the Effect of Excess Amide on the E and Z Enolate Mtio O O O O O O O O I O I I O O 68 A Study of the Effect of Ketone Structure on the E and Z malate Ratio 0 O O O O O O O O O O O 70 A Study of the Effect of Amide Structure on the Z and E Enolate Ratio . . . . . . . . . . . . 72 A Study of the Stereochemistry of Ester Enolates Formed by Deprotonation Reactions . . . . . . . . . . 75 vi TABLE OF CONTENTS--Continued Page DISCUSSION . . . . . . . . . . . . . . 79 EXPERIMENTAL . . . . . . . . . . . . . 89 I. General . . . . . . . . . . . . . 89 Spectra . . . . . . . . . . . . . 89 Gas Chromatography . . . . . . . . . . 89 II. Materials . . . . . . . . . . . . 89 Handling of Materials . . . . . . . . . . 89 Amines, Amides and n-Butyllithium . . . . . . . 89 Silylating Reagants . . . . . . . . . . 9O Ketones and Esters . . . . . . . . . . 90 III. The Reactions of Ketones and Esters With Li-Amides . . 91 a. The Reaction of Excess Ketones With Li-Amides . . . 91 b. The Reaction of Ketones With Excess Li-Amides . . . 92 c. The Reaction of Esters with Li-Amides . . . . . 92 IV. Product Analysis . . . . . . . . . . . 92 E—3—(Trimethylsilyloxy)-2-pentene (39) . . . . . 92 2-3-(Trimethylsilyloxy)-2-pentene (31) . . . . . 92 E-4-Methyl-3-(trimethylsilyloxy)-2-pentene (38) . . . 93 Z-4-Methyl-3-(trimethylsilyloxy)-2-pentene (39) . . . 93 2-Methyl—3-(trimethylsilyloxy)-2-pentene . . . . . 93 Z-4,4-Dimethy1-3-(trimethylsilyloxy)-2-pentene (31) . . 93 E-3-Phenyl-3-(trimethylsilyloxy)-2-propene (49) . . . 93 Z-3-Phenyl-3-(trimethylsilyloxy)-2-propene (41) . . . 93 O-Trimethylsilyl-O'-ethyl Methyl Ketene Acetal . . . 94 BIBLIOGRAPHY . . . . . . . . . . . . . 95 LIST OF TABLES Table Page I Reaction of N-bromoamines With Sodium Dispersion . . . 12 II Reaction of RZNLi With 2-methyl-3-pentanone . . . 21 III Reaction of RZNLi With 2-methylcyclohexanone . . . 23 IV A Study ofothe Concentration Effect on the E and Z Enolate Ratio at 0 C with 10% Excess Amide . . . . . . 60 V A Study of the Temperature Effect on the E and Z Enolate Ratio . . . . . . . . . . . . . 62 VI A Study of the Solvent Effect on the E and Z Enolate Ratio 64 VII Deprotonation of 3-Pentanone with LiTMP or LDA (1.0 mmole) in THF at 0°C . . . . . . . . . . . 69 VIII A Study of the Effect of the Amide Structure on the E and Z Enolate Ratio at 0°C . . ,. . . . . . . 73 IX The Effect of the Amide Structure on the Ratio of the Eno- lates Z and E Derived from 3-Pentanone at -78°C . . . 74 X The Effect of the Amide Structure on the Ratiooof the Eno- lates Z and E from 2-Methy1-3-pentanone at -78 C . . . 76 viii LIST OF FIGURES Figure Page I The Effect of Excess Ketone on the E and Z Enolate Ratio . 65 II The Effect of Excess Ketone in the Presence of TMEDA and HMPA on the E and Z Enolate Ratio . . . . . . 65 III The Effect of Excess TMEDA or HMPA on the E and Z Enolate Ratio 0 O O O O O O O O O O I O 66 IV The Effect of Excess Ketone in the Presence or Abgence of TMEDA or HMPA on the Z and E Enolate Ratio at -78 C . . 66 V The Effect of Both Excess Amide and TMEDA or HMPA on the Z and E Enolate Ratio . . . . . . . . . . 70 VI The Ratio of Z and E Enolates Derived from 2-Methyl-3- pentanone . . . . . . . . . . . . 71 VII The Ratio of Z and E Enolates Derived from PrOpiophenone . 72 VIII The Ratio of the Enolates Formed by Deprotonation of Ethyl Propanoate. . . . . . . . . . . . . 75 ix DMF HMPA LiTMP NCS QBBN TMEDA LIST OF ABBREVIATIONS N,N-Dimethylformamide Hexamethylphosphoramide lithium diisopropylamide lithium 2,2,6,6-tetramethylpiperidine N-chlorosuccinimide N-hydrosuccinimide 9-borabicyclo E. . 3 . 1:] nonane N,N,N,N-Tetramethylethylenediamine CHAPTER I THE SYNTHESIS AND REACTIONS OF STERICALLY HINDERED SECONDARY AMINES INTRODUCTION Secondary amines with highly branched alkyl groupings are of general importance in synthetic organic chemistry. Major applications of such amines depend on their increased substrate selectivity, resulting from steric factors. For example1 the N-chloro-derivative 1 of t-butylneopentylamine in sulfuric acid solution generates radical cation 2_which exhibits an increased selectivity for primary hydrogen over tertiary hydrogen (CH3)C\\ 30232804 (CH3)C\ + p s t 3 INC1 ‘1' CHBCH(C33)CH2CH3 T 3 lN'H 1.72 5.98 1 (CH3) CCH2 (CH3)CCH2 3 3 .1. Z n + [CH3)2CHJZNC1 + CH3CH(CH3)CHZCH3-:- [(CH3)2CH)§N H]-—-—+O.25 0.70 1 (1) 1' 2' abstraction (eq. 1). This is presumably due to the greater steric hindrance of radical 2 compared to less hindered radicals such as 2'. The alkali metal amides derived from hindered secondary amines are efficient proton abstractors but poor nucleophiles, and this char- acteristic is often of great synthetic value. For example, 2,2,6,6,- tetramethylpiperidine is probably the most hindered commercially avail- able secondary amine, and its lithium salt (LiTMP) is a very poor nucle- ophile. Although bases of moderate steric requirements such as lithium 3 diethyl- or diisopropylamide do not give a metallated derivative with alkylboranes 3, or vinylboranes_4_,2 LiTMP produces boron-stabilized carbanions in both cases (eqs. 2 and 3). The failure of the smaller benzene 11230113 + LiTMP o is. RZBCHZLi + TMP (2) 25 c ; 10-60% THE ch CH-CHBR' + LiTMP ———> RCELiCH-CHBR' + TMP (3) 2 2 o 2 25 c 5 50-70% bases to give proton abstraction is probably due to their coordination to boron (eq. 4). The bulky LiTMP is too hindered to bond to boron and /CH3 ..ZBCH3 + Base > R213 \ (4) Base therefore reacts to form the carbanion. A variety of related applications of LiTMP has been reported. Olofson3 described the use of LiTMP for a practical synthesis of aryl- cyclopropanes 5 from benzylhalides (eq. 5). LDA was less effective \. ,/ ArCH 01 + LiTMP ———-> [ArCH] ’ ‘ 2 > Ar (5) 2 (54%) (39% yield), presumably because substitution reactions of the starting benzylhalide become more likely with this less hindered amide. In all of these examples, it would be useful to know if second- ary amines with more hindered alkyl groups would show even greater sub- strate selectivity. 4 An a,a'-enolisable non-symmetrical ketone may be deprotonated to two regioisomeric enolates §_and 1 (eq. 6). The lack of regiocontrol in o o P base ' RCHZCLCHRIRZ + ncn-A-cmzlnz + RCHZC-CRIRZ (6) 2 l the formation of.§ and Z is a significant problem which limits the use of such enolates in organic synthesis. It is possible that hindered amide bases might favor proton abstraction from the less hindered side of the carbonyl function. A second reason for preparing highly hindered 20- amines is the possibility that their metal amide derivatives may be significantly stronger bases. C. A. Brown4 investigated the effect of increased alkyl group size upon alkoxide basicity. He found that the base strength of alkoxides increases with alkyl group size. Potassium tricyclohexylmeth— oxide, for example is a stronger base, by about 1.2 pKa units, than potassium t-butoxide. Brown attributed this to a decrease in solvation or ion-pair formation in the more hindered base. It seems likely that such an effect would also occur with metal amide bases. A major goal of this study was the development of a simple, in- expensive route to a secondary amine which is more hindered than second- ary amines which are presently available. We then planned to investigate synthetic applications of such an amine in reactions where steric fac- tors might lead to greater selectivity. In addition, we hoped to obtain information about the base strength of the lithium amide derived from such an amine. The best procedure for the preparation of hindered secondary amines is probably that reported by Hennions. This is illustrated by 5 the reaction of hindered primary amine 2 with tert-propargylic chloride §_to give hindered N-tert-propargylic secondary amine lg (eq. 7). 40ZKOH, H20 HC:C-C(CH3)2C1 + HC:CC(CH3)2NH2 Copper bronze, Cu2C12: [HCzCC(CH3%23§NH o 8 days, 30 C (7) _8_ 2 1_O (47%) Compound 19 was semihydrogenated to 11, using 10% palladium on charcoal as a catalyst, and then hydrogenated to the saturated amine 12, using Raney nickel in ethanol (eq. 8). Pd/C Raney-Ni E a [CHZ-CH C(CH3)232NH -——» [CHBCH2C(CH3)23-2 NH (8) Pet. ether EtOH 1 _1_2_ (41%) Recently, Kopka6 modified Hennion's procedure and pre- pared a series of highly hindered secondary amines. The sequence used to prepare this series of secondary amines is summarized in eqs. 9-13. Kopka found that the coupling procedure exemplified by equation 7 does not give significant yields when applied to more hindered R2CO + NaCECH'---——+ HCECRZCOH (9) HCECRZCOH + HCl -—-—- +'HCECRZCC1 (10) HCECRZCCI + NH2 -——-> HCECRZCNH2 (11) HCECRZCCI + HCECRZCNH2-----+ (HCECRZC)2NH (12) Nil/H2 (HC:CR2C)§ NH'-—-—-——#(CH3CH2R2C)2NH (13) reactants. The best yield (50-70%) of coupled product was obtained with an extra equivalent of the propargylamine serving as the base in place of 6 KOH (eq. 7 and 12). Kopka found also that Raney nickel W-2 was the most effective reducing catalyst (eq. 13). One disadvantage of this scheme is the large number of steps. We chose bis-(l-ethylcyclohexyl) amine (11) as a target amine because the requisite alcohol (13) and loLamine (15) are commercially available and inexpensive. This reduces the number of steps and seemed likely to make the projected synthesis of 11_both simple and inexpensive. OH HCl c1 . CECE ————> CECH (14) E H N32 CECE C':' CH + _1_5 ————> NH (15) 2 l: 19 01120113 _13 + H2 ——-> . NH (16) 2 RESULTS Synthesis of bis(1-ethylgyclohexyl) amine, 17. l-Chloro-l-ethyny1cyclohexane, 14, was prepared7 in 80% yield from the alcohol 1§_by reaction of 13 with excess cold hydrochloric acid in the presence of copper bronze powder, calcium chloride and cuprous chloride (eq. 17). The chloride was sensitive to heat and was used on HCl, Cu2C12 Cl CECH : CECH (17) CaClz, Cu 0°, 1.5 hr 13 14 without further purification. l-Ethynylcyclohexylamine 15, was coupleds’6 in dimethylformamide (DMF) solution with the chloride 14 in the presence of cuprous chloride and copper bronze (eq. 18). The product 1§_was purified by distillation H 3". NH ‘ 2 CuZCl2 C 2 cscu + 1_4 . NH (18) Cu, DMF 2 4°C, 3 days 15 16 (65%). This unsaturated amine was hydrogenated in ethanol solution with Raney nickel, W2, catalyst, activated as reported by Vexlearscheg. The hydrogenation was completed in 20 hr and GLC analysis of the crude 7 8 product showed a second, minor, component (9%) in addition to the sat- urated amine. The hydrogenation was conducted under basic (KOH) as well as acidic (CHBCOOH) conditions in efforts to obtain product of higher purity, but the results were inferior to those obtained under neutral conditions (basic media 80% purity, acidic media 87% purity). The sat- urated amine lz_was purified by distillation with a spinning band column. The isolated yield was 80%. Reactions of bis(1-Ethylcyclohexyl) amine 17. A number of experiments were done with amine 11 to compare its behavior with that of other, less hindered amines. Reactions of secondary amines with methyl iodide. The rate of reaction of diisopropylamine, 2,2,6,6,-tetramethy1- piperidine and amine H with methyl iodide was briefly examined. The three amines were mixed with one equivalent of methyl iodide in deuter- ated chloroform solution at room temperature and allowed to react over— night. Analysis by Hl NMR spectroscopy indicated that the first two amines reacted with methyl iodide to give the N-methylammonium iodide while 11_did not react. The methyl iodide NMR peak disappeared in the first two reactions, and white crystals were formed in the NMR-tube con- taining the 2,2,6,6,-tetramethy1piperidine reaction. These crystals were identified as the N-methylammonium iodide derivativelo. There was no change in the NMR spectrum of a mixture of lz_and methyl iodide, even after one week. Refluxing a deuterated ethanol solution of 11_containing excess methyl iodide for 6 hr also did not lead to reaction (NMR analy- sis). Reaction of Secondary Amines With BF3- OEtZ. Addition of equimolar amounts of boron trifluoride etherate to a hexane solution of diisopropylamine, 2,2,6,6,-tetramethylpiperidine and_11, at room temperature gave, within a few'minutes, a white preci- pitate. These reactions were repeated in CDC13. The NMR spectra showed an up-field shift of ~0.7 ppm, for the methylene hydrogens of the di- ethyl ether. This shift towards the free ether signal is an indication that BF3 is no longer coordinated to the ether. The product from the oOEt was found to be a stable white solid with a 3 2 sharp melting point, 154.5-1550C. A sample of this solid was maintained reaction of 11 with BF under high vacuum overnight with no change either in weight or NMR spec— trum. Reaction of 17 with Trimethylchlorosilane. Amine lz_was added to excess trimethylchlorosilane in CDCl3 solu- tion at room temperature and the mixture was stirred for four days. There was no evidence of reaction by NMR analysis. Reaction of 17 with N-Chlorosuccinimide (NCS). Amine lz_and N-chlorosuccinimide were stirred in methylene chloride11 solution and the reaction was followed by NMR. The signal for NCS vanished and signal for NHS appeared after one week (eq. 19). CH CH CH CH 2 3 2 3 , ‘jmn + Nus-———+ . NCl + NHS (.19) 2 2 17 18 10 The reaction was repeated under more vigorous conditions, using a 1:1 mixture of CCla and CH2C12 at reflux for 20 hrs. The signal for NCS also disappeared, as in the earlier experiment. The methylene hydrogen multiplet observed in the NMR spectrum of the residue in both reactions showed a down field shift of 0.2 ppm from the starting amine. Reaction of Secondary Amines With Bromine. A solution of bromine in CCl4 was added to a solution of 2,2,6, 6,-tetramethy1piperidine in CCl A yellow solid 12_was formed which 4. gave the N-bromo-derivative 29_on treatment with sodium hydroxide solu- tion (eq. 20). The yield of 29_was 66% but a quantitative yield was + J— m -m . 1:1 + Br ———> I Br———* -N (20) obtained by dropwise addition of bromine to a mixture of aqueous sodium hydroxide (1.18M) and a chloroform solution of the amine. A 95% yield was obtained (by NMR) with hexane as the solvent in place of chloroform. The NMR spectrum of a pure sample of 29 (obtained by distillation, 650C/0.7 mm) fortunately could be distinguished from starting amine. Benzene was used as internal standard to determine the yield by integra- ting product signals relative to the benzene signal. In a parallel experiment,11_formed a precipitate shortly after mixing with bromine, but the yield of the N-bromo-derivative of 11 was not determined because the NMR of the product is very similar to that of the starting amine. 11 Reaction of N-Bromoamines with Sodium. The reaction of 29_with sodium dispersion was investigated in some detail in an attempt to form the sodium derivative 21 (eq. 21). l I" ‘ + Na (2.1 equiv) -——+ m (21) Br Na - 20 21 Dropwise addition of 29 to 2.1 equivalent of dispersed sodium in hexane was analyzed for formation of the sodium amide gl_by means of the se- quence shown in equation 22. The yield of trimethylsilyl enol ether 0 ONa OSi(CH3)3 Q 6 AU ——---> N + + (22) H 22 23 23, as determined by GLC, ranged from 15-77 percent. Table 1 summarizes the results of a few of the reactions studied. The N-bromo-derivative of amine 11 was also reacted with sodium. The yield of trimethylsilyl enol ether 23, as determined by GLC, was 50% (eq. 23). When 29_was treated with n-butyllithium or lithium instead of CHZCH3 CHZCH3 O Si(CCH3)3 1)Cyclohexanone, wt: Br + Na ——-> . N Na 2) (CH3) 381Cl ’ 6 2 2 (23) 24 25 23 sodium, measurable amounts of 23 (eq. 24) were not obtained. 12 TABLE I. Reaction of N-bromoamines with sodium dispersion. 1)Cyclo- OSiE t1 hexanone (total Na(2.1 equiv)+N-bromoamine --———+ Na-amide -:—---+ R. T. 2):SiCl yield) (1 equiv) condition -£1 (yield) 1. hexane 50 min 77% 2. " 16 hr 60% 3. THE 30 min 43% 4. EtOEt 1 min 55% 5. " 1 hr 42% 6. hexane and n-Buli instead Na 1 hr 3% 7. hexane and Li instead Na 1 hr 0% 8. hexane, Amine instead of N-bromoamine 1 hr 15% CHZCH3 9. hexane, 'h N Br 3 hr 48% ll 2 10. hexane, overnight 52% 0 Si (CH3) 3 ' n-BuLi 1) Cyclohexanone N+ + or ——-> N "' v‘ s (24) Li 2) (CH3) 3SiCl 20 26 23 Reaction of 17 with n-Butyllithium. n-Butyllithium reacts rapidly with diisopropylamine and 2,2,6,6,- tetramethylpiperidine at 0°C. These amines were added dropwise to an equivalent amount of n-butyllithium in hexane in a flask connected to a mercury bubbler, and the evolution of butane was readily observed (eq. 25). [(CH3)2CH]2 NH + n-BuLi ——> [(CH3) 2CH :]2 NLi + n-Bu-H LDA (25) l I 1 N + n-Buli —-—-+ 1\ N + n-Bu-H -H Li When 11 was added to n—butyllithium, no butane evolution was observed, even after removal of the ice bath. The reaction mixture was allowed to stir overnight and then analyzed for the formation of amide 2_7_ by the sequence shown in equation 26. The yield of the trimethylsilyl enol CHZCH3 CI-IZCI-l3 0 Si(C113)3 _ 1) Cyclohexanone; . NH + n BuLi ———> NLi 2)(CH3)381C1 , s 2 2 (26) 17 27 23 ether, _2_3_ obtained in this case, as determined by GLC, was 21%. The re- action of _11 with n-butyllithium was repeated at reflux (60-70°C) for 5 hrs. The yield of 23, as determined by GLC, was 60%. 14 Since these experiments indicate that the rate of reaction of 11_with n-butyllithium is extremely slow, a kinetic study was carried out by gas analysis. A mercury burette was connected to the reaction flask containing two mmoles each of 11 and n—butyllithium in hexane. After 4 days, 30 ml (1.2 mmoles) of gas was evolved. The experiment was repeated with excess (3 mmoles) n-butyllithium. After 5 days, 32 ml (1.3 mmoles) of butane was obtained and this volume remained constant for two more days. The reaction of 11_with n-butyllithium (2 mmoles) was repeated at reflux (72°C) and 29 ml (1.2 mmoles) of butane was obtained after 12 hrs. This increased to 39.5 ml (1.6 mmoles) of butane after the ad- dition of 0.5 ml of water to the reaction mixture. The theoretical volume of butane gas from a 2 mmole scale reaction (eq. 25) was ”50.0 ml. We considered that the reason for the less than stoichiometric volume of butane is due to solubility of butane in the reaction mixture. Accor- dingly, the hexane solution of n-butyllithium (2 mmoles) was saturated with butane gas before the addition of the amine 11. Glass connections were used instead of rubber connections. In this case an evolution of 47 ml (1.9 mmoles) of butane after 5 days at room temperature was observed. Reaction of 17 With.0ther Organolithium Reagents. a. Methyllithium Amine lZ.(2 mmoles) was added to an equiva- lent amount of methyllithium in ether at room temperature. The methane gas evolved was measured by a mercury burette and 94 ml (3.75 moles) of gas was accumulated over a 10 hr period. In a control experiment the reaction was repeated with diisopropylamine in place of lz_and over 15 100 md,(>4mmoles) gas was obtained. The high vapour pressure of ether is possibly the reason for the excess volume observed. The reaction mix- ture of ll_and methyllithium was analyzed for the formation of the lithium amide 21. The sequence shown in equation 27 was used for the 0 Si (CH3) 3 ‘4 1)Cyclohexanonej4 analysis. The yield of 23, as determined by GLC, was only 10%. b. sec-Butyllithium. sec-Butyllithium (2 mmoles) was reacted with 11 in hexane at room temperature .45 ml (1.8 mmole) of butane was evolved after 44 hrs. Analysis for the Formation of the Lithium Amide 27. Since the yields in the analysis for the formation of the Li- amide 21, through the sequence shown in equation 27, were not large, the possibility existed that it was not accurate. The sequence was tried with LiTMP, an amide known to be formed quantitatively and a 71% yield of 2§_was obtained. Consequently, we decided to analyze for amide by re- action with tert-butyl acetate in THF at -78°C, followed by addition of cyclohexanone (eq. 28). The reaction mixture was then analyzed for the O H H0 CHZC-OC(CH3)3 NLi 1)CH 3C-OC(CH3 )3 . I7 + RLi 2)Cyclohexanone > 28 (28) B-hydroxy ester 28.. This2 method of analysis was tested first with LiTMP. LiTMP was prepared and dissolved in THF at -78°C. An equivalent amount of t-butyl acetate was added to this solution followed by cyclohexanone. The yield of the B—hydroxy ester 3§_obtained in this case as determined by 16 GLC, was 95%. Amine g was mixed with equivalent amount of methyllithium in ether overnight at room temperature and the reaction mixture was worked up (eq. 28). The yield of the B—hydroxy ester 28, as determined by GLC was 36%. To check that all _2_8 was formed by the reaction of 21 and not by any other base, the reaction was repeated but without amine (eq. 29). (No detectable amount of 28 was formed. u 3 1)CH c-oc=< ArCHZX + base ‘—> ArCH _, Ar (5) by reaction with LiTMP. LiTMP does not form a carbene with methyl iodide. We considered that the generation of carbene from methyl iodide may re- quire a stronger base or a poorer nucleophile than LiTMP. It seemed likely that 31_would meet these requirements. Therefore 21 was mixed with methyl iodide in the presence of cyclohexene and the reaction mix- ture was analyzed by GLC for norcarane. The amine 11_was observed but no norcarane was detected (eq. 31). A possible explanation for these results is that amide 21_ removes a proton from methyl iodide to give 11 and methyl iodide anion. This anion may react with methyl iodide (path a, eq. 46) or may decompose to a carbene which for some reason fails to react with cyclohexene (path b, eq. 46). Another possibility is that CH CH 2 3 cnzcn3 2 (46) _11 . NLi + c1131 » 2 CHBI (a) >>CH3CHZI _ 'éH 4: CHZI 2 > ? (b) 2_7 Cyclohexene A ——-> ? norcarane is formed but reacts with 21 (path c, eq. 46). Route (c) was ruled out because we observed in a separate experiment that 21 does not react with norcarane. Route (a) and (b) need more careful study before any final conclusion can be drawn because Olofson2b and otherslg”21 have 30 reported that the addition of carbenes or carbenoids to olefins often fails unless very specific reaction conditions are met. Brown reported that alkoxide basicity is a function of the alkyl group size. If a similar effect applies with amide basicity, it seems likely that 2_7_ is a stronger base than other less hindered amides such as LiTMP. One way of testing for this increase in basicity is to attempt metalla- tion of toluene. The pKa of toluene is reported to be 40.9, while the pKa of diisopropylamine is reported to be “'38. As expected, we found no evidence for metallation of toluene by either LDA or LiTMP (eq. 47a). Amide 21 CH3 LDA or LiTMP + l \- / CHZCH3 CH3 dim ’0 2 27 _; N. R. (47a) (47b) metallated toluene and we were able to obtain mono- as well as disily- lated toluene when the reaction mixture of toluene and 21_was quenched with trimethylchlorosilane. At first, we considered that the observed metallation was due to the presence of unreacted n-butyllithium. This was excluded by the observation that the presence of diisopropylamine (which scavenges all n-butyllithium) has no effect on the amount of metallation observed. The disilylation product is not unexpected, be- cause silicon is capable of stabilizing adjacent negative charge.22 Therefore, it is possible that mono-silylated toluene reacts rapidly with benzyl anion to form toluene and disilylated toluene (eq. 48). 31 ca Li (033)381Cl CH2S1(CH3)3 ‘2 1 CH Si(CH3)3 CH[S1(CH3)3]2 L I -— 0 (CH3)3SiCl© (48) 4. 32 CH3 LiTMP reacts with 9~methyl BBN to form the boron stabilized anion in 50% yield. Excess base (LiTMP) gives a 60% yield of the R23 CH3 + LiTMP ‘% R28 CHZLi + TM? (2) metallation product. LiTMP fails to react with triethylboron. We considered the degree of metallation would be increased if a base stronger than LiTMP were used. Unfortunately, the reaction of the amide gz_and 9~methyl BBN gave about the same results (40%) as LiTMP and no metallation was observed with triethylboron. These observations may be interpreted to mean the amide 21 is not a stronger base. But the metallation of toluene with 22_and not with LiTMP makes this inter- pretation unlikely. Another possibility is that the metallated organo- borane §§_may coordinate to the boron atom of the starting organoborane to form complex 22, The formation of,§§_will prevent at least half of the starting organoborane from participation in the reaction (eq. 49). 32 (49) When the reaction mixture is quenched with D 0, the complex §2_will 2 dissociate to give 50% yield of the deuterated product. If this is the case, use of a stronger base would have no effect on the degree of metallation. Lithium amide bases are known to deprotonate ketones to form lithium enolates. Non-symmetrical ketones may be deprotonated to two regioisomeric enolates. The lack of regiocontrol in the formation of on OLi o I R Mi 1 a 2 3 1 a 2 3 Rlcnzc CHR2R3 _2__+ R CH CCHR R + R CHZC CR R (6) these enolates is a significant problem which limits the use of enolates in organic synthesis. Amide gz_may provide an effective solution to this problem. Kinetically controlled deprotonation of 2~methy1-3- pentanone with gl_gave exclusively (99%) the less highly substituted enolate. Kinetically controlled deprotonation of 3~methy1cyclohexanone with 21 gave a 90% selectivity favoring the less highly substituted enolate. The results shown in Table II and III also indicate the 33 preferential formation of the enolate on the less hindered side as the size of the alkyl groups of the amide increases. These results can be explained by considering the chair-like transition states 69_and_§1. H CHR2R3 ” N ‘Li H, R §Q_less substituted enolate 1 more substituted enolate As the size of R increases the transition state §l_will be destabilized more than 69. Therefore, §Q_will be preferred and the regioselectivity for the less substituted enolate will increase. Because the size of R of the amides, shown in Tables II and III (other than LDA and LiTMP) is very large, exclusive regioselectivity of enolate formation towards the less substituted side of the ketone was observed. The assignment of structure to the two enolates (62 and 63) ob- tained by deprotonation of 3~methylcyclohexanone is not a simple task. 0 OLi OLi (CH3)ZSiCL OSi(CH3)3 RzNLi + llll|'-————+ lllii +. lill. ‘+ lliil + 22 fl £3 (50) 0Si(CH3)3 §§_ It seems reasonable to assign structure §2_to the major product of the reaction because hydrogens attached to C6 (leading to enolate 62) appear to be more sterically accessible than the hydrogens attached to C 2 (leading to enolate 63). In line with this, Corey23 has reported that 34 the major product formed by base-catalyzed reaction of 3~methylcyclo~ hexanone with CS2 is structure 61 and only minor amounts of the isomeric CH 31 (MeS)2 C . C(SMe)2 + cs 0 O (51) 67 68 product 68 are formed. It is likely that Corey's results reflect a kinetically controlled enolate distibution because we observe almost equal amounts of §2_and.§3 under equilibrating conditions (58% §2_and 42% 6}). Additional evidence for the assignment of structures is based on the NMR spectra of the corresponding silyl enol ethers 64_and 65, The number of protons coupled to the vinyl hydrogen of §4_should be four while only three protons should couple to the vinyl hydrogen of 62. The signal for the vinyl hydrogen of the major isomer produced in the reaction does exhibit higher multiplicity. The reactions of TM? with bromine is interesting. In our first experiments,a CCl4 solution of bromine was added to TMP. An orange solid, presumed to have structure 12 was formed, but the yield was only 66%. Examination of the mother liquor by NMR showed the presence of 29_which may be formed by the reaction of TM? with 19 as shown in + {1;}— (52) Br 35 equation 52, Modification of the procedure by addition of bromine to a heterogenous mixture of sodium hydroxide and TMP gave a quantitative yield of 20. Amine 11 likewise reacts with bromine to give the corresponding N-bromo—derivative. Reaction of the N-bromoamines with sodium to prepare the cor- responding sodium amides was at least partially successful (Table I). OSi(CH ) r. 3 3 1)Cyclohexanone 1 (53) 21 23 The yield of 2§_as determined by GLC, was between 40-77%. However, 23 was obtained in w-15% yield, with RZNH in place of R NBr. 2 A possible explanation for the formation of 22.18 a radical mechanism (eq. 54). 0Na 0 ONa OSi(CH3 )3 OSi(CI-13)3 (CH3)3SiCl L:::] Na. :6 ___. +.: -—> 1 (CH3 ) 3SiCl D|isproportionation . (54) OSi(CH3)3 /K e > o EXPERIMENTAL I. Materials l-Ethynylcyclohexyl alcohol and l-ethynylcyclohexylamine were commercially available and used without further purification. Methyl iodide, N-chlorosuccinimide, cyclohexene, toluene, a- methylstyrene, and triethylboron were commercially available and used without further purification. A11 organolithium reagents were commercially available and were used without further purification. Trimethylchlorosilane was commer- cially available and was distilled (bp 57Z/atm. pressure) prior to use. Boron trifluoride etherate was distilled (126°C/atm. pressure) under argon atmosphere. Norcarane was prepared as described by LeGoff et a1.12 Cyclohexanone, 3-methylcyclohexanone and tert-butyl acetate were commercially available and used without further purification. 2-methy1- 3—pentanone was commercially available (95%) and was purified by distil- lation with a spinning band column (120°C/atm. pressure). Diisopropylamine (bp 83°latm. pressure) was distilled and stored over molecular sieves. 2,2,6,6,-tetramethylpiperidine was commercially available and used without further purification. The last five amines listed in Table II were obtained from I. Kopka and used directly. Tetrahydrofuran was dried over sodium benzophenone ketyl, 36 37 distilled,and stored under argon over molecular sieves. Dimethyl- formamide was dried over calcium hydride, distilled under vacuum and stored over molecular sieves. U. S. P. grade absolute ethanol was used without further purification. N,N,N,N-Tetramethylethelenediamine was dried over calcium hydride and distilled before use. All other solvents were used without further purification. All inorganic reagents were commercially available and used with- out further purification. 11. Preparation of bis(1-ethy1cyclohexyl) amine (11) a. Preparation of l-chloro-ethynylcyclohexane (lg). A l-L. 3—neck flask provided with a magnetic stir—bar, ther- mometer and dropping funnel was charged with 28 g (0.25 mole) of calcium chloride, 20 g (0.20 mole) of cuprous chloride, 0.2 g of copper bronze power and 220 ml of cold concentrated hydrochloric acid. The mixture was cooled in an ice bath. 62 g (0.5 mole) of l-ethynylcyclohexanol was added and the mixture was stirred for 1.5 hr. The upper organic layer was separated and washed with two 50 ml portions of cold concen- trated hydrochloric acid and then with three 50 ml portions of distilled water. The product was dried over anhydrous potassium carbonate. Anal- ysis of the crude product by GLC (10% Carbowax 204M on chromosorb-W column) showed the sample to be 952 pure. The chloride was used without further purification. Total isolated yield of pure chloride was 80%. The infrared spectrum showed the ethynyl band at 3050 cm-1, and no band for the starting alcohol. NMR(CD13) (TMS internal standard): F 2.6 (S, 1H), 6 2.0 (bm 43), 6 1.6 (bm, 6H). 38 b. Preparation of Bis(l-ethynylcyclohexyl) amine (16). A 500 ml round bottom flask, equipped with a magnetic stir-bar, septum inlet and gas inlet value was flame dried under argon. The gas inlet value was removed. 0.22 g of copper bronze powder and 0.22 of cuprous chloride were added to the flask. Then 110 ml of DMF and 31.9 g (0.26 mole) of l-ethynylcyclohexylamine were added to the flask. The gas inlet value was reattached and the flask was flushed with argon for a few minutes. The flask was then cooled to 4°C in a cold room. 19.5 g (0.13 mole) of 95% l-chloro-l-ethynylcyclohexane was added dropwise by syringe to the stirring solution. The solution was stirred at 4°C for 72 hr. The solution was then diluted with 100 ml of water followed by 12 m1 of 50% NaOH (0.15 mole) and stirred for 10 minutes. The solu- tion was extracted with three 50 ml aliquots of ether. The ether ex- tracts were combined and dried over magnesium sulfate. The ether and the unreacted primary amine were removed under reduced pressure. The coupled amine (16) was distilled under vacuum (bp 103-1060/2 mm Hg). There was obtained 16.4 g (652) of (1_6). 0m = 3290, 2300, 1070 em’l; mp. 71-72°c, NMR (00013) 6 1.55 (11,1311), 6 2.0 (M, BR), 6 2.35 (s, 211) mass spectrum m/e 230 (M+ + 1), 229 (11*), 228 (17). 200 (21), 186 (52), 172 (73), 118 (50), 80 (100), 67 (49), 41 (58). c. Hydrogenation of bis(1-ethyny1cyclohexy1amine) withw2 Raney Nickel Catalyst. Raney nickel alloy was activated as reported previously.9 Two and a half teaspoon (10 g) of W2 Raney nickel was added to a solution of 250 ml absolute ethanol and 11.45 g (50 mmoles) of bis(l-ethynylcyclo- hexyl) amine in a 500 ml centrifuge bottle. The bottle was placed in a 39 Parr hydrogenation apparatus and purged with hydrogen 5 times. The bottle was pressurized to 60 psi and the shaker turned on. After 20 hr, the solution was filtered to remove the catalyst and the ethanol evaporated under reduced pressure. GLC analysis (5% 0N’101 Chromosorb W, acid washed, DMSC treated) showed it to be 91% pure, the impurity was not identified. The saturated amine (11) was purified by distillation with a spinning band column (110-113°c/0.3 mmHg). The isolated yield of the saturated amine was 80% (9.5 g). The hydrogenation was repeated under identical conditions ex- cept that 0.1 mole (5.6 g) of potassium hydroxide was added to the ethanolic solution of bis(l-ethynylcyclohexyl) amine before addition of the Raney nickel catalyst. After the hydrogentation was completed, the catalyst and ethanol were removed. The residue was dissolved in ether and extracted with.water. GLC analysis of the organic layer as described above indicated an 80% purity of the saturated amine. The same experi- ment was performed under identical conditions except that acetic acid (0.1 mmole) was used in place of potassium hydroxide. GLC analysis of the product under these conditions showed 87% purity of the sat- urated amine. NMR (CDC13) 5 0.8 (t, 7H), 6 1.40 (M, 24H); mass spectrum m/e 237 (14*), 184 (4), 128 (5), 98 (11), 86 (100), 72 (15), 57 (20); ele- mental analysis calculated for CI6H31N: C, 81.01; H, 13.08; N, 5.91. Found: C, 81.11; H, 12.94; N, 5.79. 40 III. Reactions of bis(l-ethylcyclohexyl) amine (11) Reaction of 17 with Methyl Iodide. Methyl iodide (2 mmoles, 0.14 ml) was added to a 10 ml round bottom flask containing a stir bar, side arm septum, gas inlet valve and 2 ml of CD013 at room temperature. Then 2 mmoles (0.5 m1) of 11 was added by a syringe. The reaction mixture was stirred overnight. NMR showed only unreacted methyl iodide and 11. The same experiment was performed with diisopropylamine and 2,2,6,6,-tetramethylpiperidine. Both amines reacted and the Ndmethylammonium iodide derivatives were formed quantitatively by NMR. Infl{(CDCl3) of N-methyl, diisorpylammonium iodide 6 1.0 (S, 13H), 6 2.43 (5, 3H), 6 3.23 (M, 2H). NMR (CD013) of Ndmethyl, 2,2,6,6,-tetramethylpiridium iodide, 6 1.03 (S, 6H), 6 1.23 (8, 6H), 6 1.46 (S, 6H), 6 2.20 (S, 3H). N-methy1-2,2,6,6,-tetramethylpiperidine iodide was also identified by a mixed melting point with an authentic10 sample m. p. 280°C. The reaction of 12_(2 mmoles) with excess methyl iodide (4 mmoles) was performed under identified conditions except the reaction mixture was stirred for one week. The NMR of the reaction mixture was identical to that of the starting amine and methyl iodide. The same experiment was performed in CZHSOD' The reaction mixture was refluxed (80°C) for 6 hrs. The NMR of the residue after the solvent was evap- orated was identical to that of the starting amine. 41 Reaction of lz_with BF3-0Et2. A 10 m1 flask, equipped with a magnetic stir bar, septum inlet and gas inlet valve was flame dried under argon. One mmole of 11 and hexane (1 ml) were added to the flask followed by BF3-0Et2 (1 mmole, 0.12 ml) was added. A.white solid was formed immediately. Evaporation of the solvent under vacuum gave a quantilative yield (0.3 g) of a white, air stable solid. Crystallization from benzene (m. p. 154.5-55). NMR (CDC13) 6 1.05 (t, 6H), 6 1.6 (bm, 20H), 6 1.95 (bq, 4H); mass spectrum m/e 237 (M+)-identical to free amine. Elemental analysis calculated for C16H31NBF3: C, 62.96; H, 10.23; N, 4.59, B, 3.54, F, 18.67. Found: C, 61.11; H, 10.43; N, 4.39; B, 4.83; F, 19.22. The same experiment was performed with diisopropylamine and with 2,2,6,6,- tetramethylpiperidine. A.white solid was formed in both cases. The same experiment with the above three amines was performed in CDCl3. The NMR spectrum of each reaction showed an upfield shift for the chemical shift of the methylene hydrogens in diethyl ether. 6 3.4 (free ether), 6 4.13 (BFBOEtZ) and 6 3.42 (reaction mixture). Reaction of 17 with trimethylchlorosilane. Essentially the same procedure was followed for the reaction of 11_with trimethylchlorosilane as that described for the reaction of 11 with excess iodide. The reaction mixture was stirred for four days. The NMR showed unreacted 11_and trimethylchlorosilane. 42 Reaction of 17 with N-Chlorosuccinimide (NCS). N-Chlorosuccinimide (0.14 g, 1 mmole) and 12 (o.25 m1, 1 mmole) were stirred in methylene chloride (2 ml) at room temperature. The reaction was followed by NMR. The NCS disappeared gradually and com- pletely after a week. The NMR spectrum of the product showed a broad- ening with slight shift of the main resonance peak of the starting amine to lower field, NMR (CCl4) 6 0.95 (t.,6H), 6 1.6 (bm, 24H). The same experiment was performed in CCl4 and CH C12 (1:1) mixture (10 ml). 2 The reaction mixture was refluxed for 20 hrs. The solvent was removed under vacuum and the residue was dissolved in 5 ml CHZClZ. The solu- tion was extracted with three 10 ml portions of water. The organic layer was dried over NaZSOA. The NMR spectrum.was identical with the one mentioned above. GLC analysis with 0V 101 column showed only the solvent peak and no amine or N-chloroamine. Mass spectrum showed m/e * 271 (14*). Reaction of 17 with Bromine. A 0.5 M standard solution of bromine in CCl4 was prepared by introducing 7.99 g (2.56 ml) bromine to a 100 ml volumetric flask then completed to 100 ml by CC14. In a 100 m1 round bottom flask fitted with magnetic stir bar, side arm septum and gas inlet valve, 1.7 ml (10 mmoles) of 2,2,6,6,-tetramethylpiperidine and 20 m1 CCl4 were introduced. To this mixture, 20 ml of the standard solution of bromine in 0C1“ was added dropwise at 0°C. The reaction mixture was stirred for 5 minutes, and an orange precipitate was formed. The solid was collected by filtration and washed three times with 5 m1 portions of 43 hexane. The solid (2 g, 66%) was transferred to an erlenmeyer flask and 10 ml of 1.1825N NaOH solution was added. The solution was extracted with three 10 ml portions of ether and the organic layer was dried over magnesium sulfate. NMR (CCla) 6 1.2 (S, 12H), 6 1.56 (S, 6H). However, the NMR of the mother liquor showed similar resonance peaks. The same experiment was repeated but with a different sequence. 2,2,6,6,-tetra- methylpiperidine (1 mmole) was added to a mixture of 1.18N NaOH (1 ml) and hexane (1 ml). To this mixture, bromine (1 mmole, 6.06 ml) was added dropwise by syringe with vigorous stirring. After 10 minutes, the organic layer was dried over MgSO4 and hexane was removed under vacuum. To the residue 1 mmole of benzene was added (internal standard) and the NMR in CDCl3 was taken. The yield of N-bromoamine was 95%. A large scale (100 mmole) experiment was performed exactly but the residue was distilled (65-700C/0.7 mm). The isolated yield was 71% with identi- cal spectral data as before. The last experiment was repeated with amine 11, but the N—bromo-derivative was not isolatedtnnrwas reacted with sodium as described in the following section. Reaction of N-Bromoamines with Sodium. The procedure given below for the reaction of sodium with N- bromo-Z,2,6,6,-tetramethylpiperidine is representative for the reactions shown in Table II. A 10 ml round bottom flask equipped with a side arm septum, stir bar and gas inlet valve was flushed with argon and charged with sodium dispersion(40% in mineral oil) (0.15 ml, 2.2 mmoles) and hexane (1 ml). To this mixture, a].PIsolution of the N-bromoamine (1 ml) in hexane was added dropwise at room temperature. The mixture was stirred for 50 44 minutes and analyzed for the Na-amide as follows: Cyclohexanone (0.1 ml, 1 mmole) was added dropwise and stirred for 10 minutes. Then tri- methylchlorosilane (0.13 ml, 1 mmole) was added all at once and C111124 (0.21 ml, 1 mmole), internal standard, was also added. The reaction mixture was diluted with pentane (1 m1) and analyzed by GLC with an SE-30 columna .A 74% yield of l-trimethylsilyloxy-1-cyclohexene was es— tablished. A sample of this compound was isolated by preparative GLC; NMR (CDC13) 6 0.16 (8,911), 61.5(bm, 4H), 6 1.97 (bs, 4H), 6 4.8 (t of t, 1H) . Reaction of 17 with Organolithium Reagents. The procedure given below for the reaction of n-butyllithium with 11 is representative. A 10 ml round bottom flask equipped with side arm, magnetic bar, fitted with a gas measuring burrette was flame dried under argon and charged with 1.6M solution of n-butyllithium in hexane (2 mmoles, 1.25 ml). The reaction mixture was saturated with n-butane and amine 11_(2 mmole, 0.5 ml) was then added at room temperature. The reaction mixture was stirred for 4 days. 49 ml (2 mmoles) of butane gas was evolved. Reaction of 17 with n-Butyllithium in the Presence of TMEDA. The same experiment described above was repeated but after the addition of the amine 11, 2 mmole (0.3 m1) of TMEDA was added. 51 ml (2.01 mmoles) of butane gas was evolved immediately. The same ex- periment was repeated with 0.2 mmole (0.03 ml) TMEDA. A total of 2 mmoles of butane gas was evolved after 4 hrs of stirring at room temperature. 45 Analysis for Li-Amide Formation. Hexane was removed from the amide solution formed above under vacuum. The residue was then dissolved in THF (2 ml) and cooled by dry ice acetone bath to -78°C. t-Butylacetate:(0.27 m1, 2 mmoles) was added dropwise and the reaction mixture was stirred for 15 minutes. Cyclo- hexanone (0.2 ml, 2 mmoles) was added dropwise and the reaction mixture was stirred for another 15 minutes. Then water (1 ml) was added. The dry ice acetone bath was removed and the reaction mixture was warmed to room temperature.n.-CISH32 (0.55 ml, 2 mmoles) was added as internal standard for glc analysis. To this mixture «0.2 g potassium carbonate was added followed by 2 ml pentane. The organic layer was dried over sodium sulfate and analyzed by GLC (SE-30 column) for the formation of the B-hydroxy ester. The yield was 99%. IV. Reactions of the lithio-bis(l—ethylcyclohexyl) amide (27) Reaction of the Amide 27 with Methyl Iodide. A 1 mmole of 21_was prepared with 10% TMEDA as described above. 0.5 ml cyclohexene was added to the amide solution followed by 0.08 ml (1.3 mmole) of methyl iodide. The reaction mixture was stirred over- night and then analyzed for the formation of norcarane by GLC (0V 101 column). The GLC trace showed only a peak for the starting amine. The same experiment was performed and the reaction mixture was analyzed for unreacted amide, but the sequence described above for formation of the B—hydroxy ester. The amount of the amide found was less than 5%. The same experiment was performed with LiTMP instead of 27. GLC analysis 46 showed 57% of the amide was methylated but no norcarane was formed. Reaction of 27 With Toluene. One mmole of 21_was prepared with 10% TMEDA as described above. 0.11 ml (1 mmole) of toluene was added to the amide solution and the mixture was stirred for 16 h. Then 0.15 mi (1.1 mmole) of trimethyl- chlorosilane was added at 0°C. The reaction mixture was warmed to room temperature and quenched with 1 m1 of water. The organic layer was dried over potassium carbonate and analyzed by GLC (SE-30 column). Rel- ative to the amount of recovered amine the recovered unreacted toluene was 80%, mono—silylated tol ene was 9% and disilylated toluene was 5%. The silylation products were identified by GLCdmass spectrum analysis. Mono—silylated toluene mass spectrum m/e 164 (M5), 149 (31), 121 (30), 91 (34), 73 (100), 65 (23), 43 (40). Disilylated toluene m/e 236 (Mf), 148 (88), 133 (17), 73 (100), 45 (58). The same experiment was performed for different periods of time prior to quenching with tri- methylchlorosilane. Reaction of 27 with aeMethylstyrene. One mmole of 21 was prepared as described above and 0.13 ml (1 mmole) of admethylstyrene was added to the amide solution. The mixture was stirred at room temperature for 20 h and quenched with trimethylchlorosilane (0.15 ml, 1.1 mmole). GLC analysis (SE-30 column) of the reaction mixture showed only unreacted admethylstyrene. The ex- periment was performed again and quenched with methyl iodide after 20 h. Analysis of the reaction mixture by GLC showed unreacted a- methylstyrene and amine 11, 47 Reaction of 27 with triethylboron. One mmole of 21_was prepared as described above. The solvent was evaporated under reduced pressure and 1 ml benzene was added. 0.14 ml of triethylboron (1 mmole) was then added to the reaction mixture and stirred for 42 hrs at room temperature. The mixture was quenched with 0.2 ml of deuterium oxide, followed by the addition of 0.2 ml 3N HCH a few minutes later. The water layer was removed with a syringe and the remaining organic layer was saturated with anhydrous potassium carbonate. The organic layer was then analyzed by GLC-mass spectra. The same experiment was performed but quenched with methyl iodide instead of D20 and was then analyzed for the methylation product by GLC-mass spectra. Preparation of B-methyl-9-borabicycloE3.3.llnonane. A 500 ml flask equipped with a reflux condenser, an addition funnel, a magnetic stirring bar, and a side-arm fitted with a rubber serum stopper was flushed with argon and maintained under a static argon atmosphere. 85 m1 of a 1.18M solution of methyllithium (100 mmoles) in ether was placed in the cleaned dropping funnel and added dropwise to the solution of 9-borabicyclo[3.3.llnonane at 0°C over a period of one hour. This was immediately followed by dropwise addition of 6.5 ml of methanesulfonic acid (100 mmoles) and approximately 100 mmoles of hydrogen was rapidly evolved. The salt was allowed to settle, and the clear solution was transferred under argon to a dis- tilling flask and the product distilled. A 74% yield of B-methyl-9-BBN b. p. 48-500C/6 mm was obtained (12 ml). 48 Reaction of 27 with B—methyl-9BBN. The same procedure described for the reaction of 21_with tri- ethylboron was repeated exactly with Bdmethyl-9BBN. The reaction mixture was quenched with D20. Reaction of 27 and other Li-amides with Unsymmetrical ketones. The procedure given below for the reaction of 21_with 3—methy1- cyclohexanone is representative of the reactions shown in Tables II and III. A 3.2 mmoles of 21 was prepared with equivalent amount of TMEDA (0.48 ml). The solvent was removed under vacuum which was broken with argon. The yellowish viscous residue was dissolved in 3 ml THF and cooled in a dry ice acetone bath. 3-Methylcyclohexanone (0.37 ml, 3 mmoles) was added dropwise and after 20 minutes, the reaction was quenched with 0.42 ml of trimethylchlorosilane (3.2 mmoles). The reac- tion mixture was stirred for 30 minutes and then the bath was removed. Upon warming to room temperature, 0.39 ml of decane (3 mmoles), internal standard, was added. The resulting solution was diluted with 3 ml pentane and then 6 ml saturated NaltCO3 was added. The organic layer was dried over anhydrous sodium sulfate before glc analysis (1/8 in x 40 ft stainless steel column packed with 20% SE-30 on Chromosorb W). The reaction mixture was analyzed by glc and/or NMR. NMR (CDClz) 6 4.82 (m, vinylic proton), 6 4.72 (m, vinylic proton), 6 1.95 (m, 6H), 6 1.02 (m, 4H), 6 0.28 (5, 9H). The NMR data for the products from the reaction of 2-methyl-3- pentanone were identical to the literature values.35 CHAPTER II STEREOCHEMISTRY 0F ENOLATE FORMATION 49 INTRODUCTION The aldol condensation is a reaction of fundamental importance in biosynthesis (eq. 1). Compound 1, for example, which is the 8 ? Bgse 2 3 RCHZCR' J3EE» RCH=CR' —R—C§°—-> ch (OH) HCH(R) CR' (1) enolate open-chain form of the aglycone of the macrolide erythromycin A,24 can be regarded as the result of a series of six aldol type condensations. The presence of ten chiral centers in this molecule requires that CH:2 0H CH3 (ll-1‘3 CH3 OH CH CH .1. sufficient control be maintained over the stereochemical outcome of each condensation. This is typical of the biosynthesis of other macrolide aglycones in which the carbon skeleton originates from a series of con- densations of acetyl- and propionyl-CoA units.25 The formation of enolates from carbonyl compounds by the action of base is the first step in the aldol condensation (eq. 1). While the regiochemical aspects of this reaction have been studied extensively,26 it is only recently that attention has been devoted to the stereochem- istry of the reaction. The correlation of enolate geometry with the 50 51 stereochemical outcome of the aldol reaction has provided a strong incentive to such studies. 27 found that addition of chelating In 1973 House and coworkers 2+ M§+ divalent cations such as Zn or to preformed lithium enolates leads to aldol product mixtures rich in the threo product 3, regardless of enolate geometry. They attributed this preference to the greater sta- bility of the threo chelate_3, in which the greater number of sub- stituents on the cyclic intermediate occupy equatorial positions. R1 R2 ' 0 {I'M ‘ O R n H .2 (erythro) Dubois28 was apparently the first to relate the stereochemical outcome of the aldol to the geometry of the enolate. He showed that the condensation is subject to kinetic stereoselection with Z-enolates giving predominantly the erythro aldol.4_(eq. 2) and E-enolates leading pre- dominantly to the threo isomer §_(eq. 3). 0H 0 CH OLi 3‘,c-c’ + R'CHO ————-> R. R (2) H ‘R l; O 31': 0. I R c-c + R'CHO -———+ (3) :11 8 .4. mi. 2 In 1977 Heathcock29 examined the use of preformed lithium enolates and found that in some cases complete kinetic stereoselection may be achieved. For aldol condensations of the type typified by equations 2 and 3, complete kinetic stereoselection was observed when R 52 is bulky. The Z-enolate gave the erythro aldol and the E-enolate gave the threo aldol. For example, the condensation of 2,2,-dimethyl-3- pentanone (100% Z-enolate) with benzaldehyde gave erythro aldol Z_with no measurable amount of threo aldol. Ethyl mesityl ketone gave a a 0113 [0L1 0H 8 ‘LDA A x a PhCHO _ 011301120 C(CH3)3 _ntfi’ THF , [c c\ ————> }\/c C(CH3)3 H C(CH3)3 Ph I 7 <4) 92% (E): 8% (Z) enolates mixture when reacted with LDA at -72°C. Reaction of this mixture with benzaldehyde generated a mixture of 92% 011 0 011 0 0 927 (8) 3R 5 3R N . ’ CH3CH2C R_720CLDATHF ¢ 8% (Z) 'P—h'C-EL Ph/lY + /Y ’ enolates ‘~ Ph R=mesity1 19, erythro (8%) _2, threo (92%) (5) threo (9) and 8% erythro (19) condensation adducts. When R is smaller (ethyl, isopropyl,. . .) stereoselectivity diminished or disappeared. In 1979 Evans and coworkers30 investigated the Stereochemistry of the aldol condensations of boron enolates. High kinetic stereo- selection was observed, regardless of the size of R, the Z-enolate (11) giving the erythro adduct 13, and the E-enolate (12) giving the threo adduct 14_(eq. 6). Evans attributed the high steroselectivity to steric parameters involved in the pericyclic transition states leading to the diastereoisomeric adducts (Scheme 1). In the case of an E-enolate, for example, transition state T2 would be destabilized relative to T by 1 non-bonded interactions between R2 and R1 and between R2 and L. 53 Scheme I IN 54 0 “ Base H3C\ [OBLZ H\ [OBLZ CH CH CR.+IL2BOSOZCF3_ ‘o > C=C + C=C 3 2 78 C,ether H/ \R H3C/ \R 11 (z) _13 (E) JRICHO [RICHO O OBL O OBL In 1976, Ireland and coworkers31 investigated the ester enolate Claisen rearrangement with a variety of allylic esters. They found a stereochemical control operating through stereoselective enolate for- mation. Scheme II demonstrates the rearrangement of E-crotyl pro- panoate. In the THF solution, the enolate anion or the derived silyl Scheme II / THF \ _, (J m 0 V N LDA OLi H - ' 16 17 0‘\“//\\\ ——- -—— 237. HMPA-THF \ / mm 0 > ___’ 0 0L1 0H 17 19 ketene acetal yielded the erythro rearrangement isomer 11. However, in the more coordinating solvent system 23% HMPA-THF, enolization took an alternative course and threo acid 12_was the major product. It was assumed that in THF the Z-type enolate 1§_was preferentially formed and 55 trapped, but in 23% HMPAeTHF the geometrically isomeric E-type enolate anion 1§_was preferentially formed. Similar results were obtained with the symmetrical ketone 3-pentanone. A high degree of selectivity for the formation of one enolate in THF [77% (E): 23% (2)] and the other 0 OLi H OLi CH3 011 CH 0011 CH 23L» \0-0/ + \ / 3 2 2 3 / " \ C= (7) 01130112 0113 01130112 H THF 777. 23% 237. HMPA-THF 57. 95% enolate in HMPA-THF [5% (E): 95% (2)] was observed. Ireland rationalized this dramatic solvent effect by considering the steric requirements for enolization of the two transition states 29 and 21, In the less coordinating solvent, THF, the interaction of the carbonyl oxygen with the lithium cation is assumed to be quite important R' OR' fl (2) .14 (E) and the carbonyl oxygen became effectively bulkier than -OR'. The re— sulting non-bonded interactions raised the energy of transition state 21, and enolization proceeded through transition state 29. The presence of HMPA, on the other hand, resulted in greater solvation of the lithium cation and an enhanced reactivity of the amide base. The lithium-car- bonyl oxygen interaction was assumed to be weaker and transition state 21, in which R becomes eclipsed with the now sterically smaller carbonyl 56 oxygen during enolization, was favored. Similar effects of HMPA on anion Stereochemistry have been reported for deprotonation reactions of a variety of ketones,298 hydrazones,32 and oxazolines.3 In 1978 Kuwajima and coworkers34 investigated the formation of the silyl enol ethers of few acyclic ketones with ethyl trimethyl silyl acetate (ETSA) and catalytic amounts of tetrabutylammonium fluoride (TBAF) in THF. Silylation of 5-nonanone at 0°C gave exclusively the Z- silyl enol ether. Silylation of 2-heptanone gave the Z—silyl enol ether 22_in 55% yield together with the regio—isomer 23 in 9% yield (the E- isomer 24_was not detected). 0 CH CH ) OSi CH on $10 " ETSA TBAF 3( 2:3 / ( 3)3 ( 3)3 ' CH3(CH2)3CH2CCH3 0’ ~+» 0-0 + CH3(CH2)4-C=CH2 -78.0h—+Rai. ’ \ H CH3 24 hr 22, 55% g_, 92 (8) / ,\ 0113(c112)3 CH 14, 074 Kuwajima and coworkers also investigated the reaction of LiTMP with 3-pentanone and obtained 84% of the E-silyl enol ether. This E- selectivity is higher than that obtained by Ireland with LDA under similar conditions (77% E ). This difference in selectivity was 57 (CH3)3SiO H 'l 1) LiTMP \ ’ 01130820 01120113 2) (CH ) 5m . 0-0 (9) 3 3 / \ -78°c, THF CH3032 CH3 84% attributed to the increased bulk of LiTMP over LDA (stabilizing_2g over 2;). In a study of the formation of silyl enol ethers from unsym- metrical ketones, House35 found that deprotonation with LDA followed by enolate quenching with trimethylchlorosilane gave a mixture in which the less highly substituted silyl enol ether (except for case 2 eq. 10) is the principal product. When trimethylamine and trimethylchlorosilane were used, mixtures of both regio- and stereoisomers (except for case d eq. 10) were obtained, but it was not clear whether these results were a R 7' H. “°’ , 1) LDA A \ _ B _ RCHZCR 2) (CH3)3SiC1 - lC—C\ /C C\ + Regio isomer H OSi(CH3)3 R OSi(CH3)3 2.5. 26. .21 a) RFME, R'-isopropyl 53% (20%) 42% (62%) 5% (18%) b) RdMe, R'fiMe 16% (24%) 13% (64%) 71% (12%) c) Ren-butyl, R'-Me 9% (29%) 7% (58%) 84% (13%) d) RPPh, R'BMe 86% (33%) 14% (67%) 0% (0%) The results between ( ) are with triethylamine. due to partial or complete equilibration of the enolates. House also observed that LDA slightly favors the formation of E—silyl enol ethers 2§_rather than the Z—isomers 26. In his study of the stereochemistry of aldol condensations of 58 boron enolates, Evans also investigated the stereochemistry of the enolates themselves. These were prepared by mixing dialkylboron tri- flates (R2BOSOZCF3) and diisopropylethylamine with one equivalent of ketone. Evans obtained Z-enolates exclusively (>99%) in most cases; however, he found that the kinetically controlled enolate ratio (22:22) was a function of the base employed in the enolization process. Thus 0 CH OBR n 0BR 3 2 I 2 011 CH .6014 CH + R BOSO CF fl» \‘c 0’ + \c c (11) 3 2 2 3 2 2 3 -78 / \ H 01120113 CH3 01120113 A8. (2) 22 (E) Diisopropylethylamine 99% 1% Lutidine 69% 31% 3-pentanone with diisopropylethylamine gave 99% (Z): 1% (E), whereas with lutidine it gave 69% (Z): 31% (E). It is clear from the above review, that enolate geometry is a major factor in determining the stereochemical outcome of the aldol reaction. Consequently, we decided to study the factors which determine the stereoselectivity of the enolate formation reaction in greater detail. RESULTS The ratio of the two enolates (E and 2) obtained from the re- action of 3-pentanone with lithium dialkyl amides was studied. The lithium dialkyl amides were prepared by reaction of the corresponding secondary amine with.n—butyllithium in hexane (eq. 12). 3-Pentanone was RZNH + n-BuLi £9993» RZN Li + n-BuH (12) .addeddropwise to a THF solution of the dialkyl amide. The resulting solution was analyzed for the two enolates 29_and‘22_by quenching with trimethylchlorosilane and analyzing by GLC for the corresponding tri- methylsilyl enol ethers 22_and.22 (eq. 13). 0 H OLi CH3 OLi II \ / \ / CH CH C CH CH + R N Li -—————-—* C=C +' C=C CH3 CHZCH3 H CHZCH3 2.9 (E) Q <2) (CH3)3SiC1 (13) H 081 (CH ) CH OSi(CH ) C=C + C=C / \ / \ CH3 CHZCH3 H CHZCH3 _32 33 59 60 TABLE IV A Study of the Concentration Effect on the E and Z Enolate Ratio at 0°C with 10% Excess Amide. 3-Pentanone + LiTMP <> E-(gg) + Z-(gl) Entry [Conc.] E(30):Z(31) Overall Yield 1 3.0M 38.5:61.5 94% 2 1.0M 86:14 100% 3 0.5M 89:11 83% 4 0.1M 90.6:9.4 100% 5 0.02M 92:8 90% 61 A Study of the Concentration Effects on the E and Z Enolate Ratio. THF was chosen as the solvent for our initial study because it is the most commonly used solvent for enolate forming reactions. LiTMP was reacted with 3-pentanone at 0°C at various concentrations in THF. The results obtained are shown in Table IV. A Study of the Stability of the Enolates, 30 and 31, and the Corresponding Silyl Ethers, 32 and 33. The stability of the enolates, 2Q_and.22, and the trimethylsilyl enol ethers, 22 and 22, was investigated. Standard solutions of enolates 29_and.22_were prepared by addition of 3-pentanone to a slight excess (1.1 equiv.) of a 1.0118olution of LiTMP in THF at 0°C. These solutions were quenched with trimethylchlorosilane after various periods of time at various temperatures and analyzed by GLC for the formation of 22_and‘22. Both enolate total yield (92-100%) and 29 to 22 ratio (82-86):(18-14) did not change over a period of 24 h at 25°C in the absence of any solvent additive, or in the presence of 1-4 equivalent of HMPA or TMEDA. A Study of the Temperature Effect on the E and Z Enolate Ratio. The temperature effect was investigated by dissolving the amide (LiTMP) in THF (110. 3-Pentanone (0.9 equiv) was then added dropwise at the specified temperature to the amide solution, followed after 15 minutes by trimethylchlorosilane. The reaction mixture was then worked up and analyzed by GLC for 22 and 22. Reaction temperature was found to have only a slight effect on the ratio of E and Z enolates (29 and_22). The ratio of 30:22_was slightly lower at reflux temperature (70°C) 62 TABLE V A Study of the Temperature Effect on the E and Z Enolate Ratio. 3-Pentanone + LiTMP 4+ E-(gg) + Z-(22) Entry Temperature E:Z Overall Yield 1 -78° 84:16 100% 2 -23° 87:13 92% O o 3 0 86:14 1002 0 a 4 70 73:27 972 5 25° 87:13 952 63 but remained constant within experimental error, at the other tempera- tures studied. The results are presented in Table V. A Study of the Solvent Effect on the E and Z Enolate Ratio. Because only aprotic solvents can be used with lithium amides, a very limited number of solvents was tried. LiTMP was dissolved in the specified solvent (hexane, benzene, t-butylamine, diethyl ether or THF) at 0°C (25°C for benzene) and 3-pentanone (0.9 equiv) was added dropwise. Trimethylchlorosilane was then added and the reaction mixture was analyzed by GLC. The highest E:Z (22222) ratio and highest enolate total yield was obtained with THF as solvent. The enolate total yields with the other solvents were low. However, a higher total yield was obtained with THF added during the silylation step. The results obtained are presented in Table VI. A Study of the Effect of Excess Ketone on the E and Z Enolate Ratio. A standard solution Alof the enolates 29 and 22_was prepared by addition of 3-pentanone to a slight excess (1.1 equiv) of a 1.0bisolu- tion of LiTMP in THF at 0°C. To this standard solution, 0.2 equivalent of 3-pentanone or 0.2 equivalent of benzophenone at 0°C was added and the solution was then quenched after various periods of time with tri- methylchlorosilane. The results are presented in Figure I. The results shown in Figure II were obtained by addition of equivalent amounts of TMEDA or HMPA (w. r. t. LiTMP) shortly after the addition of excess (0.2 equiv) 3-pentanone to solution 2_at 0°C. The results shown in Figure III were obtained by addition of either 2 or 4 equivalents of TMEDA or HMPA shortly after the addition 64 TABLE VI A Study of the Solvent Effect on the E and Z Enolate Ratio. 3-Pentanone + LiTMP 44* E-(30) + Z-(31) Entry Solvent E:Z Overall Yield 1 Hexane 12.5:87.5 <5% 2 Hexane 48:52 89% (THF added to help silylation) 3 (CH3)3CNH2 (Silylation in THF) 10:90 35% 4 Benzene 56:44 76% (THF added to help silylation) 5 Diethyl ether 73:27 78% 6 THF 86:14 99% 65 O CH CH 3 CH OH + LiTMP lgF—é Solution A 3 2 2 3 0 C - H 051(CH ) CH 081(CH ) (CH3)3SiC1 \ / 3 3 \3 I 3 3 A 4: C-C + C=C / \ / \ 31 z (tetal E E — yield) No additive 86% , 14% (100%) 0.2 equiv. of (CH ) SiCl 3-pentanone 15 min.;} 3 3 —#7 28% , 72% (100%) n n n n 1,, hr___, " 16% , 84% (90%) 0.2 equiv. of 15 min.- " . benzophenone -——--—+ 18% , 32% (100%) n u I! u ___l h_r__, H 14% ’ 86% (65%) Figure I. The Effect of Excess Ketone on the E and Z Enolate Ratio. :&-+ 0.2 equiv. 3-pentanone + TMEDA 15 min. > (CH3)3SiC1% 16%, 84% (100%) n _ILr__. .. 177:, 83% (977,) mm i934» " 102, 902 (947.) " -—1h—r—+ " 1071, 907 (907.) Figure II. The Effect of Excess Ketone in the Presence of TMEDA and HMPA on the E and Z Enolate Ratio. of excess (0.2 equiv) 3-pentanone to solution 2_at 0°C. The effect of excess ketone on the Z and E enolate ratio was also investigated in the presence and absence of TMEDA or HMPA at 66 15 min.(CH3)35101L A . _E_ , _2_ (total yield) .é’+ 0.2 equiv 3-pentanone + 2.0 TMEDA, " " 15% , 85% (83%) 4.0 TMEDA, " " 11% , 89% (65%) 2.0 HMPA, " " 7% , 93% (88%) 4.0 HMPA, " " 6% , 94% (68%) Figure III. The Effect of Excess TMEDA or HMPA on the E and Z Enolate Ratio. -78°C. In this study solution A was cooled to -78°C and then excess ketone (0.2 equiv) was added. The reaction mixture was quenched with trimethylchlorosilane at -78°C, and in the first reaction (Figure IV) TMEDA was added to activate the silylation at -78°C. The results obtained are presented in Figure IV . (CH3)38101 _ A + _E_ , a (total yield) A + 0.2 equiv 3-pentanone ‘:%§%E~r " 82% , 18% (89%) 1.0 TMEDA —};——> .. 88% , 12% (9974) 1.0 HMPA —-:———-> .. 67% , 332 (492) 1.0 mm. _E_» 72%;" 7% , 937. (83%) Figure IV. The effect of Excess Ketone in the Presence or Absence of TMEDA or HMPA on the Z and E Enolate Ratio at -78°C. The low enolate overall yield in the presence of HMPA at -78°C is possibly due to an irreversible aldol condensation (eq. 14). 67 0 (CH3) 3810 fl CH CH % CH CH + LiTMP THF-HMPA—“ 32 + 33 + (CH CH ) C CH(CH )C CH CH 3 2 2 3 (CH3)3SiC1' -—- -—- 3 2 2 3 2 3 0 We examined for this possibility by reacting acetone with a slight ex- cess (1.1 equiv) of LDA in THF at -78°C. An equivalent amount of both HMPA and 3— pentanone was then added to the reaction mixture. The solution was quenched with trimethylchlorosilane. GLC and NMR analysis indicated that the main product was the trimethylsilyl enol ether of 3-pentanone. The NMR spectrum of the reaction residue indicated that only a small amount,if any, of the aldol product 3; (eq. 15) was present. The same experiment was repeated with acetophenone in place of 1)HMPA fl 2)CHBCHZE-cnzcn3 °Si —————> 887. , 127. (777.) " 9.0 ——°i-> ———> 85% , 157. (9074) 0.45 " . 9.0 ———+ ———-> 86% , 14/. (827.) " 4.0(11118A)9=-9—+ ————> 547. , 467. (70%) " 9.0 (") E... —————-> 527 , 48% (897.) n 0.45 " 9.0 ( ) ——-> ———+ 517. , 4974 (55%) Figure V. The Effect of Both Excess Amide and TMEDA or HMPA on the Z and E Enolate Ratio. and 23. The enolate (E and 2) ratios observed were 87:13 (81% total yield) with TMEDA and 40:60 (60% total yield) with HMPA. When 0.9 equivalent of 3-pentanone was dissolved in THF (1M) and then was added very slowly (25 min) to a mixture of TMEDA and LiTMP in THF at -78°C the ratio was 91:9 (97% total yield). A Study of the Effect of Ketone Structure on the E and Z Enolate Ratio. The reaction of 2,2-dimethyl—3-pentanone with LDA or LiTMP, gave only one product presumed to have the structure 31 (eq. 17), under all conditions studied (0°C or -78°C, excess ketone or excess amide, in the presence or absence of TMEDA). The reaction of 2~methyl-3-pentanone with excess LiTMP at —78°C gave mainly the E—enolate (38). However, reaction with excess LDA at 71 a 1)R2NLi (“9381? ,CH3 (_CH ) c c ca 011 + (cu ) c=c (17) 3 3 2 3 3 3 \ ”(0113) 381Cl 11 2 37 -78°C gave mainly the Z-enolate (32). The reaction of 2~methyl-3- pentanone with LDA in the presence of excess ketone at 0°C or room temperature gave mainly the Z-enolate (39). The results obtained are shown in Figure VI (a study of the regio-isomers formed from deprotona- tion reactions of 2—methyl-3-pentanone was presented in Chapter I). 3 (CH3) 38101 11‘ ,051(CH3)3 C113 P51(CH3)3 LDA + (cu ) cnccn CH A C=C + c=c + Regio- 3 2 2 3 THF ” \ ’ \ isomers CH3 cu 7+ E—(38) + Z-(39) Amide* E:Z $33; LDA 42:58 (95%) LiTMP 75:25 (95%) CHZCH3 90:10 (85%) {*NLi C . [CH3CH2(CH3)2C]2NL1 91:9 (98%) (CH3CH2)2(CH3)CNL1C(CH3)2CH2CH3 90:10 (94%) [(CH3CH2)2(CH3)C]2NL1 92:8 (96%) (CH3CH2)3CNL1C(CH3)(CHZCH3)2 90:10 (93%) [(CHBCHZ)3C]2NL1 85:15 (91%) *All amides (1.1 equiv) were formed in the presence of equivalent amount of TMEDA. 77 Because the low overall yield of the enolates in the reactions of ethyl propanoate may be due to susceptibility of this ester to con- densation, we decided to study methyl butanoate. Ireland reported a quantitative overall yield of the enolates from the reaction of methyl butanoate with LDA, as determined by dimethyl-butylchlorosilane quench- ing (eq. 17). Ireland obtained mainly §§_(E) (91%) in THF solution and mainly 41 (Z) 84% in 23 vol % HMPAPTHF. 0 l 1 )30(CH3)28i01 H \ ,OS.1+ CH3“; [371+ ‘1‘ + H (CH3 CH CH COCH + LDA C=C C=C 3 2 3 I \ / \ CH3CH2 OCH3 H OCH3 (18) 45 32 (E) 4_7 <2) For our study of this ester, three standard solutions (A, Q, g) were prepared. Solution A was prepared by adding 5.0 mmoles of the ester 45_to 5.5 mmoles of LDA in THF (15 ml). Solution §_was prepared by adding 5.0 mmoles of the ester 45 to a 5.5 mmoles of LDA in 23 vol % HMPA-THF (15 m1). Solution §_was prepared by adding 2.5 mmoles of the ester 4§_to 5.0 mmoles of LDA in THF (15 m1). These solutions were quenched with dimethyl-tert-butylchlorosilane. 2 m1 HMPA was added to both A and §_reaction mixture to facilitate the silylation reaction. The three reactions mixtures were then analyzed by HINMR for the silyl- derivatives 4§_and.41. The spectra from solution Aland §_were identi- cal, with 46_(E) as the major product. The NMR spectrum from reaction g showed 41_(Z) as the major product. A quantitative analysis for 46 and fiz_was not possible because of incomplete resolution of NMR signals. we were unable to separate 46 and 42_by GLC even with 50' 78 column. Addition of 1.5 mmoles of the ester 45 and 4.5 ml HMPA to solu- tion A_gave a mixture which could not be analyzed by NMR because the signals for -0CH3 group of unreacted ester overlapped with product signals. When 4.5 m1 of HMPA.was added to solution A and the mixture was stirred for 10 minutes before~silylation, the NMR analysis of the resi— due indicated again that 46 (E) was the major product. The same result was obtained when 4.5 m1 of HMPA was added to solution §_and stirred for 10 minutes prior to silylation. DISCUSSION Ireland31 reported that deprotonation of 3-pentanone with LDA in THF solution gave mainly the E-enolate [77% (E):§Q, 23% (2)131J. The same sequence in a 23 vol % HMPArTHF solvent gave predominantly the Z-enolate [5% (E)-30, 95% (Z)-31]. Ireland suggested that the observed stereoselectivity arises in either case by a kinetically controlled process and that the increased Z-selectivity is a consequence of the lesser coordinating ability of lithium for carbonyl oxygen in a solvent mixture containing HMPA. Our results indicate that those deprotonation reactions of 3-pentanone with lithium amide bases which lead to pre- dominant Z-stereoselectivity operate under conditions of thermodynamic control. Under conditions where kinetic control is ensured, predominant E-selectivity is observed in the presence or absence of HMPA or the related solvent additive, TMEDA. Our initial investigation was carried out by preparing standard solutions which.contained 86% (E)1§Q and 14% (2)131. The stability of these standard solutions was studied under a variety of conditions. Both the (E):§Q:(Z)1§1 ratio and overall yield of the two enolates did not change over a period of 24 h at 25°C in the absence of any solvent additive, or in the presence of 1-4 equiv of HMPA or TMEDA. However, addition of 0.2 equivalent of 3-pentanone or 0.2 equivalent of benzo- phenon caused a rapid isomerization, complete in less than an hour at 0°C, to an equilibrium mixture of enolates with an (E):§Q:(Z)1§1 ratio 79 80 of 16:84, as shown in Figure I of the results section. The rate of this isomerization was appreciably faster in the presence of HMPA or TMEDA (complete in <10 min., Figure II of the results section). The rate of isomerization at -78°C with excess ketone was extremely slow in THF or TMEDAFTHF mixtures. With HMPA-THF mixture at -78°C the overall yield of the enolates was low, <50%, and the rate of equilibration was also low. The effect of both HMPA and TMEDA on the position of enolate equilibrium at 0°C was to increase the amount of Z-enolate (84% (2)-31 in THF alone, maximum of 94% (Z)-31 in HMPA-THF, and max 89% (2)-31 in TMEDA-THF). From these results we conclude that only unreacted ketone causes the equilibration of Ea-e Z enolates. A possible mechanism for this equilibration is a—hydrogen exchange as shown in equation 19. However, 01.1,}! a 3 OLi 0113 ____.+ \ c—c +cuc11ccncn cncnccncn+ c=c 03011” ‘08 32 23+— 32 23 0803’ \H 3 2 3 3 2 (19) this mechanism is probably too slow to account for the rapid equilibra- tion observed at 0°C (complete in less than 1 hr in THF alone). It is known36 that the equilibration of regioisomers of ketone enolates, which is assumed to occur by the same a-hydrogen exchange mechanism, requires a period of several hours at 25°C even in the presence of a substantial excess (IO-100%) of ketone (eq. 20). Benzophenone is a ketone which has no a-hydrogens and therefore cannot possibly participate in an a—hydrogen exchange mechanism. Thus 81 0L1 250 0 0L1 1 11 __.3. II | CH =-=CCH R. + CH C CH R CH C CH R + CH C-=CHR (20) 2 2 3 2 ‘r——-—— 3 2 3 several hours our observation that benzophenone promotes enolate isomerization about as efficiently as 3-pentanone provides compelling evidence against the operation of the a-hydrogen exchange mechanism. we consider that the most likely isomerization mechanism is a reversible aldol condensation as shown in equation 21. The aldol OLi H k1 0L1 \. / ——————A I C==C + R2C=O R C CH(CH )COCH CH /, \ r—1:——- 2 3 2 3 CH CH CH 2 3 2 3 k k (21) (E)1§Q 3 4 OLi CH3 \ I C=C + R C=C / \ 2 CHBCH2 H (z>-_3_1 condensation of ketones is known to be a reversible reaction with equil- 37-39 ibrium favoring the starting ketone and enolate. The role of TMEDA and, more importantly, HMPA in increasing the equilibrium amount of Z-enolate may be interpreted in the following fashion. It is known40 that lithium alkoxides are polymeric, with lithium bonded to four oxygen atoms in simple alkoxides (LiOCH3), and to three oxygen atoms in more hindered ones (LiOC(CH If a similar 3)3)‘ polymeric structure is assumed for lithium enolates then the effect of HMPA may be to ligate with lithium and decrease the effective size of 82 ;/ ((Liu" , , O/Li(HMPA) é‘I‘L1\\Iji Li(HMPA) g _E_:mMPA) _z_ gums) the lithium—oxygen grouping. This should result in an increased stability of the Z—enolate. Both HMPA and TMEDA are known to strongly activate lithium eno- lates perhaps by coordination to lithiumf‘1 This ligation of lithium should increase the value of k1 in equation 21, leading to a faster (compared to THF alone) rate of isomerization. we can explain the low yield of enolates obtained in the presence of HMPA at -78, by as- suming the formation of a stable aldol at that temperature. We were able to get equilibration (7:93) and higher overall yield of enolates, (83%), when excess ketone was reacted with LiTMP in the presence of HMPA at -78°C for one hour and then warmed to 0°C for another hour before quenching with trimethylchlorosilane. But we were unable to trap the aldol product by silylation at -78°C. The structure of the ketone has a marked effect on the composi- tion of the enolate equilibrium mixture. For 2~methyl-3-pentanone, the enolate equilibrium composition in THF is 9:91 (E:Z, R=isopropyl). 0L1 H 0L1 CH3 R CH3 R H 83 2,2-dimethy1-3-pentanone gave only one enolate at equilibrium. Heathcock reported that this is the Z-enolate (100% Z ReC(CH3)3). For propiophenone, the equilibrium enolate composition in THF is 6:94 (E:Z, Rsph). In all cases, the most stable enolate is the Z- form, because the methyl group is trans to the bulkier group, R, and cis to the smaller -OLi. Also as the size of R increases the equilibrium composition of the 2- form increases. The equilibrium mixtures of stereoisomeric enolates described above were prepared by reaction of 1-2 equivalent of the ketone with one equivalent of LDA or LiTMP in THF for 1 hr at 0°C. This is a new and very convenient method for the equilibration of enolate stereo- chemistry, and therefore it provides a simple method to study the position of enolate equilibrium. Since enolate isomerization occurs only by reaction of enolate with starting ketone, (Scheme III), a true kinetically controlled deprotonation should be favored by: a) A high amide to ketone ratio amide (RZNLi) ‘+ enolate Rate of enolate formation= k [ketone] [amide] enolate} isomerization Rate of isomerization- k [ketone] [enolate] ketone —‘ Scheme 111 b) Low temperature deprotonation or c) Rapid trapping of the enolates. A high amide to ketone ratio should increase the rate of enolate formation and decrease the rate of isomerization. This is because the strating ketone will be rapidly consumed by deprotonation and will not be available for enolate isomerization. An amide to diethyl ketone ratio of 1.0:0.9 produces a kinetically controlled enolate distribution 84 in THF (86% (E)-30, 14% (Z)-31). Increasing the amide to ketone ratio to 2:1 does not greatly alter the ratio of the enolates [882 (E)- 30, 12% (2)-31]. However a kinetically controlled enolate distribution in the presence of TMEDA or HMPA was only obtained when a higher amide to ketone ratio (1.0:0.25) was used. This effect of both TMEDA and HMEA can be explained by the ability of these solvent additives to activate enolates, in this case to the aldol condensation, which leads to enolate isomerization. The enolate distribution [66% (E)-30, 34% (2)-31] obtained in the presence of HMPA is puzzling. It is possible that the presence of HMPA does alter the kinetically controlled distribution of the two enolates. However, it is also possible that this ratio does not yet represent the results of the true kinetically controlled depro- tonation. Unfortunately, we were unable to test this possibility by the use of even higher amide to ketone ratios because both enolate total yield and the reproducibility of the results decreased drastically in the presence of HMPA under these conditions. A low temperature deprotonation should favor kinetically con- trolled deprotonation because the rate of isomerization is very slow at -78°C, as shown by the results presented in Figure IV. It is important here to emphasize that either the trapping of the enolates must occur at -78°C or that no unreacted ketone is present prior to trapping. Other- wise,unreacted ketone will cause equilibration to occur once the enolate solution is warmed to higher temperatures. It seems likely that Ireland's31 results [51% (E), 95% (2)] obtained for the deprotonation of 3—pentanone with LDA in 23 vol 2 HMPA-THF at -78°C followed by trapping ‘with dimethyl-tert-butylchlorosilane, are due to enolate equilibration. This is because of the slow reaction of the hindered dimethyl-tert- 85 butychlorosilane with enolates. Thus if unreacted ketone was present, equilibration would occur when the reaction mixture was warmed to room temperature. Rapid trapping of the enolates by trimethylchlorosilane should favor kinetically controlled deprotonation. A rapid trapping of enolate will prevent further reaction of the enolate with unreacted ketone. The data presented in Table VII of the results section are in agreement with this point. The presence of trimethylchlorosilane in the reaction mixture, changed the ratio of enolates from 17%(E):83Z(Z) to 77%(E): 23%(2), during the deprotonation step rather than addition of trimethyl— chlorosilane after deprotonation in TMEDA-THF. The kinetically controlled deprotonation of 3-pentanone at various temperature (-78°C-R. T.) in THF occurred with an almost con- stant ratio of E and Z enolates, as shown in Table V. A slight increase in the Z-selectivity was observed for the deprotonation at 70°C and this may simply be due to a slight isomerization during the deprotonation step at this high temperature. The kinetically controlled enolate distribution obtained by reaction of a variety of ketones with lithium amide bases was determined. The Z-selectivity increases as the size of R (eq. 22) increases. 1)! 0&1 /H 0L1 R' RCCH R' + R"NL1 —-—> . C=C + \c-c/ (22) 2 2 If \\ Rf, \ O R R H .12 .Z. Deprotonation of 3-pentanone with LiTMP occurs with an E:Z ratio of 86:14. Deprotonation of 2~methy1-3-pentanone occurs with an E:Z ratio of 70:30. Deprotonation of propiophenone and 2,2-dimethyl-3-pentanone 86 produces mainly Z-enolate. Heathcock reported the formation of 100% Z-enolate with the last ketone. These results can be explained by considering the two chair transition states for the deprotonation step. When R is small (ethyl or Base E Base R _ 0 R ‘IIIII.’ O H H ‘ R' R! E—transition state Z-transition state R .Li R\ Ii C\ ” OI C .II \0 1' \O ‘1 H- - .-"N'- H~‘-‘§N'—' . | / I R--C H-——C L H R isopropyl), the E—transition state, with R' in the equatorial position, is more stable than the Z-transition state, which has R' in the axial position. However, as R becomes bulkier (t-butyl or phenyl) the inter- action between R and R' becomes more severe and therefore the Z-transi- tion state becomes more stable. The results reported by Heathcock298 for the deprotonation of ethyl mesityl ketone (92% (E), 8% Z) are suprising. It is possible that the benzene ring of the exceptionally bulky mesityl 87 group is forced to take a perpendicular conformation relative to the plane defined by the carbonyl grouping. Such an orientation will effectively lower any steric interaction with R' in the §_transition state . Kuwaj ima34 reported a higher E-selectivity for deprotonation of 3—pentanone with LiTMP (84% E-enolate) compared to LDA (77% E-enolate). Our results, Table VIII also, indicate that under kinetically controlled reactions, an increase in the size of the alkyl groups of the amide (R', R") leads to a higher E-selectivity. These results can be explained 0L1 H 0Li CH u 0°C I / I / 3 R'R"NLi + R—C-CHZCH3—--+ R-C=C + R—C=C\ (22) \ H CH3 2 .2. by considering the transition states _A and 2 shown below. As the size CH3 R' 4M ___+ a I ..- _ ‘ vl’o’, ——-> E-enolate H” , " I!) Z ' N N-‘ , enolate I”"N~---‘L1HR"/ ‘Li RII/ A 13. of R' increases, transition state Q will be destabilized by increasing axial-axial interaction of R with CH3. Therefore A becomes more stable and E-selectivity will increase. However, our results obtained with more hindered amides, shown in Tables IX and X indicate that there is a limit to this effect and after a certain point is reached, E-selectivity declines. As the size of beecomes very bulky [-C(CH3)2 CH CH or 2 3 C(CHZCH3)3] it is possible that the transition state is no longer 88 chair-like and the decline in the E-selectivity results from a different transition state. A very useful conclusion for the stereochemistry of enolate formation can be mentioned here. It is possible to control the depro- tonation of 3-pentanone in THF solution alone. In order to produce predominantly E-enolate, 3—pentanone is added to 10% excess amide at 0°C or to an equivalent amount of amide at -78°C. To obtain predom- inantly Z-enolate 3-pentanone is added to a slight deficiency of the amide or a stoichiometric amount of amide is added dropwise to a solu- tion of 3-pentanone at 0°C. The same procedure with 2-methyl-3- pentanone produces similar results. Finally, the brief investigation of ester equilibration did not lead to any conclusive results. There were two main difficulties in the esters study: the instability of the silylketene acetal and the lack of a simple and easy way of analysis for the products. In fact, the first difficulty was solved by using dimethyl-tert-butylchlorosilane instead of trimethylchlorosilane. The second difficulty will probably be solved once the 250 MHz NMR will be more available for routine analysis. However, we still strongly believe that esters will behave as ketones regarding the enolate formation conditions. EXPERIMENTAL I. General Spectra. Proton magnetic resonance spectra were measured using a Varian T-60 spectometer. Gas Chromatography. Vapor phase chromatographic analysis and preparative work were carried out on a Varian Aerograph 920 thermal conductivity chromatograph. Relative peak areas were determined by a Varian disc chart integrator- model. II. Materials Handling of Materials. All reactions were carried out under an argon atmosphere and all liquids were transferred with glass syringes. All solvents were care- fully distilled42 and stored uner an argon atmosphere over molecular sieves. ' Amines, Amides and n-Butyllithium. The following amines, diethylamine, diisopropylamine, tert- butylamine, 2,2,6,6,-tetramethylpiperidine, bis(trimethylsilyl) amine, 89 90 and bis(Z-ethoxyethyl) amine were commercially available. All other amines were prepared as described in Chapter I. All Li-amides were prepared prior to use by reaction of n—butyllithium with the correspon- ding amine, as described in Chapter I. The n-BuLi was commercially available as a 1.6M solution in hexane and was used without further purification. SilylatingfiReagents. Trimethylchlorosilane, obtained commercially, was distilled and stored under argon. Dimethyl-tert-butylchlorosilane was prepared as 43 described by Corey et al. This silane was used as 3.6M solution in pentane. Ketones and Esters. 3-Pentanone, propiophenone, 2,2-dimethyl-3-pentanone and ethyl propanoate were commercially available and were distilled before use. 2-Methy1-3-pentanone was 95% commercially available. This ketone was purified by distillation through a Spinning band column. Methyl butanoate was prepared from the commercially available acid in 91% yield by the following procedure: One mole of butyric acid was refluxed with excess (1.2 equiv) thionyl chloride for two hours. The excess thionyl chloride was removed under reduced pressure. Methanol (1.5 equiv) was added dropwise at 0°C. The reaction mixture was stirred for 15 minutes at room temperature and then 200 ml of water was added. The aqueous layer was washed twice with ether (50 ml). The organic layers were combined and washed twice with saturated NaHCO3 (30 ml). After drying the organic layer over anhydrous NaZSOA, the product was purified by a 91 spinning band distillation apparatus. III. The Reactions of Ketones and Esters With Li—Amides a. The Reaction of Excess Ketones With Li-Amides. The following procedure for the reaction of 3-pentanone with LiTMP at 0°C will be representative. A 10Imlflask equipped with a stir bar, septum, gas inlet valve, and mercury bubbler was flame dried while a stream of argon was flowing through the system. A 1.6M (1.25 ml, 2 mmoles) aliquot of n-butyllithium in hexane was added to the flask. The flask was immersed in an ice bath and 0.35 ml (2 mmoles) of 2,2,6,6,-tetramethylpiperidine was added dropwise with stirring. After the evolution of butane had ceased, the hexane was removed by vacuum which was broken with argon. The white solid was dissolved in 2‘ml of THF and cooled in an ice bath. 3-Pentanone 0.19 ml, 1.8 mmoles) was added dropwise followed by (0.4 ml, 0.04 mmole) portion, and after 15 minutes, the reaction was quenched with 0.3 ml (2.3 mmoles) of tri- methylchlorosilane. The reaction mixture was stirred for 30 minutes and then the ice bath was removed. Upon warming to room temperature, 0.34 ml (2 mmoles) of nonane (internal standard) was added. The re- sulting solution was diluted with.2 m1 pentane and then 4 ml saturated NaHCO3 was added. The organic layer was dried over anhydrous sodium sulfate before glc analysis. 92 b. The Reaction of Ketones With Excess Li—Amides. The same procedures described above was used except the mmoles of 3-pentanone was less than the mmoles of LiTMP. For reactions in the presence of solvent additives (TMEDA or HMPA), these solvents were added prior to the ketone addition. c. The Reaction of Esters with Li-Amides. The same procedure described above was used except the reaction mixture was not worked up with saturated NaHCO3 in case of ethyl propanoate. The reaction mixture was diluted with pentane and then analyzed by glc. IV. Product Analysis The ratio of the enolates as well as the overall yield of the enolates was determined by glc (1/8 in x 40 ft stainless steel column packed with 20% 813-30 on Chromosorb W) analysis for the corresponding silyl enol ethers. Pure samples of each product were isolated by preparative glc and examined by NMR. E-3-(Trimethylsilyloxy)-2-pentene (30). NMR (CDC13): 6 0.18 (s, 9H), 6 1.01 (t, 3H), 6 1.54 (d, 3H), 6 2.08 (9, 2H), 6 4.6 (9, 1H) (CI-1C13 was the internal standard) Z-3-(Trimethylsilyloxy)-2+pentene (31). NMR (CD013): 6 0.18 (s, 9H), 6 1.02 (t, 3H), 5 1.52 (d of t, 3H), 93 6 2.03 (q, 2H), 4.53 (q, 1H) (CHCl was the internal standard). 3 E-4-Methyl-3-(trimethylsilyloxy)-25pentene (38). NMR (CCla): 6 0.17 (8, 9H), 6 0.94 (d, 6H), 6 1.52 (d, 3H), 6 2.60 (m, 1H), 6 4.28 (q, 1H) (TMS was the internal standard). Z-4-M ethyl-3- (trimethyl si 1y loxy) -2-pentene (39) . NMR (C014): 6 0.17 (S, 9H), 6 0.98 (d, 6H), 6 1.44 (d of d, 3H), 6 2.10 (m, 1H), 6 4.35 (q, 1H). (TMS was the internal standard). 2-Methyl-3-(trimethylsilyloxy)-2epentene. NMR (CDC13): 6 0.17 (S, 9H), 6 1.0 (t, 3H), 6 1.56 (bS, 6H), 6' 2.07 (bq, 2H) (TMS was the internal standard). Z-4,4éDimethyl-3-(trimethylsilyloxy)-2:pentene (37). NMR (CDC13): 6 0.17 (3, 9H), 6 0.97 (S, 9H), 6 1.24 (d, 3H), 6 4.50 (q, 1H). E-3—P henyl-3- (trimethylsilyloxy) -2-propene (40) . NMR (00013): 6 0.07 (s, 9H), 6 1.6 (d, 3H), 6 4.86 (q. 1H), 6 7.10 (S, SE). UV (geheptane) Amax 235 nm. Z-34Pheny1—3-(trimethylsilyloxy)-25propene (41). NMR (00013): 6 0.1 (s, 9H), 6 1.65 (d, 3H), 6 5.10 (q. In), 6 7.07 (m, 5H). UV (g-heptane) xmax 243 nm. 94 0~Trimethylsilyl4T-ethyl Methyl Ketene Acetal. NMR (CDCl : 6 0.18 (8, 9H), 6 1.19 (t, 3H), 6 1.5 (d, 3H), 3) 6 3.73 (m, 3H). BIBLIOGRAPHY 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) BIBLIOGRAPHY N. C. Deno, et al., J. Amer. Chem. Soc., _92, 2065 (1971). a) M. W. Rathke and R. Row, J. Amer. Chem. Soc., a, 2715 (1973). b) M. W. Rathke and R. Row, J. Amer. Chem. Soc., 24, 6854 (1972). a) R. A. Olofson and C. M. Dougherty, J. Amer. Chem. Soc., _9_5_, 581 (1973). b) R. A. Olofson and C. M. Dougherty, J. Amer. Chem. Soc., _9_5_, 582 (1973). Charles Allan Brown, J. C. 8. Chem. Comm., 680 (1974). F. G. Hennion and C. V. DiGiovanna, J. Org. Chem., 32, 2645 (1965). I. E. Kopka, "The Synthesis and Hydrogenation of Sterically Hindered Secondary Amines," M. S. Thesis, Michigan State University (1978). F. G. Hennion and K. W. Nelson, J. Amer. Chem. Soc., 2.9.: 2142 (1957). V. J. Shiner, Jr. and J. W. Wilson, J. Amer. Chem. Soc., fl, 2402 (1962). a) R. L. Augustine, Catalflic Hydrogenation, Marcel Dekker, Inc., New York, 1965. b) C. G. Vexlearschi, Bull. Soc. Chim. Fr., 589 (1956). H. 2. Sommer, H. I. Lipp, and L. L. Jackson, J. 03: Chem., 36, 824 (1971). a) L. Labler and F. Sorm, Collection Czechoslov. Chem. Comm, _2_4, 4010 (1959). b) T. G. Back and D. H. R. Barton, J. C. S. Perkin I, 924 (1977). E. LeGoff, J. Org. Chem., _2_9, 2048 (1964). H. C. Brown, M. Rogic, J. Amer. Chem. Soc., _8_2_, 4908 (1960). N. R. Easton, R. D. Dillard, W. J. Doran, M. Livezey, D. E. Morrison, J. Org. Chem., g, 3772 (1961). a) G. F. Hennion and C. V. DiGiovanna, J. Org. Chem., _3__1_, 1977 (1966). 95 16) 17) 18) 19) 20) 21) 22) 23) 24) 25) 26) 27) 28) 29) 30) 31) 32) 33) 96 b) C. F. Hennion and C. V. DiGiovanna, J. Org. Chem., 39, 3696 F. Klages and H. Sitz, Chem. Ber., 96, 2394 (1963). V. I. Spitsyn, I. D. Kolli, R. A. Rodionov, and T. G. Sevast'yanova, Dokl. Akad. Nauk, SSSR, 160, 1101 (1965). D. J. Peterson, J. Organometal. Chem., 8, 199 (1067). S. Bank and M. C. Prislopski, Chem. Commun., 1624 (1970). G. L. Closs and L. E. Closs, Tetrahedron Lett., 26 (1977). D. H. Hoeg and D. I. Lusk, J. Amer. Chem. Soc., 86, 928 (1964). E. W. Colvin, Chem. Soc. Rev., 1, 15 (1978). E. J. Corey and R. H. R. Chen, Tetrahedron Lett., 3817 (1973). W. D. Celmer, Pure Appl. Chem., 28, 413 (1971). M. Berry, Quart. Rev., 11, 343 (1963). Review: a. H. 0. House, Modern Synthetic Reactions, W. A. Benjamin, Menlo Park, 1972, chapter 9; b) J. d'Angelo, Tetrahedron, 32, 2979 (1976). H. 0. House, D. S. Crumrine, A. Y. Teranishi and H. D. Olmstead, J. Amer. Chem. Soc., 25, 3311 (1973). a) J. -E. Dubois and M. Dubois, Tetrahedron Lett., 4215 (1967). b) J. -E. Dubois and P. Fellman, C. R. Acad. Sci., 274, 1307 1307 (1972). c) J. -E. Dubois and P. Fellmann, Tetrahedron Lett., 1225 (1975). a) W. A. Klebshick, C. T. Buse, C. H. Heathcock, J. Amer. Chem. Soc., 99, 247 (1977). b) C. T. Buse and C. H. Heathcock, J. Amer. Chem. Soc., 22, 8109 (1977). D. A. Evans, E. Vogel, and J. V. Nelson, J. Amer. Chem. Soc., 101, 6120 (1979). R. E. Ireland, R. H. Mueller, and A. K. Willard, J. Amer. Chem. Soc., 28, 2868 (1976). G. K. Davenport, H. Eichenauer, D. Enders, M. Newcomb, and D. E. Bergbreiter, J. Amer. Chem. Soc., 101, 5654 (1979). a) M. A. Hoobler, D. E. Bergbreiter and M. Newcomb, J. Amer. Chem. Soc., 100, 8182 (1978). b) A. I. Meyers, E. S. Snyder, and J. J. H. Ackerman, J. Amer. Chem. Soc., 100, 8186 (1978). 34) 35) 36) 37) 38) 39) 40) 41) 42) 97 E. Nakamura, K. Hashimoto, and I. Kuwajima, Tetrahedron Lett., 2079 (1978). H. 0. House, L. J. Czuba, M. Gali, and H. D. Olmstead, J. Org. Chem., 34, 2324 (1969). H. 0. House and B. M. Trost, J. Org. Chem., 39, 1341 (1965). H. 0. House, Modern Synthetic Reactions, ch. 10. P. Kurath, J. OrngChem , 32, 3626 (1967). R. E. Ireland, et al., J. Org. Chem., 35, 570 (1970). D. C. Bradly, R. C. Mehrotra and D. P. Gaur, Metal Alkoxides, Academic Press Inc., 1978, p. 76. C. Agami, Bull. Soc. Chim. Fr. Rev., 1619 (1970). A. J. Gordon, R. A. Ford, The Chemist's Companion, John Wiley and Son, Inc. 1972.